Transforming Race and Class in Suburbia: Decline in Metropolitan Baltimore

Tr a nsfor m i ng R ac e a n d Cl a s s i n Su bu r bi a

This page intentionally left blank

Tr a nsfor m i ng R ac e a n d Cl a ss i n Su bu r bi a D e c l i n e i n M e t ropol i ta n Ba lt i mor e

Thom a s J. Vic i no

TRANSFORMING RACE AND CLASS IN SUBURBIA

Copyright © Thomas J. Vicino, 2008. All rights reserved. No part of this book may be used or reproduced in any manner whatsoever without written permission except in the case of brief quotations embodied in critical articles or reviews. First published in 2008 by PALGRAVE MACMILLAN™ 175 Fifth Avenue, New York, N.Y. 10010 and Houndmills, Basingstoke, Hampshire, England RG21 6XS Companies and representatives throughout the world. PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN-13: 978–0–230–60545–9 ISBN-10: 0–230–60545–1 Library of Congress Cataloging-in-Publication Data Vicino, Thomas J. Transforming race and class in suburbia : decline in metropolitan Baltimore / Thomas J. Vicino. p. cm. Includes bibliographical references and index. ISBN 0–230–60545–1 1. Suburbs—Maryland—Baltimore Metropolitan Area. 2. Suburban life—Maryland—Baltimore Metropolitan Area. 3. Social classes—Maryland—Baltimore Metropolitan Area. 4. Deindustrialization—Maryland—Baltimore Metropolitan Area. 5. Race discrimination—Maryland—Baltimore Metropolitan Area. 6. Segregation—Maryland—Baltimore Metropolitan Area. 7. Baltimore Metropolitan Area (Md.)—Social conditions. I. Title. HT352.U6V53 2008 305.5097526—dc22

2007052296

A catalogue record for this book is available from the British Library. Design by Newgen Imaging Systems (P) Ltd., Chennai, India. First edition: June 2008 10 9 8 7 6 5 4 3 2 1 Printed in the United States of America.

For My Grandparents and Parents: Dominic and Mary, Tomm and Nancy, and Thomas and Susann, Pioneers of First-Tier Suburbs

This page intentionally left blank

The semi-suburbanized and suburbanized messes we create in this way become despised by their own inhabitants tomorrow. These thin dispersions lack any regrettable degree of innate vitality, staying power, or inherent usefulness as settlements. Few of them, and these only the most expensive as a rule, hold this attraction much longer than a generation; then they begin to decay in the pattern of city gray areas. Jane Jacobs (1961, 445)

This page intentionally left blank

Con t e n t s

List of Figures

xi

List of Tables

xiii

Acknowledgments

xv

One

Suburban Evolution U.S. Suburban Development The Creation of the New Suburban Gothic Overview

1 7 11 15

Two

Suburban Frontier Crabgrass Memories The Rise of Suburban Baltimore Dynamics of Neighborhood Change Summary

19 20 37 43 53

Three Suburban Decline The Post-Suburban Era: The Study of Suburban Decline Patterns of Suburban Decline, 1970–2000 Summary

55 56 67 104

Four

Suburban Mosaic The New Suburban Landscape Typology of First-Tier Suburban Neighborhoods Suburban Transformations Summary

107 108 126 139 143

Five

Suburban Renaissance Baltimore County Confronts Suburban Decline Political Realities and Economic Imperatives Lessons Learned

147 149 160 165

x / contents

Six

Suburban Crossroads The New Metropolitan Dilemma: Saving the Suburbs State and Federal Policy Initiatives A Suburban Policy Agenda

169 170 175 181

Seven

Suburban Prospects Reflections on Baltimore Contributions Explaining Suburban Change A Suburban Research Agenda

185 186 188 189 190

Appendix

193

References

211

Index

223

Fig u r e s

1.1 Map of Baltimore’s first-tier suburbs 2.1 The Burgess model 3.1 Average value of housing units in suburban Baltimore, 1970–2000 3.2 Victorian house in Catonsville, 2005 3.3 Ranch house in Hampton, 2005 3.4 Tract rowhouse in Lansdowne, 2004 3.5 Rowhouse in Essex, 2005 3.6 Multiplex housing in Dundalk, 2004 4.1 Scree plot of eigenvalues for principal components analysis, 1970 and 2000 4.2 Typology of first-tier suburban neighborhoods, 1970 4.3 Typology of first-tier suburban neighborhoods, 2000

5 44 90 94 95 96 97 98 110 129 134

This page intentionally left blank

Ta bl e s

1.1 U.S. suburban development 2.1 Overview of twentieth-century federal policy initiatives for the suburbs 3.1 Common descriptions of first-tier suburbs 3.2 Summary of definitions for first-tier suburbs 3.3 Summary of studies on suburban change 3.4 Population in metropolitan Baltimore, 1970–2000 3.5 Population by age in Baltimore’s first-tier suburbs, 1970–2000 3.6 Age structure by race in suburban Baltimore, 1990–2000 3.7 Population by race in Baltimore’s first-tier suburbs, 1970–2000 3.8 Female-headed household population in metropolitan Baltimore, 1970–2000 3.9 Married households with children population in metropolitan Baltimore, 1970–2000 3.10 Poverty in metropolitan Baltimore, 1970–2000 3.11 Median household income in metropolitan Baltimore, 1970–2000 3.12 Age of housing stock in metropolitan Baltimore, 2000 3.13 Size of housing units in metropolitan Baltimore, 2000 3.14 Ratio of average housing values in Baltimore’s first-tier suburbs to all suburbs, 1970–2000 3.15 Manufacturing employment in metropolitan Baltimore, 1970–2000 3.16 Service employment in metropolitan Baltimore, 1970–2000 4.1 Summary of principal components analysis, 1970 4.2 Component loadings for suburban neighborhoods, 1970 4.3 Summary of principal components analysis, 2000 4.4 Component loadings for suburban neighborhoods, 2000

7 24 64 66 68 70 71 72 74 78 79 82 84 87 88 92 100 101 109 112 118 120

xiv / tables

4.5

Distinguishing characteristics of suburban neighborhoods in 1970 4.6 Distinguishing characteristics of suburban neighborhoods, 2000 5.1 Overview of Baltimore county’s community conservation projects in the Eastern first-tier suburbs, 1995–2005 5.2 Suburban government structure in metropolitan Baltimore, 2000 A.1 Census designated places, 1970–2000 A.2 Definition of Baltimore’s first-tier suburbs, 2000 A.3 Geographic characteristics of neighborhoods in metropolitan Baltimore, 2000 A.4 Overview of principal components analysis

128 135 152 163 195 196 197 203

Ac k now l ed gm e n t s

In 1676, during the peak of the Renaissance, Sir Isaac Newton wrote in a letter to Robert Hooke, “If I have seen further it is by standing on the shoulders of Giants.” If there is one thing I have learned in the Academy, it is that I, too, have benefited from the knowledge and experience of giants. When I look back on the evolution of this book, I find that it has really been a work in progress for nearly a decade. From former teachers, coaches, and directors to family, neighbors, and community advocates, I benefited from countless perspectives that shaped my thinking and ultimately influenced this book. As a young student at University of Miami, Juliet Gainsborough’s dynamic urban politics course and later her guidance of my thesis initially inspired me to study the American metropolis. This project continued to mature during my graduate studies at the University of Maryland, Baltimore County, which provided the precise formula for guidance, mentorship, fellowship, and most important, community. Scott Bass, Linda Brown, Sally Helms, Deb Geare, Linda Kassab, Freeman Hrabowski, Arthur Johnson, Marvin Mandell, Cheryl Miller, Dave Marcotte, Nancy Miller, Patricia Perillo, Anne Roland, Janet Rutledge, George Wagner, and Gay Warshaw were invaluable resources. I am particularly indebted to Donald F. Norris and John Rennie Short for their supervision of this project. Donald F. Norris oversaw the development of an earlier version of this manuscript and provided knowledge, insight, and direction for this project’s entirety. His steadfast support and belief in me brought this project to fruition. John Rennie Short helped to delicately craft the analysis, marrying geography to politics and public policy. My colleagues at the UMBC Center for Urban Environmental Research and Education including Bernadette Hanlon, Royce Hanson, Andrew Miller, Mike McGuire, Rich Pouyat, Amy Glorioso Rynes, Steve Sharkey, Sabrina Strohmier, Judy Unger, Claire Welty, and Ian Yesilonis graciously supported my research interests. This book would simply not have been possible without the center’s investment in my scholarly career. I am also indebted to a very special research synergy. My ongoing collaboration with Bernadette Hanlon and John Rennie Short has not only

xvi / acknowledgments

helped me to grow as a scholar, but it has also left me with a thirst for more discovery. The research on which this book is based was additionally made possible by the United States Environmental Protection Agency under grants R-82818201-0 and CR83105801. Although the research described in this book has been funded in part by the United States Environmental Protection Agency, it has not been subjected to the agency’s peer and policy review and therefore does not necessarily reflect the views and no official endorsement should be inferred. Of course, any and all errors rest with me. The statements in this book are based on my own conclusions, and I have made every effort to ensure the accuracy of facts and data. The School of Urban and Public Affairs at University of Texas at Arlington also played a supportive role during the final stages of the book. I am grateful for the fellowship and nurturing environment that my colleagues, including Ardeshir Anjomani, Enid Arvidson, Edith Barrett, Richard Cole, James Cornehls, Maria Martinez-Cosio, Paul Geisel, Joel Goldsteen, Carl Grodach, Rose-May Guignard, Rod Hissong, Jeff Howard, Margaret East, Jianling Li, Alejandro Rodriguez, David Tees, Robert Wegner, Robert Whelan, and Sherman Wyman, provided me. Michael Hennig provided valuable research assistance. The Urban Affairs Association (UAA) offered a valuable outlet for the dissemination of earlier ideas for this project. UAA continues to offer junior scholars a welcoming environment not only for constructive feedback of new ideas but also for an intimate collegiality. Earlier versions of this book were presented at the UAA annual meetings in Salt Lake City (2005); Montreal (2006); and Seattle (2007). David Ames, Juliet Gainsborough, Marie Howland, Paul Jargowsky, Dennis Keating, Robert Lake, Marc Levine, Peter Muller, Donna Shalala, Todd Swanstrom, and Elvin Wyly provided constructive feedback at various stages in the development of my ideas. It was a pleasure to work with the editors and staff at Palgrave Macmillan, including Luba Ostashevsky, Joanna Mericle, and Kristy Lilas. Two anonymous reviewers provided valuable critiques and improved the manuscript. A close network of family and friends also sustained me while I conducted research and prepared this book. Without the love and support of the following individuals, this project would not have come to fruition: Dominic and Mary Vicino, Tomm and Nancy Stricker, Tom and Suzi Vicino, Melissa and Jesse Kieffer, Allison Vicino, the Tippett family, the Stricker Jr. family, the Vicino family of Brazil, Effie Shockley, Stacey Morrison, Jay Tifone, and Maureen, Maura, and Tom Armstrong. I am especially thankful for the friendship and editorial assistance of Larry Pacific. Thanks are due as well to Benjamin Carson, who has always been an inspiration—he taught me how to “think big.” And finally, to all of the residents of first-tier suburban Baltimore, thank you for the opportunity to learn.

Ch a p t e r O n e Su bu r ba n Evolu t ion

During the late 1940s and early 1950s, Mary Anne Bauer used to hop on the No. 8 streetcar line along Frederick Road in Catonsville for a 30-minute ride to downtown Baltimore’s retail district. Even though Mary lived in the quiet suburb of Catonsville, there were few, if any, employment opportunities in this sleepy, quaint suburb of Baltimore. Mary worked as a part-time beautician at Hutlzer’s Department Store on Howard Street, which was one of the city’s premiere shopping venues of the day. Catonsville’s residents, just like Mary, were dependent on their central city—Baltimore City—for employment, shopping, culture, and entertainment during this era. The city was the center of life, and suburbs were peripheral. Yet, in just a short period, Mary’s family and metropolitan Baltimore’s residents were about to witness a dramatic transformation. Catonsville, Maryland is situated in the southwestern corner of Baltimore County, and it shares a border with Baltimore City. The suburb initially grew during the late 1920s as a streetcar suburb of Baltimore City. Catonsville experienced another growth spurt during the late 1940s immediately following the end of World War II. Houses in this suburb typified the “American Dream.” They featured large yards, gardens, green space, and picket fences on pristine, tree-lined streets. Half a century later, some of these “picture-perfect” suburban images still prevail. Yet, the character of Catonsville has changed in a number of ways. In 2000, its residents were significantly older than other suburban residents in the region, and its population declined from approximately 55,000 to 40,000 since 1970. The Catonsville population was primarily comprised of white residents during this period. Moreover, by 2000, various storefronts were vacant, and Main Street in Catonsville—once a bustling commercial strip—has failed to attract new businesses and retain existing ones. Its poverty rate has hovered approximately 5 percent since 1970, and the median household income grew by 5 percent to $53,000 in 2000, compared to the overall suburban median household income of $57,558.

2 / transforming race and class in baltimore

Woodlawn, Maryland, due north of Catonsville, stands in stark contrast. James Jackson considers himself a true Baltimorean. Born at University of Maryland Medical Center in downtown Baltimore, he grew up in Edmondson Village, a neighborhood located to the west of downtown. After attending public schools, he earned a business degree at a local state university, and shortly afterward, he was offered a job at the Social Security Administration in Woodlawn. James commuted from the city to his job in the suburbs for several years during the 1970s, but after his marriage, he decided to move to nearby Woodlawn. His move to suburban Baltimore allowed him to escape the increasing criminal activity in his urban neighborhood, and it shortened his commute. Woodlawn also provided a welcoming environment where many other African Americans had suburbanized from Baltimore City. Woodlawn is a suburb that straddles Interstate 695. Known as the Baltimore Beltway, this Interstate highway encircles Baltimore City and its first-tier suburbs. Like Catonsville, Woodlawn also shares a border with Baltimore City. This first-tier suburb initially grew during the early 1960s along Windsor Mill Road and Liberty Road—two important arterial roadways leading out of Baltimore City. After the desegregation of Baltimore City schools, white families in West Baltimore leaped over the city borders into nearby Woodlawn. For most of the 1960s and early 1970s, Woodlawn became known as a safe, all-white, bedroom suburb of Baltimore. Numerous housing subdivisions were built throughout Woodlawn. The houses tended to be smaller than those in Catonsville, and the housing stock in Woodlawn was more modern than the stock in Catonsville. Beginning in 1970, Woodlawn began to experience dramatic changes. In the three decades that followed, the majority-white suburb turned into a majority-black suburb. In 1970, the population was 2 percent black. Three decades later, the population was 52 percent black. During that same period, the population grew from approximately 26,000 to 36,000, which was a growth rate of 37 percent. Also, the population of the poor doubled, from 3 to 6 percent. Household income declined by nearly 17 percent, from $58,000 to $48,000. During this period, suburban Woodlawn was transformed from an all-white middle-class suburb to a majority-black suburb whose residents had a declining economic status. Heading northeast from Woodlawn along Interstate 695, Hampton is nestled at the northernmost part of the Beltway. David Klein and his wife Shirley bought a new house in Hampton in 1968 as newlyweds. The house featured 2,500 square feet, an open kitchen, four bedrooms, a spacious living room, and a large backyard. David and Shirley were able to raise their three children in this house comfortably. Hampton was an ideal location

suburban evolution / 3

to buy a house since David worked in downtown Baltimore and Shirley worked in nearby Towson; both had easy and quick access to interstate highways. Some 40 years later, and now empty nesters, they still live in the same house. This suburb was developed in the late 1960s and early 1970s in a suburban subdivision style. Hampton features a housing stock with large units and one-acre wooded plots of land. Pristine parks and open green space surround its neighborhoods. Hampton has been and remains an enclave for wealthy, white families since 1970. With a population of approximately 5,000 residents since 1970, it is the smallest first-tier suburb of Baltimore. There has been little to no growth in Hampton since its boom in the 1970s. Hampton’s poverty rate has never risen above 2 percent, but household income declined from $130,000 to $95,000 between 1970 and 2000. Although household income has declined since 1970, Hampton was still Baltimore’s most affluent first-tier suburb in 2000. Continuing eastward from Hampton along Interstate 695, Dundalk is situated at the southeasternmost part of the Beltway. It is an old manufacturing suburb. John Koch was born and raised in Dundalk, and so were his father and son. Three generations of the Koch family worked at the Bethlehem Steel plant. By 2000, John had worked in the plant as a steelworker for 25 years. As a member of the union, he earned a good living and was able to support his wife and two sons. Yet, economic hardships were all too commonplace at the nation’s largest steel plant in the twenty-first century. John was offered a corporate buyout, and he accepted. Yet, John still had 10 years left in the workforce, and like so many other steelworkers, he needed to find a job. The Dundalk that he remembered died a long time ago. For nearly a century, Bethlehem Steel Corporation was the heart of Dundalk, giving life to four generations of steelworkers. Although Bethlehem Steel built multiplex housing units throughout Dundalk for its workers, the houses were some of the smallest units in the suburbs. In its heyday, some 40,000 workers, many of them residents of Dundalk, labored at Beth Steel, and another 20,000 residents worked at other local industrial plants. Yet since 1970, Dundalk, too, has changed dramatically. Two-thirds of all manufacturing jobs in Dundalk had disappeared by 2000. Household income declined from $45,000 to $39,000, a 13 percent decline in real dollars. Over that same period, its population declined from approximately 85,000 to 62,000, a decline of one-third. In terms of age, the population was also one of the oldest in the first-tier suburbs of Baltimore. By 2000, Dundalk had suffered from many of the same symptoms of decline that afflicted neighboring Baltimore City.

4 / transforming race and class in baltimore

Among other things, this brief tour of these four first-tier suburbs around the Baltimore Beltway highlights the heterogeneity of these first-tier suburban places—streetcar suburbs, bedroom suburbs, elite suburbs, industrial suburbs. The histories of Mary Ann Bauer in Catonsville, James Jackson in Woodlawn, David and Shirley Klein in Hampton, and John Koch in Dundalk are not unique. Similar stories can be found throughout the suburbs. In fact, scholars and popular media alike have recently drawn attention to the social and economic changes in older suburban communities, particularly across the Northeast and Midwest (Wheeler 2005a). A quick peak at the latest headlines shows that there is an increasing public concern for these suburbs. For example, newspapers across the country read, “older suburbs struggle to maintain their vitality” (Christie 2005); “old ‘burbs need work’” (Ott 2006); “inner suburbs fall through the cracks” (Cohn 2006), and “suburbs nearer to cities neglected” (Ohlemacher 2006). Reports such as these suggest that urban problems now plague surrounding suburbs. Collectively, these four suburban portraits illustrate that first-tier suburbs have witnessed dramatic changes in their social and economic characteristics since 1970. Symptoms of urban decline, usually associated with the nation’s central cities, are now evident in first-tier suburbs. The decline of these suburbs is not a surprise. In her classic book The Death and Life of Great American Cities, social critic Jane Jacobs (1961) accurately predicted the decline of suburbs, and she likened it to the similarities of the erosion, or “gray areas,” of central cities. Clearly, Jacobs’ foresight was prescient—some 50 years later we are now beginning to understand this “suburbanized mess” that afflicts many first-tier suburbs today. A cursory glance at the early suburbs of Baltimore illustrates the evolution of the city’s first-tier suburbs. Figure 1.1 shows these first-tier suburbs on a map of metropolitan Baltimore. There are 21 first-tier suburbs that surround the central city (see appendix for geographic definitions). This book is about the evolution of these suburban places and their neighborhoods from 1970 to 2000. I am interested in the social, economic, demographic, and spatial changes that occurred in the first-tier suburbs relative to Baltimore City, the outer suburbs, and the metropolitan area. Therefore, this study employs spatial analysis throughout to chart suburban change. The case of metropolitan Baltimore offers an interesting and timely study of suburban change because it offers a fresh, new perspective on how an entire region’s first-tier suburbs evolve in a medium-sized Rustbelt metropolitan area at the brink of the twenty-first century. By the midpoint of the twentieth century, central cities dominated metropolitan growth in the United States. They were magnets for several newcomers from across the world, attracting some 25 million between 1900

suburban evolution / 5 Hampton

Lutherville Pikesville

Towson Overlea

Parkville Locheam Rosedale Middle River Woodlawn

Catonsville Essex

Arbutus

Dundalk Lansdowne Edgemere

Linthicum

Baltimore City

Pumphrey

Femdale

Glen Burnie Brooklyn Park 0

Figure 1.1

2.5

5

10 Miles

Map of Baltimore’s first-tier suburbs.

Source: Author; U.S. Census 2000 TIGER File.

and 1950. Central cities served as the economic backbone of the national economy. They included hubs of manufacturing, entertainment, and shopping. Furthermore, cities were the center of metropolitan gravity dominating big business, cultural and intellectual life of a burgeoning nation. The country’s urban history quickly began to change toward the end of World War II. Large-scale urban decentralization transformed the metropolitan landscape, particularly during the past 50 years (Beauregard 1989; Teaford 2006). The rim of urban development is now far from the traditional core. The result was a new metropolitan terrain that evolved with the rapid development of suburbs and the loss of central city prominence. In 1950, 60 percent of metropolitan America resided in central cities. In 1960, the U.S. metropolitan population was equally distributed between cities and suburbs. By the end of the century, only 37 percent of metropolitan residents lived in central cities. Nearly two out of every three people in metropolitan America now call suburbia home. There is no doubt that

6 / transforming race and class in baltimore

suburbs have become the favored site of development, consistently attracting more residents. Moreover, jobs, investment, and economic growth suburbanized during this period. The center of metropolitan gravity has shifted outward. In short, the United States is now a metropolitan society dominated by its multifarious suburbs. The traditional model of this metropolitan landscape posits a declining central city and expanding suburbs. A number of analyses were developed on this simple central city–suburban disparity. Owing to recent, largescale, sweeping suburbanization, however, the range of areas subsumed under the category of “suburban” has grown more heterogeneous, rendering obsolete the traditional model. Simply put, suburbia now represents a divergent set of landscapes. This complexity is explored in detail in chapter four through the presentation of the analysis of suburban neighborhood typologies in 1970 and 2000. This study complements other recent urban inquiries that have sought to identify differences among and within first-tier suburbs rather than between suburbs and central cities (Hanlon, Vicino, and Short 2006; Mikelbank 2004; Orfield 2002). In recent decades, suburbs have become sites of immense change. This presents danger and desolation as well as opportunity. The more recent spread of suburbanization has undermined the former advantages of the older suburbs. These areas were formerly able to capitalize on their proximate location to the central city. In reality, the demise of the central city has meant that these suburbs were no longer able to rely on the city for resources and amenities fully. The older suburbs, particularly those built in the 1950s and 1960s, were then no longer able to attract new development or new residents. These older places are located near the central city, and they are commonly referred to as “first-tier” suburbs. In many cases, these suburbs exhibit the very symptoms of decline that U.S. cities experienced some three decades ago. In contrast, newer suburbs, or “outer suburbs,” located further away from the core, are the primary sites of new development and investment (Hudnut 2003). This dichotomy has brought what urban scholars John Rennie Short, Bernadette Hanlon, and Thomas J. Vicino (2007) term as a “gothic” element to the older, first-tier suburbs. The term gothic refers to the grotesque or desolate—adjectives that are not typically associated with the suburbs. Yet many older suburbs, particularly those built in the postwar period, are bleak places. Many postwar suburbs struggle to survive—let alone flourish—in the new information economy. These suburban places were built in the era when smokestacks and factories dominated the economy. The suburban gothic defines the transformation of first-tier suburbs in the twenty-first century. To trace the evolution of the new suburban gothic, it is first necessary to explore the development of the U.S. suburban frontier. This volume

suburban evolution / 7

provides a review of the historical context of twentieth-century urban decentralization, and then examines three socioeconomic forces to explain the emergence of suburban decline and the new suburban gothic: the impacts of deindustrialization, the development of land-use planning, and the growth of an aging housing stock. U.S. Suburban Development Frontier is a useful analogy to apply to suburban development. It indicates the transformations of large-scale, rapid suburbanization. Scholars have used this term for U.S. urban studies for quite some time (Short 1991). In 1965, for example, Charles Abrams employed the term in the title of his book The City Is the Frontier to focus attention on the problems of declining central cities. Abrams then outlined a set of proposals to renew the city, which included increasing the amount of green space, making the city more attractive for leisurely activities, preserving neighborhoods rather than knocking them down, enhancing walkability space, making cities more female- and child-friendly. Other urbanists also have used the term. Wade (1959) used it to describe the life cycle of American Western cities. Frontier was also used as a rhetorical device to identify a new locale of biting social concern. Neil Smith (1996) applied the frontier imagery to describe the process of gentrification in selected inner-city neighborhoods. Furthermore, urban historian Kenneth Jackson (1985) used the term Crabgrass Frontier in his landmark study of American suburbanization. Joel Garreau’s (1991) book Edge City: Life on the New Frontier invoked the term to refer to forms of urban living at the edge of the metropolis and to evoke the buoyant optimism and individualism associated with the Western frontier. We can trace the historical evolution of the U.S. suburban frontier during the twentieth century through the identification and definition of a cycle of four periods: suburban utopia, suburban conformity, suburban diversity, and suburban dichotomy. Table 1.1 shows these periods and their characteristics. Table 1.1

U.S. suburban development

Period

Characteristics

Era

Suburban Utopias

Ideal image of suburbs as healthy and wholesome Suburbs are homogeneous, built in a cookie-cutter style Race, class, and ethnic divisions rise in the suburbs Declining older suburbs and growing outer suburbs

Late nineteenth to early twentieth century 1945–1960

Suburban Conformity Suburban Diversity Suburban Dichotomy

1960–1980 1980 onward

8 / transforming race and class in baltimore

Suburban Utopia During the nineteenth century, early suburbs acquired the image of ideal places for healthy living. These places provided refuge and a safe haven for society’s elite from the ever-so harsh environment of the city. The concentration of smoke-belching factories in the urban core of the central city made life unbearable for many city dwellers. Also, the elite sought to create distance from the factories and the working-class slum areas (Ashton 1978). Well-to-do families escaped to the fringes of the city. The urbansuburban fringe provided the best of both worlds for these early suburbanites. They were able to create their own utopia, utilizing the resources of their urban neighbor. During this initial phase of suburban development, what Fishman (1987, 1) describes as “bourgeois utopias,” suburbs were considered exclusive, safe, clean, and moral. According to Fishman (1987, 4), “from its origins, the suburban world of leisure, family life, and union with nature was based on the principle of exclusion.” Robert Fogelson echoes Fishman’s assessment. He observed that “[Suburbia] reflected a host of deep-seated fears that permeated much of American society” (Fogelson 2006, 24). In other words, they invoked an antiurban bias and a sense of escape from the city. Suburban Conformity Suburbia took on a new cultural form during its second period with the massive decentralization of housing and employment after World War II (Jackson 1985; Singleton 1973). In this era, an unprecedented shift occurred whereby residents of the city migrated outward, thus leading to the decentralization of jobs. In addition, there was mass production of standardized, Levittown-style housing, the age of the subdivision, all combining to produce the classic image of American suburbia as a homogeneous place of conventionality (Baxandall and Ewen 2000; Kramer 1972; Palen 1994). Examples of major suburban tract developments included Levittown, New York; Lakewood California; and Park Forest, Illinois. These subdivisions typified America’s suburbs that were celebrated in popular culture, especially during the 1950s. Television shows, such as Leave It to Beaver, Ozzie and Harriet, and Father Knows Best, were commonplace, and they disseminated the suburban dream through the airwaves one show at a time. These shows, and the American Dream they portrayed, embodied an era of domesticity and conformity that preceded the counterculture of the civil rights and feminist movements of the 1960s (Alves 2001). In addition, the role of the housewife was both celebrated and scripted by new household appliances and gadgets. Thus, television became the new medium to promote endless consumption and the

suburban evolution / 9

model family. The standardization of housing design added to this sense of conformity. As developers aimed to cut costs, they built thousands of “almost identical 800-square feet houses, with a living room, kitchen, two bedrooms, one bath, and a driveway” (Hayden 2003, 134). It is no surprise that suburbanites found comfort in conformity, and suburbs of the 1950s and 1960s provided the stable structure for that environment. Suburban Diversity The third period of suburban development is diversity. The prevalence of tract housing developments during the postwar era spurred studies on the uniformity and conformity of suburbia (Mumford 1968; Whyte 1956). Although some sought to dispel this myth of conformity, it was often assumed that all suburbs were middle class, white, and residential (Gans 1967). In reality, issues of class and race peppered America’s suburbs (Nicolaides 2002). A number of important studies in the 1960s and 1970s began to recognize the increasingly heterogeneous nature of suburbia. The identification of working-class suburbs (Berger 1968), and the rise of black suburbanization (Schnore, Andre, and Sharp 1976) contributed to the sense that the suburbs were not only the residences of white, middle-class families but also for racial minorities. These early studies emphasized racial differences (Blumberg and Lalli 1966; Farley 1964), ethnic variety (Kramer 1972), and class distinctions (Dobriner 1963; Pinkerton 1969). In addition, more recent studies have continued in this vein as the suburbs have witnessed more pronounced out-migration of middle-class blacks from central cities (Cashin 2004; Wilson 1987), large-scale immigration of the foreign-born and the development of ethnoburbs and aging suburbs (Frey 2003a; 2003b; Li 1998). A recent study identified a wide distribution of suburbs including manufacturing suburbs, black suburbs, and immigrant suburbs (Hanlon, Vicino, and Short, 2006). Despite the dominant cultural representation of suburbs as affluent, white, and residential, it is more apparent that suburbs come in various shapes and sizes with an array of demographics and economies. Since the suburbs now stretch across space and eras of building, they now comprise more diverse places. The variation among suburban places is often more striking than the difference between the central city and its surrounding suburbs. Collectively, these studies demonstrate that there are a large variety of suburbs in the United States. Suburban Dichotomy The fourth period of the suburban developed is defined by the emergence of a striking suburban dichotomy. Two contrasting elements characterize

10 / transforming race and class in baltimore

the dichotomy: the sprawling edge and the declining first-tier suburb that reflect the ongoing rolling frontier and the historical legacy of suburbanization. As the suburban frontier crept outward in both space and time, the early postwar suburbs were left behind, literally and metaphorically. The continued outward suburban expansion was coupled with the simultaneous decline of older suburbs located near the central city. The industrial, working-class suburbs, for instance, that developed during the height of the industrial revolution (Lewis 2004) have been replaced in recent decades with office park and retail developments in the outer suburbs (Lang 2003). Some scholars refer to these sprawling suburbs as “edge cities” or “boomburbs” (Garreau 1991; Lang and LeFurgy 2007). Consider the infamous case of the edge city Tyson’s Corner, Virginia. Once a quiet area located some 20 miles west of Washington, D.C., Tyson’s Corner grew substantially during the 1980s and 1990s, the growth in large part lubricated by highway construction. The growth of this suburb is characterized not only by surrounding residential growth, but also by an increase in jobs and employment infrastructure. The area boasts more than 25 million square feet of office space and 4 million square feet of retail space. Approximately 110,000 people work in the immediate area. It is the twelfth largest business district nationally. Several factors have shaped this pattern of development. The availability of cheap land on the exurban fringe of metropolitan areas has allowed, although the powerful development lobby and its political allies have promoted, unfettered growth and large-scale housing developments. The lack of mass transportation, either bus or rail, has meant that these suburbs rely exclusively on automobiles. An endless pattern of suburban sprawl is the result, best described as a “pattern of urban and metropolitan growth that reflects low-density, automobile-dependent, exclusionary new development on the fringe” (Squires 2002, 2). This brings us to the second element of the suburban dichotomy, firsttier suburban decline. It is impossible to decouple the phenomenon of suburban edge sprawl from the demise of older suburbs (Orfield 2002). Although much research focuses on the decline of U.S. central cities and the growth of sprawling suburbs, in more recent years, scholars have also turned their attention to decline in suburban areas (Bollens 1988; Hudnut 2003; Lee and Leigh 2005; Listokin and Beaton 1983; Lucy and Philips 2000; Orfield 2002; Smith, Caris, and Wyly 2001). Typically, decline occurs in older suburbs characterized as communities with slow population growth, few local resources, and declining local economies. Some mature suburbs that were once vital centers of economic activity are facing severe fiscal problems, increasing minority populations, and an aging housing stock (Lucy and Phillips 2006). The infrastructure of roads, schools, and

suburban evolution / 11

houses in these suburbs is old and decaying, and the residents are aging without a future generation to replace them. In short, the decline in these suburban areas, which can be characterized as the emergence of a new suburban gothic, contrasts sharply with the boom of newer suburbs. The Creation of the New Suburban Gothic Suburban decline is the ultimate force shaping the new suburban gothic. The persistence of metropolitan decentralization characterizes this phenomenon. During the course of the twentieth century, people and jobs continually moved away from the urban core in the central city to the suburbs. As decentralization continued, it became evident that some of the areas that were first to suburbanize were ailing by the end of the century. There are three noteworthy forces among the myriad of influences contributing to this phenomenon. First, deindustrialization of older suburbs left these places with a smaller economic base for opportunity and mobility. Second, land-use planning and exclusionary zoning at the local scale resulted in greater suburban differentiation, separating desirable suburbs from those deemed less attractive. Third, a strong demand-supply nexus that promoted substantial greenfield development. The demand for larger housing units, combined with the very powerful development lobby, influenced local land-use planning and led to the construction of new housing and commercial buildings on the outer suburban fringe. Thus capital disinvestment in the housing stock of first-tier suburbs fueled the decay and aging process. Let us now briefly examine each of these forces in more detail. Deindustrialization Working-class suburbs have long been a part of the suburbs (Nicolaides 2002), and heavy industry has been located in suburban districts outside American cities since the mid-nineteenth century (Lewis 2004). To illustrate this example, consider the case of Dundalk, Maryland, an industrial, working-class first-tier suburb of Baltimore that developed during this period. Dundalk was largely undeveloped until 1887 when the Pennsylvania Steel Company recognized its prime location as a tidewater port with important access to the Atlantic Ocean. In 1916, the area was purchased by the Bethlehem Steel Company, and the steel mill in the Dundalk area expanded to become one of the largest steel production sites in the world (Reutter 2004). A year after buying the site, Bethlehem Steel bought a thousand acres of land for expansion, formed the Dundalk

12 / transforming race and class in baltimore

Company, and built houses for its labor force. The Dundalk Company hired E.H. Bouton, designer of the prestigious, elite residence of Roland Park on the northern side of town, to design a “workingmen’s Roland Park” close enough to commute to the Bethlehem Steel plant. Therefore, Dundalk became the ideal company town. The global restructuring of the economy meant that Bethlehem Steel would begin downsizing its labor force from 1960 to 2000. Dundalk suffered from tremendous social and economic decline during the demise of its own developer. Metropolitan Baltimore is by no means unique in the early development of industrial suburbs. Boston, too, had a large suburban ring of industrial and commercial districts by 1900 (Warner 1978). Employment in Boston was spatially diffused throughout the region by the late nineteenth and early twentieth century. Likewise, Pittsburgh, Detroit, and New York witnessed substantial industrial suburbanization as early as the mid-nineteenth century. The decentralization of industry from older central cities to the suburban fringe accelerated after World War II. For example, New York City lost 6 percent of manufacturing jobs while its surrounding suburbs gained approximately 37 percent between 1947 and 1958. During the same period, Chicago lost more than 18 percent of its manufacturing employment while the suburbs gained almost 50 percent; and Cleveland lost 22 percent of manufacturing jobs while manufacturing employment in the suburbs grew by nearly 100 percent (Berry and Cohen 1973). Just as industrialization led to the growth of some suburbs, deindustrialization led to their decline. The economic shift away from heavy manufacturing resulted in the loss of relatively high-paying, unionized jobs for middle-class residents in many older suburbs. Without a doubt, deindustrialization took a toll on first-tier suburbs. The impacts on the local economies of these communities cannot be understated (Bluestone and Harrison 1982). The economic restructuring of the economy away from manufacturing and toward services favored the growing, outer suburbs as well as the central business district. Spatial inequalities resulted because greater economic opportunity relocated to outer suburbs, thus reinforcing the decentralization of people and jobs in regions. In the traditional manufacturing regions of the United States, older, blue-collar suburbs witnessed the effects of runaway industry and, in the case of metropolitan areas such as Los Angeles, the abandonment by such industries as aerospace and defense had tremendous impacts on older suburban areas (Davis 2005). Deindustrialization follows the trail of industry, and just as inner-city plants were often the first to close, it is the turn of first-tier suburban plants now. The new suburban gothic is characterized by the blue-collar job loss in the wake of deindustrialization.

suburban evolution / 13

Land-Use Planning Because of the fragmented nature of local government in the United States, the typical metropolitan area is composed of a central city surrounded by politically separate, autonomous suburban municipalities. Consequently, suburbs often compete with their central city counterpart, and each other, for new investments in the built environment. Each suburban municipality seeks to attract new development to increase its tax base. The result is the persistent development of the ever-expansive suburban frontier of the metropolis. There may be a limit to the pro-growth stance of suburbs as more established residents might lobby for slow-growth or no-growth to resist the declining quality of public life brought about by too rapid growth. Various affluent outer suburbs have generated slow-growth movements prompted by popular resistance to the increasing costs of congestion, school overcrowding, increased journey to work times, and a general senses of a decline in the quality of life. For instance, as a mature jurisdiction, Baltimore County planners sought to limit the development of their green space and rural heritage. Therefore, during the late 1960, the county implemented an “urban-rural demarcation line” to discourage new development outside of its targeted growth areas. Moreover, the case of Portland is an even more progressive case of growth restrictions. Portland’s regional government implemented a region-wide growth boundary to promote sustainable development practices (Ozawa 2004). Given the fractured and decentralized nature of planning practices like zoning, it is no surprise that these examples stand out as atypical of many other metropolitan areas, where little-to-no growth control is present. The Demand-Supply Nexus Suburbia has witnessed a housing boom that is now more than 60 years old. Once prized as the ideal location for families, many suburbs now exhibit symptoms of aging. Those housing units built during the postwar period of mass suburbanization are now particularly outdated. Innovative and highly desirable when first built, the postwar Cape Cod or suburban ramblers now represent a bygone era. This housing stock lacks the size and amenities to compete with newer housing on the outer metropolitan fringe. Housing in the postwar suburbs is far smaller than newer housing. Today, there is an unabashed desire for large developments, and the typical contemporary house is more than 2,200 square feet in size. The demand for new, large housing units is more than twice the size of the average house built in 1950. In a few locations such as the older suburb of Pimmit Hills outside Washington DC, a number of modest postwar houses are under renovation

14 / transforming race and class in baltimore

or are torn down and replaced with larger, newer housing (Straight 2005). In the 1950s, Pimmit Hills was the largest subdivision in Fairfax County, Virginia. Located inside Washington DC’s Capital Beltway, the original housing in this suburb was typically comprised of 3 bedrooms, 1 bath, and less than 900 square feet in size. Yet, in most cases, residents who pay a premium to reside in close proximity to the nation’s capital have made at least one addition to their houses. On most streets, at least one extremely large “McMansion” house replaces the original boxy structure. This type of expensive house upgrading seems to occur only in a few regions where the demand is still relatively strong for large-scale downtown employment opportunities, an affluent middle class, and an overheated housing market. In other less popular regions, small postwar suburban housing is rarely upgraded. In Lansdowne, for instance, an older industrial suburb of Baltimore, the housing stock typifies the suburban tract developments of the 1950s and 1960s and is now considered poor quality stock in need of repair. Yet, with 14 percent of local residents living in poverty, and the median household income of this older suburb 40 percent lower than the median household income in the Baltimore region, there is little immediate likelihood of private investment by homeowners or landlords to rehabilitate these aging structures. Consequently, the older postwar suburb has become the devalorized urban form. The notion of devalorization is typically used to explain the decline of central city neighborhoods. Neil Smith (1996), in his analysis of gentrification in the city, suggests that inner-city properties were devalorized or devalued because of the reallocation of capital to the suburbs. In a similar manner, the aging housing stock in the postwar suburbs is devalued. Despite the need for an injection of capital, disinvestment is widespread. The first-tier suburbs continue to lose out to edge city development and the revitalization of housing in central city neighborhoods. Caught between central city gentrification and outer suburban sprawl, many postwar suburbs are currently losing the battle for investment resources. A large pro-growth regime in the housing and transportation sectors only serves to promote this devalorization further. Development on greenfield sites is especially favored because of its ease, speed, and hence greater profitability. The new suburban gothic is one form of devalorization of older suburbs. This process began in the central city, and now, it has spread out to the first-tier suburban areas as residents and capital decentralize even further from the metropolitan core. The result is a downward spiral of declining investment and socioeconomic status of residents. First-tier suburban America has now fully matured. Initially the home of the elite, then the middle class, suburbia today houses a range of people from the wealthy to the poor and a range of local economies from the economically buoyant

suburban evolution / 15

to the economically depressed. Moreover, suburbia has emerged as a place of disparate and diverging realities. Social and economic problems are apparent in numerous first-tier suburbs, many now more than 50 years old. Population growth in the first-tier suburbs has stagnated. Compared to their counterparts in the growing outer suburbs, the residents of these areas have grown increasingly poorer since 1970, and the housing stock has aged significantly. These older neighborhoods built primarily in the immediate postwar period are part of a new suburban gothic of places of desolation and decay in the suburban landscape. By the end of the twentieth century, three forces led to the emergence of this suburban gothic: the process of deindustrialization, the nature of local land-use planning, and the demand-supply nexus. Overview This study explores the process of suburban change in metropolitan Baltimore from 1970 to 2000. Using a mixed methods approach, I analyze patterns of suburban change in the first-tier suburban communities of Baltimore from 1970 to 2000. More formally, the goals of this study are fourfold. First, it seeks to develop a spatial definition of first-tier suburbs. From an urban geographic perspective, a clear definition is lacking, and one is needed for methodological purposes. The second goal of this study aims to examine how Baltimore’s first-tier suburbs have changed since 1970. Third, this study attempts to identify and explain patterns of suburban differentiation among and within Baltimore’s first-tier suburbs. Fourth, this study then sets out to reflect and discuss implications of suburban decline for public policy and urban planning. Scholars have increasingly studied the suburbs because they are sites of immense change—just like their urban neighbors. Indeed, suburbs are worth scholarly attention for numerous reasons. Urbanized areas with large population concentrations offer a rich research observatory lab for they are complex systems with complex human interactions. Social and economic changes occur, and it becomes important to examine how and why that process unfolds. As the United States morphed into a majority suburban society, more Americans than ever live, work, recreate, and vote in the suburbs. Accordingly, as the second and third generation of suburbanites grows up, it is timely, interesting, and analytical to inquire about socioeconomic transformations taking place. The suburban century provides an opportunity to examine urban processes in suburban environments and test long-standing theories. The book is organized into seven chapters and an appendix. In chapter two, I explore the history of suburbanization. I review the literature about

16 / transforming race and class in baltimore

neighborhood change and invasion and succession processes. This literature demonstrates that suburbs, like central cities, experience neighborhood changes and life cycles, which are the product of private actions and public choices. An understanding of the public policies and private preferences that led to large-scale changes in suburban neighborhoods is integral to the analysis of suburban change. This gives context to the evolution of how suburbs change, be it growth or decline. In chapter three, I examine how Baltimore’s first-tier suburbs changed from 1970 to 2000. Using the U.S. Department of Housing and Urban Development’s State of the Cities Data System and Geolytics’ Neighborhood Change Database, I analyze data for 50 socioeconomic variables using census place level data. I conduct a descriptive spatial statistical analysis by focusing on four primary areas of suburban change: the population characteristics, income dynamics, nature of the housing stock, and labor force structure. To analyze suburban change, I stratify and compare the data in four geographic areas: first-tier suburbs, outer suburbs, central city, and region. This analytical approach provides an opportunity to examine patterns of suburban change relative to four other geographic scales in metropolitan Baltimore. Evidence of suburban decline was abundant on multiple indicators for each of Baltimore’s 21 first-tier suburbs. In chapter four, I analyze the process of first-tier suburban change at a finer geographic scale. By conducting two additional quantitative analyses at the census tract level, I capture variation of suburban change within 152 neighborhoods of first-tier suburbs. First, I analyze how patterns of suburban change evolved between 1970 and 2000 through a principal components analysis of the data at the census tract level. Then, I use a cluster analysis technique to classify first-tier suburban neighborhoods in 1970 and 2000. These analyses show the transformation of race, economic status, age, and labor force in first-tier suburban neighborhoods in metropolitan Baltimore. In chapter five, I demonstrate the implications of the findings for public policy and urban planning. In this case, the Baltimore County government confronted suburban decline by creating the Office of Community Conservation. This office was charged with revitalizing Baltimore’s first-tier suburbs. In 1994, the county began carrying out revitalization projects in the county’s declining eastern first-tier suburbs, focusing particularly on formerly heavy industrialized suburbs. I explain the political realities and economic imperatives that made the revitalization of first-tier suburbs possible, and then reflect on the lessons learned. In chapter six, I consider the prospects of state and federal policy initiatives for suburban decline, and then I argue that the new metropolitan dilemma in the twenty-first century is to save the suburbs. Policymakers

suburban evolution / 17

and planners can capitalize on the advantages of first-tier suburbs such as their infrastructure and location, affordable housing, and sustainable growth. A suburban policy agenda is presented. Last, in chapter seven, I conclude with the presentation of the study’s main contributions and a research agenda. In the end, this study documents how America’s largest frontier grew and then precipitously declined at the end of the twentieth century. This story sheds new light on a suburban landscape that we may or may not be familiar with—its disparities, its diversity, and its ultimate decline.

This page intentionally left blank

Ch a p t e r Two Su bu r ba n Fron t i e r

Suburbia conjures up many iconic images of the so-called American Dream. From the classic 1950s television show Leave It to Beaver to today’s risqué, smash-hit show Desperate Housewives, suburbs have been idealized in popular American culture. These suburban images depict a spacious house on a tree-lined street where white, middle-class Americans had a place to call home after World War II. Indeed, the pursuit of happiness in U.S. postwar history has been completely synonymous with the suburbs. A move to the suburbs symbolized many things in the American context. It was a move of social and economic mobility—a path that led away from the nation’s ailing central cities and to the emergent suburban frontier. The so-called American Dream was realized in the nation’s nascent suburbs (Baxandall and Ewen 2000; Downs 1973; Fishman 1987; Hayden 2003). Yet the suburbs did not happen by accident, and they all were not rosy, optimistic places after all. The development and subsequent growth of the suburban frontier was the product of a deliberate effort on behalf of the federal government to alter the pattern of development and shape private preferences in the marketplace. The result was not only the creation of a suburban nation, but also the evolution of a large heterogeneous suburban landscape. This chapter presents the evolution of scholarly thought on the roles that political processes and public policies played in the development of the suburban landscape. First, the history of the federal government’s involvement in promoting suburbanization and key public policies on housing and transportation are examined. Then, an overview of an emerging body of scholarship is presented to offer a revisionist suburban history, focusing on competing perspectives on the historical patterns of suburbanization. Three competing perspectives are reviewed, including the push-pull model, the metropolitan political framework, and the racespace intersection. Next, a primer on the rise of suburban Baltimore is presented, reflecting on the patterns of growth in the built environment, transportation, and urban development. Finally this chapter synthesizes

20 / transforming race and class in baltimore

the scholarship on the dynamics of neighborhood change to demonstrate the evolution of the suburbs from frontier to a full-scale metropolitan landscape. Crabgrass Memories To reflect on the political history of U.S. suburbs requires an examination of the role that public policy played in forming and shaping the suburban landscape. As early as the 1950s, critiques and negative social commentary about public policy and the suburbs were abundant. For instance, early historical accounts compared life in the suburbs to the life in central city. Social critics such as Lewis Mumford and William Whyte criticized the suburbs as banal and as places of conformity. For instance, in his book The Organization Man, Whyte (1956) argued that suburbia was plagued with classlessness. In other words, the suburbs were home to a population that was comprised of one uniform socioeconomic group—they lacked diversity. Similarly, in 1961, Lewis Mumford (1968, 486) offered a scathing indictment of the suburban environment as a place of “uniform, unidentifiable houses, lined up inflexibly at uniform distances, on uniform roads, in a treeless communal waste, inhabited by people of the same class, the same income, the same age group . . . conforming in every outward and inward respect to a common mold.” Other early critics echoed similar critiques of suburbs. In her timeless book The Death and Life of Great American Cities, Jane Jacobs (1961, 445) offered an even harsher indictment of suburban living, arguing that suburbs were “thin dispersions [that] lack any regrettable degree of innate vitality, staying power, or inherent usefulness as settlements.” Even though these critics viewed the suburbs as a negative force in 1950s American society, neither suburban growth nor its ancillary public policies dissipated. The outflow of people and jobs from central cities to suburbs suggests that this pattern was a “staying power.” Suburbia remained politically and culturally popular through the end of the twentieth century despite various social critiques. As the suburban society fully amalgamated metropolitan America in the second half of the twentieth century, studies and critiques of suburbs shifted focus. Although earlier studies in 1950s and 1960s concentrated on the social and cultural aspects of suburbia, subsequent scholarship during the 1970s and 1980s began to critically analyze the role that public policy played historically in the creation of a suburban society. Urban scholars of this generation, including historians Kenneth Jackson (1985) and Mark Gelfand (1982), argued that suburbia benefited from an assortment of public policies that helped cement a large, middle-class society located just outside central cities. Jackson and Gelfand were two

suburban frontier / 21

prominent figures who argued that the new suburban landscape would not have come to fruition without a concerted and deliberate public policy campaign. As the former hinterlands of the city, suburbia required houses for its new population and roads to travel throughout the region. Therefore, historians of suburbia during this era put forth that two major policy areas had especially important consequences for the development of suburbs. First, they held that federal transportation legislation in the wake of World War II spurred the construction of a massive road network that paved the way to suburbia. Second, they contended that a package of federal housing policies provided the means to build scores of new housing units and ultimately stimulated the market for the private development of suburbs nationwide for decades to come. Thus, scholars of the suburbs during the 1970s and 1980s primarily took a federal view of the history of suburbs, showing that public policy played a role in shaping the transformation of the metropolitan fringe from bucolic suburbs to developed suburbs. Without a doubt, the construction of an immense network of highways and roadways throughout the nation’s metropolitan areas spurred the growth of suburbs. In 1956, the 84th U.S. Congress passed the FederalAid Highway Act to plan and construct a 41,000-mile interstate highway system. Popularly known as the National Interstate and Defense Highways Act, President Eisenhower’s original public policy goal was to provide a transportation network that would allow for the quick and easy mobilization of U.S. military troops and supplies in the event of a national emergency. As Congress began to debate the legislation, Eisenhower’s policy rationale quickly became symbolic as private, corporate interests quickly realized the potential economic impact of such a large-scale public works project (Jackson 1985). The nation’s three largest automakers, Ford, General Motors, and Chrysler, successfully lobbied Congress for the passage of the transportation legislation. This lobbying coincided with the automakers’ peak production during the twentieth century. The dramatic increases in the number of automobiles nationally provide evidence of the industry’s success. In the 3 decades after World War II, the number of automobiles quadrupled from 25 million to 106 million cars in 1975. In the 3 subsequent decades, the number of automobiles doubled, reaching 210 million cars in 2005. This was not the only instance when automotive companies promoted the construction of roadways. Before the expansive highway projects of the 1950s, the automotive industry helped to accelerate the decline of streetcars and trolleys in urban centers. When governments across the country decided that mass transit should be financially self-sufficient, user fees were implemented to cover the costs of ridership. Declining passengers led to increased financial insecurity as automobiles

22 / transforming race and class in baltimore

became the preferred mode of transportation. For example, General Motors was found to be guilty of criminal conspiracy when it replaced the streetcar operations in big cities such as Los Angeles, St. Louis, Philadelphia, and Baltimore with their own company’s manufactured buses. In short, the government’s transportation policies, along with its tacit endorsement to the nation’s major automotive industrial corporations, cemented the roads that led to the new houses of suburbia. This public infrastructure was critical for the growth of suburbs nationally. Kenneth Jackson and Mark Gelfand assert the role that twentiethcentury federal housing policy played in the development of suburbs cannot be understated. At the brink of the twenty-first century, the housing sector collectively represented approximately $2 trillion or onefifth of the nation’s $9 trillion-plus gross national product (Euchner and McGovern 2003). Since 1950, furthermore, the housing sector consistently represented just under a quarter of the gross domestic product (Schwartz 2006). The postwar economic impact of housing on the U.S. economy has been colossal to say the least. Federal housing policies on subsidies such as the mortgage-interest deduction, deduction of property tax payments, and low-interest mortgages for first-time homebuyers contributed to this economic impact. These policies created undeniable incentives for urban residents to purchase a new house and join the ranks of the two-thirds of Americans who are now homeowners. Moreover, the homebuilders like the Levitt Brothers capitalized on the availability of cheap land on the metropolitan fringe and took advantage of low-interest government loans in housing to construct and mass-produce houses for the first generation of suburbanites. Last, discriminatory housing policy encouraged the private real estate practices of redlining and blockbusting to ensure the growth of a racially homogenous suburban landscape. Altogether, these policies can be attributed to the government’s long-term role in the housing sector, and scholars have recognized the overall impact of this involvement as synonymous with the process of suburbanization. Levitt and Sons, the nation’s preeminent homebuilder during the late 1940s and 1950s, played an early, yet pivotal, role in building suburbia. They were one of the first entrepreneurs to capitalize on the soaring demand for housing in the years following World War II. Postwar urban America was overcrowded with returning war veterans anxious to start new families. Unmistakably, cities were bursting at their seams, and many observers of urban America were calling for change. Whereas others viewed these conditions as the new urban dilemma, the Levitts, as astute businessmen, saw an opportunity. Abraham Levitt and his two sons, Alfred and William, helped build the way to suburbia through the development of various new suburban towns along the East Coast. Aptly named

suburban frontier / 23

“Levittown,” the first community was built on Long Island, New York on a former 1,200-acre potato farm. Demand for new houses in Levittown was enormous. In 1947, the Levitts initially built 2,000 rental homes, and within days, every unit was rented to a veteran and his family. Therefore, the Levitts announced the construction of an additional 4,000 housing units for sale. By the end of the twentieth century’s midpoint, Levitt and Sons had constructed 17,447 houses in their new suburban town. Similar Levittowns were later built in New Jersey, Pennsylvania, and Puerto Rico (Gans 1967; Kelly 1993). In 1950, Time (1950, 7) on their magazine cover article on the Levitts noted that “The houses in Levittown, which sell for a uniform price of $7,990, cannot be mistaken for castles. Each has a sharp-angled roof and a picture window, radiant heating in the floor, 12-by-16 ft. living room, bath, kitchen, two bedrooms on the first floor, and an ‘expansion attic’ which can be converted into two more bedrooms and bath. The kitchen has a refrigerator, stove and Bendix washer; the living room had a fireplace and a built-in Admiral television set.” Potential buyers only needed down payment of $90, and mortgage payments were $58 a month for a house that cost less than $8,000. On the basis of inflated dollars in 2006, this translates into a down payment of $735, and monthly mortgage payments of $475 for a $65,000 house. Remarkably affordable, this meant that homeownership was within the reach of a budding middle class, ready to blossom in the new suburban world. Indeed, a house in Levittown was simple yet adequate for new households about to give birth to the largest surge in the U.S. population—the Baby Boom. In terms of production, the Levitts refined a number of innovative home building techniques that would become the hallmark of early suburban development. It was imperative for this construction company to implement efficient uses of materials, workers, and time, for without these methods Levitt and Sons could not have offered such cheap houses. The Levitts exclusively employed nonunion construction workers to keep labor minimal. To economize the time to build a house, all lumber was precut and shipped directly to the construction site. In addition, roofs and support walls were also prebuilt and delivered to Levittown. Houses did not have basements, and the company only offered a few standard versions of the house: a ranch-style or a Cape Cod–style. This uniformity of construction allowed Levitt employees to labor in an assembly line fashion, with each worker responsible for a particular aspect of the house construction. The Levitts learned from automotive industrial pioneer Henry Ford that houses could be built in a similar way to cars. The results paid off. By 1948, Levitt and Sons produced 30 new houses a day, and demand remained consistently high through the 1950s.

24 / transforming race and class in baltimore

The achievements of Levitt and Sons were only, in part, owing to their innovative production techniques and forward thinking. Like many other new homebuilders nationally, they also enjoyed huge financial successes as the benefactors of federal housing public policies that promoted homeownership among white, middle-class residents. On the unprecedented growth of new housing in the United States, William Levitt once remarked that “if it weren’t for the Government, the boom would end overnight” (Time 1950, 7). His observation could not have been more fitting. As early as 1934, federal housing policy drove the growth of the housing industry. Capitalists such as the Levitts understood early on that the government’s housing policies would greatly benefit their industry, and ultimately the suburbs that they would build. Table 2.1 provides a summary of the major federal policy initiatives for the suburbs during the twentieth century, many of which focused on the aspect of housing. Perhaps no other public policy impacted suburban development more than the creation of the Home Owners Loan Corporation (HOLC). Historically, less than half the number of Americans owned a house. From 1900 to 1940, the home ownership rate remained stagnant at approximately 43 percent. After World War II, the rate increased to 55 percent in 1950, and then rose to 62 percent a decade later. In subsequent years, the Table 2.1 Overview of twentieth-century federal policy initiatives for the suburbs Policy initiative

Year(s)

Description

National Federal Housing Act

1934

Federal Home Loan Bank Board Redlining Policy National Federal Housing Acts

1935–1968

National Federal Housing Act

1964–present

Serviceman’s Readjustment Act (“GI Bill”)

1944–present

Federal-Aid Highway Act

1956–1991

Provided insurance to private mortgage loan company to protect lenders from risk and encourage long-term housing loans Developed “residential security maps” to segregate neighborhoods, fueling white flight from cities Created subsidies for slum clearance and urban renewal; required redevelopment of blighted areas Created incentives to streamline FHA mortgage loans to meet additional demand for housing Provided Word War II veterans with a variety of social benefits including college tuition support and low-interest, zero down payment loans for new housing in the suburbs Connected cities, suburbs, and metropolitan areas with more than 41,000 miles of roads and highways

1949; 1954

suburban frontier / 25

proportion of Americans who were homeowners continued to grow such that by 2000, two out of three households had their resident owning the house. The tremendous growth in home ownership rates can be attributed to the policy goals of the HOLC. Created in 1933, HOLC’s mission was to dramatically lower housing foreclosures that had become so commonplace during the Great Depression. Using federally issued bonds, the HOLC purchased and refinanced more than 1 million mortgages at a cost of $3 billion during its first 2 years of practice (Schwartz 2006). In his landmark book on U.S. suburbanization, Crabgrass Frontier, historian Kenneth Jackson (1985, 196) noted that HOLC “introduced, perfected, and proved in practice the feasibility of the long-term, self-amortizing mortgage with uniform payments spread over the whole life of the debt.” Jackson captures the essence of the impact of the HOLC’s practices on housing policies that subsequently led to a long-term, sustained boom of new construction of housing in the suburbs. In addition, the National Housing Act of 1934 planted the seed for a suburban housing growth spurt. The Federal Housing Authority (FHA), organized by the Roosevelt administration after the creation of the HOLC, was charged to reduce unemployment and grow the housing sector of the economy. The 30-year mortgage loan, guaranteed by the FHA, was the primary tool that the agency used to accomplish its mission. The Serviceman’s Readjustment Act of 1944 complemented FHA’s efforts. Commonly known as the “G.I. Bill,” this legislation was aimed at returning war veterans to assist them with reintegration into society (Keene 2001). Veterans were able to take advantage of a vast array for social services. For example, health care benefits and college educational opportunities provided veterans a pathway to advance into the middle class (Mettler 2005). In addition, the government provided generous housing support to aid veterans with the purchase of a new house. Low-interest, zero-down payment loans were often afforded to these veterans. Last, veterans and new suburbanites alike benefited from housing legislation to allow these new homeowners to deduct the mortgage interest from federal taxes, thus encouraging even more residents to purchase houses. Later revisions to the legislation also provided tax deductions for state and local property taxes and other home improvements. By the early 2000s, the total value of these deductions approached $120 billion—making it one of the largest domestic policy expenditures (Schwartz 2006). These housing policies thus became standard instruments in the federal government’s toolkit to both stimulate the economy and grow a new suburban frontier. In fact, they were just the stimulus that suburbia needed to hastily grow on the metropolitan fringe.

