University of Southern California Dissertations and Theses

Structure, agency, and the Kuznets Curve: observations and implications for sustainability planning in U.S. cities

(USC Thesis Other)

Structure, agency, and the Kuznets Curve: observations and implications for sustainability planning in U.S. cities

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

STRUCTURE, AGENCY, AND THE KUZNETS CURVE:

OBSERVATIONS AND IMPLICATIONS FOR

SUSTAINABILITY PLANNING IN U.S. CITIES

by

Mark Alan Hanson

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(POLICY, PLANNING, AND DEVELOPMENT)

May 2011

Copyright 2011 Mark Alan Hanson
ii
ACKNOWLEDGMENTS

Completing this dissertation signifies an important milestone in my life, and there are many
people for me to acknowledge in this achievement. I am grateful for the opportunity to
pursue graduate study at USC and to those who prepared me for this opportunity, and even
more grateful to those who stuck with me on this journey.

First, I acknowledge members of my dissertation committee for their patience and
constructive feedback on my varied (and changing) ideas. Committee members included
Tridib Banerjee (chair), Peter Robertson, and Bill Tierney. Professor Banerjee, especially,
has been instrumental in my development as a Ph.D. student from the beginning; without his
inspiration and encouragement at key points along the way, my success is unimaginable.
Several others in the School of Policy, Planning and Development helped me to observe (and
to imagine) the world in new light, with some having influenced me in ways that were not
fully appreciated until years later. These professors include Genevieve Giuliano, Dowell
Myers, Eric Heikkila, Clara Irazabal Zurita (now at Columbia University) and Niraj Verma
(now at University of Buffalo).

I also acknowledge Kenneth Richards and the late Dan Willard at Indiana University’s
School of Public and Environmental Affairs. Professor Willard inspired in me a love for and
fascination with complex environmental and social systems, through the lens of wetlands. I
count myself among the select few who were privileged to know this wonderful man. But it
iii
was Professor Richards, who, as my advisor, “insisted” that I add more policy-oriented
coursework to my schedule, advice I followed and for which I am forever grateful.

I also thank my colleagues at the RAND Corporation, especially Rae Archibald, Catherine
Augustine, Mark Bernstein (now at University of Southern California), Lloyd Dixon, Allen
Fremont, Richard Hillestad, Debra Knopman, Tom LaTourrette, Kevin McCarthy, Paul
Sorensen, Brian Stecher, and Martin Wachs. Among these individuals are some of my
biggest fans, and without their continuous support, encouragement and example, this
achievement may never have come to pass. As well, each of these individuals has helped me
better understand the craft of research and also how to conduct oneself as a researcher
working on some of the world’s most important and difficult problems.

Finally, I thank my family, especially my wife Julie, whose patience was tried most on this
journey, but whose love for me has endured these many years. Words cannot express the
gratitude I have for my parents, Alan and Marilyn, and also my children, Andrew, Anneliese
and Luke, for whom I aspire to provide the same example and opportunities afforded me, and
who motivate me in everything I do.

iv
TABLE OF CONTENTS

ACKNOWLEDGMENTS ii

LIST OF TABLES v

LIST OF FIGURES vi

ABSTRACT vii

INTRODUCTION 1
Problem Statement 4
Organization of This Dissertation 14

CHAPTER 1: LITERATURE REVIEW 15
The Emergence of Sustainable Development 15
Concepts and Challenges of Sustainable Development 21
Sustainable Cities and Theories About Them 35
Structure-Agency Dualism in Cities 43

CHAPTER 2: METHODS AND RESULTS 50
Sample Selection 50
Development of Indicators 55
Models of Kuznetsian Development 61
Modeling the Urban Regime 83

CHAPTER 3: DISCUSSION 95
A Kuznets Curve for U.S. Cities 95
Kuznetsian Development and Sustainable Cities 100
Sustainable Cities and the Urban Regime 103

CONCLUSION 113
Directions for Future Research 114
Implications for Social and Environmental Policy 123

BIBLIOGRAPHY 127

APPENDIX 138
v
LIST OF TABLES

Table 1: Summary of Basic Sustainable Development Components and Objectives 24

Table 2: U.S. Cities with Sustainability Initiatives 53

Table 3: Names and Descriptions of Community Indicators 60

Table 4: Regressions Explaining Class Shares of U.S. City Populations in 2000 by Level
of Development 63

Table 5: Regressions Explaining Class Shares of City Populations in 2000 by Level
of Development and Considering Short-Term Growth 64

Table 6: Regressions Explaining Class Shares of City Populations in 2000 by Level
of Development and Considering Community Indicators 66

Table 7: Regressions Explaining Poverty/Affluence Dissimilarity in 2000 by Level
of Development and Considering Selected Structural Variables 71

Table 8: Comparison of Cities With and Without Sustainability Initiatives for Selected
Variable Counts Normalized by 2000 Population 76

Table 9: Sustainable Cities by Kuznetsian Type 82

Table 10: Regressions Explaining Poverty/Affluence Dissimilarity in Sustainable Cities
in 2000 by Regime Associations and Public Involvement in Planning 91

Table A1: Data Sources and Derivations of Community Indicators 138

Table A2: Descriptive Statistics for Community Indicators for U.S. Cities in 2000 140

Table A3: Mean Values of Community Indicators for Selected Sustainable Cities
in 2000 141

Table A4: Descriptions of Central Objects for Regime Classes, Santa Monica,
California in 2000 142

Table A5: Correlations Among Modeled Regimes of Sustainable Cities by Kuznetsian
Type in 2000 143

vi
LIST OF FIGURES

Figure 1: The Relationship Between Economic Growth and Income Inequality 5

Figure 2: Political Influence and Government Response under Traditional Development
and Sustainable Development Scenarios 11

Figure 3: Alternative Perspectives on Sustainable Development 25

Figure 4: Class Shares of U.S. City Populations in 2000 by Level of Development 63

Figure 5: Influence of Community Indicators on Income Class Shares of U.S. City
Populations in 2000 by Level of Development 69

Figure 6: Influence of Selected Structural Variables on Poverty-Affluence Dissimilarity
in U.S. Cities in 2000 73

Figure 7: U.S. Sustainable Cities and the Kuznets Curve 74

Figure 8: Kuznetsian Development and U.S. Sustainable City Development, 1980-2000 78

Figure 9: A Kuznetsian Typology of Cities 81

Figure 10: Dendrogram of Agglomerative Hierarchical Clustering for Santa Monica,
California 86

Figure 11: Neighborhood Associations and Modeled Regimes for Santa Monica,
California 87

Figure 12: Correlations Among Regimes for U.S. Sustainable Cities in 2000 89

vii
ABSTRACT

Sustainable development has emerged as an important paradigm of planning, and its
economic, social, and environmental goals are being pursued in a growing number of
settings. This dissertation considers the interplay of structure and agency as means to better
understand sustainability planning efforts in U.S. cities. Drawing insight from Simon
Kuznets’ empiric work on the social and economic structure of developing countries
(Kuznets 1955), this study investigates the possibility that Kuznetsian development can
inform our thinking on sustainable city planning as well. Ordinary least squares regression
models reveal the classic “Kuznets Curve” among data for U.S. cities, such that income
inequality first rises then falls with economic growth. The strength of this relationship
increases as several explanatory variables are added, notably educational attainment and
labor skill of the general population, and also when segregation of residents within cities is
considered. Whether or not a city has a sustainability initiative also contributes significantly
to the relationship. These results suggest that Kuznetsian development reasonably describes
the experience of U.S. cities, and the concept provides a framework for studying structure
and agency related to sustainable development in this setting. A three-category typology of
cities is offered such that a given city’s type is determined by its position along the curve.
The typology is used to select and investigate cities for clues about why some of the
wealthier cities in the U.S. that have well-recognized sustainability initiatives appear to be
“making the turn” toward greater equality of income (a social goal of sustainable
development) while others may be at risk of “missing the turn.” Findings of this study
suggest that missing the turn may be explained by a governing elite that has effectively
viii
insulated itself from the influence of middle and lower class citizens in these cities, while
making the turn appears to associate with cities’ efforts to provide for greater public
involvement in governance, and possibly through the efforts of a “progressive middle class
regime” (Stone 1986) to promote social programs in these cities. Implications for sustainable
development theory and policy are discussed, and directions for further research are
suggested.
1
INTRODUCTION

Sustainable development is a major focus of contemporary planning research and practice
(Myers and Banerjee 2005). It remains an “object of planning’s fascination” (Campbell 1996,
p.296) for nearly four decades after the “Limits of Growth” debate of the early 1970s
(Meadows et al 1972), which “discussed whether continuing economic growth would
inevitably lead to environmental degradation and societal collapse on a global scale” (Pezzey
1992, p.1). Issues of sustainable development have stimulated research in many academic
fields and are increasingly reflected in the policies of U.S. business and government. In the
past decade alone, nearly 25,000 articles with “sustainable development” in their titles were
published, predominantly in social science fields.
1
Sustainable development concepts have
also shaped various practices in public and private sectors. Since the phrase “triple bottom
line” (TBL) was coined (Elkington 1994), for example, an increasing number of firms
embrace an expanded view of the costs of doing business along dimensions of environment
and social equity as well as economy. Today, major U.S. corporations such as Patagonia,
General Electric and Pepsi employ TBL accounting procedures and there is growing interest
by government in further promoting this trend (Federal Register 2010). In 2009 the U.S.
Environmental Protection Agency (EPA), the U.S. Department of Housing and Urban
Development (HUD) and the U. S. Department of Transportation (DOT) formed an
interagency partnership to pursue a variety of goals consistent with sustainable development

1
A Internet search on October 29, 2010 using Google Scholar returned 24,258 articles published since 2000
having the phrase “sustainable development” in their title, of which 88 percent of articles were further
categorized into subject areas called “Social Sciences, Arts, Humanities” (14,800) and “Business,
Administration, Finance and Economics” (6,470). Remaining articles (2,988) were found for fields related to
biology, chemistry, physics, engineering and medicine. Between 1990-1999, published articles having
“sustainable development” in their title numbered 11,735 in all, with 65 percent of them in social sciences, arts,
humanities, business administration, finance and economics.
2
principles, especially to improve access to affordable housing and transportation options
while at the same time protecting the environment (EPA 2009). Local governments, whether
by the encouragement by state or federal agencies or by their own initiative, demonstrate
increasing interest in incorporating sustainable development principles into their policies,
programs and plans. The International Council for Local Environmental Initiatives (ICLEI)
reports that membership has grown since this organization began in 1989 to include 1,231
local governments worldwide that have begun to pursue sustainable development; 608 of
these members are cities in the U.S. (ICLEI 2010). To date, at least 31 cities in the U.S. have
adopted “serious” sustainability initiatives (Portney 2003).

A significant branch of the sustainable development literature has focused upon cities that
have committed to sustainable development, so-called “sustainable cities,” and a number of
research themes have emerged around their various experiences. Among notable examples of
sustainable cities research are those which have focused upon the integration of sustainable
development principles into comprehensive planning efforts (Berke and Conroy 2000;
Conroy and Berke 2004), sustainable urban design movements (Garde 2004), and
comparisons of policies and programs related to the economic development, environment,
and social contexts of cities (Portney 2003). Others have considered the varied
implementations of sustainable development (Ryan and Throgmorton 2003), and also the
potential for contemporary planning tools to support sustainability planning in cities
(Campbell 1996). Sustainable cities “readers” summarize a growing literature on
sustainability theory and practice and provide a growing list of case examples relevant to
sustainable city planning (Satterthwaite 2001; Wheeler and Beatley 2004).
3
An observation is that more recent sustainable city research has considered more deeply the
political economy of sustainable cities (Portney 2003; Jepson 2007; Lubell, Feiock and
Handy 2009), suggesting that the informal associations (and dissociations) of people within
cities can provide important insights into the development outcomes of cities. This direction
implies a growing interest in human agency
2
in sustainable cities research. In the context of
cities, agency may describe the actions people take to address a variety of concerns that face
their communities. These actions likely reflect beliefs of local groups and aspirations for
cities to reflect their interests. At the same time agency must consider existing rules and
available resources, or “structures” that constrain or enable actions. In this way, city planning
reflects an interplay of structure and agency whereby participants seek to perpetuate or
change existing policies and plans, but by doing so create new structures that constrain or
enable further agency. This observation is reflected in a number of theoretical frameworks,
especially those associated with various “new institutionalism” theories
3
which have emerged
in social science research. To date, however, sustainable cities research has not fully engaged
ideas of new institutionalism, or more generally, structure and agency and their relationship.
This study suggests that considering the issues of sustainable development of cities and
framing them in light of structure-agency dualism holds promise for advancing sustainable
cities research.

2
Human agency is broadly defined as the capacity of individuals or groups of individuals to act.

3
Hall and Taylor (1996, p.936) describe new institutionalism theories as an approaches that “seek to elucidate
the role that institutions play in the determination of social and political outcomes.” They further differentiate
these theories among three fields of thought—historical institutionalism, rational choice institutionalism,
and sociological institutionalism—with each of them taking a different position on the meaning of structure,
agency and their integration, but affording these concepts great importance nonetheless.
4
Problem Statement
This dissertation considers the interplay of structure and agency as a means to better
understand sustainability planning efforts in U.S. cities. It seeks to model social and
economic structures of developing cities, discern from patterns revealed in these models
clues about human agency that may explain these structures, and thus reveal insights relevant
to further research, planning and policy in these cities. This study narrows its focus to
consider a subset of commonly promoted sustainable development goals, namely the pursuit
of greater equality of income (a social goal) with economic growth (an economic goal). This
study is limited further to the experiences of U.S. cities during the 1980-2000 period,
timeframe which spans a number of pioneering sustainable city efforts.

An initial step in this study is to consider sustainable development as a theory of
development more broadly. This study draws especially from the work of Simon Kuznets
who first identified among select developing countries an inverted U-shaped curve among
cross-sectional data describing income inequality and economic growth (Kuznets 1955). That
is, income inequality increases at lower levels of economic development, but decreases at
higher levels of economic development. The classic “Kuznets Curve” is illustrated in Figure
1. Kuznets imagined from this pattern a process of development that is reinforced over time,
and drawing from the experiences of select countries, suggested that several underlying
social and economic factors explain this pattern. These factors reflect differing abilities
among income class groups to save, urbanization processes, demographic shifts,
technological innovation, and emergent economies. Kuznets specifically identified policies
of progressive taxation, various social programs, access to education and political
5
processes—all of which tend to be found in counties demonstrating more advanced stages of
development (Kuznets 1955, 1966)—as important contributors to decreased income
inequality with further economic growth. Such policies are of course the product of human
action, and in this way the Kuznets Curve is judged to reflect possible clues about social and
economic structures and human agency.

Figure 1: The Relationship Between Economic Growth and Income Inequality

Kuznets (1955)

Since its introduction, many researchers have focused on empirically demonstrating the
curve’s existence and testing Kuznets’ claims. Among them, Ahluwalia (1976a,b) presents
several statistically significant relationships among variables that contribute to the Kuznets
relationship. Anand and Kanbur (1993) use different variables to specify the Kuznets Curve’s
form. Acemoglu and Robinson (2002) provide a political economy theory of its turning
point. Nielsen and Alderson (1997) translate the concept to models of smaller geographic
areas. And, Grossman and Krueger (1991) extend the idea to consider other measures
Income Inequality
Turning Point
Economic Growth
6
commonly associated with economic growth, namely environmental degradation. Despite
this apparent flurry of research activity, however, the Kuznets Curve has not been applied to
the experience of cities and a hope for this study is that doing so may provide some insight
into structure and agency in them, especially in those which have committed to policies of
sustainable development. Thus, a premise for this study is that a relationship in the manner of
the Kuznets Curve exists at the level of cities, and that its turning point can be explained by
economic and social structures related to sustainable development policy and planning. An
initial set of research questions ask,

1. (a) Is there a relationship between income inequality and economic growth in a cross-
section of U.S. cities? (b) Does this relationship resemble an inverted U-shaped
curve, or Kuznets Curve, in which income inequality first increases then decreases
with economic growth? (c) If so, does the curve largely reflect economic and social
structural conditions that are reinforced over the long term? And, (d) is the
relationship between income inequality and economic growth further influenced by
commitment to policies of sustainable development?

The task of specifying a Kuznets Curve for U.S. cities in this study is informed by a review
of Kuznets’ work and that of others, and adapting ideas from them appropriately to the
context of cities. In part, this task entails the development of suitable indicators of income
inequality and economic growth in cities. The task also entails judging the “goodness of fit”
of data to various modeled relationships in an attempt to determine whether the concept of
7
Kuznetsian development is generally applicable to the experience of a cross-section of U.S.
cities.

Assuming that the Kuznets relationship between income inequality and economic growth is
evident among cities more generally, a second set of explorations seeks to determine the
relevance of the Kuznets Curve to the development of sustainable cities in particular. Note
that Kuznets’ idea emerged from observations of historical data that described the
development of countries, not cities, and over a much longer period of time than has elapsed
since the sustainable cities movement began. That this idea translates to the context of cities
may be a bold assertion. Nevertheless, the Kuznets Curve is considered in this study for its
possibility to inform our understanding of the experiences of sustainable cities, potentially to
reframe our thinking about them in a manner that Kuznets was able to achieve in thinking
about developing countries. A Kuznets Curve constructed from city-level data is expected to
inform a useful taxonomy and a basis for selecting cases which, upon further consideration,
may reveal insights into the processes that lead to various development outcomes observed in
these cities. Accordingly, a second set of research questions ask:

2. (a) May a relationship between income inequality and economic growth inform a
meaningful typology of cities that can guide further investigation of human agency in
cities? (b) Could this typology be used to select U.S. cities that have committed to
sustainable development policies and to investigate why some cities appear to be
achieving social goals of sustainable development, while others are not?

8
For example, sustainable cities might position themselves as a group along the Kuznets
Curve in a manner that generally reflects lower or higher levels of economic development as
a group, by comparison to other cities. Or sustainable cities may distribute themselves more
evenly along the curve, such that some sustainable cities reflect lower levels of development,
while others may reflect higher levels of development, and among these more affluent
sustainable cities, some might be further characterized as having “made the turn” or “missed
the turn” as indicated by their level of income inequality beyond the curve’s turning point.
This bifurcation among sustainable cities by comparison to the experiences of others is
expected to usefully direct our attention to consider underlying social factors and policy
processes that explain these two contrasting outcomes.

Remaining hypotheses in this study explore aspects of human agency as may be related to
outcomes. Several political theories are considered for their relevance and plausibility in
explaining evidence from select sustainable cities. These considerations adapt and extend the
thinking of Kuznets (1955, 1966) and also Acemoglu and Robinson (2002) to the context of
cities. As suggested above, a number of mechanisms have been identified that explain the
Kuznets Curve. A focus of this research is especially upon how political mobilization of non-
elite groups may serve to “force democratization on political elites” (Acemoglu and
Robinson 2002, p.183) in a manner that makes available the opportunities created by
economic development to non-elite groups, such that income inequality ultimately decreases
with development. Kuznets’ self-described “wishful thinking” (Kuznets 1955, p.26)
describes such an outcome. But Acemoglu and Robinson (2002, p.184) suggest that
continued high level of inequality with development may persist under conditions of
9
“autocratic disaster,” whereby political mobilization and participation in governance remains
low over the development course. Given the various, often competing goals of sustainable
development, and the number and variety of cities in the U.S., it is expected that sustainable
cities in the U.S. comprise an important subset of experiences that may reflect the wishful
thinking of Kuznets in some cases, and approach autocratic disaster in others. These
potentially varied outcomes may have implications for the success (or failure) of policy
reform in these cities to achieve social goals of sustainable development.

In this research, the mechanism described above is related to the context of sustainable cities
through the “regime theory” of Stone (1986). That is, the political economy of sustainable
cities is modeled as a collection of interest-based “urban regimes” that align with city
government to influence planning processes to varying degrees, in some cases to support
development that favors the interests of a land-owning elite (Molotch 1976), and in others to
favor middle and lower class interests in policies that more aptly reflect social goals of
sustainable development (Portney 2003). According to Molotch, the former is more likely to
occur than the latter, however, making the pursuit of sustainable development in cities an
uphill political battle. Portney, on the other hand, affords great promise to the efforts of a
“progressive middle class regime” (Stone 1986) in promoting policies of sustainable
development. To explore these contrasting perspectives, the informal associations and
influences of regimes upon governance of cities are explored in this study. The associations
that create regimes are modeled on the basis of cultural similarity of members, consistent
with an “advocacy coalition” view of the policy process (Sabatier and Jenkins-Smith 1989).
Regimes are further imagined as spatially defined by the neighborhoods of cities, consistent
10
with the perspective that neighbors create “political spaces in which they are allied with like-
minded citizens to fight political battles” (O’Reilly and Webster 1998 in Gimpel and
Schuknecht 2004, p.2). Thus, a third set of research questions explored in this study is:

3. (a) Do culturally separable groups of residents form within sustainable cities? (b) Do
these groups plausibly describe interest-based regimes that sort among neighborhoods
of these cities, with least cultural difference observed between neighborhoods that
house elite class and city officials? And, (c) is greatest cultural difference observed
between the land-owning elite and lower class regimes?

In an advocacy coalition-based view, cultural differences account for differences in core
belief systems among politically active groups, hence differing policy interests, and thus
difficulty in implementing policy. Policy change becomes largely a challenge of negotiating
differences among groups to arrive at consensus decisions. This study explores the possibility
that regime associations exist within cities in a manner that explains variable success in
arriving at policies that pursue social goals through informal influences among regimes upon
these negotiations. This mechanism is considered alongside another by which city
governance is influenced through formal efforts by the city to involve various stakeholder
groups in planning. Both of these mechanisms are further imagined to occur within two
development scenarios—called “traditional” and “sustainable” in this study—and in terms of
the amenities and opportunities, or “services” delivered to various groups by cities. The first
mode of development, traditional development, is assumed to explain increasing income
inequality with development, while the second, sustainable development, is assumed to
11
explain the reverse. A conceptual diagram of political influences by urban regimes and
government response for two development scenarios is presented in Figure 2.

Figure 2: Political Influence and Government Response under Traditional Development
and Sustainable Development Scenarios
Traditional Development Sustainable Development
Political Influence

Government Response

Notes: City government (G), elite (E), middle (M) and lower (L) class regimes are shown as culturally similar
groups within a city. Arrows indicate direction of influence upon city officials by regimes and delivery of
services by the city to regime groups.

In Figure 2, regime groups are shown by the smaller circles, enclosed by larger circles that
represent the cities. The larger circles have dotted lines to reflect a given city’s relative
physical and political separation from surrounding cities and unincorporated areas, but some
relation to various concerns outside their immediate jurisdiction. These might for example
include county, state, and federal governance issues, economic and transportation linkages,
shared environmental resources and concerns, and so forth. Four regime groups—city
M
EG
L
M
EG
L
L M
E G
M
E G
L
12
government, elite, middle and lower classes—are indicated by labels, G, E, M, and L,
respectively. Their solid lines suggest an internal coherence based on their shared beliefs,
interests and common purpose. Arrows indicate direction of influence by these groups upon
others, and also delivery of services to them by city government. Dotted lines indicate weak
or indirect influence, or fewer services delivered, while solid lines indicate strong or direct
influences, or more services delivered.

Under a scenario of traditional development, the elite class is theorized to have great
influence over city government, and in return elites receive more services. By contrast, the
lower class exhibits less influence upon government and receives fewer services. Under a
scenario of sustainable development, all classes influence government and receive services
more equally because of conditions created by this mode of development. Furthermore, under
sustainable development the nature of influence is of two types. The first describes
involvement of the general public in city governance. Call this the “public participation
thesis.” The second describes influence through regime associations, especially through
efforts of a middle class that is sympathetic to lower class concerns while also having the
ability to influence the governing elite. Call this second type of influence the “political
mobilization thesis.”

Additional explorations in this study consider how these mechanisms may explain
development outcomes that reflect higher or lower income inequality in sustainable cities that
have experienced higher or lower levels of economic growth. A final set of research
questions in this study is as follows:
13

4. Is income inequality especially in more affluent sustainable cities explained by (a) the
degree of cultural separation between elite and low-income regimes? (b) Is the
potential for groups within sustainable cities to negotiate political differences,
ultimately to arrive at policies that promote greater equality, enabled by (i) efforts to
involve the general public in city planning, or (ii) influence of a “middle-class
progressive regime” (Stone 1986)? (c) Is it evident among cities that income
inequality increases with cultural separation of elite and low-income regimes, and
decreases with evidence of greater public involvement and cultural resemblance and
associations among middle class and other class groups?

In keeping with a commitment to the idea of structure-agency dualism over the development
course of cities, this study is careful not to forsake one for the other in its various
explorations. Thus in progressing from the first two hypotheses to the second two, this study
employs “methodological bracketing” (Giddens 1989). Methodological bracketing describes
a “reflexive moment of attention in which a researcher intellectually freezes the ongoing
course of structuration” (Phipps 2001, p.194). That is, the first methodological bracket holds
constant the activities of humans to investigate structure, while the second holds constant
structure to investigate agency. In this manner, the structural aspects of sustainable cities are
revealed by bracketing agency in the first two steps. In subsequent steps, structure is
bracketed as the agents within select sustainable cities are investigated further.
Methodological bracketing is facilitated in this study by the selection of sustainable cities
according to their locus on a Kuznets Curve for U.S. cities.
14

Organization of This Dissertation
I have organized this dissertation into five main sections, with several appendices. In this
section, I introduce the topic, state the problem, and disclose the organization of the
dissertation. In Chapter 1, I present the origins of sustainable development and its key
concepts and associated challenges, and also review research on the Kuznets Curve and
several theories of urban political economy. In Chapter 2, I discuss a variety of methods,
along with the selection and processing of data used in this study. I also present results in
Chapter 2. In Chapter 3, I discuss results in light of theories previously described. I conclude
by summarizing key findings and identify some contributions to the literature on sustainable
development of cities that my findings suggest. I also make some observations about social
and environmental policy in cities, and also offer some recommendations for policymakers
and future research. Appendices include additional details on data, analysis and results.
15
CHAPTER 1: LITERATURE REVIEW

In this chapter, first the origins of sustainable development are reviewed and its theoretical
linkages to the Kuznets Curve presented. Next, several concepts and challenges of
sustainable development are described. Sustainable cities research is reviewed, noting
especially the increasing application of urban theory to explain sustainability planning in the
context of cities. Finally, theories of structure and agency are presented as means by which
sustainability planning in cities may be further understood.

The Emergence of Sustainable Development
Sustainable development is an idea of development more generally. Specifically it is an
alternative theory of development that was most famously presented in a 1987 report by the
World Commission on Environment and Development (WCED), “Our Common Future”
(also known as the Brundtland Report
4
). The report sought to raise awareness among political
leaders of the world on environmental issues related to international development. The report
preceded the historic Earth Summit in Rio de Janeiro, Brazil in 1992, which, among several
important achievements, resulted in two “blueprints” for action at local, national and global

4
The reference is to Gro Harlem Brundtland who chaired the commission.

16
levels that intend to achieve goals of sustainable development. These are Agenda 21
5
and the
Kyoto Protocol
6
.

While the thinking on a variety of issues related to sustainable development predates these
historic events, these events are generally understood to have provided an important thrust
forward for the sustainable development movement worldwide. Today at least 191 countries
have ratified the Kyoto Protocol (United Nations 2010). While the U.S. is currently not
among the list of ratifying parties, 1,044 mayors across the U.S. have signed the Climate
Protection Agreement, committing to the same greenhouse gas emission reduction goals
specified in the Kyoto Protocol (USCM 2010). By 2001, a total of 6,416 municipalities (87 in
the U.S.) reported involvement in Agenda 21 processes (WRI 2010).

That sustainable development is presented as an alternative theory of development begs the
question: “Alternative to what?” In reply, sustainable development describes largely a
response to conditions created by “traditional development thinking” (Harris and Goodwin
2001, p.xxxiii), that was observed especially in development following WWII. This type of

5
Agenda 21 is a program of the United Nations aimed at global, national and local communities to address
issues related to environmental impacts. Specific areas of policy action concern social and economic
dimensions, conservation and management of resources for development, strengthening the roles of major
groups, and implementation.