26 / transforming race and class in baltimore

Not all federal housing policies were as equally beneficial for all Americans. Racial and ethnic minorities, particularly black Americans, were often subjected to and the victims of discriminatory housing policies on the part of the federal government as well as the private market. One of the earliest institutional discriminatory policies was the practice of redlining. This term was often used to describe a mortgage lender’s refusal to grant loans in particular neighborhoods of a city, or make loans on less favorable terms, on the basis of racial composition of those areas (Hillier 2003). In 1935, the HOLC developed maps with four neighborhood classifications in the largest cities throughout the nation to determine its lending practices. The maps were coded with the following names: Classification A was “Best”; Classification B was “Still Desirable”; Classification C was “Definitely Declining”; and Classification D was “Hazardous”—otherwise known as the “red” area. It was in these areas, Classifications C and D, that a red line was drawn around these neighborhoods on the map, which then made it difficult for blacks and other minorities to obtain decent loans terms, if any, to purchase a house. Kenneth Jackson showed how the federal government used its own Underwriting Manual to guide the decision making process in granting FHA mortgage loans. Jackson (1985, 208) quoted the manual as declaring, “If a neighborhood is to retain stability, it is necessary that properties shall continue to be occupied by the same social and racial classes.” Black Americans segregated in central cities were systematically excluded from the affordable housing opportunities that so many white, middle-class Americans enjoyed at the time. Redlining was not the only way that the government and real estate market discriminated against racial and ethnic minorities. The use of restrictive covenants complemented the practice of redlining in subtle ways. A restrictive covenant is a legal obligation that becomes part of the deed of a house and property that stipulates that the seller and buyer abide by certain rules (Fogelson 2006). During the first half of the twentieth century, such covenants were used to maintain the perception of stable neighborhoods. For example, covenants typically included restrictions on the racial and ethnic composition of the potential buyer, limits on property uses, limits on noise, and requirements to maintain yards and gardens. Covenants were legally binding contracts that required anyone who owned a property, then or in the future, to agree to such conditions. Black Americans were systematically excluded from purchasing houses in neighborhoods with such restrictions. Although the U.S. Supreme Court ruled restrictive covenants based on race or color were unconstitutional in 1948, de facto housing segregation persisted for many decades—and it continues to be one of the defining features of the suburban housing market (see Shelley v. Kramer 334 U.S. 1). Sociologists Douglas Massey and Nancy

suburban frontier / 27

Denton (1993) define this characteristic as a grave social problem, in their aptly titled book American Apartheid. Similarly, the real estate practice of blockbusting was an indirect housing policy that was discriminatory; it also stimulated suburban growth. The story of blockbusting in suburban Baltimore provides a case in point. In the early 1950s, Edmondson Village was an older suburb located on the western suburban fringe of Baltimore City. Real estate agents attempted to convince white residents to sell their houses at below market prices. Exploiting residents’ fears of plummeting property values, agents would tell white residents that blacks from the city were about to move into their neighborhood. Scare tactics were often employed to make the case to white residents to sell. For example, urban historians have documented that black women were paid to walk their baby carriages throughout white neighborhoods as real estate agents knocked on the doorsteps of white residents, hoping to visualize the effect of neighborhood racial change (Seligman 2005). Such practices were successful in Baltimore, among other cities. After white residents sold, real estate agents would increase the price of the house with the direct intent of selling to black city residents in search of the suburban dream. Between 1955 and 1965, nearly every white resident—some 20,000—fled Edmonson Village, and an exclusively black residential population replaced them in just a decade (Orser 1994). Both the real estate industry and the FHA failed to confront this systematic discrimination until the passage of the Fair Housing Act of 1968, which is part of Title VIII of the Civil Right Act of 1968. Looking back on the suburban frontier during the 1970s and 1980s shows that suburban historical accounts were focused on the federal government’s role in U.S. housing and transportation policy. This clearly aided the development of a suburban nation in postwar America. Jackson’s and Gelfand’s contributions represented an important step forward in our understanding in the role of government in suburbanization. These scholars effectively showed that government played an important role in shaping this transformation from the city to suburb. The federal housing administration guaranteed and subsidized mortgages for scores of returning war veterans, and advances in the technological production of housing meant that many new suburbanites had a new place to call home. Yet, not all Americans benefited equally. Racial and ethnic minorities, particularly blacks, were unable to gain access to the suburban dream owing to the discriminatory practices of redlining, restrictive covenants, and blockbusting. The government’s failure to alter the pattern of urban decentralization gave way to the creation of a suburban landscape fractured by class and race.

28 / transforming race and class in baltimore

The Push and Pull of the Frontier Scholars have developed a useful “push-pull” framework for understanding the public role in the development of the metropolitan fringe. This framework holds that a number of socioeconomic conditions in central cities pushed residents out of the urban core. Similarly, a variety of characteristics unique to the suburbs pulled residents to this new frontier. The push-pull process began after World War II, as city residents grew dissatisfied with the state of urban living. The city was increasingly home to an overcrowded, dense array of neighborhoods, coupled with an aging infrastructure, which made it an unattractive place for returning war veterans to start a family. The suburbs offered an attractive alternative environment that featured new houses with a yard and a variety of other new amenities that residents perceived the city lacked. Thus, these forces together gave birth to one of the greatest population migration movements in the history of the nation— the journey from the city to the suburb. Jon Teaford (2006) presents a compelling treatise on the transformation of the American in his book on the American metropolis entitled The Metropolitan Revolution. He implicitly puts forth the push-pull framework to understand the changes in the distribution of the metropolitan population, namely the movement from city to suburb. In the early phases of mass suburbanization, white, middle-class residents comprised the bulk of migrants to the suburbs, although black, lower-class residents remained in the central city. The decentralization of the economy also contributed to the massive growth of suburbs. Yet another factor that pushed residents to the suburbs was the lure of a new public infrastructure and modern amenities. Urban renewal efforts to save the city served as the final major push factor to the suburbs. The postwar social composition of cities and suburbs varied dramatically. The central city historically featured a healthy diversity of racial and ethnic groups. The patterns of population migration provide a case in point. For much of the twentieth century, the city served as a gateway for many of the new residents in urban America. During the great wave of immigration, it is estimated that some 25 million Europeans arrived to the United States between 1900 and 1930, and nearly all these immigrants settled in urban neighborhoods of the nation’s largest cities. Similarly, during the pinnacle of the industrialized twentieth century, nearly 5 million American blacks migrated from the rural South to the urbanized North (Harrigan and Vogel 2003). Black residents joined white residents and new immigrants to meet the surging demand for a manufacturing workforce. The city was the engine of the nation’s industrial economy, and a racially and ethnically diverse group of city residents powered the growth of that

suburban frontier / 29

engine through World War II. The city thus became a myriad of people from the reaches of the American Deep South to the shores of Europe. In short, although many groups settled in neighborhood enclaves, urban America was still nonetheless a vibrant place and home for people of many races, ethnicities, religions, and classes. Whereas the rich racial and ethnic diversity of the city was an asset during the first half of the twentieth century, it proved to be tumultuous during subsequent decades. In fact, after World War II, the social diversity of the city began to push white, middle-class residents out of the city and to the suburbs as social unrest grew. Beginning in the 1940s, racial tensions flared in cities across the nation as a world war was fought and won and veterans began to return home in 1944. Such tensions caused a series of riots to break out in cities from the East to West Coast. Two of the nation’s gravest riots, for instance, occurred in Detroit in 1943 and 1967. The Woodward Avenue Riot of 1943 was one of the earliest riots in urban America and drew attention nationally. A fight between blacks and whites over the alleged drowning of a black woman and her baby and the alleged rape of white woman on Detroit’s Belle Isle Park sparked a controversy throughout urban neighborhoods, black and white. Looting and rioting ensued along Woodward Avenue, the city’s principal commercial corridor. Blacks looted white-owned shops, setting fires to storefronts while a mob of 10,000 whites stoned and beat countless blacks. Ultimately, it took the mobilization of federal troops to quell the violence, but only after 34 people died and $2 million of property damage was caused. Unfortunately, the 1943 riot was only the precursor to an even more tragic riot in America’s Motor City. The Twelfth Street Riot of 1967 began when Detroit police raided an illegal bar and arrested all 82 patrons. Residents began to protest the arrests as excessive police brutality, and chaos quickly ensued in the surrounding neighborhoods. Thousands of residents soon joined the protests, and looting, violence, and burning buildings plagued the city for a week. All told, the riots proved devastating for Detroit: 43 residents died; more than 7,000 residents were arrested; more than 400 residents were seriously injured; more than 400 buildings were burned and later demolished; and some $40 million in property damage was reported. The turbulent story of Detroit’s experience echoes the racial violence, crime, and overall social unrest in American cities from the 1940s to the 1960s. Such riots underscore a fierce competition that existed between whites and blacks for both a limited number of jobs and an inadequate housing supply during this period. These riots occurred at the peak of the civil rights movement, the antiwar movement, and the onset of globalization. Teaford (2006, 146) captures the essence of race as a push-pull factor by noting that “many white Americans felt that the answer to the ‘urban crisis’ was

30 / transforming race and class in baltimore

to live, work, and play in the suburbs and abandon the central city to troublesome blacks.” Overall, the impact of the culmination of these social transformations made U.S. cities vulnerable and ripe for change—and suburbs appeared to be the prime benefactor from the turbulent times of the postwar urban experience. The postwar economic structure of cities also paved the way to the suburbs. In the years following World War II, millions of veterans returned home to their urban neighborhoods to find them overcrowded and dilapidated. Land was expensive, and taxes were high in the central city. Virtually no new construction of housing units occurred during the war; therefore, demand surged for new and improved housing opportunities. At the same time, large companies began to suburbanize, and regional shopping malls were built in the suburbs. The market had transformed from the historic, outdoor, downtown retail corridor in the city to the new, enclosed, suburban mall with ample, free parking. As suburbia quickly grew, it offered new residents places to work and play. The economic transformation that ultimately pushed millions to suburbanize was the decline of public transportation and the rise of the automobile. The construction of highways meant that streetcars, trains, trolleys, and buses took a backseat to the automobile (Warner 1978). New transportation technologies, coupled with a demand for new housing and jobs, facilitated rapid suburbanization. Hence, economic decentralization was a contributory force that fueled the growth of suburban America (Stanback 1991). The state of public services in the 1940s and 1950s also pulled residents from the central city and pushed them to the suburbs. The infrastructure in the central city was dismal in the decades following World War II. The basic transportation system in cities suffered from old age and a lack of regular maintenance. Potholes and cracks could be found throughout roads in the city. Furthermore, the sewer and water systems in various cities on the East Coast such as New York, Philadelphia, and Baltimore, were nearly a century old. Water was regularly contaminated and distribution was often inefficient. Electricity and gas networks were also generally disorganized and ineffective. Older cities were showing the long-term impacts of age, and this was evident in the crumbling infrastructure. Richardson Dilworth (2005), in his persuasive book The Urban Origins of Suburban Autonomy, illustrates how the development and lack of maintenance in the urban infrastructure promoted, in part, the suburbanization and fragmentation of future suburban cities. From the late nineteenth century onward, residents who grew dissatisfied with the state of urban public services were able to “break away” from the city and form their own municipality. This enabled new suburban governments to tailor an alternate package of goods and services that would better suit the desires of suburbanites.

suburban frontier / 31

Dilworth shows that in the case of the greater New York area, wealthier white residents of new suburbs could establish lower tax rates and provide higher levels of quality services than the central city counterpart provides. Thus, the public infrastructure that served as the backbone of the city was crippled by old age at the midpoint of the twentieth century. Rather than investing in the existing infrastructure for rehabilitation, many Americans chose to move to the suburbs and construct a new infrastructure for their new lives in this largely undeveloped frontier. Perhaps no other municipal service was as contentious as the public school system. A telling example is the battle for racial integration that culminated during the mid-1950s. The U.S. Supreme Court’s 1954 landmark decision in Brown v. Board of Education of Topeka held that it was unconstitutional to segregate schools, and more broadly public institutions, by race. In the Brown case, a class action lawsuit was brought by 13 black families against the Board of Education of Topeka, Kansas, holding that the separation and segregation of schools by race was unconstitutional under the Equal Protection Clause of the Fourteenth Amendment. Such separation, the plaintiffs argued, had caused decades of socioeconomic inequality. Students enrolled in white schools had long been afforded the greater resources—better teachers, more books, and superior buildings—than students enrolled in black schools. The Court unanimously agreed with them, and on May 17, 1954, the Court declared that “separate educational facilities are inherently unequal.” In doing so, the Court overturned the long-standing doctrine of “separate but equal,” that the 1896 case of Plessy v. Ferguson had established. To remedy the racial imbalance in schools throughout the nation’s metropolitan areas, the Court ordered that white and black students be bused to other neighborhoods, or even other jurisdictions, to achieve integration. This meant that groups of black students would be bused to an otherwise majority white school, and likewise, groups of white students would be bused to an otherwise majority black school. Undoubtedly, this method provoked fierce opposition among blacks and whites. In Baltimore, for instance, opposition was particularly ardent among residents in blue-collar neighborhoods. The fight over busing in Baltimore’s Southeast neighborhood was closely watched. Residents mobilized and organized the Southeast Coalition to prevent busing to and from their neighborhood. Churches, small businesses, local factories, and families rallied to “protect their neighborhood” from blacks. In 1974, they protested what was coined as “federal blackmail” and staged sit-ins in schools, and in one case, more than 2,000 students blocked traffic outside of City Hall during rush hour. Then city council member Barbara Mikulski joined in the fight and argued fervently that busing was an “assault on urban

32 / transforming race and class in baltimore

working-class families” (Durr 2003, 169). Ultimately, the community organizers were successful in their quest to keep their neighborhood a white enclave in the midst of a city with a growing black population. Southeast Baltimore remained insulated, and students were not bused. Protests over integration and busing even occurred among the nation’s highest elected officials. For example, in 1957, Orval Faubus, the governor of Arkansas, ordered the Arkansas National Guard to form a blockade to prohibit the legal entry of black students into Little Rock High School. President Dwight Eisenhower was forced to respond to the state and local opposition in Little Rock. He deployed the 101st Airborne Division from Fort Campbell, Kentucky to maintain peace and to permit the entry of black students. On July 25, 1974, the Supreme Court in Milliken v. Bradley ruled that busing across jurisdictions was unconstitutional. This decision effectively halted the practice of busing nationwide. Undoubtedly, the Court’s declaration set into motion a series of events that forever changed race relations in cities and suburbs. The foremost implication was that the Brown case cemented the decision for millions of Americans to make the move to suburbia. To attract more whites, suburban developers capitalized on the fears that white, middle-class residents held of the black population. Whites feared that if their children were forced to attend an integrated school with blacks then it would be harmful. Whether whites held blatant or latent discriminatory views toward blacks, whites perceived that suburbia provided a safe haven from undesirable residents of the city. A move to the suburbs provided an escape route from the increasingly black city. Suburbia gave new residents the opportunity to establish their own municipality, thus giving the powers of local government to white, middle-class residents. Tools such as zoning and housing codes ensured that neighborhoods and communities would remain white, and therefore ushered a modern era of de facto segregation. In essence, Brown was an impetus for white flight, and it had the unintended consequence of spurring even greater numbers of white, urban residents to flee the city for the suburbs. The last major push-pull factor relates to the contrast in the characteristics of the built environment. Urban renewal efforts in the 1950s and 1960s attempted to give central cities a makeover—a renaissance—with the hope of abetting the exodus of the white, middle-class population. In fact, numerous scholars have shown that urban renewal programs actually had the unintended consequence of pushing more residents out of the city, thus invigorating the process of suburbanization even more (Sugrue 1996; Teaford 1990). In the case of Baltimore, urban renewal had devastating effects on the population of West Baltimore, a historically black and economically deprived community. City planners sought

suburban frontier / 33

to demolish the “eyesore” of public housing projects and to “clean up the neighborhood” using a bricks-and-mortar approach. In practice, urban renewal did neither. Instead, it displaced thousands of residents and left vacant land that ultimately turned fallow. Moreover, waves of violent crime, housing abandonment, and rampant drug use added to the instability of urban communities. This project, like many others nationally, drew the attention of both city dwellers and suburbanites. City residents who had the economic means to suburbanize left behind their urban neighborhoods for safer and cleaner suburban areas although suburbanites were not lured back to the city. By many accounts, urban renewal was a failure, and it simply reinforced the strength of the magnet-like attraction to suburbia. A collection of social, economic, and physical conditions contributed to the forces that both pushed out urbanities from the city and pulled them to the suburbs. Race and class were two of the most important social factors that determined whether residents would suburbanize. The racial fissure in American society during the midpoint of the twentieth century was particularly evident in big cities and new suburbs. Fear of other people of different races, violent riots, and the prospects for the racial integration of public schools only exacerbated the racial divide. Thus, the desire to suburbanize to maintain distance was a quick and immediate action that the white, middle class took during this period. Other societal forces supplemented the social push-pull factors. The economy decentralized and became regional in nature. This gave new suburban residents a place to work and play. In addition, the suburbs offered new development of housing and infrastructure in an era when millions of city dwellers were desperate for such improvements. The collective nature of all these factors made suburbia so popular. The culmination of change in metropolitan residents’ social characteristics, economic structure, desire for public services, and an improved built environment created the perfect recipe for the making of a suburban nation. Suburbia’s New Perspectives The suburbs did not only happen because social forces pushed and pulled residents from the city to the suburb. A new generation of urban scholars has recently challenged the notion of the push-pull model of suburbanization. In fact, an emerging body of literature has begun to illuminate new perspectives on the history of suburbanization. In their book The New Suburban History, Kevin Kruse and Thomas Sugrue offer a provocative collection of essays that challenge the historical framework of the push-pull model that gave way to a white, middle-class suburban society. According

34 / transforming race and class in baltimore

to Kruse and Sugrue (2006, 6), new accounts of suburbia “challenge an older scholarship that looks at the history of suburbs largely internally and, instead, [they] examine the ideological, political, and economic issues that bound the city and suburb together in the postwar world.” Whereas the push-pull model emphasized the suburbs as separate, autonomous units from central cities, new historical perspectives offer new political insight that challenges this model and further complicates suburban history. I briefly take stock of these contributions and situate their arguments to the political history of suburbia. Numerous scholarly accounts have challenged long-standing, historical, social, and economic stereotypes of suburbs. Recently, scholars have attempted to dispel notions of race and class homogeneity in the suburbs. One such example is Andrew Wiese’s work on black suburbanization. To date, the history of the suburbanization of blacks has largely been told by the popular press. In his book Places of Their Own, Andrew Wiese (2004) provides one of the first historical accounts of the suburbanization of the rising black, middle class in the early twentieth century. Wiese documents how early blacks settled near industrial suburbs on the urban fringe, services areas, and informal clusters in the rural South. The characteristics of black suburbs paralleled those of white, working-class suburbs—especially during the streetcar era. Black suburban households, as white ones, owned houses, raised children, and took pride in a yard and garden. Although an emerging black, middle class was evident throughout the twentieth century, Wiese calls attention to the consequences of suburbanization: violence toward blacks, white flight, and discrimination toward blacks. Ultimately, this fueled a suburban environment that became increasingly segregated by race. Yet, Wiese shows that despite these challenges, black households made stunning progress during the last century as many more achieved middle-class status in the suburbs, dispelling any myth that early twentieth century blacks were exclusively rural and urban residents. In a similar vain, Becky Nicolaides (2002) traces the history of the working-class suburb of South Gate, home to big industry in Los Angeles during the twentieth century. During the prewar years, South Gate grew as a suburban enclave for young, white, working-class families. An abundance of industries in the automotive, shipping, and transit sectors meant that well-paying jobs were commonplace. As a result, South Gate grew as an affordable, stable, and close-knit community. Yet this prosperity was short-lived. Postwar America offered many new challenges for the residents. Social unrest in metropolitan Los Angeles began to threaten the security of residents in South Gate. Urban planners sought to integrate the school system, and civil rights protests were routine occurrences by the 1960s. Coupled with industrial plant closings and the threat of low-income

suburban frontier / 35

housing, South Gate became a suburban landscape for antiliberal, workingclass residents. The prospects of social and economic diversity threatened these suburbanites, and residents resisted politically as long as they could. In the end, Nicolaides shows that older suburbs—not only cities—were home to social unrest and economic turmoil. The recent work of several other urban historians—in particular, Kevin Kruse, Matthew Lassiter, and Robert Self—offers yet another fresh perspective on the historical intersections of race, class, and space in metropolitan America. In the wake of the civil rights movement and the desegregation of schools and other public institutions, suburbs were in a state of flux. Kruse, Lassiter, and Self each offer a similar critique of the push-pull model, arguing instead for a more metropolitan-oriented political framework that takes into account racial and spatial relationships in suburbs and central cities. At the core of these arguments is that the backlash thesis, originally put forth by Thomas Sugrue’s work on the urban decline of Detroit and Arnold Hirsch’s (1983) work on urban renewal in Chicago, fueled white suburbanization in the midst of racial unrest. Violent riots and failed urban renewal projects in central cities only served to exacerbate racial tensions, create ghettos, and fuel suburbanization. Continuing in the tradition of Sugrue and Hirsch, Kruse, Lassiter, and Self shed new light onto these perspectives by showing that new forms of spatial politics evolved during this era. First, in Kevin Kruse’s (2005) book White Flight, he demonstrates, through a study of metropolitan Atlanta, that whites fled the central city as a result of the civil rights movement and mounting pressures from the desegregation of public institutions. At the peak of the 1960s, fierce neighborhood battles occurred throughout Atlanta over housing desegregation and the integration of public schools. Some neighborhoods remained insulated and white, while others ultimately integrated. In areas that grew increasingly black, white Atlantans suburbanized to avoid racial integration. To maintain the suburban white stronghold, suburbanites then developed a new conservative politics that stressed individual rights and freedoms. This was typically expressed in the form of lower taxes, privatization, and total abandonment of the central city. Ultimately, Kruse shows that metropolitan Atlanta was spatially transformed into a new region because of a white backlash against black integration. Second, in his book The Silent Majority, Matthew Lassiter (2006) offers a stunningly similar account of racial and political change through suburbanization. Drawing on the example of Charlotte, North Carolina, Lassiter uses the case of court-ordered busing to illustrate how racial and political tensions shaped metropolitan Charlotte. Resistance to the integration of schools led to the large suburban sprawl of southeastern Charlotte. White,

36 / transforming race and class in baltimore

economically mobile residents suburbanized to secede from the threat of an integrated school system. As busing ensued, blacks and working-class whites grew frustrated at the disproportionate burdens that busing placed on them, and so they united politically. Ultimately, a successful coalition of “supporters of integration launched a ‘fairness and stability’ campaign that eventually drew support from a broad cross-section of the metropolis: black parents who demanded a reduction in their children’s transportation burdens, northside and westside white families mobilized for busing equalization, suburban moderates and liberals who endorsed full compliance [of Court-ordered integration], and business leaders who simply wanted the uncertainty to end” (Lassiter 2006, 138). Thus, in other words, Lassiter shows how residents in Charlotte used a new suburban politics to cater to the interests of a broad-based coalition in the city and suburbs, which mitigated the initial white backlash to the racial integration of schools. Third, in his book American Babylon, Robert Self (2003) convincingly demonstrates how a complex political-spatial relationship evolved between the city and surrounding suburbs in metropolitan Oakland. A metropolitanwide political movement evolved as the struggle to save Oakland ensued. Just as Kruse and Lassiter demonstrate, Self also shows that there was a white backlash to the efforts to revitalize Oakland and empower its disenfranchised black communities. According to Self, the backlash was more complicated than white flight from the central city. Whites sought economic security as well as racial isolation; the deindustrialization of Oakland was also perceived as a threat. Consequently, middle-class whites suburbanized to a variety of East Bay suburban communities, capitalizing on cheap land, affordable housing, and easy access to highways and jobs. Self goes on to describe that the white backlash took shape politically in the suburbs, yet it was aimed at blacks in the central city. As social and economic unrest persisted in Oakland despite the election of black mayors and revitalization attempts, suburbanites grew increasingly weary of the liberal state and high taxes. Consequently, the suburbs mobilized politically and passed Proposition 13—the largest limit on property tax in the nation. In the end, Self argues that middle-class whites not only suburbanized to escape the economic and racial insecurities of Oakland, but they also formed a new antiurban politics to avoid the escalating costs of an increasingly troubled central city. Metropolitan forces—demographic, spatial, and political—ultimately characterized the complex backlash in greater Oakland. Overall, a new body of literature emerged at the brink of the twentyfirst century that offered new historical perspectives on the making of the nation’s suburban society. This urban scholarship offered new insight on the historical patterns of suburbanization, drawing into question the

suburban frontier / 37

traditional push-pull model of suburbanization. These studies illuminated the complex historical diversity of suburbia in racial and economic terms. Moreover, they complicate the notion that suburbs were autonomous units in metropolitan America. New perspectives on suburbia ultimately show that political forces, characterized by racial, economic, and spatial terms, capitalized on the metropolitan political system through the development of city-suburban relationships. The phenomenal growth of suburban America, both in the size of the population and in the number of suburban developments, illustrates that a new spatial form of living became the clear preference for millions of residents. The era of mass suburbanization persists today, albeit in a different form. The urban history of this process illustrates that the metropolitan boundary continued to expand for the better part of the past century as the crabgrass frontier matured (Beauregard 2006). The urban fringe became increasingly popular, and the older established areas fell out of favor. This is, in part, a function of the persistence of political support for public policies that promote new fringe development. In general, residents continue to be pushed out of central cities and pulled to the suburbs. A number of factors explain this phenomenon. Although suburbs are undoubtedly more diverse today, many residents continue to live in segregated suburban areas that are afflicted with racial and class divisions. Public policies in the areas of transportation and housing also continue to serve as a ringing endorsement for the uncontrolled growth of suburbs. Competing perspectives on suburbia show that residents in suburbs also develop a spatially oriented politics on a metropolitan scale to maintain the suburban landscape. The political history of suburbanization thus represents a diverse and complex past. The consequences of this history had devastating effects on the lives of urban residents. These policies did not benefit all people equally. By and large, the winners moved to suburbia, and the losers were left behind in the central city. In too many cases, the winning group was comprised of white, middle-class residents while the losing group largely encompassed black, lower-class residents. Therefore, the interaction of private preferences and government policy ultimately shaped today’s suburbs. The Rise of Suburban Baltimore Having considered the broader national framework of U.S. suburbanization, it is now useful to hone the examination to the history of Baltimore’s suburban development. I offer a primer on the rise of Baltimore’s suburban frontier, and then I establish the context of suburban growth and the subsequent decline of Baltimore’s first-tier suburbs. This description

38 / transforming race and class in baltimore

provides a historical account of suburbanization in a now postindustrial, Rustbelt city. Baltimore’s historical development patterns illuminate the process of suburban transformation, and they set the stage for the study of the evolution of first-tier suburbs since 1970. Baltimore’s suburban experience is typical of the Northeastern and Midwestern Snowbelt cities (Arnold 1990). At the end of the nineteenth century and beginning of the early twentieth century, an extensive network of streetcars was one of the most important features that facilitated the suburbanization of Baltimore. In his landmark book Streetcar Suburbs (1978), urban historian Sam Bass Warner demonstrated that the streetcar played a critical role in the decentralization of the urban population. Warner explained that the location of streetcar tracks impacted land speculation and housing markets, which often resulted in a housing boom. The construction of a streetcar station, then, provided new suburban residents with access to the central city. Early suburbanites still largely depended on its neighboring city for economic and cultural activity. Therefore, the streetcar became the mechanism that connected suburbs to cities. It provided life for suburbs during their infancy, and it ultimately was an endorsement for growth and new development. In many cases, the streetcar station became the center of suburban communities—the very suburbs we commonly refer to today as the “old neighborhoods” of first-tier suburbs (Muller and Groves 1979). Consider the history of Baltimore’s streetcar movement as a case in point. An extensive network of tracks moved people around the city and its surrounding areas. At the height of Baltimore’s streetcar era, there were some 2,000 streetcars that traveled on 400 miles of tracks, transporting people from the central city to the immediate surrounding suburban areas (Harwood 2003). These streetcars and their tracks became the pathways that thousands would use to exit the central city during the twentieth century. Herbert Harwood, Jr., a former employee of the B & O Railroad Company, fittingly observed that “Baltimore’s urban transportation system in the first half of this century had diversity and charm almost unequaled elsewhere. Its trolleys lurched through a labyrinth of complex downtown trackwork and along single-track rural jerkwater; they threaded through steel mills and shipyards; they rolled past woods inlets, ponds, amusement park roller coasters, and—of course—endless row houses built from the 1700s to the 1930s” (Harwood 2003, ix). In metropolitan Baltimore, streetcar lines, and later the roadways, radiated in an outward fashion around the city’s core. In the first-tier suburbs to the south of Baltimore City, the No. 6 streetcars provided residents of Brooklyn Park and Curtis Bay an escape from their industrial jobs at the shipyards and heavy manufacturing plants. Residents in these suburbs

suburban frontier / 39

worked at plants such as the Domino Sugar plant of American Sugar Refining Company and Archer Daniels Midland Company’s grain elevator plant—the tallest and fastest in the world when it was built in 1923. Streetcars played an important role in transporting these working-class Baltimoreans between work and home. In the west of Baltimore City, the heavily traveled routes along Edmonson Avenue and Frederick Avenue carried passengers out of the city to various suburban places. A typical route included stops in the row house suburb of Edmondson Village, and then to the suburb of Catonsville. The western routes ended in the eighteenth-century flourmill town of Ellicott City. The streetcar lines of Nos. 9 through 14 were quite popular and well utilized, so much so that Catonsville became known as Baltimore’s streetcar suburb. Early suburban residents of Catonsville were wealthy and depended on this mode of transportation for retail and commercial activities in the city. The streetcar lines passed by Catonsville’s large Victorian and Colonial houses on tree-lined streets to carry passengers to and from downtown Baltimore. Heading in a northwestern direction out of the central city, the routes along North Avenue, Reisterstown Road, Liberty Heights Avenue, and Park Heights Avenue grew during the 1910s and 1920s. Streetcar No. 32, for instance, traveled from Liberty Heights avenue to Reisterstown Road and Gwynn Oak Avenue, arriving in the western first-tier suburbs of Woodlawn and Lochearn. During the streetcar era, these places remained small and sleepy residential suburbs of Baltimore. Perhaps the most well-known streetcar routes were located to the north of Baltimore City. The York Road corridor, which spans north and south from downtown Baltimore to Roland Park, and then to the firsttier suburb of Towson, was one of the earliest streetcar routes. As one of the earliest north-south routes, the streetcar line No. 29 transported residents from downtown Baltimore to the exclusive suburb of Roland Park. This area was an early suburb of Baltimore. It was annexed to the City of Baltimore through the Annexation Act of 1918 of the Maryland General Assembly. A substantial part of the northern areas of the City of Baltimore was suburban in nature, and most of this land was captured in the city’s last annexation in 1918. In 1896, the streetcar line No. 29 led to Roland Park’s shopping district—the nation’s first integrated suburban shopping center. Master planner Frederick Law Olmstead oversaw the development of this suburb at the turn of the twentieth century. A rapid transit streetcar line was critical for the success of his planned suburban community. The growth of suburbia continued during the late 1940s and early 1950s when streetcar No. 8 reached its peak. Traveling north and south along the York Road corridor, this route fueled Baltimoreans’ northern migration to

40 / transforming race and class in baltimore

first-tier suburbs such as Towson, Lutherville, and Timonium. Streetcar No. 8 quickly became the network’s most popular route during metropolitan Baltimore’s streetcar heyday—and this ultimately helped grow the suburb of Towson into one of the largest places in the first-tier suburbs of metropolitan Baltimore. The streetcar network also reached the northeastern first-tier suburbs of Baltimore. The streetcar No. 19 followed Hartford Road from downtown to Parkville. This first-tier suburb, primarily rural in nature during the first half of the twentieth century, housed a residential population. The residents benefited from public transportation to the city’s commercial center. Similarly, the No. 15 streetcar line along the Belair Road corridor provided access to the city for residents of neighboring Overlea. Owing to popular demand, this route survived as one of the last streetcar lines in the network. Just as the region’s other streetcars to west, northwest, north, and northeast of Baltimore City facilitated the mobility of suburban residents, the routes to the east of the central city had the same effect. The No. 26 streetcar line served Dundalk’s suburbanites along the Dundalk Avenue corridor. Until 1950, double streetcars, also known as “twin rockets,” rolled along Dundalk Avenue’s busy residential and commercial corridor. Other industrial, suburban areas also benefited from an extensive streetcar network. Due south of Dundalk, Sparrow’s Point and Turner’s Station were home to manufacturing plants and low-income African American households. Home to the bastion of steel mills, the Bethlehem Steel Company employed the majority of residents of Dundalk, and investments were made in the streetcar network to allow for the easy transport of employees to and from work daily. The No. 26 streetcar carried scores of passengers to the steel plant from Dundalk, crossing mile-wide Bear Creek estuary, to the Sparrow’s Point terminal. Finally, in some areas, streetcar rails were planned but never built. There were three significant first-tier suburban areas that never gained access to the streetcar network. First, the suburban areas due south of the city generally did not receive access to the streetcar network, especially the workingclass residential suburbs of Arbutus and Halethorpe. Second, streetcar rail also never reached the northwestern suburban sector in Pikesville and Reisterstown. Plans to extend the rails along the Park Heights Avenue corridor never came to fruition. Third, due east of the city, the large industrial suburbs of Essex and Middle River remained isolated from the streetcar network. The 10-mile stretch of rail was planned to connect Middle River and Essex to Eastern Avenue in Baltimore City’s Highlandtown neighborhood, but declining ridership and high construction costs prohibited the rail extension.

suburban frontier / 41

A vast and diverse network of rails and streetcars defined metropolitan Baltimore’s streetcar era. The heyday for the streetcar peaked before World War II, and by the onset of the 1950s, signs of an aging infrastructure and decreased ridership were apparent. The increasing popularity of automobiles quickly threatened the sustainability of the region’s streetcar network. In the end, personal cars won and public streetcars lost. Rails lost, and roads won. The very rail pathways that transported streetcars were expanded and later paved with asphalt to accommodate buses and automobiles. Baltimore’s streetcar network died in 1963 when the last streetcar was taken out of service. The charming array of metropolitan Baltimore’s streetcars is now simply just a memory as many streetcars have been preserved at Baltimore’s Streetcar Museum. One of the important features of urban development is that the transportation network linked the first-tier suburbs to the central city. The streetcar network first connected the oldest, first-tier suburbs to Baltimore. And later, the roads and highways that encircle the city provided transportation connections throughout the region. Eventually, the domination of the automobile led to the demise of Baltimore’s streetcar network. The tracks were converted to roads and highways for the automobile suburbs of the 1950s and 1960s. The streetcar shaped the first-tier suburbs of many metropolitan areas, including Baltimore. These suburbs were socially and economically dependent on the City of Baltimore. Early on, the central city provided life for these suburban areas. Suburban residents in the first-tier relied on the city not only for jobs and economic sustainability, but also for shopping and cultural amenities as well. Many of the first-tier suburbs that surround the city boundary were once served by streetcar networks. Yet, by the 1940s, buses and automotives became the preferred mode of transportation, and the streetcar era died quickly as governments paved over streetcar lines for automotive transit. Kenneth Jackson’s (1985) classic work on the history of suburbia also demonstrates the role the transportation policies played in the development of older suburbs. He notes that the construction of roads and highways led to the age of automobility, paved in part by federal highway policies. In the Baltimore region, Baltimore County’s first planner, Malcolm Dill, first envisioned a “Beltway” freeway in 1948 that would fully encircle the central city and the first-tier suburbs (Brooks and Rockel 1979). Construction of Interstate 695 (I-695) began in 1953. When it was completed in 1962, it was incorporated into the Interstate system of highways. In fact, the I-695 became the first beltway in the nation to be completed under Eisenhower’s 1956 Interstate Highway Act, at a cost of nearly $250 million. The Beltway was initially intended to be 6 lanes wide with a 36-mile circumference, but

42 / transforming race and class in baltimore

ultimately became 53 miles after the eastern portion was completed with 4 lanes. This transportation was integral for the development of Baltimore’s first-tier suburbs. Before the highway, each suburban place that bordered Baltimore City was disconnected from other suburbs since no major roadways linked them. The Beltway connected these suburbs, and it helped to stimulate a massive housing and shopping mall boom in the Baltimore region during the 1960s. This transportation network helped to establish, and ultimately foster the decentralization of people and economic activity, which would become a recurring theme throughout the history of twentieth-century urban America. Another important development factor that affected the evolution of Baltimore’s first-tier suburbs was the creation of Baltimore County’s “Urban-Rural Demarcation Line” (URDL). This was an urban growth boundary that Baltimore County adopted in 1967 at the height of the county’s population boom and still continues today. The URDL created urbanized growth areas surrounding the central city, and all the county’s first-tier suburbs were located within this boundary. It also preserved exurban areas as farmland in the northern section of Baltimore County. During the late 1960s and 1970s, the URDL allowed Baltimore’s first-tier suburban communities to grow as places strategically located near the central city. The county used zoning tools to encourage development to stay in close proximity to the central city. The land inside URDL was zoned as high-density residential and commercial uses. In contrast, the land outside of the URDL was zoned low-density residential and agriculture area. For instance, the county generally only allowed one housing unit per 50 acres in the territory outside of the URDL. The county also did not provide public infrastructure such as water and sewer services to areas outside the URDL. As a result, the areas within the URDL became fully developed, and the URDL largely kept development from encroaching into the rural areas of the county. In the regional context, the URDL promoted several unintended consequences. Once the growth areas of the URDL became fully developed, there were no areas left in the county for new development. Consequently, new population growth occurred outside of the Baltimore County. Since other counties in the region did not have a comprehensive growth plan like Baltimore County, these areas welcomed new population growth and new development. Essentially, growth and development of new suburbs leapfrogged across counties in metropolitan Baltimore. This occurred owing to the lack of a region-wide land use plan. Overall, two important historical development factors, among others, characterized the evolution of Baltimore’s first-tier suburbs. First,

suburban frontier / 43

a transportation network of streetcar lines and highways facilitated the growth of suburban areas that bordered the central city. Early streetcar lines, and then a complex highway system, provided the opportunity to transport people, goods, and services easily throughout the region. Second, Baltimore County’s decision to manage population growth led to the adoption of the URDL. This growth boundary allowed the first-tier suburbs to become fully developed by 2000. These two development trends, as well as many others, characterized the historical patterns of development in the Baltimore region. By 2000, the urban landscape in metropolitan Baltimore had grown into a large, diverse metropolitan region. Several key characteristics defined this landscape. The metropolitan population reached two and a half million in 2000. Politically, the region consisted of six large county government units. Economically, the region transitioned from an industrial city–based economy to a suburban-based services economy. Socially, greater Baltimore remained a black and white metropolis. In terms of development, growth occurred in rings, and multiple suburban nodes were established. Dynamics of Neighborhood Change Let us now turn briefly to the theoretical elements of how the suburban frontier evolves. Several strands of scholarly work provide the theoretical foundation for this study. Specifically, this study draws on a body of multidisciplinary literature from the fields of urban studies in geography, political science, and sociology to review socioeconomic change in American suburbs critically. First, I review the major contributions that examine the process of neighborhood change. A significant body of work has focused on explaining how and why neighborhood change occurs in central cities. I extend that work and apply it to the context of suburban neighborhoods. Next, I review the key models of suburban change from the sociological perspective. These models help to explain the process of change in suburban settings, and I elaborate on the implications of these models for first-tier suburbs. Studies on residential differentiation, otherwise known as neighborhood change, is a useful body of work that informs us about the distinguishing characteristics of changing and declining urban and suburban areas. In particular, there has been a strong emphasis on the processes that explain how and why neighborhoods change. In 1925, Robert Park and Ernest Burgess of the Chicago School of urban studies were two of the earliest pioneers to study the dynamics of neighborhood change. The “Chicago School” is commonly used to describe a group of urban scholars at the University of Chicago who studied the socioeconomic structures and

44 / transforming race and class in baltimore

geographic patterns of the residents of Chicago. Drawing on the principles of the discipline of ecology, Park and Burgess developed an approach that has come to be known as “human ecology.” Using this perspective, scholars have examined people and their behavior on the basis of their spatial organization on the land. The ecological perspective views humans as part of a large natural system, competing for resources. Park and Burgess referred to this as a “balance of nature,” a condition in which social and economic changes constantly occur between humans and the landscape. In essence, urbanized populations continually move between various neighborhoods and communities to improve socioeconomic status. During the twentieth century, many urbanists, particularly those in the Chicago School, focused on developing land use models based on the ecological perspective. Scholars subsequently used these models to develop a method for studying the distribution of population and the socioeconomics of place in urbanized areas. Using the concept of zones, Ernest Burgess (1925) developed one of the first urban land use models for the U.S. city. Using Chicago as a research laboratory, Burgess and his colleagues identified a pattern of “concentric zones” for which urban growth and form could be modeled. Figure 2.1 depicts the concentric zone model, which is composed of five distinct zones that encircle the core of the central city. The first zone, known the as “the Loop” in the case of Chicago, is the central business district (CBD) and the center for

1. Central Business District 2. Transitional Zone Recent Immigrant Groups - Deteriorated Housing - Factories - Abandoned Buildings 3. Working Class Zone - Single Family Tenements 4. Residential Zone - Single Family Homes - Yards and Garages 5. Commuter Zone - Outer Suburban Areas

Figure 2.1

The Burgess model.

Source: Adapted from Burgess (1925).

suburban frontier / 45

commercial activity. In many ways, Burgess’ Loop represents the power of downtown agglomeration economies and the economic centralization of business services. The second zone immediately encircles the CBD as a place of social transition. It houses the slums, the ghetto, and first generation immigrants. Burgess witnessed the development of places such as Chinatown and Little Sicily in this transitional zone. The third zone is coined “working-class zone,” which houses blue-collar workers and shops, and most residents in this zone are second generation Americans. The fourth zone is entirely a residential area composed of single-family dwellings and yards. Fifth, the outermost zone is known as a suburban commuter area. Burgess noted that this zone was exclusively comprised of the social elite and wealthy. The Burgess concentric ring model principally reflects the socioeconomic spatial arrangement of an American industrial city in the twentieth century. Although other important land use models were subsequently developed (see Harris and Ullman 1945; Hoyt 1939), an understanding of the Burgess model is critical for application of the concept of “invasion and succession” to urbanized areas. The Chicago School adopted this term to refer to the process of neighborhood change whereby one social group moves into, or “invades” the territory (residential area) of another group, thus displacing and succeeding the previous group, hence “succession.” Burgess and his colleagues observed this phenomenon in Chicago as newly arrived immigrants settled in central cities and displaced working-class residents in neighborhoods around the city’s core. They argued that residential movement occurred between the concentric zones. Displaced residents moved outward, which created a strong force for the decentralization of urban populations (von Hoffman and Felkner 2002). The process of invasion and succession can also be thought of as the process of suburbanization when urban populations decentralize in a spatial fashion from the urban core (Bourne 1981). Neighborhoods change when high socioeconomic status residents leave one zone for an adjacent zone as lower socioeconomic status residents enter. In turn, this causes more high socioeconomic status residents to move out to zones beyond the city, often resulting in a self-reinforcing cycle of change. Displaced residents move outward in a series of stages. Abrahamson’s (1996) work on this process provides us with a practical guide to understanding the stages of invasion and succession. Specifically, he notes that neighborhoods move through six stages of invasion and succession. They include the following: 1. Pre-invasion. This is the initial stage of succession. Neighborhoods are generally stable and intact. The population is both racially and economically homogenous.

46 / transforming race and class in baltimore

2. Invasion. During this stage, a new population begins to enter the neighborhood in very small numbers. Downs (1982) refers to this population as “pioneers.” The new population settles in a residential area with different socioeconomic characteristics from the majority of residents. 3. Early succession. New residents continue to settle into a neighborhood. The invasion constitutes approximately 20 percent of residents. The neighborhood still remains largely homogeneous. 4. Middle succession. During this stage, neighborhoods are in a state of flux, and rapid invasion and succession occurs. The invasion of new residents accounts for approximately 20–60 percent of residents. This leads to the creation of a new majority population. 5. Late succession. During this stage, the population becomes increasingly more homogenous again. But, the new population is entirely different from the previous population in the socioeconomic context. The invasion population is approximately 60–80 percent of the neighborhood’s overall population. 6. Consolidation. In this last stage, the population is once again homogenous, but the socioeconomic fabric of the neighborhood significantly alters. The new population that succeeded the neighborhood is typically of a lower economic class and is composed of a minority racial or ethnic group. In short, the neighborhood has transformed into a new place. Once the process of neighborhood invasion and succession begins, it often creates a chain reaction or ripple effect. Neighborhoods can progress through these six stages in a short period. In the end, neighborhoods in close proximity to other “invaded” neighborhoods become more vulnerable, and thus they are more likely to experience suburban change. There are two exceptions to the invasion and succession process. First, “elite enclaves” can develop in neighborhoods (Abrahamson 1996). They are very stable neighborhoods that preserve a high degree of population and economic homogeneity. In other words, they are places that remain impermeable to neighborhood change for a long period. Second, another exception is the process of gentrification. This process is essentially the opposite of succession. The population in these neighborhoods moves from a low to high socioeconomic status (Wagner 1995). Neighborhoods that experience an influx of socioeconomic elites displace lower socioeconomic status residents. In addition, these areas have clear, delineated boundaries that typically protect them from the external processes of other neighborhoods (Wagner 2001). In general, this trend is not the norm in the suburban environment, and these neighborhoods are exceptions to the succession and invasion process.

suburban frontier / 47

The filtering of housing is a process embedded in invasion and succession. Although there is an abundance of scholarly descriptions of filtering, Ahlbrandt and Brophy (1975) provide a succinct definition that still has salience and meaning today. They hold that “filtering is the term used to describe the process through which existing housing gradually declines in value, thereby making it available to groups of lower socioeconomic rank” (Ahlbrandt and Brophy 1975, 9). During the process of invasion and succession, many changes occur on a neighborhood scale that impact the filtering of housing. In their classic work on neighborhood change, Grigsby, Baratz, Galster, and Maclennan (1987) offer some important observations on the process. They note that the following neighborhood changes spur filtering; occupancy; income of occupants; value or rent of dwelling units; price per unit of housing services; or quality and quantity of services provided by housing units or neighborhood environs. These changes in the characteristics of housing units and the population create an environment that is conducive for the filtering of housing units from higher-status economic groups to lower-status economic groups. Grigsby et al. (1987) speculate six causes of neighborhood succession and filtering, which include: 1. Changes in real income of households can lead to significant changes in the physical infrastructure and population in a neighborhood. Rising incomes have two consequences. They allow residents to invest in housing improvements, or they allow residents to move to better housing in another neighborhood. In contrast, declining incomes have the opposite effect. The lack of financial resources causes residents not to invest in housing improvements, which facilitates the deterioration of neighborhoods. This scenario has a compiling effect. When housing units decay, socially mobile residents will be more likely to move to another neighborhood in search of better socioeconomic conditions. 2. Growth in the number of households can stimulate the market to provide a larger supply of housing for the neighborhood. In this scenario, housing vacancies decrease as the demand for new housing increases. The increased demand for housing then creates a situation for rents and housing prices to increase, since the supply of housing in any given neighborhood is limited owing to the physical constraints of space. 3. Decrease in the number of households can stimulate the invasion process. If a neighborhood loses residents, upper-income groups are likely to leave. As they leave, homeowners and landlords are forced to lower the price of their housing units. As a result, in-migrants are able to purchase and rent cheaper housing units. The invasion of

48 / transforming race and class in baltimore

new, lower-income households further provokes the flight of upperincome households, and vacancies of housing units increase. 4. Obsolescence occurs when a neighborhood’s housing stock and environs change for the worse—for instance, “structures fall into disrepair or trash fills the streets” (Grigsby et al., 1987, 38). There are numerous forms of obsolescence, including style, space and layout, structural, site, or location. Housing styles may go out of fashion, housing units age and may lack regular upkeep, and access from the neighborhood to other amenities may become limited over time. As a neighborhood becomes obsolete, highincome households will leave and low-income households will replace them. 5. Change in housing demand and supply resulting from public intervention can impact the rate and degree of neighborhood succession. Federal government policies have had a major impact on the demand for and supply of housing. Grigsby et al. (1987) argue that federal tax policy (namely the mortgage deduction tax incentive) and cheap real estate have fueled the persistence of suburbanization. The consequence is that neighborhood growth occurs in zones further away from the urban core, which in turn drives up the demand for housing in the suburbs. 6. Neighborhood deterioration is the “absolute negative change” of the social and physical qualities of a neighborhood. Grigsby et al. (1987, 41) go on to provide a concise view of this negative change: Physical deterioration could result from inadequate housing maintenance, intrusion of blighting commercial or industrial uses, increased traffic, disappearance of aesthetically pleasing open space in favor of trash or rubble-strewn empty lots, or building abandonment. Deterioration of a neighborhood’s social environment might be manifested in such changes as a growing proportion of “problem” families, a rising crime rate, or a drop in the quality of neighborhood schools.

Grigsby et al.’s (1987) six causal factors that lead to neighborhood succession are important to bear in mind during the analysis of how and why socioeconomic change occurs in first-tier suburbs. The process of neighborhood change can easily be applied to the study of suburban decline. This is a useful vantage point for understanding the flow and movement of population between city and suburb, and suburb to suburb. From this literature, a broad definition for suburban decline can be developed. Drawing on the factors of neighborhood succession and decline in central cities, we can interpret the decline of suburbs in a similar fashion. Like urban decay, we can infer from the literature that suburban decline

suburban frontier / 49

can be generally understood as the weakening of the socioeconomic status of neighborhoods and the deterioration of the physical infrastructure in the built environment. This literature also suggests that neighborhood change is a complex process involving many factors, internal and external to neighborhoods. Factors internal to the neighborhood include the composition of race and ethnicity, socioeconomic status, and occupation. Factors external to the neighborhood include location and proximity to the central city and other suburbs, transportation networks, and access to the region’s economic amenities. These factors impact neighborhoods differently, and analyses of neighborhood change should appropriately account for them. Overall, Short (1978, 420) holds that an understanding of the invasion and succession process is fundamental to the study of neighborhood change; he summarizes that “[it is] designated an important research topic for the understanding of urban residential patterns.” In other words, it is essential to understand the process of how residents move so that we can differentiate suburbs and analyze why such patterns exist. By applying this understanding of the invasion and succession process to the study of how suburbs change, it is now useful to explore various models of suburban change. Early Models of Suburban Change In response to suburban growth during the 1960s and 1970s, sociologists applied the study of invasion and succession to suburbs. A body of literature emerged to explain if, how, and why suburbs change. Logan and Schneider (1981) and Stahura (1984) categorize this literature into three themes based on three competing theories of suburban change: the persistence model, the ecological model, and the stratification model. Each of these models is critically reviewed in turn. Persistence Model Farley (1964) was one of the first scholars to study the process of suburban change. As early as the 1950s, scholars began to document the changing nature of suburban populations. Wealthy suburbs, industrial suburbs, and black suburbs, just to name a few, were some of the diverse types of suburbs that scholars identified (Masotti and Hadden 1973; Schwartz 1976). Yet, few researchers inquired about the process driving these social forces related to diversity. Farley asked whether suburbs were really different from one another, and if so, why. More formally, Farley (1964, 39) asked, “[did] rapid population growth alter the socioeconomic characteristics of a suburb or [did] suburban places retain their peculiar characteristics even after great population increases?”

50 / transforming race and class in baltimore

On the basis of a study of 137 suburbs of 24 cities, which contained a sample of 54 percent of the nation’s urbanized population in 1960, Farley used population growth and levels of educational attainment to determine whether suburbs retained their initial status from 1920 to 1960. He found overwhelming evidence that suburbs retained their original socioeconomic characteristics over four decades, despite dramatic growth. In other words, the characteristics of suburbs “persisted” over time, and suburbs did not change substantially. He coined the term “suburban persistence” to refer to this phenomenon, which was defined as the ability of suburban areas to maintain their original socioeconomic status over a long period. Ultimately, Farley (1964, 39) concluded that “[A]s individual entities, suburbs demonstrate a stability of characteristics relatively little affected by population growth. This suggests that the characteristics of a suburb may be fixed relatively early in that suburb’s history and subsequent growth reinforces the existing residential patterns.” In a similar study, Guest (1978) replicated Farley’s findings of strong suburban persistence. In a national sample of 661 incorporated suburbs, Guest analyzed changes in socioeconomic status from 1920 to 1970. Like Farley, Guest used education to measure changes in suburban socioeconomic status, though unlike Farley, this study controlled for the age of a suburb and location to the central city’s core. He found evidence to support Farley’s theory of suburban persistence. Guest (1978, 292) concluded that “most suburbs retained a very high stability of relative ranking in social status.” In other words, similar to Farley’s study, there was little socioeconomic change in the status of suburbs over a period of 50 years. These two studies were significant because the findings challenged the long-standing urban ecological theories of the Chicago School. Namely, the concentric zone theory, originally purported by Burgess (1925), held that there were constant cycles of growth and decline in urbanized areas as higher socioeconomic groups moved outward toward the periphery, leaving the area to lower socioeconomic groups to take over. Essentially, highstatus neighborhoods would decline as decentralization of the population occurred. Farley and Guest held that the invasion and succession process did not apply to suburban areas. Naturally, this sparked the attention of urban scholars, and subsequent studies were conducted to assess the relevance of the Chicago School (Logan and Molotch 1987). Ecological Model The ecological model of suburban change grew in response to the Farley and Guest’s assertion that suburbs did not experience socioeconomic change. That viewpoint stood in stark contrast with the geographical theories of urban change that dated back to the 1920s of the Chicago School. It also

suburban frontier / 51

diverged from the neighborhood life cycle theories of succession and invasion. As Hoover and Vernon (1962), and later Grigsby et al. (1987) showed, neighborhoods deteriorate over time as the population and housing ages, which results in the invasion of an increasingly poorer population. One group of scholars sought to debunk the persistence model with a landmark study in 1980. Choldin, Hanson, and Bohrer (1980, 972) challenged the suburban persistence model. They argued that “the findings [of the persistence models] are inconsistent with previous theories of urban development, which state that residential areas tend to change and decline over time.” The researchers studied the 157 suburban communities of Chicago from 1940 to 1970 and found evidence that suburbs followed the neighborhood life cycle, just as cities did. In an indictment of the persistence studies, Choldin, Hanson, and Bohrer (1980, 973) held that “[Farley and Guest] employed deficient research designs” because they used inappropriate measures of socioeconomic status. Instead of using education as a proxy for socioeconomic status, Choldin, Hanson, and Bohrer used household income. Their study demonstrated that the change from high- to low-status in the suburbs over three decades was a function of age and declining incomes. So, as Chicago’s suburbs aged, they got poorer and evolved into different places. In the end, Choldin, Hanson, and Bohrer (1980, 973) concluded, The question of suburban stability versus change and decline has ramifications in urban sociological theory as well as public policy . . . In the policy realm, stability implies that the suburban belt will not suffer the long-run problem attendant upon status decline. Presumably central cities could continue to deteriorate and, perhaps, to become slums, while suburban persistence would guarantee a rosy future outside the city . . . yet, [in this study] we found that suburbs do change in status over time, typically moving gradually downward. (Emphasis added.)

Stratification Model In response to both the persistence and ecological models of suburban change, Logan and Schneider (1981) used a political economy perspective to lend insight onto the early debate about how suburbs change. They noted that “political theories of suburbanization interpret the metropolis as a stratified system of places, competing for the benefits (ranging from social status to high land values to public service and tax advantages) which accrue to residents of more successful communities” (176). In other words, high-status suburbs benefit from the ability to use political tools such as zoning and taxation to keep unwanted residents out. Logan and Schneider

52 / transforming race and class in baltimore

thought that this would cause poor suburbs to further deteriorate and lead to increased stratification over time. Ultimately, they reasoned that stratification, which they equated with socioeconomic inequality, would cause suburbs to change over time. To test their stratification model of suburban change, Logan and Schneider (1981) collected socioeconomic census data for 52 metropolitan areas from 1960 to 1970. Similar to the ecological studies, they used median household income as a proxy for socioeconomic status. Their statistical analysis demonstrated that income inequality and racial inequality grew during the 1960s. Inequality was greater in politically fragmented regions, and inequality tended to decline in suburbs with a fast-growing population. Logan and Schneider concluded that this provided evidence that suburbs do change, but unlike the ecological model that held that suburbs experience change owing to life cycles, they linked the suburban socioeconomic changes to the structure of local government and to suburban employment opportunities. Overall, scholars developed three main models to explain suburban change: the suburban persistence model, the suburban ecological model, and the suburban stratification model. An understanding of these models is useful because it provides an understanding of how early scholars theorized about suburban change. These models also provide valuable information about the analysis of socioeconomic change. For instance, we can glean from these studies what socioeconomic data are important for gaining a firmer understanding of change. In addition, the application of geographic and political theories of place can guide future analyses of suburban change. Yet, this body of work is limited in a number of ways. In general, these studies do not account for more recent social forces, such as the suburbanization of minorities and immigrants. Such a dynamic change in the composition of the population may yield new findings. Another limitation relates to the data collection methods. Most of the early studies collected data only on suburban municipalities. These studies assumed that most suburbs were incorporated. They ignored suburbs in a metropolitan area without municipalities, like Baltimore. Last, these studies are limited because, by and large, they do not account for the development of newer spatial relationships between cities and suburbs, and suburbs to other suburbs (Beauregard 1989; Logan and Schneider 1981). These models primarily relied on older census data from 1950, 1960, and 1970, and census geographic definitions for suburbs during that era were limited. In plain terms, there is a need to analyze suburban changes with newer data since today’s suburban landscape is dramatically different from previous decades.

suburban frontier / 53

Summary It is important to reflect on the future prospects for suburbia since there is increasingly more evidence that “trouble in paradise” exists (Baldassare 1986, 1). Urban historian Robert Fishman (2000) conducted a survey of scholars and found that the decline of older, inner-ring suburbs emerged as one of the most important new challenges facing metropolitan America in the new century. These inner-ring, or first-tier, suburbs refer to the developments that grew up in the first suburban generation of the 1950s—the Levittowns of yesterday. A half century later, these suburbs are mature and show signs of social and economic decay. Indeed, the aging process has decentralized from city to first-tier suburb, and the very challenges that the urban core faced three decades ago are now omnipresent in these suburbs today. Recently, urban scholars have recognized the importance of the study of suburban America. One prominent scholar noted that “[T]he field of urban politics demands an anthology that is current and covers major topics of research such as . . . suburban policy” (Pelissero 2003, 3). The analysis of suburban change is one example of an important contribution to suburban policy. Indeed, it is time to supplement our knowledge with additional research that is explicitly focused on redefining America’s first-tier suburbs and that considers the policy implications of socioeconomic suburban change. The process of suburban transformation offers scholars of suburbs a ripe opportunity to push the boundaries of the current scholarship on suburbs. Since suburbs now exhibit symptoms of a life cycle, social scientists have a chance to examine the legacies of these places and reflect on the factors that gave way to the process of decline. Historians McManus and Ethington (2007, 318) echo this sentiment, and go ever further in their critique of studies: “It is not at all clear, however, that the current wave of revisionism in suburban studies has transcended the limitations of the dichotomies it seeks to refute . . . With very few exceptions, the field of suburban studies has ignored the question of what happens to a suburban seedbed after it has been planted: after it ceases to occupy the leading edge of a metropolis, once it no longer stands as the historically typical suburban form” (emphasis added). The history of our cities provides insight into the destiny of ailing first-tier suburbs in metropolitan America. Whether they become the cities of tomorrow remains an important question.

This page intentionally left blank

Ch a p t e r Th r e e Su bu r ba n D ec li n e

In a recent article, the New York Times declared that suburbs are “awaiting a ‘senior tsunami’ as the first baby boomers turn 65” (Roberts 2007). The article went on to highlight that “America’s suburbs, historically a haven for young families with children, are aging more rapidly than the nation’s central cities as the first suburban generation grows older.” The article is interesting as it takes note of an important phenomenon in suburbia today—the social and economic change of its population. The nation’s oldest suburbs have matured and reached their midlife point. Some have even reached a crisis stage while others are struggling with symptoms of old age. Just like many others nationally, the rising tide of suburban decline has drifted to Baltimore’s suburbs. Yet, there were still some bright spots in the suburbs. In Money Magazine’s “Annual Best Places to Live,” the editors ranked Catonsville, Maryland as the 49th best place in the nation to live in 2007 on its “Top 100 Great American Towns” list. With approximately 40,000 residents, Catonsville’s housing affordability, ease of living, leisure and arts, and education each scored an “A”; jobs, economy, and diversity scored a “B”; and safety scored a “C” on the magazine’s grading index of places. This was a remarkable distinction since Catonsville was a declining first-tier suburb in metropolitan Baltimore. Despite the socioeconomic decline Catonsville experienced since 1970, the suburb had weathered its downfall relatively well compared to many of its neighboring first-tier suburban communities. A bit older, Catonsville still remained a desirable place of residence after many years of transition although other Baltimore suburbs did not. These popular accounts of suburbs paint two very different portraits of first-tier suburbs today. In 1970, first-tier suburbs began to experience dramatic social and economic transitions as urban growth began to decentralize even further outward in the metropolitan area. The development and expansion of the outer suburbs fueled new housing, residential, and employment opportunities. The consequence was that first-tier suburbs, as

56 / transforming race and class in baltimore

fully built places, could no longer grow, began to age, and could no longer attract new population. Thus, the process of neighborhood transition began to take hold throughout these suburban communities. The population, income, housing, and labor force were the primary agents of change. Suburbs changed at various rates, but they were each impacted differently. Nonetheless, suburban decline affected every first-tier suburb in metropolitan Baltimore. Only through a systematic analysis of geographic census data can we determine the extent and proliferation of that decline. The Post-Suburban Era: The Study of Suburban Decline To study the decline of suburbs is to analyze the socioeconomic changes that occurred in these suburbs relative to other suburbs (both first-tier suburbs and outer suburbs), a central city, and a region—the geographic units that comprise metropolitan America. In this study, I analyze how the suburban landscape in metropolitan Baltimore evolved from 1970 to 2000, with particular emphasis on how the first-tier suburbs transformed relative to other parts of the region. In particular, this chapter focuses on four main areas of suburban change, including population characteristics, income dynamics, nature of the housing stock, and labor force structure. In each of these areas, I present the results of what I term a descriptive spatial statistical analysis. This technique focuses on how suburban changes occurred at different geographic scales and compares them relative to one another. Before discussing the results presented in this chapter, it is first important to make some notes about the descriptive spatial statistical analysis. In this chapter, I utilize census place level data to conduct a comparative analysis that allows me to analyze the suburban changes in the first-tier. The analysis of change is only meaningful when the first-tier suburban changes are compared to other geographic scales. Therefore, I stratify the data in this chapter into four geographic scales: first-tier suburbs, outer suburbs, central city, and region. This analytical approach provides an opportunity to examine patterns of suburban change relative to four other geographic scales in metropolitan Baltimore. During the period of the 1970s and the 1980s, there was a lull in suburban scholarship. In fact, it was not until the 1990s that urban scholars began to revisit the state of the suburbs by focusing on suburban change. During the 1990s, urban scholars took a fresh look at the state of suburbia. Much had changed since the initial studies on the first wave of mass suburbanization after World War II. Suburbs had become more complex places—home to a diverse population, new politics, and a dynamic economy (Gainsborough

suburban decline / 57

2001a; Oliver 2001; Stanback 1991). By 1990, suburbs were the dominant place of residence in metropolitan America (Thomas 1998). As more people populated the suburbs, they grew more diverse and complex. Scholars began to revisit the study of suburban change by examining patterns of socioeconomic decline in first-tier suburbs. Before presenting the patterns of suburban decline in metropolitan Baltimore, it is first important to take stock of a small, yet emerging, body of literature that analyzes more recent socioeconomic changes in first-tier suburbs. The analysis of change in first-tier suburbs can be classified into three groups of studies based on their methodological approaches. The first group of studies is based on quantitative assessments of suburban change that primarily focus on measuring the extent of socioeconomic decline in first-tier suburbs nationally. The second group of studies is based on qualitative assessments of suburban change that provide valuable accounts of how various suburbs have changed, and they reflect on the implications for public policy and planning. The third group of studies is based on defining first-tier suburbs that focus on conceptualizing a definition of first-tier suburbs. Quantitative Assessments Two urban planners at the University of Virginia, William Lucy and David Phillips, have become well known nationally for their cogent analysis of suburban decline. In a recent study, they used recent socioeconomic census data to examine recent suburban changes. They were among the first scholars to find that suburbs were declining. In a study of socioeconomic change based on a sample of 554 suburbs in 24 states from 1960 to 1990, they analyzed census place level data to identify trends on suburban decline. They defined suburban decline as income decline, focusing on the median family income of a suburb relative to median family income of the metro area. Lucy and Phillips (2000) argued that a suburb with a declining income was a suburb experiencing overall socioeconomic decline. Lucy and Phillips (2000) found that approximately one-third of the first-tier suburbs experienced greater decline than the other suburban places in their sample. A number of other findings were significant. Nearly every declining suburb lost population, and none experienced growth. They found that income was a polarizing force in the suburbs between 1960 and 1990. For example, the ratio of high-income households to low-income households grew each decade, and by 1990, the ratio of highest-to-lowest income suburbs was 3:4. In other words, the wealthiest suburb had 3.4 times the income of the poorest suburb in their study. The characteristics of suburbs in decline varied, yet they shared one similar

58 / transforming race and class in baltimore

feature: during the 1990s, on any measure of decline, older suburbs as a whole fared worse than their central city counterpart did. Specifically, Lucy and Phillips identified five distinguishing features of “declining suburbs,” including 1. Population loss. Declining suburbs experienced substantial population loss. The greater the population loss, the more socioeconomic decline that a suburb experienced. The stagnation of population growth—in other words, no growth—was also a factor that contributed to the decline of suburbs. 2. Income decline. Any suburb that experienced a loss of family income in constant dollars from 1960 to 1990 also experienced an overall general socioeconomic decline. Approximately three-quarters of the suburbs in the sample experienced some real income loss, relative to the other suburbs in each respective metropolitan area in the study. 3. Racial and ethnic change. Since 1960, Lucy and Phillips argued that the presence of racial and ethnic minorities increased substantially. Suburban decline occurred simultaneously as the racial and ethnic minority populations expanded into the suburbs, particularly black and Latino populations. Significant racial and ethnic changes accompanied declines in real incomes in some suburban areas from the 1960s to the 1990s. 4. Value of the housing stock. The median value of a housing unit was an important predictor of suburban decline. It was closely associated with median family income. Since 1960, the value of the housing stock declined with each subsequent decade. 5. Age of the housing stock. The suburbs that experienced rapid decline had the largest portion of housing units that were at least 30 years old. Socioeconomic decline was greatest in suburbs with housing built in the 1940s and the 1950s. Very old housing, pre-1939, was not clearly associated with socioeconomic decline. On the basis of the factors of suburban decline, they concluded that the nation had entered the “post-suburban era.” Lucy and Phillips argued that the era of suburban dominance had ended. Suburbs had become complex places with complex problems. Central cities were no longer, as a whole, the only areas to experience socioeconomic decline. Older suburbs outside of central cities were now a casualty of urban decay. Lucy and Phillips concluded that only metropolitan-wide planning efforts would avert the future suburban socioeconomic decline. The recent work of Myron Orfield (2002) has also influenced the scholarly thought about suburban decline. In a similar fashion to Lucy and

suburban decline / 59

Phillips, Orfield sought to measure the amount of socioeconomic decline in suburbs. Using descriptive statistics, principal components analysis, and a cluster analysis technique, Orfield analyzed census place level data to study socioeconomic changes in American suburbs during the 1990s. He analyzed cities and suburbs of the largest 25 metropolitan areas. The sample of 4,741 places included 4,606 incorporated municipalities (including central cities) and 135 unincorporated areas. Orfield collected data on a wide range of socioeconomic variables, including population distribution, racial composition, municipality tax capacity, and elementary school free lunch eligibility. Orfield developed a classification of suburbs, which he referred to as the “new suburban typology.” On the basis of these analyses, the typology classified places in the sample into four major categories based on the type of place. The first type was classified as “central cities,” and they constituted just 1 percent of the sample. The second type was classified as “bedroom developing suburbs,” and they constituted the largest portion of the sample at 45 percent. Orfield noted that these places experienced rapid population growth and had high public expenditures for infrastructure such as roads and schools. The third type was classified as “affluent job centers” suburbs, and they represented 15 percent of the places in the sample. Large commercial and retail centers typified these places. The fourth type was classified as “at-risk communities,” accounting for 40 percent of the sample of suburban places. This group of suburbs had an aging and diversifying population with increasing social needs as well as population decline or stagnation. During the 1990s, out of all urban and suburban places in this study, at-risk communities witnessed the largest socioeconomic decline. More specifically, Orfield distinguished three types of at-risk communities. First, he noted that segregated communities were terrible places to live in America, and they constituted about one-fifth of at-risk communities. During the 1990s, these places witnessed dramatic increases in poverty and minority populations, while simultaneously experiencing drops in tax capacity (the ability of a local government to raise revenue through taxation). Second, older communities were largely nondescript suburbs in the northeast with few cultural and shopping amenities and an aging housing stock. Third, low-density places saw slight increases in poverty and minority populations during the 1990s. They also had low tax capacities and little-to-no population growth. On the basis of these interpretations of the results, Orfield (2002, 35) concluded that “compelling evidence challenges the notion of a monolith known as the suburbs . . . [and] once poverty and social instability permeate communities just outside the central city, decline accelerates and intensifies.”

60 / transforming race and class in baltimore

This work is significant for several reasons. It is one of the first attempts to incorporate new evidence from Census 2000 to define the social and economic condition of American suburbs. These findings demonstrate the sheer heterogeneity of suburban places. Orfield’s research suggests that nearly 40 percent of suburbs in the nation’s most populated metropolitan areas are now suffering from and showing signs of urban decay. Particularly noteworthy is the breakdown of what Orfield calls at-risk suburban communities. These are the very suburbs that once embodied the idealized suburb—what Fishman (1987) coins “bourgeois utopias.” This evidence clearly shatters the notion of any utopia. The social ills of the central city have spread out and decentralized from the core, creating a very heterogeneous metropolis. Similar to Orfield, Mikelbank (2004) used census place level data to better define the various types of suburban diversity. He collected data on race, class, and function for 3,567 suburban places in a geographically representative sample of the nation. For each suburban place, Mikelbank accounted for four dimensions of suburban development: population, place, economy, and government type. Using a cluster analysis technique, Mikelbank developed a hierarchical classification of 10 distinct types of suburbs. The entire sample was separated into two major types of suburbs: “Middle America suburbs” and “healthy suburbs.” First, Middle America suburbs constituted just less than two-thirds of all suburbs in the sample. Mikelbank distinguished two major subcategories of Middle America suburbs. “White bedrooms,” which constituted approximately half of those Middle America suburbs, are small suburbs in terms of population and employment. Three types of suburbs further differentiated white bedroom suburbs. “Seasonal wealth suburbs” had high incomes and high vacancy rates during winter; “traditional suburbs” had a large population of families with children; and “small retail suburbs” had small employment sectors based on retail. Mikelbank called the other half of Middle America suburbs “manufacturing suburbs.” They included “black suburbs,” based on a large black population in the manufacturing sector. Manufacturing suburbs also consisted of “struggling suburbs,” based on low-household income and high poverty rates. Second, healthy suburbs accounted for one-third of the overall sample. Mikelbank classified these places as either “suburban success” or “working diversity.” Specifically, high-household income and high rates of bachelor’s degrees characterized suburban success places. Three distinctions could be made about these places. “Prosperity” suburbs had high-household incomes and high housing values. “Working stability” suburbs had large populations and a variety of employment centers. “Aging” suburbs had a large elderly population and an old housing stock. Mikelbank defined the other category as working diversity. These places, on the basis of their

suburban decline / 61

location in the country, had high employment rates and large pockets of immigrants and ethnic minorities. Mikelbank’s study is noteworthy because it further defines how suburbs relate to one another, and to central cities. His detailed typology illustrates that suburbs are a mosaic of urbanized places undergoing rapid socioeconomic change. His work shows that the cluster analysis technique allows scholars to differentiate between many types of suburbs. In particular, this research reinforces the notion that comparative analysis is useful between not only city and suburb but also among suburban places. Building on the previous work of Lucy and Phillips, Orfield, and Mikelbank, Puentes and Warren (2006) conducted the most recent and comprehensive quantitative analysis of socioeconomic change in first-tier suburbs. They analyzed 64 urban counties that housed the majority of first-tier suburbs in nation. The county level geographic scale allowed them to collect and analyze socioeconomic census data over a span of 50 years, from 1950 to 2000. Comparing the first-tier suburbs to primary cities, newer suburbs, and the nation allowed these scholars to provide four benchmarks of suburban change. First, the most striking finding was that one-fifth of Americans lived in first-tier suburbs in 2000. This meant that the first-tier suburbs housed a significant number of residents, nearly 50 million. Second, over 50 years, the first-tier became home to a very diverse population. Specifically, large enclaves of immigrants and elderly populations developed in the first-tier suburbs. Third, income declined and poverty rose during the 1990s in all the first-tier suburban counties. Fourth, the infrastructure of first-tier suburbs grew outdated as housing aged and commercial retail strips grew obsolete. Based on these findings, Puentes and Warren (2006, 1) concluded that first-tier suburbs were “neither fully urban nor completely suburban,” and thus a specialized policy agenda was necessary to combat socioeconomic decline of these areas. It is important to identify two limitations of the Puentes and Warren study. First, the use of county level geography masks substantial variation in these areas. For instance, Baltimore County, Maryland was one of the counties in their sample. They treated all suburbs in the county as “first-tier suburbs.” Yet, the reality is that Baltimore County houses a diverse range of places, including first-tier suburbs, outer suburbs, and rural areas (Outen 2005). Second, Puentes and Warren did not compare first-tier suburbs with the trends in their respective metropolitan areas. This limits the ability to control for regional differences across the nation such as cost of living standards. Despite the limitations, this study stands out as the most comprehensive national study of suburban change. The breadth of study allowed these scholars to make broad generalizations about the trends in first-tier suburbs.

62 / transforming race and class in baltimore

Qualitative Assessments The second set of studies on suburban change involves qualitative assessments. Building on the quantitative suburban studies, two scholars have recently given perspective to the process of suburban change by observing first hand on how the socioeconomic transitions in suburbs affect people and their built environment. Using qualitative methods such as direct observation and interviews, William Hudnut and Dennis Keating give us a portrait of the conditions of the first-tier suburbs at the beginning of the twenty-first century. First, William Hudnut (2003), the former mayor of Indianapolis, studied the decline of older suburbs throughout the nation. Although Hudnut is not a scholar, per se, his journalistic account of the socioeconomic state of first-tier suburbs provides an in-depth understanding about the current challenges that these places face. Hudnut visited 60 first-tier suburban communities across the nation. During these visits, he spoke with public officials, residents, and drove around many suburban neighborhoods. His book is based on these conversations and observations. Four themes of suburban change characterized his findings. Based on his observations, Hudnut developed the following themes to help explain the suburban changes that he witnessed: 1. An aging housing stock. First-tier suburbs tended to have a very old housing stock. Hudnut observed that many houses were deteriorating and were in need of repair. Residents failed to maintain the housing stock. 2. Investment in infrastructure of sprawling, outer suburbs. In all the metropolitan areas that Hudnut visited, he observed new construction of large houses, schools, roads, and water and sewer systems. Substantial public investments went to fund these projects while the first-tier suburbs deteriorated. 3. An aging population from the Baby Boom. First-tier suburbs housed many residents who were born during the population boom during the late 1940s and 1950s. Some five decades later, suburban residents had aged significantly compared to those in the outer suburbs. 4. Deconcentration of poverty from city to suburb. Hudnut drove along the city-suburban political border in many metro areas. He speculated that residents of high-poverty neighborhoods in the central city were beginning to spill over into adjacent first-tier suburbs. Second, Dennis Keating’s (2005) work on the decline of Cleveland’s first-tier suburbs also demonstrates that older suburbs, in close proximity

suburban decline / 63

to the central city, experienced rapid socioeconomic decline relative to all other suburbs in the metropolitan area. He analyzed census data on population and economic characteristics to chart the history of growth and decline of Cleveland’s first-tier suburban population. He found that all the first-tier suburbs that shared a border with Cleveland declined from their socioeconomic status in 1960. The population and household income declined and poverty increased over four decades in these first-tier suburbs. Next, Keating interviewed various first-tier suburban mayors in greater Cleveland to help illuminate the causes and implications of suburban decline. He found evidence that Cleveland’s first-tier suburbs joined politically to confront suburban decline. In 1998, the First Suburbs Consortium of Cleveland was formed. It comprised of the mayors of Cleveland’s first-tier suburbs. Keating argued that the consortium was an ideal forum to discuss the problems of decline, but little could be done without the mandate of some regional policy or government. Keating concluded that three main themes prevailed from his interviews about suburban decline: 1. Lack of community investment. Suburban municipalities suffered from a dwindling tax base. As a result, they were not able to raise the revenue that they had grown accustomed to raising and therefore could not sustain public service levels. 2. Loss of jobs. The deindustrialization of Cleveland’s economy meant that there were fewer jobs in the manufacturing sector. Many of the residents in older suburbs depended on these jobs for generations. 3. Pressure for new development in the outer metropolitan fringe. The outer suburbs grew and the first-tier suburbs did not. There was a high demand for housing and new infrastructure to be built in growing areas. Politicians paid attention to growth and did not focus on first-tier suburbs. The work of Hudnut and Keating represents an important step toward furthering our understanding of the issues facing many older, deindustrialized suburbs. Their research helps to tell the story of suburban decline, which is important for attracting the attention of policymakers, planners, and other leaders. Defining First-Tier Suburbs The third set of studies on suburban change focuses on developing definitions for first-tier suburbs. Specifically, scholars have attempted to label first-tier suburbs and develop criteria for their identification. This has been an important aspect for the study of suburbs because the technique

64 / transforming race and class in baltimore

for identifying first-tier suburbs determines, in part, the outcome of studies on suburban change. Scholars have not yet settled on a technique for identifying first-tier suburbs, nor have they settled on a definition of first-tier suburbs. In fact, numerous academic studies and other reports have used a wide variety of terms to refer to the suburbs immediately outside of central cities. Table 3.1 lists some of the various terms that have been used to describe the older suburban neighborhoods located near the urban fringe of central cities. For instance, these terms describe the characteristics of first-tier suburbs based on their location in the region (e.g., “close-in suburbs,” “inner-ring suburbs,” or “inner suburbs”); based on their age (e.g., “older suburbs”); and based on the infrastructure of the place (e.g., “bedroom suburbs,” or “streetcar suburbs”). Although it is evident that scholars have used a variety of terms to describe these places, there are several characteristics that all of these terms share. They all refer to suburban places that, in general, are the oldest suburbs within the region. In addition, they refer to suburbs that are in close proximity to their respective central city. For the purposes of this study and its application to Baltimore, I adopt the term “first-tier suburbs.” Hudnut (2003) popularized the use of first-tier suburbs in his book Halfway to Everywhere. This term is useful because it recognizes that suburban development occurred in phases, and it does not presume an urban form like other commonly used terms such as “inner-ring.” Three substantive studies developed criteria for defining first-tier suburbs. In the first study, Puentes and Warren (2006) developed a definition Table 3.1

Common descriptions of first-tier suburbs

Term

Source

“Bedroom suburbs”

Muller, (1981); Baxandall and Ewen (2000); Orfield (2002); Lang (2003) Fishman (1987) Downs (1973); Pascal (1987) Puentes and Warren (2006) Hudnut (2003) Orfield (1997); Rusk (1999) Hanlon and Vicino (2007) Barnes (2006) Mikelbank (2004); Vicino, Hanlon, Short (2007) Listokin and Beaton (1983) Warner (1978)

“Bourgeois utopia” “Close-in suburbs” “First suburbs” “First-tier suburbs” “Inner-ring suburbs” “Inner suburbs” “Inurbs” “Middle America suburbs” “Older suburbs” “Streetcar suburbs”

suburban decline / 65

based on the county level geography. They collected data on three criteria, which included the age of the place, the location to central city, and the size of the population, to develop the definition. First, they included all counties that were part of the 1950 standard metropolitan area. Puentes and Warren (2006, 3) argued that “the time these places developed is integral to how they developed” (emphasis in the original). Second, a county had to share a border with one of the top 100 cities of the study’s sample in 1950. Third, a county had to have a minimum of 120,000 residents in 1950. After analyzing these data, Puentes and Warren found that 64 counties in the United States satisfied these three criteria, and these counties contained all of America’s first-tier suburbs, according to these scholars. In the second study, after reviewing the works of scholars who previously studied specifically first-tier suburbs, Hanlon and Vicino (2007) found that (1) a clear definition of first-tier suburbs was lacking; and (2) few scholars had studied empirically the first-tier suburbs. Hanlon and Vicino developed a definition for first-tier suburbs. They identified first-tier suburbs based on two criteria: temporal and spatial. The temporal criterion referred to when the suburb was built, and the spatial criterion referred to the suburb’s physical proximity to the central city. They analyzed the year the housing units were built in the region to specify the temporal criteria. If more than half of the census designated place’s (CDP’s) housing stock was built before 1970, then the suburb satisfied the temporal criterion. Yet, the age of housing alone was not sufficient to define first-tier suburbs because it did not account for the spatial aspect—that is, the degree of proximity to the central city. Older suburbs developed many decades ago, and they frequently developed right on the fringe of the central city. Thus, to be considered a “first-tier suburb” of Baltimore, CDPs also had to share a boundary with Baltimore City, or share an adjacent boundary of a first-tier suburb whose housing stock was older than 30 years. Hanlon and Vicino used 1970 as the threshold for the temporal criterion because it captured the suburbs that had already been fully developed by that date. Since the beginning of the 1970s was also a critical turning point in the development of a metropolitan economy and outer suburban growth (Judd and Swanstrom 2005), this temporal criterion identified suburbs that existed before these regional changes. In the third study, Leigh and Lee (2005) studied suburban change in the Philadelphia region to develop a definition of first-tier suburbs. They relied on two criteria, location and age, to determine what constituted a first-tier suburb. Using census tract data, they analyzed the distribution of the age of the housing stock in the region to develop a definition. They found that the two immediate rings (within 20 miles) surrounding the central city housed the majority of the region’s middle-aged housing stock,

66 / transforming race and class in baltimore Table 3.2

Summary of definitions for first-tier suburbs

Study

Criteria

Scale

Advantages

Disadvantages

Puentes and Warren (2006)

County

Hanlon and Vicino (2007)

Age Location Population Age Location

Sacrifices in-depth analysis of place and neighborhood variation Sacrifices in-depth analysis of neighborhood variation

Leigh and Lee (2005)

Age Location

Tract

Allows comparative broad national scope of suburban change Allows comparative analysis among suburbs within a region Allows in-depth analysis of neighborhood variation

Place

Sacrifices national comparative analysis

built between 1950 and 1969. As a result, they classified all the census tracts in those two rings as first-tier suburbs. Leigh and Lee (2005, 15) argued that “inner-ring suburbs [are] single family, residential suburban areas built between 1950 and 1969.” The development of a definition of first-tier suburbs is a critical starting point for the study of first-suburbs because it determines the scale at which the data are collected, and it helps to develop generalizations about changes in the first-tier suburbs. Table 3.2 summarizes the main elements of these studies and reviews the principal advantages and disadvantages to their definitions. It shows that location and age were common criteria that all three studies used to define first-tier suburbs. Puentes and Warren (2006) also used population size as a criterion. Table 3.2 also shows that the three studies used three different scalar approaches, in hierarchical order: county, place, and tract. The advantage of using a coarse geographic scale, such as counties, facilitates the national comparison of first-tier suburbs, yet the disadvantage is that it sacrifices the in-depth analysis of place and neighborhood variation. In contrast, the advantage of using a fine geographic scale, like tracts, captures significant variation of neighborhoods, yet the disadvantage is that it sacrifices the ability to compare in a broader, national context. Suburban Studies It is clear from the literature that first-tier suburbs experienced socioeconomic decline and diversity at the beginning of the twenty-first century. Recent studies on suburban change can be grouped into three categories. First, the quantitative studies measured the patterns of socioeconomic changes in first-tier suburbs. These studies found that a significant portion

suburban decline / 67

of suburbs experienced socioeconomic decline, which could be characterized by population loss, income decline, and an aging housing stock. Next, the qualitative assessment studies provided an account of how first-tier suburbs changed and how local political leaders think about those changes. Last, several studies focused on developing definitions for first-tier suburbs. Scholars generally agree that both a spatial and temporal criteria is necessary to define first-tier suburbs in a region adequately. Table 3.3 provides a summary of the recent studies on suburban change, and it compares and contrasts the various approaches and findings of the studies. It shows that during the 2000s, there were eight analytical studies on suburban change. The majority of them used place level geography to examine suburban change. There were two exceptions. Leigh and Lee (2005) used both place and tract level geography, and Puentes and Warren (2006) used county level geography. Three of the studies were based on the suburbs of a single region (e.g., Baltimore, Cleveland, and Philadelphia), while the remaining six studies were based on a broad national sample of suburbs. Most of the studies used a variety of quantitative methods, and a few studies integrated qualitative methods. In general, all these recent studies found some evidence of suburban decline. The national studies estimated that approximately one-third to one-half of the nation’s first-tier suburbs were in decline. The regional studies found that some first-tier suburbs declined more than others in their respective metro areas. Overall, the aging of the population and housing stock, accompanied by income loss, characterized the principal findings of these studies. In depth, regional analysis among suburban places was lacking in these studies, as well as neighborhood level analysis. Patterns of Suburban Decline, 1970–2000 The analysis of patterns of suburban decline is divided into four primary sections related to themes on population, income, housing, and labor force. In terms of the population characteristics, I analyze four elements of change in the population of Baltimore’s first-tier suburbs. First, I examine how the size of the population changed over three decades and compare how the distribution of population changed among suburbs. I then explore how the racial composition changed, giving particular attention to the trends of diversity and segregation. Next, I investigate the patterns of change in the age structure of metropolitan Baltimore and last consider the family composition, analyzing differences between married families with children and female-headed households in the suburbs. In terms of the income dynamics, I examine the distribution of median household income and income ratios, and then I explore how the poverty rate is distributed

Table 3.3

Summary of studies on suburban change

Author

Year

Scale

Sample

Approach(s)

Method(s)

Key finding

Lucy and Phillips

2000

Place

24 MSAs

Orfield

2002

Place

25 MSAs

Quantitative Qualitative Quantitative

Descriptive spatial statistics and ground truth analysis Principal components and cluster analysis

Hudnut

2003

Place

Qualitative

Mikelbank

2004

Place

60 first-tier suburbs 25 MSAs

Quantitative

Direct observation and interviews Cluster analysis

Hanlon and Vicino

2007

Place

Baltimore

Quantitative Qualitative

Descriptive spatial statistics and ground truth analysis

Keating

2005

Place

Cleveland

Quantitative Qualitative

Descriptive spatial statistics and interviews

Leigh and Lee

2005

Place and tract

Philadelphia Quantitative

Descriptive spatial statistics

County

64 Counties Quantitative

Descriptive spatial statistics

Found that one-third of suburbs experienced socioeconomic decline Found that 40 percent of suburbs were “at-risk” for more socioeconomic decline Found many suburbs struggling with new demographics and fiscal realities Found 40 percent of suburbs were “struggling” and 29 percent of suburbs were “aging” Developed definition of first-tier suburbs using spatial and temporal criteria Found first-tier suburbs were declining yet working together politically to discuss decline Developed definition of first-tier suburbs using census tracts and spatial criteria Found that first-tier suburbs were one-fifth of the nation’s population

Puentes and Warren 2006

suburban decline / 69

among the population in metropolitan Baltimore. In terms of the nature of housing stock, I observe the age, size, value, tenure, and style of the housing stock. Finally, in terms of structure of the labor, I evaluate the changes in the composition of the industries, the education of the labor force, and the unemployment trends. Collectively, these patterns allow an overall assessment of suburban decline. Population Characteristics Metropolitan Baltimore witnessed a modest population growth during the past three decades. Overall, the region grew 23 percent, from approximately 2 million to 2.5 million since 1970. The region’s central city, Baltimore, continued to lose population since its peak in 1950 of nearly 1 million. Since 1970, the city lost approximately one-third of its residents to the suburbs. Suburban Baltimore grew by 1 million from 1970 to 2000, doubling its population. Despite these growth patterns in the suburbs, the distribution of the growth was uneven. The outer suburbs experienced overwhelming growth although the population of the first-tier suburbs stagnated. A more detailed examination of the spatial distribution of population change over time reveals several distinct patterns. The population of the first-tier suburbs and the outer suburbs mirrored one another in 1970—both stood at approximately 1.5 million. However, since then, a diverging pattern of slow growth, no growth, and population loss occurred, as table 3.4 reveals. In the aggregate, suburbs in the first-tier shrunk by 11 percent with 65,615 persons left between 1970 and 2000. In contrast, the outer suburbs grew significantly during the same period. The population soared by 138 percent in 3 decades, and just more than 800,000 residents moved to the outer suburbs. Since 1970, the population trend among the first-tier suburbs was a pattern of slow growth or loss. Table 3.4 also demonstrates that 15 first-tier suburbs lost population since 1970. In particular, the loss was dramatic in four of these suburbs. Catonsville, Dundalk, Lutherville, and Towson each lost nearly one-third of its population. Only six of the first-tier suburbs experienced population growth since 1970, but the growth in Essex, Glen Burnie, and Hampton is negligible. Ferndale and Woodlawn, located to the west and south of the city, grew substantially. There were 6,170 residents who moved to Ferndale since 1970, which represented a 62 percent increase. Similarly, 9,795 residents moved to Woodlawn, showing a 37 percent population growth rate. To the northeast of the city, Middle River grew by 20 percent. Two trends were especially apparent among the majority of these places between 1970 and 2000. First, they failed to attract additional residents,

70 / transforming race and class in baltimore

Table 3.4

Population in metropolitan Baltimore, 1970–2000 1970

1980

1990

2000

Change Number

Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs Outer Suburbs Baltimore City Region

22,724 20,163 13,847 11,508 54,983 33,208 85,267 71,293 10,346 9,078 38,112 39,614 9,886 14,314 38,547 37,263 4,358 5,220 16,922 16,759 9,878 7,457 29,056 26,908 24,042 17,854 19,917 26,756 13,186 12,965 33,935 35,159 25,252 22,555 6,387 5,666 19,430 19,956 77,825 51,083 26,284 29,453 580,184 514,232 583,652 871,844 905,759 786,775 2,069,595 2,172,851

19,750 20,116 22,608 10,987 10,938 22,909 35,233 39,820 215,163 65,800 62,306 222,961 9,226 9,248 21,098 40,872 39,078 966 16,355 16,056 6,170 37,305 38,922 375 4,926 5,004 646 15,509 15,724 21,198 7,547 7,539 22,339 25,240 25,269 23,787 16,442 15,814 28,228 24,616 23,958 4,041 12,137 12,148 21,038 31,617 31,118 22,817 24,815 29,123 3,871 5,483 5,317 21,070 18,703 19,199 2231 49,445 51,793 226,032 32,907 36,079 9,795 504,915 514,569 265,615 1,141,579 1,387,271 803,619 736,014 651,154 2254,605 2,382,508 2,552,994 483,399

Percent 211 221 228 227 211 3 62 1 15 27 224 213 234 20 28 28 15 217 21 233 37 211 138 228 23

Source: U.S. census.

and when they did, the growth was small relative to the growth in the outer suburbs. Second, many of these communities could not maintain their population base, and they actually lost people. The changes in the spatial distribution of the population during the 1970s, 1980s, and 1990s in Baltimore’s first-tier suburbs indicate that these places lagged behind the population boom in the outer suburbs. Baltimore’s first-tier suburbs are aging at a faster rate compared to the outer suburbs. These places witnessed the number of residents aged more than 65 years more than double from just below 40,000 residents in 1970 to more than 85,000 residents by 2000. In contrast, the number of residents aged less than 18 decreased by 37 percent, and residents aged 18 to 64 decreased by 2 percent. The population aged more than 65 represents at

suburban decline / 71

least 40 percent of the residents in 5 first-tier suburbs, including Catonsville, Hampton, Lutherville, Pikesville, and Towson. Indeed, these aging patterns are most prevalent in the first-tier suburbs. A closer examination of the changes in the age structure reveals several important trends among the first-tier suburbs. The population aged more than 65 increased at a faster pace than the other age cohorts in every first-tier suburb. Table 3.5 indicates that in 6 of these places, the population aged more than 65 quadrupled, which includes Ferndall, Glen Burnie, Hampton, Linthicum, Lutherville, and Middle River. In another 7 places, the population aged more than 65 doubled, which includes Dundalk, Essex, Landsdowne, Overlea, Pikesville, Pumphrey, and Rosedale. Thus, two-thirds of Baltimore’s first-tier suburbs experienced extraordinary increases in the elderly population over a three-decade period. In contrast, Woodlawn is the only first-tier suburb that experienced growth in all age cohorts, but that growth has been the greatest for residents

Table 3.5

Population by age in Baltimore’s first-tier suburbs, 1970–2000

Place

Aged less than 18 Percent

Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs Source: U.S. census.

233 237 233 253 247 226 222 238 248 231 236 231 247 224 229 249 213 260 235 244 1 237

Number 21,701 21,306 22,815 212,871 21,488 22,727 2831 24,223 2634 21,453 2773 22,323 22,950 21,429 2489 25,037 2618 21,192 22,320 25,245 90 252,335

18 to 64 Percent 26 226 11 227 213 21 35 9 26 22 22 28 23 14 28 224 20 215 28 5 43 22

Number 2789 22,272 2,419 213,681 2843 2142 2,598 1,981 2165 2164 2107 21,474 2296 1,769 1,023 25,895 2,535 2562 21,175 1,463 6,569 27,208

Aged more than 65 Percent

Number

21 86 47 168 51 143 263 201 218 100 205 74 213 266 197 85 166 196 168 70 79 112

511 820 2,572 6,936 503 3,153 1,441 3,401 657 814 1,048 1,537 2,945 2,100 905 2,387 3,805 628 2,554 4,280 1,634 44,631

72 / transforming race and class in baltimore

aged more than 65. The population aged more than 65 grew by 1,634, a 79 percent increase since 1970. There was a net increase of 90 residents aged below 18, a slight increase of only 1 percent. Among the adult working population, Woodlawn had the most growth out of any first-tier suburb. More than 6,000 residents aged 18–64 moved to the suburb, bringing the total adult working population to just more than 21,000. More than half of Baltimore’s first-tier suburbs are aging in place; that is, there is growth in the elderly population while all other age cohort populations decline. This trend is especially alarming because it means that there is population loss. No new residents locate to these suburbs, and the residents who remain are aging rapidly. Eventually, as the residential population ages and dies—without growth in younger generations— the suburbs potentially become well poised to decline ever further as the population continually dwindles. Consider the experience of Dundalk, a suburb located on the eastern fringe of Baltimore City. Dundalk had the highest population aged more than 65 increase of any first-tier suburb. Since 1970, Dundalk’s elderly population increased by 7,000 residents. Moreover, the size of the two younger generation populations decreased simultaneously, adding a “double whammy” effect to Dundalk’s population. The youth population represented more than one-third of Dundalk residents in 1970, yet by 2000 less than 19 percent of the population was in the youth cohort. Similarly, the adult working population (ages 18–64) shrunk by 27 percent; a staggering 13,681 people left. At the dawn of the new century, Dundalk was not only aging in place, but it was also losing people without building a successive generation of residents. During the 1990s, a distinct aging pattern emerged among the white and black population in suburban Baltimore. The pattern of aging in the first-tier suburban white population is much more pronounced than the pattern of aging of the black population, as shown in table 3.6. In the Table 3.6

Age structure by race in suburban Baltimore, 1990–2000 Age cohort

White Percent

Black

Number

Percent

Number

First-tier suburbs

Under 18 18–64 More than 65

28 215 1

26,580 240,887 821

87 56 78

13,413 20,494 2,655

Outer suburbs

Under 18 18–64 More than 65

12 7 40

18,098 27,986 21,200

68 57 105

18,600 35,021 4,018

Source: U.S. census. The census data by race and age at the place level geography could only be compared in 1990 and 2000 owing to differences of definitions in these categories.

suburban decline / 73

past decade, there was a net loss of 46,646 white residents aged below 64 in first-tier suburbs. The majority of the population loss was in the white, 18–64 age cohort, which accounted for 88 percent of the overall loss since 1990. In contrast, the black population migrated to the first-tier suburbs in large numbers. These places witnessed a net gain of 36,562 black residents, or an increase of 40 percent since 1990. Moreover, most of the population increases in the black suburban population occurred among residents aged 18–64, with significant increases in the youth population as well. Although the first-tier suburban elderly black population increased 78 percent, the absolute number is small, amounting to some 2,600 residents. Still, there is a trend of significant growth in the black population, of all ages, in Baltimore’s first-tier suburbs. This occurred as the first-tier white population shrunk. The black population partially offset the white population loss. In contrast to the first-tier, Baltimore’s outer suburbs witnessed marked population growth among white and black residents of all ages since 1990. The outer suburban population is significantly younger, and it is aging at a slower rate than the first-tier suburban population. The outer suburban black population aged 18–64 grew more than the white population, adding 57,639 residents in the past decade. Among both whites and blacks, the change in the 18–64 age cohort was the largest, and the youth cohort grew at approximately the same rate for whites and blacks during the 1990s. There were distinct patterns of aging among the first-tier suburbs and outer suburbs of Baltimore between 1970 and 2000. The first-tier suburbs saw marked changes in the elderly population since 1970. In all 21 first-tier suburbs, the population aged more than 65 increased significantly more than in the outer suburbs. At the same time, the youth and working adult populations shrunk in many first-tier suburbs. This suggests that the population aged without successive generations to replace previous ones. There were some exceptions to these trends. In the 1990s, younger black residents migrated to the first-tier suburbs, which offset some of the losses among the white population. Just as the population and age structure transitioned remarkably during the three decades since 1970, so too has the suburban racial landscape in metropolitan Baltimore. During this 30-year period, there were large, and often dramatic, increases in racial diversity throughout the first-tier suburbs. Yet these changes did not occur uniformly; some suburbs witnessed significant diversification of race and ethnicity while others experienced very little change. A closer examination in detail of the changes in the composition of race and ethnicity in the first-tier suburbs reveals several interesting and noteworthy patterns.

74 / transforming race and class in baltimore Table 3.7

Population by race in Baltimore’s first-tier suburbs, 1970–2000 1970

1980

1990

2000

Change

Number Percent Number Percent Number Percent Number Percent White 524,947 Black 18,401 Hispanic 3,923 Other 1,450

96 3 1 0

465,690 37,699 3,699 7,176

91 7 1 1

434,576 54,700 5,254 10,385

86 11 1 2

390,946 91,403 9,576 22,644

76 18 2 4

Percent 226 397 144 1462

Source: U.S. census. The terms “white” and “black” refer to non-Hispanic individuals whose self-identify as a member of only one race on the short form of the census. The “other races and ethnicities” category includes Asian, Native American, and multiple racial groups that the U.S. census bureau uses. This analysis includes only racial and ethnic data that can be geographically compared over a 30-year period owing to changes in the definitions of these groups.

The predominant trend since the 1970s has been one of diversification. There was an increase in all minority populations in the Baltimore metropolitan area from 1970 to 2000. Table 3.7 displays the cumulative racial changes for all of the first-tier suburbs. The greatest changes occurred between the white and black population. The aggregate data for 1970 show that first-tier suburbs were 96 percent white, 3 percent black, and less than 1 percent Hispanic and other races and ethnicities. By 2000, these places were no longer bastions for white suburbanites. The white population dropped to 76 percent while the black population increased noticeably to 18 percent. The size of the Hispanic and other races population was negligible in 1970 and 2000. Over 3 decades, these groups collectively grew from 5,000 residents to 30,000 residents. Although in 2000, they still only represented 4 percent and 2 percent, respectively, of the overall first-tier suburban population. The Baltimore metropolis was primarily a white and black region during the last quarter century Further analysis of the changes in the racial composition among the first-tier suburbs reveals two important trends: increased racial diversity and persistent segregation. What is first most important to underscore is that the white population decreased in every first-tier suburb over the past 30 years. With the exception of two suburbs (Lochearn and Pumphrey), Baltimore’s first-tier suburbs were at least 95 percent white three decades ago. By 2000, 12 suburbs still had a white population of more than 80 percent. However, there remained even greater enclaves of the white population secluded throughout various first-tier suburbs. In particular, four suburbs remained overwhelmingly white. Brooklyn Park, Edgemere, Hampton, and Linthicum still maintained a white population of more than 90 percent by the end of the twentieth century. These suburban places defy the overall trend of increased racial diversity in the suburban landscape.

suburban decline / 75

Second, between 1970 and 2000, 10 of the first-tier suburbs had black populations that grew more than 10 percent. Most of the growth in the black population occurred during the 1990s, and in several cases that growth increased substantially. For instance, in Lansdowne, the black population tripled in the 1990s alone, when the black population grew by 2,011 residents in this suburb, which accounted for one-fifth of the place’s total residential population. Similarly, in various other first-tier suburbs, the black population doubled in a single decade during the 1990s. For instance, on the eastern fringe of Baltimore County, Essex’s black population increased by 4,683 residents, which accounted for one-fifth of its population in 2000. To the northeast, Parkville and neighboring Rosedale each added an additional 3,000 black residents to account for a quarter of their respective population bases in 2000. In yet three other first-tier suburbs of Ferndale, Glen Burnie, and Middle River, the black population also grew, but the growth was more modest. These suburbs each witnessed their black population double during the 1990s alone, but it represented less than 15 percent of its total residential population. Without a doubt, it is evident that the black population increased systematically throughout the first-tier suburbs, and that growth was especially marked during the 1990s. The third pattern in the change in racial composition relates to the spectacular growth of the black population in the western first-tier suburbs. This growth is noteworthy because it equals the growth of the black population that occurred in all other first-tier suburbs combined since 1970. On the western fringe between Baltimore City and the bordering first-tier suburbs, the migration or turnover in the white population was quite marked. Two suburbs located on the western fringe of the city, Woodlawn and Lochearn, are cases in point. Both suburbs were majority white places in 1970, and by 2000, they both were majority black places. In Woodlawn, the white population declined from 97 percent in 1970 to only 38 percent in 2000. In 3 decades, some 12,000 white residents—47 percent of the population—left Woodlawn. As white residents departed Woodlawn, black residents made this suburb a new home. More than 18,000 blacks moved here during that same period. The racial change in Woodlawn did not occur evenly over the course of the past three decades. In the 1970s and 1980s, the racial change was slower than during the 1990s. Approximately three-quarters of Woodlawn’s population remained white until 1990. The fastest rate of racial change occurred in the 1990s when the white population decreased to 35 percent as the black population increased to 51 percent. Indeed, the 1990s ushered in a turbulent decade of racial change for Woodlawn. There was a similar trend in neighboring Lochearn. In 1970, Lochearn had a similar look and feel to many of Baltimore’s other first-tier suburbs.

76 / transforming race and class in baltimore

The suburb was overwhelmingly white and segregated; 91 percent white and 8 percent black. Yet as early as 1980, racial change in the composition of Lochearn’s population was evident. By 1980, the population was halfwhite and half-black—making it the only suburb in the entire metropolitan area to achieve an even racial composition. This balance quickly dissipated during the 1990s. By 2000, Lochearn’s white population declined to 18 percent, and the black population climbed to 78 percent. The changes in the racial composition of these two suburbs are significant because together Lochearn and Woodlawn represent two very large suburban communities containing some 61,000 residents in 17 neighborhoods to the west of Baltimore City. Woodlawn and Lochearn evolved into completely different places for the racial composition of their population had substantially changed in character and nature in very recent history. These dramatic racial changes in the western first-tier suburbs provide evidence of resegregation in first-tier suburbs such as Woodlawn and Lochearn. This process had begun as early as the 1950s in the western neighborhoods of Baltimore City, bordering the suburbs (Orser 1994). By 1970, these places had high levels of white segregation. During the 1980s and 1990s, they integrated to an extent as the black population grew in the first-tier suburbs. Yet, in 2000, the majority of the white population had left, and a black population became the majority. The result was a racial resegregation of the population. In 2000, these places now had high levels of black segregation. The black population invaded the western first-tier suburbs, and they succeeded the white population from 1970 to 2000. Overall, while there has been increased racial diversification, the suburban landscape is still vastly segregated by race and place. In terms of the foreign-born population, metropolitan Baltimore stands out as a region that, by and large, failed to attract significant portions of this growing U.S. demographic since 1970. Although the number and percentage of foreign-born residents increased in the region, the city, the first-tier suburbs, and the outer suburbs, this growth was subtle. In each of these geographic areas, the growth on a decade-by-decade basis was very similar. For example, for the entire Baltimore region, just fewer than 3 percent, or 62,000 residents, were immigrants in 1970. By 2000, that figure grew to represent 6 percent of population, or 153,000 residents. Similar trends were apparent in both the first-tier suburbs and the outer suburbs. In 1970, Baltimore’s first-tier suburbs had 2.8 percent of the population, or 16,245 residents, comprised of immigrants. Over the next three decades, the foreign-born population grew steadily, just as it did in the region. The majority of growth occurred during the 1990s. By 2000, 6.2 percent of the population, or 31,903 residents, were immigrants in the first-tier suburbs. Overall, the largest increases occurred in the outer suburbs. In

suburban decline / 77

1970, 3 percent of the outer suburban population, or 17,509 residents, were immigrants, and by 2000, 7 percent, or 97,180 residents were immigrants in the outer suburbs of Baltimore. However, it should be noted that very little distinguished these growth patterns from those of the region or the first-tier suburbs. With several exceptions, the foreign-born population among Baltimore’s 21 first-tier suburbs was relatively small with little-to-no growth over 30 years. Small increases of 1 percent occurred in half of the first-tier suburbs. There was virtually no growth in the eastern suburbs of Dundalk, Edgemere, and Overlea. In the areas that attracted immigrants, two small enclaves developed in the first-tier suburbs since 1970. First, in the suburb of Pikesville, just to the northwest of Baltimore City, a large settlement of Russian-Jewish immigrants developed. In 1970, 9 percent of the population, or 2,272 residents were foreign-born; 3 decades later, 20 percent, or 5,853 residents, were foreign-born. In other words, nearly one in four residents of Pikesville was foreign-born. Pikesville’s surge in its immigrant population did not occur until the 1990s when the number of immigrants doubled in a single decade. Second, due north of the city is Hampton, the smallest first-tier suburb of Baltimore. During the 1990s, Hampton became home to a growing number of Asian immigrants. In 2000, 11 percent of Hampton’s population, or 550 residents, were immigrants. Aside from these two small immigrant enclaves, first-tier suburban Baltimore did not attract many foreign-born residents. While four other first-tier suburbs, Arbutus, Lutherville, Towson, and Woodlawn, had their relative population share of immigrants increase to 8 percent by 2000, this still represented a small portion of the overall suburban population. Metropolitan Baltimore had a comparatively smaller number of immigrants, unlike the neighboring Maryland suburbs surrounding Washington DC where nearly one out of three residents was foreign-born (Frey 2003b). The last major transformation of metropolitan Baltimore’s population characteristics occurred in the structure of the households and families. Since 1970, there have been striking changes in the composition of households comprised of families with children. The dramatic rise in female-headed households is the most obvious trend. Table 3.8 shows that the overall population trend for the Baltimore region reflects a rise in the households headed by females. For instance, in 1970, less than 50,000, or 16 percent of all households in the region, were female-headed households. By 2000, the number of families headed by females nearly doubled, accounting for just less than one-third of households in the region. A similar, albeit more defined pattern, was also manifest in Baltimore City. Three decades ago, less than one-third of households in the central city were headed by females; yet by 2000, more than half of all households

78 / transforming race and class in baltimore Table 3.8 Female-headed household population in metropolitan Baltimore, 1970–2000 1970

1980

1990

2000

Number Percent Number Percent Number Percent First-Tier Suburbs Outer Suburbs Baltimore City Region

Number Percent

7,629

9

11,133

16

14,191

23

17,986

27

7,900

8

18,207

13

25,221

16

35,151

18

32,604

27

45,606

45

51,763

53

42,863

54

48,133

16

74,946

25

91,175

28

96,000

28

Source: U.S. census.

were headed by females in Baltimore City. More important, two distinct trends among the suburbs were especially apparent. First, there were very few female-headed households in suburban Baltimore in 1970. In both the outer suburbs and the first-tier suburbs, there were approximately 16,000 female-headed households, which constituted 8 percent of the region’s entire suburban population. By 1980, there was a noticeable gap between the female-headed household population in first-tier suburbs and outer suburbs. Although there were a greater number of residents living in female-headed households in the outer suburbs in 1980, there was a greater concentration of female-headed households, relative to the population, in the first-tier suburbs. In 1980, 16 percent of all households were headed by females, and over the next two decades that population cohort only climbed. Since 1970, the female-headed household population grew throughout metropolitan Baltimore. In the suburbs, that population increase was more pronounced in the first-tier suburbs rather than the outer suburbs—and by 2000, nearly one-third of residents in Baltimore’s first-tier suburbs lived in female-headed households. Second, a comparison among first-tier suburbs reveals substantial variation in the female-headed household population trend since 1970. This population cohort grew especially concentrated in the western suburbs of Lansdowne and Lochearn. In 1970, female-headed households represented 11 percent of Lansdowne’s population and 6 percent of Lochearn’s population. Three decades later, nearly half of the population in these suburbs was comprised of female-headed households—putting them on par with Baltimore City. In addition, the female-headed household population tripled in seven other first-tier suburbs, including Dundalk, Essex, Lansdowne, Lochearn, Middle River, Parkville, and Woodlawn. In contrast, three suburbs stand out as outliers, defying the trends of increasing female-headed households over time. Hampton, Linthicum,

suburban decline / 79 Table 3.9 Married households with children population in metropolitan Baltimore, 1970–2000 1970

1980

1990

2000

Number Percent Number Percent Number Percent Number Percent First-Tier Suburbs Outer Suburbs Baltimore City Region

53,826

85

51,273

80

41,590

73

39,501

66

57,605

91

86,407

84

134,058

80

175,755

76

194,285

65

142,799

55

114,440

43

109,905

39

519,054

76

523,439

73

411,048

65

342,857

64

Source: U.S. census.

and Lutherville each had less than 13 percent of a female-headed household population, and the increases over each decade occurred at a slower growth rate than the rest of the first-tier suburbs. These trends reveal that, while there was variation among the first-tier suburbs, overall, these areas witnessed marked increases in the number of female-headed households over the past three decades. As the female-headed population soared throughout metropolitan Baltimore, the size of the population of married families with children decreased substantially, in what perhaps was a rhythmic response to the evolutionary nature of U.S. household and family structure. Table 3.9 shows that the region had more than half a million households, or 76 percent, that were married with children. By 2000, only two-thirds of households were comprised of married families with children. In Baltimore’s suburbs, a comparable trend was evident. In 1970, more than 100,000 households—nearly 9 in 10—were comprised of married families with children in the suburbs. There was little variation between the first-tier and outer suburbs. Yet over time, the first-tier suburbs failed to retain the proportion of its population of married families with children households. In 1980, the cohort declined to 80 percent, and it continued to plummet until 2000 when only 66 percent of households were married with children in first-tier suburbs. By comparison, the outer suburbs had a slightly higher proportion of the population of married families with children than the first-tier suburbs. By 2000, the population in the outer suburbs still contained significantly more married families with children than all other areas of the region. Moreover, it is no surprise that in the same firsttier suburbs that had the largest increase in the female-headed household population, the household population of married families with children decreased in a similar fashion. The first-tier suburban population of married families with children fell by one-third since 1970. Lansdowne and

80 / transforming race and class in baltimore

Lochearn, for example, both witnessed that population declined in half over three decades. The overall trend is that metropolitan Baltimore, its central city, and surrounding suburbs, all witnessed declines in the household population comprised of married families with children. Although these trends mirrored larger national demographic patterns, the changes since 1970 in the first-tier suburban population were much more dramatic compared to the counterpart in the outer suburbs. In summary, Baltimore’s first-tier suburbs experienced significant increases in racial diversity over the 30-year period of this study. Two trends stand out. First, with two exceptions, the black population increased in every first-tier suburb. Second, the western first-tier suburbs experienced a pattern of resegregation as the population transformed from all white to a balanced racial population to all black. Yet, the aggregate pattern over three decades showed that Baltimore’s suburbs remained overwhelmingly white, and the central city was still predominately black. The first-tier suburbs were slightly more diverse than the outer suburbs. The racial diversification was marked and distributed in an uneven spatial pattern. The western suburban frontier contained the highest degree of black segregation. In contrast, the eastern suburban fringe housed large portion of an exclusively white population. Furthermore, metropolitan Baltimore as a whole failed to attract immigrants, lagging behind other neighboring regions in its ability to compete for new residents. The first-tier suburban population was also home to an increasingly older population than anywhere else in the suburbs. Suburban Baltimore experienced noteworthy changes in the composition of households and families over the past three decades. The most important change was the shift from households comprised of married families with children to female-headed households. The onset of the 1980s brought the first wave of significant changes to suburban families. During the 1990s, these changes became even more pronounced as the gap widened between outer suburbs and the first-tier suburbs. Faster rates of change in the firsttier suburbs continued to outpace the rates of changes in the region, the outer suburbs, and even Baltimore City. The rise of nontraditional families became a distinguishing characteristic of first-tier suburbs. In short, the predominant population feature of suburban Baltimore can be characterized as a suburban frontier that witnessed dramatic change in its racial composition, age structure, and household composition in a short period. Income Dynamics The socioeconomic status of the residents is another important barometer for assessing the well-being of a population. Since 1970, metropolitan

suburban decline / 81

Baltimore witnessed a substantial transformation in the spatial distribution of its high-income and low-income population. For most of the twentieth century, there was a clear decentralization of high-income residents to the suburban frontier. Low-income residents remained spatially isolated in or near the urban core. Yet, the onset of the 1970s began to challenge these long-held, bedrock assumptions about the location and distribution of metropolitan income dynamics. It became apparent that low-income residents—and indeed families living in poverty—would move to the suburbs. In other cases, suburban residents grew poorer as their neighbors in the outer suburbs flourished. To capture the evolution of these changes in metropolitan Baltimore, I examine three important measures of income dynamics. An overview of the analysis of poverty, household income, and income ratio change is presented. One of the most revealing changes in the economic status of residents in metropolitan Baltimore was the increase of the poverty level since 1970. Table 3.10 displays the poverty trends through the region. Overall, the region witnessed no increase in the proportion of its population living in poverty. Nearly 230,000 residents, or 12 percent of the regional population, have lived in poverty since 1970 and continue to today. During that same period, there were major differences in the changes in the poverty level in the suburbs and central city. Baltimore City remained home to the largest number and proportion of residents living in poverty in the region; by 2000, one in four residents was poor. In the suburbs, a more multifaceted portrait of poverty emerged over the past three decades. For instance, the number of residents living in poverty in the first-tier suburbs doubled between 1970 and 2000. The poverty level stood at 4 percent in 1970, and by 2000, more than 8 percent of the population was living in poverty. More than 12,000 residents grew poor during these 30 years in the first-tier suburbs. The outer suburbs, in contrast, witnessed a slight decline in poverty during the same period. In fact, they actually had a higher poverty level in 1970 than the first-tier suburbs. Three decades ago, the outer suburbs were relatively undeveloped. As such, the population and landscape was largely rural, and thus these areas typically had higher poverty rates than urbanized areas (Lopez, Adelaja, and Andrews 1988). By 1980, poverty levels in both the first-tier and outer suburbs converged at 6 percent. During the 1980s, a diverging spatial pattern in the distribution of the poverty population emerged in the suburbs. Poverty increased in the first-tier suburbs as it simultaneously decreased in the outer suburbs. The gap continued to widen in the 1990s. Poverty began to level off in the outer suburbs by 2000, and it continued to increase in the first-tier suburbs. At the end of the twentieth century, the first-tier suburbs had the highest concentration of a poverty-stricken population in suburban Baltimore.