6
The Kyoto Protocol (United Nations 1998) provides means by which to stabilize “greenhouse gas
concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the
climate system” (United Nations 1992). It is a legally binding protocol to the international environmental treaty,
the United Nations Framework Convention on Climate Change that establishes a commitment to reduce
greenhouse gases.

17
development has been described as being inspired by ideas of “modernization” and
“modernism.”
7

Traditional development approaches focus primarily on social improvement of the masses,
notably through a linear conception of development whereby countries pass through a series
of economic stages (Rostow 1960). According to Rostow, the process begins with a less
developed society, one in which agriculture is the dominant mode of production and where
output is consumed by producers rather than traded. In the next stage, increased
specialization of labor generates surpluses that can be traded, and infrastructure for trade
emerges. Then, economic “take-off” occurs, in which manufacturing outgrows agricultural
production. Further growth is concentrated in a few regions and industries. Technological
innovation generates new goods and services. Investment opportunities grow, and a more
diverse and mature economy develops. In its advanced stage, the economy is characterized
by mass consumption, flourishing consumer industries, and an active service sector.

As Harris and Goodwin (2001) explain, the vision represented by this final stage, and the
model that leads to it, was widely accepted and reflected in development policies of
organizations which emerged after WWII, such as the International Monetary Fund, the
World Bank and the United Nations. These organizations provided the substantial capital
investment required to transform postwar economies in the manner described, especially to
those countries willing to promote the social and political institutions that supported the

7
According to Berman (1982, p.16), “modernization” is the transformation of human society and the natural
environment that traces to great discoveries in the physical sciences and industrialization of production, most
conspicuously in the twentieth century. “Modernism” is the collection of “visions and values” associated with
modernization.
18
large-scale industrialization also required (i.e. foreign aid to these countries was tied to their
achieving the various stages above.)

Progress by the traditional development approach has generally been monitored at national
scale, in terms of overall health, education and Gross Domestic Product (GDP), especially by
use of the United Nations Development Programme’s Human Development Index which
combines these measures. The results of postwar development are summarized by Harris and
Goodwin (2001, p.xxxii):

Globally, most countries have made significant advances both in GDP and in
Human Development Index measures. But overall, the record of development
on a world scale is open to two major criticisms:

• The benefits of development have been distributed unevenly, with income
inequalities remaining persistent and sometimes increasing over time. The
number of extremely poor and malnourished people has remained high,
and in some areas has increased, even as a global middle class has
achieved relative affluence.

• There have been major negative impacts of development on the
environment and on existing social structures. Many traditional societies
have been devastated by overexploitation of forests, water systems, and
fisheries. Urban areas in developing countries commonly suffer from
severe pollution and inadequate transportation, water, and sewer
infrastructure. Environmental damage, if unchecked, may undermine the
achievements of development and even lead to collapse of essential
ecosystems.

Among the earliest observations of the impacts of development were those of Simon
Kuznets. Recall that the Kuznets Curve, shown previously in Figure 1, describes an inverted
U-shaped curve that depicts increasing income inequality with initial economic growth, and
decreasing income inequality after a turning point in the curve. While this curve is
constructed using cross-sectional data, thus representing a “snapshot” view of the
19
experiences of developing countries, it is often interpreted to inform thinking on
development processes that are reinforced over time. According to Kuznets (1955), two main
explanations contribute to the observed pattern. First, greater savings by people in upper
income classes concentrates wealth in these groups. Lower income classes are not able to
achieve these same levels of savings, and this unequal distribution of wealth is perpetuated in
future generations. Second, the processes of industrialization and urbanization reinforce this
pattern. This latter mechanism is in part explained by a shift from agricultural to industrial
economies—the “take off” that Rostow describes—such that wealth in the new economy
accrues largely to an elite class. Because elite are more likely to live in cities, income
inequality is observed to be greater in urban areas than in rural areas (income levels are
generally lower in rural areas, but more evenly distributed.) But eventually, continued growth
and development decreases overall income inequality, such that lower income classes,
especially in cities, are able to take advantage of the opportunities created by development.

Several mechanisms explain patterns of decreasing income inequality observed especially in
more advanced stages of development. One mechanism identified by Kuznets (1955) is that
of demographic shift in the general population, such that members of upper income class
groups account for a decreasing share of the total population. Total population of an area
grows, more by the additions to middle and lower class groups, especially by those who seek
greater opportunities. Related to this, higher-income industries emerge, thus creating
opportunities for workers to move into them from lower-income industries. Note that the
availability of education and training are important contributing factors that help to make
these shifts among industries possible, especially toward higher-income industries that
20
require greater training and skill. And, according to Kuznets, an obvious mechanism that
explains decreasing income inequality is “legislative interference and political decisions”
(Kuznets 1955, p. 9). This mechanism limits the accumulation of wealth by elites or
otherwise serves to redistribute wealth in society, for example by progressive taxation, rent
control, and lowering of interest rates.

Moreover, Kuznets identifies three basic political requirements that ultimately allow
economic growth to proceed: stability, flexibility, and representativeness. “One could hardly
expect much economic growth under conditions of political turmoil, riots, and unpredictable
changes in regimes” (Kuznets 1966, p.451) for example. Flexibility, especially to “resolve
the conflicts of interest that inevitably arise in the course of development” and
representativeness, to afford “some means of expression so that their interests can be
considered by government when its broad decisions are made” are also identified as
“minimum requirements” to advance development. Yet,

Too much stability of political structure, which usually results in
inflexibility or unresponsiveness, may be as inimical to modern economic
growth as too little stability; too much flexibility—response to economic
change without careful consideration of costs to those affected by it—may
be as destructive of economic as too little flexibility; and unselective
representativeness—greater concern with short-term interests of some
groups than the longer-term future—may constrain economic growth as
much as too little representativeness. (Kuznets 1966, p.452)

Thus, Kuznets identifies a role for civil society to create conditions under which development
can proceed, especially to respond to needs of multiple interests of communities over the
course of their development. Participation of various groups in governance processes and
21
thoughtful decision making are important, and without these, concerns about development—
especially the social environmental concerns noted previously by Goodwin and Harris—are
likely to emerge. Indeed the importance of public involvement in development is highlighted
in many “failed” development projects reviewed by Scott (1998). Scott notes that various
problems of development become particularly acute in the context of an authoritarian state
and a disengaged civil society. Acemoglu and Robinson (2002) also identify low political
mobilization as leading to development outcomes described as “autocratic disaster,” an
outcome characterized as having high income inequality and low output. Political action of
groups, and human agency more generally, thus emerges as an important mechanism by
which the wishful thinking of Kuznets can be realized. Although the term sustainable
development was not in use at the time Kuznets published his work, many proponents of
sustainable development appear to have similar aims for development, including the more
egalitarian conditions that may exist beyond the turning point of the Kuznets Curve.

Concepts and Challenges of Sustainable Development
Campbell (1996, p.296) writes that, appropriately defined and “incorporated into a broader
understanding of political conflicts in industrial society, [sustainable development] can
become a powerful and useful organizing principle for planning.” Yet, while sustainable
development advocates generally agree that the uneven distribution of development benefits,
impacts on the environment and social structures are unacceptable, agreement on what
sustainable development should achieve, how to achieve it, and even what sustainable
development really is, continue to be matters of considerable debate.

22
Nearly twenty years ago, a World Bank report (Pezzey 1992) listed 33 different definitions of
sustainable development (and the author claimed the list not to be exhaustive). In the decades
that have followed, it is difficult to know how the actual number of definitions may have
changed, but the number of definitions today is certainly not one. Inferring from a growing
ICLEI membership, for example (see Introduction), commitments to sustainable
development by communities worldwide have increased since release of the World Bank
report. Local adaptations and new operational definitions for sustainable development are
likely to have increased as well. Highlighting the variety among local implementations of
sustainable development, Ryan and Throgmorton (2003) review the experiences of two
cities—Chula Vista, California and Freiburg, Germany—whereby efforts to apply
sustainable development concepts resulted in nearly opposite land use and transportation
policies. The promise of sustainable development to serve as an “organizing principle” brings
with it many challenges, if our common future implies also a pursuit of common goals.

To avoid the problem of defining sustainable development for this study, I will follow the
lead of many scholars, resigning as others have done, to consider the most famous definition
of sustainable development cited in the Brundtland Report (WCED 1987, p.43):

Sustainable development is development that meets the needs of the present
without compromising the ability of future generations to meet their own
needs. It contains within it two key concepts:

• the concept of ‘needs’, in particular the essential needs of the world’s
poor, to which overriding priority should be given; and
• the idea of limitations imposed by the state of technology and social
organization on the environment’s ability to meet present and future
needs.

23

Several aspects of this definition are found in the many other definitions available, and from
the Brundtland Report definition some generally accepted characteristics of sustainable
development can be inferred. From this definition, for example, sustainable development is
inherently normative, despite its origin in empiric observations about social and
environmental impacts of development, as described previously. It names the poor as those
for which priority “should be given,” thus highlighting social justice as an issue of central
importance. It also recognizes limits, suggestive of an environmentally deterministic nature
of the concept, and a role for science and technology especially to address social and
environmental concerns.

More generally, sustainable development is portrayed in a manner that implies a systems
view
8
that extends across both time and space, especially in that intergenerational needs must
be met, and that “local concerns are connected to broader concerns” (Berke and Manta-
Conroy 2000). Furthermore, sustainable development is generally presented as three distinct,
yet interrelated components associated with economic, environmental, and social concerns.
These are commonly understood to be the “three Es” of sustainability (“economy,”
“environment,” and social “equity”). The view is generally depicted either as a “planner’s
triangle” (Campbell 1996) or as a system of interlocking rings, such that each “E” is a vertex
or ring. Sustainable development is generally shown or implied to occupy the center of such
illustrations as a goal to be pursued or a balance point to be achieved. The three Es model of
sustainability thus seeks to broaden our focus on multiple and related goals of development

8
Systems thinking describes that which places greater emphasis on holism than on reductionism, where
individual problems are viewed as part of a set of related problems that must be solved together.
24
and pursue them simultaneously, rather than those which traditional development thinking
has too narrowly and linearly defined. The maxim “Think globally, act locally” calls
attention to the multiple scales and issues of relevance to sustainable development thinking
(and to take action with these in mind). Table 1 summarizes general components and
objectives of sustainable development that are often discussed in the literature.

Table 1: Summary of Basic Sustainable Development Components and Objectives
Component Objectives
Economic • Produce goods and services
• Maintain appropriate levels of government and debt
Environmental • Maintain ecosystems and biological diversity within them
• Maintain a stable resource base, especially to avoid over-exploitation of non-
renewable natural resources
Social • Achieve distributional equity
• Provide appropriate levels of health, education, and other social services
• Achieve appropriate levels of political participation and accountability

Challenges over competing objectives of sustainable development
But a much larger list of development concerns exist than can be considered or acted upon.
We live in a world of limited resources and “bounded rationality.”
9
The numerous issues and
greater political involvement that sustainable development requires present great challenges
for deciding priorities and appropriate policy actions to be taken. This challenge is not helped
by the ambiguity of the concept. Levett (1998) notes in particular that the most commonly
recognized interlocking rings model provides little information on the relative importance of
economic, environmental or social objectives, and therefore an order in which they should be
pursued. Levett has called this model the “three ring circus” model of sustainability.

9
Basic human intellectual and psychological limitations require that “the capacity of the human mind for
formulating and solving complex problems is very small compared with the size of the problems whose solution
is required for objectively rational behavior in the real world” (Simon 1957, p.198).
25
Invoking the concept of limits, Levett (1998, p.295) promotes instead a rearrangement of the
three Es to an alternative “Russian dolls” configuration, arguing that

…environment is a precondition for the other two. Without the planet’s basic
life-support systems there can be no economy or society…economy…is a
social construct. It only works as it does because humans have created the
institutions and inculcated the assumptions, expectations and behaviors which
make it so. The only reason for keeping it thus is if we think that it will be
good at meeting our needs.

Figure 2 illustrates these two contrasting perspectives on sustainable development. Note that
the Russian dolls configuration of sustainable development turns traditional development
thinking on its head, placing greatest importance on environmental considerations rather than
economic growth.

Figure 3: Alternative Perspectives on Sustainable Development

“Three Ring Circus”

“Russian Dolls”

Source : Levett (1998)

Economy
Environment Society

Economy
Society
Environment
26
Levett is not alone in his thinking about the primary importance of the environment to
society. Diamond (2005) reviews 15,000 years of human history to provide evidence that this
is so. In his study, Diamond identifies five sets of factors that contribute to societal
collapses.
10
Environmental factors were common to all major collapses (although
environmental factors were not necessarily the major cause for collapse.) Indeed observations
about increasing pressures of population and economic growth upon environmental resources
were the focus of great attention by the Club of Rome,
11
whose first report highlighted
various scenarios for the longer term future of society and the planet, and was an important
catalyst to the sustainable development movement.

The environmental concerns described by Diamond and others have their fundamental basis
in the ecological concept of resilience. Resilience refers to the ability to recover from threats
to production, such as disturbances by human development. According to Holling (1994),
resilience is directly related to the genetic diversity of species, individually and together, that
allows ecosystems to adapt to change, and therefore to survive. To ecologists, it is this ability
to adapt to change over time, thus a focus on maintaining ecosystem resilience that should be
the primary objective of sustainable development. Sustainable development definitions, such
as that of the Brundtland report, recognize a very specific concern related to resilience, that
of environmental limits. Holling (1994, p.84) notes that “organisms are exuberantly over-
productive, and that limits set by time, space, and energy are invariably encountered.” That

10
These factors are local environmental damage, climate change, attack by hostile neighbors, decreased support
by friendly neighbors, and response to societal problems (Diamond 2005).

11
The Club of Rome is a global think tank that focuses upon issues of global importance, particularly issues
related to economic development and environmental resources. Scientists affiliated with this organization
authored its most famous publication, “The Limits of Growth” (Meadows et al 1972).
27
is, there is an inherent “carrying capacity”
12
of an ecosystem that imposes limits on
production. Whereas plants and animals routinely bump up against these limits on their own,
often with adverse consequences to individuals and populations,
13
an objective of planning
for sustainable development according to an ecological perspective is to design and manage
human-environmental systems such that these adverse consequences are largely avoided.
Designing and managing human-environmental systems is expressed in the theory of “design
with nature” (McHarg 1969), which is built around the concept of a “thermodynamic
imperative,” described as follows:

All systems are subjected to the necessity of finding the fittest environment,
adapting it and themselves to make it more fitting. A fit environment is
defined as that where the maximum needs of a user are provided by the
environment as found, requiring the least work of adaptation. Successful
evolution contains a least work solution. The achievement of evolutionary
success reveals syntropic fitness and health of species and ecosystems.
Excessive pathology and morbidity reveal entropic misfit—a system unable to
find the fittest environment, unable to adapt to itself (McHarg 1969, p.v).

Accordingly, an important focus of planning for sustainable development is on achieving the
fitness that is required to maintain resilience of ecological systems, especially to recognize
the finite nature of many resources required to sustain life on earth, including land, food,
clean air and water, and creating development designs with these specifically in mind.

12
Carrying capacity describes limits imposed by environmental resources—e.g. food, water, habitat—that can
sustain a population of given size in the long term.

13
A common example in ecology texts is that of the St. Paul reindeer (Scheffer 1951) illustrating both
exponential growth, followed by crash, when a population outstrips its resources.

28
But such ecological views have been criticized for not considering social systems and
practical needs of human society. The “entropic misfit” of depleted soil is well-known, for
example, to the subsistence farmer who nevertheless continues to work his soil despite
decreasing yields. Education on this point does not implore him to fallow his land, if this
means starving his family. Indeed the most commonly accepted definition of sustainable
development, i.e. the Brundtland Report definition, gives priority to those in greatest
economic need. As Rolston (1996, p. 248) explains, the ethical dilemma is pinpointed on a
bumper sticker which reads “Hungry loggers eat spotted owls.”

Nonetheless, overexploitation of environmental resources may ultimately lead to societal
collapse in the long term. Sustainable development thus comprises both long-term and short-
term views that reflect conflicting imperatives for planning. Attempting to reconcile multiple
objectives of development can lead to some counterintuitive, even provocative policy
recommendations for different situations. For example, despite the growing consensus over
human causes and consequences of climate change, Schelling (1997, p.8) remarks about
developing countries: “Their best defense against climate change and vulnerability to
weather in general is their own development, reducing their reliance on agriculture and other
such outdoor livelihoods.” Verma (1996) likens the challenges of sustainable development to
the dilemma faced by a motorist approaching an intersection when the light changes to
yellow. Does the yellow light mean slow down, or speed up? Depending on the situation it
could mean either.

29
Challenges over competing perspectives on human welfare
Human welfare is a topic that is considered in many fields, and the challenge it presents for
development reflects many viewpoints as well. To the neoclassical economist, human
welfare is believed to be maximized through consumption of goods and services through an
efficient
14
allocation of resources. Important to this perspective is the idea of a discount rate,
such that the cost of something now, expressed in monetary terms, will increase in the future.
This perspective also assumes a limitless supply of inputs (including substitutes) and
generally focuses on those benefits and costs that are more easily measured. From the
perspective of neoclassical economics, then, borrowing from future generations (in the form
of resource depletion and environmental damage in the current generation) may be
acceptable. These central ideas—discounting, limitless inputs, measurement—lie at the heart
of what sustainable development scholars have termed the “expansionist paradigm” (Rees
1995). The concern, of course, is that many environmental resources are limited, therefore
production cannot be sustained indefinitely. Social and environmental costs, though difficult
to quantify, are important to consider and without them true costs of production are
underestimated in this view. Mostly, however, sustainable development advocates agree that
passing on social and environmental problems to future generations is not acceptable. For
these reasons, many argue that neoclassical economics is particularly ill-suited to support
sustainable development thinking.

14
Economic efficiency implies (at least) two important concepts for purposes here, first that no more output is
produced without further inputs, and second that production proceeds first, and in order, using inputs with
lowest unit cost.
30
The opposition between the “expansionist paradigm” and the alternative “steady-state
(ecological) alternative” (Rees 1995) is prominent in the sustainable development literature.
One notable attempt to resolve these opposing views has been by Herman Daly, who
separates various types of capital into categories such as natural, built, and human. Natural
capital, in particular, becomes the focus of attention in ecological economics. Natural capital
can be further divided into both renewable and non-renewable resources. That capital which
can be perpetually input to production without diminishment, i.e. renewable resources, is by
ecological economic definition sustainable. The focus of sustainable development according
to an ecological economic perspective becomes one of maintaining steady-state consumption
with respect to resource renewal, consuming only renewable resources, protecting non-
renewable resources, and therefore maintaining a constant stock of natural capital (Daly
1994). Yet, as Common and Perrings (1992 in Krishnan, Harris and Goodwin 2001, p.112)
note, “ecological economics of sustainability privileges the needs of the system over those of
the individuals.” Thus a focus upon environment and economy in sustainable development,
even through the lens of ecological economics, remains open to critique on social grounds.

The challenges with respect to the last E of sustainable development—social equity—are
perhaps greatest of all and its representation in the literature most diffuse, largely because of
the multiplicity of views on social justice and considerable debates that surround them. Barr
(1998) differentiates among three of views: libertarian, liberal, and collectivist. While
libertarians generally favor “a reward system that lacks public intervention in the
redistribution of private endowment” (Deka 2004, p.333), for example, liberals favor
redistribution in a manner that either maximizes total benefit to society (a utilitarian view of
31
liberalism), or that which provides greatest benefit to the least advantaged such that basic
liberties and opportunities are provided for all (a Rawlsian view of liberalism [Rawls 1971;
2001]). Collectivists essentially extend Rawlsianism to consider not just equality of
opportunity, but equality of outcome as well. Treatment of social justice concerns described
in the Brundtland Report definition appear to call for Rawlsian liberalism, namely to focus
attention upon the “essential needs of the world’s poor” (WCED 1987, p.43), but as
suggested previously, this definition is not the only one in use and many communities (and
members within them) likely embrace different perspectives on social justice.

Overcoming challenges of sustainable development
Given the challenges presented so far, actually achieving sustainable development may seem
impossible. Nevertheless, Toman remarks that sustainable development discussions often do
resolve economic and ecological positions according to a socially determined “safe minimum
standard” regarding “intergenerational fairness, resource constraints, and human impact” that
depend on “moral judgment about moral imperatives and the value of discounting” (Toman
1992 in Krishnan, Harris and Goodwin 2001, p.90). But these resolutions vary considerably
from place to place (Ryan and Throgmorton 2003) and time to time (Howarth and Norgaard
1994). Sustainable development, like development, is a process; outcomes are intermediate in
nature and depend upon the continuous action of people involved. As Campbell explains,
sustainable development “cannot be reached directly, but only approximately and indirectly,
through a sustained period of confronting and resolving the triangle’s conflicts” (Campbell
1996, p.296).
32
Communicative approaches have generally been promoted as promising for supporting the
ongoing deliberation that Toman and Campbell describe, especially to apply the moral
judgment to arrive at minimum standards. This promotion may not be surprising as the
“communicative turn” in planning, noted by Healey (1996), coincided with growing interest
in sustainable development. Campbell (1996) has suggested in particular that “consensus
building” and “mediated negotiation” are important communicative approaches available to
the planner that can be used in the pursuit of sustainable development objectives. These
approaches are essentially built from the theoretical work of Habermas (1973; 1984) and
initially applied to planning by Forester (1989). Essentially, “communicative action” is
theorized to resolve the “legitimation crisis” created when there is “colonization of the
lifeworld by the system” (Verma and Shin 2004, p.132). Examples of the “system” include
those based in traditional development thinking, for example those associated with the state-
initiated projects reviewed by Scott (1998). The transformation of society by these systems—
with social and environmental consequences—exemplifies the “colonization” Verma and
Shin describe. “Lifeworld” describes collectively the members of society and environment
thus transformed. The legitimation crisis describes an outcome in which the interests of
society’s members are not represented in the policies and programs of the system. As
explained by Verma and Shin (2004) communicative action provides a mechanism for
“reinstatement of the lifeworld” (p. 138) through participation in “speech communities”
whereby participants contribute freely in uncoerced discussion about shared concerns.
Importantly, this situation of uncoerced discussion, the “ideal speech situation” describes
conditions whereby intersubjective agreement can be reached among participants (Habermas
1973).
33
But communicative action theory is often criticized for its failure to consider the reality of
political power in practice (Flyvberg 1998). The conditions under which public deliberation
occurs tend not to resemble those of ideal speech such that agreement can occur among all
affected parties. Some members of society have stronger voices in decision making, often
tracing to personal relationships with politically powerful people or owing to their personal
wealth. Others do not have these same advantages. Communicative approaches are also
challenged in their ability to address regional environmental and intergenerational concerns
for sustainable development, which often involve issues that are not or cannot be known by
citizens involved in planning decisions. Among sustainable city researchers, Brugmann
(1997) focuses his concern specifically on the capacity of the lay public to develop
meaningful sustainability indicators programs,
15
notably for technical measures and
education purposes related to long-term sustainability. The concern is similar to that of
Dryzek (1995) who criticizes communicative action for its inability to accommodate a
meaningful role for the environment in speech communities. Brugman and Dryzek both
question whether some participants in public deliberations may be technically qualified to
participate, and whether they may be sincere with respect to environmental goals of
sustainable development. Indeed the empowerment of individuals in open, public discourse
regarding sustainable development entrusts them with great responsibility to act as stewards
of the environment and fellow citizens, and in the interest of generations of people whom
they have never met. To be “green cities, growing cities, just cities?” (Campbell 1996) may
simply be too much to ask. To the point, Pepper (1996, p.323) writes:

15
Sustainability indicators are variables that measure conditions relevant to an issue of interest, such as water
quality, traffic congestion, availability of affordable housing, and so forth. Indicator programs are common
elements of local sustainability initiatives, and are often developed through public participatory processes.
34
What if a community felt that it did not want to be green? What if people
decided, in their face-to-face democratic town meetings, that they wanted to
release plumes of sulphurous smoke into the atmosphere? Or that they wanted
to live in inegalitarian patriarchy?

Nevertheless, planning for sustainable development appears to be headed increasingly in the
direction of greater public involvement and trust in community partnerships to resolve the
various tensions in planning (Mazmanian and Kraft 2001). And despite the concerns for
communicative approaches, there is growing evidence that not applying them in an effort to
involve the public in decision making creates greater concern. Scott (1998) notes various
failures of schemes to improve the human condition where the public was not engaged. The
experience of transportation planning in the U.S. reinforces the notion that without
involvement of the public, plans can be ill-advised and in some cases may simply never come
to pass. Black (1990), for example, presents the Chicago Area Transportation Study of the
1950s as a mostly successful exercise conducted by transportation agency staff without
public input, progressing through all steps of the planning process…except implementation.
By contrast, communicative models of planning, particularly that of “planning through
consensus building” (Innes 1996), however imperfect are generally recognized in practice as
state-of-the-art. As Innes (1996, p.469) notes, “consensus building, properly designed, can
produce decisions that approximate the public interest.” But the question remains as to
whether sustainable development, pursued by means of communicative planning,
appropriately addresses the various economic, environmental, and social challenges of
development. Its various ideas and interpretations are likely to take many forms arising out of
the stakeholder deliberations in the growing number of cities, and observations related to
these may reflect varied success in achieving sustainable development goals.
35
Sustainable Cities and Theories About Them
Many cities in the U.S. experienced substantial economic growth following WWII, while
also witnessing social and environmental impacts. As described previously, some conditions
associated with development, such as changes in income inequality, are expected to be more
pronounced in urban areas. Although the Kuznets relationship has generally been explored at
national scale (an exception is Nielsen and Alderson [1997] who apply it to the models of
counties in the U.S.), attention to the processes within cities that may also reflect Kuznetsian
development is generally lacking. A hypothesis of this study is that the development
experience of cities may reveal the classic Kuznets relationship, and that policy reform in
cities consistent with sustainable development principles contributes to changes reflected in
the turning point of the Kuznets Curve. To provide some background, selected ideas and
experiences of U.S. city development during the postwar period are reviewed below.

An important commonality of the experience of U.S. cities over the past 50 years has been
that of “urban renewal”
16
and the effects of increased automobility that accompanied
construction of the interstate highway system. Jacobs (1961) was among the first to critique
large-scale urban renewal programs in the U.S., identifying the destruction of neighborhoods,
dispersion of residents, and development benefits that were unequally shared as a result of
these programs. Altshuler and Luberoff (2003) provide many examples of the social
upheavals they triggered. Related to urban renewal, increased automobility also emerged as a

16
Urban renewal describes the process by which public agencies reclaim private land for public projects,
notably in the 1950s and 1960s in the U.S., and the large-scale demolition of established neighborhoods and
businesses to make room for freeways during that time.
36
means by which people were separated and cultural differences were reinforced. According
to Schaeffer and Sclar (1975, p.119),

With massive auto transportation, people have found a way to isolate
themselves…They have stratified the urban landscape like a checkerboard,
here a piece for the young married, there one for health care, here one for
shopping, there one for the swinging jet set, here one for industry, there one
for the aged.

Since the 1950s, “urban sprawl” has generally characterized city development in the U.S.,
resulting from the greater mobility and freedom provided by the automobile, especially away
from the inner city neighborhoods divided by newly constructed highways. Importantly, the
greater freedom to travel has been enjoyed only by those who can afford an automobile, with
an effect of creating demand for suburban residential development by more affluent members
of society, thus leaving older, poorer central city neighborhoods, their residents and their
limited resources behind. With this new urban form and social and economic structure of
postwar cities created in part by highways designed to support an increasing numbers of
automobiles,

many planners, scholars, policymakers and interest groups have decried the
greater income inequality, crime, and health risks faced by those who remain in inner cities
of America.
17
Research on “environmental justice”—the exposure of disadvantaged groups
to hazards, based upon their location in their living environment—is active and growing
(UCC 1987, Cutter 1995, Rhodes 2003, Pulido 2004) as is a growing body of literature on
health effects and urban form (McCann and Ewing 2003, Sturm and Cohen 2004). Too, the

17
Note that social problems cannot be entirely attributed to sprawling urban form or investment in highway
projects. The period of large-scale highway development in the U.S. also coincided with desegregation (which
has been associated with “white flight” [Armor 1986]) for example. Contributing factors may also include local
practices of exclusionary zoning, a federal tax system and lending practices that encourage home ownership,
availability of land and private developers seeking to meet demand of a growing population. All of these likely
contribute to and result from postwar transformation of urban form in the U.S.
37
environmental impacts more generally—particularly air pollution, habitat destruction, and
surface water runoff—are increasingly well-documented (Ewing 1994, Forman 2003, Ewing
et al 2005). The critique of urban development in the U.S., especially as related to the
highway program and urban sprawl, as well as the conditions and economic incentives
behind this development, mirror well the images of postwar global economic development
described earlier. And just as the nations of the world have responded to the call for an
alternative development theory—sustainable development—so too have various
communities, interest groups, and policymakers in the U.S.

The experience of cities in particular has provided rich material for studying sustainable
development. Sustainable cities, or cities which have embraced principles of sustainable
development in their planning, have become an important focus of research. Some of this
research has sought to frame the identity and experiences of sustainable cities. Haughton
(1999) for example has categorized sustainable cities into four types according to their
purpose: “externally dependent cities”, “self reliant cities,” “redesigning cities” and “fair
share cities.” Self reliant cities seek to minimize consumption, waste streams and trade with
those outside their boundaries, while externally dependent cities do largely the opposite.
Redesigning cities embrace largely principles of urban design to make cities attractive places
to live and work. Fair share cities seek balance among the competing objectives of
sustainability, and contribute to regional solutions for social, economic and environmental
concerns. Other sustainable cities researchers have interpreted urban design movements in
light of sustainable development (Garde 2004) or identified sustainable urban forms
(Jabareen 2006). Still others have focused their attention more closely on the content of city
38
policies and planning documents (Berke and Conroy 2000; Conroy and Berke 2004, Jepson
2004), or developed performance indexes that use the city as unit of analysis (Portney 2003;
Lubell, Feiock and Handy 2009).

A general observation about some of the more recent sustainable cities research is that many
researchers have turned to urban theory for further inspiration. The turn may suggest a
realization that, by selecting cities as objects of investigation, we can leverage what we know
about cities to advance our understanding of sustainable development in them. Indeed a long
history of theorizing about cities more generally exists that can be considered in research on
sustainable development in cities.

Jepson (2007) reviewed sustainable development in light of the “Childe thesis” (Childe 1936;
1950), a theory of city formation. The Childe thesis attempts to explain the “neolithic
revolution” describing prehistoric transformation of hunter-gatherer societies to agriculture-
based settlements. Accordingly, areas that could support agriculture became settled and
populations grew. These settlements were characterized by specialization and division of
labor, which led to agricultural surplus. The surplus led to organization by a ruling class, and
the emergence of writing, art, science, trade and government. The idea has been extended to
an “urban revolution” in which larger, more complex cities evolved, not altogether unlike
Rostow’s stages model of economic development. The Childe thesis names four
interdependent factors—population, organization, environment, and technology—that form
the basis of community and its development as a city. Jepson (2007) considers the “POET”
framework to explain adoption of sustainable development policies, citing in particular a
39
collective response to environmental pressures, capacity for community organization and
education of its population.

Other urban theories are also available for consideration. For example, a view of city
formation that opposes Childe’s is one in which cities preceded agriculture, and cities instead
emerged as centers of trade. That is, agriculture (or trade for food) grew as a result of the
need to support a city’s growing population. This is the “trade thesis” of Jacobs (1969), an
idea that appears to generate Haughton’s “externally dependent city.” A view of the city that
is less materially oriented is that of Mumford (1938; 1961) who promotes the spiritual basis
of cities that attracted its first settlers and continues to allow cities to flourish and grow.
Mumford (1938 in Kostof 1991, p.37) writes that a city is a “point of maximum
concentration for the power and culture of a community”. This point of view seems
consistent with the notion of a city as comprising ideals capable of cohering individuals in a
movement, such as the sustainable development movement, to pursue common goals.
Haughton might describe this as a motivation that expresses “redesigning” or perhaps “fair
shares” sustainable city types. Whatever their differences, these urban theories share in
common their expression of complex and dynamic internal structures that reflect purposeful
associations and actions of people, and where people are involved in an ongoing process that
shapes the cities in which they live.

Also inherent in the urban theories presented above is the idea of growth. Cities add
population over time as some function of economic transformation, material production or
trade, and cultural identity that attracts new residents, or maintains them over time. Lubell,
40
Feiock and Handy (2009) highlight the importance of growth in our understanding of
sustainable development in cities by calling attention to Tiebout’s Hypothesis. Tiebout
(1956, p.418) suggests that “the consumer-voter may be viewed as picking that community
which best satisfies his preference pattern for public goods.” Heikkila (1996) and also
Banerjee and Verma (2004) use Census data to demonstrate the existence of various
“Tieboutian clubs” in Los Angeles County that differentiate themselves according to
population characteristics and land uses. The result is a metropolitan landscape similar to that
described previously by Schaeffer and Sclar (1975). Also, in their study of cities in the
Central Valley of California, Lubell, Feiock and Handy (2009) observe that size and growth
correlate with adoption of sustainability policies by cities, and that higher socioeconomic
status and higher levels of intellectual capital
18
among city residents explain this observation.
Their results suggest that sustainable cities are indeed Tieboutian Clubs of particular types
and, further, that income, educational attainment, and employment skill level may be
important factors that explain their existence. Tiebout sorting begins to explain how
individuals aggregate in urban space on the basis of common interests and preferences, and
as such provide some initial clues about human agency in cities, or at least among them. But
Tiebout’s Hypothesis stops short of explaining how agents interact to create policy and
development outcomes within them. Stone’s “regime theory” of holds promise in this regard.

Regime theory recognizes a divide within the American political economy between market
and state, in that public and legal powers along with overall citizen well-being are the
responsibility of the state, while ownership and control of productive assets are largely

18
Lubell, Feiock and Handy (2009) defined intellectual capital as proportion of businesses that are professional,
scientific, educational, managerial, and health and social services according to the U.S. Census Bureau.
41
privately held (Ferman 1996). Effective governance requires bridging this divide, and this is
the purpose of regimes. A regime describes the “informal arrangements by which public
bodies and private interests function together in order to be able to make and carry out
governing decisions” (Stone 1989, p.6). Regime theory proposes that various social groups—
urban regimes—engage to shape the politics of cities. Stone (1989) identifies four types of
regimes: (1) the “maintenance regime” which focuses on routine service delivery of city
government, (2) the “development regime” which seeks to align with city leaders for the
purpose of economic growth and development, (3) the “middle class progressive regime”
which concerns itself with a variety of environmental and social issues, and (4) the “lower
class opportunity expansion regime” which seeks primarily to overcome social inequities
faced by poorer members of society. As Portney (2003) notes, the ease by which the various
goals of these regimes can be achieved decreases in the order the regimes are listed above.
That is, the type and amount of resources required to overcome problems faced by the lower
class regime is hardest to achieve. By contrast, interests in economic growth are most easily
accommodated and incorporated into the routine activities of the maintenance regime. This
relative “ease” of achieving regime goals, has a political explanation, in that the success of
regimes depends on those who “do the best job of delivering political resources to local
elected officials being most likely to see their preferred policies adopted” (Lubell, Feiock and
Handy 2009, p.297). This last observation about regimes leads to the theory of cities as
“growth machines” Molotch (1976), and this theory is reflected in more recent sustainable
cities research (Portney 2003; Lubell, Feiock and Handy 2009).

42
Growth machine theory presumes that cities are subject to rising levels of population and
financial activity and to Molotch, growth is the defining feature and common interest of all
cities, the level of which determines a city’s success. According to Molotch (1976, p.310),

The desire for growth provides the key operative motivation toward consensus
for members of politically mobilized elites, however split they might be on
other issues, and that a common interest in growth is the overriding
commonality among important people in a given locale.

Growth machines are explained by strong political alliances between local government
officials and a land-based elite that seeks to accumulate wealth through development. Within
sustainable cities, however, there is the possibility that a “middle class progressive regime”
will hold narrow development interests at bay. The middle class progressive regime, as
Portney (2003, p.27) explains, focuses on issues of relevance to sustainable development,
especially “historic preservation, affordable housing, urban design, affirmative action, and
linkage of funds for social purposes.” The overlap of interests, namely affirmative action and
affordable housing between middle and lower class regimes further suggests the possibility
that political resources can be pooled among these regimes to influence governance of cities.

Absent in the study by Lubell, Feiock and Handy (2009) and not explicitly considered by
regime theory is the possibility that neighborhood based groups behave as regimes (or that
regimes are manifest in neighborhoods) to promote (or resist) growth. Davidoff (1965 in
Campbell and Fainstein 1996, p.217)) was among the first to recognize the important role
neighborhood associations play “to combat a renewal plan, a zoning change, or the proposed
location of a public facility,” continuing to suggest that “such organizations may seek to
develop alternative plans which would, if effected, better serve their interests.” O’Reilly and
43
Webster (1998 in Gimpel and Schuknecht 2004, p.2) report that geographic locations are
capable of creating for people “political spaces in which they are allied with like-minded
citizens to fight political battles.” In this way, neighborhood-based regimes have the potential
to promote (or oppose) the designs of cities, and thus the action of regimes emerges as a
potentially important mechanism related to city development.

Structure-Agency Dualism in Cities
Issues of structure and agency have characterized a longstanding debate among social
scientists, and the interplay of structure and agency is a well-established theme of
sociological research, as advanced by Giddens in his theory of “stucturation” (1976, 1984),
and also Bourdieu (1977, 1984, 1990), among others. In the context of planning and policy,
Emirbayer and Mische (1998) describe agency as a temporally embedded social engagement,
informed by the past, yet oriented toward the future, while also being constrained by the
contingencies of the present.

An important inspiration for this study is structuration theory, an idea that addresses the basic
concern for whether structural relationships or the individual should be accorded primacy in
explaining the development of society. Previous approaches in social analysis focused
separately on macrological concerns for the conditions that determine social outcomes, or
micrological concerns for subjective interpretations of society. Previous sustainable cities
research appears to have followed these same conventions. (On the one hand, a focus upon
planning documents, urban form, and so forth reflect interest in the structural conditions that
are believed to determine social outcomes. On the other hand, recent sustainable cities
44
research reflects growing interest in urban political economy and strategic conduct of
agents). In cities, agency describes people acting to address a variety of concerns that face
their communities. Their actions further reflect beliefs and aspirations for their cities to better
serve their interests. At the same time, these actions must also reflect the influence of
existing rules and resources (structures) that constrain or enable human action. And these
actions are ultimately reflected in new rules that constrain or enable further actions. City
development, thus, reflects an ongoing interplay between structure and agency whereby
participants seek to perpetuate or change existing plans, but by doing so also create new
structures that constrain or enable further agency. Together, structure and agency explain a
dynamic, ever developing city. The two are inextricably linked and should not be considered
apart from one another.

Structuration theory has been increasingly applied to urban analysis and theories of relevance
to the concerns of cities. Dear and Moos (1986; Moos and Dear 1986) investigated the utility
of structuration in an analysis of the built environment, for example, focusing in particular on
the ghettoization of ex-psychiatric patients. Bernard et al (2007) applied the theory to
conceptualize urban neighborhoods and health disparities in urban neighborhoods. Kidder
(2009) applied structuration theory to explain how bike messengers interact in an urban,
social environment. While the idea of structuration theory appears to be highly relevant to the
experience of cities, it is less apparent in the sustainable cities literature. Most relevant to
thinking about structure-agency dualism, urban regimes and the alternative approach to
development that sustainable development offers, may be the work of Savitch and Kantor
45
(2002, p.53) in their review of the development experience of ten world cities in Europe and
North America, writing that

While regimes work within contextual constraints, they can change those
contexts. Agency can work upon structure, enabling a city to move into
another context and change its development strategy.

As Harris and Goodwin (2001) explain, sustainable development emerged first as a popular
reaction to economic development that too narrowly distributed benefits of development and
created negative impacts on the environment and existing social structures. Pioneers of
sustainable development sought to change these policies. Sustainable development advocates
in cities continue to take this approach, seeking to confront existing plans to create new ones,
thus moving these cities into new contexts built upon development strategies which moderate
economic development with respect to broader social and environmental concerns.

Human agency and policy change in cities
That urban regimes are capable of policy change makes two assumptions, first that individual
actors aggregate to form such groups, and second that these groups are responsible for
change. A notable representation of these ideas has been in the “Advocacy Coalition
Framework” or ACF (Sabatier and Jenkins-Smith 1989, 1993). The ACF is a theoretical
framework built upon the idea that individuals have affinity for one another and form
advocacy coalitions. Advocacy coalitions are separable from one another according to
differences among core beliefs and policy interests. That policy debates arise out of these
differences, but can be resolved by a number of mechanisms, are predicted by the ACF.
Specifically, the ACF predicts that the policy process can advance when there is agreement
46
among coalitions that conditions are unacceptable, thus creating a “hurting stalemate.” Also,
policy can advance through the influence of external forces, such as new regulatory
requirements or changes in economic conditions, introduction of information that offers new
insight and options, and by the actions of a “policy broker” who serves to mediate differences
among coalitions. The ACF and its various hypotheses fall into a class of ideas termed
“rational choice institutionalism”—one of several “new institutionalism” theories described
previously by Hall and Taylor (1996), which theorizes that the behavior of individuals is
strongly influenced by the institutions in which they occur—institutions being broadly
defined as systems of rules, customs or other mechanisms that govern behavior. Another
notable theoretical framework of this class is the Institutional Analysis and Development
Framework (IADF), first proposed by Kiser and Ostrom (1982). The centerpiece of the IADF
is the “action arena” in which individuals interact. During these interactions, individuals are
influenced by others, along with various rules and resources of their physical world, and
individuals behave strategically given these influences to produce various outcomes. The
outcomes feed back to influence further actions. Both the ACF and the IADF have been used
to frame thinking about the conditions and changes observed in many social and
environmental systems (Sabatier 2007, Ostrom and Moran 2005). Neither of them have been
applied to sustainable cities research.

In another framework resembling in many ways the ACF and IADF, Savitch and Kantor
(2002) consider the circumstances that structure the development of cities, who determines
the course of development, and the interaction of these. They introduce the idea of a
“bargaining context” in which policy action related to development occurs. This context
47
describes essentially the market conditions and governance structures of cities, and the
implications these have for cities in negotiation with private investors. The varying contexts
of cities result in varied development outcomes, with some cities adopting free-market
approaches to address issues such as employment, growth and environment. In others, cities
adopt strict governmental regulation to address these same issues. Savitch and Kantor further
introduce ideas of “driving” and “steering” variables that enable cities to pursue different
development strategies. Driving variables describe such things as market conditions and
intergovernmental support—forces that are external to local urban systems. Steering
variables are embedded within urban systems and include “popular control” and “local
culture.” Where popular control is decentralized (e.g., governance responsibility shared by
neighborhood or civic associations) and local culture reflects great interest in public
amenities, environmental protection, historic protection, etc., development outcomes are
described as “social-centered.” Where popular control is centralized and characterized as
having weak civic associations, and local culture reflects little or no interest in taxing
business, restricting land development, and so forth, development outcomes are described as
“market-centered.”

Social and economic structures and the Kuznets Curve
Savitch and Kantor highlight the importance of regimes to work within contextual
constraints of cities, often to change these contexts, and in this way to achieve more
or less social- or market-centered development outcomes. These outcomes, argued to
reflect the influence of regime actions are reflected in the structures of cities. Savitch
and Kantor (2002, p.167) point out that “cities take into account their structural
48
context and pursue policy strategies that are roughly consistent with their resource
capabilities.” Exploring the policy pursuits of sustainable cities, thus requires some
attention to the resources and social and economic context of cities. As introduced in
in the previous section, an approach for specifying social and economic structures of
cities may be found from an adaptation of the Kuznets Curve.

The relationship between income inequality and economic growth in the Kuznets
Curve has been represented in a number of ways in the literature. Economic
development is most commonly measured as per capita income and shown on the x-
axis, while the income inequality, shown on the y-axis, has taken many forms.
Ahluwalia (1976a,b) considers the logarithm of per capita income for different class
shares of the income distribution. Robinson (1976) considers the variance of the
logarithm of income. Others, including Neilsen and Alderson (1997) measure income
inequality according to Gini coefficient.
19
Anand and Kanbur (1993) consider six
different measures, including those mentioned above, and others. Empirical tests of
the Kuznets relationship has been the focus of much research since its initial
presentation, especially to verify its promise for explaining patterns of development.
Demonstration, and critique, of other Kuznets Curves, especially the Environmental
Kuznets Curve (EKC), has become an active pursuit of researchers (Grossman and
Krueger 1991; Yandle et al 2004; Winslow 2007). Overall the Kuznets relationship
has held up in some cases, not so in others, but remains an inspiration for continued

19
The Gini coefficient is a statistical measure of dispersion, commonly used to measure inequality of wealth.

49
empirical and theoretical work, and a promising lens through which structure and
agency in sustainable cities can be imagined.

As discussed previously, Kuznets suggests that a variety of social, economic and
political factors explain the relationship. This study focuses upon the political
economy explanations, including that of Acemoglu and Robinson (2002, p.183) who
suggest that increasing income inequality induces political instability such that lower
classes are ultimately able to “force democratization on political elites.” But along
with the decreased income inequality that ultimately reflects the “wishful thinking” of
Kuznets, Acemoglu and Robinson explain that other outcomes are possible. This
research considers that various non-democratic paths can be observed in the
experience of cities, to include for example one which leads to “autocratic disaster”
characterized by continued high inequality with economic growth. In this way, a
valuable contribution by Acemoglu and Robinson is to clarify the range of
possibilities for development that the Kuznets Curve elucidates, to consider more
deeply their underlying mechanisms, and also to encourage us to ponder which
outcomes may be preferred. The implication here is that the value of the Kuznets
Curve to our understanding of development may be less in predicting what is, but
instead in framing our thinking on what should and should not be, and to further
direct attention on how to achieve one or the other. As will be described more fully in
the next chapter, the Kuznets Curve informs a potentially useful typology of cities
that can frame further research on sustainable cities, especially to consider the actions
of people within them to achieve sustainable development goals.
50
CHAPTER 2: METHODS AND RESULTS

This chapter presents both methods and results of this study. Overall, investigation proceeded
in several steps to test the hypotheses listed in the Introduction. First, samples of U.S. cities
were identified for study, including those with well recognized commitments to sustainable
development. Indicators were developed based on available data from the U.S. Census
Bureau. Statistical models of income inequality and economic growth were explored in a
search for the classic Kuznets relationship among cross-sectional data for U.S. cities.
Statistical relationships among variables, including a variable for the existence of a
sustainability initiative, were summarized and inspected for preliminary clues about social
and economic structures in developing cities. A typology of cities was developed according
to cities’ positions along a modeled Kuznets Curve for cities. Well-recognized sustainable
cities in the U.S. were classified according to this typology. In a sub-sample of these cities—
those which have achieved relatively high levels of economic development—urban regimes
in these cities were investigated further. Influences upon income inequality by various
regime associations in these cities, and also by business and political leaders, and the general
public involved in sustainability planning were statistically tested. Results are described
briefly in this chapter and discussed further in the next.

Sample Selection
As discussed in Chapter 1, past sustainable cities research has focused largely on cities
known to have committed to policies of sustainable development. But among the 5,340 U.S.
cities reported by the U.S. Census Bureau in 2000, only a very small number of them are
51
known to have actually committed to such policies. Indeed much research on sustainable
cities to date has generally involved in-depth case studies of a few pioneering cities, such as
Santa Monica, California (Staley 2006), or supported theory development using a small
number of purposefully selected cities (Ryan and Throgmorton 2003), or slightly larger
comparative studies based on convenience samples
20
(Lubell, Feiock and Handy 2009). A
review of popular student readers on sustainable development (Wheeler and Beatley 2004;
Satterthwaite 1999) reveals a short list of some of the most commonly recognized sustainable
cities, typically those with sustainability initiatives that emerged in the mid-1990s.

To be able to generalize from the experiences of these cities to others and to make
recommendations for policy is made difficult given the limitations of these studies. A
challenge of sustainable cities research remains in identifying a suitable sample of cities to
consider. To date, researchers have often turned to various lists of sustainability planning
case studies, award-winning plans, and so forth, for their samples. One such list is derived
from case studies of city planning efforts related to sustainable development published since
2002 by the U.S. Environmental Protection Agency (EPA) Smart Growth Office. Another
such list, also from the EPA, is based upon the dozens of “National Award for Smart Growth
Achievement” awards given to tribal, state, local and regional governments that exemplify
principles of Smart Growth. The American Planning Association (APA) also awards the
preparers of high-quality plans each year, many of which are consistent with selected
principles of sustainable development. Some of the larger samples used in sustainable cities
research are based on these lists from the EPA and APA. Berke and Conroy (2000) and

20
A “convenience sample” is a sample of cases this is not randomly drawn, but instead determined by practical
considerations such as cost, time, and availability of data.
52
Conroy and Berke (2004) for example, used these sources to develop a sample of 42 U.S.
cities, 25 of which incorporated sustainable development as a central organizing principle in
their community plans (the remaining 17 were used as a control group in their study). A
slightly larger number of cities in the U.S., 31 in all, that are recognized for having actually
adopted sustainability initiatives was reported by Portney (2003), with most of these cities
also being considered by Conroy and Berke. The cities on Portney’s “reasonably
comprehensive list” also include those represented in two popular student readers on
sustainable cities (Wheeler and Beatley 2004; Satterthwaite 1999). Table 2 presents
Portney’s list of cities. Besides their relatively small number (representing less than 1 percent
of all U.S. cities), the studies based on these present important challenges related to their
selection into the samples which may affect researchers’ abilities to make valid inferences
from results. There is, for example, the possibility that by limiting his sample to the relatively
few cities in the U.S. that are furthest along in their planning for sustainable development,
studies based on Portney’s list ignore important experiences of cities that are just starting
down the path. Also, the cities identified by Portney are generally among the larger cities in
the U.S. and thus their experiences may not represent smaller, growing cities that may face
different types of issues in their development. There is also the possibility that the sample of
cities used by Berke and Conroy (2000) and Conroy and Berke (2004) may not represent
cities that do not meet APA or EPA performance standards, but may be important to our
understanding of sustainable cities nonetheless. Finally, the study by Lubell, Feiock and
Handy (2009) considers only those experiences of cities in California’s Central Valley, a
region which may not be representative of others across the U.S.

53
Table 2: U.S. Cities with Sustainability Initiatives
City Name of Sustainability Initiative
Chattanooga, TN
Jacksonville, FL

Orlando, FL
Tampa, FL

Seattle, WA
Olympia, WA
Portland, OR
Milwaukee, WI
Santa Monica, CA
San Francisco, CA
San Jose, CA
Santa Barbara, CA
Austin, TX

Indianapolis, IN
Boulder, CO
Cambridge, MA
Boston, MA
Brookline, MA
Scottsdale, AZ
Tucson, AZ
Phoenix, AZ
Brownsville, TX
Cleveland, OH
Lansing, MI
Ithaca, NY
Burlington, VT
New Haven, CT
Annapolis, MD
Oklahoma City, OK
Granstville, UT
Stuart, FL
Sustainable Chattanooga
Jacksonville Indicators Project, Jacksonville Community
Council
Sustainable Communities
The Tampa/Hillsborough Sustainable Communities
Demonstration Project
Sustainable Seattle/The Comprehensive Plan
Sustainable City Indicators/Sustainable Community Roundtable
The Comprehensive Plan
Campaign for Sustainable Milwaukee
Santa Monica Sustainable City Program
The Sustainability Plan
Sustainable City Programs (Sustainable City Major Strategy)
The South Coast Community Indicators Project
Sustainable Communities Initiative and Sustainability
Indicators Project of Hays, Travis and Williamson Counties
IndyEcology
The Sustainability Program
Sustainable Cambridge, Cambridge Civic Forum
Sustainable Boston Initiative
Comprehensive Plan
Scottsdale Seeks Sustainability
The Livable Tucson Vision Program
Comprehensive Plan, Environmental Element
Eco-Industrial Park
Sustainable Cleveland Partnership, EcoCity Cleveland
Sustainable Lansing
EcoVillage at Ithaca
No apparent name
Vision for a Greater New Haven
Alliance for Sustainable Community
Possibilities: Neighbors in Action
Grantsville General Plan for Sustainable Community
Sustainable Community
Portney (2003)

For these reasons, conclusions drawn from evidence using all of these samples may be
especially prone to committing a Type I error,
21
inferring from a group of cities erroneously
believed to represent the broader sample of cities, when in fact they do not.

21
Neyman and Pearson (1928) first recognized two types of errors analysts commit when attempting to make
inferences from data, so-called “Type I” and “Type II” errors. Formally, in statistical hypothesis testing, Type I
errors describe rejecting the null hypothesis when the null hypothesis is true. The opposite, accepting the null
hypothesis when it is in fact false, is a Type II error.
54
There is also the possibility of committing a Type II error, or failure to make an inference
from evidence of sustainable development when in fact such evidence exists, given
especially that among the more than five thousand U.S. cities that are not known to have
committed to sustainable development since publication of the latest “reasonably
comprehensive list,” there may be some that have done so. The city of Ventura, California is
one such case. According to their city’s website (City of Ventura 2010):

In 2007, Ventura's City Council adopted the Green Initiative. The Green
Initiative is the first of many steps toward the City's goal of becoming a
national model for sustainability. Since its adoption, the City has completed a
Green House Gas Inventory which was certified and registered by the
California Climate Action Registry, as well as joined the Sierra Club Cool
Cities Program and ICLEI Local Governments for Sustainability.

The commitment to sustainable development by cities is likely much larger than what
samples in sustainable cities research reflect. Recall from the Introduction that ICLEI
membership includes hundreds of U.S. cities that have demonstrated interest in sustainable
development, and this list is growing by 10 percent annually (ICLEI 2010). The number of
cities with adopted sustainability initiatives is likely to increase in coming years as well,
potentially to decrease chances of committing both Type I and Type II errors in sustainable
cities research over time. Note however that research based on samples drawn from an
ICLEI list of cities too may be prone to selection bias and error, especially where ICLEI
membership may have been pursued by cities merely as a token gesture of commitment to
sustainable development, not indicative of intent to adopt a sustainability initiative in the
future.

55
An important premise of this research is that sustainable cities are, first, cities and that our
understanding of sustainable cities can be informed by deeper consideration on this point.
Sustainable development in cities should be considered within the context of city
development. Another premise is that city development can be informed by research on
development more generally. To explore the possibility that Kuznetsian development
provides insight generally into city development, and specifically into sustainable city
development, this study considers two samples of U.S. cities. The first sample considers all
cities in the U.S. (5,340 in the year 2000) for evidence of the Kuznets Curve more generally.
The second sample considers the “reasonably comprehensive list” compiled by Portney
(2003) for further insight into these cities. Portney’s list remains the largest list of cities in
the academic literature to date that are recognized to have committed to sustainable
development.

Development of Indicators
Several sets of indicator variables are developed for this study. The first set is intended to
support the search for the classic Kuznets relationship between income inequality and
economic growth in cities. As discussed in Chapter 1, Kuznets (1955) originally considered
per capita Gross National Product (GNP) for 5 ordinal groups (quintiles) over time for three
developed countries. In a cross section of 62 countries, Ahluwalia (1976a,b) demonstrated a
statistically significant relationship between the income shares of three of these groups—the
lowest 40 percent, lowest 60 percent, upper 20 percent—and the quadratic of the logarithm of
per capita income. Using county-level data for the U.S., Neilsen and Alderson (1997) also
demonstrated the classic Kuznets relationship between inequality and economic growth using
56
the Gini coefficient
22
and the logarithm of median family income. In each of these studies,
ordinary least squares (OLS) regression methods were employed. Also in each, several
additional variables were shown to influence models of Kuznetsian development. One
important observation is that some variables differ in effect between Ahluwalia’s studies and
that of Neilsen and Alderson, indicating that while the overall Kuznets relationship between
inequality and economic growth generally holds, other variable relationships may change as
the geographic scale is varied. The implication is that while ideas and methods employed in
published research on the Kuznets Curve provide general guidance, extending these ideas
and methods to a study of U.S. cities may require some careful adaptation.