Table 3.10

Poverty in metropolitan Baltimore, 1970–2000 1970

Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs Outer Suburbs Baltimore City Region Source: U.S. census.

1980

1990

2000

Change

Number

Percent

Number

Percent

Number

Percent

Number

Percent

1,100 677 2,221 4,141 772 2,261 468 2,221 127 779 169 952 302 1,042 696 1,110 899 685 721 3,512 678 25,533 14,765 163,486 229,317

5 5 4 5 7 6 5 6 3 5 2 3 1 5 5 3 4 11 4 5 3 4 6 18 12

1,143 707 1,794 3,977 477 3,808 813 2,765 60 1,405 194 892 367 2,221 515 1,799 843 278 826 1,982 1,159 28,025 33,028 176,476 237,529

6 6 6 6 5 10 6 8 1 8 3 3 2 8 4 5 4 5 4 4 4 6 6 23 13

1,197 866 1,664 4,392 515 5,296 657 2,287 151 1,463 78 1,388 434 2,373 346 1,816 1,124 337 849 2,498 1,269 31,000 32,634 156,284 219,918

6 8 5 7 6 13 4 6 3 10 1 6 3 10 3 6 5 6 5 6 4 6 4 22 11

1,362 877 1,655 5,641 644 4,748 1,247 2,977 90 2,159 268 1,751 554 2,217 619 2,268 1,972 404 1,033 3,523 2,257 38,266 47,210 143,514 228,990

7 8 5 10 7 12 8 8 2 14 4 7 4 9 5 7 7 8 5 8 6 8 5 23 12

Number Percent 262 200 2566 1,500 2128 2,487 779 756 237 1,380 99 799 252 1,175 277 1,158 1,073 2281 312 11 1,579 12,733 32,445 219,972 2327

24 30 225 36 217 110 166 34 229 177 59 84 83 113 211 104 119 241 43 0 233 50 220 212 0

suburban decline / 83

A comparison of poverty status among Baltimore’s first-tier suburbs demonstrates considerable variation. Nineteen of the twenty-one first-tier experienced increases in the proportion of the population living in poverty since 1970. The two exceptions were Hampton and Pumphrey, whose poverty population decreased by one-third. Table 3.10 furthermore shows that a number of suburbs had much larger increases in poverty than others. For example, Dundalk and Essex, each located on the waterfront to the east of Baltimore City, reached a 10 and 12 percent poverty level, respectively, by 2000. Similarly, Lansdowne experienced the largest increase in poverty and approached a 15 percent poverty rate by 2000—the highest poverty rate of any suburb in metropolitan Baltimore. Unlike the overall poverty trend, three suburbs stand out as remarkably stable. Hampton, Linthicum, and Lutherville maintained a very low poverty population over three decades; the poverty rate remained less than 4 percent for 30 years. These trends reveal that suburban poverty emerged in concentrated pockets throughout neighborhoods in first-tier suburbs. A further examination of the change in the median household income confirms an analogous pattern to the trend of poverty status. The analysis of median household income change reveals that metropolitan Baltimore and its suburbs improved each decade since 1970 in the accrual of household income. Table 3.11 shows that three decades ago, the region’s median household income stood just below $40,000, and by 2000, it had risen 27 percent to nearly $50,000. Suburban Baltimore was the primary benefactor of the region’s rising tide of median household income. Over 30 years, the suburbs grew 23 percent more income, or $10,837, reaching the region’s highest level of $57,558 in 2000. In contrast, median household income in Baltimore City fell dramatically since 1970, plummeting by some $5,000, or 11 percent, to $35,438 by 2000. Some interesting observations apply to the first-tier suburbs. The median household income in Baltimore’s first-tier suburbs was $49,669 in 2000, which represented a 9 percent decline from $54,530 in 1970. Furthermore, there was a diverging trend among first-tier suburbs and the region’s suburbs. Median household income for all suburbs consistently rose each decade from $46,721 in 1970 to $57,558 in 2000. Yet, in 1970, the first-tier suburbs were better off than all other suburbs in the region. By 1980, median household income was equally distributed between the suburbs—both the first-tier suburbs and outer suburbs had a median household income of approximately $50,000. During the 1980s, the first-tier suburbs began to lose income relative to all other suburbs. By the 1990s, median household income in the first-tier suburbs continued to deteriorate even more as the income in the outer suburbs carried on its growth.

84 / transforming race and class in baltimore Table 3.11

Median household income in metropolitan Baltimore, 1970–2000 1970 ($)

1980 ($)

1990 ($)

2000 ($)

Change Percent

Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs All Suburbs Baltimore City Region

45,141 45,986 50,250 45,450 43,418 41,473 49,595 46,550 129,917 42,223 55,800 64,716 68,473 43,618 48,405 51,550 67,605 44,241 50,668 51,236 58,822 54,530 46,721 40,016 39,289

43,676 44,494 48,039 46,740 48,786 40,687 47,862 47,122 98,660 38,880 61,775 52,444 70,547 42,163 49,469 44,131 62,005 49,637 51,398 56,377 50,646 52,168 50,620 36,076 42,949

47,412 42,740 52,948 41,884 48,381 36,993 50,394 48,495 91,222 41,326 63,817 53,323 68,997 41,382 48,833 44,740 65,175 49,047 53,026 56,036 55,102 52,442 56,946 37,911 49,107

47,792 42,207 53,061 39,789 46,928 34,978 45,816 45,281 95,546 37,160 61,479 49,517 61,573 37,900 48,242 41,410 58,598 45,321 47,801 53,775 48,878 49,669 57,558 35,438 49,938

6 28 6 212 8 216 28 23 226 212 10 223 210 213 0 220 213 2 26 5 217 29 23 211 27

Net ($) 2,651 23,779 2,811 25,661 3,510 26,495 23,779 21,269 234,371 25,063 5,679 215,199 26,900 25,718 2163 210,140 29,007 1,080 22,867 2,539 29,944 24,861 10,837 24,578 10,649

Source: U.S. census. Constant 1999 dollars.

Despite these aggregate trends, there are some important differences to note regarding household income among the first-tier suburbs. Table 3.11 shows that, with the exception of four suburbs, every suburb experienced a loss of real income. That is, after taking into account changes of inflation, losses occurred in many suburbs. A finer grain analysis of median household income in the first-tier suburbs demonstrates that these changes varied to quite an extent. There were pockets of great wealth in the first-tier suburbs. For example, Hampton was the wealthiest first-tier suburb over the past three decades; yet, it experienced the largest income loss between 1970 and 2000. Median household income in Hampton dropped from $130,000 to $95,000. Despite this loss, Hampton remained a bastion of affluence— the only such case in the first-tier suburbs of metropolitan Baltimore. Like Hampton, the other suburbs located to the north of Baltimore City,

suburban decline / 85

including Pikesville, Towson, and Lutherville, all had higher income levels in 1970 than the other first-tier suburbs, but none paralleled the wealth of Hampton. This contrasts dramatically with the suburbs at the other end of the income spectrum. Four suburbs particularly stood out as poor places. Dundalk, Middle River, Essex, and Lansdowne lost a significant amount of income over 30 years. In 1970, these suburbs were largely middle-class places, with median household incomes more than the $40,000 level for metropolitan Baltimore. Yet by 2000, residents lost an average 15 percent of their income, and their income fell to $35,000. Further, to illustrate the income disparity between suburbs, it is useful to analyze the income ratios. This is a useful method for measuring the income status of suburban areas. It compares income levels of individual first-tier suburbs with the income level for all suburbs. The median household income of a particular first-tier suburb is compared with the suburban median household income in the form of a ratio. If the median household income in the first-tier suburb is the same as the overall suburban median household income, the ratio is one. A number lower than one indicates that the suburb lags behind all other suburbs, and a ratio greater than one indicates that the suburb has more income than the median suburban income level. The average income ratio for the first-tier suburbs steadily declined over 30 years. In 1970, the first-tier suburban income ratio to all suburbs was 1.17, indicating that the first-tier suburbs actually had 17 percent more income than the overall suburban median income for the region. In 1980, the income ratio was 1.03, which suggested that household income was evenly distributed between the first-tier suburbs and all other suburbs. During the 1990s, the ratio continued to decline to 0.92, and by 2000, the income ratio reached a low of 0.86. This demonstrates that in 2000 the first-tier suburbs collectively had 14 percent less income than all other suburbs. Therefore, over 30 years, the first-tier suburbs lost 31 percent of the share of their income relative to all other suburbs in the region. Income ratios fell in every first-tier suburb since 1970, and in many cases, the change in the income ratio from 1970 to 2000 was dramatic. Dundalk, Essex, Middle River, and Lansdowne each had approximately 10 percent less income than all other suburban households in 1970. However, by 2000, each of these suburbs had 40 percent less income than all other suburbs in the region. Catonsville and Towson were the only two suburbs in 2000 that approached an even income ratio with all other suburbs. All of Baltimore’s first-tier suburbs had significantly lower income ratios in 2000 than in 1970, indicating that these places fell far behind their suburban counterparts in the rest of the metropolitan area.

86 / transforming race and class in baltimore

In summary, spatial transformations defined the income dynamics of metropolitan Baltimore between 1970 and 2000. The first-tier suburbs were formerly middle-class areas up to the 1970s. Many of these suburbs were indeed more affluent than the region. Yet each subsequent decade after 1970, residents in the first-tier suburbs grew increasingly worse-off economically compared to their neighbors in the outer suburbs. Poverty increased, and household income declined. A review of the income ratios showed just how disproportionate these losses were. By 2000, the first-tier suburban reality was that the socioeconomic status of residents was deteriorating in a similar manner that Baltimore City had a generation ago. Nature of the Housing Stock Five main characteristics define the nature of the housing stock in metropolitan Baltimore. In particular, the age, size, value, tenure, and style of the housing stock in first-tier suburbs are important characteristics for assessing the nature of the housing stock in the first-tier suburbs. These trends are important for assessing the nature and quality of the housing stock (Green and Malpezzi 2003), and they are each reviewed in turn. The age of housing is an important first measure that determines when the housing unit was built. In metropolitan Baltimore, table 3.12 shows that more than half of the housing stock was constructed before 1970, and approximately one-third was built during the 1950s and 1960s. The housing stock in the central city is by far the oldest. More than 85 percent of the housing stock in Baltimore City was built before 1970, and more than one-third was built before 1939. A mere 3 percent of the housing stock was built during the most recent decade. On the other hand, there were prominent differences in the age of the housing stock throughout suburban Baltimore. Housing units in first-tier suburbs were considerably older than other suburbs. The bulk of the housing stock—45 percent—in the first-tier suburbs was constructed during the 1950s and 1960s. Moreover, one-third of the housing stock in the first-tier was very old since it was built before 1950, whereas the number of pre-1950 housing units in the outer suburbs was negligible. The housing stock in the outer suburbs was substantially younger than the stock in the first-tier suburbs. Just below one-third of the housing stock in the outer suburbs was built before 1970, compared to more than two-thirds of the housing stock in the first-tier suburbs. In addition, more than a quarter of the stock in the outer suburbs was built in the 1990s alone. Examining the housing age among first-tier suburbs shows the housing stock was both older and newer. Table 3.12 shows that the majority of houses in the first-tier suburbs, except for Pikesville, were built before

suburban decline / 87 Table 3.12 Age of housing stock in metropolitan Baltimore, 2000 Pre-1939 Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs Outer Suburbs Baltimore City Region

21 14 21 13 22 8 4 4 4 13 11 5 5 6 18 8 4 11 7 14 4 11 4 37 17

1940–1949 1950–1969 1970–1989 1990–1999 17 34 12 25 17 19 6 13 4 10 9 10 6 25 16 19 4 11 8 15 8 15 3 18 11

42 36 32 52 31 39 46 46 62 58 50 55 68 28 36 56 37 39 53 46 43 45 22 30 30

14 13 21 7 21 26 36 27 22 13 20 24 20 31 23 15 36 32 25 21 36 22 46 12 29

7 3 14 2 9 8 8 10 7 6 10 6 2 10 7 2 19 7 8 5 9 7 25 3 13

Source: U.S. census.

1970. New housing units constituted a very small portion, 10 percent of the housing stock, in the first-tier suburbs, with Catonsville and Pikesville leading new housing units. Four suburbs stood out as having particularly older housing stocks. Arbutus, Catonsville, Edgemere, and Overlea each had approximately one-third of the housing stock built in the 1940s or earlier. In other suburbs, suburban tract developments were built immediately following the conclusion of World War II to accommodate returning G.I. veterans. Such suburbs included Brooklyn Park, Dundalk, and Middle River, where more than one-quarter of the housing stock was built during this time both for workers during the war and for war veterans. In short, the age of the housing stock is an important factor for charting the suburban changes in the first-tier suburbs. Most of the housing stock in the first-tier suburbs was considerably older than the stock in the outer suburbs. By 2000, the first-tier suburbs were fully built out.

88 / transforming race and class in baltimore

The size of housing is a second measure that specifies the number of rooms per housing unit. To examine the size of housing units, the number of rooms is calculated using a standard grouping of three or less rooms, four to six rooms, and seven or more rooms. In metropolitan Baltimore, there were many differences in the size of housing units. Table 3.13 shows that approximately half of all housing units in the region had between four and six rooms, while just more than one-third of the houses had seven rooms or more. The central city was home to the region’s smallest houses. One in five houses in Baltimore City had three or fewer rooms, and three-quarters had fewer than six rooms. In the suburbs, housing units in the first-tier suburbs were considerably smaller than those in the outer suburbs. Two-thirds of the houses in first-tier suburbs had six or fewer rooms. Conversely, 91 percent of the housing units in the outer suburbs had more than 4 rooms. The suburbs had the region’s largest houses, but Table 3.13

Size of housing units in metropolitan Baltimore, 2000 3 rooms or less

Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lutherville Lochearn Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs Outer Suburbs Baltimore City Region Source: U.S. census.

4–6 rooms

7 or more rooms

Number

Percent

Number

Percent

Number

Percent

841 284 2,252 2,152 398 2,905 741 2,122 38 518 60 460 863 1,181 343 1,609 1,893 179 564 3,336 1,666 24,405 32,807 62,915 122,781

10 7 14 8 10 17 12 13 2 9 2 7 8 12 7 12 14 8 8 15 11 11 9 21 14

3,911 2,387 6,228 17,483 2,315 10,318 3,800 8,848 339 4,416 1,098 2,217 5,062 6,706 3,042 7,666 6,156 1,120 3,902 9,385 6,822 113,221 149,821 161,290 433,403

47 56 39 66 61 61 59 56 17 73 37 34 49 67 58 57 46 52 53 43 47 53 43 54 49

3,628 1,609 7,528 6,746 1,097 3,804 1,891 4,906 1,567 1,107 1,811 3,924 4,322 2,090 1,841 4,250 5,310 847 2,932 9,304 6,087 76,601 168,050 76,272 325,390

43 38 47 26 29 22 29 31 81 18 61 59 42 21 35 31 40 39 40 42 42 36 48 25 37

suburban decline / 89

the favored quarter was certainly located in the outer suburbs, and modest sizes characterized the first-tier suburban housing stock. A comparison within first-tier suburban Baltimore shows even greater variation in the size of the housing units. Table 3.13 demonstrates that half of the first-tier suburbs had at least 10 percent of their housing stock with fewer than 3 rooms. Overall, there were almost 25,000 housing units in Baltimore’s first-tier suburbs with less than 3 rooms. In particular, five suburbs stood out as places with the smallest housing stock. Dundalk, Edgemere, Essex, Lansdowne, and Middle River each contained housing stocks that had at least 70 percent of the entire stock with 6 or fewer rooms. Lansdowne had the smallest housing units of all first-tier suburbs—82 percent of the stock had less than 6 rooms. In stark contrast, three suburbs had exceptionally huge housing units compared to other first-tier suburbs. In Linthicum and Lutherville, 60 percent of the housing stock had at least 7 rooms. Hampton, the suburb with the largest houses in the first-tier suburbs, had 80 percent of its housing stock with 7 or more rooms. To demonstrate these trends further, a brief look at the square footage of housing units is useful. Both the western and eastern suburbs bordering Baltimore City housed the bulk of units less than 1,500 square feet. Arbutus and Lochearn, for example, located on the western suburban fringe, as well as Dundalk, Essex, Middle River, Overlea, and Parkville on the eastern suburban fringe, all contained virtually no housing units of more than 1,500 square feet. On the northern suburban fringe, a different trend was notable. Pikesville, Towson, Lutherville, and Hampton had predominately larger units, some as large as 5,000 square feet. Above all, Hampton’s housing stock was the largest. It was the only suburb that had practically no housing units below 1,500 square feet. It remained Baltimore’s first-tier suburb for exclusively large housing units. Overall, first-tier suburbs contained the majority of the suburbs’ smallest housing units. They were quite small compared to today’s standards. Among first-tier suburbs, there was considerable variation in the size of housing units. Yet, the ubiquitous trend was that larger houses were located further away from the urban core while smaller houses were located in the first-tier suburbs near the urban core. The value of housing is a third important characteristic in the assessment of the nature of the housing stock. Figure 3.1 charts the value of housing in suburban Baltimore between 1970 and 2000. Houses in the first-tier suburbs were worth approximately $20,000 less than outer suburbs in 1970. During the 1970s, Baltimore’s suburban housing stock gained substantial value, adding some $40,000 in net worth for homeowners. By the 1980s, the gap in the value of houses in the suburbs reached its lowest

90 / transforming race and class in baltimore $180,000

Value of House (1999 Dollars)

$163,978 $160,000

$157,842 $142,049

$140,000 $139,550 $129,961

$120,000

$126,019

$105,847 $100,000

$80,000

$84,596

$60,000 1970

1980

Outer Suburbs

Figure 3.1

1990

2000

First-Tier Suburbs

Average value of housing units in suburban Baltimore, 1970–2000.

Source: U.S. Census.

point. The disparity in housing value between the first-tier suburbs and outer suburbs was only $13,000, making it the lowest during the 30-year period. During the 1990s, however, house values in the outer suburbs continued to climb gradually, growing by 15 percent in one decade. The first-tier suburbs increased by 7 percent, and the average value peaked at $139,550. In 2000, the disparity in housing value between the suburbs was the greatest, standing at some $31,000. Among first-tier suburbs, there were two extremes in the trends of housing value. First, there were only 5 suburbs that had values more than $150,000 by 2000. Housing values in Hampton, Linthicum, Lutherville, Pikesville, and Towson peaked in 1990. Although the value of these places sunk during the 1990s, they were still worth the most in 2000 among all other first-tier suburbs. Hampton stood out as having the most valuable housing stock, which was still worth an average of a quarter of a million dollars per house in 2000. Second, four suburbs with the least valuable housing stock lost value during the 1980s and 1990s. Dundalk, Essex, Lansdowne, and Middle River each had housing values that were below $100,000 in 1970, and they remained below $100,000 in 2000. In fact,

suburban decline / 91

Dundalk’s housing stock on average was only worth $82,500 in 2000, making it the least valuable among the first-tier suburbs. Examining the ratio of housing values reinforces the trend that the housing stock in Baltimore’s first-tier suburbs was losing value relative to the growth in value of all other suburbs. Similar to the income ratio statistic, the housing value ratio measures the value of the housing stock for an individual first-tier suburb relative to the housing value for all suburbs. The housing value of a particular first-tier suburb is compared with the suburban housing value in the form of a ratio. If the housing value in the first-tier suburb is the same as the overall suburban housing value, the ratio is one. A number lower than one indicates that the suburb has a smaller value relative to all other suburbs, and a ratio greater than one indicates that the suburb has a greater housing value than all other suburbs in the region. Table 3.14 shows the change in the ratio of average housing value in the first-tier suburbs to all suburbs from 1970 to 2000. These data provide a telling illustration of the spiral of declining house value in the first-tier suburbs. With the exception of Edgemere, the housing stock in every first-tier suburb was worth less in 2000 than in 1970. In other words, the first-tier housing stock lost value relative to the worth of the housing stock in all suburbs. These losses were significant. In the aggregate, the first-tier housing stock was worth on average 5 percent more in the 1970s and 1980s. Yet by 2000, the first-tier was worth 13 percent less than all other suburbs. In 1970, more than half of suburbs in the first-tier had housing values worth more than the median value of houses in suburban Baltimore. By 2000, only five suburbs had ratios more than one, meaning that they had values worth more than all other suburban houses. Hampton’s housing stock was worth one and a half times more than that of the suburban housing stock, although Dundalk’s houses were worth nearly half as much as houses in the suburbs as a whole. By any measure, the housing stock in Baltimore’s first-tier suburbs was worth considerably less than other suburbs in the region. During a period in American history of unprecedented economic growth in the housing sector, the first-tier suburbs did benefit as much as other suburbanites did. Indeed, housing trends worsened over time, and on average housing values plummeted relative to the outer suburbs each decade since 1970. Housing tenure is a fourth element of the housing stock, which refers to the composition of housing units including owner-occupied units, renter-occupied units, and vacant units. Collectively, they comprise the entire housing stock of a place. Several important developments occurred since 1970 in the tenure of the housing stock. Throughout the metropolitan area, Baltimore City, and the outer suburbs, three specific trends in

Table 3.14 Ratio of average housing values in Baltimore’s first-tier suburbs to all suburbs, 1970–2000 1970

Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs All Suburbs Source: U.S. census.

1980

1990

2000

Change

Value ($)

Ratio

Value ($)

Ratio

Value ($)

Ratio

Value ($)

Ratio

Value

Ratio

73,636 64,091 84,091 53,636 61,818 58,182 76,818 73,636 207,234 53,636 94,091 85,860 134,091 50,455 71,818 79,545 115,000 75,455 82,727 95,000 85,708 84,597 80,621

0.91 0.79 1.04 0.67 0.77 0.72 0.95 0.91 2.57 0.67 1.17 1.06 1.66 0.63 0.89 0.99 1.43 0.94 1.03 1.18 1.06 1.05 1.00

121,839 97,931 137,701 86,897 105,287 97,701 131,494 121,839 238,851 90,115 149,655 120,920 176,322 92,874 115,402 111,264 186,437 122,989 128,736 161,839 133,103 129,962 125,799

0.97 0.78 1.09 0.69 0.84 0.78 1.05 0.97 1.90 0.72 1.19 0.96 1.40 0.74 0.92 0.88 1.48 0.98 1.02 1.29 1.06 1.03 1.00

123,553 108,883 148,318 91,521 119,919 103,365 128,937 125,034 292,598 100,404 166,218 114,939 201,211 97,174 117,766 116,016 204,441 135,666 126,918 180,081 127,591 139,550 149,736

0.83 0.73 0.99 0.61 0.80 0.69 0.86 0.84 1.95 0.67 1.11 0.77 1.34 0.65 0.79 0.77 1.37 0.91 0.85 1.20 0.85 0.93 1.00

112,700 99,200 141,300 82,500 124,800 98,400 126,800 119,500 248,000 96,400 150,000 104,600 170,700 88,500 107,300 100,200 168,000 126,600 114,100 157,100 109,700 126,019 144,511

0.78 0.69 0.98 0.57 0.86 0.68 0.88 0.83 1.72 0.67 1.04 0.72 1.18 0.61 0.74 0.69 1.16 0.88 0.79 1.09 0.76 0.87 1.00

39,064 35,109 57,209 28,864 62,982 40,218 49,982 45,864 40,766 42,764 55,909 18,740 36,609 38,045 35,482 20,655 53,000 51,145 31,373 62,100 23,992 41,422 63,890

20.13 20.11 20.07 20.09 0.10 20.04 20.08 20.09 20.85 0.00 20.13 20.34 20.48 20.01 20.15 20.29 20.26 20.06 20.24 20.09 20.30 20.18 0.00

suburban decline / 93

housing tenure became evident since 1970: (1) increasing homeownership rates, (2) decreasing rentership rates, and (3) decreasing housing vacancy rates. Baltimore City is an exception to the trend in housing vacancy rates. The number of vacant units in the city consistently grew for 3 decades, reaching nearly 40,000 units by 2000. In suburban Baltimore, home ownership rates increased, and rentership rates decreased since 1980. In contrast, first-tier suburbs witnessed slight reductions in homeownership rates and growth in rentership and vacancy rates. Overall, home ownership rates declined from 70 percent to 66 percent since 1970. Rentership rates increased from 29 percent to 31 percent. Housing vacancy rates rose steadily between 1970 and 2000. The number of vacant units increased by one-third since 1970, adding some 4,500 vacant units to the overall housing stock. There were four assorted trends in the changes in housing tenure among the first-tier suburbs. First, there were especially high home ownership rates—more than 70 percent—in Brooklyn Park, Catonsville, Dundalk, Edgemere, Overlea, Pumphrey, and Rosedale, but almost all of those rates declined from their 1990 levels. Second, low home ownership rates and high rentership rates—more than 40 percent—were particularly prevalent in Essex, Lansdowne, and Middle River. Third, four suburbs, including Brooklyn Park, Edgemere, Overlea, and Pumphrey, had comparatively low rentership rates that hovered approximately 20 percent. Fourth, housing vacancy rates increased steadily in four suburbs, including Dundalk, Essex, Lansdowne, and Middle River. In each of these suburbs, the vacancy rates were well over the suburban average of 4 percent. It is important to note that the tradition of home ownership was common for most residents of first-tier suburbs between 1970 and 2000. This held true for affluent suburbs such as Hampton and Lutherville, as well as poorer suburbs such as Brooklyn Park and Dundalk. With the onset of the 1980s, changes in the tenure of the housing stock slowly began. By 2000, the tenure of housing was significantly different from the tenure characteristics in 1970. A variety of housing types and styles characterizes the fifth element of the housing stock in Baltimore’s first-tier suburbs. Ranging from rowhouses and bungalows to mansions, first-tier housing development reflects the era in which the houses were built. Five distinct styles characterize this wide array of housing in the first-tier suburbs. The data that I rely on to characterize these housing styles and stocks is based on field visits and ground truth. Catonsville, located just over the southwest city line, was one of the earliest suburbs of Baltimore. It was a classic “streetcar suburb,” and it provided housing options to many of Baltimore’s early class of elites.

94 / transforming race and class in baltimore

Various houses were built at the beginning of the twentieth century in a Victorian style. A substantial number of houses are as large as 5,000 square feet, and they are all detached housing units on about a half acre of land. Victorian neighborhoods in Catonsville feature copious amounts of green space, and the streets are lined with lush trees and shrubs. Now more than a century old, many of these houses are some of the oldest in the suburbs, and they are in need of repair. Reinvestment and maintenance of housing is commonplace in this first-tier suburb. A number of houses have been converted into apartments to cater to the student body at nearby University of Maryland, Baltimore County. Figure 3.2 shows an older Victorian house under renovation. The suburb also experienced a second housing boom during the late 1940s, and smaller tract developments are scattered throughout Catonsville. Situated due north of Baltimore, Hampton was the newest and smallest first-tier suburb. In 1929, the Hampton Development Corporation founded the suburb, yet very few houses were initially built (Brooks 1979). Hampton remained largely rural until the 1970s when five neighborhoods were subsequently built to surround the Historic Hampton Mansion,

Figure 3.2

Victorian house in Catonsville, 2005.

Source: Original photograph by author.

suburban decline / 95

Figure 3.3

Ranch house in Hampton, 2005.

Source: Original photograph by author.

a landmark estate built in 1790. In total, there were 636 detached, single-family housing units built. Figure 3.3 illustrates a typical house in Hampton, which resembled a one-story ranch style house. They had amenities such as large rooms, multiple-car garages, and generous yard space. Hampton was unlike any other first-tier suburb as it shared many similar characteristics with the outer suburbs. Lansdowne stood in stark contrast to both Catonsville and Hampton. An old historic iron ore-mining town during the nineteenth century, Lansdowne was located on the Baltimore and Ohio “B & O” railroad line. It was not built until two large neighborhoods, Baltimore Highlands and Riverview, were developed in the early 1950s in a “cookie-cutter” development pattern. The houses were built quite small, and most of the units are less than 1,000 square feet. Many houses are attached rowhouses with one bedroom, a tiny yard, and front porch. The style was typically bland with nondescript brick features. As figure 3.4 shows, the roofs were flat, and houses lacked green space. More than 50 years old now, the housing stock had fallen into disrepair and exhibited many symptoms of utter decay. On the eastern suburban fringe, Essex’s housing stock shared many of the characteristics of Lansdowne’s stock. The Taylor Land Company

96 / transforming race and class in baltimore

Figure 3.4

Tract rowhouse in Lansdowne, 2004.

Source: Original photograph by author.

began building in the suburb in 1909, and development soared until the start of the World War II (Brooks 1979). The developer constructed rowhouses in a similar style to the Baltimore City rowhouse (Hayward and Belfoure 2001). Essex’s housing stock was roomier with ample garden space as figure 3.5 demonstrates. In addition, these houses featured two bathrooms, something that was unheard of during the 1920s and 1930s. Housing continued to grow in Essex as the nearby Glen L. Martin Company employed some 50,000 residents during the 1940s and 1950s. By 1960, Essex was fully built. By the beginning of the twenty-first century, it showed significant signs of age. Dundalk, just east of the Port of Baltimore, had been the most populated first-tier suburb since World War II. It was a classic “industrial suburb,” home of the former Bethlehem Steel Company. Founded in 1917, Bethlehem Steel commissioned the construction of some 10,000 housing units during the 1920s to provide a residence in close proximity to its factory (Reutter 2004). Dundalk was designed in a similar fashion to Baltimore City’s exclusive neighborhood of Roland Park. It featured parks, dense housing, and walkable streets to build a sense of community

suburban decline / 97

Figure 3.5

Rowhouse in Essex, 2005.

Source: Original photograph by author.

(Brooks 1979). Figure 3.6 illustrates an example of the typical multiplex housing style in Dundalk. These houses were smaller and older, yet they were relatively well-maintained as compared to neighboring industrial Essex or Lansdowne. In summary, metropolitan Baltimore’s housing stock was transformed in a number of important ways since 1970. The oldest and smallest housing stock was located in the urban core. In the suburbs, older and smaller housing was located in close proximity to the urban core. With few exceptions, newer and larger housing was reserved exclusively for the outer suburbs. Similarly, the value of housing stock increased as the distance increased from the urban core. The first-tier suburban housing stock lost value as the suburbs as a whole gained value. The housing stock was also comprised of a diversity of style. Housing units in the first-tier suburbs were small and large, attached and detached, and old and young. The housing stock reflected the style of the day, when the first-tier suburbs were built. Whereas homeowners were commonplace a generation ago, renters and vacant units grew increasingly widespread in the suburbs. The housing stock in the firsttier suburbs thus grew more diverse and more complex over time.

98 / transforming race and class in baltimore

Figure 3.6

Multiplex housing in Dundalk, 2004.

Source: Original photograph by author.

Labor Force Structure Employment sustains life and provides social and economic opportunities for residents of metropolitan areas. Jobs and education are means of this pursuit as they provide for basic foundations of well-being. Three important components of the labor force structure uncover details about who works. First, I analyze the changes in the composition of industries in the Baltimore region over three decades. Second, I present evidence about how the levels of educational attainment for suburban residents changed during this period. Third, I show the variation in the unemployment rate during the three past decades in Baltimore’s suburbs. The economy of metropolitan Baltimore radically transformed since 1970. Typical of a Rustbelt city, a considerable portion of residents in the region held manufacturing jobs in 1970 (Sassen 1990). Yet, with each decade since then, the number of manufacturing jobs declined remarkably. At the same time, employment in the service sector increased continuously between 1970 and 2000. There was evidence of this transformation throughout suburban Baltimore. However, in many cases these experiences of first-tier suburbs varied significantly. Thus, it is useful to examine the change in

suburban decline / 99

the composition of the industry in metropolitan Baltimore. The industrial composition refers to the industry that a resident of a place worked in during census year. It does not measure where the jobs are located; rather, it measures where the resident lives and who works in a particular industry. In 1970, a considerable portion of residents in metropolitan Baltimore held jobs in the manufacturing sector. Table 3.15 demonstrates that approximately one-third of the labor force in the first-tier suburbs and Baltimore City was employed in the manufacturing industry in 1970. In the outer suburbs, the portion of residents employed in the manufacturing industry was considerably less; 18 percent of residents held manufacturing jobs in 1970. Over the next three decades, the entire region witnessed sharp declines in the number of residents employed in the manufacturing sector. By 2000, only 7 percent of central city residents and 8 percent of outer suburban residents held manufacturing jobs. Combined, the central city and the first-tier suburbs lost more than 100,000 manufacturing jobs during this period. In the first-tier suburbs, 11 percent of the labor force worked in the manufacturing industry, which was the highest in the region in 2000. Further examination of the loss of manufacturing jobs in the first-tier suburbs reveals several important trends. Baltimore’s first-tier suburbs experienced a 200 percent loss in the number of residents who worked in the manufacturing industry—amounting to some 32,000 jobs over 30 years. Table 3.15 shows again that there were sharp declines over 30 years in the number of residents who held manufacturing jobs for every first-tier suburb. In 1970, approximately one-third of residents in the first-tier suburbs were employed in the manufacturing industry. By 2000, no first-tier suburb had more than 20 percent of its residents employed in manufacturing. In several suburbs, the losses were substantial. For example, nearly half of the population in the eastern first-tier suburbs of Dundalk and Edgemere were employed in manufacturing in 1970, and by 2000, under one-fifth of the population held manufacturing jobs. Although these two places maintained the largest manufacturing population in 2000, there was a threefold decrease in the number of residents with manufacturing jobs since 1970. In the midst of the losses in the manufacturing industry, employment in service industries became increasingly more prevalent throughout metropolitan Baltimore. Table 3.16 shows that there was dramatic growth in the service job sector in the outer suburbs. In 1970, approximately one in five jobs were in service sector, which accounted for some 70,000 jobs; 3 decades later, nearly 200,000 jobs, or approximately half of all workers, were employed in the service sector in the outer suburbs. The first-tier suburbs witnessed similar growth, albeit at a slower pace. Every first-tier

Table 3.15

Manufacturing employment in metropolitan Baltimore, 1970–2000 1970

Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs Outer Suburbs Baltimore City Source: U.S. census.

Percent

Number

29 33 19 48 56 39 29 25 18 33 27 18 23 43 30 26 11 30 36 19 18 32 18 28

2,108 1,561 2,326 12,673 1,798 6,260 1,831 3,881 451 2,128 871 2,124 1,599 4,484 1,382 3,382 1,132 752 2,742 2,966 2,251 58,702 41,779 101,126

1980 Percent 19 27 14 34 44 30 25 20 19 27 21 14 16 32 18 17 9 27 26 11 13 25 15 16

1990

2000

Change

Number

Percent

Number

Percent

Number

Percent

Number

1,916 1,487 2,215 11,117 1,605 5,743 1,761 3,696 454 2,027 822 2,042 1,494 4,040 1,171 3,103 1,110 737 2,493 2,746 2,144 53,923 40,562 69,516

13 17 10 21 27 19 15 14 10 15 16 11 9 22 13 10 7 15 17 8 10 14 12 10

1,405 903 1,696 6,592 1,171 3,811 1,307 2,829 244 1,142 638 1,553 791 2,633 889 1,625 930 438 1,652 1,898 1,993 36,140 46,021 43,408

11 10 8 17 20 14 10 9 6 14 10 9 10 16 8 9 7 12 13 6 8 11 8 7

1,152 494 1,522 4,720 873 2,529 807 1,856 147 980 376 1,090 753 1,822 525 1,405 1,028 286 1,205 1,436 1,472 26,478 43,206 27,605

218 223 211 231 236 225 219 216 212 219 217 29 213 227 222 217 24 218 223 213 210 221 210 221

2956 21,067 2804 27,953 2925 23,731 21,024 22,025 2304 21,148 2495 21,034 2846 22,662 2857 21,977 2104 2466 21,537 21,530 2779 232,224 1,427 273,521

Table 3.16

Service employment in metropolitan Baltimore, 1970–2000 1970

Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn First-Tier Suburbs Outer Suburbs Baltimore City Source: U.S. census.

1980

1990

2000

Change

Percent

Number

Percent

Number

Percent

Number

Percent

Number

Percent

21 14 30 12 11 13 14 16 35 10 18 27 26 15 23 23 36 19 18 36 23 21 22 26

1,839 1,428 2,149 10,450 1,509 5,341 1,691 3,511 440 1,885 789 1,960 1,389 3,878 1,124 2,948 932 715 2,368 2,636 2,058 51,042 69,043 101,126

25 18 33 18 17 20 18 21 38 17 22 31 33 19 27 28 41 22 23 40 27 26 31 40

1,916 1,487 2,215 11,117 1,605 5,743 1,761 3,696 454 2,027 822 2,042 1,494 4,040 1,171 3,103 1,110 737 2,493 2,746 2,144 53,923 76,714 100,124

32 24 38 23 24 26 21 23 44 23 28 34 41 22 33 34 46 23 28 47 35 31 38 36

1,405 903 1,696 6,592 1,171 3,811 1,307 2,829 244 1,142 638 1,553 791 2,633 889 1,625 930 438 1,652 1,898 1,993 36,140 129,979 120,512

40 32 46 30 28 34 31 33 52 30 34 42 47 32 36 42 53 31 34 54 40 38 46 28

1,152 494 1,522 4,720 873 2,529 807 1,856 147 980 376 1,090 753 1,822 525 1,405 1,028 286 1,205 1,436 1,472 26,478 190,038 115,962

19 18 16 18 17 21 17 17 17 20 16 15 21 17 13 19 17 12 16 18 17 17 24 2

Number 2687 2934 2627 25,730 2636 22,812 2884 21,655 2293 2905 2413 2870 2636 22,056 2599 21,543 96 2429 21,163 21,200 2586 224,564 120,995 14,836

102 / transforming race and class in baltimore

suburb experienced growth in the number of residents with jobs in service industries. In 1970, services jobs comprised between 10 and 20 percent of the labor force. By 2000, the percentage of residents with service jobs was at least 30 percent in every first-tier suburb. In particular, service employment was more than 50 percent in three first-tier suburbs, including Hampton, Pikesville, and Towson. On the contrary, the four suburbs with the largest number of residents employed in the manufacturing sector, including Dundalk, Edgemere, Essex, and Middle River, had the lowest portion of residents employed in service industries, which did not exceed one-third of the labor force in any of these suburbs. Thus, there was a marked transformation in the metropolitan economy of greater Baltimore. The distinguishing characteristic of this change was the transition of the labor force employment in the manufacturing sector to the service sector. As employment rate declined manufacturing industries, it systematically rose in the service sector. Residents with services jobs overwhelmingly resided in outer suburbs while residents with manufacturing jobs resided in first-tier suburbs. Another useful benchmark of the labor force is to examine the education of workers. Over 30 years, the level of education of the labor force changed in several notable ways. Among all suburbs in metropolitan Baltimore, the prevalent trend between 1970 and 2000 was a steady increase in the number of years of education attained. Three trends became more pronounced in each decade. Yet, further analysis of educational attainment shows several important differences between first-tier and outer suburbs. These changes are analyzed by categorizing educational attainment based on the number of years attained. The definitions of the three categories are (1) 11 years or less represents high school dropouts; (2) 12 years represents high school graduates; and (3) 16 years or more represents college graduates. First, the number of high school dropouts fell steadily in suburban Baltimore. During the 1970s, the number of dropouts decreased significantly, and in subsequent decades, it fell steadily. In 1970, the outer suburbs had about a quarter fewer high school dropouts than the firsttier suburbs. Nearly half of residents in the first-tier suburbs had no high school diploma in 1970. By 2000, the first-tier suburbs had twice as many high school dropouts than the outer suburbs. Approximately one-fifth of residents in the first-tier suburbs and one-tenth of residents in the outer suburbs had 11 years or fewer of education. Second, the number of high school graduates declined steadily in the outer suburbs, and it remained stagnant in the first-tier suburbs. In the outer suburbs, the portion of the population with only a high school education declined from 33 to 20 percent between 1970 and 2000. In contrast, the

suburban decline / 103

portion of residents with only a high school education in the first-tier suburbs remained fixed at approximately one-third of the overall population. Third, the number of suburban residents with at least a four-year college education increased since 1970. Yet, there were large increases in the outer suburbs and small increases in the first-tier suburbs. In 1970, only a minority of suburban residents had a college education. Approximately 18 percent of outer suburban residents and 12 percent of first-tier suburban residents had at least 16 years of education. The number of residents with a college degree in the outer suburbs more than doubled to 45 percent, 30 years later. In contrast, the first-tier suburbs had less than a quarter of its population with 16 years or more of education by 2000. The labor force became more educated during the past three decades. However, those gains were uneven across the suburbs. Residents in the outer suburbs were more educated than residents in the first-tier suburbs were. Unemployment trends also offer some insight into the structure of the labor force. Since 1970, there has been substantial variation in unemployment throughout the region. In Baltimore City, unemployment doubled over the past three decades, reaching 12 percent in 2000. In suburban Baltimore, there were modest differences between the first-tier suburbs and outer suburbs. In 1970, the unemployment rate in both the first-tier suburbs and outer suburbs stood at 3 percent. Over three decades, unemployment in the outer suburbs remained low, hovering between 3 and 4 percent. The unemployment in the first-tier suburbs continued to grow each decade so that by 2000, the unemployment rate was 8 percent, double that of the outer suburbs. Among the first-tier suburbs, the trends were more diverse. The unemployment rate grew in every first-tier suburb over 30 years. For many suburbs, the unemployment rate peaked in 1980; the same year that stagnation hit the nation. The unemployment rate was moderately high in some suburbs and quite low in others. For example, Lansdowne had the highest unemployment rate in 2000, which approached 8 percent. Dundalk and Essex also approached 7 percent unemployment rates by 2000. In contrast, Hampton and Lutherville maintained very low unemployment rates since 1970, fluctuating between 2 and 3 percent. In summary, there were marked changes in the structure of the labor force in metropolitan Baltimore. They were related to the composition of industries, the level of educational attainment, and the unemployment rate. Between 1970 and 2000, two trends were prevalent in the composition of industries. First, the number of residents with manufacturing jobs declined in the entire region. First-tier suburbs lost some 32,000 jobs alone. Second, employment in service industries increased, yet those increases varied among first-tier suburbs. Despite the overall gains in the education levels for suburban Baltimore, the growth was largely uneven from 1970

104 / transforming race and class in baltimore

to 2000. The population of the outer suburbs grew increasingly more educated than the first-tier suburbs. Slower gains occurred in the first-tier suburbs. The outer suburbs had twice as many college graduates than the first-tier. Conversely, the first-tier suburbs had twice as many residents with only a high education or less. Unemployment in the first-tier suburbs grew over 30 years, and it remained low for the outer suburbs. By 2000, the unemployment rate in the first-tier suburbs was nearly double the rate in the outer suburbs of metropolitan Baltimore. Summary Evidence of suburban decline abounds in each of Baltimore’s 21 first-tier suburbs. They experienced decline in four principal areas that were related to the (1) population characteristics; (2) income dynamics; (3) nature of the housing stock; and (4) labor market structure. The decline of a suburban community occurred when there were marked changes in these four main areas of focus during the 30-year period of study. These measures of decline were all based on relative spatial comparisons to the central city, the outer suburbs, and the metropolitan area. First, population growth of first-tier suburbs stagnated since 1970. They failed to attract additional residents, and when they did, the growth was small relative to the growth in the outer suburbs. Furthermore, many of these communities could not only maintain their population base, they actually lost residents. The first-tier suburbs also aged significantly between 1970 and 2000 as compared to the outer suburbs. In fact, they aged without a successive generation to replace older residents. Baltimore’s suburbs were highly segregated by race and place. Although the suburbs remained overwhelmingly white, the first-tier suburbs were slightly more diverse than the outer suburbs. There were some increases of racial diversity, and it was distributed in an uneven spatial pattern— enclaves of black and white suburbs were commonplace. One particular trend that occurred in the western first-tier suburbs stood out. Between 1970 and 2000, there was a near total resegregation of these places. In 1970, they were all majority white suburbs. By 2000, the very same places had become majority black suburbs. The invasion of a black population characterized the racial changes in these first-tier suburbs. The first-tier suburbs also experienced changes in the composition of families over the past three decades. The family structure shifted away from married families with children to female-headed households. The first-tier suburbs had more nontraditional families than the outer suburbs. The rise of nontraditional families became a distinguishing characteristic of the changes in first-tier suburbs.

suburban decline / 105

Second, first-tier suburbs grew increasingly poorer over 30 years. The gap in poverty status between the outer suburbs and the first-tier suburbs continued to widen, especially during the 1990s. Only a few suburbs had virtually no poverty, and the poverty population grew in many of the first-tier suburbs. There was substantial variation in the poverty status among Baltimore’s first-tier suburbs. Several of the poorest suburbs contained triple the amount of poverty as others. A handful of suburbs were very poor, and a few were affluent. The eastern suburbs of Dundalk, Essex, Lansdowne, and Middle River stood out as the poorest suburbs. In contrast, Hampton, Linthicum, and Lutherville were wealthy suburbs in 1970 and 2000, yet they lost some of their wealth compared to all other suburbs. The majority of Baltimore’s first-tier suburbs had lower incomes in 2000 than in 1970, after adjusting for inflation. These suburbs also had lower income ratios in 2000. This indicated that these places fell far behind their suburban counterparts in the rest of the metropolitan area. The emergence of an increasingly poor population defined these suburban changes. Third, the first-tier suburbs were fully built by 2000. Most of the housing stock was considerably older than the outer suburbs. First-tier suburbs contained the majority of the smallest housing units in the suburbs. The overall trend was that newer, larger houses were located further away from the urban core while older, smaller houses were located in the first-tier in close proximity to the urban core. Housing units in the first-tier were the least valuable houses in the suburbs. By any measure, the housing stock in Baltimore’s first-tier suburbs was worth considerably less than housing in the outer suburbs of Baltimore. The trend worsened over time, and on average, housing values plummeted each decade since 1970. The tradition of home ownership was common for most residents of first-tier suburbs. This held true for wealthy suburbs such as Hampton and Lutherville, as well as poorer suburbs such as Brooklyn Park and Dundalk. Moreover, the style of first-tier suburban housing stock was diverse. The initial postwar housing boom produced a series of rowhouses in industrial areas. Other suburbs had large houses on sizeable plots of land while others had much smaller footprints. Fourth, the structure of the labor force transitioned in a marked fashion in the first-tier suburbs over 30 years. The deindustrialization of the U.S. economy transformed Baltimore and its suburbs. Baltimore City, and particularly the first-tier suburbs, lost their industrial might over three decades. Manufacturing jobs became less prevalent as the employment in service industries grew. During this period, the education levels of the labor force changed. The population of the outer suburbs grew increasingly more educated, and the first-tier suburbs had

106 / transforming race and class in baltimore

fewer residents with higher levels of education. Unemployment in the first-tier continued to grow each decade so that by 2000, the unemployment rate was double that of the outer suburbs. The shift away from an industrial labor economy to a service-oriented economy characterized this change. The first-tier suburbs are not what they used to be. The wave of socioeconomic decline crashed in the suburbs like a tsunami. In some suburban communities such as Lansdowne, it has eroded away the population, infrastructure, and the economic well-being of the residents. Still in other suburban places such as Dundalk, it has washed away the economic successes of an entire generation from the first-tier suburbs to the outer suburbs. In all, the data showed a pattern of decline on a number of measures between 1970 and 2000. These patterns of decline related to changes in the characteristics of the population, the income dynamics, the nature of housing stock, and the labor structure of the first-tier suburbs. Yet, in other suburbs such as Catonsville and Hampton, they remained relatively insulated from the ills of suburban decline. These suburbs attracted national attention for their suburban legacies and charm for the very reason that they stood out as anomalies in a declining first-tier suburban landscape. In the end, first-tier suburbs could not escape their apparent fate by the end of the twentieth century from the urban decay that afflicted the nation’s central cities during the 1960s and 1970s. It now encroaches suburbs.

Ch a p t e r Fou r Su bu r ba n Mo sa ic

Over a century ago, Baltimore’s German-bred National Brewing Company opened its doors in 1885 in the Brewers Hill and Canton neighborhoods as one of the city’s premier beermakers. The brewery capitalized on the demand for alcohol before the Prohibition years, and after World War II, the company developed its famed “National Bohemian” brand of beer. The drink would become known locally as “Natty Boh,” appealing to Baltimoreans’ charm and sense of community. A successful advertising campaign helped launch the beer’s success throughout the Mid-Atlantic States. “Mr. Boh” became the mascot, a one-eyed handlebar-mustachioed beer animated figure on television that Baltimoreans came to love. Mr. Boh regularly appeared in commercials and in the newspaper to promote the beer. His famed slogan was “From the Land of Pleasant Living,” a culture reference made to appeal to the great life on the Chesapeake Bay. Such a tasteful beer, coupled with an enticing advertising campaign, ensured that Natty Boh would remain the beer of choice for Baltimoreans for most of the twentieth century. National Brewing Company provided thousands of jobs for residents of eastern Baltimore during early-to-mid twentieth century. It was a local engine for the economy of the neighborhoods in this part of the city—just as Bethlehem Steel was for Dundalk. Yet, as the national economy began to restructure in the 1970s, so did the beer brewing industry. National Brewing Company closed its facility in Baltimore City in 1978 and moved to first-tier suburban Arbutus in the neighborhood of Halethorpe. The company brewed its beer until 2000 when it could no longer compete with the diversity of beer companies. Natty Boh and its company was sold to Pabst Brewing Company and moved to Pennsylvania. This illustration highlights that even a local institution as strong as Natty Boh could not withstand the forces of diversity. A century ago, there was little diversity of beer, and few companies competed in the marketplace. As technology improved and access to the market increased, the selection and choice of beer increased dramatically. By the end of the twentieth century, there was truly a mosaic of beer choices. In fact, there

108 / transforming race and class in baltimore

were more than 10,000 beer selections. Similarly, suburbs have witnessed a parallel metamorphosis. As postwar suburbanization exploded after 1945, the population continued to grow, and by the 1970s, a remarkable diversification in the suburbs began to occur in their neighborhoods. Baltimore’s first-tier suburbs flourished during the 1950s and 1960s as they grew up. Yet during the 1970s and 1980s, these suburbs were in a state of flux as the early stages of large-scale social and economic change materialized. Without a doubt, Baltimore’s 21 first-tier suburban communities that surround Baltimore City experienced decline relative to the outer suburbs since 1970. In some cases, suburbs such as Dundalk, Essex, Lansdowne, and Middle River actually declined substantially during the period from 1970 to 2000. In other cases, suburbs such as Hampton, Linthicum, and Lutherville declined from their status in 1970, but their decline was small compared to other first-tier suburbs. How do the experiences of these first-tier suburbs vary so much? What are the characteristics of suburbs in decline, and how do they vary over time? The answers can be found in the neighborhoods of these suburbs. For this part of the study, it is important to hone the analysis to a finer geographic scale to answer these questions. The 21 first-tier suburban communities of metropolitan Baltimore are comprised of 152 census tracts, which can be thought of as the neighborhoods of these suburbs. By focusing on these smaller areas within the first-tier suburbs, it is possible to explore the relationship between neighborhood change and suburban transformation. Moreover, this analysis provides the opportunity to detect changes to and variations within the suburban spatial structure of metropolitan Baltimore over a period of 30 years. In short, this approach offers a novel opportunity to examine—with a finer grain analysis—the process of neighborhood change in the first-tier suburbs. In this chapter, I present the results of two multivariate statistical analyses on first-tier suburban neighborhoods. First, I analyze how patterns of suburban structure evolved between 1970 and 2000 through a principal components analysis of socioeconomic data at the census tract level. Second, I use a cluster analysis technique to classify first-tier suburban neighborhoods in 1970 and 2000. Then, I present typologies of first-tier suburban neighborhoods, giving particular attention to the variety of characteristics of suburbs in decline. I conclude with a discussion on four main suburban transformations, in which I synthesize the results of this chapter. The New Suburban Landscape Two detailed analyses of the suburban landscape show that suburbs were a mosaic of many different types of suburbs—they were not a monolithic unit in 1970, let alone in 2000. First-tier suburbs could be distinguished

suburban mosaic / 109

primarily by class structure three decades ago. “Professional class households” and “fragile family, renter households” were commonplace among first-tier suburban neighborhoods in metropolitan Baltimore. Less common, though present during the 1970s, included “poor, black households,” “new, suburban family households,” “older, white households,” and “young, single households.” Socioeconomic diversity was clearly at hand with the onset of the 1970s. By 2000, the suburb mosaic was fully formed and crystallized. Class structure not only defined the suburbs, but race, age, and housing also stratified neighborhoods in the first-tier suburbs. “Poor households” and “minority, renter households” were by far the most common suburban neighborhood structure by 2000. “Black middle-class households” and “older, white households” were also prevalent throughout suburban neighborhoods, suggesting that the transformation of race was a large factor in the suburban mosaic. Finally, “younger, service worker households” and “older house, no growth households” also defined suburbs. Without a doubt, the new suburban landscape was transformed through a marked socioeconomic structure of class, race, age, housing, and labor. These forces created a mosaic of suburban neighborhoods. In the sections that follow, I review the results of two principal components analyses (PCA), one conducted for suburban neighborhoods in 1970, and the other in 2000. The PCA method detects the relationship between the socioeconomic data and their spatial location. This is commonly referred to as suburban spatial structure (Beauregard 1989). PCA also detects patterns in that structure and reports the results as scores for each suburban neighborhood. For the analyses in 1970 and 2000, I characterize the patterns of suburban spatial structure in Baltimore’s first-tier suburbs using this statistical geographical methodology. Patterns of Suburban Spatial Structure in 1970 The 1970 PCA generated six principal components. Table 4.1 lists these six components and provides definitions for them. I developed these Table 4.1

Summary of principal components analysis, 1970

Component and definition

Extraction sums of squared loadings Eigenvalue

1 2 3 4 5 6

Professional class households Fragile family, renter households Poor, black households New, suburban family households Older, white households Young, single households

13.45 5.83 3.70 2.82 2.45 2.08

% of Variance Cumulative % 30.56 13.26 8.40 6.40 5.57 4.74

30.56 43.82 52.22 58.62 64.19 68.93

110 / transforming race and class in baltimore

definitions based on the interpretation of the results for the 1970 PCA. In this section, I explain the procedure for interpreting the results, and then I define the main characteristics of each component. First, it was important to determine how many components to extract from the dataset. Since one of the goals of PCA was to reduce the number of variables into a smaller, more manageable set of components, it was necessary to determine how many components to use for the smaller set. One way to determine how many components to extract was to analyze the eigenvalues for each component on a graph. Figure 4.1 plots the eigenvalue for each component for the 1970 PCA. The plot illustrates that there were six components with an eigenvalue more than two. In general, eigenvalues of two or more could be interpreted as having at least twice the explanatory power as the original set of variables, and so they were a meaningful representation of the larger dataset (Kline 1994). Therefore, the first six components were extracted for the 1970 PCA since each of their eigenvalues was more than two. Second, it was necessary to identify the explanatory power of the six components to determine whether they explained a large extent of the 16 14 12

Eigenvalue

10 8 6 4 2 0 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 Component 1970

Figure 4.1 and 2000.