Indicators of economic growth and income inequality
This study uses the data from the 2000 census, the most recent census in the U.S. for which
data are available. The measure of economic growth for this study is average household
income in a cross-section of U.S. cities, a measure that is reasonably consistent with research
described above, and also conveniently derived from data available from the U.S. Census
Bureau. Also consistent with previous research, the logarithm of average household income
for U.S. cities is considered in both first-order and second-order polynomial form. Ahluwalia
suggests that the logarithmic transformation gives “equal weight to equal proportional
differences in GNP in measuring ‘level of development’ [which has] intuitive appeal since
growth occurs at a compound rate over time” (Ahluwalia 1976a, p.2). Quadratic form is
expected to better fit the expected curvilinear relationship between income and inequality.

22
The Gini coefficient is a measure of statistical dispersion and is usually defined mathematically based on the
ratio of the area that lies between the line of equality and the Lorenz curve over the total area under the line of
equality. Values can range from 0 to 1, where 0 corresponds to complete equality and 1 corresponds to complete
inequality.
57

Rather than limit the dependent measure to income shares of different groups, as by
Ahluwalia, or to indices such as the Gini coefficient, as by Neilsen and Alderson, this study
develops models using measures of both types, with some further modification of each to
consider the unique contexts of cities and available data. In this study, four different
dependent variables are employed, three to characterize development by income classes in
cities—poor, middle class, and affluent—and the fourth to characterize the divide between
poor and affluent. The first three dependent variables are shares of population accounted for
by income class groups, adapted from Ahluwalia. The fourth is an affluence-poverty
dissimilarity index. The poverty-affluence dissimilarity index is a measure of evenness of
which two income groups (namely, poor and affluent) are distributed across an area. Like the
Gini coefficient, it is commonly used to measure inequality of wealth. But, according to
Massey and Fischer (2003) dissimilarity indices have become standard in research on
geographic segregation, which is an important characteristic of cities in the U.S., given the
separation of land use and differentiation among communities and neighborhoods within
them.

In this research, poor is defined according to the Federal Poverty Level (FPL)
23
and
affluence describes household incomes exceeding $100,000 in 2000. Middle class
describes remaining households that are classified neither as poor nor affluent. Note

23
FPL is one measure of poverty and used as basis for eligibility for government assistance programs, as
published in guidelines by the Department of Health and Human Services (DHHS). In 2000, FPL for a family of
four in the 48 contiguous U.S. was an annual income of $17,050, below which the family is said to be living in
poverty.

58
that the three dependent measures for income classes differ slightly from those used
previously by Ahluwalia, but are judged nonetheless to provide similar opportunity to
observe the “impact of the development process over different ranges of the income
distribution.” (Ahluwalia 1976a, p.1)

Poverty-affluence dissimilarity is computed from tract-level data from the U.S. Census
Bureau
24
as follows:

½ Σ | a
i
/ A – p
i
/ P |
where
a
i
= the poor population of the i
th
tract
A = the total poor population of city
p
i
= the affluent population of the i
th
tract
P = the total affluent population of city

Like the Gini coefficient, the affluence-poverty dissimilarity index varies between 0 and 1,
decreasing when poor and affluent are more evenly distributed, and increasing when poor
and affluent are more segregated. In this study, affluence-poverty dissimilarity indices were
computed for cities in 1980 and in 2000—years that span a period of increasing interest in
sustainable development. In both years poverty was defined according to FPL at that time.
Affluent households in 1980 were made comparable to affluent households in 2000 by
adjusting household income according to the Consumer Price Index (CPI), and the value

24
According to the U.S. Census Bureau (1994, p.10-1), “census tracts are small, relatively permanent
geographic entities within counties (or the statistical equivalents of counties) delineated by a committee
of local data users. Generally, census tracts have between 2,500 and 8,000 residents and boundaries that follow
visible features. When first established, census tracts are to be as homogeneous as possible with respect
to population characteristics, economic status, and living conditions.”
59
matched to nearest cutoff values available for aggregated data from the U.S. Census Bureau.
Accordingly, affluent households in 1980 were those for which household income exceeded
$50,000. While the definitions for poverty and affluence are somewhat arbitrary, they are
judged to be sufficient for general comparisons among cities. Alternative approaches (e.g.,
grouping by percentile groups) would be similarly arbitrary.

Indicators of economic and social structure
Many additional measures were developed in this study, especially to be considered as
additional explanatory variables in models of Kuznetsian development. These include
indicators for age, race/ethnicity, family structure, education, language, employment,
housing, income and expenses, and mobility, as shown in Table 3. These indicators were
derived using tract-level data for the entire U.S. from the 2000 Census, and aggregating their
values to city level according to their overlap with municipal jurisdictions. That is, counts for
cities were based on sums of counts for census tracts that spatially intersected or were
enclosed by city boundaries in 2000. These indicators were developed using ArcGIS
software and Census data compiled by Geolytics in their Neighborhood Change Database
(Geolytics 2003). This database provides comparable data for prior years that are normalized
using 2000 Census geography (this normalization is necessary to support the temporal
illustrations presented in the next section.)

60
Table 3: Names and Descriptions of Community Indicators
Variable Description
Age (percent of total population)
POP517 5-17 years old
POP1864 18-64 years old
POP65P 65+ years old
Race/ethnicity (percent of total population)
NHWHITE Non-Hispanic White
NHBLACK Non-Hispanic Black
NHASIAN Non-Hispanic Asian
NHSOTH Non-Hispanic Other
HISPANIC Hispanic
Family Structure (percent of family households)
MCWKID Married couple with children
MCNKID Married couple without children
FHWKID Female household with children
FHNKID Female household without children
Education (percent of population > 25 years old)
NOHSD No high school diploma
HSD High school diploma only
PSE Post-secondary education
Language (percent of population > 18 years old)
ENGL Speak English at home
NOENGL Do not speak English "well" or "very well" at home
Employment (percent of civilian labor force > 16 years old)
UNEMP Unemployed
LOWSK Employed in "low skill" occupations
MEDSK Employed in "medium skill" occupations
HIGHSK Employed in "high skill" occupations
Housing (percent of housing units)
OWNOCC Owner-occupied
RNTOCC Renter-occupied
SFR Detached single family residence
NEWCON Constructed within last 10 years
GT1PR More than one person per room in renter-occupied housing
Income and Expense (percent of households)
HIN100P Household income $100,000+
WELFARE Households receiving public assistance
POVERTY Persons living below Federal Poverty Level
GT35RNT More than 35% household income goes to pay rent
Mobility (percent of civilian employed population > 16 working outside the home, except as noted)
PBTRAN Commute to and from work by public transit
AUTO Commute to and from work by automobile
CMT25 Commute less than 25 minutes one-way
CMT2544 Commute 25-44 minutes one-way
CMT45P Commute 45+ minutes one-way
FORCNTR Resided in foreign country 5 years ago (% pop > 5)
SMHSE Resided in same house 5 years ago (% pop > 5)

61
Table A1 in the Appendix presents specific data sources used to develop these variables.
Table A2 in the Appendix presents descriptive statistics for these variables. Given the wide-
ranging issues relevant to sustainable development in cities, clearly more variables could
have been imagined for this study. Indeed many more datasets are available. However a
challenge for the empiric work envisioned in this study is that many of these other sets
cannot match the same broad geographic extent (entire U.S.) and same fine scale (tract) that
Census data are capable of supporting.

Models of Kuznetsian Development
The first hypothesis in this study is that a relationship exists between income inequality and
economic growth in U.S. cities, namely an inverted U-shaped curve in which income
inequality first increases with economic growth then decreases, as shown previously in
Figure 1. Ahluwalia (1976a,b) explored this relationship separately for different class shares
of national income by ordinary least squares regression.
25
Ahluwalia justified the
decision to model the dependent variable as income share rather than a summary index of
income inequality—such as Gini coefficient or poverty-affluence dissimilarity index—by his
interest in focusing on impacts upon different income classes as well as concern for the
relative insensitivity of index measures, which may obscure statistically significant
relationships among variables. His models showed declining share of GNP followed by
rising share of GNP with increasing per capita GNP for the lowest income groups, while the
upper income group described the opposite pattern. He provided evidence of a nonlinear

25
Ordinary least squares regression is a method for estimating the unknown parameters in a linear model,
namely y = βx + ε. where βis the parameter to be estimated and ε is a constant error term of the model. This
method minimizes the sum of squared distances between the observed data, and the data predicted by the linear
approximation.
62
relationship between income shares by class and per capita GNP by demonstrating improved
model fit (as measured by coefficient of determination, R
2
) when entering per capita GNP in
quadratic form. Ahluwalia further reasoned that the observed relationship is explained not by
short-term economic growth but by social structural factors reinforced over the long-term
after demonstrating that overall growth in GNP over the previous 10 years failed a test of
statistical significance. This study seeks to replicate Ahluwalia’s demonstration of the
Kuznets Curve for developing countries, using data for U.S. cities instead. Ahluwalia also
included categorical (dummy) variables for socialist countries in his models. Models in this
study include a dummy variable for cities that have well-recognized sustainability initiatives,
and also for participation in governance in these cities.

Ahluwalian models of Kuznetsian development
To test the first hypothesis of this study, several tests of statistical significance were
employed, namely to explore the relationship between economic growth and income
inequality in cities, the contribution of additional social structural variables, and commitment
of cities to policies of sustainable development. Table 4 presents results of three initial
modeling exercises based on Ahluwalia’s approach, as described above. Dependent measures
are shares of total city population for three income class groups defined previously. In these
models, explanatory variables include average household income measures and a constant
term. As shown in Table 4, model fit (as measured by coefficient of determination, R
2
) does
indeed improve when entering per capita GNP in quadratic form, similar to what Ahluwalia
found.

63
Table 4: Regressions Explaining Class Shares of U.S. City Populations in 2000 by Level
of Development
Dependent Variable: Share of Population by Income Class
Poor Middle Class Affluent

Explanatory Variable
eq. 1a eq. 1b eq. 2a eq. 2b eq. 3a eq. 3b
1. Constant
2. log[Average household income in
2000]
3. log[Average household income in
2000]
2
2.03*
-0.41*

17.15*
-6.75*

0.67*

1.46*
-0.14*

-28.23*
12.31*

-1.31*

-2.49*
0.55*

12.07*
-5.56*

0.64*

R
2

F
N
0.47
4,745
5,340
0.58
3,617
5,340
0.06
309
5,340
0.44
2,063
5,340
0.83
26,634
5,340
0.93
34,006
5,340
* statistical significance at the 10 percent level for a two-tailed test

Equations 1b, 2b, and 3b in Table 4 are presented graphically in Figure 4. Note that the
model of the poor’s class share generally decreases with development while the model of
affluent’s class share generally increases—an intuitively reasonable result. The model of
middle class share rises then falls over the development course. This result—an inverted U,
and preliminary evidence of a Kuznets Curve for cities—is discussed in the next chapter.

Figure 4: Class Shares of U.S. City Populations in 2000 by Level of Development
0.0
0.2
0.4
0.6
0.8
1.0
4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6
log[average household income in 2000]
share of city population in 2000
eq. 3b
eq. 2b
eq. 1b

64
Table 5 presents results of models that include additional variables related to short-term
growth. Variable 4 is the growth rate of average household income over the 1990-2000
period for each city in the sample.
26
The effect on class share of population growth during
that same period (variable 5), as well as total population (variable 6) and population density
(variable 7), are also illustrated. By comparison to Ahluwalia, who demonstrated that short
term economic growth did not explain patterns of development with statistical significance in
a cross-national sample, growth in average household income between 1990 and 2000 is
statistically significant in explaining increased shares for poor and affluent for U.S. cities in
2000.

Table 5: Regressions Explaining Class Shares of City Populations in 2000 by Level of
Development and Considering Short-Term Growth
Dependent Variable: Share of Population by Income Class
Poor Middle Class Affluent

Explanatory Variable
eq. 1b eq. 1c eq. 2b eq. 2c eq. 3b eq. 3c
1. Constant
2. log[Average household income in
2000]
3. log[Average household income in
2000]
2
4. Growth rate of average household
income (1990 to 2000)
5. Growth rate of population (1990 to
2000)
6. log[Population in 2000]
7. log[Population density as persons/
mile
2
in 2000]
17.15*
-6.75*

0.67*

19.94*
-7.95*

0.79*

0.01*

0.01*

0.04*
-0.001
-28.23*
12.31*

-1.31*

-31.43*
13.69*

-1.44*

-0.004

-0.01*

-0.05*
0.002*
12.07*
-5.56*

0.64*

12.49*
-5.74*

0.66*

-0.002*

0.01*

0.01*
-0.001*
R
2

F
N
0.58
3,617
5,340
0.64
1,554
5,340
0.44
2,063
5,340
0.51
919
5,340
0.93
34,006
5,340
0.93
11,567
5,340
* statistical significance at the 10 percent level for a two-tailed t-test

Overall growth in population between 1990 and 2000 also makes a statistically significant
contribution to this model. But the influence of these variables on income class shares of

26
The growth rate is computed as (X
2000
– X
1990
) / X
1990
x 100% where X is the parameter, e.g. average
household income, for which the growth rate is estimated for the 1990-2000 period.
65
population is small. Population density of cities (variable 7) appears not to create important
and/or statistically significant effect on population shares, but the effect of overall city size
(variable 6) is significant and at least as influential as population growth in this model. That
is, the share of both poor and affluent (especially the poor) appear to increase with city size
while the share of middle class decreases, and these relationships are statistically significant.
These observations are discussed further in Chapter 3. Table 6 presents results for models
that include several additional variables for various demographic, socioeconomic, and
housing and household characteristics compiled by the U.S. Census Bureau in 2000,
aggregated for each U.S. city. New variables are numbered continuously to aid comparison
among models, and equations 1b, 2b and 3b are also carried forward for comparison. The
units of additional variables (8-42) are logarithms of counts within cities. For example,
variable 8 is the logarithm of the number of people aged 5-17 in cities. Support for this
transformation of these variables is by the same logic applied to that for income measures,
namely that “growth occurs at a compound rate over time” (Ahluwalia 1976a, p.2) This same
observation that is reflected in studies of population growth since Malthus (1798). Note the
improved model fit for models specified by equations 1d, 2d, and 3d, with R
2
values
increasing from between 0.44 and 0.93 to between 0.82 and 0.95. Improved model fit
suggests that the added measures explain substantial remaining variation in the relationship
between average household income in cities and income class shares. Some observations
about these parameters are noted below, and discussed further in Chapter 3.

66
Table 6: Regressions Explaining Class Shares of City Populations in 2000 by Level of
Development and Considering Community Indicators
Dependent Variable: Share of Population by Income Class
Poor Middle Class Affluent

Variable
Group

Explanatory
Variable
eq. 1b eq. 1d eq. 2b eq. 2d eq. 3b eq. 3d

Age

Race and
Ethnicity

Family
Structure

Education
Level

Language

Employment

Housing and
Expenses

Mobility
1. Constant
2. log[AVHHIN]
3. log[AVHHIN]
2

8. log[POP517]
9. log[POP1864]
10. log[POP65P]
11. log[NHWHITE]
12. log[NHBLACK]
13. log[NHASIAN]
14. log[NHOTHER]
15. log[HISPANIC]
16. log[MCWKID]
17. log[MCNKID]
18. log[FHWKID]
19. log[FHNKID]
20. log[NOHSD]
21. log[HSD]
22. log[PSE]
23. log[ENGL]
24. log[NOENGL]
25. log[UNEMP]
26. log[LOWSK]
27. log[MEDSK]
28. log[HIGHSK]
29. log[OWNOCC]
30. log[RNTOCC]
31. log[SFR]
32. log[NEWCON]
33. log[GT1PR]
34. log[WELFARE]
35. log[GT35RNT]
36. log[PBTRAN]
37. log[AUTO]
38. log[CMT25]
39. log[CMT2544]
40. log[CMT45P]
41. log[SMHSE]
42. log[FORCNTR]
17.15*
-6.75*
0.67*

5.82*
-2.15*
0.20*
0.12*
0.30*
0.01*
0.01
0.01*
-0.01*
0.00*
0.00*
-0.13*
0.04*
0.01*
0.00*
0.00
-0.16*
-0.15*
-0.02*
0.00
0.01*
-0.01*
0.04*
-0.14*
0.03*
-0.01
0.00
0.01*
0.01*
0.04*
0.03*
0.00
-0.01
-0.05*
-0.02*
-0.01*
0.03*
0.00
-28.23*
12.31*
-1.31*

-14.32*
6.54*
-0.71*
-0.18*
-0.33*
0.00
-0.01
-0.01*
0.00*
0.00*
0.00*
0.12*
-0.11*
-0.01*
-0.01*
0.00
0.20*
0.18*
0.01
0.00*
-0.01*
0.05*
-0.09*
0.19*
0.00
0.04*
-0.01*
-0.01*
0.00*
-0.04*
-0.04*
0.00
0.00
0.07*
0.02*
0.01*
-0.04*
0.00
12.07*
-5.56*
0.64*

9.50*
-4.39*
0.51*
0.06*
0.02*
-0.01*
0.00
0.00*
0.00*
0.00
0.00*
0.01
0.07*
0.00
0.01*
0.00
-0.03*
-0.03*
0.00
0.00*
0.00
-0.03*
0.05*
-0.05*
-0.03*
-0.03*
0.01*
0.00
-0.01*
0.00
0.01*
0.00
0.01
-0.02*
-0.01*
0.00*
0.01
0.00*
R
2

F
N
0.58
3,617
5,340
0.88
1,094
5,340
0.44
2,063
5,340
0.82
638
5,340
0.93
34,006
5,340
0.95
2,496
5,340
* statistical significance at the 10 percent level for a two-tailed t-test

In all models, the population of working adults (18-64 years of age, variable 9) contribute
substantially, although much less so for models of affluent class share, and differently for
67
models of poor and middle class. While the share of poor (equation 1d) increases as the
working age population increases, the share of the middle class (equation 2d) decreases, and
the share of affluent (equation 3d) remains relatively unchanged. The share of non-Hispanic
white population does not significantly affect any of the models and while other
race/ethnicity variables (variables 11-15) contribute with statistical significance, the
magnitudes of effects are very small. Language ability also has small or statistically
insignificant effect in each model of income class share shown in Table 6. Family structure
explains poor and middle class shares in opposing ways. The middle class share increases
with married couple households, while the share of poor decreases. The opposite pattern is
observed for all other family types for these two income classes. Housing and expense
variables have greatest effect on the model for poor, with greater crowding (variable 33),
public assistance (variable 34) and rental cost burden (variable 35) increasing with share of
poor in cities. Mobility variables, especially those related to commuting by automobile to
work decrease with share of poor and this effect is statistically significant. Again, the
opposite pattern is observed for models of middle class. The effect of these variables on
explaining share of affluent class in cities is either small or not statistically significant.

The greatest contribution of all structural variables observed for all models relate to
education and employment skill level, all of which have statistically significant relationships
with class shares. In particular, higher rates of receiving a high school diploma (variable 21)
or a college education (variable 22), and having either low or high skill jobs (variables 26,
68
28)
27
consistently decrease the share of poor (and affluent) in cities, while increasing the
share of the middle class. By contrast, having medium skill jobs (variable 27)
28
explain
greater share of both poor and affluent classes and a lower share of middle class.
Figure 5 adapts an illustration by Ahluwalia showing the overall effect of additional
structural variables on patterns of development by displaying graphs of equations with and
without these additional variables. Dotted lines show traces of equations 1b, 2b and 3b, while
solid lines show equations 1d, 2d and 3d. The curves for equations 1d, 2d, and 3d are
computed by holding all explanatory variables other than average household income constant
at their mean values for the sample.

Ahluwalia interprets a flattening of ascending and descending phases of income shares to be
the effect of structural changes that favor greater equality of income. In this manner, the
contribution of many structural variables considered in this study decrease the share of poor
in cities with economic growth. These additional structural influences also tend to stabilize
the share of middle class, while shares of both poor and affluent decline. The implications of
these observations are discussed in Chapter 3.

27
“Low skill” jobs considered in this study include operators, assemblers, transportation, and material moving
workers, non-farm laborers and service workers as defined by the U.S. Office of Management and Budget
(OMB) “Standard Occupation Codes.” High skill jobs include professional and technical occupations,
executives, managers, and administrators. (OMB 2000)

28
“Medium skill” jobs considered in this study according to OMB Standard Occupational Codes include sales
workers, administrative support and clerical workers, precision production, craft and repair workers.
69
Figure 5: Influence of Community Indicators on Income Class Shares of U.S. City
Populations in 2000 by Level of Development
0.0
0.2
0.4
0.6
0.8
1.0
4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6
log[average household income in 2000]
share of city population in 2000
eq. 1d
eq. 2d
eq. 3d

Note: dotted lines depict traces of equations 1b, 2b and 3b presented in Figure 4.

Segregation models of Kuznetsian development
A poverty-affluence dissimilarity index is also employed as a dependent measure in this
study. However, an attempt to replace the dependent variable in the class-share models
specified above with a poverty-affluence dissimilarity index realizes the concerns noted by
Ahluwalia. That is, when modeling poverty-affluence dissimilarity using structural variables
defined previously for a subset of cities
29
, R
2
decreases from values of 0.88-0.95 in equations
1d, 2d and 3d, by more than half to a value of 0.42. Ahluwalia describes this problem as one
of relative insensitivity of index measures to variation among the explanatory variables.

29
A meaningful value for the index requires at least two census tracts over which counts can be distributed.
There were 820 U.S. cities in 2000 for which this was not the case and therefore dissimilarity indices could not
be computed, leaving N = 3,048 for a sample of cities that meets this requirement. Note that the excluded cities
were very small, with average total population of 4,387 in 2000. The share of total U.S population in cities in
2000 accounted for by the excluded cities was 6.47%.
70
Additional steps were taken to improve goodness of fit of models to the data. Table 7
summarizes several models with poverty-affluence dissimilarity as the dependent measure
along with a variety of explanatory measures. As before, the goodness of fit improves when
the income variable is entered in quadratic form (R
2
value increases by an order of
magnitude, from 0.003 to 0.02, from equation 4a to 4b.) Considering selected structural
variables for race/ethnicity, education and employment skill, which were observed to be
statistically significant achieves R
2
= 0.37, as shown in equation 4c. But the overall decrease
in fit for models in Table 7 by comparison to models of class share (Tables 4-6) suggests that
dissimilarity index-based models are not explaining equally well the variation among data for
cities. A strong hunch is that the decrease in R
2
may be because the explanatory variables
that contribute to Kuznetsian development in cities themselves have a strong geographical
component that is not represented in models specified by equations 4a-4c. To explore this
possibility, equation 4d summarizes results of modeling select variable groups, namely
race/ethnicity, education, and employment themselves as indices of dissimilarity. That is,
dissimilarity indices were constructed for white/black segregation and non-Hispanic
white/Hispanic segregation. Low/high and medium/high levels of education were also
considered, where low level represents those who have not received a high-school diploma,
medium level represents those who have received a high-school diploma but not attended
college, and high level represents those who have attended college. And, low/high,
low/medium and medium/high skill employment were also considered, where these
categories of low, medium and high skill correspond to those described before. Goodness of
fit for equation 4d improves substantially (R
2
= 0.62), even for this limited set of indicators,
when modeling these variable groups as dissimilarity indices, suggesting that doing so has
71
better explained the geographic distribution of wealth in cities that poverty-affluence
dissimilarity also describes. Equation 4e includes also a categorical variable that indicates
whether or not the city has a sustainability initiative according to Portney (2003).

Table 7: Regressions Explaining Poverty/Affluence Dissimilarity in 2000 by Level of
Development and Considering Selected Structural Variables
Dependent Variable: Poverty/Affluence Dissimilarity Index
Explanatory Variable eq. 4a eq. 4b eq. 4c eq. 4d eq. 4e
1. Constant
2. log[average household income in
2000]
3. log[average household income in
2000]
2
6. log[total population in 2000]
11. log[NHWHITE]
12. log[NHBLACK]
15. log[HISPANIC]
20. log[NOHSD]
21. log[HSD]
22. log[PSE]
26. log[LOWSK]
27. log[MEDSK]
28. log[HIGHSK]
43. Non-Hispanic white/black
dissimilarity
44. Non-Hispanic white/Hispanic
dissimilarity
45. Low/high education dissimilarity
46. Medium/high education
dissimilarity
47. Low/high skill employment
dissimilarity
48. Low/medium skill employment
dissimilarity
49. Medium/high skill employment
dissimilarity
50. Dummy variable for sustainability
initiative
0.66*
-0.07*

-16.60*
7.16*

-0.76*

-2.36
1.03

-0.14

0.14*
0.12*
0.03*
-0.01
0.07*
-0.18*
-0.02
0.13*
0.12*
-0.16*

-5.16*
2.23*

-0.25*

0.06*

0.01

0.05*

0.47*
0.10*

0.18*

0.50*

0.22*

-5.19*
2.21*

-0.25*

0.07*

0.01

0.05*

0.47*
0.10*

0.18*

0.50*

0.22*

-0.06*

R
2

F
N
0.003
8
3,048
0.02
32
3,048
0.37
151
3,048
0.62
491
3,048
0.62
449
3,048
* statistical significance at the 10 percent level for a two-tailed t-test

While poverty-affluence dissimilarity models have lower R
2
than class share models—
consistent with Ahluwalia’s concern for the sensitivity of summary indices—equations 4d
72
and 4e do present some statistically significant relationships among variables implicated in
the development process. All explanatory variables in equations 4d and 4e are statistically
significant except white-black dissimilarity (variable 43). And while also significant, the
contribution of white-Hispanic dissimilarity (variable 44) to the model is small; these suggest
that factors other than race better explain income inequality in U.S. cities. Again the greatest
contribution to models of income inequality is disparity in education, particularly differences
between those who have attended college and those who have not received a high school
diploma (variable 45). Differences between low skill and medium skill employment (variable
48) also contribute substantially to the model, as does city size (variable 6). And, despite
accounting for only 1 percent of sampled cities, the presence of a sustainability initiative
(variable 50) makes a small, statistically significant contribution to the model specified by
equation 4e, and in the direction of decreasing income inequality. The overall contribution of
the structural variables to the model of poverty-affluence dissimilarity specified by equation
4d is demonstrated graphically by comparing it to equation 4b, as shown in Figure 6. As
before, the curve for equation 4d is computed by holding all explanatory variables other than
average household income constant at their mean values for the sample. In Figure 6, as in
Figure 5, the effect of these additional structural variables is to “flatten” the curve described
by income variables alone, similar to what Ahluwalia demonstrated. Another observation is
that the “turning point” of the curve given by equation 4d precedes that of equation 4b. That
is, the descending phase of the Kuznets Curve begins at lower income level when the
influence of structural variables is considered.

73
The difference between turning point values can be quantified.
30
In the case of 4b and 4d, for
example, turning points occur at average household incomes of $54,473 and $30,548
respectively. Thus, along dimensions of economic growth, structural factors appear to have
an effect of shifting the descending phase of the curve by $23,925 per household on average.
The meaning of this observation is discussed in Chapter 3.