2000

Scree plot of eigenvalues for principal components analysis, 1970

suburban mosaic / 111

original dataset. Table 4.1 shows that the 1970 PCA had substantial explanatory power. The six components that were extracted explained more than two-thirds (68.93 percent) of the variation from the original dataset. Component 1 explained approximately one-third of the total variation, and component 2 explained 13 percent. The last 4 components each explained less than 10 percent of the total variation. Thus, the data for 50 variables on 152 cases (census tracts) could be largely explained by only 6 components. Third, it was important to determine whether the variables selected for this analysis were relevant for the study of suburban change. The PCA produced a communality value to ascertain the relevance of the variables selected. The communality is a numerical estimation of the variance and strength of each variable that is reduced into a set of fewer components. Each communality value indicated how well the variable correlated with the six extracted components. For the 1970 PCA, the communality values ranged from 0.46 to 0.96. The communality values could be interpreted as percents, which were measures of correlation. Together, the 50 variables were, on average, 80 percent correlated with the 6 extracted components. More than two-thirds of the variables (36 of 50) had values more than 0.70. This value was a common threshold used in PCA to demonstrate a strong correlation among the variables (Wyly 1999). What this meant was that the variables selected for this study shared a common variance with one another. Simply put, this reinforced the validity of the analysis. Fourth, it was necessary to analyze the loadings to interpret the meaning of each extracted component. The loadings are measures of the degree to which each variable in the dataset contributed to the meaning of each new component that the PCA produced. In other words, the loadings were used to interpret the characteristics of each component. Table 4.2 shows the loadings for the 1970 PCA. The table only displays the loading values greater than 0.30 and less than 20.30. Loadings greater than and less than these values, respectively, did not significantly contribute to the meaning of the component. To demonstrate the interpretation, consider the household income loading on component one. The loading value was 0.91. Loadings with a high or positive value, such as this example, suggest that income was a defining characteristic of that component. Conversely, loadings with low or negative values suggest that wealth, in this example, was not a defining characteristic of that component. Similar to this example, all the loadings for each of the six components were analyzed to develop definitions for them. Fifth, the PCA scores were used to display the spatial patterns in the data. The scores were used to create an index of positive and negative values that measured the degree to which each neighborhood related to the

Table 4.2

Component loadings for suburban neighborhoods, 1970

Variable description Population size Population ages 18–24 Youth population (17 and under) Population ages 18–64 Population ages 65 and over White, non-Hispanic population Black, non-Hispanic population Hispanic (any) population Foreign born population Married families with children households Single, never married households Female-headed households Divorced families households Poverty population Average household income Average household income ratio to all suburbs Average family income Average family income ratio to all suburbs High school drop-out (11 or fewer years) High school graduate (12 years) Some college (13–15 years) 4-year college graduate or higher (16 years) One or no bedrooms in housing unit Two bedrooms in housing unit

Component one

Component two

Component three

20.69 0.57 0.6

Component four 0.45

Component five

Component six

20.37 0.61 0.53

20.62 0.62

20.58

0.39 0.63 20.63

0.33 0.52 0.33 0.36 0.56

0.54

0.39 0.91 0.91 0.92 0.92

0.8 0.92

0.34 0.67

20.52

Three bedrooms in housing unit Four bedrooms in housing unit Housing units built in 1960s Housing units built in 1950s Housing units built in 1940s Housing units built in 1939 or earlier Average value of housing units Renter-occupied housing units Owner-occupied housing units Vacant housing units Professional and technical occupations Executives, managers, and administrators occupations Sales occupations Administrative support and clerical occupations Production, craft, and repair workers occupations Operators, assemblers, transportation occupations Nonfarm laborers occupations Service workers occupations Farm, fishing, forestry occupations Unemployed population

0.68 0.53 0.33 0.61

0.44 0.37

0.9 0.82 20.82 0.41

0.31 20.31 0.31

0.86 0.91 0.75 0.32 20.81 20.82 20.6

0.46 0.44 0.34 0.34

114 / transforming race and class in baltimore

components that the PCA produced. In other words, the scores told how well each neighborhood represented the meaning of each of the six components that the 1970 PCA produced. A neighborhood with a high score meant that the component was the defining characteristic of that area; lower scores meant that the component was not a defining characteristic. Let us now consider the patterns of suburban spatial structure in 1970. Component one detects a strong presence of what I term “professionalclass households.” It was the strongest measure of suburban structure in 1970, which explained nearly one-third of the variation of the original dataset. Table 4.2 shows that the loadings were the highest for this component. Above all other variables, they indicated that high socioeconomic status defined household structure in first-tier suburbs in 1970. High levels of household and family income dominated the trend. In addition, the loadings on educational attainment were very high. Two housing characteristics also stood out. The loadings showed that large housing units and high housing values were defining traits. The labor force composition also helped to explain the high socioeconomic status. The loadings for the first component indicated that there was a clear inverse relationship between white-collar and blue-collar occupations. The loadings were very high for professional, technical, managerial, and executive administrative occupations. Conversely, the loadings showed high negative values for production and labor-intensive occupations. The spatial trend for component one is very clear. High scoring neighborhoods are located to the north and west of Baltimore City. These neighborhoods were professional-class households, and they scored high on this component. In contrast, low scoring neighborhoods are located to the east and south of Baltimore City, and they scored very low for component one. Component one did not define these neighborhoods. The existence of a strong and weak professional-class was thus particularly apparent. For instance, the northern suburban Baltimore neighborhoods in Pikesville, Towson, Lutherville, and Hampton represented the location of the strongest professional-class households. The neighborhoods in the eastern suburbs of Dundalk, Essex, Middle River, and to the west in Landsdowne, scored the lowest on the professional-class households component. There was a clear demarcation between high and low socioeconomic status neighborhoods in the first-tier suburbs in 1970. Component two captures clusters of what I define as “fragile family renter households.” Accounting for 13 percent of the variation in the original dataset, this component strongly related to housing tenure and family structure. Table 4.2 shows that the loadings for renter-occupied housing units with one bedroom were the dominant characteristics. Housing vacancy was also prevalent, which explained the turnover in

suburban mosaic / 115

rental housing units. The loading was 0.82 for rentership and 20.82 for ownership, suggesting that renter households are common and homeowner households were uncommon. In addition, households were primarily comprised of divorced, female-headed families, and singles. The labor force loadings showed that these households tend to work in administrative support and clerical occupations. In spatial terms of component two, there were small clusters of fragile family renter households throughout first-tier suburban Baltimore. High scoring neighborhoods contained fragile family renter households whereas low scoring neighborhoods did not. The largest group of fragile family neighborhoods was located directly north of Baltimore City in suburban Towson. Small clusters of these neighborhoods were also evident along the western suburban fringe. The southern and eastern suburban neighborhoods tended to not have fragile family renter households. Component three reveals the emergence of what I term “poor black households.” This component represented approximately 9 percent of the variation in the original dataset. Table 4.2 shows that race was the highest loading value. It was 0.62 for black and 20.62 for white. The presence of a black population was one of the defining characteristics of component three. The third component was also the only component in the 1970 PCA in which the poverty loading value was significant, at 0.39. The loadings also showed that the presence of older housing units built before1940s was a defining characteristic. Service and labor occupations also loaded high. A brief look at the spatial distribution of component three shows a mosaic of poor black households among most of the first-tier suburbs. High scoring neighborhoods that contained poor black households were concentrated in the eastern and southwestern neighborhoods of Dundalk, Essex, Middle River, and Lansdowne. In addition, there were several northern suburban neighborhoods in Towson that scored even higher on this component for poor black households, indicating that were was a significant concentration of poor black residents in these neighborhoods. In contrast, low scoring neighborhoods in the western suburban fringe contained relatively few poor black households. There was little evidence that poor black households were located in the western suburban neighborhoods such as Woodlawn and Lochearn in 1970. Component four provides evidence of what I define as “new suburban family households.” The fourth component accounted for approximately 7 percent of the variation in the original dataset. Table 4.2 shows that the fourth component loaded high on variables relating to the age of the population, housing age, and housing tenure. The variable for families with young children loaded high while the variable for households aged more than 65 loads very low. Also, new housing loaded the highest, at 0.53.

116 / transforming race and class in baltimore

There was also evidence that residents primarily rented new housing units. This indicated that new housing units were built in the 1960s, and young families with children resided in these neighborhoods in 1970. In spatial terms, the largest concentration of new suburban family tracts was located due north of Baltimore City in Pikesville, and to a lesser degree, east of Baltimore City in Middle River and Essex. These neighborhoods scored high on component four and contained many new suburban family households. At the northwestern part of the central city, Pikesville and Lochearn stood out as the suburbs that scored the highest on new suburban family households. Also, the northernmost and easternmost parts of the first-tier suburbs contained new residents. In contrast, low scoring tracts did not contain any new suburban family households; as such, component four did not define these neighborhoods. For instance, the southwestern and northeastern corners around Baltimore City did not contain new suburban family households. Similarly, the western suburbs of Catonsville and Arbutus, as well as the old core of suburban Towson to the north, scored the lowest on this component. Component five detects what I define as “older, white households.” This component accounted for 6 percent of the variation in the original dataset. Table 4.2 shows that the age and race of the population and age of housing were the only characteristics that had significant loadings on the fifth component. In terms of population age, the young children variable loaded very low at 20.37 while the persons aged more than 65 loaded high, at 0.39. Similarly, the white population loaded at 0.63, and the black population loaded at 20.63. This demonstrates that the presence of a white population and the lack of a black population defined component five. Last, the age of housing loaded high for housing units built before 1950. Together, these loadings suggest that an elderly, white population resided in older housing units in these neighborhoods in 1970. Older, white households were distributed throughout the majority of first-tier suburban neighborhoods in metropolitan Baltimore. High scoring neighborhoods on component five contained older, white households. The largest spatial concentration of these neighborhoods was located on the eastern suburban fringe, particularly in Dundalk, Parkville, Overlea, and Rosedale. There were also pockets of these neighborhoods in the northern suburbs around Towson, and in southwestern suburbs around Catonsville and Arbutus. Low scoring neighborhoods contained few older, white households. In particular, there were three pockets of neighborhoods without any such households. They could be found at the very northern and southern tips of the suburbs, as well as the suburbs due west of the city. Component six captures what I term “young, single households.” It accounted for just less than 5 percent of the variation in the original

suburban mosaic / 117

dataset. Table 4.2 shows that the age of population and family status were the only two variables with significant loadings. Both the youth population variable (ages 18–24) and the workforce population variable (ages 16–64) loaded very high. In addition, the never married variable loaded high, at 0.54. Together these variables suggest that there was an emerging trend of young, single household developing in the first-tier suburbs in 1970. Neighborhoods that contained young, single households were generally limited to four suburbs in 1970: Essex, Pikesville, Towson, and Catonsville. Overall, high scoring neighborhoods on component six were located on the outermost fringe of the first-tier suburbs. In each of these four suburbs, there was a large university student population in 1970 at University of Maryland, Baltimore County (UMBC), Towson University, and the Community College of Baltimore County. Low scoring neighborhoods did not contain young, single households, and they comprised the majority of suburban neighborhoods. Overall, Baltimore’s first-tier suburbs were largely comprised of a diversity of neighborhoods in 1970. Class boundaries separated first-tier suburban neighborhoods. The context of place in determining the socioeconomic status of a suburb was important three decades ago. The northern first-tier suburbs were places of affluence, which distinguished these areas from others. The eastern and western first-tier suburbs housed a solidly middle-class, white population, with some exceptions. Specifically, three characteristics differentiated Baltimore’s first-tier suburban landscape. First, class was a distinguishing feature of Baltimore’s first-tier suburbs in 1970. A clear spatial pattern separated the professional-class households from the working-class households. Northern and western first-tier suburbs stood out as the most socially and economically advantaged areas around the Baltimore region in 1970. A working-class population worked in the manufacturing economy in the eastern and southern suburbs of Baltimore. Second, the age of housing in first-tier suburbs varied largely; there were new and old housing units. Newer first-tier suburbs housed new, young families and provided improved housing conditions. Older first-tier suburbs housed an elderly population. The newer areas were located in the northern and western suburbs. The older areas were located in the eastern and southern suburbs. In 1970, the overall spatial trend generally showed that the oldest and most established neighborhoods were located closer to Baltimore City while the younger and newly established neighborhoods were located further away from the urban core. Third, the population of the first-tier suburbs was somewhat heterogeneous in 1970. Small pockets of black residents were scattered throughout the first-tier suburbs in a few instances. Yet, in the eastern and western

118 / transforming race and class in baltimore

suburbs, the population was predominately white. It is important to highlight that a near-contiguous pattern of western suburban neighborhoods primarily housed a white population. This is the exact location, on the western fringe, that experienced substantial racial turnover and resegregation in subsequent decades. Finally, the seed of suburban decay had been planted by 1970. The emergence of fragile renter families, poor black households, and older white households laid the foundation for an environment that would be ripe for suburban change, particularly decline, in the future. Patterns of Suburban Spatial Structure in 2000 The 2000 PCA generated six principal components. Table 4.3 lists these six components and provides definitions for them. I developed these definitions based on the interpretation of the results for the 1970 PCA. In this section, I explain the procedure for interpreting the results, and then I define the main characteristics of each component. First, the eigenvalue for each component was plotted to determine how many components to extract from the analysis. Figure 4.1 shows that there were six components with an eigenvalue more than two. In general, eigenvalues of two or more could be interpreted as having at least twice the explanatory power as the original set of variables, and so they were a meaningful representation of the larger dataset (Kline 1994). Therefore, the first six components were extracted for the 2000 PCA since each of their eigenvalues was more than two. Second, it was necessary to identify the explanatory power of the six components to determine whether they explain a large extent of the original dataset. Table 4.3 shows that the 2000 PCA had substantial explanatory power. The six components that were extracted explained more than two-thirds (66.83 percent) of the variation from the original dataset. Table 4.3 also shows that component 1 explained 27 percent of the Table 4.3

Summary of principal components analysis, 2000

Component and Definition

Extraction sums of squared loadings Eigenvalue

1 2 3 4 5 6

Poor Households Minority, renter households Black middle-class households Older, white-households Younger, service worker households Older housing, no growth households

13.58 7.84 4.17 3.46 2.24 2.12

% of Variance Cumulative % 27.16 15.68 8.34 6.92 4.47 4.25

27.16 42.85 51.19 58.11 62.58 66.83

suburban mosaic / 119

total variation in the dataset, and component 2 explained 16 percent. The last 4 components each explained less than 10 percent of the total variation. Thus, the data for 50 variables on 152 cases (census tracts) could be largely explained by only 6 components. Third, it was important to determine whether the variables selected for this analysis were relevant for the study of suburban change. Each communality value indicated how well the variable correlates with the six extracted components. For the 2000 PCA, the communality values ranged from 0.30 to 0.96. The communality values could be interpreted as percents, which were measures of correlation. Together, the 50 variables were, on average, 81 percent correlated with the 6 extracted components. More than three-quarters of the variables (42 of 50) had values more than 0.70. This value was a common benchmark used in PCA to demonstrate a strong correlation among the variables (Wyly 1999). What this meant was that the variables selected for this study shared a common variance with one another. This reinforced the validity of the analysis. Fourth, it was necessary to analyze the loadings to interpret the meaning of each extracted component. Table 4.4 shows the loadings for the 2000 PCA. The loadings were measures of the degree to which each variable in the dataset contributed to the meaning of each new component that the PCA produced. In other words, the loadings were used to interpret the characteristics of each component. The table only displays the loading values greater than 0.30 and less than 20.30. Loadings greater than and less than these values, respectively, did not significantly contribute to the meaning of the component (Kline 1994). To demonstrate the interpretation, consider the household income loading on component one. The loading value was 20.89. Loadings with low or negative values, such as this example, suggest that income was not a defining characteristic of that component. Conversely, loadings with a high or positive value, suggest that the variable was a defining characteristic of that component. Similar to this example, all the loadings for each of the six components were analyzed to develop definitions for them. Fifth, the scores from the 2000 PCA were used to display the spatial patterns in the data. The scores were used to create an index of positive and negative values that measured the degree to which each neighborhood related to the components that the PCA produced. In other words, the scores told how well each neighborhood represented the meaning of each of the six components that the 2000 PCA produced. A neighborhood with a high score meant that the component was the defining characteristic of that area; lower scores suggest that the component was not a defining characteristic. Let us now consider the patterns of suburban spatial structure in 2000.

Table 4.4

Component loadings for suburban neighborhoods, 2000

Variable description Population size Population ages 18–24 Youth population (17 and under) Population ages 18–64 Population ages 65 and over White, non-Hispanic population Black, non-Hispanic population Hispanic (any) population Foreign born population Married families with children households Single, never married households Female-headed households Divorced families households Poverty population Average household income Average household income ratio to all suburbs Average family income Average family income ratio to all suburbs High school dropout (11 or fewer years) High school graduate (12 years) Some college (13–15 years) 4-year college graduate or higher (at least 16 years) One or no bedrooms in housing unit Two bedrooms in housing unit

Component Component Component Component Component Component one two three four five six 0.37 0.33 0.39

20.7 0.49 0.63 0.53 0.39 20.89 20.89 20.9 20.9 0.78 0.76 0.4 20.87

0.65 0.57 20.33 20.48 0.4 0.41 0.55 20.31 0.74

0.61

0.47 0.52

20.31 0.31

20.4

20.31 0.34

0.33 0.63 20.64 20.32

0.49 0.41 20.47

0.42 0.44

0.5 0.41

Three bedrooms in housing unit Four bedrooms in housing unit Housing units built in 1990s Housing units built in 1980s Housing units built in 1970s Housing units built in 1960s Housing units built in 1950s Housing units built in 1940s Housing units built in 1939 or earlier Average value of housing units Renter-occupied housing units Owner-occupied housing units Vacant housing units Professional and technical occupations Executives, managers, and administrators occupations Sales occupations Administrative support and clerical occupations Production, craft, and repair workers occupations Operators, assemblers, transportation occupations Nonfarm laborers occupations Service workers occupations Farm, fishing, forestry occupations Manufacturing occupations Public administration occupations Unemployed population

20.52

20.64

20.61

20.5 20.54 20.6

0.76 0.4 0.45 0.42 20.83 0.45 20.45 0.4 20.85 20.86

0.54 20.54

0.53 20.53 0.39

0.33

0.31 0.66 0.67 0.65 0.55 0.6

0.42

0.43 0.36 0.36

0.36

0.4

122 / transforming race and class in baltimore

Component one detects a strong presence of what I term “poor households.” It was the strongest measure of suburban structure in 2000, which explained 27 percent of the variation of the original dataset. Table 4.4 shows that there are many variables that loaded either high or low for the first component. Collectively, the loadings for these variables described the classic characteristics of the underclass (Jargowsky 1997; Wilson 1987). For instance, female-headed households and divorced families had high loadings, and married-parents families loaded very low, at 20.70. This suggests that the traditional married-parents family was not common in these neighborhoods. All variables for household and family income loaded very low, at 20.90.In addition, poverty loaded high. So, the income loadings showed that the population was comprised of poor residents. In terms of education, the loadings showed that households had low levels of educational attainment. The housing loadings demonstrated that units were small and did not have much market value. They also showed that the population resided in rental units. The loadings for the labor force variables demonstrate that residents did not hold professional class occupations. Instead, they held manufacturing and low-wage service occupations. Unemployment also loaded high, at 0.36. The first component captured the essence of the characteristics of poor households. Suburban neighborhoods with poor households follow a distinct spatial trend. High scoring neighborhoods contained many poor households. Located in the eastern suburbs, they were uniformly home to the largest number of poor neighborhoods. In particular, a group of ten neighborhoods in both Dundalk and Essex scored especially high. Also, the suburban neighborhoods directly south of the city contained a sizeable stock of poor households, especially around the Glen Burnie area. Directly west of the city, Woodlawn, Lansdowne, and Lochearn also contained a group of poor neighborhoods. In contrast, low scoring neighborhoods contained few poor households. Two areas in the first-tier suburbs were insulated from the poor neighborhoods. Due north of the city, Towson, Lutherville, Hampton, and Pikesville each contained virtually no poor households in their neighborhoods. Similarly, the outer fringe of the western suburbs in Catonsville and Linthicum did not contain poor households in their neighborhoods. It was apparent by 2000 that the spatial trend was the one that segregated nonpoor households from poor households in the first-tier suburbs. Component two captures what I define as “minority renter households.” This component accounted for 16 percent of the variation in the original dataset. Table 4.4 shows that the second component loaded highly on race, housing tenure and age, and poverty variables. The white population loaded very low at 20.48, and all other races and ethnicities loaded

suburban mosaic / 123

highly as well. For example, black, Hispanic, other races, and foreign-born populations all loaded more than 0.40. This suggested that there was a strong presence of minority households. In terms of housing, the loadings indicated that the population rented one-bedroom housing units that were primarily built during the 1970s. The poverty loading value was 0.61, making it the highest loading value on the poverty variable out of all 6 principal components in 2000. There was a clear spatial pattern for the second component. High scoring neighborhoods contained minority renter households. The majority of minority renter households were located in the northern and western suburbs. In particular, Arbutus, Woodlawn, Lochearn, and Towson each contained a substantial population of minority renter households. Conversely, the southern and eastern suburbs scored low on component two and did not contain a large minority renter population. These included the suburbs of Catonsville, Dundalk, Overlea, and Middle River. There were two exceptions. First, a cluster of neighborhoods in the eastern suburbs around Essex indicated the presence of a concentrated minority renter population. Second, another area of minority renters was located at the very southern tip of the first-tier suburbs around Glen Burnie. This reinforced the spatial trend of the first component of poor neighborhoods. Many of the poor neighborhoods were also the same places that housed the minority renter population. Component three reveals the emergence of what I define as “black middle-class households” in suburban Baltimore. This component accounted for 9 percent of the variation in the original dataset. Table 4.4 shows that the main variables with high loadings related to the population age and size, race, poverty, and occupation. The population size variable had a loading value of 0.37, and the child population variable loaded at 0.33. These two loadings indicated that these neighborhoods were growing. In terms of race, the black population loading value was high while the white population value was low. This implies that these neighborhoods had a considerable black population relative to other suburban neighborhoods. Also, the loading for the poverty variable was very low, at 20.47. This signifies that poverty was not widespread. In terms of the labor force, the loadings show that public sector employment was significant. Government employment for the residents in these neighborhoods was the distinguishing feature of the third component. The third component demonstrated the growth of a black middle class in first-tier suburban neighborhoods. The third component highlights several noteworthy spatial trends. High scoring neighborhoods contained black middle-class households. The most important trend was the strength of black middle-class neighborhoods on the western and northwestern fringe of suburban Baltimore.

124 / transforming race and class in baltimore

A contiguous area of 45 neighborhoods was a prominent feature of the landscape. These neighborhoods were exclusively located in Woodlawn, Lochearn, Pikesville, and Towson. There were also parts of suburban Baltimore that did not house the black middle class. These low scoring neighborhoods contained few households that were part of the black middle-class population. Along the eastern suburban fringe, the majority of suburban neighborhoods lacked a black middle-class residential population. Yet, there was a pocket of black middle-class households along the waterfront in the eastern first-tier suburb of Middle River. The suburbs to the south of Baltimore contained a diversity of neighborhoods, and the pattern varied significantly. By 2000, a new socioeconomic class emerged in the first-tier suburbs, and they largely segregated to the western suburban fringe. Component four identifies what I define as “older, white households.” It represented 7 percent of the variation in the original dataset. Table 4.4 shows that the loadings for the age and race of the population and housing size variables were significant. The race variables loaded the highest out of all variables for this component. The white population loaded very high, at 0.63. In contrast, the black population loaded very low, at 20.64. This suggests that the residents of these neighborhoods were mainly white. Also, the loading value for the population age variable was low for the child population and high for the elderly population. This implies that the population was aging without a younger generation. Last, the loadings showed that there were small houses in these neighborhoods. Specifically, one and two bedroom housing units were the most common. The fourth component provided evidence of patterns of aging in older, white residents among suburban neighborhoods in metropolitan Baltimore. A brief look at the spatial layout for the fourth component reveals some important trends. High scoring neighborhoods contained older, white households. The most distinguishable feature of the fourth component was the polarization of white neighborhoods. The older, white population was the defining characteristic of the eastern suburbs of Baltimore. Dundalk, Essex, and Middle River were the highest scoring suburbs on this component, which suggested that their neighborhoods had the highest degree of older, white households. On the other side of Baltimore City, the western suburbs, and to a lesser degree the southern suburbs, did not have an older, white population. Their neighborhoods scored low on component four, and they contained an older, white population. For instance, there was a large cluster of 18 neighborhoods due west of Baltimore that provided tangible evidence of this pattern. There were virtually no older, white residents in this area because this was the place that housed the black middle-class population. Otherwise, the northern central suburbs around

suburban mosaic / 125

Towson housed an older, white population with small pockets of a diverse population. Even though older, white households could be found throughout suburban Baltimore, the first-tier suburbs were still largely a mosaic of people in 2000. Component five detects what I define as “younger service worker households.” This component explained 5 percent of the variation in the original dataset. Table 4.4 shows that the fifth component had four significant loading values. First, the youth population loaded high, at 0.34. Second, the married-parents with children variable was very low, at 20.32. Third, the service occupation variable was the highest loading variable, at 0.42. Fourth, the unemployment loading value was high, at 0.40. Together, these variables showed that a portion of the younger population was employed in the services sector of the economy. The spatial pattern for the fifth component is a mosaic. High scoring tracts contained younger service worker households. Younger service workers tended to be located in neighborhoods that were further away from the central city. Woodlawn, Pikesville, and Middle River contained a service worker population that was considerably younger. Most of the neighborhoods that bordered Baltimore City did not have younger service workers. There were two large pockets of suburban neighborhoods that lacked younger service workers, which were located around the northeastern and southwestern corners of the city. This population was distributed through suburban Baltimore, and there were no large concentrations in any single neighborhood. Component six identifies what I define as “older housing, no growth households.” This component explained 4 percent of the variation in the original data. This was the lowest out of the six principal components. Table 4.4 shows that there were six variables with significant loadings. They related to population size and housing age. Specifically, the loadings could be boiled down to two important trends. First, the population size variable loading was low, at 20.31. Second, the housing age variables loaded very low for newly built housing units, and they loaded very high for older housing units. Together, these loadings demonstrated that neighborhoods scoring high on the sixth component were places that lacked population growth. They were also neighborhoods with a very old housing stock that dated before 1950. The spatial trends for the sixth component present several findings. High scoring neighborhoods contained older housing, no growth households. The older, no growth tracts were primarily clustered around the northern central suburbs and southwestern suburbs. Neighborhoods in Pikesville and Towson as well as Catonsville and Arbutus included older housing, no growth households in their residential populations. Low

126 / transforming race and class in baltimore

scoring neighborhoods did not contain such neighborhoods. They were located in the outermost areas to the north, east, west, and south of the region. Overall, the 2000 PCA showed that there was a diversity of firstsuburban neighborhoods in Baltimore. This analysis detected six components in first-tier suburban neighborhoods: poor households, minority renter households, black middle-class households, older white households, younger service worker households, and older housing no growth households. Collectively, these six components had substantial explanatory power. They represented two-thirds of the variation from the original dataset. The analysis demonstrated that there was a spatial relationship between poor households and nonpoor households. Poor households were generally located in the eastern and western first-tier suburban neighborhoods. In contrast, nonpoor households were located in northern and southwestern first-tier suburban neighborhoods. Economic class separated the population. Residents of different classes were located in different suburban neighborhoods throughout the region. With few exceptions, minority renter households were located in the northern and western first-tier suburban neighborhoods. There were very few minority renter neighborhoods in the eastern and southern first-tier suburbs. The presence of a minority population and renter households distinguished these areas from other neighborhoods. Similarly, the analysis showed the emergence of black middle-class households. These households were primarily located in the northwestern suburban neighborhoods of metropolitan Baltimore. A significant portion of residents in these neighborhoods was black, and public sector employment was a distinguishing characteristic. Above all, the analysis showed that the population was middle class in these neighborhoods. The analysis also detected older white households, younger service worker households, and older housing no growth households. These three components were generally clustered together and scattered throughout suburban tracts in the first-tier suburbs. Older residents and older housing units were generally located in the northern and eastern suburbs. The 2000 PCA demonstrated that class, age, and racial characteristics stratified first-tier suburban neighborhoods. Typology of First-Tier Suburban Neighborhoods The PCA provided important baseline information about the patterns of suburban spatial structure in 1970 and 2000. The PCA scores established the basis for differentiating among neighborhoods in the first-tier suburbs. To create a typology of first-tier suburban neighborhoods, a cluster

suburban mosaic / 127

analysis was employed to better the variation of neighborhood types with suburbs. The clustering technique analyzes the relationships between the PCA scores for each neighborhood and then classifies them into smaller, more manageable groups based on a set of common characteristics. This technique facilitates the interpretation of the principal component scores and allows for the differentiation of suburban neighborhoods. In the following sections, I present the results that define each of the clusters for the analyses in 1970 and 2000. Suburban Neighborhoods in 1970 The cluster analysis that I selected for 1970 yielded five clusters. For the analysis, I clustered 912 individual PCA scores from 6 principal components for 152 census tracts in 1970. Table 4.5 summarizes the distribution of the five cases for the cluster analysis in 1970. Figure 4.2 is a map of the typology of first-tier suburban neighborhoods in 1970. I use this map to describe the location of the suburban neighborhoods for each cluster. In this section, I review the results of the cluster analysis for 1970. For each cluster, I present the distinguishing characteristics that stood out relative to the other clusters in the analysis of the data. Cluster one detects what I define as “minority presence neighborhoods” in Baltimore’s first-tier suburbs in 1970. Only 11 of the 152 neighborhoods in the first-tier suburbs are classified in this group, making it one of the smallest clusters. There were approximately 27,478 residents, or 5 percent of the first-tier suburban population, that resided in minority presence neighborhoods in 1970. The average neighborhood size was approximately 2,500 residents, the smallest of all the clusters. The distinguishing characteristics of minority presence neighborhoods related to the race, occupation, and economic status of residents. At the far southeastern corner of Baltimore, the two largest minority presence neighborhoods were located in Dundalk, as figure 4.2 shows. Blue-collar neighborhoods fully surrounded them. This area was traditionally a large industrial base for the suburbs. For example, Bethlehem Steel Corporation was located in these two black neighborhoods. One neighborhood is Turner’s Station, which was 86 percent black in 1970. The other minority presence neighborhood was 80 percent black. The Bethlehem Steel plant was located in this neighborhood, which was a massive 3,100 acre industrial site located on the Chesapeake Bay. Together, these two neighborhoods supplied a large, low-wage labor base—some 8,000 workers— for the steel plant. In 1970, the average household income in these two minority presence neighborhoods stood at $40,000, and the poverty rate was 9 percent.

Table 4.5

Distinguishing characteristics of suburban neighborhoods in 1970

Cluster (N)

Population

Income

Housing

Labor force

Minority presence neighborhoods (11)

• • • • • • •

• $64,000 • 10% more income than all suburbs • 6% poverty • $61,000 • 5% more income than all suburbs • 5% poverty • $50,000 • Same income level as all suburbs • 5% poverty • $58,000 • 5% more income than all suburbs • 3% poverty • $124,000 • 113% more income than all suburbs • 2% poverty

• 50% built in 1950s and 1960s • 80% homeownership

• 50% service occupations • 40% high school education • 20% manufacturing occupations • 40% high school education • 48% manufacturing occupations • 43% high school dropouts • 26% manufacturing occupations • 44% high school education • 55% professional occupations • 40% college education

Older middle-class neighborhoods (27)

25% black 75% white 17% ages 18–24 2,498 avg. pop. size 98% white 16% over age 65 3,466 avg. pop. size

Blue-collar neighborhoods (56)

• 98% white • 30% below age 17 • 3,787 avg. pop. size

Newer middle-class neighborhoods (48)

• 97% white • 65% ages 18–64 • 3,785 avg. pop. size

Wealthy neighborhoods (10)

• 99% white • 62% ages 18–64 • 3,042 avg. pop. size

• 32% built prior 1939 • 68% homeownership

• 40% built in 1950s • 77% homeownership

• 87% built in 1950s and 1960s • 65% homeownership • 50% built in 1960s • 79% homeownership

suburban mosaic / 129

Legend Minority Presence Older Middle Class

N

Blue Collar Newer Middle Class Wealthy

Figure 4.2

0

3.5

7

14 Miles

Typology of first-tier suburban neighborhoods, 1970.

Source: Author; U.S. Census TIGER File.

Elsewhere, on the western suburban fringe, there were six minority presence neighborhoods scattered among blue-collar and other middleclass neighborhoods. These neighborhoods were on average 25 percent black at a time when the black population in all other first-tier suburban

130 / transforming race and class in baltimore

neighborhoods averaged 2 percent. The economic status and race of the population distinguishes these neighborhoods from others. The average household income was $64,217, which was 10 percent higher than all other suburbs. Moreover, 1 percent of the labor force was unemployed, and half of the residents in these neighborhoods worked in the servicerelated industries. Cluster two identifies what I define as “older middle-class neighborhoods” in the first-tier suburbs of metropolitan Baltimore in 1970. There are 27 neighborhoods, or 1 in 6 neighborhoods in the first-tier suburbs, classified in this group. These suburban neighborhoods housed approximately 93,568 residents, or 18 percent of the population in 1970. The population was 98 percent white. The distinguishing characteristics of older middle-class neighborhoods relate to the age and economic status of the population and the age of housing. Figure 4.2 shows that older middle-class neighborhoods were primarily located in the western and northern suburbs. Specifically, there were two large pockets of these neighborhoods to the west in Catonsville and to the north in Towson. By 1970, older middle-class neighborhoods showed signs of aging. These neighborhoods housed the largest portion of residents more than the age of 65, and they had the smallest portion of residents less than the age of 17. Furthermore, the housing stock was old. Half of the housing stock was built before 1950. Despite these patterns of aging, the residents in these neighborhoods were middle class. With an average household income of $61,000, these neighborhoods had 5 percent more household income than all other suburbs. Cluster three detects what I define as “blue-collar neighborhoods” in Baltimore’s first-tier suburbs in 1970. There are 56 neighborhoods grouped in this cluster, which makes it the largest one. Approximately 40 percent of the entire first-tier suburban population, or 212,094 residents, lived in blue-collar neighborhoods in 1970. These neighborhoods were enclaves for the white population; they were 98 percent white. Blue-collar neighborhoods were exclusively located in the eastern and southwestern suburbs in 1970, as figure 4.2 illustrates. All the neighborhoods that shared an eastern border with Baltimore City were classified as blue-collar neighborhoods. With a few exceptions, the eastern suburban fringe was entirely blue-collar, as indicated by figure 4.2. To the southwest of the city, the blue-collar neighborhoods were scattered among newer middle-class neighborhoods. The occupation, education levels, and economic status of blue-collar residents were the distinguishing characteristics for these neighborhoods. Half of the population worked in manufacturing jobs in 1970. These occupations included laborers, production workers, repairers, operators,

suburban mosaic / 131

assemblers, and material movers. In general, residents did not have high levels of education. Nearly half of the labor force lacked a high school diploma. These jobs earned the residents an average annual income of approximately $50,000, which made these residents the lowest wageearners among all other clusters in 1970. Cluster four identifies what I define as “newer middle-class neighborhoods” in the first-tier suburbs of Baltimore in 1970. This is the second largest cluster with 48 neighborhoods classified in the group. Approximately one-third of the population, or 181,719 residents, lived in newer middleclass neighborhoods 3 decades ago. The residents were 97 percent white. Many of the newer stable neighborhoods were located in the western and southern first-tier suburbs in 1970. Figure 4.2 illustrates that these neighborhoods were primarily located to the west and north of the central city. The greatest concentration of newer middle-class neighborhoods was located in the western suburbs. In particular, a contiguous section of 15 neighborhoods was located in the suburbs of Lochearn and Woodlawn. Also, they were scattered in the southern first-tier suburbs of Anne Arundel County. Still a few other neighborhoods were situated in Towson, just to the north of Baltimore City. Only four newer middle-class neighborhoods were located in the eastern suburbs. The distinguishing characteristics of newer middle-class neighborhoods relate to the age and economic status of the population and the age of housing. The population in these neighborhoods was generally the youngest of all the clusters. Nearly two-thirds of the population was between the ages 18 and 64, and there were very few older residents. In addition, these areas featured the youngest housing stock. Four out of five housing units were built within the past two decades. The residents were largely middle class. Average household income was $58,000, and poverty was below 3 percent. Cluster five detects what I define as “wealthy neighborhoods” in Baltimore’s first-tier suburbs in 1970. There were only ten neighborhoods classified in this group, which made it the smallest and most exclusive cluster. Approximately only 6 percent of the first-tier suburban population, or 30,437 residents, lived in wealthy neighborhoods in 1970. The population was almost completely white, and the residents lived exclusively in an area directly north of Baltimore City in 1970. Hampton, Lutherville, Pikesville, and Towson contained wealthy neighborhoods. In other words, there were no wealthy neighborhoods in metropolitan Baltimore’s southern, western, or eastern suburbs. The economic status, education level, and occupation of residents were distinguishing features of wealthy neighborhoods. By far, these neighborhoods had the highest household income level in the entire metropolitan

132 / transforming race and class in baltimore

area in 1970. The average household income stood at $124,000—more than double the suburban average of $47,721. Moreover, the poverty level was less than 2 percent, which made it the lowest in the region. In terms of the labor force, nearly two-thirds of the residents held professional occupations, and 40 percent of them had at least a 4-year college degree. Overall, the 1970 cluster analysis demonstrates that several factors differentiated the first-tier suburban neighborhoods. They include characteristics about race, income, housing, and occupation of residents. Table 4.5 provides a summary of these distinguishing characteristics of the five clusters of suburban neighborhoods. For each cluster, table 4.5 reports data on the characteristics that set that cluster apart from others, and together these characteristics define each of the clusters. A number of other factors characterize the cluster analysis. First, race was a distinguishing feature of first-tier suburban neighborhoods. With the exception of minority presence neighborhoods, all of the clusters were white enclaves. Although minority presence neighborhoods still had a majority white population, these areas were distinct because a quarter of suburban population in these neighborhoods was black in 1970. The minority population was only located in 11 neighborhoods in the first-tier suburbs. In stark contrast, 3 percent of the rest of the first-tier suburban population was black. In essence, minority residents were the suburban pioneers of the 1970s. Second, the variation of household income was also a distinguishing feature. Most of the first-tier suburban neighborhoods had higher income levels relative to all other suburbs in the region. Moreover, not a single neighborhood in the first-tier suburbs was below the suburban average income level of $47,721 in 1970. Wealthy neighborhoods had twice the income as all other suburban neighborhoods. They also stood out as an anomaly not only compared to all other first-tier suburbs but also to the metropolitan area in 1970. This evidence suggests that the residents of wealthy first-tier suburban neighborhoods formed the upper echelon of Baltimore’s society. In the other 4 clusters, household income varied between $50,000 and $64,000, and the range of household income was only $14,000. Third, the age of housing in the first-tier suburban neighborhoods was diverse. Wealthy and newer middle-class neighborhoods had the youngest housing stock. These areas had the youngest housing stock and were the newest locations to feature the latest suburban subdivisions. In contrast, blue-collar and older middle-class neighborhoods had the oldest housing stock. These neighborhoods often had a housing stock that was built before World War II, and it was exhibiting symptoms of decay by the 1970s. The newer housing stock was generally located to the north and west of the

suburban mosaic / 133

city, and the older housing stock was located to the east and south of the central city. Fourth, the most common occupations that first-tier suburban residents held were in the manufacturing and service industries in 1970. In three of the clusters, manufacturing employment represented at least a quarter of all jobs. In blue-collar neighborhoods, nearly half of all residents were employed in the manufacturing sector. In minority presence neighborhoods and newer middle-class neighborhoods, service sector employment was the most common. Professional occupations were only common in wealthy neighborhoods. In terms of education, most of the labor force had a high school diploma. In some cases, there were pockets of extremes, which included high school dropouts at one end of the spectrum and college educated residents at the other end of the spectrum. Suburban Neighborhoods in 2000 The cluster analysis for 2000 yielded five clusters. For the 2000 analysis, 906 individual PCA scores from 6 principal components for 151 neighborhoods were clustered. One neighborhood was eliminated from the analysis since that census tract that existed in 1970 no longer existed in 2000; it simply became part of another tract. Figure 4.3 is a map of the typology of first-tier suburban neighborhoods in 2000. This map describes the location of the suburban neighborhoods for each cluster. In this section, I review the results of the cluster analysis for 2000. For each cluster, the distinguishing characteristics that stood out relative to the other clusters in the analysis of the data are presented. Cluster one identifies what I define as “poor neighborhoods” in Baltimore’s first-tier suburbs in 2000. Table 4.6 shows that there are 24 neighborhoods in this cluster. Together, these neighborhoods housed approximately one-fifth of the first-tier suburban population, or 113,096 residents. The average neighborhood size contained 4,712 residents. The population was one-third black and 60 percent white. Figure 4.3 illustrates that poor neighborhoods were scattered throughout every quadrant of the first-tier suburbs in 2000. The distinguishing characteristics of poor neighborhoods related to the economic status and family composition of the population and the age and tenure of housing. They were the poorest neighborhoods in the first-tier suburban areas of Baltimore. Average household income was $45,000, which was 35 percent lower than all other suburbs. The poverty rate in these neighborhoods was the highest among all first-tier suburbs. It averaged 12 percent, but the range was between 3 and 30 percent. More than half of the population had a high school education or less. In addition,

Legend Poor Wealthy Blue Collar University

N

Black Middle Class Middle America No Data

Figure 4.3

0

3.5

7

Typology of first-tier suburban neighborhoods, 2000.

Source: Author; U.S. Census TIGER File.

14 Miles

Table 4.6

Distinguishing characteristics of suburban neighborhoods, 2000

Cluster (N)

Population

Income

Housing

Labor force

Poor neighborhoods (24)

• • • • • • • • • • • • • • •

• $45,000 • 35% less income than all suburbs • 12% poverty • $96,000 • 36% more income than all suburbs • 4% poverty • $48,000 • 32% less income than all suburbs • 8% poverty • $13,000 • 76% less income than all suburbs • 61% poverty • $55,000 • 22% less income than all suburbs • 6% poverty • $59,000 • 17% less income than all suburbs • 5% poverty

• 66% built between 1950s and 1970s • 38% homeownership

• 60% service occupations • 51% high school education or less

• 38% built after 1970 • 77% homeownership

• 70% professional occupations • 60% college education

• 85% built prior 1970 • 75% homeownership

• 30% manufacturing occupations • 69% high school education or less

• 76% built during 1970s • 33% rentership

• 60% service occupations • 60% at least some college education

• 70% built between 1950s and 1960s • 75% homeownership

• 40% professional occupations • 53% at least some college education

Wealthy neighborhoods (26)

Blue-collar neighborhoods (48)

University neighborhoods (3)

33% black 60% white 55% female-headed 4,712 avg. pop. size 91% white 9% foreign-born 13% female-headed 2,961 avg. pop. size 93% white 21% more than age 65 26% female-headed 3,007 avg. pop. size 67% white 25% black 2,284 avg. pop. size

Black middle-class neighborhoods (13)

• 71% black • 38% female-headed • 3,484 avg. pop. size

Middle America neighborhoods (37)

• • • •

86% white 11% black 19% female-headed 3,540 avg. pop. size

• 64% built before 1970 • 22% manufacturing occupations • 83% homeownership • 50% at least some college education

136 / transforming race and class in baltimore

more than half of the households were female-headed, which was the highest of all the clusters. These neighborhoods also had an older housing stock. Two-thirds of the stock was built between 1950 and 1970. Approximately one in three residents owned a house. Cluster two detects what I define as “wealthy neighborhoods” in the first-tier suburbs of Baltimore in 2000. Table 4.6 shows that there are 26 neighborhoods in this cluster. Approximately 15 percent, or 76,992 residents, of the population in first-tier suburbs resided in wealthy neighborhoods. The residents were 91 percent white. These places were the only neighborhoods to house a notable immigrant population. On average, 9 percent of the population was foreign-born. These neighborhoods were located primarily in the northern suburbs, as figure 4.3 demonstrates. There was also a small pocket of six wealthy neighborhoods in Catonsville, near the southwest corner of the first-tier suburbs. Economic status, education, and occupation were the distinguishing characteristics of wealthy neighborhoods. These suburban areas were the only ones in the first-tier suburbs that had a higher income level relative to all other suburbs. The average household income was $96,000. These neighborhoods were approximately 36 percent wealthier than all other suburbs. Only 4 percent of the population lived in poverty, which was the lowest of all the clusters. The labor force was highly educated, and the residents predominantly maintained professional occupations. Cluster three identifies what I define as “blue-collar neighborhoods” in Baltimore’s first-tier suburbs in 2000. One-third of all neighborhoods are classified in this group, which makes it the largest cluster. More than one-fifth of the total first-tier suburban population, or 114,373 residents, lived in blue-collar neighborhoods. These areas had the largest portion of white residents, which was 93 percent white in 2000. Figure 4.3 shows that these neighborhoods were located mainly in the eastern suburbs, and to a less extent, in the southeastern suburbs of metropolitan Baltimore. The distinguishing characteristics of blue-collar neighborhoods is related to the economic status, education, age, and occupation of the population and the age of housing. These areas were 32 percent poorer than all other suburbs. The average household income was $48,000 in 2000. Poverty was also higher in these neighborhoods than others throughout metropolitan Baltimore. Eight percent of the population lived in poverty. Blue-collar neighborhoods also had a large elderly population and few younger residents. Approximately one-quarter of the population was more than the age of 65, and the youth population comprised only 6 percent of residents. Furthermore, the housing stock was significantly older than all other neighborhoods in first-tier suburban Baltimore. Nearly 85 percent of the entire housing stock was built before 1970.

suburban mosaic / 137

In manufacturing, 30 percent of labor force was employed. Blue-collar residents had the lowest education levels. Approximately 70 percent of residents had a high school education or less. Cluster four detects what I define as “university neighborhoods” in the first-tier suburbs of Baltimore in 2000. Only three neighborhoods are classified in this group, which makes it the smallest cluster. One percent of the first-tier population resided in university neighborhoods, which amounted to some 7,000 residents in 2000. Figure 4.3 displays the location of these three university neighborhoods. Located in two neighborhoods in the western first-tier suburbs of Catonsville and Arbutus, UMBC is Baltimore’s public national research university. Enrollment stood at approximately 12,000 students in 2006. Directly north of Baltimore City, Towson University is Baltimore’s regional university, enrolling some 18,000 students in 2006. The student bodies of these two universities were the primary residents of these three neighborhoods. The distinguishing characteristics of university neighborhoods were related to the racial and ethnic composition and economic status of the population as well as housing tenure. They stood out as anomalies as compared to all other neighborhoods in the first-tier suburbs. The population was the most racially and ethnically diverse group in metropolitan Baltimore. The student bodies of UMBC and Towson Universities are both widely known for their racial and ethnic diversity. For example, UMBC President Freeman Hrabowski frequently comments, “the diverse campus community often resembles the halls of the United Nations” (personal communication, August 21, 2005). The residents were also among the poorest in the region. Average household income was $13,000, which was 76 percent lower than all other suburbs. Clearly, the presence of a large student population contributed to the low incomes. In addition, nearly two-thirds of the population lived in poverty. Renter-occupied units were the most common form of housing tenure in university neighborhoods. Again, given the transit-oriented nature of student living, these findings are not surprising. Cluster five identifies what I define as “black middle-class neighborhoods” in Baltimore’s first-tier suburbs in 2000. There are 13 neighborhoods in this group, housing approximately 10 percent of the first-tier suburban population, or 45,301 residents. The average neighborhood contained 3,484 residents. Black middle-class neighborhoods were almost all located on the western suburban fringe, as figure 4.3 illustrates. One neighborhood was located to the north and another to the east. A large, near-contiguous section of 10 black middle-class neighborhoods is located on the western suburban fringe and illustrated on the cluster map. These 10 neighborhoods were only located in 2 first-tier suburbs: Woodlawn and Lochearn.

138 / transforming race and class in baltimore

The race, economic status, education, and occupation of the population were distinguishing features of black middle-class neighborhoods. In 2000, black middle-class neighborhoods were 71 percent black. Residents had an average household income of $55,000. More than half of the labor force had at least a four-year college degree, and the most common type of occupation was professional employment. Three-quarters of residents were homeowners, and they typically lived in houses that were built during the 1950s and 1960s. Cluster six detects what I define as “middle America neighborhoods” in the first-tier suburbs of Baltimore in 2000. Other urban scholars have also used the term “Middle America” to describe these particular socioeconomic characteristics of a spatial landscape (Mikelbank 2004; Vicino, Hanlon, and Short 2007). The 37 neighborhoods in this cluster make it the second largest. Approximately one-quarter of the first-tier suburban population, or 130,990 residents, lived in Middle America neighborhoods in 2000. The population was 85 percent white and 11 percent black. Figure 4.3 shows the location of Middle America neighborhoods. They were primarily located around the southwestern and northeastern corners of Baltimore City. The distinguishing features of Middle America neighborhoods were related to the economic status and age of the population, as well as the age and tenure of housing. In 2000, approximately one-fifth of the population was more than the age of 65. The residents of Middle America neighborhoods were largely middle class. They earned an average household income of $59,000 but that was 17 percent less than the suburban average. In 2000, 5 percent of the population lived in poverty. Homeownership in Middle America neighborhoods was the highest of all the clusters. In 2000, 83 percent of the population owned a home. Residents lived in older housing units. Two out of three houses were built during the 1970s. Overall, the cluster analysis for 2000 demonstrates that there was substantial variation in the socioeconomic structure of the first-tier suburban neighborhoods. Table 4.6 provides a summary of the distinguishing characteristics of the five clusters of suburban neighborhoods. For each cluster, table 4.6 reports data on the characteristics that set that cluster apart from others. Overall, four prevailing characteristics defined the clusters of Baltimore’s first-tier suburban neighborhoods in 2000. First, the black and white residential populations were segregated in first-tier suburban neighborhoods. Without a doubt, there was an increase in the racial diversity of the first-tier population, yet racial segregation was prevalent. With the exception of poor and university neighborhoods, all the other neighborhoods in first-tier suburbs had high levels of racial segregation in 2000. Wealthy, blue-collar, and Middle America neighborhoods were all overwhelmingly white, while black middle-class neighborhoods

suburban mosaic / 139

had a large majority black population. These patterns echo the findings of sociologists Douglas Massey and Nancy Denton (1993), who found that residential racial segregation was a persistent feature of the suburban landscape in their landmark book American Apartheid. Second, there was considerable income variation in the typology of six clusters for 2000. Average household income ranged from $32,000 to $161,000 in 2000, a span of $129,000 between the wealthiest and poorest neighborhood. The higher income neighborhoods were spatially clustered together in the northern and southwestern suburbs. Likewise, a similar spatial pattern was evident for lower income neighborhoods in the eastern and western suburbs. Despite the large differences in levels of wealth, nearly all neighborhoods in the first-tier suburbs lagged behind their counterparts in the outer suburbs. The neighborhoods in 5 of the 6 clusters all had income levels below the suburban average income level of $57,557 in 2000. Third, the structure of the labor force was also a defining characteristic in many neighborhoods. Again, there was tremendous variation in type of occupations and levels of education for residents in the first-tier suburbs of metropolitan Baltimore. Manufacturing sector employment was prevalent in blue-collar and Middle America neighborhoods, which ranged from one quarter to one half of the population. In contrast, professional sector employment was common in only two types of neighborhoods: wealthy and black middle class. The range of residents with professional jobs was between 40 and 70 percent of the population in these neighborhoods. Service occupations were common in poor, university, and Middle America neighborhoods. Education levels also varied tremendously. Wealthy and black middle-class neighborhoods had the highest portion of collegeeducated residents. Poor, blue-collar, and Middle America neighborhoods had attained the least amount of education in the first-tier suburbs. Fourth, the age and tenure of the housing stock also differentiated the clusters of suburban neighborhoods. Overall, the housing stock was the oldest in the suburbs in 2000. For every cluster, the majority of housing units were built between the 1950s and 1970s. In addition, high levels of homeownership were a defining feature of the first-tier suburban neighborhoods. In the wealthy, Middle America, blue-collar, and black middle-class clusters, at least three out of four residents owned a house. In contrast, the majority of residents in poor and university neighborhoods rented housing units. Suburban Transformations The two statistical analyses presented in this chapter demonstrate that Baltimore’s first-tier suburban neighborhoods witnessed dramatic change

140 / transforming race and class in baltimore

from 1970 to 2000. There are four themes of change that were common to both the principal components and cluster analyses. The transformation of race, economic status, labor force, and age are the defining features of Baltimore’s first-tier suburban changes. It is important to reflect on these four transformations briefly, which allow us to characterize metropolitan Baltimore’s overall suburban transformation by providing specific examples of how these changes occurred in four of Baltimore’s first-tier suburbs. Although there are interdependent relationships among these four transitions, it is useful to reflect on the role they played in four particular suburbs. Woodlawn: The Transformation of Race First, one of the most evident transformations in the first-tier suburbs was the change in the racial composition of the population. There was an obvious increase in the number of black residents in first-tier suburbs. In particular, there was a dramatic decrease in the white population and large increase in the black population in the western first-tier suburbs of Baltimore. The case of Woodlawn typifies the racial transformation that occurred in western suburban Baltimore. The racial composition of Woodlawn’s population changed dramatically between 1970 and 2000. In 1970, western first-tier suburbs were 95 percent white. By 2000, they were two-thirds black. Some 12,000 white residents left Woodlawn, and more than 17,000 black residents replaced them in this 30-year period (Keating 1994). These analyses demonstrated that in the western suburban neighborhoods, there was a distinguishing spatial pattern that separated the black and white population. For example, six of Woodlawn’s eight neighborhoods were nearly three-quarters black, and the other two neighborhoods had a majority white population. These new black residents, who initially desegregated Woodlawn, were living in resegregated neighborhoods by 2000. The same pattern was evident for the other 20 western suburban neighborhoods surrounding Woodlawn. In total, some 32,000 white residents left Baltimore’s western first-tier suburban neighborhoods, and 35,000 black residents settled in these same neighborhoods over a period of 3 decades. The emergence of a black middle class was also a distinguishing feature of the western suburban transformation. The results of this study suggest that public sector employment played a role in this transformation. The growth of many local, state, and federal agencies during the period of 1970–2000 meant that there was a plethora of employment opportunities for residents in the western and northwestern suburbs. Within a 30-mile

suburban mosaic / 141

radius of these suburbs, more than 1,000 independent government agencies employed an army of suburban residents in metropolitan Baltimore by the end of the twentieth century (Light 1999). In fact, there were some 15,000 public sector jobs located in Woodlawn alone. The headquarters for two of the largest federal agencies were housed in this first-tier suburb of Baltimore. For example, the headquarters for the Social Security Administration and the Centers for Medicare and Medicaid moved to Woodlawn in the early 1960s, and since then, they have employed many suburban Baltimore residents. In short, the Baltimore-Washington corridor served a hub for government activity, and the results of this analysis suggest that this played a role in shaping the formation of a black middle class in suburban Baltimore. The experience of Woodlawn captured the essence of this suburban transformation during the 1980s and 1990s. Essex: The Transformation of Economic Status Second, Baltimore’s first-tier suburbs witnessed a geographic concentration of poor economic status residents. The analysis showed that poor neighborhoods were primarily located in the eastern first-tier suburbs. Essex is an example of a first-tier suburb with the greatest concentration of lower economic status residents in metropolitan Baltimore. Located on the eastern fringe of suburban Baltimore, Essex was largely a middle-class suburb in 1970. Household income was $41,473, which was close to region’s suburban income level of $46,721. By 2000, real household income declined by 16 percent, reaching a low of $34,978. This meant that the residents of suburban neighborhoods in Essex were 40 percent poorer than all other first-tier suburbs in the region. Along with the income losses, poverty in Essex increased significantly. The poverty rate in Essex’s 12 neighborhoods more than doubled between 1970 and 2000. By 2000, 13 percent of Essex’s population lived in poverty. In three of Essex’s neighborhoods, the poverty rate was more than 15 percent in 2000. Essex’s economic woes began with the industrial restructuring of the national economy. The demise of heavy manufacturing industries took a particularly heavy toll on this suburban community. During the height of the World War II industrial years, the Glenn L. Martin Company, located in the heart of Essex, employed some 53,000 employees. The company proved to be the life and soul of Essex for decades after the war, but it slowly downsized, and after several mergers and the purchase by Lockheed Martin, the plant no longer employed the blue-collar workforce that it once heralded from Essex and other industrial first-tier suburbs on the eastern suburban fringe. The results were devastating for Essex. Its population

142 / transforming race and class in baltimore

declined significantly, and as this study showed, the economic status of the remaining residents declined rapidly. Essex and its residents suffered from suburban deindustrialization. This account shows that Essex transformed into a new place. Essex was no longer a middle-class suburb in 2000. Relative to other first-tier suburbs and the outer suburbs, Essex stood out as one of the places whose economic status changed the most between 1970 and 2000. At the beginning of the twenty-first century, the economic prospects for Essex appeared bleak. Dundalk: The Transformation of Labor Force Third, the case of Dundalk highlights how the labor force changed since 1970. This first-tier suburb was home to two industrial powerhouses. The first industry was Bethlehem Steel Corporation, which operated its steel plant in Dundalk since 1916. The second industry was General Motors, which opened its Broening Plant in 1934, manufacturing a variety of vehicles for the company. Scores of residents in Dundalk and other eastern first-tier suburbs worked for these manufacturing giants. Yet by 2000, Bethlehem Steel Corporation’s work force dwindled from approximately 28,000 employees in 1970 to 2,500 in 2000—a 91 percent decrease (Connolly 2006). The General Motors’ Broening production plant closed in 2005. In Dundalk, half of the labor force, or approximately 16,000 residents, was employed in manufacturing industries in 1970. In 10 of Dundalk’s neighborhoods, three-quarters of residents in these neighborhoods held manufacturing jobs. They defined what it meant to be a bluecollar neighborhood. Three decades later, manufacturing employment declined regionally by 50 percent. Dundalk experienced greater losses relative to the region. There was a loss of 10,000 jobs, which represented a decrease of 70 percent. In 2000, only one in five residents had manufacturing jobs—a dramatic local impact on Dundalk residents. Other neighborhoods in Dundalk also experienced a significant decrease in the number of residents who had manufacturing jobs. Consider the example of a minority presence neighborhood in Dundalk named Turners Station. The cluster analysis showed that this area was a blue-collar neighborhood in 1970. Half of residents held manufacturing jobs. By 2000, 13 percent of the residents were unemployed, which was the one of the highest rates in the first-tier suburbs. Also, poverty grew from 9 to 25 percent. The cluster analysis classified Turners Station as a poor neighborhood in 2000. This neighborhood was Dundalk’s home for its disenfranchised black population. This example shows that the loss of manufacturing jobs was part of the changes in the labor force that occurred between 1970 and 2000. This transformation was one of the defining characteristics of the

suburban mosaic / 143

changes in the labor force in the eastern first-tier suburbs of metropolitan Baltimore. Parkville: The Transformation of Age Fourth, the age of the population was a defining feature of suburban change. Consider the experience of Parkville, a first-tier suburb located at the northeastern corner of Baltimore City. Between 1970 and 2000, Parkville’s population aged significantly more than other first-tier neighborhoods and outer suburbs in the region. Parkville’s 9 suburban neighborhoods contained 38,000 people in 1970. The population was primarily comprised of young families with children. Approximately 28 percent of the population was comprised of children under 17, and two-thirds were adults aged between 18 and 64. Only 6 percent of the population was aged more than 65 in 1970. Parkville’s population was very different 30 years later. The cluster analysis originally classified Parkville’s neighborhoods as newer or older middle-class neighborhoods. By 2000, the analysis showed that the majority of places were Middle America neighborhoods. The aging of the population was the distinguishing feature that determined the change in clusters. All the suburban neighborhoods in Parkville witnessed increases in the number of residents more than the age of 65, and there were decreases in the residents less than the age of 64. Specifically, there was a population loss of 5,037 residents less than the age of 18, which was a 50 percent decline between 1970 and 2000. Similarly, there was approximately a loss of 6,000 residents between the ages of 18 and 64, which was a 25 percent decrease during that same period. By 2000, nearly a quarter of the population was more than the age of 65. In three of Parkville’s neighborhoods, the more than one-third of population was more than the age of 65. The Parkville experience demonstrates that the transformation of age was a distinguishing characteristic of this first-tier suburban population. Summary Just as the ancient Romans used mosaics to elegantly decorate and adorn the buildings of the sundry and vast empire of Rome, the mosaic has also come to represent the diversity of the suburban landscape in the twentieth century. Whereas the Roman mosaics featured colored glass and stones pieced together in great detail to depict a history, painting, or culture, today’s American suburban mosaic is an amalgam of many different types of suburbs mixed together in the metropolis, whether it be rich suburbs, poor suburbs, white suburbs, black suburbs, industrial suburbs, or university

144 / transforming race and class in baltimore

suburbs. The suburban landscape is full of an array of places that speckle the landscape. The suburban mosaic represents the socioeconomic realities of metropolitan America in the twenty-first century. Americans live in a decentered, demographically diverse metropolis with suburbs that are increasingly vulnerable as growth occurs at the urban fringe. The key ingredient of the suburban mosaic is diversity—diversity of people, housing, income, and jobs. Urban historian Jon Teaford captures the essence of this suburban diversity in his book The American Suburb: The Basics. One of “the basics” of America suburbs, he contends, is that suburbs are remarkably diverse, and they have been for a long time. He aptly asserts that “suburbia thus includes not only the stereotypical shopping mall and housing subdivision, but virtually everything else as well. It is as varied as the nation as a whole, with landmarks from every century, people of all types, and a full range of businesses, employment, and shopping. Sneering comments about bland suburbs reveal the prejudiced ignorance of the observer rather than the reality of suburbia” (2008, 219). Indeed, the reality today, just as it was in 1970, is that diversity defines the suburbs as it does the city. The two statistical analyses presented in this chapter demonstrated that there was a significant spatial restructuring of Baltimore’s first-tier suburbs since 1970. The PCA in 1970 and 2000 illustrated the spatial transformation of the race, the economic status, the labor force, and the age of the population. These were the distinguishing characteristics of suburban change. The 1970 PCA revealed that class boundaries differentiated the first-tier suburban neighborhoods. A clear spatial pattern separated the professional-class households from the working-class households. The northern first-tier suburbs contained affluent neighborhoods, and the eastern and western first-tier suburbs contained middle-class neighborhoods. There were few instances of poor neighborhoods. They stood out in the analysis as distinguishing features because they did not share similar characteristics of most other neighborhoods. The 1970 PCA detected the very characteristics of suburban decline that were previously demonstrated in chapter three. These included the emergence of fragile renter families, poor black households, and older white households in first-tier suburban neighborhoods. The 2000 PCA showed that the class and racial boundaries differentiated the first-tier suburban neighborhoods. The northern first-tier suburbs generally contained affluent and white neighborhoods, and the eastern first-tier suburbs contained a large white population whose residents had a variety of economic statuses. The western first-tier suburbs contained a large black population whose residents were mainly middle class. The

suburban mosaic / 145

2000 PCA provided evidence that poor households, minority households, renter households, and older households were all more prevalent in 2000 than in 1970. The two cluster analyses created a typology of first-tier suburban neighborhoods in 1970 and 2000 based on the PCA scores. The 1970 cluster analysis showed that blue-collar and newer middle-class suburban neighborhoods were the most common. Together, they accounted for 70 percent of the cluster cases. There were smaller clusters of minority presence neighborhoods and older middle-class neighborhoods, and wealthy neighborhoods were the smallest cluster. In 2000, the clusters changed in a number of ways. Blue-collar neighborhoods were the most common, followed by Middle America neighborhoods. Manufacturing employment distinguished the blue-collar cluster, and age of housing and population distinguished the Middle America cluster. They accounted for 57 percent of the cluster cases in 2000. Also, the number of wealthy clusters doubled to 26 tracts, and 24 tracts of poor clusters emerged in 2000. Last, two distinct clusters became apparent in 2000. The analysis detected three university neighborhoods and 12 black middle-class neighborhoods. These statistical analyses demonstrated that first-tier suburban neighborhoods were differentiated in 1970. They were differentiated in a different pattern in 2000. Over the 30-year period of study, the analyses showed that every first-tier suburban neighborhood changed, and the degree of change varied throughout the suburbs. The changes were related to the transformation of the population’s race, age, economic status, as well as changes in the labor force. There was a spatial relationship between each of the characteristics of change in 1970 and 2000. Above all, both these statistical analyses confirmed that suburban transformation is a spatial phenomenon. First-tier suburban neighborhoods changed from their status in 1970, and many of these changes were related to patterns of suburban decline. Thus, the suburban mosaic provides a lens for understanding the complex spatial patterns of socioeconomic diversity in Baltimore’s firsttier suburbs. Key to understanding the suburban mosaic is contrast. The creation of image and outline provides the finer details and the depth needed to uncover a vibrant picture of diversity. This creates a rich texture for understanding not only the maps of diversity, but also the suburban mosaic of people and their built environments—and their differences.