Figure 6: Influence of Selected Structural Variables on Poverty-Affluence Dissimilarity
in U.S. Cities in 2000
0.0
0.2
0.4
0.6
0.8
1.0
4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6
log[average household income in 2000]
poverty-affluence dissimilarity in 200
eq. 4f
eq. 4b

30
Given a mathematical specification of the Kuznets relationship
y= ax
2
+ bx +c
where
y income inequality, i.e. poverty-affluence dissimilarity index value
x = average household income
and turning point occurs at dy/dx = 0, or
dy/dx = 2ax + b = 0
Solving, the turning point thus occurs at
x =-b/2a.
Note that x-values are log transforms of average household income. The corresponding average household
income values are 10
x
. For two curves with turning points at x
1
and x
2
, the difference in average household
income is thus 10
x1
– 10
x2
.
eq.4d
74
The relevance of Kuznetsian development to sustainable cities
To this point, consideration has been given to either all U.S. cities in 2000 (N = 5,340) for
example to create the models represented by Figures 4 and 5, or to a smaller though still
substantial number of U.S. cities for which poverty-affluence dissimilarity indices can be
computed (N = 3,048) to create the models represented in Figure 6. Figure 7 presents
observations about sustainable cities highlighted by Portney (2003) and their relationship to
models of Kuznetsian development presented above. These observations are germane to the
second hypothesis of this study. In Figure 7, sustainable cities are overlaid as points in
relation to the curve presented in Figure 6 according to their computed values for income
inequality and level of economic development. Accordingly, sustainable cities deviate
substantially from the curve describing U.S. cities overall.

Figure 7: U.S. Sustainable Cities and the Kuznets Curve
0.0
0.2
0.4
0.6
0.8
1.0
4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6
log[average household income in 2000]
poverty-affluence dissimilarity in 200
eq. 4f
eq. 4b

75
With the exception of Stuart, Florida given by point (4.63, 0.33) and Annapolis, Maryland
given by point (4.82, 0.11), sustainable cities exceed by 41 percent the levels of poverty-
affluence dissimilarity observed elsewhere in the U.S. This is a statistically significant
difference in income inequality (p < 0.0001, two-tailed test). Furthermore, sustainable cities
have average household incomes ($53,881 in 2000) that exceed those of U.S. cities that do
not have sustainability initiatives ($49,925 in 2000), but this difference is not statistically
significant (p = 0.30, two-tailed test).

Considering income class shares, sustainable cities have a 25 percent larger share of poor (p
= .02, two-tailed test) and a six percent smaller share of middle class (p = 0.0005, two-tailed
test) on average. Sustainable cities also have an 18 percent larger share of affluent
households on average, but this difference just misses statistical significance at the 10 percent
level (p = 0.13, two-tailed test). Overall, income class groups appear to be more polarized by
income class in sustainable cities than in other cities in the U.S.

Comparing sustainable cities to other U.S. cities according to variables in equation 4d reveals
statistically significant differences (p < 0.10, two-tailed test) for all explanatory variables
except measures of household income (variables 2 and 3). Mean values in sustainable cities
exceed those of other cities in all cases. Considering that this may be a systematic difference
explained by differences in city size,
31
analogs for variables 8-42 were developed that
normalize counts for a given city by the number of people, households, or housing units as

31
Indeed the populations of sustainable cities significantly exceeded those of other cities considered in models
of poverty-affluence dissimilarity by 187,617 persons on average in 2000 (p < 0.0001, two-tailed test).

76
appropriate. Table 8 provides results of comparing these variables for sustainable cities and
others, further summarized to include only those where statistically significant differences
occurred (i.e., p < 0.10, two-tailed test).
32

Table 8: Comparison of Cities With and Without Sustainability Initiatives for Selected
Variable Counts Normalized by 2000 Population
Variable
Group
Explanatory Variable Difference of Means t-Statistic
Age

Race and
Ethnicity

Family
Structure
Education
Level
Language

Employment

Housing and
Expenses

Mobility

8. POP517
9. POP1864
10. POP65P
11. NHWHITE
12. NHBLACK
13. NHASIAN
14. NHOTHER
16. MCWKID
17. MCNKID
21. HSD
22. PSE
23. ENGL
24. NOENGL
26. LOWSK
27. MEDSK
28. HIGHSK
29. OWNOCC
30. RNTOCC
31. SFR
33. GT1PR
35. GT35RNT
36. PBTRAN
37. AUTO
39. CMT2544
41. SMHSE
42. FORCNTR
-0.04
0.07
-0.03
-0.15
0.06
0.04
-0.01
-0.06
-0.07
-0.10
0.13
-0.10
0.03
-0.03
0.11
-0.06
-0.15
0.18
-0.23
0.02
0.09
0.07
-0.11
0.04
-0.09
0.03
0.0002
<0.0001
0.002
0.0001
0.04
0.003
0.001
<0.0001
<0.0001
<0.0001
<0.0001
0.002
0.02
0.006
<0.0001
0.0001
<0.0001
<0.0001
<0.0001
0.04
<0.0001
0.0004
0.0003
0.01
<0.0001
<0.0001

For each variable in Table 8, the magnitude and direction of difference of means is shown,
such that a positive difference of means indicates that the average value for sustainable cities

32
Excluded from Table 8 (because differences were not statistically significant) are shares of Hispanic
population (variable 15), female-headed households (variables 18 and 19), and people who have not received
high school diplomas. Also excluded are rates of unemployment (variable 25), share of people receiving public
assistance (variable 34), and shares of population that commute less than 25 minutes to work, or more than 45
minutes to work.
77
exceeds that of others. A negative difference of means indicates that the average value for
sustainable cities is less than others. Accordingly, sustainable cities are shown to have
greatest difference for a number of measures of race/ethnicity, education, language,
employment, housing and mobility. Namely, sustainable cities have significantly lower
overall shares of single family detached homes and higher shares of renters. Sustainable
cities have significantly lower shares of non-Hispanic whites and English-speaking
households, and significantly higher shares of those who have attended college. Sustainable
cities have significantly higher rates of employment in “medium skill” jobs in sales,
administrative and clerical support, as well as precision production, craft and repair work.
Sustainable cities have significantly lower rates of automobility. These observations are
discussed further in Chapter 3.

Temporal illustrations of “making the turn” or “missing the turn”
Figure 7 illustrated the overall relationship of U.S. sustainable cities to an average
development path for other U.S. cities. The substantial deviation by sustainable cities from
that path directed attention to several statistically significant differences between sustainable
cities and others. Yet, while Figure 7 and Table 8 suggest that there may be important
differences between sustainable cities and others according to cross-sectional analyses, these
do not necessarily provide insight into cities which may be in the process of “making the
turn” or “missing the turn” toward greater income equality, hence instances of agency that
may explain important policy changes over time. A temporal (or longitudinal) perspective
may reveal this. Figure 8 presents a temporal view of sustainable cities according to
dimensions of Kuznetsian development over the 1980-2000 period. Each sustainable city is
78
shown in Figure 8 as a point with a tail.
33
The point indicates values in 2000, while the end of
the tail indicates corresponding CPI-adjusted values for 1980. Selected cities have been
highlighted and labeled because they exemplify a variety of paths along dimensions of
Kuznetsian development. For example, New Haven, Connecticut and Olympia, Washington
experienced substantial lowering of income inequality during the 1980-2000 period (25
percent and 17 percent respectively), but these cities appear to have achieved relatively lower
levels of economic development.

Figure 8: Kuznetsian Development and U.S. Sustainable City Development, 1980-2000
Austin
Annapolis
Boulder
Cambridge
New Haven
Olympia
Phoenix
Portland
S. Francisco
S. Barbara
S. Monica
Scottsdale
Stuart
Tampa
0.1
0.2
0.3
0.4
0.5
0.6
0.7
4.5 4.6 4.7 4.8 4.9
log[average household income in 2000]
poverty-affluence dissimilarity in 200
eq. 4e
eq. 4b

Among the cities exhibiting highest levels of economic development, Scottsdale, Arizona,
achieved substantial economic growth over the period (53 percent), but appears to have done

33
The inspiration for this illustration is from the work of Charles Wilson, inventor of the “cloud chamber.” The
cloud chamber provides insights into forces by and upon alpha and beta particles as they pass through super-
cooled liquids. The idea is that the paths of Kuznetsian development shown in Figure 8 may similarly reveal
“forces” related to structure and agency in cities, just as paths in cloud chambers generated insights into the
physics of particles.
79
so with considerable impact upon income inequality, which also increased (by 23 percent)
during that time. Tampa, Florida and Phoenix, Arizona followed similar paths, demonstrating
the highest values of poverty-affluence dissimilarity of all sustainable cities in this sample.
Austin, Texas also demonstrated high poverty-affluence dissimilarity in 2000, having
apparently done little to change this value over the 20-year period. In 2000, these three cities
(Tampa, Phoenix and Austin) exceeded 98 percent of all U.S. cities on this measure of
income inequality, while average household income grew by 56, 29 and 50 percent
respectively during that time. Portland, Oregon demonstrated a stable, intermediate level of
income inequality over the period, remaining unchanged while average household income
increased by 40 percent. Similarly, household income grew by 40 percent in Boulder,
Colorado while income inequality increased by only 2 percent during that time. Cambridge,
Massachusetts, San Francisco and Santa Monica, California emerge among sustainable cities
that have demonstrated greatest progress toward achieving greater equality of income while
also achieving substantial economic growth over the 1980-2000 period, decreasing poverty-
affluence dissimilarity by 36, 16 and 27 percent, while increasing average household income
by 84, 88 and 74 percent respectively during that time. Indeed measures of income inequality
demonstrated by Santa Monica and Cambridge over the period correspond with the least
deviation from the curve among sustainable cities that have achieved highest levels of
economic development.

Methodological bracketing and a Kuznetsian typology of sustainable cities
While structuration implies an inseparability of structure and agency, structuration analyses,
such as those by Dear and Moos (1986) and Moos and Dear (1986), consider the structural
80
aspects of the research setting separately from the strategic conduct of agents involved,
ultimately to integrate findings in the end to present a fuller understanding of the structure-
agency dualism. The approach is supported by the idea of “methodological bracketing”
offered by Giddens (1989). Phipps (2001, p.194) describes methodological bracketing as a
“reflexive moment of attention in which a researcher intellectually freezes the ongoing
course of structuration” and further that

…two methodological brackets are for (1) the analysis of institutionalized
properties of systems to identify properties that are chronically reproduced
over the long term; versus (2) the analysis of the strategic conduct of actors to
identify their social practices that regenerate or alter the institutionalized
properties of systems for day-to-day life.

Methodological bracketing is used in this study as a means by which to focus attention on
selected sustainable cities that may provide insights into the strategic conduct of agents,
which may further explain (or be explained by) associated structures evident in these cities.
To this end, the second hypothesis of this study suggests that Kuznetsian development can
inform a typology of sustainable cities that reflects differences among social and economic
institutions of cities, from which cases can be selected for further investigation. In this way,
selected cases are bracketed with respect to institutionalized properties of cities such that a
clearer look at the strategic conduct of agents in cities is possible.

A conceptual typology based on Kuznetsian development of U.S. cities is shown in Figure 9.
Three types of cities are shown—A, B and C—described as follows:
• Cities found in area A are “Type A” cities. These cities have relatively lower income
inequality, and lower average household income. These cities appear to be in an early
81
Figure 9: A Kuznetsian Typology of Cities

• stage of economic development, a stage in which substantial economic growth has
not occurred. Substantial benefits associated with economic growth have not been
generated that would lead to uneven distribution of these benefits in these cities.
• “Type B” cities have higher average household incomes than Type A cities (but lower
than “Type C” cities to be described next), which suggests that Type B cities have
experienced higher levels of economic growth than Type A cities. With this growth,
benefits have generally accrued more to affluent households than to poor households,
as reflected in higher levels of income inequality in these cities. The tension created
by this uneven distribution of income may serve to motivate policy change in these
cities, but currently Type B cities remain in a transitional state, as observed by their
locus near the turning point of the Kuznets Curve.
• Type C cities have the highest average household incomes of all cities. While
Kuznetsian development predicts that income inequality is lower than that observed
Poverty-Affluence Dissimilarity
Average Household Income
A B C
82
for Type B cities, for some Type C cities, this may not be the case. Type C cities are
differentiated further into one of two types, those which have “made the turn,” thus
fulfilling the wishful thinking of Kuznets, and those which have “missed the turn,”
thus demonstrating high levels of income inequality with further economic growth.

Table 9 presents results of classifying the sustainable cities compiled by Portney (2003)
according to this typology. The three groups are defined according to terciles of the average
household income distribution for U.S. cities. While many rules for assigning cities to groups
might be applied, this simple approach seems most appropriate for the initial explorations
envisioned in this study. Accordingly, five sustainable cities are categorized as Type A cities,
while 11 and 13 sustainable cities are categorized as Type B and Type C respectively. Where
available, scores for Portney’s “Taking Sustainable Cities Seriously” index
34
are also shown
in parentheses.

Table 9: Sustainable Cities by Kuznetsian Type
Type A Type B Type C
Brownsville, TX (7)
Cleveland, OH (14)
Milwaukee, WI (6)
New Haven, CT (8)
Tucson, AZ (18)
Stuart, FL (na)
Burlington, VT (na)
Chattanooga, TN (18)
Lansing, MI (na)
Indianapolis, IN (9)
Ithaca, NY (na)
Jacksonville, FL (15)
Oklahoma City, OK (na)
Olympia, WA (8)
Orlando, FL (11)
Tampa, FL (19)

Austin, TX (17)
Boston, MA (14)
Boulder, CO (26)
Cambridge, MA (14)
Phoenix, AZ (15)
Portland, OR (25)
San Francisco, CA (23)
San Jose, CA (26)
Santa Barbara, CA (10)
Santa Monica, CA (25)
Scottsdale, AZ (26)
Seattle, WA (30)
Annapolis, MD (na)

34
The Taking Sustainability Seriously Index ranges from 6-30, with median value of 15. For a given
sustainability initiative of a city, the index measures presence/absence of 34 program elements within seven
broad categories of sustainable development policies (Portney 2003).
83
Note that the trends shown in Figure 8 suggest that Type C cities have followed a
considerable variety of paths over the 1980-2000 period, much more so than cities of other
types. The paths of Type C cities range from maintaining consistently high or increasing
income inequality (e.g., Phoenix and Austin) to substantially decreasing income inequality
(e.g. Santa Monica and Cambridge), while others maintained intermediate levels (e.g.
Boulder and Portland) during that time. Table A3 in the Appendix presents mean values of
community indicators for these six illustrative Type C sustainable cities (Phoenix, Austin,
Boulder, Portland, Santa Monica and Cambridge). If Type C cities are further classified into
those which have “made the turn” as may be indicated by poverty-affluence dissimilarity
values that match (or are lower) than average values for the U.S., only one U.S. city—
Annapolis—can be classified as such. Santa Monica and Cambridge appear most likely to
claim this achievement among those sustainable cities remaining.

Modeling the Urban Regime
As illustrated above, the categorization of sustainable cities according to the income
distribution of U.S. cities is skewed toward higher levels of economic development,
suggesting that the conditions created by economic growth may have helped to motivate
policy changes related to sustainable development. Several Type C sustainable cities in
particular appear to be challenged in reconciling economic development with achievement of
social goals, while others appear to have overcome this challenge in their further
development. A deeper look into this variability of Type C sustainable cities may be
instructive. Therefore this group of cities is selected for investigation of human agency that
may explain differences in achieving greater (or lesser) equality of income. Consideration of
84
human agents in this study draws largely from theories of urban political economy and
geography. As described in Chapter 1, these theories generally assume that people associate
within cities either to align with or confront the governing elite in a manner that translates to
policy outcomes. Additional steps in this study seek to identify politically active groups and
specify their associations and alignments (to investigate the third hypothesis of this study)
and relate these associations and alignments to observed patterns of development (to
investigate the fourth hypothesis).

Cluster analysis of sustainable cities
Agglomerative hierarchical clustering (Bailey 1994) is employed to model groups of interest.
The approach seeks to represent similarities of members which may reflect shared interests
and beliefs, and which Sabatier and Jenkins-Smith (1989) have argued form the basis of
“advocacy coalitions” that influence the policy process. The inspiration for this approach also
draws from growth machine theory (Molotch 1976), which suggests that a natural alignment
between government officials and the land-owning elite largely explains growth and
development of cities, and also from the proposition by Acemoglu and Robinson (2002) that
the turning point of the Kuznets Curve may be the result of the political mobilization of
citizens. Considering also the reasoning of Portney (2003), the “progressive middle class
regime” (Stone 1986) is hypothesized to play a particularly important role in making this turn
in sustainable cities. Agency is further imagined upon a backdrop of deliberative planning
processes that are open, to varying degrees, to participation by various groups in city
governance.

85
The cluster analysis in this study seeks first to identify groups based on shared
characteristics, and second to support reasoning that these groups plausibly identify regimes
capable of influencing city governance. As an illustrative example, results for Santa Monica,
California are shown as a dendrogram in Figure 10.
35
Santa Monica is selected because it is
generally regarded as a model sustainable city in the literature and is familiar to this author.
Clusters were created using tract-level data from the 2000 Census for the entire set of
community indicators previously described in Table 3 in an attempt to model groups based
on a rich set of characteristics beyond merely income class. The x-axis in Figure 10 shows
the similarity (correlation coefficient) value upon which each successive cluster is based, and
the y-axis shows census tract numbers for each entity, as grouped into three clusters—C1,
C2, and C3. In this example, clusters emerged where similarity of successive candidate
members and cluster members achieved correlation coefficient values of approximately 0.75.
The correlation coefficient used for clustering was the Spearman correlation coefficient ( ρ), a
non-parametric measure of statistical dependence.
36
The linkage method used for clustering
was complete.
37
Cluster analysis was facilitated by use of XLSTAT software—an extension
available for Microsoft Excel.

35
A dendrogram summarizes graphically the combining of entities with successive stages of clustering.

36
This choice of measure was made after observing that the community indicators in this study are not normally
distributed, thus correlation measured by the more familiar Pearson correlation may not be appropriate. The
Spearman correlation coefficient measures relationship among variables on an ordinal (rank-order) scale. This
scale has an intuitive appeal, as sorting of individuals among groups seem more likely to be based on relative
than absolute characteristics.

37
Complete linkage methods, also called “furthest neighbor” methods, admit successive entities to the growing
cluster based on more rigorous membership criteria than single linkage methods, such that members are similar
to one another along more measures, thus producing “tighter, more compact” clusters (Bailey 1994). While this
approach may exaggerate differences among resultant clusters, it is judged to more clearly identify regimes for
further testing.

86
Figure 10: Dendrogram of Agglomerative Hierarchical Clustering for Santa Monica,
California

Results are mapped by tract for the city in Figure 11. C1 is shown in white, C2 in light gray,
and C3 in dark gray. The tract numbers shown are for the central objects
38
of the cluster
classes, and further indicated by asterisks in Figure 10. Table A4 in the Appendix provides
additional descriptive statistics for these tracts. From Figure 11, several observations can be
made regarding the plausibility of identifying regimes by this approach. That boundaries of
neighborhood associations
39
(shown as darker lines) do not divide cluster classes across
similar groups, and instead reinforce group distinctions, suggests that clustering by this
method has reasonably identified culturally and politically separable groups.

38
The central object of a cluster class is that data object which best approximates the class centroid. The class
centroid describes the average values of the group of objects that comprises the cluster.

39
Neighborhood Association boundaries are provided by the City of Santa Monica (2010).

6037701701
6037701502
6037701602
6037701400
*6037701501
6037701302
6037701202
6037702300
6037702201
6037702202
6037702100
6037702000
6037701601
6037701301
*6037701201
*6037701802
6037701801
6037701702
6037701900
0.40 0.50 0.60 0.70 0.80 0.90
C2
C1
C3
ρ = 0.75
87
Figure 11: Neighborhood Associations and Modeled Regimes for Santa Monica,
California

Note however that regime theory describes four regimes, not three. Upon deeper inspection,
public records for 2000 indicate that four of seven council members lived in one of two
north-westernmost tracts shown (tracts comprising the North of Montana Association). Two
more council members lived in the adjacent neighborhood labeled the Wilshire/Montana
Neighborhood Coalition. According to the descriptive statistics for the central objects of the
classes that define these areas, C1 and C2 describe the wealthiest landowners in the city. The
result suggests that the development regime and the maintenance regime are largely
inseparable among neighborhoods in Santa Monica. From a political geography perspective
this city’s landowning elite and government officials appear to be well aligned—a finding
that is consistent with growth machine theory. For purposes of this study, cluster C1 (shown
in white in Figure 11) therefore represents a hybrid “development/city maintenance” regime,

06037701802
Northeast
Neighbors
Wilshire/Montana
Neighborhood
Coalition
Mid-City
Neighbors
Pico
Neighborhood
Association
Friends of
Sunset Park
Ocean Park
Association
06037701201
06037701501
North of Montana
Association
06037701802
Northeast
Neighbors
Wilshire/Montana
Neighborhood
Coalition
Mid-City
Neighbors
Pico
Neighborhood
Association
Friends of
Sunset Park
Ocean Park
Association
06037701201
06037701501
North of Montana
Association
C2 (Progressive
Middle Class
Regime)
C3 (Low-Income
Expansion
Regime)
C1 (Development/City
Maintenance Regime)
88
a group of landowning and governing elite that is presumed to associate not only in the
coffee shops, parks and schools and other shared spaces of their neighborhoods, but in the
meeting rooms of city hall to make governance decisions.

A political landscape consistent with growth machine theory is further corroborated by the
fact that no council members live in Santa Monica’s Pico neighborhood area, the poorest and
most ethnically diverse area of the city. Divided by the Santa Monica Freeway, which opened
in the 1960s, the Pico neighborhood presents a classic example of an area directly impacted
by the interstate highway program in the U.S. Among the greatest concerns of residents in
this area, i.e. members of the Pico Neighborhood Association, are poverty, lack of affordable
housing and gang violence. Residents of the Pico neighborhood plausibly defines the C3
cluster, the “low-income expansion regime” as shown in dark grey in Figure 11. The C2
cluster, representing the “middle class progressive regime” occupies areas that are light
grey.
40

Having employed cluster analysis to plausibly identify the existence of groups which may
represent active urban regimes in Santa Monica, a next step is to investigate whether linkages
among these groups explain observed patterns of development as theory would predict.
Recall that influence on city governance and policy change is theorized to depend on the
informal associations among interest-based regimes. Given that the elite class and city
maintenance regimes are largely inseparable using census tract data, this regime association
is assumed to have a correlation coefficient of approximately 1. Figure 12 summarizes values

40
Note that illustrations do not depict non-residential (i.e. commercial and industrial) land uses in the city, a
substantial number of which occur in the C3 cluster area.
89
of correlation coefficients among clusters defined using all community indicators listed in
Table 3 for remaining regime associations among elite, middle and lower class regimes in
U.S. sustainable cities. Values are presented as box plots, further arranged by sustainable city
type. The “whiskers” in Figure 12 represent the overall ranges of correlation coefficient
values while the boxes represent the interquartile ranges. Horizontal lines dividing the boxes
represent the median values of the coefficients for the various city types. Table A5 in the
Appendix presents correlation coefficients among modeled regimes for all sustainable cities
considered in this study.

Figure 12: Correlations Among Regimes for U.S. Sustainable Cities in 2000

Note: Y-axis denotes correlation coefficient (Spearman) values, while x-axis denotes correlations among Elite
(E), Middle (M), Lower (L) classes for variables listed in Table 3.

From Figure 12, sustainable cities have highly variable and generally overlapping ranges of
values for correlation coefficients. Type A cities appear to have a generally high degree of
association among elite and lower, and also middle and lower classes. Type B and Type C
cities both have a generally high degree of association between elite and middle classes.
0.40
0.50
0.60
0.70
0.80
0.90
E-L E-M M-L E-L E-M M-L E-L E-M M-L
Type A Type C Type B
90
Inspecting more closely the patterns of among regime associations listed in Table A5 in the
Appendix, one intriguing pattern emerges. From the median values shown in Figure 12 and
also Table A5, a “low, high, high” pattern can be seen among several Type C sustainable
cities, especially those having highest average household incomes—Santa Monica,
Cambridge, San Francisco, and Scottsdale. The same pattern is not observed at all among
Type A or B sustainable cities. The pattern indicates a higher degree of association between
middle class regimes with both elite and lower class regimes, and lower degree of association
between elite and lower class regimes in these cities. The pattern is observed not only in the
most affluent sustainable cities considered in this study but also among many of those
affluent cities having the lowest measures of income inequality, including those which have
experienced decreasing levels of income inequality over the 1980-2000 period (an exception
is Scottsdale.)

Given a possible dual association the middle class has with both elite and lower classes—
indicated by shared cultural characteristics of group members—an interpretation is that
where planning for sustainable development in cities translates to achieving social goals such
as equality of income, a mediating influence of the middle class progressive regime may
have contributed to this effect. Table 10 summarizes results of further testing this claim. The
model specified by equation 5c in particular explores the influence of regime associations (as
indicated by the value of the correlation coefficients among them) on income inequality (as
poverty-affluence dissimilarity.)

91
Table 10: Regressions Explaining Poverty/Affluence Dissimilarity in Sustainable Cities
in 2000 by Regime Associations and Public Involvement in Planning
Dependent Variable: Poverty/Affluence Dissimilarity Index
Explanatory Variable eq. 5a eq. 5b eq. 5c
1. Constant
2. log[average household income in
2000]
3. log[average household income in
2000]
2
43. Non-Hispanic white/black
dissimilarity
44. Non-Hispanic white/Hispanic
dissimilarity
45. Low/high education dissimilarity
46. Medium/high education
dissimilarity
47. Low/high skill employment
dissimilarity
48. Low/medium skill employment
dissimilarity
49. Medium/high skill employment
dissimilarity
51. Correlation between elite and
lower class regimes
52. Correlation between elite and
middle class regimes
53. Correlation between middle and
lower class regimes
54. Dummy variable for involvement
of political and/or business leaders
55. Dummy variable for general public
involvement
-50.49
21.23

-2.23

-0.05

-0.25

1.38*
-1.30*

2.10*

-1.24

2.73*

-0.36
0.08

0.00

0.01

-0.54*

1.84*
-2.66*

10.82*

-9.39*

10.42*

2.54
-0.47

0.00

0.21

-0.18

0.31

-0.08

0.23*
R
2

F
N
0.72
5
29
0.95
10
13
0.83
4
12
* statistical significance at the 10 percent level for a two-tailed t-test

Equation 5a in Table 10 resembles that of equation 4d (Table 7) but considers data only for
the sustainable cities identified by Portney (N = 29). Results suggest that these same
variables fit data for sustainable cities as well as, in fact slightly better than data for U.S.
cities more generally (R
2
= 0.72 for equation 5a, by comparison to R
2
= 0.67 for equation
4d), however the influence of income variables just miss significance at the 10 percent level
(p = 0.13, two tailed test) casting some doubt on whether Kuznetsian development
fundamentally describes the experiences of these sustainable cities. Equation 5b focuses
92
attention further upon Type C sustainable cities (N = 13), and goodness of fit increases even
further (R
2
= 0.95) although the salience of the Kuznets process for this sample diminishes
further as well, as indicated by less significant income variables. It may be that the number of
cities is too few and their similarity to great to detect the curvilinear relationship more
generally observed among inequality and income variables for U.S. cities. Yet, poverty-
affluence dissimilarity continues to be explained well by dissimilarity indices for education
and employment skill level rather than by race-ethnicity separation, as was previously
demonstrated by models summarized in Table 7. In fact, dissimilarity of education and
employment skill contribute greatly to this model, with employment skill contributing most
of all among explanatory variables.

Turning attention to influences of regime associations on governance in Type C sustainable
cities, and also formal efforts by cities to involve the general public in planning, equation 5c
includes variables for these factors. Variables 51-53 describe correlations among various
income class groups. Variables 54 and 55 are categorical variables describing city efforts to
involve stakeholders. The overall fit of the data to the model represented by this equation is
good (R
2
= 0.83; p = 0.08, F-test) suggesting that poverty-affluence dissimilarity is generally
well explained by a combination of regime associations and public involvement in Type C
sustainable cities. Note that while the magnitude of the coefficients suggest that regime
associations contribute substantially to this model, they are not statistically significant at the
10 percent level (p = 0.34 and p = 0.45, and p = 0.43, t-test, for variables 51, 52 and 53
respectively). Interestingly, greater association between elite and lower classes increases
93
income inequality, as does greater association between middle and lower classes. Only the
elite-middle class association is observed to decrease income inequality in these cities.

Efforts to involve various members of the community in city governance of Type C
sustainable cities also contribute to the model represented by equation 5c. Data for variables
54 and 55 are provided by Portney (2003). According to Portney, involvement of political
and business leaders of cities (variable 54) included efforts by cities to engage mayors,
council members and members of the business community (e.g., Chamber of Commerce) in
the sustainability initiative. General public involvement (variable 55) included efforts by
cities to encourage participation in public hearings and “visioning” processes and to involve
neighborhood groups or associations in the various sustainability initiatives. The degree to
which community members were engaged in a given sustainability initiative was not
identified by Portney, only that some level of effort was made to involve these members of
society. Furthermore, Portney reported data on involvement in governance for only 23
cities
41
with only 12 of them being classified as Type C sustainable cities. Results suggest
that while the associations among regimes and involvement of political and business leaders
do not explain income inequality in sustainable cities with statistical significance (p = 0.34, p
= 0.45, p = 43, and p = 0.18, each in two-tailed tests, for the various regime associations and
for involvement of political and business leaders, respectively), general public involvement
does contribute significantly to decreased income inequality. Note also that the direction of
influence of political and business leaders opposes that of the general public in this model.

41
Portney (2003) does not report data for public involvement in Stuart, Burlington, Ithaca, Lansing, Oklahoma
City, or Annapolis. Therefore these cities are removed from models that consider public involvement as an
explanatory variable.
94
That is, involvement of political and business leaders appears to increase income inequality
(although not significantly), while the involvement of the general public appears to
significantly decrease income inequality. Implications of these results are discussed further in
Chapter 3.

95
CHAPTER 3: DISCUSSION

This chapter discusses the results presented in Chapter 2 with respect to four sets of research
questions presented in the Introduction. Specifically the existence of the Kuznets relationship
for U.S. cities is determined, and various model specifications are explored further. The
relevance of Kuznetsian development to sustainability planning in cities is considered and its
promise as a theoretical lens to help focus attention in sustainable cities research and policy
is discussed. Finally, several ideas about human agency are revisited for selected sustainable
cities, given also their relationship to the Kuznets Curve.

A Kuznets Curve for U.S. Cities
Just as Ahluwalia showed for developing countries, and Neilsen and Alderson showed for
U.S. counties, the classic Kuznets Curve emerges in cross-sectional data for U.S. cities. That
is, an empiric relationship exists between income inequality and economic development in
U.S. cities such that income inequality first increases with economic growth, then decreases
among these data. The finding provides a stepping-off point from which to explore aspects of
structure and agency in cities that may explain this relationship, as well as to consider the
relevance of these findings for sustainable cities research and policy.

The Kuznets Curve for U.S. cities is revealed in this study through a series of regression
model refinements adapted from Ahluwalia (1976a,b). These models considered income
class share and average household income using 2000 Census data for 5,340 U.S. cities. As
in the demonstrations by Ahluwalia, goodness of fit increased when entering average
96
household income in quadratic form (Table 4), thus indicating the classic Kuznets
relationship. The improvement was greatest in models where the middle class share was the
dependent measure (equation 2b). Somewhat curiously, the signs of equations for poor and
affluent (equations 1b and 3b) did not oppose one another as Ahluwalia showed for income
shares of upper and lower quintile groups in developing countries. Instead signs of equations
for both poor and affluent opposed that of the middle class. As shown in Figure 4, the
similarity of curves for poor and affluent may be explained mathematically by the fact that
shares of both of these classes occur along sections of upward-opening parabolas, similar in
orientation but out of phase with one another, such that as average household income
increases in cities, the share of poor decreases while the share of affluent increases. By
contrast, the model for middle class share is a parabola that opens downward, first increasing
and then decreasing across the data range. The pattern suggests that a different mechanism of
development may be at work than is observed at larger scales of aggregation (counties and
nations). It is also, perhaps, a clue that the middle class plays an especially important role in
the transformation of social institutions of cities during their growth and economic
development.

By comparison to Ahluwalia, who demonstrated that short term economic growth did not
explain patterns of development at the national level with statistical significance, a different
result was found for U.S. cities (Table 5). That is, growth in average household income
between 1990 and 2000 is statistically significant, at least in explaining growing shares of
poor and affluent classes in 2000 (and for these, the contributions to modeled relationships
are small). The result suggests that a greater polarization between classes may have occurred
97
during that time. Such finding is consistent with the observation by Harris and Goodwin
(2001) that alternative approaches to development, such as sustainable development, are
motivated by class differences that emerge out of traditional development thinking. Overall
growth of city population during that time period also makes statistically significant but
small contributions to models of development (equations 1c, 2c, and 3c). While the impact of
short term growth in cities cannot be dismissed (at least not that between 1990-2000), results
suggest that the Kuznets Curve largely reflects social structural conditions that have been
reinforced over a much longer term, as Ahluwalia also showed. In Table 6 several
explanatory variables for age, race/ethnicity, family structure, language, education,
employment, housing and mobility were shown to be statistically significant in models of
Kuznetsian development (equations 1d, 2d and 3d), with the greatest contributions of all
variables being made by educational attainment and employment skill level. Specifically,
higher rates of receiving a high school diploma or college education consistently decreased
the shares of poor (and also affluent) while increasing the share of the middle class in cities.
These results are similar to those reported by Ahluwalia for developing countries, and
support Kuznets’ claim that development that achieves greater equality of income may be
explained by movement by workers from lower to higher skill occupations, as further
enabled by education and training.

The general results described above are essentially replicated when changing the dependent
measure to one of geographic segregation, however the R
2
values decrease from those
ranging from 0.82-0.95 in models for class share (equations 1d, 2d and 3d) to 0.62 for a
model of income inequality that employs dissimilarity indices as both dependent and
98
independent measures (equation 4d). Nonetheless, goodness of fit is judged to be quite high
in this model. For comparison, Ahluwalia reports R
2
values of 0.46-0.76 for class share
models using cross-national data, while Nielsen and Alderson report R
2
values of 0.40-0.79
for models of Gini coefficient using cross-county data. The goodness of fit for models of
cities reported in this study fall squarely in these ranges, suggesting that results of this study
are generally consistent with comparable results reported elsewhere in the literature.

Observations about the poverty-affluence dissimilarity model are remarkable for several
additional reasons. First, model fit improves substantially when explanatory variables are
expressed as dissimilarity indices. This suggests that the geographic segregation observed for
income classes is further reinforced by social and economic structures that exhibit similar
segregation in cities. Decisions about the physical design of cities, especially land use and
zoning decisions, may serve to reinforce or disrupt the social and economic structures
therein. Certainly the history of urban renewal and automobility since the 1950s in the U.S.,
as discussed in Chapter 1, are consistent with this claim. Second, a relatively high R
2
value
(0.62) is achieved with only a handful of variables (eight in addition to the two income
variables) for race/ethnicity, educational attainment and employment skill, suggesting that
these factors may be of central importance to the development of cities with respect to social
goals. Decisions to invest in education and employment skill development may serve to
transform residents of cities, and therefore the cities themselves over the long term. Third,
the influence of race/ethnicity is either not significant (as in the case of black-white
segregation) or it is small (as in the case of Hispanic-white segregation) by comparison to the
influence of variables for education and employment skill. Finally, the presence of a
99
sustainability initiative (equation 4e) makes a small but statistically significant contribution,
and in the direction of decreasing income inequality. This last observation suggests possible
relevance of the sustainable development movement to the transformation of cities especially
toward social goals. To the extent policies in sustainable cities promote certain social
programs (e.g. education and vocational training), observations are generally consistent
Kuznetsian development and the nonlinear course development takes.

Overall, the effect of adding various structural variables listed in Table 6 is a flattening of
ascending phases of modeled curves describing poor and affluent class shares in cities, and a
flattening of both ascending and descending phases of curves describing middle class shares
(Figure 5). From Figure 6, structural variables also decrease overall income inequality by
comparison to models that do not include additional structural variables. One interpretation is
that policies and programs that target social structures related to these variables may be
responsible for these effects. One might expect that policies supporting local schools and job
skill development, for example, increase the share of middle class in cities, especially in
cities that appear to be in earlier stages of development. Policies or conditions that favor
living near work, or address concerns for housing affordability, may also contribute to these
effects. Furthermore, from Table 7 and Figure 6, policies that focus on education and job
skill development, and others that may be included in various sustainability initiatives,
appear to hasten the occurrence of the turning point of the Kuznets Curve, such that progress
towards decreased income inequality begins to occur at much lower average income levels in
U.S. cities than would otherwise occur without such policies.

100
Together, results provide evidence that is relevant to the first set of questions addressed in
this study. There is in fact a relationship between income inequality and economic growth in
U.S. cities, namely an inverted U-shaped curve in which income inequality first increases
with economic growth then decreases. This relationship is most clearly observed in models of
middle class share and poverty-affluence dissimilarity, which consider average household
income and also education level, employment skill and commitment by cities to policies of
sustainable development.

Kuznetsian Development and Sustainable Cities
That the presence of a sustainability initiative makes a small but statistically significant
contribution to models presented in this study is relevant to the second set of questions in this
study. Kuznetsian development may indeed provide a suitable theoretical window from
which to observe at least some of the many issues relevant to planning for sustainable
development in cities. Note that the Kuznets Curve estimated in this study describes an
average path defined by measures of income inequality and growth for all U.S. cities in 2000.
Points for individual cities almost never lie on the curve itself. Instead some points may lie
above or below the modeled curve, or before or after its turn. This study proposes that
classifying sustainable cities by their locus on the Kuznets Curve can inform our thinking on
development in cities, and further direct our attention to instructive sustainable cities
experiences.

Figure 7 indicates that sustainable cities in the U.S. generally lie above the Kuznets Curve.
This is curious since the regression coefficient for the model that considers sustainability
101
initiatives (equation 4e, variable 50, Table 7) is negative, which suggests that the influence of
sustainability planning is in the direction of decreasing income inequality. This observation
may be explained in two parts. First, sustainable cities as a group may be cities that face
considerable challenges in reconciling social and economic goals, more so than other cities.
Second, these cities may have turned to consider policies of sustainable development to face
these challenges, but for now sustainability remains more an aspiration than an
accomplishment. Indeed results presented in Table 8 give an overall impression that
sustainable cities—by comparison to other cities—describe mostly cities with a large labor
force of workers who may be struggling to achieve economic goals despite relatively high
educational attainment and job skill. This is indicated in these cities by their relatively greater
shares of non-white, working age persons, including immigrants, who are generally less able
to buy a home, face relatively greater rent burden, many of whom are dependent upon public
transit to get to work, despite many of them also being college educated and possessing
considerable skill.

A three-category typology of cities directs us to consider the varied experiences of
sustainable cities along their development course. According to Table 9, five sustainable
cities were classified in this study as Type A. Type A sustainable cities are characterized as
having achieved a lower level of economic development. Eleven sustainable cities were
categorized as Type B, occupying a more transitional stage of development, with several of
them—Tampa, Oklahoma City, and Indianapolis demonstrating particularly high income
inequality with economic growth (see Table A5 in the Appendix). Thirteen sustainable cities
were classified as Type C. Type C cities are characterized as having achieved highest levels
102
of economic development among sustainable cities. Interestingly, from Figure 8, most
sustainable cities have been heading in the direction of achieving greater income equality
over the 1980-2000 period, as indicated by the general downward trend observed among the
paths tracing these cities’ development courses. That is, most sustainable cities appear to be
in the process of “making the turn.” That the foray by cities into sustainable development
began largely in the early 1990s suggests that related policies of these cities may have
contributed to the general downward trend in income inequality observed over the 1980-2000
period for these cities. Advocates of sustainable development may find this interpretation
heartening.

But the apparent success of sustainable cities in decreasing inequality appears also to
diminish among many sustainable cities that have experienced greatest economic growth.
Figure 8 shows that Type C sustainable cities reflect a much wider range of experiences than
Type A or Type B cities since 1980, and with respect to the average path of U.S. cities, some
of these cities appear to be further “off course” than others in the pursuit of social goals of
sustainable development. Type C sustainable cities may be particularly important for
sustainable cities research; that is, if policy change is necessarily shaped by the structures
created by, and creating, the experiences of these cities, the variation of experiences of Type
C cities may hold greatest promise for further study. These cities may also provide greatest
insight for practice. Type C sustainable cities may be particularly instructive for planners in
less developed cities whose visions for further development match the successes seen in
some of these cities, while their concerns may have been realized in others. For example,
Cambridge and Santa Monica, despite their relatively high poverty-affluence dissimilarity
103
appear also to be in the process of “making the turn” toward greater equality of income. By
comparison, cities such as Phoenix and Austin appear to be continuing down a path of high
income inequality with further growth. Type C sustainable cities may provide greatest insight
into the conditions and policies that explain different development paths.

To summarize, results demonstrate that U.S. cities that have committed to sustainable
development, so-called sustainable cities, do indeed reveal an interesting relationship to the
Kuznets Curve, with most positioned above it (but some below), and some before and some
after its turn. Sustainable cities that have achieved highest levels of economic development
demonstrate a substantial variety of social outcomes, with some making and others missing
the turn toward decreased income inequality. A typology of sustainable cities in light of
Kuznetsian development of cities shows promise for directing further analysis and theorizing.

Sustainable Cities and the Urban Regime
Ahluwalia admonishes researchers about the relationships observed from cross-sectional
analyses. Such relationships “…do not establish causality. They are better described as
‘stylized facts,’ which can be observed but which then need to be explained by an appropriate
theory” (Ahluwalia 1976b, p.129). The remainder of this discussion seeks to build from clues
from the “stylized facts” presented previously to explain some patterns observed from the
experiences of U.S. sustainable cities.

Recall that Acemoglu and Robinson (2002, p.183) present a political economy explanation of
the Kuznets Curve. Accordingly, the turning point depends upon the political mobilization of
104
lower and middle classes ultimately to “force democratization upon political elites.” But two
outcomes are possible according to this mechanism. One outcome reflects lower inequality
and higher economic growth, in other words the “wishful thinking” of Kuznets (1955), while
the other describes higher inequality and lower growth, or “autocratic disaster” (Acemoglu
and Robinson 2002). Sustainable cities considered in this study are listed by type in Table 9
and the temporal “cloud chamber” view (Figure 8) provides additional insight into the paths
followed by these cities. Figure 8 suggests that sustainable cities that have experienced
higher levels of economic growth (i.e. Type C sustainable cities) comprise an interesting
sample of cities to consider for further insights into human agency in sustainable cities. Cities
such as Cambridge and Santa Monica for example may be characterized by outcomes that
reflect “wishful thinking” while outcomes for Phoenix and Austin approach those nearer
“autocratic disaster.” A focus of this study is upon the possible influence of lower and middle
classes upon the governing elite (i.e. the political mobilization thesis) to explain these
alternative outcomes.

To investigate further, income class groups were modeled by agglomerative hierarchical
clustering of tract-level Census data. Figures 10 and 11 present results for Santa Monica.
Results suggest that the existence of groups differentiated by income class, culture and
neighborhoods in this city is plausible. Furthermore results suggest that Santa Monica’s elite
live in one of two adjacent neighborhoods north of Wilshire Boulevard, while the city’s poor
reside in neighborhoods nearest the city’s center, further divided by the Santa Monica
Freeway. That results are consistent with maps of known neighborhood associations suggest
105
that clustering has done more than to simply identify groups based on class and culture, but
has indeed identified groups that are also organized politically.

Furthermore, public records for Santa Monica for 2000 indicate that city council members
were largely found within or immediately adjacent to neighborhoods of the wealthiest
homeowners in the city. These findings are relevant to growth machine theory (Molotch
1976), which suggests that city policy results from the alignment of “development regime”
and “maintenance regime” (Stone 1986) where these regimes represent largely the interests
of a land-based elite class and local officials. No such alignment involving middle- or lower-
class groups is apparent. Results shed some light on the third set of research questions in this
study, suggesting that cultural factors account for discernible differences among residents of
sustainable cities, especially between those describing elite and non-elite groups.

The final set of questions of this study concern whether groups act as human agents in the
context of sustainable cities and, according to the political mobilization thesis, informal
regime associations account for various development outcomes of sustainable cities. Another
thesis—the public participation thesis—suggests that formal efforts by cities to involve
citizens in planning also contributes to development outcomes. Recall from Figure 2, two
possible theoretical explanations for development outcomes, one following from traditional
development thinking, and another following from ideas of sustainable development. Under
traditional development, the exclusion of lower class regimes from city governance translates
to this segment of the population being underserved, i.e. amenities and opportunities not
provided to them, thus contributing to greater inequality between elite and lower classes.
106
Both lower political mobilization and lower public participation could account for this result.
Judging from the results described above, it appears that neighborhoods comprising the
development regime and the city maintenance regime are allied for political battles in Santa
Monica (owing to their shared geography), potentially giving wealthier residents an
advantage over poorer residents in the city in representing their interests in city policy
decisions. Figure 2 suggests that outcomes of increasing income inequality with economic
growth may be expected under the traditional mode of development, especially where the
political alignment between development regime and city maintenance regime is strong. In
the context of sustainable development, however, the influence of lower class groups on city
governance translates to policy decisions that favor greater equality of income over the
course of city development, owing to political mobilization and/or public participation in
governance. This study provides some evidence that at least one of these influences may be
present in Santa Monica.

With regard to the political mobilization thesis, this study explores the possibility that
development outcomes result from either the efforts of regimes to “marshal local resources
and coordinate local institutional elites for the purpose of pursuing traditional economic
growth and development” (Portney 2003, pp.26-27), or instead to “focus on such measures as
environmental protection, historic preservation, affordable housing, the quality of urban
design, affirmative action, and linkage funds for various social purposes” (Stone 1993, in
Portney 2003, p.27). The former describes primary interests of the “development regime”
while the latter describes that of the “middle class progressive regime” (Stone 1986).
According to Portney, the middle class progressive regime supports a position that is likely
107
most consistent overall with sustainable development thinking. Were efforts by non-elite
groups effective in shaping policies of sustainable cities, development outcomes would be
expected to be reflected by these cities having successfully navigated the turn on the Kuznets
Curve. Furthermore, one would expect to see the middle class regime to figure prominently
in explaining these outcomes.

Indeed Kuznetsian models suggest that middle class shares are greatest near the turning point
(Figures 4 and 5) for U.S. cities overall, suggesting that, at least according to their number,
the middle class regime may represent an influential group in cities. Curiously, however,
cities with sustainability initiatives are observed to have significantly smaller middle class
shares by comparison to other U.S. cities, and several Type C sustainable cities have missed
the turn towards greater equality of income entirely. On the one hand, this observation is
consistent with the notion that sustainable cities are troubled cities that face considerable
challenges in their development, compounded further by relatively fewer citizens capable of
promoting a more progressive policy agenda. But if sustainable development constitutes a
project of the middle class, as Portney suggests, how is it that sustainability initiatives exist at
all in cities where the middle class is significantly diminished in size? One answer may be
that sustainable development in these cities is not a project of a progressive middle class, but
owing to their relatively greater number—especially in Type C sustainable cities—a project
of the elite class. If sustainable development is indeed a project of the elite class, one might
expect that policies are conceived with much less emphasis on social goals of sustainability,
and possibly greater emphasis on economic or environmental goals. The greater income
inequality with economic growth observed in cities such as Phoenix and Austin (Figure 8) is
108
consistent with this proposition. A greater leaning towards environmental interests over those
of individuals, or possibly environmental protectionism that equates to policies that favor
property interests of the elite, is not fully explored in this study but is presented by Frieden
(1976). In his study of several northern California counties, environmental protectionism has
served to limit growth, increase housing prices, and exclude lower income classes from many
communities. Where development does occur, so-called “green building” requirements and
incentives endorsed by the governing elite create new opportunities enjoyed by select
developers, such that wealth accrues to a select few. One might describe the conditions
created in such settings as those of a “green growth machine.”

Another interpretation of the relatively lower middle class shares observed in sustainable
cities may be that despite their diminished size, their influence upon city governance remains
significant. That is, the existence of sustainability initiatives in these cities is in fact
motivated by increasing environmental and social concerns, and a potent middle class
progressive regime is both present and active in these cities. Cities such as Cambridge and
Santa Monica may be notable examples to consider here. This perspective is more consistent
with the political mobilization thesis, not to mention the primary motivation behind the
sustainable development, as explained by Harris and Goodwin (2001). Perhaps both
explanations of middle class shares are reflected in the experience of U.S. sustainable cities.
The varied outcomes presented in Figure 8 suggest that many factors are likely at play, and
multiple explanations are possible.

109
Results of this study allow us to imagine further how regime associations may influence
governance structures in sustainable cities, possibly to explain their various positions along
the Kuznets Curve. Figure 12 summarizes these associations by city type for U.S. sustainable
cities. Table A5 in the appendix presents correlation coefficients of regime associations for
all sustainable cities considered in this study. These variables model the cultural similarity of
two groups according to the Spearman correlation coefficient, considering the community
indicators listed in Table 3. Cultural similarity is meant to represent shared beliefs and
interests, and therefore provide basis for political organization and mobilization to influence
policy outcomes that favor these interests. Focusing especially upon Type C sustainable
cities, a characteristic “low, high, high” pattern is apparent for those that have made the turn
toward greater equality of income. That is, cities such as Cambridge, Santa Monica, San
Francisco and Annapolis demonstrate low correlation between elite and lower classes, high
correlation between elite and middle classes, and high correlation between middle and lower
classes. The pattern suggests a story that is consistent with the political mobilization thesis.
That is, a tension between elite and lower classes (as indicated by low correlation between
the two) motivates interest by non-elite groups to promote policies of sustainable
development (in the words of Acemoglu and Robinson [2002] to “force democratization
upon the elites.”) The pattern suggests further that the tension is resolved in planning
processes through efforts of a middle class regime. The mechanism by which this occurs is in
this regime’s dual association with elite and lower classes and ability to bridge cultural
differences between elite and lower classes, mediate differences in interests, and thereby
promote political consensus in policy decisions made by the governing elite. This bridging
and mediating capacity is indicated by relatively higher correlations between both elite and
110
middle class, and also higher correlation between middle and lower class. An interpretation is
that members of the middle class in sustainable cities are both sympathetic to the issues faced
by the lower class while at the same time, hold great sway over elites, and therefore
government officials. The capacity for dual association with the middle class translates to
city policy that better responds to broader social concerns, ultimately to decrease income
inequality with further economic growth.

To test these ideas of policy change further, several regression models were constructed to
explore the statistical relationships between income inequality, regime associations, and the
involvement of various specific stakeholder types in the development of the sustainability
initiative for their cities. Results are summarized in Table 10. The model represented by
equation 5c is found to explain quite well poverty-affluence dissimilarity according to regime
associations and public involvement in advanced sustainable cities overall (R
2
= 0.83; p =
0.08, F-test). In particular, regime associations contribute substantially to this model,
although coefficients for individual associations are not statistically significant at the 10
percent level. Equation 5c also includes variables for the involvement of two stakeholder
groups—political and business leaders of cities, and the general public. Of all the individual
factors that explain income inequality in Type C sustainable cities, only the direct
involvement of the general public contributes significantly to these models, and the
contribution is in the direction of decreased income inequality. The implication is that efforts
to involve the general public explain best the relatively greater progress along the Kuznets
Curve, especially in sustainable cities that have achieved highest levels of economic
development.
111

The regression results are reinforced by results presented in Figure 8 for cities such as
Cambridge and Santa Monica, by comparison to cities of Phoenix and Austin. In Phoenix and
Austin, Portney (2003) reports that the general public was largely excluded from the planning
process for the sustainability initiative. Accordingly, income inequality remained high or
increased during the 1980-2000 period. In Cambridge and Santa Monica, where income
inequality instead decreased during that same period, the public was reported by Portney to
be directly involved in these cities’ sustainability initiatives. The implication is that whatever
the associations and efforts by regimes, efforts by city officials to provide a public forum for
members of all income classes and interest groups to be heard directly by city leaders may
matter most for achieving social goals of sustainable development in cities.

Thus, responses to the fourth set of questions addressed in this study are mixed. Despite a
relatively greater dissociation between elite and lower class regimes in some of the wealthier
sustainable city examples, there appears to be no statistically significant effect of this
dissociation that explains income inequality in these cities. More generally, informal regime
associations do not influence patterns of income inequality in sustainable cities with
statistical significance, in spite of the pattern of low association between elite-lower, high
association between elite-middle, and high association between middle-lower class regimes,
that emerges among sustainable cities that have achieved the highest levels of economic
growth and also the lowest levels of income inequality.

112
As in Kuznets’ original work, the evidence provided by this study in support of either
political mobilization or public participation (or both) in a manner that reflects promise for
sustainable development in cities is “perhaps 5 per cent empirical information and 95 per
cent speculation, some of it possibly tainted by wishful thinking.” Kuznets himself was
careful not to present his findings with great confidence but sincerely hoped that “little harm
and much good” would come of his “collection of hunches” (Kuznets 1955, p. 27). Results
especially for the final questions addressed in this study are presented in the same spirit and
with great hope that they may inspire further, meaningful research on sustainable cities.
While this study has not produced strong support for the political mobilization thesis, this
study has however, presented statistically significant evidence in favor of the public
participation thesis. Accordingly, efforts by city leaders to involve the general public in its
sustainability plans can have important effect on achieving social goals of sustainable
development. And, the “low, high, high” pattern of regime associations in more advanced
sustainable cities, while not statistically significant in its ability to explain patterns of income
inequality, persists as basis of a strong hunch that the middle class progressive regime plays
an important role in bridging cultural differences among groups, mediating differences in
policy interests, and helping sustainable cities to achieve social goals.

113
CONCLUSION

This study contributes to a growing body of literature on sustainable development, reporting
in particular on the experience of cities in the U.S. that have joined the sustainable
development movement. As an alternative approach to development, the pursuit of
sustainable development has been motivated by the effect of traditional development
approaches of creating social inequities and environmental damages that, as advocates of the
movement promote, may be addressed by policy reform. This study employs data from the
U.S. Census Bureau, methods such as regression analysis and cluster analysis as well as
theories of development and the political economy of cities to account for the wide variety of
local implementations of sustainable development outcomes observed in U.S. cities. An
important aspect of this study is its interpretation of city experiences as the interplay of
structure and agency. Specifically, this study attempts to infer from observable social and
economic structural differences among U.S. cities evidence of human agency that may
explain different outcomes, and to make some plausible statements about structure, agency
and sustainable development in cities that warrant further research and consideration in
policymaking.

A feature of this study is its application of the empirical and theoretical work of Simon
Kuznets, who made important observations about the experiences of developing countries, to
the contexts of cities. The Kuznets Curve stands as an important artifact among data
describing income inequality and economic growth, directing researchers to consider
underlying aspects of structure and agency. As in previous research, this study first sought to
114
confirm that such a curve exists. Results from a series of regression models using 2000
Census data for U.S. cities provided evidence that the classic Kuznets relationship does
indeed exist among these data. In this way Kuznetsian development appears to be generally
applicable to the experience of U.S. cities. Several additional factors such as educational
attainment, employment skill, and segregation of neighborhoods emerge as especially
important contributors to the modeled relationship. In light of normative aspects of the curve,
embodied by the “wishful thinking” of Kuznets himself that economic development should
eventually arrive at social goals including greater equality of income among members of
society, this study indicates that while a number of sustainable cities have appear to have
worked diligently toward this goal, several of the more affluent cities considered here have
“missed the turn” toward greater equality of income among their residents. The main
contribution of this study to the literature on sustainable cities is in demonstrating how the
Kuznets Curve can be used to explore relationships among data for developing cities and to
identify potentially interesting cases for further study. Below are several ideas on how the
ideas in this study can be expanded further.