This page intentionally left blank

Ch a p t e r Fi v e Su bu r ba n R e na is sa nc e

Throughout the Dark Ages, Western Europe was largely a desolate and simple place. There was a lack of a written history during this period, and many forms of human inquiry stagnated including language, music, art, architecture, and technology. Population decline also characterized the Dark Ages. To stimulate societal growth of all kinds, the Italians and French alike began to search for innovative ways to invigorate the population through cultural and scientific avenues. The result was the European Renaissance, which began in the fourteenth century and lasted through the seventeenth century. The word renaissance means “rebirth” in French and Italian, and it offered an ailing Europe an opportunity to reinvent itself. For Europe, renaissance meant revival of people; recovery from disease; resurgence from death; and reawakening from stalemate. Many human advances were made during the Renaissance’s three centuries, and such hallmark contributions came from a variety of scholars such as da Vinci, Michelangelo, and Medici. The Renaissance remade a declining Europe into a flourishing empire. The transformation of Europe parallels the twentieth-century suburban transformation of metropolitan America. Just as Europe was a gothic place during the Dark Ages, the suburban gothic has emerged as a dominant force in U.S. regions. Suburbs, especially first-tier suburbs, suffer from some of the same symptoms of decline and stagnation that medieval Europe experienced a millennium ago. The first-tier suburb of Catonsville is an example of a place that had stagnated during the 1970s and 1980s relative to other suburbs in the region, yet with the onset of the 1990s, this suburb began to experience a resurgence of people and place. Money magazine took note of this transformation, observing that “once a quiet bedroom community outside Baltimore, Catonsville is undergoing a renaissance, thanks in large part to the University of Maryland at Baltimore County [sic], which has transformed itself from a commuter school into a dynamic research center . . . In the process, the city has attracted a number of high-tech firms—and the jobs they come with” (Money, August 2007,

148 / transforming race and class in baltimore

90; emphasis added). Catonsville is a suburban destination in the twenty-first century since its rebirth through renaissance. Yet the blossoming of this suburb, among many others, would not have been possible without public intervention. Similar to Europe, Catonsville declined, stagnated, and then rejuvenated itself with the assistance of government and its citizens. Therefore, a growing number of influential academic and popular leaders have focused on the socioeconomic state of the nation’s older suburbs. Scholars and practitioners have increasingly paid attention to the changing nature of American suburbs and their likelihood to experience decline in recent decades—a familiar story for many regions (Hudnut 2003; Lucy and Phillips 2000; 2006; Wheeler 2005a). A brief glimpse at some latest headlines shows that there is public concern for the topic of suburban decline. Newspaper headlines nationally read, “older suburbs struggle to maintain their vitality” (Christie 2005); “old ‘burbs need work’” (Ott 2006); “inner suburbs fall through the cracks” (Cohn 2006); and “suburbs nearer to cities neglected” (Ohlemacher 2006). Reports such as these suggest that not only central cities, but also their surrounding first-tier suburbs, suffer from widespread social and economic problems. As a result, this has prompted a variety of inquiries into the state of first-tier suburbs today. These suburban areas were once the icons of the American Dream. Yet with the massive growth of metropolitan areas and the decentralization of people and jobs, these areas are no longer places of destination. The lure of the outer suburbs (Garreau 1991) and the rebound of many downtowns (Grogan and Proscio 2000) have left the first-tier suburbs caught in between two stronger forces of place. Decades of deindustrialization, coupled with an aging population, an obsolete housing stock, and a diversifying population made first-tier suburbs fully mature by the beginning of the twenty-first century. Across the nation, these issues affected many older suburban areas to varying degrees (Bourne 1996; Hudnut 2003). Just like the national trends, suburbs in metropolitan Baltimore are no exception. For three decades, Baltimore’s first-tier suburbs witnessed dramatic social and economic changes. In particular, this study showed that the first-tier suburbs experienced marked transformations both relative to each other and to the outer suburban areas of the region, and the majority of Baltimore’s first-tier suburbs witnessed moderate to severe decline between 1970 and 2000. In general, the stagnation of population and income, the lack of a marketable housing stock, and changes in the labor market characterized Baltimore’s first-tier suburban transformation. Baltimore’s suburbs—like so many others in the Rustbelt—had reached a crossroads by the end of the twentieth century, and their future prospects were tenuous at best. The time for change had arrived, and suburban residents turned to their government for help.

suburban renaissance / 149

Indeed, what ultimately happens to the first-tier suburbs is, in part, a question of public policy. What are the roles of public policy as they apply to first-tier suburbs? What can we learn from the metropolitan Baltimore experience? What political and policy tools can be used to revitalize these suburban communities, and how are they limited? Drawing on the experience of Baltimore’s confrontation with suburban decline, I reflect on how the political structure and actors that constitute local government affect what those local governments can do and choose to do about suburban decline. I present a discussion of the lessons learned in confronting suburban distress and their relationship to local policy. Without a doubt, evidence of suburban decline abounds. It is clear that first-tier suburbs now have urban problems, and an unmistakable pattern of social and economic decline has emerged in the first-tier suburbs. Problems of socioeconomic decline have often been associated with the city or urban areas, and these areas have been commonly depicted “as place[s] of crime and discontent” by the media (Vale 1995, 646). Thus, popular representations of decline have been associated in a negative light as urban decline morphs into suburban decline. Although popular parallelisms and social constructions of suburban decline have been related to urban decline, the experiences are vastly different given the difference in the age of development and the cycle of decline between city and suburb. In short, the wrath of social and economic decline has wreaked havoc in U.S. central cities for decades, and it now creeps into suburbs (Short 2007). Between 1970 and 2000, the outer suburbs flourished while the first-tier suburbs declined. In other words, these disparities were spatial in nature. In 1970, the first-tier suburbs were new places and the outer suburbs were still largely rural in character. Yet by 2000, the spatial character of suburbs had changed. First-tier suburbs were built out and older, although the outer suburbs were newer, developing places. In essence, the life cycle of suburbs decentralized in an outward spatial pattern (Lee and Leigh 2005). The political and policy responses in first-tier suburbs have thus transformed from growth to decline issues. Let us now reflect on how the public sector confronted the problem of suburban decline in metropolitan Baltimore, focusing particularly on the first-tier suburbs in Baltimore County. Baltimore County Confronts Suburban Decline Baltimore’s first-tier suburbs evolved into different places as the socioeconomic status of its residents declined between 1970 and 2000. What, if anything, can be done to prevent further decline of these places and to put them on the track to recovery? In this chapter, I present an overview of the current political and policy responses at the local level of government that

150 / transforming race and class in baltimore

deal with the decline of older suburban communities, and then I offer a critical appraisal about the realities of confronting suburban decline. Community Conservation Baltimore County, Maryland stands out as a national leader among local governments in confronting suburban decline (Outen 2005). The county has attracted national attention for confronting its problems of suburban decline through the development of a strategy called “community conservation.” The primary focus here is to review a sample of revitalization projects under the county’s community conservation efforts and to present a critical assessment of the realities and limitations of the county’s suburban revitalization efforts. Beginning in the mid-1990s, Baltimore County developed a strategy to systematically address the problem of suburban decline. Baltimore County Executive “Dutch” Ruppersberger created the Office of Community Conservation on July 1, 1995. When Ruppersberger ran for public office in 1994 and throughout the following eight years of his term in office, he focused on four strategic goals for Baltimore County: education, crime, economic development, and community conservation. The idea was to build human capital throughout this maturing jurisdiction by conserving areas that were already built. Accordingly, this office was charged with stabilizing and revitalizing the county’s older neighborhoods and commercial areas in first-tier suburbs. The county’s mandate to address this problem continued into the 2000s when the next county executive, James “Jim” Smith, following Ruppersberger, also made first-tier suburban renewal a platform issue for his administration. Following in the spirit of former County Executive Ruppersberger, Smith acknowledged the importance of revitalization by noting that “the role of government is to protect the individual character of our neighborhoods, and safeguard our resources and adjust to the changing development needs of our communities” (Smith 2003, 1). The mission of the Office of Community Conservation was thus “to preserve, stabilize and enhance the human, physical, and economic conditions of the County’s urban communities” (Baltimore County 2000, 1). To carry out this mission, the office created the Renaissance Development Initiative. This initiative was a countywide revitalization plan that sought to create a “renaissance”—or a new beginning—for Baltimore’s aging suburban communities. When it began in 1995, the plan called for redevelopment projects that addressed the problems of a dilapidated housing stock and struggling commercial strips throughout first-tier suburbs. The office used a map of the community conservation areas to guide

suburban renaissance / 151

its planning process and public investments. Funding for revitalization projects was directed exclusively to these areas. All of the county’s first-tier suburbs were located within the conservation boundaries. Furthermore, all of Baltimore County’s first-tier suburbs are classified as census designated places (CDPs). These CDP boundaries are delineated to collect data on unincorporated areas with concentrations of population, housing, and commercial sites, and a degree of local identity. These boundaries are established in cooperation with local and state officials, and they define the areas for Dundalk, Essex, and Middle River in this chapter. With a public mandate from the county executive, the Office of Community Conservation focused its energy and resources on these three communities to prevent and even reverse a longtime pattern of socioeconomic decline (Baltimore County 1999). In the decade between 1995 and 2005, the Office of Community Conservation carried out a variety of projects to address the problem of an aging housing stock and struggling commercial areas. The office’s planning efforts and funding largely focused on the eastern suburbs of Baltimore, particularly in Dundalk, Essex, and Middle River. Table 5.1 provides an overview of the community conservation projects from 1995 to 2005 in Dundalk, Essex, and Middle River. By 2000, these three firsttier suburbs had experienced the most socioeconomic decline, and they became the county’s primary planning focus for targeted revitalization. Therefore, the review is limited to projects in these three suburbs for the first decade of revitalization from 1995 to 2005. Dundalk First, consider the case of Dundalk. This first-tier suburb is located on the eastern fringe of the central city. It is one of the largest and oldest suburbs of metropolitan Baltimore. Dundalk’s transition from a stable to declining community was quite dramatic. The suburb was home to some of the largest industrial plants in the world, including the behemoth Bethlehem Steel Corporation and General Motors (Reutter 2004). Dundalk’s Bethlehem Steel Plant opened in 1917 and employed some 40,000 workers at its peak in 1950, and today it only employs approximately 2,500. Similarly, Dundalk’s General Motors Broening Highway Assembly Plant opened in 1935 and employed some 10,000 workers at its peak in 1950. It was closed in 2005 when only 1,100 jobs remained. The plant was demolished in 2007. The massive deindustrialization of the economy hit the area hard because so many of its residents were employed with industrial jobs. Two out of three manufacturing jobs disappeared between 1970 and 2000. The poverty rate nearly doubled from 5 to 10 percent; household income declined

152 / transforming race and class in baltimore Table 5.1 Overview of Baltimore county’s community conservation projects in the Eastern first-tier suburbs, 1995–2005 Area of emphasis

Project

Location

Investment ($)

WaterView Homes Hopewell Pointe Homes Miramar Landing Shelter Harbor Housing—Redevelopment Chesapeake Village (property acquisition, Riverdale Village demolition of structure, Villages of Tall Trees and relocation of residents) Tidewater Village York Park Apartments Housing—Owner Settlement Expense Loan Assistance Program (SELP) Commercial WaterView Town Center Historic Dundalk Village Shopping Center Martin Plaza Shopping Center

Essex Essex Middle River Dundalk Middle River Essex Middle River Middle River Dundalk Conservation Areas Essex Dundalk

45 million* (1) 60 million* 20 million N/A 2 million 0.5 million 17 million 0.5 million 17.2 million 10 million

Middle River

Infrastructure

Essex

25 million* 15 million

Housing—New

Planning

Eastern Boulevard Enhancement Maryland Route 43 Extension Streetscaping of Primary and Secondary Corridors

Middle River Dundalk; Essex; Middle River Tall Trees Park and Eastern Essex; Regional Park Enhancements Middle River Plan to Realize the Waterfront Conservation Potential Areas Urban Design Assistance Dundalk Team (UDAT) ReDiscovery Campaign Dundalk; Essex

45 million* a 7.7 million*

60 million 50 million 11 million 0.5 million 1.5 million 0.5 million

Source: Baltimore County Office of Community Conservation. * denotes private-public dollars. Note: a total project.

by 12 percent, falling from $45,000 to $39,000 in 2000. Consequently, the population declined by one-third; approximately 22,000 residents left. Dundalk became an aging community, and the number of residents aged more than 65 tripled since 1970. By the end of the 1990s, Dundalk faced especially challenging issues. The enduring deindustrial legacy of Dundalk became a familiar chronicle that many other first-tier suburbs across the nation also experienced (Niedt 2006). These trends of socioeconomic decline suggested that Dundalk would continue to deteriorate unless something was done to thwart the suburban decay. In response, Baltimore County Executives Dutch Ruppersberger and Jim Smith took a special interest in the revitalization of Dundalk. Since

suburban renaissance / 153

2002, the Office of Community Conservation has invested $70 million into Dundalk. This public investment was aimed at demolishing rundown apartment complexes, renovating older housing units, and refurbishing Dundalk’s town square. At a press conference, Smith continued his pledge to give Dundalk a suburban makeover. On March 27, 2006, Smith announced the $7.6 million purchase of the first 20 buildings of the complex of York Park Apartments. The remaining 36 buildings were subsequently purchased for an additional $3.4 million. The housing complex, located in the center of Dundalk, was decrepit and an eyesore. It became a magnet for criminal activity—there were more reported crimes in this area than in any other part of the county, and Baltimore County police responded to 3,800 calls in 2005 (Smith 2006). The county worked for several years to condemn the property for demolition. County Councilor John Olszewski Sr. (7th District, Dundalk), played a pivotal role in the condemnation of the property. As a chair of the council, he was able to convince other councilors to vote with him to condemn the property. Plans to raze the property came to fruition in January 2007, yet not without the costs of acquisition, demolition, and relocation of residents escalating to more than $17 million for the entire project. Despite the costs, spirits were not dampened. Olszewski said, “[B]ringing this type of significant change to my community is why I chose to be in public service” (Mosher 2006). Smith (2006, 1) followed suit, declaring, “[T]he wait is over . . . today Dundalk’s renaissance shines like never before, and I promise you that as the York Park complex comes down, a wonderful new community will be born.” Dundalk residents, including the suburb’s community development corporation, the Dundalk Renaissance Corporation (DRC), rallied behind county planners and Smith for this milestone. Residents felt that the demolition of York Park would help Dundalk confront its challenges, including fighting crime and providing attractive, quality housing. DRC President Scott Holupka commented that “the creation of desirable housing options for families in the area is critical for Dundalk’s continued renaissance. The county’s commitment to changing the landscape along Yorkway is a major step forward in that effort” (Smith 2006, 1). Carolyn Jones, president of the Greater Dundalk Alliance, echoed Holupka’s sentiments. “We’ve wanted them down for years. It’s finally coming to fruition. It’s a sense of hope” (Malarkey 2007). Other projects in Dundalk were related to commercial and infrastructure areas of redevelopment emphasis. In 2000, the county commissioned local architect Jane Willeboordse to spearhead an urban design assistance team (UDAT) to develop a community-based vision for Dundalk in the twenty-first century. Willeboordse, with public funding and logistical support from the county, assembled a team of planners, academics, and

154 / transforming race and class in baltimore

community advocates to create the new plan for the Dundalk. Two years later, the results were presented to the county and a series of recommendations was given to the Office of Community Conservation. Among other suggestions, the UDAT urged the renovation of the Dundalk Village Center; streetscaping of Dundalk Avenue; large-scale renovations of the Dundalk Village Apartments; and the expansion of the Heritage Trail to other neighborhoods, various parks, and the waterfront (Willeboordse, personal communication, October 24, 2004). Accordingly, in March 2005, the Office of Community Conservation worked with the county council, particularly Councilor John Olszewski Sr. of Dundalk and Council Chair Joe Bartenfelder, to sell the historic Dundalk Village Shopping Center to a private developer to spur its revitalization. The village was designed by Frederick Law Olmstead and is considered Dundalk’s focal point. County planners consider its redevelopment to be the anchor that stimulates future projects. In March 2005, developer Jack Jacob of JMJ Dundalk Properties, LLC, purchased the property for $3.7 million and promised $2 million in renovations of office, residential, and retail space. The county provided an additional $2 million in loans and grants. Baltimore County Council Chair Joe Bartenfelder viewed the deal as positive step, commenting that “this partnership is really in the best spirit of government working together with business and the community to bring about positive change that benefits everyone” (Mosher 2005). In addition to the historic village, other developers have sought to capitalize on the waterfront development recommendations that the UDAT presented to Baltimore County. There are currently three developers moving forward with plans to develop Dundalk’s waterfront. In 2005 land speculator John Riehl, of Obrecht-Riehl Properties, announced plans to build a large housing development on the water with a 50-slip marina as well as 79 single-family homes, 56 mid-rise condominiums, and 14 townhomes on 63 acres near Peninsula Expressway. In October 2005, T. Kevin Carney, president of Thomas Builders, announced the construction of 3, 4-story buildings of 144 condominiums and a 60-slip marina in Dundalk’s Bear Creek. Last, yet another mixed-used development and 300-slip marina proposal remains pending. Baltimore County Director of Economic Development summarized aptly, “the magic is in the appeal of the water, and the water is all over the place there” (Graham 2005). Essex Second, consider the case of Essex. By 2000, this first-tier suburb had become the most troubled place in metropolitan Baltimore. The home of the former Glenn L. Martin Company, a defense contractor and longtime

suburban renaissance / 155

benefactor of the cold war, it employed some 50,000 residents in its heyday. The fortunes of the middle-class population who were born during Essex’s industrial era were lost with the onset of suburban decline. For instance, household income declined by 16 percent between 1970 and 2000, declining from $42,000 to $34,000. In addition, poverty doubled from 6 to 12 percent during this period, which made Essex home to the largest poverty population in the county. The housing stock was very old; two-thirds of housing units were built during the 1940s for the Glenn L. Martin Company’s work force. By 2000, Essex faced a critical point because the suburb had experienced three consecutive decades of the most severe socioeconomic decline, relative to all other suburbs in the Baltimore region. The Office of Community Conservation attempted to tackle the decline of Essex. A decrepit housing stock became the center of planners’ attention. A large number of apartment complexes in the suburban neighborhoods of Essex dated back to the World War II era. These apartment complexes had fallen into disrepair after some 50 years of neglect. They lacked the modern amenities that apartments feature nowadays, such as central air conditioning, large bathrooms, and multiple bedrooms. Making matters worse, many of the complexes were poorly managed, becoming places that harbored crime (Smith 2006). In short, they were plagued with a variety of urban housing problems. One of the first complexes that the Office of Community Conservation focused on was the Riverdale Village complex on Eastern Avenue in Essex. For half a century, the 1,140-unit apartment complex housed thousands of low- to middle-income families. In the 1990s, the complex was entirely composed of Section 8 renters. A lack of upkeep allowed the community to become infested with drug use and violence. A HUD foreclosure caused the complex to become fully vacant in 2000, and by 2002, the Office of Community Conservation spent approximately $0.5 million to acquire and demolish the property for redevelopment. In its place, the county sought to stimulate the redevelopment of a new suburban town center named WaterView Town Center. The county partnered with Mark Builders Company to construct the town center, described by the developer as a “neo-traditional community.” The $45 million project featured 175 single-family housing units, linked with sidewalks and bikeways to the 96,000 square feet commercial center that included a supermarket, a bank, and other local retailers. Although this development was originally billed as “affordable housing,” the waterfront property soared in value. Therefore, many of the originally displaced residents were no longer able to reside in this development. In another nearby project, the county attracted Ellwood Building Corporation to Hopkins Creek in Essex to develop a

156 / transforming race and class in baltimore

similar suburban town center to WaterView Town Center. The $60 million development featured 262 housing units comprised of 86 single-family units and 176 condominiums as well as a community swimming pool, a restaurant, and a private marina. Both of these housing developments in Essex took advantage of their location on the waterfront, and other public investments such as new highways, freshly paved streets, and ample sidewalks to attract and retain residents. Ultimately, county planners aimed to have these projects serve as anchors for the county’s revitalization plan for eastern suburban Baltimore, and to date, planners continue to invest in strategies to capitalize on the suburban waterfront as the greatest asset. Middle River Third, consider the case of Middle River. Similar to Essex and Dundalk, Middle River is located on the eastern fringe of the county. This first-tier suburb experienced many of the same patterns of socioeconomic decline that Essex and Dundalk faced. Similar to Essex, the Office of Community Conservation made housing redevelopment and infrastructure its primary goals. The county acquired, condemned, and demolished the Victory Villa Gardens Complex in the Glenmar neighborhood. In its place, Office of Community Conservation Director Mary Harvey and County Councilor Joseph Bartenfelder (6th District, Essex) enticed Ryland Homes, metropolitan Baltimore’s largest homebuilder, with $20 million to redevelop the site. The new development is called Miramar Landing and will feature approximately 800 housing units, of which 100 will be for seniors. Priced between $250,000 and $400,000, the development is out of reach for many Essex residents. The county also paid attention to an area called Tidewater Village. This was another troubled apartment complex, located in the center of Middle River. The property suffered from similar problems of decay. The Office of Community Conservation razed about half of the property to make it a less dense environment. The property was reduced from 981 units to 549 units. The remaining units were all renovated and brought up to modern housing standards. Green space and recreation areas were also integrated into the redevelopment. A playground was installed, and open space trails gave access to the Eastern Regional Park. Two other dilapidated apartment complexes were completely demolished after succumbing to years of abandonment. The county used a different approach in redeveloping these two properties. They were converted into parks. The Village of Tall Trees became a 50-acre neighborhood park for residents in Middle River. Chesapeake Village was transformed into a park with access to the Chesapeake Bay. Last, streetscaping also became a common

suburban renaissance / 157

approach for the revitalization of older suburbs. In Middle River, the Office of Community Conservation invested $5 million to refurbish Eastern Boulevard. The project improved transportation access, created pedestrian walkways, and provided general aesthetic improvements along the main corridor. The Fallout over SB 509 It is important to note that process of redevelopment in first-tier suburbs has not received uniform community support. In 2000, Maryland Senator Michael J. Collins (Democrat, 6th District) of the Baltimore County Administration, sponsored Senate Bill (SB) 509, entitled “Baltimore County Neighborhood Renewal Authority.” The bill sought to authorize Baltimore County to undertake and carry out projects for residential, commercial, and industrial development and redevelopment; authorizing Baltimore County to exercise the power of eminent domain; limiting the power of Baltimore County to undertake specified renewal projects to specified geographic areas within the County; limiting the County’s exercise of eminent domain to specified properties within the County; requiring the County to provide specified compensation to displaced persons.

The Maryland General Assembly, with the endorsement of Baltimore County’s Office of Community Conservation, Department of Planning, and Office of the County Executive, passed SB 509, a bill that permitted the county to use eminent domain for economic redevelopment. The House of Delegates passed the bill 91 to 35 votes and the Senate passed the bill 37 to 5 votes. Governor Parris Glendening signed the bill on April 25, 2000. Under the mandate of SB 509 and the Baltimore County Renaissance Initiative, County Executive Ruppersberger and planners sought to condemn approximately 300 properties in the eastern first-tier suburbs of Essex and Middle River for a new waterfront mixed-use development. Residents of these suburbs strongly opposed the county’s attempts to seize private property for private redevelopment. They formed a community group named “Essex-Middle River Community in Action” to oppose the redevelopment initiative. Rick Impallaria, a resident of Essex and owner of a body shop on Eastern Avenue, commented, “we’re having an effect . . . we’ll pull Baltimore County together on this. I tell you, from what I hear, there are a lot of people in different neighborhoods who have interest in what we’re doing who don’t like Dutch Ruppersberger” (Anft 2000, 1). Many of the residents of Essex and Middle River mobilized to try to save numerous homes, 40 businesses, and hundreds of rental apartments in their neighborhoods from the bulldozer. The residents argued

158 / transforming race and class in baltimore

that their houses provided an affordable alternative for housing in an overpriced region as well as family-owned businesses that have passed from multiple generations. The Essex-Middle River Community in Action argued that the county was robbing them of their “blue-collar waterfront” and was using government to condemn private land for economic gain at the expense of longtime residents. Essex-Middle River Community in Action member and Essex resident Bob Delsignore commented that “we all want to see the area revitalized. Basically, people just want the option of being part of the process instead of being subjects of the county. Nobody’s against cleaning up the area. It’s the way they’ve done it—by punishing the people who have paid their dues and didn’t create the problem in the first place” (Anft 2000, 2). Robert Hannon, who was the county’s director of the Office of Economic Development in 2000, noted the importance of revitalizing eastern Baltimore County. “There is a rare convergence of opportunity. This area had been bypassed by development interests in the past because of the physical and economic decline there. Our intention is to bring benefit to underachieving areas” (Anft 2000, 2). Indeed, the neighborhoods in Essex and Middle River had long fallen behind the rest of the county, with incomes at least $16,000 less than the median household income for the county in 2000. But what properties would be considered “blighted” would become the source of much controversy. On November 7, 2000, the group mobilized 70 percent of the electorate to vote to remove county’s right to use eminent domain for economic revitalization purposes. Despite SB 509’s unpopularity and the mounting legal challenges, the Maryland State Attorney General J. Joseph Curran Jr. reviewed SB 509 and interpreted the law to be fully constitutional in 2000. He relied on a passage that was added to the Maryland constitution in 1960 that authorized the use of eminent domain to remove blight and a 1996 law that allowed private land to be condemned for development. “Even in the absence of slums or urban blight, the use of the power of eminent domain to promote certain types of development has been upheld,” the opinion stated (2000, 1). “These cases recognize that where the development serves public purposes, such as providing employment opportunities and general economic benefit for the community, a public use is present even though the public may not literally or physically be able to use the property,” he said. Curran (2000, 1) went on to opine that “[t]he necessity of a particular condemnation is [an issue] for the legislative or executive branch and will not be disturbed by a court unless the decision is so oppressive, arbitrary, or unreasonable as to suggest bad faith.” The U.S. Supreme Court largely agreed with Curran and the State of Maryland. The Court subsequently reaffirmed the right to use eminent

suburban renaissance / 159

domain for economic development in Kelo v. City of New London 545 U.S. 469. In the aftermath of the community’s opposition to SB 509, the Office of Community Conservation developed a participatory planning process for the redevelopment of its other eastern first-tier suburb, Dundalk. The county provided funding and logistical support for the development of the Dundalk Renaissance Corporation, a community development corporation that would serve as the suburb’s community liaison and voice. As a result, subsequent efforts to revitalize the county’s suburbs proceeded with tepid community support. The fallout over SB 509 also caused Ruppersberger political problems, including many right in his own backyard. Ruppersberger had been expected to present himself as a formidable primary challenger to Lieutenant Governor Kathleen Kennedy Townsend in 2002. “He can run for governor,” observed political analyst Frank DeFilippo, “but it’s going to be awfully tough. If he ran right now, he might not even win his own county.” He went on to note that “Dutch was playing an insider’s game in Annapolis, but he forgot what was happening with his folks back home. It’s bad timing for him. He could have probably run for governor [in 1998] and got it. [Governor Parris] Glendening was weak that year. But SB 509’s going to make it tough for him this time” (Anft 2000, 8). Yet another political observer, political scientist Matthew Crenson, noted that Ruppersberger’s aspiration for the governorship was over after the fallout over 509. “The constituency with which he may have hurt himself is state legislators,” Crenson said. “He got them to vote for this bill, and then it backfired. If they’re looking for someone to put up for [governor], they may think twice about him” (Anft 2000, 8). Ruppersberger never ran for governor of Maryland. Instead, he launched a successful bid for the Second District of the U.S. House of the Representatives in 2002, and since then he has represented many of the same residents of the eastern Baltimore County who were impacted by SB 509. Overall, table 5.1 presents a typology and critical review of Baltimore County’s revitalization efforts in the county’s eastern first-tier suburbs. It serves as a broad representation of the types of projects with areas of emphasis in housing, commerce, infrastructure, and planning that the Office of Community Conservation carried out over the past decade. Although the revitalization of eastern suburban Baltimore was a slow and lengthy process spanning more than 10 years, these projects offer evidence of progress (Malarkey 2006). Despite the progress and modest success, several limitations prevailed. These projects used a “bricks and mortar” approach. Policymakers and planners chose to focus on the renewal of physical infrastructure rather than the development of human capital and the social and economic well-being of residents in first-tier suburbs.

160 / transforming race and class in baltimore

Although the built landscape was revitalized, these projects did very little to raise the economic standing for residents of first-tier suburbs. In fact, the county’s approach raises a number of concerns. First, the county’s redevelopment strategy raises questions about gentrification and displacement (Niedt 2006). The lack of affordable housing remains a dominant trend for Baltimore County and the region. A recent study found that 25 percent, or 70,000 of all households, and 70 percent, or 30,000 of all low-income households, lack affordable housing in Baltimore County (Vicino, Sharkey, Regan, et al. 2005). The county’s strategy of demolishing affordable housing units and replacing them with significantly more expensive units only serves to further deplete the stock of affordable housing units. Despite these concerns, private developers have generally viewed the transformation of eastern suburban Baltimore County in a positive light. One developer commented, “that area [eastern Baltimore County] has a lot of positive things going on . . . the Route 43 extension; the county cleaning up some of the older apartments; the whole area is undergoing a renaissance” (Graham 2005). Second, the county and private developers have focused in large part on developing the land on the waterfront. In the case of WaterView in Essex, a suburban town center dubbed “affordable, middle-class housing” quickly morphed into upper, middle-class expensive housing as the value of land soared. Revitalized housing projects are increasingly out of reach for many residents of eastern suburban Baltimore County, and the majority of Baltimore County residents will not be able to afford waterfront property, especially residents of declining first-tier suburbs. Third, the county’s policies largely do not tackle one of the core issues of suburban decline: income decline. The Office of Community Conservation does offer limited homeowner grants called “Settlement Expense Loan Program,” which provides a $10,000 loan for closing costs that is forgiven after 15 years. Overall, income assistance programs were negligible and did little to impact the socioeconomic status of residents in declining first-tier suburbs. Despite these limitations, Baltimore County’s community conservation approach was one of the most comprehensive local approaches in the nation to confronting suburban decline at the brink of the twenty-first century. Political Realities and Economic Imperatives These accounts of the county’s revitalization projects demonstrate that Baltimore County was able to confront the decline of its first-tier suburbs from 1995 to 2005. However, the effectiveness and impact of suburban revitalization projects are not yet fully known. Since accurate microlevel data are not yet available to test the impacts of these projects on the

suburban renaissance / 161

outcomes of the population, it is important to reflect on the process of first-tier suburban revitalization. This raises several important questions. What made these projects possible? What are the limitations of the county’s revitalization efforts? What is the likelihood that these efforts will actually prevent future decline? I address these questions with a critical assessment of the realities and limitations of confronting suburban decline in metropolitan Baltimore. Strong County Government Without a doubt, the characteristics of Baltimore County, namely the large size, the ability to redistribute funds, and its power, helped to facilitate these projects and make suburban revitalization possible. Baltimore County’s revitalization projects were expensive. They required substantial public investment from the local government. In the state of Maryland, county governments are very strong. In Baltimore County, the county executive maintains strong budget authority, which allows the executive to link his or her policy priorities to the budget process. In this case, Baltimore County Executives Dutch Ruppersberger and Jim Smith made suburban revitalization important policy issues for their administrations. Both supported the Office of Community Conservation by providing an annual operating budget of $55 million to support a professional staff of approximately 30 urban planners. This gave Office of Community Conservation Director Mary Harvey a mandate to engage in suburban renewal (personal communication, May 12, 2005). The governance structure of Baltimore County is a political and corporate body that carries out the public functions of the jurisdiction. The county is fully suburban and surrounds Baltimore City, an independent political incorporation in the state. There are no political subdivisions, and Baltimore County has been operated as a charter county since 1957 under a county executive and a seven-person legislative county council. The General Assembly of Maryland vests the county council with all lawmaking powers granted under its charter. The seven councilors are elected from each of seven equally populated and spatially contiguous districts respectively. They each represent approximately 107,000 residents. Districts 6 and 7 represent the first-tier suburbs of eastern Baltimore County. John Olszewski Sr. has represented Dundalk and parts of Essex since 1998 in District 7. Joseph Bartenfelder has represented Middle River and parts of Essex since 1994 in District 6. Both Olszewski and Bartenfelder have been key government actors in facilitating the redevelopment of first-tier suburbs in their districts. They have worked together closely in a coalition comprised of County Executives Dutch Ruppersberger and Jim Smith,

162 / transforming race and class in baltimore

Office of Community Conservation Director Mary Harvey, and various private developers in the region to make the first-tier suburban renaissance initiative a reality. Two implications of Baltimore’s strong county government stand out as key to making these projects possible. First, Baltimore County maintained several important powers that allowed the county to confront suburban decline. Since there were no municipalities, the county had exclusive zoning and planning powers over the entire territory of Baltimore County, including the first-tier suburbs. Many of the projects required substantial land use authority. These powers included the ability to rezone land to other uses, condemn properties, and conduct countywide planning. These powers were essential to Baltimore County’s suburban revitalization efforts. Second, Baltimore County was able to redistribute fiscal resources on a countywide level. This gave the county a substantial “throw-weight,” or the ability to redirect revenue from one source to another. The size of Baltimore County and the relative affluence of the outer suburbs enabled the county to redistribute funds for suburban revitalization. The county was able to capitalize on its large tax capacity because it levied both income and property taxes over a large territory for a large and relatively affluent population. With more than 760,000 residents, Baltimore County had a substantial tax base that generated more than $1 billion in revenue for the county in 2000. Essentially, the redistribution of public funds from the wealthier outer suburbs of Baltimore County partially subsidized the renewal projects in the poorer first-tier suburbs. This occurred because of the absence of municipalities in Baltimore County. All these suburbs were under a single jurisdiction—the county government. Hence, this provided Baltimore County the opportunity to direct a portion of its revenue toward suburban renewal. Lack of Political Fragmentation Baltimore’s first-tier suburbs have only two local governments, making them atypical among the nation’s larger metropolitan areas. There are no municipalities, independent school districts, or special districts in Baltimore County. Anne Arundel County has only two incorporated cities and no independent school districts or special districts. This means that there is virtually no political fragmentation in Baltimore’s first-tier suburbs. In other words, local government is centralized at the county level. Table 5.2 displays the number of local governments and the corresponding population for each suburban jurisdiction in the Baltimore region. There are five county governments, which contain 13 municipalities. As table 5.2

suburban renaissance / 163 Table 5.2

Suburban government structure in metropolitan Baltimore, 2000

Jurisdiction Anne Arundel County Baltimore County Carroll County Harford County Howard County Total

Number of municipalities County population Municipal population 2 0 8 3 0

489,656 754,292 150,897 218,590 247,842

33,339 0 35,224 32,621 0

13

1,861,277

101,184

Source: U.S. census.

shows, only a small fraction of Baltimore’s suburban population resides in these municipalities. As I discuss, the absence of political fragmentation is fundamentally important for addressing the decline of first-tier suburbs (Culver 1982). In stark contrast, examples from two other regions underscore the role of the structure of local government in confronting suburban decline. The St. Louis and Pittsburgh regions are two of the most politically fragmented metropolitan areas in the nation, and they have also been extensively studied. For instance, in greater St. Louis, Orfield (2002) estimated that there were 312 independent local governments, or 12.2 local governments per 100,000 residents. In St. Louis County alone, there were 91 municipalities, 28 townships, 24 school districts, and 24 fire districts. Greater Pittsburgh shares a similar dynamic. Orfield (2002) estimated that there were 418 local governments, including 412 independent municipalities and townships. This amounts to 17.7 governments per 100,000 residents in greater Pittsburgh. Such widespread political fragmentation throughout metropolitan America may very well undermine any government’s ability to confront suburban decline. In fragmented regions, the population is divided among many small municipalities (Burns 1994; Miller 2002; Teaford 1979). The consequences of this feature of American local government have been the topic of immense scholarly debate for decades (Nelson and Foster 1999). Indeed, the academic and policy debates on fragmented versus centralized regional governments still range from centrist and polycentrist perspectives to regionalist views (Phares 2004). The centrists hold that large, consolidated governments are more efficient for the administration of local services while polycentrists argue that small, fragmented governments are preferred since competition among local units attracts economic growth (Ostrom, Tiebout, and Warren 1961). In the 1990s, the “new regionalist” perspective emerged as a new wave of local governance (Norris 2001a).

164 / transforming race and class in baltimore

New regionalists hold that a larger regional governance system is needed to manage local problems that transcend political boundaries such as traffic and air pollution. With this perspective, we can view suburban decline as a regional problem since it spreads through neighborhoods of numerous first-tier suburbs. Without a regional system or some centralization of local government, potential challenges rise when confronting suburban decline. First, local governments may not benefit from economies of scale, which allows larger areas to collect the revenue necessary for suburban revitalization. In some instances, it should be noted that there is evidence that without a regional system or local centralization of government, costsavings and efficiency of public programs still might be possible through techniques such as interlocal agreements and informal cooperative network (Carr and Feiock 2004; Feiock 2004; 2007; Frederickson, Johnson, and Wood 2003; Ostrom, Gardner, and Walker 1994; Savitch and Vogel 2000; Visser 2002). Second, redistribution of public funds may not occur to the same degree that it would in a centralized government structure, and greater fiscal inequality among local governments is likely. Third, in fragmented systems, each local government maintains independent zoning and planning authorities. This means that suburbs that want to confront decline are able to although other suburbs do not have to do anything. Last, the political will to confront suburban decline varies largely among the elected officials of the many different municipalities. In a sense, the first-tier suburbs of fragmented regions can best be described as competing in an environment of the “survival of the fittest” places—a perverse outcome of the Tiebout (1956) world. In stark contrast, Baltimore County has no municipalities, and there are several advantages that single local governments have in confronting suburban decline. First, the centralization of local government means that Baltimore County was able to collect revenue from a single, broad tax base and redistribute resources from wealthier to poorer communities. Second, Baltimore County’s single government status means that the county had various planning powers such as the ability to zone land, guide development (and redevelopment), and determine public infrastructure (e.g., water and sewer) for the entire territory of Baltimore County, including the first-tier suburbs. So, Baltimore County did not have to persuade other local governments to participate in the planning process for revitalization. Third, the lack of political fragmentation meant that Baltimore County leaders could set the regional agenda for the Renaissance Initiative. Since the county had jurisdiction over all of its first-tier suburbs, it avoided the arguments and in-fighting that are often typical between county and municipal governments and among municipalities within fragmented regions. Thus, in a region without widespread political fragmentation, such as Baltimore,

suburban renaissance / 165

it will be easier for the local government to confront suburban decline. However, in both consolidated and fragmented regions, this still requires the political will, public and private resources, commitment and time to confront this problem. For the past decade, there has been strong political will to address suburban decline in Baltimore County. Yet even then, the first-tier suburbs still declined relative to the outer suburbs, suggesting that there are limitations to the extent in which local governments are capable of responding to suburban decline. Lessons Learned The implications of a strong county government structure and the lack of political fragmentation suggest that these features of local government facilitated suburban revitalization projects in the Baltimore region. Although the Baltimore region’s structure of local government may be the exception to the American norm, there are some lessons that can be drawn from the experience of how Baltimore County confronted suburban decline. A caveat is in order. The single case nature of this study does make it difficult to draw wider lessons for metropolitan America. But a reflection on the characteristics that made Baltimore’s experience with suburban renewal a success is warranted if we are to understand the political and economic dynamics associated with confronting suburban decline in a systematic method as Baltimore did. Let us consider three lessons from the metropolitan Baltimore experience about the realties of dealing with first-tier suburban decline. The first lesson is that local governments are able to confront suburban decline and revitalize aging communities, but the ability to carry out and sustain revitalization will depend on the political actors. The Baltimore case informs us about what a large jurisdiction can do under a single political unit. The centralized government structure of Baltimore County streamlined revitalization because the multiple strong county executives supported the vision of community conservation, the county council remained engaged in the process, developers were willing to invest capital, and the bureaucracy of county government was made more efficient to support suburban renewal. In short, all political actors shared the goal of renewing suburbs and worked together. These characteristics ultimately allowed the county to develop a “semi-regional” plan to address the problem because the county governed all of its first-tier suburbs. This was only possible because Baltimore County had jurisdiction over all of its territory. The second lesson is that confronting suburban decline requires the political willingness to confront the problem. Baltimore County’s leaders

166 / transforming race and class in baltimore

were serious about doing something. They recognized the problem of suburban decline, and they acknowledged decline as a public problem. For a decade, multiple county executive administrations supported the Office of Community Conservation. This gave county planners an explicit mandate to confront the reality of suburban decline and develop a strategy of renewal. Support for these projects depended on the political will of county leaders to commit to the renewal of older suburban communities. If future county executives do not support the Office of Community Conservation as strongly as the Ruppersberger and Smith administrations did, then the county’s ability to deal with suburban decline may be substantially limited. Other jurisdictions that want to renew older suburbs would be well-advised to develop the political will necessary to confront this issue. The third lesson is that when a local government has control over a significant tax base, it has a broader range of options for addressing suburban decline. Since the county levied a local income tax, Baltimore County had substantial public funds that it used for the purpose of suburban revitalization. The county’s outer suburbs housed a quarter of a million affluent residents, which generated considerable funds for the government. Baltimore County used these funds, in part, for redistributive purposes. Whether public spending is higher on a per capita basis in a consolidated political structure such as Baltimore County (versus a fragmented structure) remains a question of scholarly debate. This debate has a long scholarly tradition (Boyne 1992; Dolan 1990). Nonetheless, according to Office of Community Conservation Director Mary Harvey, there is evidence that higher levels of spending were used for distributive programs in the Renaissance Initiative (personal communication, May 12, 2005). In the Baltimore context, this meant that wealthier outer suburbs subsidized renewal projects in the poorer first-tier suburbs. This has allowed the county to invest more than $1 billion in its declining suburbs since 1995. Such a large public investment in the revitalization of older, first-tier suburbs helped sustain projects such as town center revitalization, infrastructure improvements, and homeowner initiatives in Baltimore County’s Office of Community Conservation. Without a funding mechanism and control over a significant tax base, the “menace” of suburban decline is likely to persist (Smith, Caris, and Wyly 2001). These lessons show that Baltimore County’s ability to confront suburban decline depended on several factors. A favorable economic climate and the political will to tackle the issue made suburban revitalization projects feasible. Confronting suburban decline also depended on the lack of political fragmentation, which gave the county the ability to generate substantial public funds and then redistribute them from wealthier outer suburbs to

suburban renaissance / 167

poorer first-tier suburbs. Any changes to these characteristics might very well threaten Baltimore County’s ability to carry out revitalization projects in the future. These lessons are consistent with the findings of a similar study, which concluded that regions were better off when governed by few government actors in a consolidated local structure for the regional benefit (Nelson and Foster, 1999).

This page intentionally left blank

Ch a p t e r Si x Su bu r ba n Cro s sroad s

Baltimore’s first-tier suburbs reached a crossroads with the onset of the 2000s. Urban scholar Hudnut (2003, 419) succinctly observed that the future prospects for the first-tier suburbs were unknown and could follow several paths, arguing that If [we] allow deterioration to continue, it will gradually infect other nodes of development in the region. But if they can stem the flight of blight by becoming stronger and healthier through the practice of urban acupuncture, if public policies can focus resources on their redevelopment, they will become brighter lights on the regional horizon and show others how renewal can be accomplished.

The future of the first-tier suburbs could follow a number of different roads. The first road is one that leads to the full renewal of suburbs. This road will require substantial public and private institutions to invest extensively in suburban communities to make them viable places again. The second road is one that continues the status quo. Governments that assist suburbs may continue to provide some assistance to the communities that they deem “savable,” although others are left to the private market. The third road is one that leads to more suburban decline. Doing nothing to assist suburbs that are already declining may result in the persistence of suburban decline. The fate of Baltimore’s first-tier suburbs is still unknown, but the policy responses suggest that Baltimore County has chosen to confront the realities of suburban decline. Only time will tell if that road leads to success. There is a window of opportunity to prevent additional decline. The decay of suburban neighborhoods on the urban fringe is not inevitable. Action is required. Residents must have a commitment to neighborhood stability, and politicians need the resolve to confront the issue. Strong leadership is necessary to bridge the political dialogue and public policy.

170 / transforming race and class in baltimore

Acknowledging the critical challenges of the suburban crossroads that this mature jurisdiction faces, Baltimore County (2000, 15) planners concisely summarized: Baltimore County is a county at the crossroads. It is a county that has grown substantially in population and has also seen tremendous growth in the poverty rate . . . It is an aging county, both in population and physical infrastructure. More than 70 percent of the county’s housing stock is aged 30 years or older and much of it lacking in modern amenities or suffering from some type of obsolescence. It is a county where 100,000 people are without health insurance and 47,000 households are living close to the poverty level.

Without a doubt, Baltimore County and its surrounding neighborhoods in Anne Arundel County have reached a crossroads. The road to more decline may be an easy path to follow. It requires doing nothing and maintaining the status quo. The road to success depends on the political and social willingness to confront the decline of its suburbs. Ultimately, the fate of first-tier suburbs rests with the people who call them home. The New Metropolitan Dilemma: Saving the Suburbs The first-tier suburbs are worth saving. Policymakers and planners are faced with the dilemma of determining how to save these fragile suburban communities. Despite the suburban social and economic transitions in metropolitan Baltimore, most of these communities have not deteriorated beyond repair. Although first-tier suburbs experienced socioeconomic decline since 1970, many of these places are still in-tact communities. The neighborhoods are populated with families, young and old. In fact, more than half a million people in metropolitan Baltimore still call the first-tier suburbs home. These residents have a strong sense of community, and they do not want to see their neighborhoods decline any further. In other words, they have a stake in the investment of their community. Although the results of this study showed that Baltimore’s first-tier suburbs experienced three decades of socioeconomic decline, there is evidence to suggest that the future trajectory for the first-tier suburbs is a pattern of continued decline. But public policy and planning can play a role to prevent some of the future decline. Let us consider the benefits and amenities of first-tier suburbs that make them worth saving. First-tier suburbs can capitalize on a number of features that set them apart from other suburban areas. Three features stand out as main assets for suburban Baltimore. First, the established suburban neighborhoods in Baltimore feature a strong built infrastructure. Roads, sewers, and

suburban crossroads / 171

schools are already built. Second, first-tier suburbs are home to the region’s most affordable housing stock, offering an attractive opportunity for new homeowners and younger families. Third, investing in first-tier suburbs promotes sustainable, “smart” growth by retaining Baltimore County’s Urban-Rural Demarcation Line (URDL), which offers the opportunity to encourage the revitalization of first-tier suburbs. Public policies that capitalize on these assets of Baltimore’s first-tier suburbs are likely to be effective. Infrastructure and Location The built infrastructure is one of the greatest assets of the first-tier suburbs. The physical infrastructure that supports first-tier suburbs is already in place. It does not need to be built from scratch, as it does in the outer suburbs. One of the unattractive features of landscape in the outer suburbs is that the entire landscape needs to be newly built. In contrast, the transportation network of roads and highways already exists in the first-tier suburbs. In addition, many of the first-tier suburbs already have a sturdy housing stock that was well constructed. The neighborhoods feature parks and open green spaces that are often connected with sidewalks. First-tier suburban neighborhoods also contain established hospitals and schools in close proximity to residential areas. These public investments in the infrastructure have been long paid off over several decades. The upkeep and even expansion of the existing infrastructure remains a considerably less expensive alternative than constructing a new network of infrastructure from the ground up in the outer suburbs (Puentes and Orfield 2002). Although it is necessary to provide maintenance and modernization of the infrastructure, this is still an attractive fiscal option. First-tier suburbs are also well-positioned in the metropolitan area. Hudnut (2003, 1) fittingly refers to this advantage as “halfway to everywhere.” He also observes that first-tier suburbs are “regional pivot points, centrally located in the metropolitan mosaic” (Hudnut 2003, 419). Indeed, Baltimore’s first-tier suburbs are no exception. They are strategically located halfway to nearly every amenity in the region. For instance, they are located halfway to regional shopping centers and employment nodes in the outer suburbs. They are also located halfway to downtown Baltimore, providing easy access to a premiere health care system as well as cultural and entertainment amenities. The proximity to highway networks and mass transit alternatives makes the first-tier suburbs’ location to “everywhere” even easier to access the amenities throughout the region. The promotion and revitalization of first-tier suburbs can build on these features. Making the most of these two assets can facilitate those processes.

172 / transforming race and class in baltimore

In short, a readymade infrastructure and a strategic location are profitable assets for these places. Affordable Housing The presence of affordable housing is another asset for the first-tier suburbs. They boast the largest supply of affordable housing in the region. I briefly chart out recent trends in the regional housing market, and then I examine the trends of affordability in the first-tier suburbs. The national housing boom had a large impact on the real estate market in the Baltimore region. Fair market rents (FMR)—a common metric used as a barometer for measuring housing affordability—soared throughout metropolitan Baltimore in recent years. The FMR is a regional dollar figure that HUD establishes annually for metropolitan areas, which is publicly available in HUD’s State of the Cities Data System (2006). The HUD FMRs from 1996 to 2006 for a 2-bedroom apartment in metropolitan Baltimore ranged from $599 to $950, an increase of 59 percent over a decade. During the 1990s, FMRs rose very slightly each year. But, in the early 2000s, FMRs increased significantly. In contrast, personal incomes did not keep up with the rising cost of housing. For example, between 2002 and 2003, there was a 21 percent increase in the FMR in metropolitan Baltimore although regional median household income only rose by 2 percent during that same period. In other words, the average householder’s outlay for housing expenses outpaced the intake of income tenfold. Thus, the cost of living increased substantially more than personal incomes—paving the way for a regional housing market that fails to produce sufficient levels of affordable housing (Vicino, Sharkey, Regan, et al. 2005). Residents of suburban Baltimore were increasingly priced out of the market and could no longer afford the very housing units they occupied. Some 26 percent, or 49,151 households, in first-tier suburbs lacked affordable housing in 2000. Similarly, 24 percent, or 76,935 households, in the outer suburbs lacked affordable housing. Overall, one-fifth of homeowners and one-third of renters lacked affordable housing in suburban Baltimore. In a recent study of affordable housing in Baltimore, Vicino, Sharkey, Regan, et al. (2005) found that a disproportionate number of low-income residents experienced a housing burden. Half of suburban renters earning less than $34,000 lacked affordable housing, and among very low-income renters, earning less than $19,000, more than 80 percent lacked affordable housing. Despite the overall lack of affordable housing in suburban Baltimore, there are marked differences in the quantity of affordable housing units among various suburbs. First-tier suburbs are home to some of the region’s

suburban crossroads / 173

most affordable housing units. A recent in-depth analysis by the Baltimore Sun revealed that older suburbs near Baltimore City contained more affordable housing than outer suburbs located in other counties (Hopkins, 2005). The study’s main finding was that Howard and Anne Arundel Counties—home to the region’s outer suburbs and most expensive housing stock—contained the most number of houses priced more than $600,000 and the least number of houses priced less than $199,000. In fact, the study also found that Anne Arundel County contained only 2 zip codes in which housing was priced less than $199,000, and Howard County contained none. Approximately two-thirds of housing units in the firsttier suburbs were priced less than $199,000 in 2005. More than 90 percent of the housing units were priced less than $299,000. Only two zip codes in the northern first-tier suburbs had housing prices that approached those in the outer suburbs. Overall, a growing number of residents in suburban Baltimore lacked affordable housing during the late 1990s and early 2000s. Low- and middle-income groups were disproportionately impacted, and as a result, they experienced a housing financial burden or were squeezed out of the market. At the same time, first-tier suburbs housed many more affordable units compared to the outer suburbs. Given these trends, the first-tier suburbs are much more affordable places to live. In a tight, high-priced regional housing market like Baltimore, the first-tier suburbs can take advantage of the affordability of its housing stock. They can provide homeownership to families who would otherwise be unable to own a house. The first-tier suburbs can provide starter homes for young couples and new families. The affordability of the first-tier suburbs is an important advantage in a high-demand real estate market in Baltimore. Sustainable Growth The last asset of the first-tier suburbs is that they promote sustainability. Investing and promoting first-tier suburbs allows for the overall sustainability of the Baltimore metropolitan area. According to the United Nations’ World Commission on Environment and Development (1987, 1), sustainability refers to “development that meets the needs of the present without compromising the ability of future generations to meet their own needs.” In the Baltimore context, investment in the redevelopment of first-tier suburbs provides an opportunity to develop the region in a sustainable manner. Baltimore County has used a growth boundary to promote sustainable practices. The implication for first-tier suburbs is that urban growth boundaries encourage infill development and promote the investment of first-tier suburban areas (Johnson 2001; Stoel 1999).

174 / transforming race and class in baltimore

Consider the case of one of Baltimore County’s sustainable growth policies. The URDL divides urban and rural land in Baltimore County through land use policies and zoning (Baltimore County, 2000). In effect, the URDL acts as an urban growth boundary. It is a boundary that encourages development within a defined urbanized area while preserving land outside of the URDL. The county government uses the URDL for three specific purposes: (1) preserve land and green space; (2) protect water resources; and (3) limit development to already established communities (Outen 2005). Baltimore County’s (2000, 219) Master Plan 2010 reinforces the importance of the URDL as mechanism “to protect agricultural and sensitive environmental areas of the rural county from development encroachment.” Rural areas do not have public water and sewer systems, and large lot zoning requirements effectively limit urbanization. The URDL’s important role in shaping suburban development cannot be understated. All of the first-tier suburbs in Baltimore County are located within the URDL. The growth boundary meant that approximately 90 percent of the county’s population remained inside the URDL since its inception in 1967. For nearly three decades, the URDL allowed Baltimore’s suburbs to remain highly urbanized places, clustered around the Baltimore Beltway. In particular, the first-tier suburbs in the URDL were hubs of economic activity and home to the majority of the region’s population. Yet, as this study demonstrated, many socioeconomic transitions unfolded during this period. In particular, newer residents that sought newer development left Baltimore County. They leapfrogged into surrounding counties since most of Baltimore County was off-limits to new development. Consequently, Baltimore County became the slowest growing county in the metropolitan area by 2000. The URDL has been a mixed blessing for Baltimore County. It allowed the region’s first-tier suburbs to flourish during their growth spurt. In addition, green space and water resources remained protected. Yet once first-tier suburbs fully matured, they were no longer places of destination for all suburban residents. This study demonstrated that some residents stayed in the first-tier suburbs while many more located in newer developments in suburbs further away from both the first-tier suburbs and the urban core of Baltimore City. Thus it is necessary to provide several caveats on Baltimore County’s URDL. Although the majority of first-tier suburbs are located in Baltimore County, 5 of the 21 suburbs are located in Anne Arundel County. A separate political unit governs these suburbs, and they are autonomous from Baltimore County. As a result, local government functions such as planning and zoning are carried out in a different fashion. They may not be coordinated with Baltimore County, and the URDL’s primary impact

suburban crossroads / 175

on first-tier suburbs is limited to those under the jurisdiction of Baltimore County. Another caveat relates to the strength of political support for maintaining the URDL. Recently, there has been increased pressure to expand the boundary outward—and thus weaken—Baltimore County’s URDL. For example, every four years, the Baltimore County Council reviews all petitions to change the zoning and use of land outside of the URDL. During the latest review cycle in 2004, there were a record number of proposals to change the boundary. More than 200 independent requests to push the URDL out farther into the county’s rural areas were submitted to the county council (Wheeler and Goldberg 2005). For instance, a request to extend public water and sewer lines outside of the URDL for a new 200-unit suburban development was narrowly defeated. These cases demonstrate that the URDL was increasingly fragile and vulnerable to new development pressures. In summary, the policy and planning process can be used to improve first-tier suburbs. Baltimore’s first-tier suburbs have a number of assets that make them worthy as places of investment. They feature an infrastructure of roads, sewers, and schools that is already built. They are home to the region’s most affordable housing stock. In addition, they help to promote sustainable growth. First-tier suburbs can capitalize on meeting the needs of today’s residents, and investing in these communities can ensure that they remain places of destination for future generations to come. Policymakers and planners should take advantage of these assets to promote first-tier suburbs and stabilize socioeconomic decline. The new metropolitan dilemma is to figure out creative and innovative ways to halt the cycle of suburban decline. First-tier suburbs have suffered from socioeconomic decline in metropolitan Baltimore for three decades, and policymakers and planners need to confront this dilemma to save the suburbs. State and Federal Policy Initiatives Taking a broader perspective, Baltimore County was still relatively limited in what it could do about suburban decline. Its efforts were focused on physical infrastructure in primarily three first-tier suburbs. The county benefited from state programs that aided its ability to revitalize firsttier suburbs, and federal programs also offered assistance. In the case of Baltimore County, the approach has largely been local. In other words, local leaders have used local funds to address what is perceived to be local issues. Yet other possibilities for state intervention exist, and interactions between state and federal government with local governments represent other paths to confronting suburban decline. They merit brief attention,

176 / transforming race and class in baltimore

and in the following sections, I reflect on the role and interaction of state and federal government with local government in confronting the decline of first-tier suburbs. State Government During the 1990s, Maryland’s state government pursued several policy initiatives that sought to change the pattern of urban development and revitalize older suburban areas. In particular, Maryland’s Smart Growth Initiative explicitly focused on promoting first-tier suburbs as places of residence and investment. In this case, the state implemented urban development policies that sought to guide new development and investment to the already established communities, many of them located in the first-tier suburbs. Here, I review these policies, and then I offer a critical assessment on the role of state government in confronting the decline of first-tier suburbs. For nearly a decade, Maryland has been heralded as a national leader in managing the growth of its urbanized areas (Burchell, Listokin, and Galley 2000; Cohen 2002; Downs 2001). In 1997, under the mantra of “smart growth,” then Governor Parris Glendening developed a state program to fight urban sprawl. The Smart Growth Areas Act of 1997 sought to encourage the investment and revitalization of cities and established suburbs. This legislation provided state funding for transportation, housing, economic development, and environmental projects to “priority funding areas” (PFAs). In Baltimore County, the PFAs were strictly located only in the first-tier suburbs. These were areas that already had existing development, and they covered all of the first-tier suburbs of Baltimore. To enforce the smart growth policies, the legislation allowed the governor to deny state funding for projects that demonstrated the potential to exacerbate sprawl (Cohen 2002). Despite the national recognition that Maryland received, various scholars have noted that Maryland’s smart growth policies are severely limited in what they can realistically do. Cohen (2002) points out two critical limitations in Maryland’s smart growth program: (1) it preserves local autonomy and (2) it favors incentives over regulations. Similarly, Knaap (2001) argues that smart growth will continue to be largely ineffective as long as local governments maintain exclusive zoning and planning powers. Indeed, smart growth policies are only as strong as the willingness of Maryland’s counties and municipalities to support them. For example, if a jurisdiction does not want to embrace smart growth’s plan for mixeduse, high-density communities, then it can simply prohibit that type of development through its own zoning laws. In effect, this diminishes the

suburban crossroads / 177

state’s role in smart growth. Given the popular support for control over local zoning, it is unlikely that county governments will relinquish these coveted powers any time soon (Gainsborough 2001b). Furthermore, smart growth policies in Maryland do not regulate land use; they only provide incentives by making funds available for projects. According to Porter (1999), without the state explicitly regulating the use of land, smart growth policies are too weak to alter the pattern of development. If the state withholds funding to a local government for a project that it deems as a stimulus for sprawl, there is nothing to stop the local government from pursing alternative sources of revenue to fund the project. Therefore, financial incentives alone are not strong enough to slow urban sprawl. Owing to these two limitations, the state’s first-tier suburbs have largely not benefited from smart growth policies. Maryland’s version of “smart growth” does not appear to be an appropriate tool for stopping suburban decline unless substantial changes are made to overcome these limitations. It is important also to note that Maryland’s smart growth initiatives have not been immune to politics. Although former Governor Glendening voiced large support and provided substantial funding for smart growth initiatives from 1995 to 2003, his Republican successor did not. Under the administration of Republican Governor Robert Ehrlich from 2003 to 2007, support and funding for smart growth significantly waned. In 2003, Ehrlich authorized the Governor’s Office of Smart Growth to be dismantled. In addition, in the wake of a recession and a reallocation of the state’s resources based on the governor’s priorities, Maryland’s Departments of Housing and Community Development (DHCD) and Planning suffered from large budget cuts during the early 2000s (Wheeler 2005b). By 2007, the state’s newly elected Democratic governor, Martin O’Malley, reinstituted the Governor’s Office of Smart Growth and provided additional funding for such programs with the support of a Democratically controlled state house (Green 2007). This example highlights the political fights and philosophical divides between Republicans and Democrats in Maryland. It also demonstrates that support for smart growth, in part, is a function of the state’s economic climate and the political will of the state’s leaders to use the program as a tool for suburban revitalization. Without substantial increases in funding, explicitly for first-tier suburbs, and a strong commitment from the state’s leaders, the impact of smart growth on first-tier suburban revitalization is tenuous at best. As a result of the lack of land use regulation and a regional zoning authority, Maryland’s smart growth initiative remains a weak tool for revitalizing first-tier suburbs. Thus far, the state has not been able to prevent the decline of first-tier suburbs because it has not provided the

178 / transforming race and class in baltimore

proper zoning powers that are necessary to alter the pattern of urban development, and political leaders throughout the state have not maintained a continuous commitment to these programs. Federal Government In national terms, the federal role on urban policy has not only focused on the development of metropolitan infrastructure and the renewal of older central cities (Wolman and Agius 1996), but it has also historically focused on metropolitan development (Teaford 2008). Specifically, the federal government spurred the development of suburbs through a number of policy initiatives including Federal Housing Authority (FHA) loans, mortgage interest tax deduction, and the construction of major highways. Urban historian Kenneth Jackson (1985) provides a detailed description of these policy initiatives in his landmark book Crabgrass Frontier. Thus, although the development of suburbs garnered substantial federal attention, neither the decline nor the revitalization of first-tier suburbs has received the attention of the federal government until very recently (Puentes and Orfield 2002). In 2005, a bill was introduced in the 109th U.S. Congress to revitalize first-tier suburban communities. This policy proposal has significant implications for first-tier suburbs because it represents the first time that the federal government explicitly recognized the problem of suburban decline. On May 12, 2005, Senator Hillary Rodham Clinton (D-NY) introduced the Suburban Core Opportunity Restoration and Enhancement (SCORE) Act (S.1024) with a companion bill in the House of Representatives (H.R. 2347), which was cosponsored by Representatives Peter King (R-NY) and Carolyn McCarthy (D-NY). This was a federal policy proposal that specifically targeted declining first-tier suburban communities by providing grants for revitalization projects. Clinton and her colleagues pointed to a growing body of research that demonstrated that first-tier suburbs were declining and exhibiting symptoms of urban decay. The impetus for this legislative proposal was to call attention to the problem and prompt federal action, with the goal of stalling the cycle of suburban decline that became prevalent in many northeastern and midwestern metropolitan areas by the 2000s. Senator Clinton acknowledged that federal government action was critical because first-tier suburbs were caught in a “policy blindspot.” In other words, current public policies did not address the problem of suburban decline. Clinton recently commented that “we need to spur revitalization in declining suburban areas before they hit rock bottom, but most first-[tier] suburbs don’t quality for existing federal programs . . . the

suburban crossroads / 179

SCORE Act is what federal economic development programs should be: send money and a roadmap giving communities the momentum they need to help themselves” (Clinton 2006). Moreover, Clinton even went on to say that first-tier suburbs were “frozen out” of the majority of the federal government’s support programs. These suburban communities, for example, have typically been excluded from the largest federal redevelopment programs during the 1990s, such as Empowerment Zones, HOPE VI, and Community Development Block Grants (CDBG). Each of these federal programs provided substantial funding to urban areas that faced critical challenges such as extreme poverty, decrepit housing stock, and lack of employment opportunities, but many suburban areas with similar problems have been excluded. Yet it should be noted that several other federal urban policies, in particular transportation policies, have benefited suburbs during the postwar boom. For example, one of the largest grants-in-aid programs is the federally funded Highways Trust Fund program, which funds the construction of the state-owned Interstate Highway System. Approximately $32 billion is provided as grants-in-aid to states to construct and operate highways, and other funds are provided for mass transit as well. Thus, many national urban policies have been aimed at central cities, and some federal policies with substantial funding in housing and transportation contributed to the growth of suburbs. Although first-tier suburbs exhibited symptoms of urban decay, they faced a unique set of issues that were suburban in nature—many of them had not declined as much as their central city counterpart (Hudnut 2003). Since the federal government had not taken a large role in addressing these issues, this prompted Clinton and her colleagues to introduce the SCORE legislation in 2005 to the U.S. Congress. As the first federal policy that would directly target first-tier suburbs, the SCORE Act set out to provide economic and tax incentives for the redevelopment of ailing first-tier suburbs throughout the nation. At the heart of the legislation was a $250 million “Reinvestment Fund.” Similar to other federal trust funds, the SCORE Act aimed to provide grants to first-tier suburban communities. The goals of the act were to create employment opportunities, develop housing, and expand business opportunities for residents of first-tier suburban neighborhoods. To meet these goals, the legislation sought to create “SCORE Project Areas” to determine the eligibility of suburban communities. To qualify for funding, the legislation held that first-tier suburban communities had to satisfy several criteria, including: 1. The community had to be ineligible to be designated a “renewal community” by the IRS Code of 1986 under Section 1400E;

180 / transforming race and class in baltimore

2. A representative of the first-tier suburb’s local government had to be nominated from the community to the Secretary of Housing and Urban Development; 3. The first-tier suburb could not be smaller than a census designated place; 4. The first-tier suburb had to be located near some of the following: existing development and infrastructure, substandard or underutilized residential, commercial, an economically obsolescent regional mall (known as a “greyfield mall”) or industrial properties; and 5. A SCORE Advisory Committee of local stakeholders had to be created. The SCORE Act did not advance in the 109th Congress. Although it was read aloud twice before the U.S. Senate Finance Committee, a vote was never held on the legislation. In essence, the SCORE Act would have begun to promote a makeover of first-tier suburban neighborhoods. It was the first cut at creating a national program that could potentially become a comprehensive policy and planning tool to tackle the decline of first-tier suburbs. This bill is relevant because it represents the first step toward confronting a growing set of socioeconomic problems in first-tier suburbs. Through the legislative process, the problem of suburban decline was defined and recognized as a public problem (Kingdon 1997; Rochefort and Cobb 1994). The implication is that at the very least, the bill prompted a national discussion on the problem of suburban decline. At most, the SCORE Act may be revamped and reintroduced into the 110th Congress. In short, the issue of suburban decline transcends local and state boundaries—it is a federal issue that has yet to be tackled systematically at the federal level (Weir, Wolman, and Swanstrom 2005). Despite this initial step in confronting suburban decline, a number of major shortcomings plagued the SCORE Act. Foremost, the problem is that the SCORE Act was simply a bill—it did not become law. The bill’s failure did not bring any substantive policy to first-tier suburbs other than the definition of a public problem and raising the visibility of first-tier suburban issues. Another shortcoming was that the legislation provided only a negligible amount of public money. Thus far, federal legislators have lacked the political will to commit the funds necessary to engage in a large-scale revitalization effort in first-tier suburbs. If the federal government is serious about confronting suburban decline, then legislation to confront suburban decline is needed. Indeed, a larger federal role in first-tier suburban revitalization might counter the limitations of the ability of local and state government to respond to suburban decline.

suburban crossroads / 181

In summary, the state of Maryland’s initiatives have focused on smart growth and the federal government historically focused on central cities as well as housing and transportation systems in the suburbs. New federal legislation on suburban decline has yet to become law. Given the overall lack of progress in confronting suburban decline at the state and federal levels of government, it is worth contemplating briefly how state and federal policies might interact with local policies to revitalize first-tier suburbs. For instance, are different levels of government more suitable for funding or administering certain types of suburban revitalization efforts? Similarly, could federal government funding for infrastructure projects compensate underfunded, yet politically popular, local revitalization efforts in jurisdictions with smaller tax bases than Baltimore County’s? Questions of this nature abound. It does seem appropriate for the federal government to consider a similar grants-in-aid program to declining suburbs for expensive revitalization infrastructure projects as well as social programs to improve the economic status of residents in declining areas. Suburban decline is an emergent social problem, and its path of destruction knows no political, economic, or social boundaries. In short, these are important questions and issues that state and federal policymakers need to consider to confront suburban decline. A Suburban Policy Agenda Baltimore County stands out as a national example of what a centralized local government is capable of doing in terms of first-tier suburban revitalization. The government was able to launch a decade-long suburban revitalization effort after three decades of social and economic decline. While this case may be atypical for the very reason that few other jurisdictions or states in the nation have developed such an approach to tackle the decline of older suburbs, policymakers and planners elsewhere might still examine the Baltimore experience to learn about the realities and limitations that a mature county faced in confronting suburban decline. A number of implications prevail for various levels of government. At the local level, suburban political leaders need to figure out ways to cooperate, especially to address suburban decline (Foster 2001; Katz 2000; Orfield 2002; Rusk 1995; 1996; 1999). In the case of metropolitan Baltimore, confronting suburban decline depended on the willingness of various political actors in a single jurisdiction to address the problem. This was easier to do in Baltimore County because political leaders used the county’s large tax-base to redistribute funds for first-tier revitalization projects. So, the challenge for other more fragmented regions is to provide funding mechanisms that will be redistributive in nature. One such example

182 / transforming race and class in baltimore

is the tax-base revenue sharing program in Minneapolis. In this case, a metropolitan council redistributes public funds from high-tax capacity suburbs to low-tax capacity suburbs. Wealthier communities provide poorer communities with more resources to deal with decline (Orfield 1997). However, the Minneapolis case is only one instance, among 322 metropolitan areas nationally, where such a program worked. The Metro Council in the Minneapolis-St. Paul region redistributes tax revenue. In various cases, Minneapolis has actually been a net loser based on the redistribution of a portion of the industrial and commercial property tax (see Orfield 1997; Rusk 1999). Furthermore, Nunn and Rosentraub (1997, 17) call attention to the challenges of metropolitan political cooperation, observing that “the greatest success or returns from interjurisdictional programs accrue, as might be expected, from those that present the greatest challenge to local jurisdictions (for example, revenue sharing).” Other regions have overcome these challenges by developing voluntary tax-base sharing agreements. For example, the New Jersey Meadowlands Commission was established in 1970 to pool the resources of 14 municipalities in 2 urbanized counties to grow as a common region. The goal was to reduce inequality among local governments and to decrease the fiscal incentives for them to squander resources competing for new development. The commission ensures even regional development by redistributing a proportionate share of the property taxes from new development, regardless of municipality. In a similar case, Montgomery County, Ohio, including Dayton and its 29 surrounds municipalities, established a program to share the benefits of economic development. It provides two methods for redistributing funds: a countywide funding pool for economic development projects and a government equity fund that shares a portion of growth in tax revenues. In this case, local governments share in the economic benefits that result from new economic development among the jurisdictions in Montgomery County, Ohio. A number of other regions and states have similar tax-base and revenue sharing programs such as Louisville-Jefferson County, Kentucky, Ohio, Michigan, Virginia, and Wisconsin. These voluntary programs, although small in scale, do serve as both a model and a mechanism for local governments to share the benefits of regional cooperation. Yet the reality is that to confront suburban decline effectively, there is a need to develop a politically feasible mechanism fully to fund initiatives like tax-base sharing. However, an important caveat is in order. The scale and scope of voluntary, local programs is quite limited. The political will to maintain local autonomy is likely stronger than the will to eliminate the real barriers—fragmented governments—to regional cooperation and renewing first-tier suburbs (Norris 2001a; 2001b). These limitations

suburban crossroads / 183

on local jurisdictions suggest that other levels of government might also provide an opportunity to address the problem of suburban decline. At the state level, it is essential for Maryland, as well as other states confronting similar issues, to realize that the decline of first-tier suburbs will continue in the absence of meaningful regional growth management policies. For instance, a growth boundary that is truly regional in nature would limit leapfrog development and uneven growth that contributes to the demise of first-tier suburbs. Without a true regional growth management policy, this pattern will certainly continue. It is plausible that greater public investments will benefit growing outer suburbs rather than declining first-tier suburbs. Lucy and Phillips (2000, 21) aptly refer to this as the “tyranny of easy development.” The new suburban reality is that market-driven development will continue to move outward as the aging process in first-tier suburbs continues. Furthermore, without a regional land use policy that overrides local zoning powers and without generous public funding for programs like the Smart Growth Initiative, it is likely that Maryland’s efforts, and those of other states, will not effectively combat the decline of first-tier suburbs in the long term. Although regional policies are not panaceas for urban growth, there is tremendous pressure to continue the decentralization of population and to build out new suburban areas. In the absence of such regional tools, it is imperative that governments provide substantial funding for the reinvestment of established communities in the first-tier suburbs. At the national level, a larger federal role may overcome the funding limitations and zoning issues that limit local and state governments from renewing older suburbs. The SCORE Act does not provide ample funding to confront suburban decline across the country practically. Although the future of the SCORE legislation is still unknown, introducing the SCORE Act in the U.S. Congress was a valuable effort on the part of legislators. It prompted a national discussion on first-tier suburbs, and it brought attention to the problems of declining first-suburbs to many national leaders. An increased federal role in suburban revitalization may also very well begin with the development of new institutional arrangements for dealing with problems such as suburban decline that transcend federal, state, and local government (Colman 1965). An advisory commission to the congress and president or special assistant secretaries in executive agencies such as HUD may help develop new creative intergovernmental solutions to suburban decline. Finally, all three levels of government should also pay particular attention to the impacts of revitalization policies on residents in first-tier suburbs. A well-balanced approach should incorporate not only physical, place-building policies, but it should also integrate people-based, human

184 / transforming race and class in baltimore

capital policies that seek to improve the social and economic well-being of residents. This study, as well as other national studies, has shown that suburban decline is fundamentally a housing problem and an income problem. Residents of declining suburbs live in houses that are losing value and in desperate need of repair and renovation. Moreover, the same residents are the poorest suburban population in the region. Housing loan and grant programs provide a basis for building stronger communities, but income assistance programs would also help stabilize residents with declining incomes. An intergovernmental approach is needed for such a complex problem that spans multiple political jurisdictions and multiple facets of one’s life. Grave social and economic problems now plague suburban areas. Without substantial changes to public policies and proactive, forwardlooking policies and planning, the forces of suburban decline and the outward decentralization of that decline are likely to continue. This is the defining new feature of the suburban landscape, and these are the realities of confronting suburban decline. First-tier suburbs offer many features that make them attractive places to save and rebuild. Bridging the new metropolitan dilemma requires thoughtful and calculated policies and planning that take advantage of first-tier suburbs’ location and infrastructure, affordable housing stock, and potential for sustainable growth. At the end of the twentieth century, suburbs had reached a crossroads, and they were confronted with multiple roads to drive toward their future. Whether these suburbs ultimately decline or sustain still remains a question of public policy, planning, and political leadership.

Ch a p t e r Se v e n Su bu r ba n P ro sp e c ts

American suburbanization is a social, political, economic, and above all, a spatial phenomenon. It is the process of human population movement that is the result of private actions and public decisions. The suburb is the distinct and defining characteristic of American urban development that came to life during the second half of the twentieth century. Early studies of American suburbs focused on comparisons between city life and suburban life. Social and cultural critiques abound. Gans (1967), Whyte (1956), and Mumford (1968), for instance, criticized suburbia as conservative, monotonous, and even superficial. Indeed, suburbs in the 1950s and 1960s were overwhelmingly white and middle class, and some places deserved these early criticisms. For example, the Levittown suburban developments in New York, New Jersey, and Pennsylvania exemplified the early suburbs as collections of houses, streets, and cars—there was virtually nothing else. These suburbs were dependent on the central city for jobs, recreation, shopping, and services. Early critics of the suburbs alleged that the banality and the lack of diversity characterized the first wave of suburbanization immediately following World War II (Kelly 1993). Yet, the period from the 1970s to the 2000s challenges these classic images of American suburban life. There is little doubt that suburbs have been changing for several decades. During these years, numerous studies, including this one, detailed the remarkable social and economic diversification of American suburbs. The growth of the suburbanization of the black population, the immigrant population, and the poor population was documented in studies as early as the mid-1970s (Masotti and Hadden 1973; Schwartz 1976). Then during the 1980s, scholars focused on the impacts of the dramatic population growth of the suburbs (Baldassare 1986; Kelly 1989). The movement to the suburbs did not falter during the 1990s. By 1990, the majority of Americans lived in the suburbs. Political scientist William Schneider (1992) fittingly observed that the United States had become a “suburban nation.” The Census 2000 revealed that some 168 million, or 6 out of 10 Americans, resided in suburbs. Since

186 / transforming race and class in baltimore

World War II, American suburbs have experienced phenomenal growth, and that growth ultimately ushered more social and economic diversity. The evidence of suburban diversity—a true mosaic—shows that the classic image of 1950s suburban life is now just a distant memory. Reflections on Baltimore The case of Metropolitan Baltimore’s first-tier suburbs offers a prime example of suburban change. The region’s suburbs witnessed a marked transformation between 1970 and 2000. The descriptive spatial statistical analysis charted suburban changes in Baltimore’s first-tier suburbs by comparing them with other geographic scales, which included the outer suburbs, the central city, and the region. In conclusion, the analysis revealed four trends about the population, income dynamics, nature of the housing stock, and structure of the labor force. First, the characteristics of the population changed in several noteworthy ways. Population growth stagnated in Baltimore’s first-tier suburbs since 1970. Many of these suburbs failed to attract additional residents, and those that did grow experienced a relatively small amount of growth. The majority of these communities not only failed to maintain their population bases, but they actually lost residents. The population of first-tier suburbs also grew much older compared to the outer suburbs. The proportion of residents aged more than 65 years was significantly greater in the first-tier suburbs than the outer suburbs. Residents were aging in place; that is, as they grew older, the elderly population statistically increased while the other age cohorts declined. In terms of the race of the population, first-tier suburbs were highly segregated by race and place in 1970 and 2000. The increases of racial diversity were modest and distributed in an uneven spatial pattern. The western first-tier suburbs contained the highest degree of black segregation. In fact, the population of the western suburbs changed from majority white to majority black over the three decades. In contrast, the eastern suburban fringe housed the largest portion of a white population, although the number of black residents grew. Second, since 1970, the economic status of residents in the first-tier suburbs changed remarkably. Baltimore’s first-tier suburbs grew increasingly poorer between 1970 and 2000. The poverty rate doubled from 4 percent (25,533 residents) in 1970 to 8 percent (38,266 residents) in 2000. Household income declined by 9 percent between 1970 and 2000, from $54,530 in 1970 to $49,669 in 2000. Moreover, when compared to all other suburbs in metropolitan Baltimore, the first-tier suburbs had a higher household income than the median suburban income level in 1970. By 2000, they had a lower household income than the median suburban

suburban prospects / 187

income level. First-tier suburbs were known as desirable places to live in 1970. Three decades later, that was no longer the case. Third, the nature of the housing stock changed substantially between 1970 and 2000. The value of the housing stock in Baltimore’s first-tier suburbs declined relative to the value of housing in the overall suburbs. For example, the housing stock in the first-tier suburbs was worth 5 percent more than the average value of housing in suburban Baltimore in 1970. Three decades later, the first-tier suburban housing stock was worth 13 percent less than the average value of housing in suburban Baltimore. By 2000, first-tier suburbs had the least valuable housing stock of the region’s total suburban area. The first-tier suburbs also had a disproportionate share of older, smaller housing units. Conversely, the outer suburbs had a disproportionate share of newer, large housing units. The housing stock in Baltimore’s first-tier suburbs was no longer as desirable as it was in 1970. The analysis demonstrated that the characteristics of declining first-tier suburbs included a housing stock that was smaller, worth less, and older than the housing stock in other parts of suburban Baltimore in 2000. In short, the housing stock in the first-tier suburbs lacked the caché that it once held. Fourth, the structure of the labor force changed in a number of ways between 1970 and 2000. In 1970, a considerable portion of the labor force consisted of residents working in manufacturing and labor occupations, with a smaller group of residents working in professional occupations. By 2000, the number of residents working in manufacturing or labor related occupations declined by two-thirds. Over the 30-year period, employment in the services sector became increasingly prevalent. In general, first-tier suburban residents employed in the services sector tended to work in low-wage occupations in retail and food establishments rather than professional occupations. The level of educational attainment of residents in the labor force throughout first-tier suburban Baltimore changed over three decades relative to the outer suburbs. In 1970, the population of the first-tier suburbs had the greatest number of residents with the greatest number of years of education out of all suburbs in the region. By 2000, the population of the outer suburbs had the greatest number of residents with the greatest number of years of education. Residents in first-tier suburbs did not have as many years of education as residents in the outer suburbs. The principal components analysis and cluster analysis reinforced the findings of a first-tier suburban transformation in Baltimore. The 1970 analyses showed that class defined Baltimore’s first-tier suburbs. A clear spatial pattern separated professional-class households from the workingclass households. A working-class population resided in the eastern and

188 / transforming race and class in baltimore

southern suburbs of Baltimore, and a professional-class population resided in the western and northern suburbs. The age of housing in first-tier suburbs varied to an extent in 1970; there were new and old housing units alike. Households were mainly comprised of two-parent, homeowner families in the first-tier suburbs. The 2000 analyses showed how dramatically the first-tier suburbs had changed in 30 years. The existence of poor places became a distinguishing characteristic. With a few exceptions, housing filtered from a higher class to a lower class as the neighborhoods aged over three decades. Minority renter households were increasingly more common since 1970. The emergence of a black middle class characterized the transformation of the western first-tier suburbs. Together these analyses demonstrated that suburban places and neighborhoods were differentiated in both 1970 and 2000. In 1970, class was the differentiating feature of first-tier suburbs. Thirty years later, both class and race were differentiating features since the economic status and racial composition of the population changed largely. Contributions This study offers three important contributions to the body of work on suburban change. First, this study developed an analytical definition for first-tier suburbs. I developed a method for identifying first-tier suburbs since previous studies disagreed on what constituted a first-tier suburb. To begin, I used the census place level geography to analyze suburbs. This geography captured the social and economic cohesiveness that defined an entire suburban community. Next, I utilized two criteria for the identification of first-tier suburbs: spatial (location relative to the central city) and temporal (age of development measured by housing). In the case of metropolitan Baltimore, the application of these spatial and temporal criteria captured all the suburban places that were located within eight miles of the outer boundary of the central city. This method is a valuable contribution because it can potentially be replicated for the study of suburbs in other urbanized areas in the United States. Second, this study used a novel approach to analyze first-tier suburbs. I employed a spatially orientated analytical approach to the study of suburban change. Using data at multiple geographic scales, I was able to examine change among first-tier suburbs and within individual first-tier suburbs. The rationale for using geographical analysis was twofold. Using place level data provided the opportunity to characterize the changes among first-tier suburbs within a metropolitan area. The use of tract-level data allowed me to investigate neighborhood change within first-tier suburbs. This captured the rich socioeconomic variation of first-tier suburbs. Previous studies did

suburban prospects / 189

not detect as much first-tier suburban differentiation as this study because they did not employ urban geographical analysis using microlevel data. Moreover, the use of GIS tools allowed for the visualization, display, and analysis of patterns and differentiation. This approach serves as a methodological contribution to the study of suburban change. Third, this study characterized the process of change in the first-tier suburbs of metropolitan Baltimore from 1970 to 2000. During this period, the results of this study showed that the aggregate trend was a pattern of decline in the first-tier suburbs. The socioeconomic status of suburban places and neighborhoods changed over 30 years. Contrary to the earliest studies on suburban change (Farley 1964; Guest 1978), socioeconomic status in metropolitan Baltimore did not persist over time. Instead, a number of the patterns of change characterized the decline of first-tier suburbs. As the population aged in place and did not grow, as the household income of residents declined, as the housing stock aged and failed to compete with newer housing in outer suburbs, and as the labor market replaced well-paying jobs with lower-paying service jobs, then first-tier suburbs declined. The consequence of these changes was the replacement of these suburbs’ higher socioeconomic status populations by increasingly lower socioeconomic status populations. This pattern ultimately produced a self-reinforcing, downward spiral of decline. Explaining Suburban Change Although this study did not establish cause and effect relationships about suburban change, it is however possible to draw on the literature and this study’s finding to speculate why suburban change, particularly the specific changes discussed here, occurred. Baltimore’s first-tier suburban transformation is similar to the large-scale changes that occurred in the neighborhoods of America’s older central cities. For example, the aging of residents and a declining population, coupled with declining socioeconomic status, characterized the changes that urban neighborhoods witnessed during the mid-twentieth century. First-tier suburbs experienced similar changes. Of course, Baltimore’s first-tier suburbs did not suffer the same extent of socioeconomic decline, as did the central city. But the 30-year trends identified in this study suggest that first-tier suburban places are headed down a similar path. The trajectory forecasts a continued pattern of decline. In a sense, the decline of first-tier suburbs is just another manifestation of the aging process. Since there is no regionwide urban containment policy, new development continues to decentralize in an outward expansion to the exurbs. Under the current market system of development, new investment to the outer suburbs will continue in greater proportions than in

190 / transforming race and class in baltimore

the first-tier suburbs. Although some areas in the first-tier suburbs have recently seen growing public investments, most first-tier suburban areas continue to experience a lack of investment, or even disinvestment. This same process first took place in central city neighborhoods and produced socioeconomic decline (Beauregard 2003). Now several decades later, first-tier suburbs are showing clear signs of a similar decline—they have aged and failed to capture new suburban growth although newer outer suburbs have developed and have captured a higher socioeconomic status population. In the end, the suburban changes that were demonstrated in this study suggest that the future prospects for first-tier suburbs are tenuous at best. All things being equal, these suburbs may very well share the same fate as many declining neighborhoods in Baltimore City. A Suburban Research Agenda Although this book raises questions about suburban decline, the study of suburbs still does not garner substantial attention among U.S. urban scholars. In an era when more Americans than ever live, work, and recreate in suburbs across the nation, there continues to be a deficit of scholarly work about suburbs in general and first-tier suburban change in particular (Hanlon, Vicino, and Short 2006; Palen 1994). Most of the contemporary suburban research, especially during the 1990s, examined the impacts of urban sprawl and the role of outer suburbs or exurbs in metropolitan America (Garreau 1991; Squires 2002). In contrast, the early 2000s witnessed a small, yet blossoming body of scholarship that explicitly focused on the evolution and changing nature of first-tier suburbs (see Hudnut 2003; Lucy and Phillips 2000; 2006; Orfield 2002). A recent cover story in the Chronicle of Higher Education provides a case in point. The article calls attention to a new generation of scholars that is revisiting suburbia and its changing dynamics. The author appropriately notes that “a revisionist-minded, cross-disciplinary group of researchers [are] rereading the suburban landscape in ways that unsettle much of the received wisdom about its history and political economy” (Howard 2006, A16). Indeed, it is high time for scholars to revisit and expand the study of suburbia, in at least the following five areas. Future research should first address the role that political fragmentation (or its absence) plays in the process of suburban decline. Why do Baltimore’s suburbs still experience decline in a centralized county government structure? Do similar regions, such as St. Louis, Pittsburgh, or Cleveland, decline more because of political fragmentation? Does suburban decline vary regionally across the nation? If so, how and why do the causes of suburban decline vary? In other words, does political

suburban prospects / 191

fragmentation exacerbate the process of suburban decline? Answers to these questions would help scholars identify how and why suburban decline in different regions varies under different political structures. Developing a deeper understanding of this relationship is fundamental to the study of first-tier suburbs. It would not only contribute to the debate about whether political fragmentation undermines the social and economic well-being of metropolitan areas, but it would also inform us about the capability of metropolitan areas to address and respond to the problem of suburban decline. Second, additional research is needed to identify the determinants of suburban decline. It is plausible to develop questions and hypotheses about why first-tier suburbs experience socioeconomic decline. For instance, to what extent does the loss of manufacturing jobs, or the nature and age of the housing stock, or the economic status of residents contribute to the decline of first-tier suburbs? Although most of the studies on suburban decline suggest that these factors contribute to suburban decline, causal relationships have not been identified. Thus, the development of models that explain why these changes occur is one of the next logical steps in the progression of the study of first-tier suburbs. Third, future research should focus on comparative studies of suburbs in different metropolitan areas. Does suburban decline vary regionally across the nation? If so, how and why do the causes of suburban decline vary? Employing both quantitative and qualitative approaches, including geographic techniques, would assist scholars in examining the variation of suburban decline. Similarly, it would be interesting to compare patterns of suburban decline and growth to other regions in an international context. There is generally a deficit of comparative suburban studies, and this body of work presents many opportunities to branch out the study of changing suburbs in a broader international framework. Fourth, additional work should focus on demographic analysis and spatial change in suburban environments. Suburbs are increasingly sites of immense social and economic change, and the spatial dynamics and demographic characteristics of those changes are not well understood in the suburban context. Questions abound about the nature of place and change. For instance, are residents of first-tier suburbs becoming poor in place? Or, are new poor residents moving from other suburbs or the central city to the first-tier suburbs causing these suburbs to become poor? So, are poor residents from the central city “spilling over” into the first-tier suburbs, or are first-tier suburban residents becoming poorer in place? This study offered some evidence to suggest that first-tier suburban residents became poorer in place in the eastern suburban fringe of Baltimore, but this evidence was preliminary at best.

192 / transforming race and class in baltimore

The next decennial census in 2010 offers exciting and rich opportunities to answer some of these challenging demographic and spatial questions. For the first time, the U.S. Census Bureau now recognizes new complex metropolitan geographies. The census data categories that stratified the central city and the suburb will no longer be utilized in the upcoming census. Instead, the U.S. Census Bureau will identify larger regions as Core Based Statistical Areas (CBSAs), which will contain Metropolitan Statistical Areas (MetroSAs) and Micropolitical Statistical Areas (MicroSAs). Places and central cities will longer be used. Principal cities will be utilized to identify “core” cities within the CBSA. These new geographic data categories will allow urban scholars to make use of data that more accurately reflect today’s metropolitan landscape. As such, these new data sources will enable scholars to delve further into the complex spatial relationships in the suburbs (Frey, Wilson, Berube, et al. 2004). A fifth area of future research should investigate the policy and planning responses to suburban decline. Specifically, what are the effects of the suburban revitalization efforts on human capital or people-based policies versus physical or place-based policies? Similar to the case of Baltimore City’s revitalization and gentrification of the 1980s and 1990s, undoubtedly questions such as these can be raised in the suburban context as well (Levine 1987; Wagner 1995). Since most of Baltimore County’s key initiatives to help declining suburbs did not begin until the late 1990s and early 2000s, the full effects of first-tier suburban revitalization are still unknown. In fact, for Baltimore County, many of the revitalization projects were just under way by 1999, which was the latest year that reliable neighborhood and suburban level census data were available. Thus, it is important to develop criteria to determine whether revitalization projects worked and why. Then, the comparison and analysis of policy and planning responses at all three levels of government to suburban decline and their effectiveness would provide more insight into these questions. A rich body of scholarly work is emerging on the first-tier suburbs. It is not only reshaping how we think about the process of suburbanization but also the future prospects for the suburbs in the metropolitan landscape. It is imperative that a future suburban research agenda investigate the socioeconomic dynamics in first-tier suburbs to establish a more refined and nuanced understanding of the urban processes shaping suburban environments in the twenty-first century.

A pp e n di x

In this study, a mixed methods approach was employed to investigate how metropolitan Baltimore’s first-tier suburbs changed between 1970 and 2000. In this appendix, I present a discussion of the research questions, geographic data, data collection and organization, analysis of suburban change, and limitations.

Research Questions Three main research questions drove this study. First, how did Baltimore’s first-tier suburbs change between 1970 and 2000? Second, what did the changes mean for first-tier suburbs and their residents? Third, what were the implications of these findings for public policy and planning? Several other questions were related to these main ones. They included the following questions: 1. What is a first-tier suburb? 2. What are the principal factors associated with suburban change? (e.g., the role of economic restructuring and demographic changes related to race, age, and socioeconomic status). 3. How did the socioeconomic characteristics among first-tier suburbs, relative to the outer suburbs and the region, vary between 1970 and 2000? 4. How did the socioeconomic characteristics vary within first-tier suburbs between 1970 and 2000? 5. What can public policy and planning do about suburban change? Just as Isaac Newton once famously remarked, it is by standing on the shoulders of giants that we all benefit from seeing further. Similarly, I developed the research questions for this by synthesizing the previous work of many giants. The literature indicated that there was no clear definition of first-tier suburbs, and so I began the research for this study by crafting a definition of first-tier suburbs. Also, a clear understanding about the process of suburban changes, specifically in all the first-tier suburbs of a region, was lacking in previous studies. Furthermore, little was known about the socioeconomic variation within the first-tier suburbs. So, in this study, I asked questions about the evolution of first-tier suburbs, giving particular attention to the process of suburban change and to the implications of these changes for public policy and planning.

194 / appendix These questions are relevant to the study of suburban change for several reasons. This study is timely because it utilizes the most recent census data to analyze the socioeconomic differentiation in first-tier suburbs between 1970 and 2000. Since the Census 2000 revealed that metropolitan areas had grown more diverse than ever, both socially and economically (Berube, Katz, and Lang 2005; 2006), it is important to understand how these changes occurred in first-tier suburbs so that policy and planning implications can be developed. An understanding of the process of suburban change may also help to prevent suburban changes that are related to decline. For example, the identification of the characteristics of changing first-tier suburbs might assist planners and policymakers to develop a plan to confront such change. Finally, these questions get at the root of the problem that this study addresses: the decline of first-tier suburbs. Before moving on to a discussion of the research methods, it is important to first review several important concepts about the geographic data, which are critical for understanding how the data were collected and analyzed.

The Geographic Data In this study, it was necessary to first develop a definition for first-tier suburbs using census geographic criteria before analyzing them. This was because most urban data are scale-specific to a given geographic area, and in many cases, the same data collected at different scales can show very different trends (Brenner 2001; Short, Yeung, Kuus, et al. 1996). After defining first-tier suburbs, the quantitative data were then collected at two geographic scales: census place level and census tract level. According to the U.S. Bureau of the Census, there are three types of places: census designated places (CDPs), consolidated cities, and incorporated places. CDP boundaries are delineated to collect data on unincorporated areas with concentrations of population, housing, and commercial sites and a degree of local identity. For example, places such as Columbia and Towson are CDPs. Consolidated cities consist of two or more local governments that have merged to form a regional government. A consolidated place is a unit of local government from which the functions of an incorporated place and its county or minor civil division unit have merged. There are no examples of consolidated places in metropolitan Baltimore. Incorporated places are municipal incorporations that operate a local government other than the central city in a metropolitan area. For example, Baltimore City is an incorporated place in the Baltimore region. An incorporated place is established to provide governmental functions for a concentration of people as opposed to a minor civil division unit, which is generally created to provide services or administer an area without regard to population. For the purposes of this study, these three designations were combined to identify what some have termed “urban places” (Birch 1975; Vicino, Hanlon, and Short 2007). In metropolitan Baltimore, place level census data were available for all the geographic areas in the region. I was able to rely on place level data because the boundaries for Baltimore’s census places remained consistent for the entire period of this analysis. Several census geographies were combined to define the Baltimore region. The definition of the region varied each decade as the number of CDPs

appendix / 195 continually grew with the passing of each subsequent decade. The region comprised of five counties and Baltimore City. Queen Anne’s County was excluded from the regional definition because of the Chesapeake Bay geographical separation from the rest of the metropolitan area. The sum of all CDPs for each respective decade constituted the regional definition. In other words, the region definition was the sum of all suburbs (first-tier suburbs and outer suburbs) plus the central city. Although the U.S. Census Bureau classifies the state’s capital Annapolis as a central city, it was treated as a suburb for this study to facilitate the comparison and interpretation of the data at various geographic scales. Table A.1 shows how the number of census defined places changed over three decades. The number of first-tier suburbs and the central city (Baltimore City) remained the same over the 30 years of this study, but the number of outer suburbs grew substantially. In 1970, there were an equal number of first-tier suburbs and outer suburbs in the region. By 2000, the number of outer suburbs had grown by 376 percent, from 21 CDPs in 1970 to 79 CDPs in 2000. The growth in the number of CDPs reflected the continued population decentralization away from the urban core to the outer areas of the region. In chapter one, Figure 1.1 depicts the names and boundaries of census places for each of Baltimore’s 21 first-tier suburbs. Since there were no suburban municipalities around the first-tier of Baltimore, it was not possible to use political boundaries to define the first-tier suburbs. Therefore, I used CDPs as baseline geography for developing a definition of first-tier suburbs. The northwestern and northeastern corners of Baltimore City divide the first-tier suburbs into smaller geographic groups of western, northern, and eastern suburbs. Separating first-tier suburbs into smaller groups based on their location to the central city allowed for the generalization of findings based on these geographic cohorts. Table A.2 shows the data for the two criteria that I used to determine whether a suburb would be classified in the first-tier. Scholars have argued that first-tier suburbs, by their very nature, are located near the central city and tend to be the oldest suburban areas of a metropolitan area (Jackson 1985; Sternlieb and Lake 1975). Drawing from the literature, I developed a twofold strategy for classifying first-tier suburbs. The first criterion related to the spatial location of suburbs. The suburbs were automatically classified in the first-tier if they shared a border with Baltimore City. In this study, Table A.2 also shows that there were 12 places that bordered Baltimore City, and 9 places were adjacent to a first-tier suburb that bordered the city. The second criterion applied only to suburbs that did not share a border with Table A.1

Census designated places, 1970–2000

Geographic Scale

1970

1980

1990

2000

First-tier suburbs Outer suburbs All suburbs Central city Region

21 21 42 1 43

21 70 91 1 92

21 76 97 1 98

21 79 100 1 101

Source: U.S. Census 2000 TIGER File.

196 / appendix Table A.2

Definition of Baltimore’s first-tier suburbs, 2000

Census designated place Arbutus Brooklyn Park Catonsville Dundalk Edgemere Essex Ferndale Glen Burnie Hampton Lansdowne Linthicum Lochearn Lutherville Middle River Overlea Parkville Pikesville Pumphrey Rosedale Towson Woodlawn

Temporal criterion (Percent of housing stock built before 1970) 79 84 65 91 70 66 56 63 70 81 70 70 79 59 70 84 44 61 67 75 55

Spatial criterion (Shares a border with Baltimore or first-tier) Central City Adjacent Central City Central City Adjacent Adjacent Adjacent Central City Adjacent Central City Adjacent Central City Adjacent Adjacent Central City Central City Central City Adjacent Central City Central City Central City

Source: U.S. Census 2000 TIGER File.

the city. To satisfy the temporal criterion, I used the age of the housing stock as a proxy for the development age of a suburb. Suburban places that shared a boundary with a place that was adjacent to the central city were classified as a first-tier suburb if more than half of the housing stock was built before 1970. Table A.2 shows that with the exception of Pikesville, the majority of the housing stock in every first-tier suburb was built before 1970. In this analysis, 9 suburbs shared a border with another suburb that bordered the city and at least half of the housing stock was built by 1970. On an average, 70 percent of the housing stock in these 21 first-tier suburbs was built before 1970. I used that year as a threshold because half of the region’s housing development was already established by 1970. These spatial and temporal criteria captured all first-tier suburbs, and these places were located within eight miles of the border of Baltimore City. Baltimore’s first-tier suburbs were located near the central city and had already experienced substantial development by 1970. After classifying the first-tier suburbs, I then endeavored to determine which census tracts fell within the boundaries of each first-tier suburb. Since only broad generalizations could be drawn about suburban changes using place level data, I utilized census tract data to analyze suburban change within first-tier suburbs. Place level data were useful for comparing aggregate trends among first-tier

appendix / 197 suburbs. Yet, they were not useful for comparing trends within first-tier suburbs because the place level scale was too coarse to capture any socioeconomic variation within these places. Simply put, the place geography did not account for the rich variation within first-tier suburbs. I continued the analysis of suburban change using census tract level data. In the context of this study, it was useful to think of census places as suburbs. Each census place contained a collection of census tracts. Within each first-tier suburb, the census tracts represented the subdivisions of these places. In other words, suburbs were comprised of multiple neighborhoods. Table A.3 shows the number of census places and census tracts for the Baltimore region. Among the 21 first-tier suburbs, 152 census tracts were identified, or first-tier suburban neighborhoods. Using Geographic Information Systems (GIS) software, I identified all of the census tracts that fell within the boundaries of each first-tier suburban place in the Baltimore region. I identified suburban tracts by overlaying the census place boundary layer over the census tract layer in the GIS software. In some cases, all the tracts were an exact spatial match to the place boundary. Yet in Table A.3 Geographic characteristics of neighborhoods in metropolitan Baltimore, 2000 Location

Place

Western

Arbutus Brooklyn Park Catonsville Ferndale Glen Burnie Lansdowne Linthicum Lochearn Pumphrey Woodlawn Hampton Lutherville Parkville Pikesville Towson Dundalk Edgemere Essex Middle River Overlea Rosedale First-Tier Outer Suburbs Baltimore City Region

Northern

Eastern

Region

Number of neighborhoods

Source: U.S. Census 2000 TIGER File.

5 4 14 3 9 4 2 9 2 8 1 6 9 6 18 21 6 12 6 3 6 152 263 200 615

County Baltimore Anne Arundel Baltimore Anne Arundel Anne Arundel Baltimore Anne Arundel Baltimore Anne Arundel Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Baltimore Region

198 / appendix other cases, several census tracts overlapped the boundary of two places. When this occurred, I assigned the tract to the place where the majority of the tract boundary fell. This technique maintained the identity of each suburban place while capturing the socioeconomic variation within each first-tier suburb. These data allowed me to examine the process of neighborhood change within first-tier suburbs more closely. Tract level data were more appropriate to analyze socioeconomic variation within each first-tier suburb since they encompassed a smaller geography. Further, to demonstrate the relationship between places and tracts, consider the case of the Catonsville CDP, a first-tier suburb on the western fringe of Baltimore. There were 14 census tracts within this census place in 2000. In other words, this signified that there were 14 major neighborhood subdivisions in the first-tier suburb of Catonsville. Both these geographic scales contained socioeconomic data that were collected and analyzed. For example, at the place level, data were obtained for the entire census place. In this case, there would be one value, such as the poverty level, for the Catonsville CDP. In contrast, at the tract level, there would be data for each of the 14 tracts within the Catonsville CDP. So in this example, there would be 14 poverty values. While place level data provided one measurement of any given variable (i.e., poverty), the tract level data provided multiple measurements of any given variable for each of the census tracts. This provided a rich socioeconomic variation within suburban places. Finally, throughout this study, the first-tier suburbs were organized into three smaller geographic units: western suburbs; northern suburbs; and eastern suburbs, as Table A.3 displays. For analytical purposes, the first-tier suburbs were divided into these groups based on the suburbs’ respective location relative to the central city. Each first-tier suburb was classified into one of these three geographic units. There were 10 places grouped in the western suburbs, 5 places in the northern suburbs, and 6 places in the eastern suburbs. There were 16 suburbs in Baltimore County, and 5 suburbs in Anne Arundel County. This facilitated the analysis of suburban change by allowing me to generalize spatial patterns based on larger, regional patterns of urban development and socioeconomic structure in Baltimore.

Data Collection and Organization Continuing in the tradition of urban studies, I used both quantitative and qualitative data to study suburban change in the first-tier suburbs of Baltimore. This twotiered approach provided the opportunity to quantitatively measure the social and economic transitions from 1970 to 2000 and then discuss the policy implications of these changes. Using this approach required the collection and organization of a vast amount of data. Here I describe the types of data that I collected, and then I discuss how the data were organized for the analysis. The primary source of quantitative data that I relied on was from the U.S. Bureau of the Census, 1970 to 2000, and the U.S. Department of Housing and Urban Development’s (HUD) State of the Cities Data System 1970 to 2000. These publicly available datasets contained a rich collection of census data on a wide range of demographic, socioeconomic, and housing characteristics of suburban places and tracts. I compiled data on 50 variables for 101 CDPs in the region, which included the central city, the outer suburbs, and the first-tier suburbs. In

appendix / 199 addition, I compiled data at the census tract level on the same 50 variables for the 152 tracts in Baltimore’s first-tier suburbs. I collected these data for places and tracts at four points in time: 1970, 1980, 1990, and 2000. I organized these data into six broad categories that were related thematically: population characteristics, family structure, income characteristics, education attainment, housing characteristics, and labor market characteristics. The first set of variables was related to characteristics about the demographics of the population. I included variables that provided relevant information about the population size and racial composition of each place. The literature suggested that population trends, such as size and age of the population, as well as the racial and ethnic composition of a community, were important indicators for measuring decline (Orfield 2002). Scholars often used such demographic data as a metric to determine the socioeconomic condition of suburbs (Lucy and Phillips 2000). This set of data comprised of 4 main population variables, including size, age, racial and ethnic composition, and foreign-born status of the population. The age variables were divided into 4 cohorts to capture various aging patterns, including the children population (ages 17 and below), the young adult population (ages 18–24), the workforce population (ages 18–64), and the older population (ages 65 and more). The racial and ethnic variables captured non-Hispanic white, non-Hispanic black, Hispanic, and Asian races. The other races categories were ignored because the prevalence of these groups was negligible in metropolitan Baltimore. The second set of variables dealt with the composition of families. This category classified households into one of four mutually exclusive family types. Households were classified as “married families with children,” “single, never married,” “female-headed households,” or “divorced family households.” The third set of variables related to the income characteristics of households and families. The literature suggested that these data were important for identifying trends in poverty and economic segregation, two principal factors for diagnosing the decline of suburban areas (Swanstrom, Casey, Flack, et al. 2004). I included a series of variables that told me about three main income characteristics: 1. The number of poor suburban residents; 2. The income levels for households and families; and 3. The disparities between the rich and poor in each suburb. Accordingly, I calculated three main income variables for both households and families. First, I included the average income for each observation. Second, I included the aggregate income for the entire place or tract. This provided a measure of economic disparities in the region. Third, I computed an average income ratio for each suburban place or tract to all suburbs in the region. This provided the information about the economic standing of each suburb or tract relative to all other suburbs in the region. I computed all income data using 1999 constant dollars to control for inflation. In addition, I calculated the percentage of residents living in poverty. For all of the income data, it was necessary to compute averages instead of medians because the census data containing median income figures were not available at comparable geographic scales over 30 years.

200 / appendix The fourth set of variables related to the education attainment of households. The education variables measured the level of education attainment by adding the number of years that a person spent seeking an education. I classified the population into one of four mutually exclusive categories, which included: 1. 2. 3. 4.

Did not graduate from high school (less than 11 years of education); High school graduate (exactly 12 years of education); Some college (between 13 and 15 years of education); or College graduate (at least 16 years of education).

For each of these categories, I calculated the percentage of residents who attained each level of education for each census place and tract in the study. These variables were important because the literature suggests that neighborhood stability and educational status are related (e.g., Wilson 1987). The fifth set of variables related to housing characteristics in the region. There were four main housing variables that I grouped based on the housing size, age, value, and tenure. In terms of housing size, I calculated the percentage of housing units that had a particular number of bedrooms, using groups of one or no bedrooms, two bedrooms, three bedrooms, or four bedrooms as a proxy for the size of a housing unit. In terms of housing age, I calculated the percentage of housing units that were built in a particular decade, which spanned from the 1990s to the pre-1940s. In terms of housing value, I calculated the average housing value into 1999 constant dollars for each place and tract in the region. In terms of housing tenure, I calculated the percentage of the housing stock that was classified as either renter-occupied units, owner-occupied units, or vacant units. This information was important for comparing and contrasting characteristics of suburban housing in the first-tier suburbs. Recent research focused on the role that smaller, older housing played in the suburban decline process, so I attempted to capture as much information as possible about the nature and quality of the housing stock (Hudnut 2003). These data were useful in identifying trends over time in Baltimore’s suburban housing market. The last set of variables related to labor market characteristics in metropolitan Baltimore. I collected data on the employment status of residents and the occupation of employed workers. Specifically, I calculated the percentage of the labor force that was employed in 1 of 11 different occupations. These occupations represented the major sectors of the economy. I used these particular occupation categories because they were the only ones that were comparable across 30 years at the same geographic scale. In addition, I computed the percentage of the population that was unemployed each decade. This information provided a portrait of the type of work that residents had in first-tier suburbs, as well as the level of labor force participation. These data measured the types of jobs that residents of each place or tract had during each decade. The data did not measure the location of the jobs. Overall, I collected and organized data from a set of 50 socioeconomic variables that were available over a 30-year period from either the U.S. Census or U.S. HUD. I organized the data into six main categories, which included population characteristics, family structure, income characteristics, education attainment, housing characteristics, and labor market characteristics. These data provided a

appendix / 201 wealth of information about the process of the suburban change in the first-tier suburbs of Baltimore. These variables enabled me to assess the process of suburban change over three decades in metropolitan Baltimore. Many of the variables that I included in this study were based on well-established dimensions of urban differentiation. For instance, Shevky and Bell (1955), early pioneers in the field of urban studies, identified three primary dimensions of urban space in their famous study of Chicago: economic status, family status, and ethnic status. Their analysis, as well as subsequent studies, suggested that the structure of cities could be understood with the analysis of data on these three dimensions (Berry and Rees 1969; Knox 1994; Murdie 1969). Furthermore, urban scholars still view Shevky and Bell’s three primary dimensions as important elements of urban differentiation in the United States because the same three dimensions have been established in suburban areas, too (Davies 1984; Hughes 1993; Knox 1994; Perle 1981; Wyly 1999). The variables that I selected each relate to the primary dimensions that differentiate suburban areas. As such, they are appropriate for uncovering the suburban socioeconomic structure of Baltimore.

Analysis of Suburban Change I employed three quantitative techniques and one qualitative technique to analyze suburban change in Baltimore’s first-tier suburbs from 1970 to 2000. In the first quantitative analysis, I conducted a descriptive spatial statistical analysis of census place level data. In the second quantitative analysis, I performed a principal components analysis (PCA) on census tract level data to determine patterns of urban structure in first-tier suburban tracts for 1970 and 2000. In the third quantitative analysis, I used a cluster analysis technique to create a typology of first-tier suburban tracts. Then, I used ground truth analysis to verify the results and give additional context to suburban change. These four analyses allowed me to examine the process and nature of suburban change in Baltimore’s first-tier suburbs.

Descriptive Spatial Statistical Analysis I conducted a descriptive statistical analysis using census place level data in Baltimore’s first-tier suburbs. I refer to this method as a “descriptive spatial statistical analysis” because it was more than a simple exercise of describing statistical changes. Instead, I employed a geographic approach that allowed me to compare first-tier suburban changes relative to other places in the region. I utilized census place level data to conduct a comparative analysis of suburban change relative to four other geographic scales: 1. 2. 3. 4.

Other first-tier suburbs in Baltimore; Outer suburban areas; The central city (Baltimore City); and The Baltimore Primary Metropolitan Statistical Area (PMSA).

This technique was both descriptive and analytical. In descriptive terms, I portrayed how first-tier suburbs in Baltimore changed since 1970. In analytical

202 / appendix terms, I demonstrated how first-tier suburbs changed by examining the changes relative to other geographic units in the metropolitan area. To analyze geographic data over time, I used Geolytics’ Neighborhood Change Database, a commercially available database developed by the Urban Institute and the Rockefeller Foundation. Since the census tract boundaries changed each decade, it was necessary to use this tool to control for geographical boundary changes. This tool normalized census tract boundaries from 1970, 1980, and 1990 to the 2000 tract boundaries, thus facilitating change-over-time analysis of first-tier suburban tracts in the Baltimore region. To begin, I constructed a comprehensive dataset that contained all the 50 socioeconomic variables for the 101 census defined places in the region. I used these variables to analyze suburban change at four points in time: 1970, 1980, 1990, and 2000. I examined how first-tier suburbs changed by comparing them with one another. Then, I compared them to other geographic units, which included the outer suburbs, the central city, and the region. This provided a benchmark for determining the extent of suburban change relative to the Baltimore metropolitan area. Next, I manipulated the dataset in a number of ways. Using the 50 variables in the dataset I compiled, I made four calculations so that the data were easily comparable over time and simple to understand each geographic scale. They included 1. Calculation of summary statistics for each variable on the first-tier suburbs; 2. Calculation of the percent changes for each variable (from 1970 to 2000); 3. Inflation calculation for all monetary data based on the 1999 consumer price index (CPI); and 4. Calculation of income ratios for first-tier suburbs to all other suburbs. It was also useful to visualize many of the findings from the descriptive spatial statistical analysis. I displayed the data in a variety of charts and graphs, and then I interpreted and explained the results in the narrative of chapter four in this study. This analysis served as the baseline for understanding what suburban changes occurred between 1970 and 2000.