Directions for Future Research

Campbell and Fainstein (1996, p.1) claim that the central question of planning theory is
“What role can planning play in developing the good city and region within the constraints of
a capitalist political economy and a democratic political system?” Innes (1998, p.vi) claims
that planning theory in recent years

115
more often tells us in a nuanced way how practice of various kinds has
worked, permitting readers to draw their own lessons for their own situations.
Planning theory is now much more about helping planners see themselves and
what they do than it is about providing prescriptions.

One hope is that this study provides insight on how to reveal the lessons Innes describes, and
if not to help planners see themselves, perhaps at least to help them see the plans they
produce in new and useful light. As Brooks (2002) notes, “planning does not occur in a
vacuum, but in a social, political, and economic context.” The ability to understand
sustainability planning especially must seek to better understand not only plans, but the
context in which they occur. The application of urban theory to sustainable development that
has more recently emerged in the sustainable cities literature appears to hold promise in this
regard, and this study has attempted to promote that direction of the field. The idea of
structuration appears to be a useful device for contemplating the dynamics and the evolving
order of cities that pursue sustainable development. And the Kuznets Curve seems a useful
construct for further directing our attention to aspects of structure and agency. With these in
mind, this study demonstrated several means by which to inform and organize thinking and
to better understand the context in which planning for sustainable development in cities
occurs.

That cities in the U.S., on average, generally achieve greater equality of income with further
economic development is a finding not previously reported, and thus a potentially important
contribution to the planning literature. It is also evidence that Kuznetsian development may
be relevant to our thinking about cities. That those cities which have committed to policies of
sustainable development appear in some cases to be able to accelerate Kuznetsian
116
development, thus achieving greater equality at lower levels of economic growth, may be
encouraging news for advocates of sustainable development. However, it is important to note
that given the methods employed in this study, these claims will require further
substantiation—an important project of future research. Regression analysis of cross-
sectional data do not, and cannot, establish causation. Nonetheless, results do suggest several
hypotheses about the causal mechanisms at work in sustainable cities that deserve further
attention. In particular, this study has provided some insight into the explanation for the
acceleration of Kuznetsian development, namely by a possible influence of education and
employment skill level on development processes, neither of which appear to be areas of
active pursuit by sustainable cities researchers, nor foci of sustainable city policy. Sustainable
development planners and scholars may do well to consider the specific contributions
education and employment make in cities. A much richer set of variables describing
education and employment circumstances in models of Kuznetsian development may help to
isolate specific factors that explain income inequality in cities. The mechanisms by which
education and employment can contribute to the observed decrease in income inequality with
city development deserve much closer attention. Results could inform policy decisions
regarding education and employment in cities and regions.

Another contribution this study has made is in its classification of cities by Kuznetsian type.
Of course, grouping objects by type is not new to research. And, according to Verma (1996),
planning for sustainable development (and planning practice more generally [Verma 1995])
is largely a search for suitable categories. But doing so for sustainable cities is made more
challenging by the dynamics and multiple issues of concern that are involved. Classifying
117
cities by Kuznetsian type is suggested to allow the researcher to “pause” the ongoing course
of structuration of cities and direct attention to instances of human agency. This approach
may hold great promise for informing case study research designs that can uncover important
subtleties about cities that cannot be read from gross measures that seem to puzzle
sustainable cities researchers. As Portney (2003, p.241) reflects,

…indeed the cities that seem to be taking sustainability seriously differ greatly
from one another. Some are large, and some are relatively small. Some
experienced very rapid population growth, and some much less so. Some
faced significant environmental problems, and some have not. Some are very
densely populated areas, and some are less so. Some are on the west coast, but
many are not. Some are politically liberal places, and some are pretty
conservative. In short, the broad-brush characteristics do not seem to
characterize very well the cities that take sustainability seriously.

The categorization of sustainable cities by income levels, and further differentiation by their
trends in income inequality for example helped to call attention to some possible underlying
explanations which are linked to urban political theory. Can further research in this direction
help to clarify the manner in which sustainable cities differentiate among one another?

Categorizing sustainable cities by Kuznetsian type may provide empiric grounding for work
following Verma (1995, 1996) who notes that categories can be either structural or
purposeful. The typology presented in this study is both. It is structural in that it informs with
analytic rigor objects for further analysis based on economic and social characteristics. It is
purposeful in that it focuses attention on outcomes in cities that should be aspired to, and also
those which might be avoided by deliberate action of stakeholders. A Kuznetsian framework
118
could be fruitfully applied to both normative theory development and also to study the
experiences of a growing number of sustainable cities.

This study reflects the observation by Innes (1995, p.183) that planning researchers
increasingly take “practice as the raw material of their inquiry,” while at the same time
suggests that research on sustainable cities is in need of some new raw material. Portland,
Boulder, Santa Monica, and so forth, are likely to continue to inform sustainable cities
research for many years as their experience continues to unfold. But continuing to describe
the experiences of these cities (and few others) is of limited value to those who, as Innes
describes, seek to “draw their own lessons for their own situations.” (Innes 1998, p.vi) Do
Portland, Boulder, and Santa Monica provide blueprints relevant to the development of other
U.S. cities? The small sample size employed in this study, and others, presents an important
challenge to sustainable cities research and the methods available to sustainable cities
researchers. The typology presented may help planners and researchers know where to look
for new research material and how to select from it for further study. If we cannot gather
detailed information on all cities, can we know where to begin? Do some cities appear to
hold more promise in revealing important trends and variations? What additional Type C
cities are available for sustainable cities research? Are some of these cities experimenting
with policies of sustainable development? Do these experiences describe patterns of making
or missing the turn towards sustainable development? Who has been involved in making
these policies in these cities and how can we characterize them? Adding to the sample of
sustainable cities used in research would also serve purposes of better analytic generalization,
as well as the statistical generalization that policy audiences often require. The growing
119
membership in ICLEI, while not used as a sample in this study—or others—appears to hold
promise for future sustainable cities research. Including ICLEI members would increase the
sample size of known sustainable cities used in research by an order of magnitude. An
important aspect of this group may also be that few of them have actually adopted
sustainability initiatives, and may therefore be able to provide insight into the transformation
of cities along a decision path towards greater commitment to sustainable development—a
path that may also reflect the Kuznets relationship.

The specification of the Kuznets Curves developed for this study might be revisited to
consider several analytic enhancements. For example, future specifications might consider
cost-of-living differences among cities. Doing so may rearrange sustainable cities along the
Kuznets Curve in important ways, thus directing attention to different examples than were
highlighted in this study. Also, the temporal animation presented in Figure 8 might be
extended to consider 2010 Census data (which was not yet available for this study). Given
that the earliest sustainable city initiatives dated only to the 1990s, giving little time for
sustainable city policies to be reflected in the structures of cities reported in this study, 2010
data may depict a more accurate picture of the outcomes related to sustainable development
policies.

Future research might also consider more deeply the results of this study in light of others
who have sought to broadly compare sustainable cities. Table 9 provides both a
categorization of sustainable cities by their Kuznetsian type and also their overall scores
according to Portney’s “Taking Sustainability Seriously” index (Portney 2003). Recall that
120
Portney reviewed 31 U.S. cities having recognized sustainability initiatives, creating his
index based on the presence or absence of program elements in seven categories that
encompassed 34 program types. From Table 9, sustainable cities which appear to be in early
or transitional stages of development (Type A and Type B cities) have generally lower scores
on Portney’s index, whereas Type C sustainable cities have generally higher scores. Do
sustainable cities that appear to reflect more advanced stages of economic development take
sustainability more seriously? It appears so, according to Portney. However, the range of
outcomes for income inequality among Type C cities identified in this study suggest that the
seriousness measured by Portney does not necessarily reflect success in the pursuit of social
goals of sustainable development. Indeed none of the elements studied by Portney indicate
commitment to social programs (with the exception of inner city public transit.) Given the
relatively greater resources of wealthier cities (as indicated by higher average household
incomes, and therefore greater tax base) it seems likely that Type C sustainable cities are in a
better position to support social programs in general. Why then do we see such wide
variation in income inequality among these cities? Do some of these cities (e.g. Tampa,
Phoenix, and Austin) purposely commit less to programs that may achieve social goals of
sustainable development, such as education and job skills development aimed at providing
greater opportunity for lower income groups? This study has not provided evidence for or
against such a claim, but results of this study have suggested the issue deserves a closer look.
It also appears that sustainability plans can score reasonably well on measures of
environmental sustainability, for example by having programs that provide tax incentives for
environmentally friendly development as in Austin, or zoning to delineate environmentally
sensitive areas in Phoenix, while at the same time performing poorly on social dimensions of
121
sustainable development. Importantly, communities do not require public involvement to
adopt and implement sustainable city plans. Plans can just as easily (perhaps more easily) be
adopted if they are embraced by a governing elite that controls the planning process (as
Portney reports is the case in these two cities.) However, where social goals of sustainable
development are being more successfully achieved, this study has suggested that public
involvement in the planning process is an important contributor. By not including the public
in sustainability planning, plans run the risk of not “approximating the public interest” (Innes
1996, p.469), and may instead protect interests of a wealthy land-based elite in sustainable
cities. “Green growth machine” politics, in which elite interests are couched in policies of
environmental protection deserves continued attention in future sustainable cities research.

This study also has suggested that a middle-class progressive regime is present and culturally
resembles both elite and lower classes in many sustainable cities, and because of this duality
is more likely to engage both of these groups based on shared interests and concerns resulting
in policy that is likely to be more socially centered than would otherwise occur. While
statistical evidence for this idea was not found in this study, it remains an idea worth
pursuing. In this study, models of income inequality that included regime associations as
explanatory variables fit data describing Type C sustainable cities well, and coefficients for
regime variables contributed substantially to these models. These same models may achieve
more compelling results with larger samples of cities. What factors contribute most to the
“low, high, high” pattern uncovered in this study, especially for cities which appear to be
making the turn toward greater equality of income? Cultural resemblances, interests and
concerns of these groups could be investigated more deeply, and new model specifications
122
could be tried. How do cultural resemblances among groups translate to the planning process
in sustainable cities? Do these predictably explain alliances among groups that can influence
governance in these cities? Each of these questions could be addressed in further research.

Another project for future research may be to attach more meaningful labels to the three
types identified by the Kuznetsian typology presented in this study, and by doing so enable
further exploration into development processes in cities. Haughton’s sustainable city types,
for example, might be reconsidered in light of results of this study. Do Haughton’s self-
reliant cities map to one of the types identified here, such as Type A? Externally dependent
cities as Type B? Redesigning cities and fair shares cities as variants of Type C? If so,
Haughton’s sustainable city types may be related by a series of decision paths that describe
sustainable city development as Kuznetsian development.

Finally, just as Kuznets original work (Kuznets 1955) inspired a search for a similar inverted
U-shaped relationship among various indicators of environmental degradation and income
levels among developing countries (Grossman and Krueger 1991; IBRD 1992), an analogous
search for an Environmental Kuznets Curve (EKC) among data for U.S. cities seems an
interesting possibility for future research. The integration of both the classic Kuznets Curve
describing social concerns, alongside an EKC describing various environmental concerns
may provide a more comprehensive view of the structures, policies and underlying politics
of sustainable cities over their development course.

123
Implications for Social and Environmental Policy
Development of a city given “constraints of a capitalist political economy and a democratic
political system” (Campbell and Fainstein 1996, p.1) is likely to be challenged if the “take
off” that Rostow describes leaves some members of society behind. This study provides a
snapshot of U.S. cities engaged in their various takeoffs and landings, and also some insight
into how sustainable development policies and policymaking in cities may contribute to each
of these.

A typology of sustainable cities based on the Kuznets Curve may be useful for comparative
policy analysis, as it can frame thinking especially on sustainable cities which have reached
more advanced stages of development and whose initiatives might be emulated, and also
those that should not be followed. Continued attention to the experiences of cities and novel
programs that are responsible for social and environmental achievements, and encouraging
other cities to adapt the most promising among them to their own settings, seems a
reasonable pursuit for practice, and a goal for policymakers. Identifying cities with methods
inspired by this study, and encouraging best practices that may be found from them, may be
particularly relevant to the collective missions of the U.S. Environmental Protection Agency,
U.S. Department of Housing and Urban Development and the U. S. Department of
Transportation, which, as described in the Introduction, recently formed an interagency
partnership for the purpose of pursuing goals related to sustainable development.

A challenge however for state and federal policymakers is in providing support and guidance
in a manner that does not follow the “high modernist ideology” that Scott (1998) warns
124
against. Mazmanian and Kraft (2001) too remind that we are well past a period dominated by
command-control regulatory policymaking by the state, and now well into a period better
characterized by partnerships, local/regional collaborations and consensus building
approaches that arrive at policy. State and federal policymakers increasingly find themselves
in a role of providing guidance and support, largely consigning cities to that of decision
making authority. Yet, the local perspective on global concerns is often like that from a tree
in a forest. And because, as this study has suggested, sustainable cities are largely troubled
cities that have experienced (or are experiencing) generally higher inequality with economic
growth, the local view on some policy concerns—especially regional environmental policy—
is perhaps obscured further. State and federal policymakers may be in a better position to
guide longer term decisions concerning common pool resources, such as environmental
resources, which as Levett urges are preconditions for all else. This study provides some
means by which state and federal policymakers might identify cities where social problems
are most acute, and to help them guide city leaders toward more comprehensive policy based
on insights learned from sustainable development experiences elsewhere. The temporal view
of sustainable cities’ experiences depicted in Figure 8 may prove useful in this regard.

This study has further suggested that reversing conditions of inequality may in some cases
involve a middle-class progressive regime in city planning. This group may be an important
focus of state and federal policymakers to consider more deeply. Culturally, this regime
occupies a middle-ground between powerful elites that are believed to control city planning,
and also the lower class, whose many interests such as affordable housing, transportation
accessibility, and so forth, are held in common with progressives and are also the foci of
125
government agencies (e.g. DOT, HUD). Various interests of elites and progressives,
including environmental interests, are also likely to be held in common as they may also be
the focus of government agencies such as the EPA. This study suggests that the middle class
progressive regime bridges a crucial divide among neighborhood groups of cities, and is
potentially in a position help negotiate agreement on city plans that may serve broader
interests. But it is possible that middle class interests in some cities may not be progressive at
all, and the data hint at a possible exodus of middle class citizens from larger, more
sustainable to smaller, less sustainable U.S. cities. Such a trend is consistent with that
observed in the suburbanization of the U.S. since the 1950s. Whereas earlier periods of
surbanization were believed to be motivated by desegregation policies and the coming of age
of the automobile, this study suggests that its continuation in U.S. sustainable cities may be
explained by lack of affordable housing and “green growth machine” politics such that the
middle class is systematically excluded, along with the lower income class, from city
governance. The implication is that this potentially influential group turns instead to consider
that which is available to them outside sustainable cities, namely the suburban areas of cities
that do not commit to sustainable development but offer instead more affordable housing and
higher quality of life. Consistent with this, the National Association of Homebuilders reports
that the majority of homebuyers continue to prefer suburban options for new homes, and that
these homes should be larger and adjacent to parks, playgrounds, trails, open space, and
water (NAHB 2002). Since the 1950s especially, homebuilders have generally tended to
supply this demand, building new homes on undeveloped parcels further and further from
urban centers. An implication loosely interpreted from results of this study is that state and
federal policy makers might slow this trend by seeking to encourage the middle class to
126
remain in more urban areas and to continue to support efforts that promote principles of
sustainable development in cities. This might be achieved by helping them to have their
voice heard in local politics and to encourage efforts to build more attractive cities that
provide more of the amenities they seek elsewhere. Government grant programs, technical
support, and awards for exemplary professional planning efforts might be refined to consider
more fully the potential for the progressive middle class in cities to effect desired social and
environmental policy that promotes housing affordability, transportation access, and
environmental protection. State and federal policy makers might also take greater interest in
location efficient mortgage programs which provide reduced rate loans to individuals who
live close to their work, selective investment in educational systems and parks of cities that
commit to policies of sustainable development, and tax advantages that make living and
working in these cities more attractive. Incentives for homebuilders to build a greater variety
of home products, including smaller, more affordable homes within sustainable cities may
also help ensure that progressives remain a vital part of sustainable cities such that they are
growing, just and green.
127
BIBLIOGRAPHY

Acemoglu, D. and J.A. Robinson. 2002. “The Political Economy of the Kuznets Curve.”
Review of Development Economics. 6(2):183-203.
Ahluwalia, M.S. 1976a. Inequality, Poverty and Development. World Bank. Washington,
D.C.
Ahluwalia, M.S. 1976b. “Income Distribution and Development: Some Stylized Facts.”
American Economic Review. 66(2):128-135.
Aiken, W. and H. LaFolette. 1996. World Hunger and Morality, 2
nd
Edition. Prentice Hall.
Upper Saddle River, New Jersey.
Altshuler, A. and D. Luberoff (eds.) 2003. Mega-Projects: The Changing Politics of Urban
Public Investments. Brookings Institution Press, Washington, D.C.
Altshuler. A. 1965. “The Intercity Freeway.” In A.A. Altshuler. The City Planning Process.
Cornell University Press. Ithaca, New York.
Anand, S. and S.M.R. Kanbur. 1993. “The Kuznets Process and the Inequality-Development
Relationship.” Journal of Development Economics. 40:25-52.
Armor, D.J. 1995. Forced Justice: School Desegregation and the Law. Oxford University
Press. Oxford.
Bailey, F.D. 1994. Typologies and Taxonomies: An Introduction to Classification
Techniques. Series 07-102: Quantitative Applications in the Social Sciences. Sage
Publications. Thousand Oaks, California.
Banerjee, T. and N. Verma. 2004. “Sprawl and Segregation: Another Side of the Los Angeles
Debate.” In Varady, D. (ed). Desegregating the City: Ghettos, Enclaves, and
Inequality. SUNY Press. Binghamton, New York.
Barr, N. 1998. The Economicis of the Welfare State, 3
rd
Edition. Stanford University Press.
Stanford, California.
Berke, P.R. and M.M. Conroy. 2000. “Are We Planning for Sustainable Development? An
Evaluation of 30 Comprehensive Plans.” Journal of the American Planning
Association. 66(1):21-33.
Berman, M. 1982. All That is Solid Melts Into Air: The Experience of Modernity. Penguin
Books. New York.
Bernard, P., R. Charafeddine, K.L. Frohlich, M. Daniel, Y. Kestens. L. Potvin. 2007. “Health
Inequalities and Place: A Theoretical Conception of Neighborhood.” Social Science
& Medicine. 65:1839-2852.
128
Black, A. 1990. “The Chicago Area Transportation Study: A Case Study of Rational
Planning.” Journal of Planning Education Research. 10(1):
Bourdieu, P. 1977. Outline of a Theory of Practice. Cambridge University Press. London.
Bourdieu, P. 1990. The Logic of Practice. Polity Press. Cambridge.
Bourdieu, P. and L.J.D. Wacquant. 1992. An Invitation to Reflexive Sociology. University of
Chicago Press. Chicago.
Brooks, M.P. 2002. Planning Theory for Practitioners. Planners Press. Chicago, Illinois.
Brugman, J. 1999. “Is There Method in Our Measurement? The Use of Indicators in Local
Sustainable Development Planning.” Local Environment. 20(1):59-72.
Campbell, S. 1996. “Green Cities, Growing Cities, Just Cities? Urban Planning and the
Contradictions of Sustainable Development.” Journal of the American Planning
Association. 62(3):296-312.
Campbell, S. and S.S. Fainstein. 1996. Readings in Planning Theory, 2
nd
Edition. Blackwell
Publishing. Malden, Massachussetts.
Chaves, E., E. Garcia, and S. Gilmore. 2006. “Consensus Building in Transportation
Planning Practice: Case Evaluation of Process and Outcomes.” Journal of the
Transportation Research Board. 1981:76-83.
Childe, V.G. 1936. Man Makes Himself. Watts & Co. London.
Childe, V.G. 1950. “The Urban Revolution.” Town Planning Review. 21:3-17.
City of Ventura, California. 2010. Green Ventura. http://cityofventura.net/greenventura.
Accessed October 29.
Common, M. and C. Perrings. 1992. “Towards and Ecological Economics of Sustainability.”
Ecological Economics. 6(1):7-34.
Conroy, M.M. and P.R. Berke. 2004. “What Makes a Good Sustainable Development Plan?
An Analysis of Factors That Influence Principles of Sustainable Development.”
Environment and Planning A. 36: 1381-1396.
Cutter, S. 1995. “Race, Class and Environmental Justice.” Progress in Human Geography.
19(1):111-122.
Daly, H. 1994. Beyond Growth: The Economics of Sustainable Development. Beacon Press.
Boston.
Davidoff, P. (1965). “Advocacy and Pluralism in Planning.” Journal of the American
Institute of Planners. 31(4):331-338.
129
Dear, M.J. and A.I. Moos. 1986. “Structuration Theory in Urban Analysis: 1. Theoretical
Exegesis.” Environment and Planning A. 18:231-252.
Deka, D. 2004. “Social and Environmental Justice Issues in Urban Transportation.” In
Hanson, S. and G. Giuliano (eds.) 2004. The Geography of Urban Transportation, 3
rd

Edition. The Guilford Press. New York.
Diamond, J. 2005. Collapse: How Societies Choose to Fail or Succeed. Viking Books. New
York.
Dryzek, J.S. 1995. “Political and Ecological Communication.” Environmental Politics. 4:13-
31.
Elkington, J.1994. “Towards the sustainable corporation: Win-win-win business strategies
for sustainable development.” California Management Review. 36(2):90-100.
Emirbayer, M. and A. Mische. 1998. “What is agency?” American Journal of Sociology
103(4):962–1023.
Ewing, R. and J. Kostyack. 2005. Endangered by Sprawl: How Runaway Development
Threatens America’s Wildlife. National Wildlife Federation, Smart Growth America,
Nature Serve.
Ewing, R.H. 1994. “Characteristics, Causes, and Effects of Sprawl: A Literature Review.”
Environmental and Urban Studies. 21(2):1-15.
Federal Register. 2010. Solicitation of Applications for the Research and Evaluation
Programs: FY 2010 Triple Bottom Line Accounting Competition—A Method for
Identifying Impacts of Economic Development Initiatives on Regional Economies.
July 10. http://www.federalregister.gov. Accessed October 29.
Ferman, B. 1996. Challenging the Growth Machine: Neighborhood Politics in Chicago and
Pittsburgh. University Press of Kansas. Lawrence, Kansas.
Flyvberg, B. 1998. Rationality & Power: Democracy in Practice. University of Chicago
Press. Chicago, Illinois.
Forman, R.T.T., D. Sperling, J.A. Bissonette, A.P. Clevenger, C.P. Cutshall, V.H. Dale, L.
Fahrig, R. France, C.R. Goldman, K. Heanue, J.A. Jones, F.J. Swanson, T.
Turrentine, T.C. Winter. 2003. Road Ecology: Science and Solutions. Island Press.
Washington, D.C.
Forrester, J. 1989. Planning in the Face of Power. University of California Press. Berkeley,
California.
Frieden, B.J. 1979. The Environmental Protection Hustle. MIT Press. Cambridge,
Massachussetts.
130
Frieden, B.J. 1983. “The Exclusionary Effect of Growth Controls.” Annals of the American
Academy of Political and Social Science. 465:123-135.
Friedmann, J. 1987. Planning in the Public Domain: Knowledge to Action. Princeton
University Press. Princeton, New Jersey.
Garde, A.M. 2004. “New Urbanism as Sustainable Growth: A Supply Side Story and Its
Implications for Public Policy.” Journal of Planning Education and Research. 24(2):
154-170.
Geolytics. 2003. Neighborhood Change Database: 1970-2000. Geolytics, Inc. East
Brunswick, New Jersey.
Giddens, A. 1976. New Rules of Sociological Method: A Positive Critique of Interpretive
Sociologies. Hutchinson Press. London.
Giddens, A. 1984. The Constitution of Society. Polity Press. Cambridge.
Gimpel, J.G. and J.E. Shuknecht. 2004. Patchwork Nation: Sectionalism and Political
Change in American Politics. University of Michigan Press. Ann Arbor, Michigan.
Grossman, G.M. and A.B. Krueger. 1991. “Environmental Impacts of a North American Free
Trade Agreement.” NBER Working Paper 3914.
Habermas, J. 1973. Theory and Practice. Beacon Press. Boston.
Habermas, J. 1984. The Theory of Communicative Action. Beacon Press. Boston.
Hall, P.A. and R.C.R. Taylor. 1996. “Political Science and the Three New Institutionalisms.”
Political Studies. XLIV:936-957.
Hanson, S. and G. Giuliano (eds). 2004. The Geography of Urban Transportation, 3
rd