Principal Components Analysis For the second quantitative analysis, I used this dataset to conduct a PCA, which is a form of factor analysis. This is a quantitative technique that educational psychologist Harold Hotelling developed in 1933 (Wyly 1999). The basic premise of PCA is to understand the relationships among many different variables in a large dataset. Specifically, it is a statistical technique designed to 1. Reduce the number of variables in a dataset to several main factors; and 2. Detect structural relationships between the variables. PCA does this by reducing the dataset’s variables and transforming them into a smaller set of new variables, called principal components.

appendix / 203 PCA has a long history in urban studies, and it has been an important tool in deciphering the spatial organization of urban places (Berry and Horton 1970; Berry and Kasarda 1977; Janson 1980; Perle 1981; Wyly 1999). During the 1960s and 1970s, PCA was a commonly used methodology for urbanists. The availability of computing power allowed urbanists for the first time to quantitatively test the prevailing social theories of the Chicago School of the 1920s (see Berry and Kasarda 1977). As a result, urban research flourished as scholars conducted numerous PCAs on cities around the world. Yet, by the 1980s, the popularity of PCA faded as scholars embraced other quantitative and qualitative methods in the field of urban studies (Knox 1991). In the last decade, urban scholars once again turned to PCA as a method to gain a firmer understanding of the processes shaping suburbia. Some of the first empirical studies to analyze the decline of suburbs have employed PCA. Other scholars have effectively used PCA to study suburban change. For example, Lucy and Phillips (2000) analyzed change using a sample of 24 metropolitan areas; Orfield (2002) examined change using the top 25 most populated regions; and Wyly (1999) explored suburban change in the Minneapolis metropolitan area. These three recent studies effectively used PCA to gain a better understanding of the patterns of suburban change in U.S. metropolitan areas. They represented some of the first attempts to quantitatively and systematically redefine large-scale changes in the suburban landscape. Table A.4 presents an overview of the main elements of PCA. Each of these elements will be explained in this section as I describe of the process of PCA that was utilized for this study. Table A.4

Overview of principal components analysis

PCA

Element

Definition

Unit of measurement

Input

Dataset

A collection of data with cases or observations (census tracts), with variables for each case A new variable that results from the reduction of a large set of variables; the component represents a collection of variables A number representing the order of importance for each component A correlation between the variables and the components; used to attach meaning to the component An estimate of the variance in each variable accounted for by all of the components A numerical index of the strength of the meaning attached to each component

Attributes from census data

Outputs Component

Eigenvalue Component loading

Communality Component score

Based on the number of components selected

Greater than zero (If over two, then component was selected) 21.00 to 11.00

0–1.0 (higher is better) Any negative or position number

204 / appendix Similar to Lucy and Phillips (2000), Orfield (2002), and Wyly (1999), I conducted two PCAs on Baltimore’s first-tier suburban places using census tracts, one for 1970 and one for 2000. Using the same 50 variables compiled for the descriptive spatial statistical analysis, I constructed a data matrix containing 50 variables for the 152 census tracts (152 3 50) in Baltimore’s first-tier suburbs. The matrix contained 7,600 separate numerical values, called attributes. The 1970 PCA contained 152 census tracts, but the 2000 PCA contained 151 tracts. I deleted a tract in Southeastern Baltimore County from the analysis in 2000 because there were no data associated with the tract. This area, formerly mixed use residential with industry in 1970, was located in Sparrows Point. By 2000, it only contained industrial sites, so it was not appropriate to include in the analysis. To begin, I loaded the data matrix into SPSS, a statistical software package that computes the PCA. This procedure produced four sets of statistical outputs. First, the PCA produced a communality value for each of the 50 variables. The communality was a numerical estimation of the variance and strength of each variable that was reduced into a set of fewer components. In other words, the communality provided a value of the strength of the relationship between each variable and component that the PCA produces. Other studies have shown that the majority of the variables in a study should have a communality of at least 0.70 to ensure that the variables are related to one another (Perle 1981; Wyly 1999). The communality value could be interpreted as a percent. In this study, the majority of the communalities produced were greater than 0.70. This meant that the variables selected for the analysis were highly correlated with one another, thus providing a strong range of socioeconomic measures. Second, the PCA generated component loadings. They were measures of the degree to which each variable in the dataset contributed to the meaning of each new component that the PCA produced. In other words, the loadings were used to interpret the characteristics of each component. For example, loadings with a high value on the income variables suggested that wealth was a defining characteristic of that component. Conversely, loadings with low or negative values suggested that wealth was not a defining characteristic of that component. The loadings were correlation coefficients that calculated the association between the original variables and the newly derived components from the PCA. By definition, the component loading values varied between 21.00 and 11.00. For example, a component loading of 0.70 could be interpreted as being 70 percent positively correlated with the component. The third statistical output that the PCA produced was the component score. The PCA generated one score for each observation in the dataset (census tract). The scores were an index of positive and negative values that measured the degree to which each observation related to the components that the PCA produced. In other words, the scores informed me how well each observation represented each of the six components that the PCA produced for the 1970 and 2000 analyses. For example, a census tract that scored very high on component one would score low on the other five components in the dataset. To facilitate that PCA interpretation, I mapped the PCA scores for each component to examine spatial patterns using GIS. This allowed me to differentiate the patterns of suburban characteristics within and among first-tier suburbs in metropolitan Baltimore.

appendix / 205 The fourth statistical output that the PCA generates was the eigenvalue. This was a positive number that represented the variance explained by each of the principal components, in decreasing order of importance. The eigenvalue helped to determine two features of the PCA: (1) the order of importance for the components and (2) the number of components to extract from the dataset. Components with high eigenvalues indicated that the component explained more of the variation in the dataset than components with lower eigenvalues. Previous studies suggested that all the components with eigenvalues more than two should be extracted from the dataset (Wyly 1999). What this meant was that each component extracted from the dataset with an eigenvalue more than two had considerable explanatory power (Kachigan 1991). In both the 1970 and 2000 analyses, I extracted six components from each analysis because the eigenvalue was more than two for each of the six components that I extracted. This was necessary because I needed to limit the number of components that I used from the PCA. I graphed the eigenvalues and the components to visually display both the importance and the number of principal components that I used for this study. PCA was advantageous for several reasons. It was a useful methodology for this study because it transformed a large number of variables into a new, smaller set of composite variables, or principal components. These components shared a common set of socioeconomic characteristics. With such a large number of observations, it was cumbersome to describe and analyze the data without a tool to make sense of the patterns in the dataset. This procedure reduced the dataset to a manageable size so that I could analyze suburban change. The PCA also helped to determine the distinguishing characteristics among and within first-tier suburban places in the Baltimore PMSA. This method allowed me to characterize and determine the socioeconomic variation in suburban Baltimore, which was critical for defining the key changes in the first-tier suburbs between 1970 and 2000. This technique acted as both a data reduction method and a process for identifying the most important empirical elements embedded in a dataset.

Cluster Analysis For the third quantitative analysis, I conducted a cluster analysis to create a typology of suburban tracts in Baltimore’s first-tier suburbs. Recently, numerous urban scholars have used this grouping technique to create classifications of suburban places and tracts. The approach has been utilized to identify “suburban immigrant enclaves” (Logan and Zhang 2004); “at-risk suburbs” (Orfield 2002); “edgeless cities” (Lang 2003); “healthy suburbs” (Mikelbank 2004); and “Middle America suburbs” (Vicino, Hanlon, and Short 2007), just to name a few. In this study, I built on the tradition of clustering to create categories of suburban neighborhoods that depict the characteristics of variation in Baltimore’s first-tier suburbs. This study was among the first of its kind to differentiate all of the different types of first-tier suburbs in an entire metropolitan area. Cluster analysis is a statistical technique that categorizes data into a series of smaller groups that share a set of similar characteristics. The members of each group are clustered so that they have a strong level of association with each other and a weak level of association with the members of other clusters. For the purposes of this study, I clustered the component scores generated for each census

206 / appendix tract from the PCA. To begin, I loaded the dataset into SPSS to group the scores using a two-step clustering method. This clustering procedure is also known as expectation maximization cluster analysis. In the first step, the cases were initially clustered based on the nearness criterion, which was automatically calculated by determining the hierarchical relationships of the cases. In the second step, the cluster analysis generated a data structure tree to refine the associations between the cases, and then the procedure assigned them to a similar cluster accordingly. This clustering procedure allowed me to specify the number of clusters that I desired. Scholars have typically defined between four to seven clusters (Hartigan 1975; Orfield 2002). Drawing on previous studies, I specified alternative cluster sizes of three, four, five, and six clusters for the 1970 and 2000 analyses. I compared each specification for ease of interpretation, and I selected the most appropriate cluster size. For the 1970 analysis, five clusters were selected, and for the 2000 analysis, six clusters were selected. This was an optimal clustering strategy because it allowed me to interpret the meaning of each cluster by specifying the number of clusters that I desired. It was an ideal grouping procedure for this analysis owing to the large size of the data matrix of first-tier suburban tracts (Bailey 1975; Orfield 2002). The results of the cluster analysis were valuable for several reasons. The cluster groupings enabled me to create a typology of neighborhoods in Baltimore’s firsttier suburbs, drawing on the results of the PCA. The suburban typology, for instance, illuminated the socioeconomic variation within suburban places in metropolitan Baltimore. I gained a clearer understanding about the distinguishing characteristics of neighborhoods within first-tier suburban places. In his national study of American suburbs, Mikelbank (2004, 936) fittingly observed that “typologies serve as a springboard from which the behavior of complex and diverse phenomena can be more clearly understood. Classification research can help bridge the conceptual gap between the seemingly unique character of an individual observation and the well-understood behavior of groups of similar observations.” Indeed, in a similar vain, I was able to distinguish among the many types of suburban neighborhoods in the region. This was important for developing new theories about the patterns of urban restructuring in the metropolitan landscape, as well as new forms of spatial organization.

Ground Truth Analysis The fourth research approach that I used was ground truth analysis. This analysis built upon and supplemented the quantitative analyses, and I integrated the findings throughout my study. In this part of the study, this approach was used to validate the results of my quantitative analyses. Also, my goal was to gain insight into the process of suburban change by observing first hand the evidence of governmental responses to the changing nature of first-tier suburbs. Social scientists often use ground truth analysis as a mechanism to verify the findings of a study. The term “ground truth” literally means to observe on the ground to determine whether the results of other analyses are true (Yin 2003). It is also a tool that researchers use to bring the results of a study to life—a way of giving meaning to the data. In essence, a supplemental method adds reliability to the results of the research (Pickles 1994).

appendix / 207 I collected qualitative data from three sources, which included documentation, archival records, and direct observation. The first source of data that I relied on was documentation. This was an important source of data that allowed me to confirm and strengthen the evidence gathered from the quantitative analyses. I relied on newspapers and other written media articles as the primary source of documentation. For example, I conducted a search for articles during the 1990s and 2000s that related to suburban change and decline covered in the Baltimore Sun, the region’s leading daily newspaper. Also, I supplemented the document search by searching for articles in community newspapers and newsletters in the first-tier suburbs. Yin (2003) provides a number of advantages for supplementing studies with documentation sources. He holds that the information is stable and exact—names, references, and details of events can be easily verified. In addition, Yin notes that there is broad coverage of the topic under study, which helps to uncover a wealth of information over a long period. These searches provided useful, historical background information on suburban decline in Baltimore, and it was helpful for chronicling Baltimore County’s revitalization efforts in first-tier suburbs. The second source of data that I utilized was archival records. I searched an array of public documents for information pertaining to suburban change and decline. In particular, local and state planning archives proved quite useful. I reviewed planning documents in the state’s Department of Housing and Community Development that pertained to neighborhood revitalization, suburban decline, and sustainable growth. I also reviewed Baltimore and Anne Arundel County’s master plans. These documents contained valuable maps and charts detailing the characteristics of aging suburbs. The county’s plans for revitalization of older suburban areas were especially informative. They assisted with the analysis of political, planning, and policy responses to suburban decline. The third source of data was direct observation. Urban scholars have used this method for centuries, and researchers even today favor this method as a way of explaining complex social and economic interactions. For example, in a recent study on urban decentralization, Robert Bruegmann (2005, 9) notes that “[t]he best source of information was the built environment itself. A great deal of my research has consisted of going out and looking around.” Building on this longstanding tradition in social science, I, too, went out and observed the environment that I studied. I began by making multiple site visits to each of the first-tier suburbs in Baltimore. During the site visits, I drove through the major corridors and “main streets” in each community. I visited commercial and industrial locations as well as parks and schools, when possible. Also, I visited numerous neighborhoods in each suburb. Dabbs (1982) notes that photographing one’s observations is an especially useful method as it helps to convey important characteristics about the study to outsider observers and nonspecialists. Accordingly, at each site, I took a series of photographs, documenting the housing stock, the built infrastructure, and the general state of local economic conditions. This method provided evidence of the current socioeconomic state that I was able to relate to the quantitative analyses. Overall, the benefits of direct observation were twofold: they gave “reality” to the topic under study, and they provided “context” for comparing first-tier suburbs to one another (Berg 2004).

208 / appendix

Geographic Information Systems Throughout my research, I utilized GIS, an important tool to portray data trends on maps. Beginning in the 1990s, the implementation of GIS software for analytical purposes became recognized as a respected method in the social sciences (Greene 2004). For example, Findlay and Hoy (2000) argue that GIS technology is useful to study social, economic, and demographic changes in metropolitan areas. Still others have used this technique to identify dimensions of suburban tract differentiation such as economic class and ethnic group (Logan and Zhang 2004). Goodchild and Janelle (2004, v), summarizing the emergence of GIS in the social sciences, aptly state that [t]he advent of geographic information systems (GIS) has enabled an explosion of interest in and ability to study the spatial patterns of behavior . . . It provides a powerful new tool that has stimulated new and exciting social science research using geographical concepts and data. At last, long-held but unverified hypotheses about the importance of locational and spatial variables can be tested. We are at the dawn of a revolution in a spatially oriented social science. Similarly, I used GIS technology to assist me in creating maps. This method served several purposes. First, I was able to identify first-tier suburbs using geographic data in the GIS program. Second, GIS facilitated the explanation of suburban change because I was able to identify spatial patterns using maps from the results of the quantitative analyses. Third, the maps allowed the typology of first-tier suburbs to be visually displayed. GIS supplemented the analysis of suburban change by providing the opportunity to show patterns in the data geographically.

Limitations No study in the social sciences is without limitations. This study was limited in at least five ways. First, this study was limited to a single metropolitan area that was located in the northeastern Rustbelt. Studying the Baltimore experience did not allow me to generalize about all first-tier suburbs nationally. Instead, I was only able to observe and comment on the suburban changes in Baltimore. Nevertheless, and even with this limitation, the results of this study were generalizable to larger “theoretical propositions” about the evolution of first-suburbs since 1970 (Yin 2003). Second, this study sacrificed breadth for depth. Rather than focusing, for example, on first-tier suburbs nationally, I examined in-depth the 21 first-tier suburbs in a single region. This approach enabled me to pay closer attention to the socioeconomic differences among and within first-tier suburbs, which gave me a fuller and richer understanding of neighborhood change in the first-tier suburbs. This was particularly important because it provided the opportunity to analyze suburban change relative to other geographic areas such as the outer suburbs, the central city, and the region.

appendix / 209 Third, as is often the case in the social sciences, the data sources were limited. I relied on U.S. Census and U.S. HUD data since they were the only reliable data sources of socioeconomic information over a 30-year period. For example, I was not able to include data on school performance and crime for several reasons. Those data were not available at the first-tier suburban geographic scale. In fact, the finest geographic scale for these data was at the county level—far too coarse of a scale to apply to individual suburbs. Similarly, reliable data were not available before 1990 in an accessible format. Therefore, I limited the data sources to the socioeconomic data that was available from 1970 to 2000 in either the decennial censuses or U.S. HUD datasets. Fourth, this study analyzed first-tier suburban change during a 30-year time frame, from 1970 to 2000. I limited the analysis to these years for several reasons. First, reliable census data were available only during the time frame of this study. Census geographies before 1970 were not comparable to present ones. In addition, the Neighborhood Change Database, the commercial database that I used, normalized census tract boundaries to the 2000 tract boundaries. Second, one of the goals in this study was to gain a better understanding of the process of first-tier suburban change in a modern, postindustrial era. On the basis of a review of the literature that demonstrated that major economic restructuring had occurred by 1970, I used that year as the starting point for the study (Bluestone and Harrison 1982). Furthermore, since the data did not permit me to describe the socioeconomic condition of suburbs before 1970, I relied on previous literature and historical accounts to supplement the analysis on an as-needed basis throughout the study. Fifth, this study was limited to the census geographic definitions that were developed and adopted in 1999 for the Census 2000. I did not use newer geographic definitions that the U.S. Office of Management and Budget (OMB) developed during the early 2000s to better reflect the metropolitan landscape. These newer geographic entities, such as “micropolitan areas” and “principal cities” could not be compared to previous geographies that were statistically defined. In a recent report, Frey, Wilson, Berube, et al. (2004) demonstrated that new census geographic definitions for the Census 2010 would greatly impact how urbanized areas are studied. One of their main points was that it would be difficult to compare older census data to newer census data. Therefore, I used the older definitions. In essence, limiting the analysis to OMB’s 1999 geographic definitions for metropolitan areas facilitated the change-over-time analysis from 1970 to 2000.

Summary To recap, I presented an overview of the research methodologies that I employed in this study. Using a mixed-methods approach that incorporated four methodologies allowed me to gain an understanding about how and why suburban change occurred in Baltimore’s first-tier suburbs from 1970 to 2000. I developed a definition of first-tier suburbs that used both spatial and temporal criteria. Then, I employed a descriptive spatial statistical analysis that made it possible to measure suburban changes in first-tier suburbs since 1970. Next, I conducted a PCA, which revealed the characteristics of the socioeconomic structure in first-tier suburbs in

210 / appendix 1970 and 2000. Then, I performed a two-step cluster analysis that allowed me to create a typology of the different types of suburban tracts in Baltimore’s first-tier suburbs. Collectively, these three quantitative analyses guided me in conducting ground truth analysis to verify my results. Using the direct observation technique and making field visits to each first-tier suburb, I gathered data on the current socioeconomic condition of suburbs. This methodological approach provided me with background information about nature and process of suburban change. In this study, I integrated these four approaches to gain a thorough understanding of the process of suburban change so that I could reflect on the implications for public policy and planning.

R e f e r e nc e s

Abrahamson, M. 1996. Urban enclaves: Identity and place in America. New York: St. Martin’s. Abrams, C. 1965. The city is the frontier. New York: Harper Colophon Books. Ahlbrandt, R., Jr. and P.C. Brophy. 1975. Neighborhood revitalization: Theory and practice. Lexington, KY: D.C. Heath and Company. Alves, T. 2001. Some enchanted evening: Tuning in the amazing fifties, switching off the elusive decade. American Studies International 39 (3): 25–40. Anft, M. 2000. A river of trouble. City Paper. July 5. Arnold, J.P. 1990. Baltimore: Southern culture and a northern economy. In Snowbelt cities: Metropolitan politics in the Northeast and Midwest since World War II, edited by R.M. Bernard, 25–39. Bloomington: Indiana Univ. Press. Ashton, P.J. 1978. The political economy of suburban development. In Marxism and the metropolis, edited by W.K. Tabb and L. Sawers, 64–89. New York: Oxford Univ. Press. Bailey, K.D. 1975. Cluster analysis. Sociological Methodology 6: 59–128. Baldassare, M. 1986. Trouble in paradise: The suburban transformation in America. New York: Columbia Univ. Press. Baltimore County. 1999. Strategic conservation plan. Towson, MD. ———. 2000. Master plan 2010. Towson, MD. Barnes, F. 2006. A congress of mayors: The GOP’s new suburban strategy. The weekly standard 11 (16): 25, January 2. Baxandall, R. and E. Ewen. 2000. Picture windows: How the suburbs happened. New York: Basic Books. Beauregard, R.A., ed. 1989. Atop the urban hierarchy. Totowa, NJ: Rowman & Littlefield. ———. 2003. Voices of decline: The postwar fate of U.S. cities. 2nd ed. New York: Routledge. ———. 2006. When America became suburban. Minneapolis: Univ. of Minnesota Press. Berg, B.L. 2004. Qualitative research methods for the social sciences. 5th ed. Boston: Allyn & Bacon. Berger, B.M. 1968. Working class suburb: A study of auto workers in suburbia. Berkeley: California Univ. Press. Berry, B.J.L. and F.E. Horton, eds. 1970. Geographic perspectives on urban systems. Englewood Cliffs, NJ: Prentice-Hall. Berry, B.J.L. and J.D. Kasarda. 1977. Contemporary urban ecology. New York: Macmillan.

212 / references Berry, B.J.L. and P.H. Rees. 1969. The factorial ecology of Calcutta. American Journal of Sociology 74 (5): 445–91. Berry, B.J.L. and Y.S. Cohen. 1973. Decentralization of commerce and industry: The restructuring of metropolitan America. In The urbanization of the suburbs, edited by L.H. Masotti and J. K. Hadden, 431–57. Beverly Hills, CA: Sage. Berube, A., B. Katz, and R.E. Lang, eds. 2005. Redefining urban and suburban America: Evidence from Census 2000 Volume 2. Washington, DC: Brookings Institution Press. ———. 2006. Redefining urban and suburban America: Evidence from Census 2000 Volume 3. Washington, DC: Brookings Institution Press. Birch, D.L. 1975. From suburb to urban place. Annals of the American Academy of Political and Social Science 422: 25–35. Bluestone, B.P. and B. Harrison. 1982. The deindustrialization of America. New York: Basic Books. Blumberg, L. and M. Lalli. 1966. Little ghettoes: A study of Negroes in the suburbs. Phylon 27 (2): 117–31. Bollens, S.A. 1988. Municipal decline and inequality in American suburban rings, 1960–1980. Regional Studies 22 (4): 277–85. Bourne, L. 1981. The geography of housing. New York: John Wiley & Sons. Bourne, L.S. 1996. Reinventing the suburbs: Old myths and new realities. Progress on Planning 46 (3): 163–84. Boyne, G.A. 1992. Is there a relationship between fragmentation and local government costs? A comment on Drew Dolan. Urban Affairs Quarterly 28 (2): 317–23. Brenner, N. 2001. The limits of scale? Methodological reflections on scalar structuration. Progress in Human Geography 25 (4): 591–614. Brooks, N. and E. Rockel.1979. A history of Baltimore County. Towson, MD: Friends of Towson Library. Bruegmann, R. 2005. Sprawl: A compact history. Chicago: Univ. of Chicago Press. Burchell, R.W., D. Listokin, and C.C. Galley. 2000. Smart growth: More than a ghost of urban policy past, less than a bold new horizon. Housing Policy Debate 11 (4): 821–80. Burgess, E.W. 1925. The growth of the city: An introduction to a Research Project. In The city, edited by R.E. Park, E.W. Burgess, and R.D. McKenzie, 47–62. Chicago: Univ. of Chicago Press. Burns, N. 1994. The formation of American local governments: Private values in public institutions. New York: Oxford Univ. Press. Carr, J.B. and R.C. Feiock, eds. 2004. City county consolidation and its alternatives. Armonk, NY: M.E. Sharpe. Cashin, S. 2004. The failures of integration: How race and class are undermining the American dream. New York: Public Affairs. Choldin, H.M., C. Hanson, and R. Bohrer. 1980. Suburban status instability. American Sociological Review 45 (6): 972–83. Christie, L. 2005. Older suburbs are new again. CNN Money. June 9. Clinton, H.R. 2006. Opening remarks on First Suburbs Symposium. Brookings Institution. Washington, DC. February 15.

references / 213 Cohen, J. 2002. Maryland’s Smart Growth: Using incentives to combat sprawl. In Urban sprawl: Causes, consequences and policy responses, edited by G. Squires, 293–324. Washington, DC: Urban Institute Press. Cohn, D. 2006. Inner suburbs fall through the cracks. Washington Post, February 15, sec. A. Colman, W.G. 1965. The role of the federal government in the design and administration of intergovernmental programs. Annals of the American Academy of Political and Social Science 359 (1): 23–34. Connolly, A. 2006. Feeling pressure for profits. Baltimore Sun, May 14. Culver, L.W. 1982. The politics of suburban distress. Journal of Urban Affairs 4 (1): 1–18. Curran, J.J., Jr. 2000. Opinion on S.B. 509. Office of the Attorney General. Annapolis, MD. Dabbs, J.M., Jr. 1982. Making things visible. In Varieties of qualitative research, edited by J. Van Maanen, 31–63. Thousand Oaks, CA: Sage. Davies, W.K.D. 1984. Factorial ecology. Aldershot, UK: Gower. Davis, M. 2005. Ozzie and Harriet in hell: On the decline of inner suburbs. In Sprawl and suburbia, edited by W.S. Saunders, 27–33 Minneapolis: Univ. of Minnesota Press. Dilworth, R. 2005. The urban origins of suburban autonomy. Cambridge, MA: Harvard Univ. Press. Dobriner, W.M. 1963. Class in suburbia. Englewood Cliffs, NJ: Prentice Hall. Dolan, D. 1990. Fragmentation: Does it drive up the costs of government? Urban Affairs Quarterly 26 (1): 28–45. Downs, A. 1973. Opening up the suburbs: An urban strategy for America. New Haven, CT: Yale Univ. Press. ———. 1982. Neighborhoods and urban development. Washington, DC: Brookings Institution Press. ———. 2001. What does “Smart Growth” really mean? Planning, April. Durr, K.D. 2003. Behind the backlash: White working-class politics in Baltimore, 1940–1980. Chapel Hill: Univ. of North Carolina Press. Euchner, C. and S. McGovern. 2003. Urban policy reconsidered: Dialogues on the problems and prospects of American cities. New York: Routledge. Farley, R. 1964. Suburban persistence. American Sociological Review 29 (1): 38–47. Feiock, R.C., ed. 2004. Metropolitan governance: Conflict, competition and cooperation. Washington, DC: Georgetown Univ. Press. ———. 2007. Rational choice and regional governance. Journal of Urban Affairs 29 (1): 47–63. Findlay, A.M and C. Hoy. 2000. Global populations: Towards a geographic research agenda. Applied Geography 20 (3): 207–19. Fishman, R. 1987. Bourgeois utopias: The rise and fall of suburbia. New York: Basic Books. ———. 2000. The American metropolis at century’s end: Past and future influences. Housing Policy Debate 11 (1): 199–213. Fogelson, R. 2006. Bourgeois nightmares: Suburbia 1870–1930. New Haven, CT: Yale Univ. Press.

214 / references Foster, K.A. 2001. Regionalism on purpose. Cambridge, MA: Lincoln Institute of Land Policy. Frederickson, H.G., G.A. Johnson, and C.H. Wood. 2003. The adapted city: Institutional dynamics and structural change. Armonk, NY: M.E. Sharpe. Frey, W.H. 2003a. Boomers and seniors in the suburbs: Aging patterns in Census 2000. Washington, DC: Brookings Institution Metropolitan Policy Program (January). ———. 2003b. Melting pot suburbs: A study of suburban diversity. In Redefining urban and suburban America: Evidence from Census 2000 Volume 1, edited by B. Katz and R.E. Lang (eds.), 155–180. Washington, DC: Brookings Institution Press. Frey, W.H., J.H. Wilson, A. Berube, and A. Singer. 2004. Tracking metropolitan America into the 21st century: A field guide to the new metropolitan and micropolitan definitions. Metropolitan Policy Center. Washington, DC: Brookings Institution. Gainsborough, J. 2001a. Fenced off: The suburbanization of American politics. Washington, DC: Georgetown Univ. Press. ———. 2001b. Smart growth and urban sprawl: Support for a new regional agenda? Urban Affairs Review 37 (5): 728–44. Gans, H.J. 1967. The Levittowners: Ways of life and politics in a new suburban community. New York: Random House. Garreau, J. 1991. Edge city: Life on the new frontier. New York: Anchor Books. Gelfand, M. 1982. Cities, suburbs, and government policy. In Reshaping America: Society and institutions 1945–1960, edited by R.H. Bremner and G.W. Reichard, 261–82. Columbus: Ohio State Univ. Press. Goodchild, M.F. and D.G. Janelle. 2004. Spatially integrated social science. New York: Oxford Univ. Press. Graham, S. 2005. Baltimore County waterfront teeming with development. Baltimore Business Journal, March 11. Green, A. 2007. Office of Smart Growth will be revived in Maryland under Gov.-Elect O’Malley. Baltimore Sun. January 8. Green, R.K. and S. Malpezzi. 2003. A primer on U.S. housing markets and housing policy. Washington, DC: Urban Institute Press. Greene, R.W. 2004. GIS in public policy: Using geographic information systems for more effective government. Redlands, CA: ESRI. Grigsby, W., M. Baratz, G. Galster, and D. Maclennan. 1987. The dynamics of neighborhood change and decline. London: Pergamon. Grogan, P.S. and T. Proscio. 2000. Comeback cities: A blueprint for urban neighborhood revival. Boulder, CO: Westview. Guest, A.M. 1978. Suburban social status: Persistence or evolution? American Sociological Review 43 (2): 251–64. Hanlon, B.F. and T.J. Vicino. 2007. The fate of inner suburbs: Evidence from metropolitan Baltimore. Urban Geography 28 (3): 1–27. Hanlon, B.F., T.J. Vicino, and J.R. Short. 2006. The new metropolitan reality in the US: Rethinking the traditional model. Urban Studies 43 (12): 2129–43. Harrigan, J. and R. Vogel. 2003. Political change in the metropolis. 7th ed. New York: Longman.

references / 215 Harris, C.D. and E.L. Ullman. 1945. The nature of cities. Annals of the American Academy of Political and Social Science 242: 7–17. Hartigan, J. 1975. Clustering algorithms. New York: John Wiley & Sons. Harwood, H.H., Jr. 2003. Baltimore’s streetcars: The postwar years. Baltimore: Johns Hopkins Univ. Press. Hayden, D. 2003. Building suburbia: Green fields and urban growth: 1820–2000. New York: Pantheon. Hayward, M.E. and C. Belfoure. 2001. The Baltimore rowhouse. Princeton, NJ: Princeton Architectural Press. Hillier, A. 2003. Redlining and the home owners’ loan corporation. Journal of Urban History 29 (4): 394–420. Hirsch, A. 1983. Making the second ghetto: Race and housing in Chicago 1940–1960. Chicago: Univ. of Chicago Press. Hoover, E.M. and R. Vernon. 1962. Anatomy of a metropolis: The changing distribution of people and jobs within the New York metropolitan region. New York: Anchor Doubleday. Hopkins, J.S. 2005. Rising tide and the affordable housing crunch. Baltimore Sun. April 5. Howard, J. 2006. Suburban makeover, scholarly edition. Chronicle of Higher Education, March 24. Hoyt, H. 1939. The structure and growth of residential neighborhoods in American cities. Washington, DC: Federal Housing Administration. Hudnut, W.H., III. 2003. Halfway to everywhere: A portrait of America’s first-tier suburbs. Washington, DC: Urban Land Institute. Hughes, H.L. 1993. Metropolitan structure and the suburban hierarchy. American Sociological Review 58 (3): 417–33. Jackson, K.T. 1985. Crabgrass frontier: The suburbanization of the United States. Oxford: Oxford Univ. Press. Jacobs, J. 1961. The death and life of great American cities. New York: Vintage Books. Janson, C.G. 1980. Factorial social ecology: An attempt at summary and evaluation. Annual Review of Sociology 6: 433–56. Jargowsky, P.A. 1997. Poverty and place: Ghettos, barrios, and the American city. New York: Russell Sage Foundation. Johnson, M. 2001. Environmental impacts of urban sprawl: A survey of the literature and proposed research agenda. Environment and Planning A 33 (4): 717–35. Judd, D. and T. Swanstrom. 2005. City politics: The political economy of urban America. 4th ed. New York: Longman. Kachigan, S. 1991. Multivariate statistical analysis: A conceptual introduction, 2nd ed. New York: Radius. Katz, B., ed. 2000. Reflections on regionalism. Washington, DC: Brookings Institution Press. Keating, D. 1994. The suburban racial dilemma: Housing and neighborhoods. Philadelphia, PA: Temple Univ. Press. ———. 2005. Fighting suburban decline: Cleveland’s first suburbs consortium. Paper presented at the 35th Annual Meeting of the Urban Affairs Association. April 16. Salt Lake City, NV.

216 / references Keene, J. 2001. Doughboys: The great war and the remaking of America. Baltimore: Johns Hopkins Univ. Press. Kelly, B.M., ed. 1989. Suburbia re-examined. Westport, CT: Greenwood. ———. 1993. Expanding the American dream: Building and rebuilding Levittown. Albany: State Univ. of New York Press. Kingdon, J.W. 1997. Agendas, alternatives, and public policies. 2nd ed. New York: Pearson. Kline, P. 1994. An easy guide to factor analysis. London: Routledge. Knaap, G.J. 2001. Land market monitoring for smart urban growth. Cambridge, MA: Lincoln Institute of Land Policy. Knox, P. 1991. The restless urban landscape: Economic and sociocultural change and the transformation of metropolitan Washington, DC. Annals of the Association of American Geographers 81 (2): 181–209. ———. 1994. Urbanization: An introduction to urban geography. Englewood Cliffs, NJ: Prentice Hall. Kramer, J. 1972. North American suburbs: Politics, diversity and change. Berkeley, CA: Glendessary. Kruse, K. 2005. White flight: Atlanta and the making of modern conservatism. Princeton, NJ: Princeton Univ. Press. Kruse, K. and T. Sugrue, eds. 2006. The new suburban history. Chicago: Univ. of Chicago Press. Lang, R.E. 2003. Edgeless cities: Explaining the elusive metropolis. Washington, DC: Brookings Institution Press. Lang, R.E. and J. LeFurgy. 2007. Boomburbs: The rise of America’s accidental cities. Washington, DC: Brookings Institution Press. Lassiter, M. 2006. The silent majority: Suburban politics in the sunbelt south. Princeton, NJ: Princeton Univ. Press. Lee, S. and N.G. Leigh. 2005. The role of inner ring suburbs in metropolitan smart growth strategies. Journal of Planning Literature 19: 330–46. Leigh, N.G. and S. Lee. 2005. Philadelphia’s space in between: Inner-ring suburb evolution. Opolis: An International Journal of Suburban and Metropolitan Studies 1 (1): 13–32. Levine, M.V. 1987. Downtown redevelopment as an urban growth strategy: A critical appraisal of the Baltimore renaissance. Journal of Urban Affairs 9 (2): 103–23. Lewis, R., ed. 2004. Manufacturing suburbs: Building work and home on the metropolitan fringe. Philadelphia, PA: Temple Univ. Press. Li, W. 1998. Anatomy of a new ethnic settlement: The Chinese ethnoburb in Los Angeles. Urban Studies 35 (3): 479–501. Light, P.C. 1999. The true size of the federal government. Washington, DC: Brookings Institution Press. Listokin, D. and W.P. Beaton. 1983. Revitalizing the older suburb. New Brunswick: State Univ. of New Jersey Center for Urban Policy Research. Logan, J.R. and H. Molotch. 1987. Urban fortunes: The political economy of place. Berkeley: Univ. of California Press. Logan, J.R. and M. Schneider. 1981. The stratification of metropolitan suburbs, 1960–1970. American Sociological Review 46 (2): 175–86.

references / 217 Logan, J.R. and W. Zhang. 2004. Identifying ethnic neighborhoods with census data: Grouping concentration and spatial clustering. In Spatially integrated social science, edited by M.F. Goodchild and D.G. Janelle, 113–26. New York: Oxford Univ. Press. Lopez, R.A., A.O. Adelaja, and A.S. Andrews. 1988. The effects of suburbanization on agriculture. American Journal of Agricultural Economics 70 (2): 347–58. Lucy, W.H. and D.L. Phillips. 2000. Confronting suburban decline: Strategic planning for metropolitan renewal. Washington, DC: Island. ———. 2006. Tomorrow’s cities, tomorrow’s suburbs. Chicago: American Planning Association Planner’s Press. Malarkey, J. 2006. Housing, business boom in county. Baltimore Examiner, April 24. ———. 2007. Demolition of blighted apartments to begin. Baltimore Examiner, January 22. Masotti, L.H. and J.K. Hadden, eds. 1973. The urbanization of the suburbs, Vol. 7. Thousand Oaks, CA: Sage. Massey, D. and N.A. Denton. 1993. American apartheid: Segregation and the making of the underclass. Cambridge, MA: Harvard Univ. Press. McManus, R. and P.J. Ethington. 2007. Suburbs in transition: New approaches to suburban history. Urban History 34 (2): 317–30. Mettler, S. 2005. Soldiers to citizens: The G.I. bill and the making of the greatest generation. New York: Oxford Univ. Press. Mikelbank, B.A. 2004. A typology of U.S. suburban places. Housing Policy Debate 15 (4): 935–64. Miller, D.Y. 2002. The regional governing of metropolitan America. Cambridge, MA: Westview. Money. 2007. The best places to live. August. Mosher, J. 2005. Sale of Dundalk Village Shopping Center hailed as redevelopment. Daily Record, March 14. ———. 2006. Baltimore County to buy troubled Dundalk apartment complex. Daily Record, March 28. Muller, E.K. and P.A. Groves. 1979. The emergence of industrial districts in mid-nineteenth century Baltimore. Geographical Review 62 (2): 159–78. Muller, P.O. 1981. Contemporary suburban America. Englewood Cliffs, NJ: Prentice-Hall. Mumford, L. 1968. The city in history: Its origins, its transformations, and its prospects. New York: Harvest Books. Murdie, R. 1969. Factorial ecology of metropolitan Toronto, 1951–1961. Research Paper No. 116. Department of Geography, University of Chicago. Nelson, A.C. and K.A. Foster. 1999. Metropolitan governance structure and income growth. Journal of Urban Affairs 21 (3): 309–24. Nicolaides, B.M. 2002. My blue heaven: Life and politics in the working class suburbs of Los Angeles, 1920–1965. Chicago, IL: Univ. of Chicago Press. Niedt, C. 2006. Gentrification and the grassroots: Popular support in the revanchist suburb. Journal of Urban Affairs 28 (2): 99–120. Norris, D.F. 2001a. Prospects for regional governance under the new regionalism: Economic imperatives versus political impediments. Journal of Urban Affairs 23 (5): 557–71.

218 / references Norris, D.F. 2001b. Whither metropolitan governance? Urban Affairs Review 36 (4): 532–50. Nunn, S. and M.S. Rosentraub. 1997. Dimensions of interjurisdictional cooperation. Journal of the American Planning Association 63 (2): 205–19. Ohlemacher, S. 2006. Report: Suburbs nearer to cities neglected. Washington Post, February 15, sec. B. Oliver, J.E. 2001. Democracy in suburbia. Princeton, NJ: Princeton Univ. Press. Orfield, M. 1997. Metropolitics: A regional agenda for community and stability. Washington, DC: Brookings Institution Press. ———. 2002. American metropolitics: The new suburban reality. Washington, DC: Brookings Institution Press. Orser, W.E. 1994. Blockbusting in Baltimore: The Edmondson Village story. Lexington: Univ. Press of Kentucky. Ostrom, E., R. Gardner, and J. Walker. 1994. Rules, games, and common-pool resources. Ann Arbor: Univ. of Michigan Press. Ostrom, V., C.M. Tiebout, and R. Warren. 1961. The organization of government in metropolitan areas: A theoretical inquiry. American Political Science Review 55: 831–42. Ott, T. 2006. Cuyahoga’s old “burbs need work,” report says. Cleveland Plain Dealer, February 15, sec. A. Outen, D. 2005. Baltimore County, Maryland: Using the entire toolkit for habitat protection. In Nature-friendly communities: Habitat protection and land use planning, edited by C.Duerksen and C. Snyder, 152–73. Washington, DC: Island. Ozawa, C.P. 2004. The Portland edge: Challenges and successes in growing communities. Washington, DC: Island. Palen, J.J. 1994. The suburbs. New York: McGraw-Hill. Pascal, A. 1987. The vanishing city. Urban Studies 24 (6): 597–603. Pelissero, J.P., ed. 2003. Cities, politics, and policy: A comparative analysis. Washington, DC: CQ. Perle, E.D. 1981. Perspectives on the changing ecological structure of suburbia. Urban Geography 2 (3): 237–54. Phares, D., ed. 2004. Metropolitan governance without government? Aldershot, UK: Ashgate. Pickles, J., ed. 1994. Ground truth: The social implications of geographic information systems. New York: Guilford. Pinkerton, J.R. (1969). City-suburban residential patterns by social class. Urban Affairs Quarterly 4 (4): 499–519. Porter, D.R. 1999. Will “Smart Growth” produce “Smart Growth?” Abell Report 12 January: 1–8. Puentes, R. and D. Warren. 2006. One-fifth of America: A comprehensive guide to America’s first suburbs. Washington, DC: Brookings Institution Metropolitan Policy Center. Puentes, R. and M. Orfield. 2002. Valuing America’s first suburbs: A policy agenda for older suburbs in the Midwest. Washington, DC: Brookings Institution Metropolitan Policy Center. Reutter, M. 2004. Making steel: Sparrows Point and the rise and ruin of American industrial might. Urbana and Chicago: Univ. of Illinois Press.

references / 219 Roberts, S. 2007. Suburbs are graying faster than big cities. New York Times, June 12. Rochefort, D.A. and R.W. Cobb, eds. 1994. The politics of problem definition: Shaping the policy agenda. Lawrence: Univ. Press of Kansas. Rusk, D. 1995. Cities without suburbs. 2nd ed. Washington, DC: Woodrow Wilson Center. ———. 1996. Baltimore unbound. Baltimore: Abell Foundation. ———. 1999. Inside game, outside game: Winning strategies for saving urban America. Washington, DC: Brookings Institution Press. Sassen, S. 1990. Economic restructuring and the American city. Annual Review of Sociology 16: 465–90. Savitch, H.V. and R. Vogel. 2000. Metropolitan consolidation versus metropolitan governance in Louisville. State and Local Government Review 32 (3): 198–212. Schneider, W. 1992. The suburban century begins: The real meaning of the 1992 election. Atlantic Monthly, July. Schnore, L.F., C.D. André, and H. Sharp. 1976. Black suburbanization, 1930– 1970. In The changing face of the suburbs, edited by B. Schwartz, 69–94. Chicago: Univ. of Chicago Press. Schwartz, A. 2006. Housing policy in the United States. New York: Routledge. Schwartz, B., ed. 1976. The changing face of the suburbs. Chicago: Univ. of Chicago Press. Self, R.O. 2003. American Babylon: Race and the struggle for postwar Oakland. Princeton, NJ: Princeton Univ. Press. Seligman, A. 2005. Block by block: Neighborhoods and public policy on Chicago’s west side. Chicago: Univ. of Chicago Press. Shevky, E. and W. Bell. 1955. Social area analysis. Westport, CT: Greenwood. Short, J.R. 1978. Residential mobility. Progress in Human Geography 2 (1): 419–47. ———. 1991. Imagined country: Environment, culture, and society. London: Routledge. ———. 2007. Liquid city: Megalopolis and the contemporary Northeast. Washington, DC: RFF. Short, J.R., B.F. Hanlon, and T.J. Vicino. 2007. The decline of inner suburbs: The new suburban gothic. Geography Compass 1 (3): 641–56. Short, J.R., K. Yeung, M. Kuus, and H. Wells. 1996. The dirty little secret of world cities research. International Journal of Urban and Regional Research 20: 697–717. Singleton, G.H. 1973. The genesis of suburbia: A complex of historical trends. In The urbanization of the suburbs, edited by L.H. Masotti and J.K. Hadden, 29–50. Beverly Hills, CA: Sage. Smith, J.T. 2003. Renaissance redevelopment press release, December 15. Office of the County Executive. Towson, MD. ———. 2006. Smith seeks council approval to purchase York Park Apartments press release, March 27. Office of the County Executive. Towson, MD. Smith, N. 1996. The new urban frontier: Gentrification and the revanchist city. London and New York: Routledge.

220 / references Smith, N., P. Caris, and E. Wyly. 2001. The “Camden syndrome” and the menace of suburban decline: Residential disinvestment and its discontents in Camden County, New Jersey. Urban Affairs Review 36 (4): 497–531. Squires, G.D., ed. 2002. Urban sprawl: Causes, consequences, and policy responses. Washington, DC: Urban Institute Press. Stahura, J.M. 1984. A research note on the metropolitan determinants of suburban persistence. Social Forces 62 (3): 767–74. Stanback, T.M. 1991. The new suburbanization: Challenge to the central city. Boulder, CO: Westview. Sternlieb, G. and R.W. Lake. 1975. Aging suburbs and black homeownership. Annals of the American Academy of Political and Social Science 422: 105–17. Stoel, T.B., Jr. 1999. Reining in urban sprawl: What can be done to tackle this growing problem? Environment 41 (4): 6–33. Straight S. 2005. 1950s suburb evolves with time. Washington Post, November 12. Sugrue, T. 1996. The origins of the urban crisis: Race and inequality in postwar Detroit. Princeton, NJ: Princeton Univ. Press. Swanstrom, T., C. Casey, R. Flack, and P. Dreier. 2004. Pulling apart: Economic segregation among suburbs and central cities in major metropolitan areas. Living cities census series. Metropolitan Policy Program. Washington, DC: Brookings Institution Press. Teaford, J.C. 1979. City and suburb: The political fragmentation of metropolitan America, 1850–1970. Baltimore: Johns Hopkins Univ. Press. ———. 1990. The rough road to renaissance: Urban revitalization in America 1940–1985. Baltimore: Johns Hopkins Univ. Press. ———. 2006. The metropolitan revolution: The rise of post-urban America. Baltimore: Johns Hopkins Univ. Press. ———. 2008. The American suburb: The basics. New York: Routledge. Thomas, G.S. 1998. The United States of suburbia: How the suburbs took control of America and what they plan to do with it. Amherst, NY: Prometheus Books. Tiebout, C. 1956. A pure theory of local expenditures. Journal of Political Economy 64 (5): 416–24. Time Magazine. 1950. Up from the potato fields. July 3. U.N. World Commission on Environment and Development. 1987. Our common future. Oxford: Oxford Univ. Press. U.S. Department of Housing and Urban Development. 2006. State of the cities data system. Washington, DC. Vale, L.J. 1995. The imaging of the city. Communication Research 22 (6): 646–63. Vicino, T.J., B.F. Hanlon, and J.R. Short. 2007. Megalopolis 50 years on: The transformation of a city region. International Journal of Urban and Regional Research 31 (2): 344–67. Vicino, T.J., S. Sharkey, A. Regan, E. Shockley, L. Gottlieb, B. Hwang, and J. Wilson. 2005. Affordable housing in metropolitan Maryland: A policy analysis. Governor’s Commission on Housing Policy. Annapolis, MD. Visser, J.A. 2002. Understanding local government cooperation in urban regions: Toward a cultural model of interlocal relations. American Review of Public Administration 32 (1): 40–65.

references / 221 von Hoffman, A. and J. Felkner. 2002. The historical origins and causes of urban decentralization in the United States. Report W02-1. Joint Center for Housing Studies. Cambridge, MA: Harvard University. Wade, R.C. 1959. The urban frontier: The rise of Western cities, 1970–1980. Cambridge, MA: Harvard Univ. Press. Wagner, G. 1995. Gentrification, reinvestment, and displacement in Baltimore. Journal of Urban Affairs 17 (1): 81–96. ———. 2001. Gentrification in Baltimore. PhD diss., Univ. of Maryland Graduate School, Baltimore. Warner, S.B. 1978. Streetcar suburbs: The process of growth in Boston, 1870–1890. Cambridge, MA: Harvard Univ. Press. Weir, M., H. Wolman, and T. Swanstrom. 2005. The calculus of coalitions: Cities, suburbs, and the metropolitan agenda. Urban Affairs Review 40 (6): 730–60. Wheeler, T. 2005a. Baltimore’s inner suburbs show signs of aging. Baltimore Sun, May 12. ———. 2005b. Smart growth pace slower in Maryland. Baltimore Sun, March 6. Wheeler, T. and L. Goldberg. 2005. Outcry reinforces importance of Baltimore County growth line. Baltimore Sun, April 3. Whyte, W.H. 1956. The organization man. New York: Doubleday. Wiese, A. 2004. Places of their own: African American suburbanization in the twentieth century. Chicago: Univ. of Chicago Press. Wilson, W.J. 1987. The truly disadvantaged: The inner city, the underclass, and public policy. Chicago: Univ. of Chicago Press. Wolman, H.L. and E.J. Agius, eds. 1996. National urban policy: Problems and prospects. Detroit, MI: Wayne State Univ. Press. Wyly, E. 1999. Continuity and change in the restless urban landscape. Economic Geography 74 (4): 309–38. Yin, R.K. 2003. Case study research: Design and methods. 3rd ed. Thousand Oaks, CA: Sage.

This page intentionally left blank

I n de x

Abrams, Charles, 7 Affordable housing, 36, 155, 160, 171–175, 184 Arbutus, 5, 40, 70–71, 77, 82, 84, 87–88, 89, 92, 100–101, 107, 116, 123, 125, 137, 196–197 Baltimore County Executives Ruppersberger, C.A. Dutch, 150, 152, 157, 159, 161, 166 Smith, James, 150, 152–153, 155, 161, 166 Bartenfelder, Joseph, 154, 156, 161 Bedroom suburbs, 64 Bethlehem Steel Corporation, 3, 12, 40, 96, 107, 127, 142, 151 see also Bethlehem Steel Company, 11 Black suburbs, 9 Boomburbs, 10 Bouton, E.H., 12 Brooklyn Park, 5, 38, 70–71, 74, 82, 84, 87–88, 92–93, 100–101, 105, 196–197 Brown v. Board of Education of Topeka, 31 Burgess, Ernest, 43–44 Cape Cod houses, 13, 23 Carney, T. Kevin, 154 Catonsville, 1, 2, 4, 39, 55, 69, 85–88, 92–95, 100–101, 106, 116–117, 122–123, 125, 130, 136–137, 147–148, 196–198

Chicago School of Sociology, 43–44 Clinton, Hillary, 178–179 Collins, Michael J., 157 Colonial houses, 39 Community Development Corporations (CDCs) Dundalk Renaissance Corporation, 153 Greater Dundalk Alliance, 153 Curran Jr., J. Joseph, 158 Data sources Neighborhood Change Database, 16, 202, 209 State of the Cities Data System, 16, 172, 198 Decline, suburban economic status, 80–86, 186 housing stock, 86–98, 187 labor force, 98–104, 187 population, 69–80, 186 Developers, real-estate Carney, T. Kevin, 154 Jacob, Jack, 154 Riehl, John, 154 Discrimination blockbusting, 27 redlining, 26–27 riots, 29–30 schools, 31–32 urban renewal, 32–33 Dundalk, 141 Dundalk Renaissance Corporation, 153

224 / index ecological model, 50–51 Edge cities, 10 Edgemere, 5, 70–71, 74, 77, 82, 84, 87–88, 89, 91–93, 99–102, 196–197 Edmondson Village, 2, 23, 39 Ehrlich, Robert, 177 Eisenhower, Dwight, 21 Essex, 5, 40, 69–71, 75, 78, 82–97, 100–103, 105, 108, 114–117, 122–124, 141–142, 151–152, 154–161, 196–197 Farley, Reynolds, 49–50 Ferndale, 5, 69–71, 75, 82, 84, 87–88, 92, 100–101, 196–197 first-tier suburbs Arbutus, 5, 40, 70–71, 77, 82, 84, 87–88, 89, 92, 100–101, 107, 116, 123, 125, 137, 196–197 Brooklyn Park, 5, 38, 70–71, 74, 82, 84, 87–88, 92–93, 100–101, 105, 196–197 Catonsville, 1, 2, 4, 39, 55, 69, 85–88, 92–95, 100–101, 106, 116–117, 122–123, 125, 130, 136–137, 147–148, 196–198 Dundalk, 141 Edgemere, 5, 70–71, 74, 77, 82, 84, 87–88, 89, 91–93, 99–102, 196–197 Essex, 5, 40, 69–71, 75, 78, 82–97, 100–103, 105, 108, 114–117, 122–124, 141–142, 151–152, 154–161, 196–197 Ferndale, 5, 69–71, 75, 82, 84, 87–88, 92, 100–101, 196–197 Glen Burnie, 5, 69–71, 75, 82, 84, 87–88, 92, 100–101, 122–123, 196–197 Hampton, 2–5, 69–71, 74, 77–78, 82–85, 87–95, 100, 102, 103, 105, 106–108, 114, 122, 131, 196, 197

Lansdowne, 5, 14, 70–71, 75, 78–79, 82–85, 87–90, 92–93, 95–97, 100–101, 103, 105–106, 108, 115, 122, 196–197 Linthicum, 5, 70–71, 74, 78, 82–84, 87–90, 92, 100–101, 105, 108, 122, 197–197 Lochearn, 5, 39, 70–71, 74–76, 78–79, 82–84, 87–97, 92, 100–101, 115–116, 122–124, 131, 137, 196–197 Lutherville, 5, 40, 69–71, 77, 79, 82–85, 87–90, 92–93, 100–101, 103, 105, 108, 114, 123, 131, 196–197 Middle River, 5, 40, 69–71, 75, 78, 82, 84–85, 87–90, 92–93, 100–102, 105, 108, 114–116, 123–125, 151–152, 156–158, 161, 196–197 Overlea, 5, 40, 70–71, 77, 82, 84, 87–89, 92, 94, 100–101, 116, 123, 196–197 Parkville, 5, 40, 70–71, 75, 78, 82, 84–85, 87–89, 92, 100–101, 116, 143, 196–197 Pikesville, 5, 40, 70–71, 77, 82, 84–90, 92, 100–102, 114, 116–117, 122–125, 131, 196–197 Pumphrey, 5, 70, 71, 74, 82–84, 87–88, 92–93, 100–101, 196–197 Rosedale, 5, 70, 71, 74, 82–84, 87–88, 92–93, 100–101, 116, 196–197 Towson, 3, 5, 39–40, 69–71, 77, 82, 84–85, 87–90, 92, 100–102, 114–117, 122–125, 130–131, 137, 194, 196–197 Woodlawn, 2–5, 39, 69–72, 74–78, 82, 84–88, 92, 100–101, 115, 122–125, 131, 137, 140–141, 196–197

index / 225 Fishman, Robert, 8, 53, 60 Fogelson, Robert, 8

Hrabowski, Freeman, 137 Hudnut, William, 62–64, 169, 171

Galster, George, 47 Garreau, Joel, 7 Gelfand, Mark, 20, 22, 27 General Motors, 21–22, 142, 151 Glen Burnie, 2, 69, 70–71, 75, 82, 84, 87–88, 92, 100–101, 122–123, 196–197 Glendening, Parris, 157, 159, 176–177 Government role, suburbanization housing, 22–25 National Interstate and Defense Highways Act, 21 transportation, 21, 39–41 Greater Dundalk Alliance, 153 Grigsby, William, 47–50 Guest, Avery, 50–51

Immigrant suburbs, 9 Industrial suburbs, 9

Hampton, 2–5, 69–71, 74, 77–78, 82–85, 87–95, 100, 102, 103, 105, 106–108, 114, 122, 131, 196, 197 Hanlon, Bernadette, 6, 65–68 Hannon, Robert, 158 Harvey, Mary, 156, 161–162, 166 Hirsch, Arnold, 35 Holupka, Scott, 153 housing, filtering deterioration, 48 household chance, 47 income change, 47 invasion, 44–51 obsolescence, 48 succession, 44–51 housing styles, suburban Cape Cod, 13, 23 Colonial, 39 McMansion, 14 Multiplex, 98 Rambler, 13 Ranch, 95 Rowhouse, 94–95 Victorian, 39, 94

Jackson, Kenneth, 7, 20, 22, 25–27, 41, 178 Jacob, Jack, 154 Jacobs, Jane, 4, 20 Jones, Carolyn, 153 Keating, Dennis, 62–63, 68 Kelo v. City of New London, 159 Kennedy Townsend, Kathleen, 159 Kruse, Kevin, 33–36 Lakewood, CA, 8 Lansdowne, 5, 14, 70–71, 75, 78–79, 82–85, 87–90, 92–93, 95–97, 100–101, 103, 105–106, 108, 115, 122, 196–197 Lassiter, Matthew, 35–36 Lee, Sugie, 65–67 Legal history Brown v. Board of Education of Topeka, 31 Kelo v. City of New London, 159 Milliken v. Bradley, 32 Plessy v. Ferguson, 31 Leigh, Nancy, 65–67 Levitt, Abraham, 22–24 Levitt, Alfred, 22–24 Levitt, William, 22–24 Levittown, NY, 8, 23, 53, 185 Linthicum, 5, 70–71, 74, 78, 82–84, 87–90, 92, 100–101, 105, 108, 122, 197–197 Lochearn, 5, 39, 70–71, 74–76, 78–79, 82–84, 87–97, 92, 100–101, 115–116, 122–124, 131, 137, 196–197 Logan, John, 49, 51–52 Lucy, William, 57–58, 61, 68, 183, 204

226 / index Lutherville, 5, 40, 69–71, 77, 79, 82–85, 87–90, 92–93, 100–101, 103, 105, 108, 114, 123, 131, 196–197 Malcolm, Dill, 41 Manufacturing suburbs, 9 McMansion houses, 14 Middle America suburbs, 64 Middle River, 5, 40, 69–71, 75, 78, 82, 84–85, 87–90, 92–93, 100–102, 105, 108, 114–116, 123–125, 151–152, 156–158, 161, 196–197 Mikelbank, Brian, 60–61 Milliken v. Bradley, 32 models of change, suburban ecological model, 50–51 persistence model, 49–50 stratification model, 51–52 Multiplex houses, 98 Mumford, Lewis, 20, 185 Neighborhood change, 43–53 see also resident differentiation Neighborhood Change Database, 16, 202, 209 neighborhood clusters in 1970 blue collar, 129–130 minority presence, 129–130 newer middle class, 129, 131 older middle class, 129–130 wealthy, 129, 131 neighborhood clusters in 2000 black middle class, 134, 137 blue collar, 134, 136 middle America, 134, 138 poor, 133–134 university, 134, 137 wealthy, 134–136 Nicolaides, Becky, 34–35 Norris, Donald F., 163, 182

Olmstead, Frederick Law, 39, 154 Olszewski Sr., John, 153–154, 161 Orfield, Myron, 58–60 Overlea, 5, 40, 70–71, 77, 82, 84, 87–89, 92, 94, 100–101, 116, 123, 196–197 Park Forest, IL, 8 Park, Robert, 43–44 Parkville, 5, 40, 70–71, 75, 78, 82, 84–85, 87–89, 92, 100–101, 116, 143, 196–197 Pennsylvania Steel Company, 11 persistence model, 49–50 Phases of development, suburban conformity, 8 decentralization, 11 dichotomy, 9 diversity, 9 frontier, 7 utopias, 8 Phillips, David, 57–58, 61, 68, 183, 204 Pikesville, 5, 40, 70–71, 77, 82, 84–90, 92, 100–102, 114, 116–117, 122–125, 131, 196–197 Plessy v. Ferguson, 31 Puentes, Robert, 62, 65–57 Pumphrey, 5, 70, 71, 74, 82–84, 87–88, 92–93, 100–101, 196–197 Rambler houses, 13 Ranch houses, 95 Regionalism cooperation, 182 Minneapolis, 182, 203 New Jersey, 182 Ohio, 182 revenue sharing, 182

index / 227 Renaissance Initiative Community Conservation, 150–151 Dundalk, 151–154 Essex, 154–156 Middle River, 156–157 Riehl, John, 154 Roland Park, 12, 39, 96 Rosedale, 5, 70, 71, 74, 82–84, 87–88, 92–93, 100–101, 116, 196–197 Rowhouses, 94–95 Ruppersberger, C.A. Dutch, 150, 152, 157, 159, 161, 166 SB 509, 157–160 Schneider, Mark, 49, 51–52 Schneider, William, 185 SCORE Act, 178–180 Self, Robert, 35–36 Senate Bill 509 see also SB 509 Short, John Rennie, 6, 49 Smart Growth Initiative, 171, 176, 181, 183 Smith, James, 150, 152–153, 155, 161, 166 Smith, Neil, 7, 14 Stahura, John, 49 State of the Cities Data System, 16, 172, 198 stratification model, 51–52 Suburban gothic decentralization, 11 deindustrialization, 11–12 demand-supply nexus, 13–15 gentrification, 14 sprawl, 14 Suburbs decline, 67–104 growth, 37–43 history, 20–37 revitalization, 147–167 social critiques, 185–186 Sugrue, Thomas, 33–35

Teaford, Jon, 28–29, 144 Towson, 3, 5, 39–40, 69–71, 77, 82, 84–85, 87–90, 92, 100–102, 114–117, 122–125, 130–131, 137, 194, 196–197 Towson University, 117, 137 Tract developments, suburban Lakewood, CA, 8 Levittown, NY, 8, 23, 53, 185 Park Forest, IL, 8 Tyson’s Corner, VA, 10 Types of, suburbs black, 9 bedroom, 64 boomburbs, 10 edge cities, 10 immigrant, 9 industrial, 9 manufacturing, 9 middle America, 64 working-class, 9 Urban growth boundaries Baltimore County’s Urban-Rural Demarcation Line, 42–43, 171, 174–175 see also URDL Portland’s UGB, 13 University of Maryland, Baltimore County, 94, 117, 147 see also UMBC Vicino, Thomas J., 6, 65–68, 172 Wade, Richard, 7 Warren, David, 62, 65–67 Willeboordse, Jane, 153–154 Wiese, Andrew, 34 Woodlawn, 2–5, 39, 69–72, 74–78, 82, 84–88, 92, 100–101, 115, 122–125, 131, 137, 140–141, 196–197 Whyte, William, 20, 185