Edition. The Guilford Press. New York.
Hardin, G. 1968. “The Tragedy of the Commons.” Science. 162(3859): 1243-1248.
Harris, J.M. and N.R. Goodwin. 2001. “Volume Introduction.” In Harris, J.M., T.A. Wise,
K.P. Gallagher and N.R. Goodwin (eds.) 2001. A Survey of Sustainable Development:
Social and Economic Dimensions. Frontier Issues on Economic Thought Series.
Island Press. Washington, D.C.
Harris, J.M., T.A. Wise, K.P. Gallagher and N.R. Goodwin (eds.) 2001. A Survey of
Sustainable Development: Social and Economic Dimensions. Frontier Issues on
Economic Thought Series. Island Press. Washington, D.C.
Haughton, G. 1999. “Environmental Justice and the Sustainable City.” Journal of Planning
Education and Research. 18(3):233-43.
131
Healey, P. 1996. “The Communicative Turn in Planning Theory and its Implications for
Spatial Strategy Formations.” Environment and Planning B: Planning and Design.
23(2):217-234.
Heikkila, E.J. 1996. “Are Municipalities Tieboutian Clubs?” Regional Science and Urban
Economics. 26:203-226.
Holling, C.S. 1994. “An Ecological View of the Malthusian Conflict.” In H. Landsberg and
K. Lindahl-Kiessling (eds.) Population, Economic Development and the
Environment: The Making of Our Common Future. Oxford University Press. Oxford.
Howarth, R.B. and R.B. Norgaard. 1992. “Environmental Valuation Under Sustainable
Development.” The American Economic Review. 82(2):473-477.
Innes, J. 1995. “Planning Theory’s Emerging Paradigm: Communicative Action and
Interactive Practice.” Journal of Planning Education and Research. 14(3):183-189.
Innes, J. 1996. “Planning Through Consensus Building: A New View of the Comprehensive
Planning Ideal.” Journal of the American Planning Association. 62(4):460-72.
Innes, J. 1998. “Challenge and Creativity in Postmodern Planning.” Town Planning Review.
69(2):viii-ix.
International Bank for Reconstruction and Development (IBRD). 1992. World Development
Report 1992: Development and the Environment. Oxford University Press. Oxford.
International Council for Local Environmental Initiatives (ICLEI). 2010. ICLEI Global
Members. http://www.iclei.org/index.php?id=11454. Accessed October 29.
Jabareen, Y.R. 2006. “Sustainable Urban Forms: Their Typologies, Models, and Concepts.”
Journal of Planning Education and Research. 26:38-52.
Jacobs, J. 1961. The Death and Life of Great American Cities. Vintage Books. New York.
Jacobs, J. 1969. The Economy of Cities. Random House. New York.
Jepson, E.J. 2004. “The Adoption of Sustainable Development Policies and Techniques in
U.S. Cities: How Wide, How Deep, and What Role for Planners?” Journal of
Planning Education and Research. 23:229-241.
Jepson, E.J. 2007. “Sustainability and the Childe Thesis—What Are the Effects of Local
Characteristics and Conditions on Sustainable Development Policy?” Cities.
26(6):434-447.
Jones, E. 1996. “Social Polarization in Post-Industrial London.” In O’Loughlin, J. and J.
Friedrichs (eds.) 1996. Social Polarization in Post-Industrial Metropolises. William
de Gruyter. Berlin.
132
Kantor, P., H.V. Savitch and S.V. Haddock. 1997. “The Political Economy of Urban
Regimes: A Comparative Perspective.” Urban Affairs Review. 32(3):348-377.
Kidder, J. 2009. “Appropriating the City: Space, Theory and Bike Messengers.” Theory and
Society. 38:307-328.
Kiser, L. and E. Ostrom. 1982. “The Three Worlds of Action: A Metatheoretical Synthesis of
Institutional Approaches.” In Ostrom, E. (ed.) Strategies of Political Inquiry. Sage.
Beverly Hills, California.
Kostof. S. 1991. The City Shaped: Urban Patterns and Meanings Through History. Bullfinch
Press. Boston.
Krishnan, R., J. Harris, and N.R. Goodwin. 2001. A Survey of Ecological Economics. Island
Press. Washington, D.C.
Kuznets, S. 1955. “Economic Growth and Income Inequality.” The American Economic
Review. 45(1):1-28.
Kuznets, S. 1966. Modern Economic Growth: Rate, Structure, and Spread. Yale University
Press. New Haven, Connecticut.
Kuznets, S. 1973. “Modern Economic Growth: Findings and Reflections.” The American
Economic Review. 63(3):247-258.
Levett, R. 1998. “Sustainability Indicators—Integrating Quality of Life and Environmental
Protection.” Journal of the Royal Statistical Society. Series A (Statistics in Society).
161(3):291-302.
Lubell, M., R. Feiock and S. Handy. 2009. “City Adoption of Environmentally Sustainable
Policies in California’s Central Valley. Journal of the American Planning
Association. 75(3):293-308.
Malthus, T. 1798. An Essay on the Principle of Population. J. Johnson. London.
Massey, D.S., M.J. Fischer. 2003. The Geography of Inequality in the United States, 1950-
2000. Brookings-Wharton Papers on Urban Affairs, pp. 1-40. The Brookings
Institution. Washington, D.C.
Mazmanian, D.A. and M.E. Kraft. 1999. “The Three Epochs of the Environmental
Movement.” In Toward Sustainable Communities: Transition and Transformation in
Environmental Policy. MIT Press. Cambridge, Massachussetts.
McCann, B.A. and R.H. Ewing. 2003. Measuring the Health Effects of Sprawl: A National
Analysis of Physical Activity, Obesity and Chronic Disease. Smart Growth America,
Surface Transportation Policy Project. September.
133
McHarg, I. 1992. Design with Nature (25
th
Anniversary Edition). John Wiley & Sons. New
York, New York.
Meadows, D.H. and D.L. Meadows, J. Randers, and W.H. Behrens. 1972. The Limits of
Growth. Universe Books. New York.
Molotch, H. 1976. “The City as a Growth Machine: Toward a Political Economy of Place.”
American Journal of Sociology. 82(2):309-332.
Moos, A.I. and M.J. Dear. 1986. “Structuration Theory in Urban Analysis: 2. Empirical
Application.” Environment and Planning A. 18:351-373.
Moran, E. and E. Ostrom. 2005. Seeing the Forest and the Trees: Human-Environment
Interactions in Forest Ecosystems. MIT Press. Cambridge, Massachussetts.
Mumford, L. 1938. The Culture of Cities. Harcourt Brace and Company. New York.
Mumford, L. 1961. The Cities in History: Its Origins, Transformations and Its Prospects.
Harcourt, Brace and World. New York.
Myers, D. and T. Banerjee. 2005. “Toward Greater Heights for Planning: Reconciling the
Differences Between Profession, Practice and Academic Field.” Journal of the
American Planning Association. 71(2):121-129.
National Association of Homebuilders (NAHB). 2002. What 21
st
Century Home Buyers
Want: A Survey of Customer Preferences. Home Builder Press. Washington, D.C.
Neilsen, F. and A.S. Alderson. 1997. “The Kuznets Curve and the Great U-Turn: Income
Inequality in the U.S. Counties: 1970-1990.” American Sociological Review.
62(1):12-33.
Neyman, J. and E.S. Pearson. 1928. “On the Use and Interpretation of Certain Test Criteria
for Purposes of Statistical Inference, Part I.” Biometrika. 20A:175-240.
Norgaard, R.B. 1994. Development Betrayed: The End of Progress and a Coevolutionary
Revisioning of the Future. Routledge. New York.
O’Loughlin, J. and J. Friedrichs (eds.) 1996. Social Polarization in Post-Industrial
Metropolises. Walter de Gruyter. Berlin.
O’Reilly, K. and G. Webster. 1998. “A Sociodemographic and Partisan Analysis of Voting in
Three Anti-Gay Referenda in Oregon.” Professional Geographer. 50(4):498-515.
Pepper, D. 1996. Modern Environmentalism: An Introduction. Routledge. New York, New
York.
134
Pezzey, J. 1992. Sustainable Development Concepts: An Economic Analysis. World Bank
Environment Paper Number 2. The World Bank. Washington, DC.
Phipps, A.G. 2001. “Empirical Applications of Structuration Theory.” Geografiska Annaler.
83B(4):189-204.
Portney, K.E. 2003. Taking Sustainable Cities Seriously: Economic Development, the
Environment and Quality of Life in American Cities. MIT Press. Cambridge,
Massachussetts.
Pulido, L. 2004. “Environmental Racism and Urban Development.” In Wolch, J., M. Pastor,
Jr. and P. Dreier (eds.) Up Against the Sprawl: Public Policy and the Making of
Southern California. University of Minnesota Press. Minneapolis.
Rawls, J. 1971. A Theory of Justice. Harvard University Press. Cambridge, Massachussetts.
Rawls, J. 2001. Justice as Fairness: A Restatement. Harvard University Press. Cambridge,
Massachussetts.
Rees, W.E. 1995. “Achieving Sustainability: Reform or Transformation.” Journal of
Planning Literature. 9(4):343-361.
Rhodes, E.L. 2003. Environmental Justice in America: A New Paradigm. Indiana University
Press. Bloomington, Indiana.
Robinson, S. 1976. “A Note on the U-Hypothesis Relating Income Inequality and Economic
Development.” American Economic Review. 66(3):437-440.
Rolston, H., III. 1996. “Feeding People Versus Saving Nature?” In Aiken, W. and H.
LaFolette. 1996. World Hunger and Morality, 2
nd
Edition. Prentice Hall. Upper
Saddle River, New Jersey.
Rostow, W. 1960. The Stages of Economic Growth: A Non-Communist Manifesto.
Cambridge University Press. Cambridge.
Ryan, S. and J.A. Throgmorton. 2003. “Sustainable Transportation and Land Development
on the Periphery: A Case Study of Freiburg, Germany and Chula Vista, California.”
Transportation Research Part D. 8:37-52.
Sabatier, P. (ed.) 2007. Theories of the Policy Process, 2
nd
Edition. Westview Press.
Cambridge, Massachussetts.
Sabatier, P.A. and C.M. Weible. 2007. “The Advocacy Coalition Framework: Innovations
and Clarifications.” In Sabatier, P.A. (ed.) 2007. Theories of the Policy Process.
Westview Press. Boulder, Colorado.
135
Sabatier, P.A. and H. Jenkins-Smith. 1993. Policy Change and Learning: An Advocacy
Coalition Approach. Westview Press. Boulder, Colorado.
Santa Monica, City of. 2010. Neighborhood Organizations.
www.smgov.net/content.aspx?id=4241. Accessed October 29, 2010.
Satterthwaite, D. (ed.) 1999. The Earthscan Reader in Sustainable Cities. Earthscan
Publications. London.
Savitch, H.V. and P. Kantor. 2002. Cities in the International Marketplace: The Political
Economy of Urban Development in North America and Western Europe. Princeton
University Press. Princeton, New Jersey.
Schaeffer, K.H. and E. Sclar. 1975. Access for All: Transportation and Urban Growth.
Penguin Books. New York.
Scheffer, V.B. 1951. “The Rise and Fall of a Reindeer Herd.” Scientific Monthly. 73:356-
362.
Schelling, T.C. 1997. “The Cost of Combating Global Warming.” Foreign Affairs. 76(6):8-
14.
Schneider, M. 1986. “Fragmentation and the Growth of Local Government.” Public Choice.
48:255-263.
Scott, J.C. 1998. Seeing Like a State: How Certain Schemes to Improve the Human Condition
Have Failed. Yale University Press. New Haven, Connecticut.
Simon, H. 1957. Models of Man. John Wiley & Sons. New York.
Staley, S.R. 2006. “Sustainable Development in American Planning: A Critical Appraisal.
Town Planning Review. 77(1):99-126.
Stone, C.N. 1989. Regime Politics: Governing Atlanta 1946-1988. University Press of
Kansas. Lawrence, Kansas.
Sturm, R. and D. Cohen. 2004. “Suburban Sprawl and Physical and Mental Health.” Public
Health. 118:488-496.
Tiebout, C. 1956. “A Pure Theory of Local Expenditures.” The Journal of Political
Economy. 64(5):416-424.
Toman, M.A. 1992. “The Difficulty in Defining Sustainability.” Resources. 106:3-6.
U.S. Census Bureau. 1994. Geographic Areas Reference Manual. November.
http://www.census.gov/geo/www/garm.html. Accessed October 29, 2010.
136
U.S. Department of Health and Human Services (DHHS). 2010. The 2000 HHS Poverty
Guidelines. http://aspe.hhs.gov/poverty/00poverty.htm. Accessed October 26, 2010.
U.S. Environmental Protection Agency (EPA). 2009. HUD-DOT-EPA Partnership for
Sustainable Communities. http://www.epa.gov/dced/partnership/index.html. Accessed
October 29, 2010.
U.S. Office of Management and Budget (OMB). 2000. Standard Occupational Classification
Manual 2000. National Technical Information Service. Springfield, Virginia.
United Church of Christ (UCC). 1987. Toxic Wastes and Race in the United States: A
National Report on the Racial and Socioeconomic Characteristics of Communities
with Hazardous Waste Sites. Commission for Racial Justice. New York.
United Nations. 1987. Kyoto Protocol to the United Nations Framework Convention on
Climate Change. http://unfccc.int/resource/docs/convkp/kpeng.pdf. Accessed October
29, 2010.
United Nations. 1992. United Nations Framework Convention on Climate Change.
http://unfccc.int/resource/docs/convkp/conveng.pdf. Accessed October 29, 2010.
United Nations. 2010. Status of Ratification of the Kyoto Protocol. http://unfccc.int/kyoto-
protocol/status_of_ratification/items/2613.php. Accessed October 29.
United States Conference of Mayors (USCM). 2010. Mayors Climate Protection Center List
of Participating Mayors. http://www.usmayors.org/climateprotection/list.asp.
Accessed October 29, 2010.
Verma, N. 1995. “What is Planning Practice? The Search for Suitable Categories.” Journal
of Planning Education and Research. 14:178-182.
Verma, N. 1996. “Understanding What We Can’t Define? The Epistemology of Sustainable
Development.” TRIALOG. 48:6-11
Verma, N. and H.R. Shin. 2004. Communicative Action and the Network Society: A
Pragmatic Marriage. Journal of Planning Education and Research. 24(2):131-140.
Wheeler, S.M. and T. Beatley (eds.) 2004. The Sustainable Urban Development Reader.
Urban Reader Series. Routledge. New York.
Winslow, M. 2007. “The Environmental Kuznets Curve Revisited Once Again.” Forum for
Social Economics, pp. 1-18.
Wolch, J., M. Pastor, Jr. and P. Dreier (eds.) 2004. Up Against the Sprawl: Public Policy and
the Making of Southern California. University of Minnesota Press. Minneapolis.
137
World Commission on Environment and Development (WCED). 1987. Our Common Future.
Oxford University Press. Oxford.
World Resources Institute. 2010. Agenda 21: Number of Formally Committed Municipalities.
http://earthtrends.wri.org/searchable_db/index.php?theme=10&variableID=661&acti
on=select_countries. Accessed October 29.
Yandle, B., M. Bhattarai, and M. Vijayaraghavan. 2004. Environmental Kuznets Curves: A
Review of Findings, Methods, and Policy Implications. The Property and
Environment Research Center. Research Study 02-1 Update, pp. 1-38.
138
APPENDIX

Table A1: Data Sources and Derivations of Community Indicators
Variable Derivation from 2000 Census (Numerator Table:Row; Denominator Table:Row)
Age (percent of total population)
POP517 P8:8-20,47-59; P1:1
POP1864 P8:21-34,60-73; P1:1
POP65P P8:35-40,74-79; P1:1
Race/ethnicity (percent of total population)
NHWHITE P4:5,14,17,35; P6:1
NHBLACK P4:6,13,18-21,29-32,39-44,50-55,60-63,66-69,71,73; P6:1
NHASIAN P4:7,24; P6:1
NHSOTH P4:10; P6:1
HISPANIC P7:10; P6:1
Family Structure (percent of family households)
MCWKID P12:5,20; P15:1
MCNKID P12:6,21; P15:1
FHWKID P12:12,27; P15:1
FHNKID P12:13,28; P15:1
Education (percent of population > 25 years old)
NOHSD P37:7-10,24-27; P37:1
HSD P37:11,28:P37:1
PSE P37:12-18,29-35: P37:1
Language (percent of population > 18 years old)
ENGL P19:25,47; P8:21-34,60-73
NOENGL P19:29-30,34-35,39-40,44-45,51-52,56-57,61-62,66-67; P8:21-34,60-73
Employment (percent of civilian labor force > 16 years old)
UNEMP P43:7,14; P43:5,12
LOWSK P50:23,44,46-48,70,91,93-95; P43:6,13
MEDSK P50:32,79,33,80,36,40,42,83,87,89; P43:6,13
HIGHSK P50:10,45,57,92,5,7,52,54; P43:6,13
Housing (percent of housing units)
OWNOCC H7:2; H1:1
RNTOCC H7:3; H1:1
SFR H30:2; H1:1
NEWCON H34:2-4; H1:1
GT1PR H49:13; H7:2
Income and Expense (percent of households)
HIN100P P76:14-17; P15:1
WELFARE P63:2,P64:2; P64:1
POVERTY P87:2; P87:1
GT35RNT H73:7,14,21,28,35,42,49; H62:2

139
(Table A1 continued)
Mobility (percent of civilian employed population > 16 working outside the home, except as noted)
PBTRAN P30:5; P30:2,5,12-15
AUTO P30:2; P30:2,5,12-15
CMT25 P31:3-7,15; P30:2,5,12-15
CMT2544 P31:8-11; P30:2,5,12-15
CMT45P P31:12-14; P30:2,5,12-15
FORCNTR P24:16; P24:1
SMHSE P24:2; P24:1

Note: Each indicator is a rate variable for which the numerator is the sum of counts indicated by rows in 2000
Census tables before the semi-colon. The denominator is the sum of counts from rows after the semi-colon.
140
Table A2: Descriptive Statistics for Community Indicators for U.S. Cities in 2000
Variable
Group Variable Minimum Maximum Mean
Standard
Deviation
POP517 0.5 36.1 19.1 3.6
POP1864 27.8 99.5 59.4 5.1
POP65P 0.0 65.0 14.9 5.6
NHWHITE 0.0 100.0 77.6 23.1
NHBLACK 0.0 98.2 9.2 16.0
NHASIAN 0.0 63.7 2.2 4.5
NHSOTH 0.0 95.9 1.3 5.6
HISPANIC 0.0 99.0 9.8 16.9
MCWKID 0.0 61.4 24.1 7.6
MCNKID 0.0 58.1 28.6 5.6
FHWKID 0.0 55.9 7.2 3.3
FHNKID 0.0 22.9 4.5 2.4
NOHSD 0.0 40.9 12.4 5.4
HSD 3.4 55.7 31.0 8.4
PSE 6.0 94.6 48.1 14.8
ENGL 4.9 100.0 87.4 15.8
NOENGL 0.0 73.8 3.2 5.8
UNEMP 0.0 77.6 6.1 3.7
LOWSK 3.3 53.8 23.0 5.8
MEDSK 4.8 76.6 29.0 10.2
HIGHSK 18.5 68.0 45.4 6.7
OWNOCC 0.0 97.0 61.5 12.4
RNTOCC 0.8 100.0 29.4 11.5
SFR 1.0 100.0 67.7 15.0
NEWCON 0.0 85.3 15.0 11.7
GT1PR 0.0 60.1 4.2 5.6
HIN100P 0.0 71.5 7.8 8.2
WELFARE 0.0 51.2 8.4 5.1
POVERTY 0.0 90.2 13.3 8.1
GT35RNT 0.0 41.0 8.7 5.1
PBTRAN 0.0 58.8 1.5 3.4
AUTO 3.9 100.0 93.5 6.9
CMT25 22.3 115.6 71.8 14.1
CMT2544 0.0 52.4 19.8 9.8
CMT45P 0.0 53.9 11.9 7.0
FORCNTR 0.0 48.1 1.8 2.2
Age

Race and
Ethnicity

Family
Structure

Education
Level

Language

Employment

Housing

Income and
Expenses

Mobility

SMHSE 8.0 82.9 54.4 8.8

Note: Values are expressed as percentages of citywide counts, as described in Table A1. N = 5,340.
141
Table A3: Mean Values of Community Indicators for Selected Sustainable Cities in
2000
Variable
Group Variable Phoenix Austin Boulder

Portland
Santa
Monica Cambridge
Number of
Census
Tracts 291 155 21 145 19 30
POP517 20.3 15.4 10.4 15.1 10.3 9.1
POP1864 63.1 70.9 77.7 67.3 71.4 77.6
POP65P 8.1 6.6 8.1 11.6 14.3 9.2
NHWHITE 56.6 53.7 85.0 76.9 72.9 65.7
NHBLACK 5.2 10.1 1.4 7.4 4.3 13.1
NHASIAN 2.4 5.2 4.9 7.8 8.7 13.0
NHSOTH 1.8 0.4 0.5 1.1 0.7 0.9
HISPANIC 34.0 30.6 8.2 6.9 13.5 7.3
MCWKID 25.4 19.4 14.6 17.4 11.6 12.6
MCNKID 22.9 19.9 18.6 22.0 16.8 17.8
FHWKID 7.7 6.4 4.1 6.0 3.9 5.4
FHNKID 4.6 4.1 2.1 4.3 3.2 3.9
NOHSD 12.4 8.3 2.9 9.1 5.3 5.4
HSD 22.9 17.3 8.8 22.1 12.0 12.2
PSE 53.8 66.1 85.7 63.6 79.0 77.3
ENGL 69.2 69.3 85.8 83.9 71.6 69.7
NOENGL 11.4 8.3 3.2 5.1 4.6 4.0
UNEMP 5.6 4.4 6.8 6.5 7.4 6.1
LOWSK 21.2 16.6 16.7 20.9 11.1 10.7
MEDSK 31.1 42.8 53.0 37.4 60.3 66.8
HIGHSK 47.5 40.4 30.2 41.4 28.5 22.4
OWNOCC 57.1 43.4 47.4 53.0 27.8 30.7
RNTOCC 36.8 52.6 49.8 41.3 65.2 64.6
SFR 57.7 46.7 43.6 60.6 19.4 8.8
NEWCON 20.3 21.1 9.6 10.2 5.7 4.7
GT1PR 11.4 8.4 3.6 4.9 4.9 4.1
HIN100P 9.4 10.3 12.5 8.0 14.4 11.4
WELFARE 6.4 4.2 3.8 8.2 6.2 6.1
POVERTY 15.7 14.3 17.9 13.0 10.4 12.9
GT35RNT 12.1 18.3 22.9 14.4 21.1 22.2
PBTRAN 3.4 4.5 9.1 12.8 4.5 26.5
AUTO 92.2 90.8 72.6 79.2 88.5 42.6
CMT25 57.4 69.0 84.1 68.9 64.2 63.7
CMT2544 32.0 25.7 13.7 26.0 31.3 30.4
CMT45P 14.0 8.9 9.2 9.7 13.1 11.4
FORCNTR 6.3 6.6 6.2 4.2 4.8 10.6
Age

Race and
Ethnicity

Family
Structure

Education
Level

Language

Employment

Housing

Income and
Expenses

Mobility

SMHSE 43.4 36.3 34.4 45.3 52.6 38.7
Note: Values are expressed as mean percentages of citywide counts, as described in Table A1.
142
Table A4: Descriptions of Central Objects for Regime Classes, Santa Monica,
California in 2000

Variable Group Variable
Development/
Maintenance
Regime
Middle Class
Progressive
Regime
Low-Income
Opportunity
Regime

Census Tract 06037701201 06037701501 06037701802
POP517 948 313 626
POP1864 2452 3501 3387
POP65P 690 644 420
NHWHITE 3844 3903 1729
NHBLACK 20 98 751
NHASIAN 272 377 376
NHSOTH 34 0 19
HISPANIC 184 277 1784
MCWKID 537 210 240
MCNKID 548 452 172
FHWKID 71 96 160
FHNKID 35 117 137
NOHSD 87 152 364
HSD 233 405 626
PSE 2624 3350 1785
ENGL 2461 3230 2138
NOENGL 48 92 471
UNEMP 59 213 297
LOWSK 58 345 399
MEDSK 1386 1673 716
HIGHSK 338 930 863
OWNOCC 1384 616 319
RNTOCC 185 2146 1461
SFR 1515 83 289
NEWCON 154 83 126
GT1PR 13 75 335
HIN100P 933 360 55
WELFARE 59 136 333
POVERTY 129 421 936
GT35RNT 42 572 516
PBTRAN 12 73 111
AUTO 1525 2523 1476
CMT25 1068 1739 1197
CMT2544 524 832 474
CMT45P 179 350 251
FORCNTR 51 175 331
Age

Race and
Ethnicity

Family Structure

Education Level

Language

Employment

Housing

Income and
Expenses

Mobility

SMHSE 2921 2146 2261

Note: Values are expressed as counts of population (for Age, Race and Ethnicity, Family Structure, Education,
Language variables), members of civilian labor force (for Employment variables), housing units (for Housing
variables), households (for Expenses variables) and employed civilians (for Mobility variables) for the given
Census tract.

143
Table A5: Correlations Among Modeled Regimes of Sustainable Cities by Kuznetsian
Type in 2000

Correlation Coefficient
(Spearman) between Regimes
City Type
Elite-
Lower
Elite-
Middle
Middle-
Lower
Poverty-Affluence
Dissimilarity Index
Tucson, AZ A 0.80 0.76 0.82 0.58
Brownsville, TX A 0.89 0.72 0.70 0.55
Milwaukee, WI A 0.85 0.85 0.86 0.54
Cleveland, OH A 0.23 0.47 0.69 0.50
New Haven, CT A 0.69 0.61 0.86 0.48
Stuart, FL A 0.80 0.82 0.81 0.33
Tampa, FL B 0.69 0.72 0.95 0.70
Oklahoma City, OK B 0.83 0.61 0.74 0.66
Indianapolis, IN B 0.66 0.74 0.65 0.63
Jacksonville, FL B 0.57 0.79 0.85 0.58
Orlando, FL B 0.83 0.82 0.68 0.57
Ithaca, NY B 0.75 0.92 0.74 0.56
Burlington, VT B 0.80 0.81 0.76 0.55
Chattanooga, TN B 0.85 0.85 0.75 0.53
Lansing, MI B 0.59 0.90 0.62 0.50
Olympia, WA B 0.91 0.83 0.90 0.40
Phoenix, AZ C 0.65 0.90 0.74 0.70
Austin, TX C 0.74 0.76 0.76 0.68
Boulder, CO C 0.84 0.52 0.67 0.59
Portland, OR C 0.55 0.84 0.86 0.57
Santa Barbara, CA C 0.72 0.60 0.90 0.54
Boston, MA C 0.63 0.57 0.77 0.52
Scottsdale, AZ C 0.73 0.87 0.88 0.51
San Jose, CA C 0.77 0.81 0.80 0.51
Seattle, WA C 0.67 0.67 0.70 0.51
San Francisco, CA C 0.64 0.87 0.81 0.46
Santa Monica, CA C 0.45 0.74 0.73 0.38
Cambridge, MA C 0.76 0.87 0.88 0.37
Annapolis, MD C 0.85 0.91 0.91 0.11

Note: Type A, B, and C represent low, medium and high levels of economic growth according to terciles of the
distribution of average household income for cities in the U.S. for 2000. Correlations are based on Spearman
correlation coefficient values for cities based on consideration of variables presented in Table 3. Poverty-
affluence dissimilarity is calculated by the formula presented on p.59.

Abstract (if available)

Abstract Sustainable development has emerged as an important paradigm of planning, and its economic, social, and environmental goals are being pursued in a growing number of settings. This dissertation considers the interplay of structure and agency as means to better understand sustainability planning efforts in U.S. cities. Drawing insight from Simon Kuznets’ empiric work on the social and economic structure of developing countries (Kuznets 1955), this study investigates the possibility that Kuznetsian development can inform our thinking on sustainable city planning as well. Ordinary least squares regression models reveal the classic “Kuznets Curve” among data for U.S. cities, such that income inequality first rises then falls with economic growth. The strength of this relationship increases as several explanatory variables are added, notably educational attainment and labor skill of the general population, and also when segregation of residents within cities is considered. Whether or not a city has a sustainability initiative also contributes significantly to the relationship. These results suggest that Kuznetsian development reasonably describes the experience of U.S. cities, and the concept provides a framework for studying structure and agency related to sustainable development in this setting. A three-category typology of cities is offered such that a given city’s type is determined by its position along the curve. The typology is used to select and investigate cities for clues about why some of the wealthier cities in the U.S. that have well-recognized sustainability initiatives appear to be “making the turn” toward greater equality of income (a social goal of sustainable development) while others may be at risk of “missing the turn.”

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The role of CALGreen codes and sustainable rating systems in practicing sustainability

PDF

Do sustainability plans affect urban sustainability outcomes in Santa Monica, San Francisco, and San Jose?

PDF

Productive frictions and urbanism in transition: planning lessons from traffic flows and urban street life in Ho Chi Minh City, Vietnam

PDF

Governance and urban design: Omaha by design

PDF

Urban conservation in the Middle Eastern historic cities: globalization and lack of identity

PDF

The crisis of potable water in Mexico City: institutional factors and water property rights as conditions for creating adequate metropolitan water governance

PDF

Urban spatial structure, commuting, and growth in U.S. metropolitan areas

PDF

The Wenchuan earthquake recovery: civil society, institutions, and planning

PDF

Embedding sustainability: a change management guide for ports

PDF

The spatial economic impact of live music in Orange County, CA

PDF

Green healthcare, an environmentally sustainable methodology: an investigation of the ecological impacts of the healthcare industry and the role of green initiatives in sustainable medical services

PDF

Resilient and equitable urbanism by design: insights from the collaborative process to reimagine the SF Bay Area

PDF

Evergreen economies: institutions, industries and issues in the green economy

PDF

Social construction of the experience economy: the spatial ecology of outdoor advertising in Los Angeles

PDF

Place production in globalizing middle eastern cities: a study of Cairo and Dubai

PDF

The impact of social capital: a case study on the role of social capital in the restoration and recovery of communities after disasters

PDF

Community development agreements: addressing inequality through urban development projects

PDF

Beyond the limits to planning for equity: the emergence of community benefits agreements as empowerment models in participatory processes

PDF

Why go green? Cities' adoption of local renewable energy policies and urban sustainability certifications

PDF

Essays on the economics of cities

Asset Metadata

Creator Hanson, Mark Alan (author)

Core Title Structure, agency, and the Kuznets Curve: observations and implications for sustainability planning in U.S. cities

School School of Policy, Planning, and Development

Degree Doctor of Philosophy

Degree Program Policy, Planning, and Development

Degree Conferral Date 2011-05

Publication Date 02/01/2011

Defense Date 11/29/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag agency,economic,Economic development,environmental,Kuznets,OAI-PMH Harvest,political mobilization,public participation,social,structure,sustainability planning,sustainable cities,sustainable development,urban regimes

Place Name USA (countries)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Banerjee, Tridib (committee chair), Robertson, Peter John (committee member), Tierney, William G. (committee member)

Creator Email mhanson@rand.org,mhanson@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3634

Unique identifier UC1249311

Identifier etd-Hanson-4316 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-428075 (legacy record id),usctheses-m3634 (legacy record id)

Legacy Identifier etd-Hanson-4316.pdf

Dmrecord 428075

Document Type Dissertation

Rights Hanson, Mark Alan

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu