Citation
Inequities in young people's access to urban parks

Material Information

Title:
Inequities in young people's access to urban parks an environmental justice investigation in Denver
Creator:
Rigolon, Alessandro ( author )
Language:
English
Physical Description:
1 electronic file (727 pages). : ;

Subjects

Subjects / Keywords:
Environmental justice -- Colorado -- Denver ( lcsh )
Parks -- Planning -- Colorado -- Denver ( lcsh )
Recreation areas -- Colorado -- Denver ( lcsh )
Youth -- Recreation -- Colorado -- Denver ( lcsh )
Environmental justice ( fast )
Parks -- Planning ( fast )
Recreation areas ( fast )
Youth -- Recreation ( fast )
Colorado -- Denver ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Given the benefits of repeated contact with nature for young people inequities in park spatial distribution between young people of different socio economic status and ethnicity are a serious environmental justice issue. A growing body of literature has documented that low income young people of color have access to fewer acres of parks and to parks with lower quality than wealthier Non Hispanic White young people. However very few studies have investigated the planning processes that have contributed to these inequities. Also no previous investigation on access to parks has combined qualitative research on young people s use and perception of parks in a geospatial analysis of park spatial distribution. In this dissertation I integrated the analyses of equitable decision making processes of equitable park distributions and of park quality in relation to young people s needs to understand how planning policies and practices contributed to park provision for different income and ethnic groups with a particular focus on parks that can be meaningful for young people. Drawing from geospatial methods interviews with experts and policy analysis I uncovered a complex picture of environmental injustice including inequities in the spatial distribution of parks and in the policies and practices that led to such distribution. My geospatial analysis showed that low income ethnic minority young people have very low access to safe parks and to parks with excellent levels of quality which might substantially limit their park visitation. Also the Poundstone Amendment a state policy limiting Denver s annexations from surrounding counties aimed to avoid ethnic integration in schools combined with the lack of an impact fee for parks in infill developments has significantly hindered Denver s capacity to improve park provision particularly for its low income ethnic minority people. The mixed research integration of qualitative and quantitative datasets showed that most cases of distributional injustice are linked to unjust processes while very few cases of partial distributional equity are related to just processes. This dissertation has important implications for park and land use planning including current barriers and possible solutions to increasing park equity for environmental justice research and practice and for the use of mixed research in planning.
Thesis:
Thesis (Ph.D.)--University of Colorado Denver.
Bibliography:
Includes bibliographic references.
System Details:
System requirements: Adobe Reader.
General Note:
College of Architecture and Planning
Statement of Responsibility:
by Alessandro Rigolon.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
930293604 ( OCLC )
ocn930293604

Downloads

This item has the following downloads:


Full Text
INEQUITIES IN YOUNG PEOPLES ACCESS TO URBAN PARKS: AN
ENVIRONMENTAL JUSTICE INVESTIGATION IN DENVER
By
ALESSANDRO RIGOLON
B. Arch., M. Arch. University of Bologna, 2007
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Design and Planning
2015


11
2015
ALESSANDRO RIGOLON
ALL RIGHTS RESERVED


This thesis for the Doctor of Philosophy degree by
Alessandro Rigolon
has been approved for the
Design and Planning Program
by
Jeremy R. Nemeth, Chair
Louise Chawla, Advisor
Nancy L. Leech
Sharon E. Sutton
Willem van Vliet


IV
Rigolon, Alessandro (Ph.D., Design and Planning)
Inequities in Young Peoples Access to Urban Parks: An Environmental Justice
Investigation in Denver
Thesis directed by Associate Professor Jeremy R. Nemeth
ABSTRACT
Given the benefits of repeated contact with nature for young people, inequities in
park spatial distribution between young people of different socio-economic status and
ethnicity are a serious environmental justice issue. A growing body of literature has
documented that low-income young people of color have access to fewer acres of parks
and to parks with lower quality than wealthier Non-Hispanic White young people.
However, very few studies have investigated the planning processes that have contributed
to these inequities. Also, no previous investigation on access to parks has combined
qualitative research on young peoples use and perception of parks in a geospatial
analysis of park spatial distribution. In this dissertation, I integrated the analyses of
equitable decision making processes, of equitable park distributions, and of park quality
in relation to young peoples needs to understand how planning policies and practices
contributed to park provision for different income and ethnic groups, with a particular
focus on parks that can be meaningful for young people. Drawing from geospatial
methods, interviews with experts, and policy analysis, I uncovered a complex picture of
environmental injustice, including inequities in the spatial distribution of parks and in the
policies and practices that led to such distribution. My geospatial analysis showed that
low-income ethnic minority young people have very low access to safe parks and to parks
with excellent levels of quality, which might substantially limit their park visitation. Also,


V
the Poundstone Amendment, a state policy limiting Denvers annexations from
surrounding counties aimed to avoid ethnic integration in schools, combined with the
lack of an impact fee for parks in infill developments, has significantly hindered Denvers
capacity to improve park provision, particularly for its low-income ethnic minority
people. The mixed research integration of qualitative and quantitative datasets showed
that most cases of distributional injustice are linked to unjust processes, while very few
cases of partial distributional equity are related to just processes. This dissertation has
important implications for park and land use planning, including current barriers and
possible solutions to increasing park equity, for environmental justice research and
practice, and for the use of mixed research in planning.
The form and content of this abstract are approved. I recommend its publication.
Approved: Jeremy R. Nemeth


VI
Para Sofia


ACKNOWLEDGEMENTS
At the end of this four-year journey, I am overwhelmed by feelings of gratitude. I
am grateful for the beautiful state of Colorado, with its mountains, its sunsets, and its
people, who have welcomed me, supported me, and made me feel at home. At the
University of Colorado, the Children, Youth, and Environment Center and its wonderful
people have been a safe and stimulating harbor for a foreign student looking for an
anchor and new perspectives. Willem, Louise, Tori, Mara, and other fellow students and
interns have made this journey an exciting and smooth ride. Jeremy has been a recent and
inspiring addition to the great team of mentors I was lucky to work with, and I wish our
paths had crossed earlier. Sharon and Nancy have been exceptional for my growth as a
researcher, from two different perspectives. Sharon welcomed me at the University of
Washington in Seattle five years ago, when this journey started, as I walked into her
room with an Italian-English dictionary. My fellow PhD students, particularly Travis and
Mehdi, have also been incredible travel mates, and greatly helped with my research.
Colorado has also gifted me with a family and a new home. I would not have met
my wife, Sofia, if I had not started a PhD program in Colorado. She is the greatest among
all great gifts of life. Sofia and her family, Suse, Manuel and Fabian, have become my
family. They have become another home, away from home, and helped me cope with
missing my parents and friends in my native land, Italy. My parents, who are also coping
with their only child living oceans away, deserve the greatest gratitude. They accepted
my dream of pursuing an academic career and a life full of adventures in the United
States. I hope to make the most of the next journeys that Sofia and I will share.


Vlll
TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION.........................................................1
Denver as a Compelling Case Study................................8
Denvers Population Growth and Park System................8
Denvers Demographics.....................................9
Denvers History of Ethnic Residential Segregation.......11
Denvers Environmental Justice Issues....................12
Generalizability.........................................13
A Complex Picture of Environmental Injustice....................14
Relevance to Urban Planning.....................................17
Definition of Key Terms.........................................19
II. THEORETICAL FRAMEWORK, LITERATURE REVIEW, AND
RESEARCH QUESTIONS.....................................................23
Theoretical Framework...........................................23
Identification of Constructs.............................23
Relationships between Constructs.........................33
Transformative Paradigm and Environmental Justice........35
Literature Review...............................................43
Park Spatial Distribution and Demographics...............44
Policies, Practices, and Park Spatial Distribution
61


IX
Young Peoples Demographics and Perceptions and Use of Parks
.............................................................71
Park Availability and Young Peoples Use of Parks............79
Park Availability and Environmental Justice..................82
Literature Gaps and Inconsistencies.................................83
Main Gap: Lack of Focus on Just Processes, Just Outcomes, and
Park Quality.................................................83
Less Substantial Gaps........................................84
Methodological Variations and Inconsistencies................86
Research Objectives and Questions...................................88
Research Objectives..........................................89
Research Questions...........................................90
Mixing Purposes..............................................92
III. RESEARCH METHODS......................................................94
Population and Sampling Frame.......................................94
Park Planning Processes Analysis.............................95
Park Spatial Distribution Analysis...........................97
Research Design.....................................................98
Data Collection Instruments........................................101
Park Planning Processes.....................................102
Park Spatial Distribution...................................106
Data Analysis......................................................107


X
Park Planning Process Data Analysis.......................107
Park Spatial Distribution Data Analysis...................110
Mixed Research Integration................................128
Data Legitimation................................................131
IV. PARK PLANNING PROCESS FINDINGS......................................135
An Introduction to Denvers History..............................138
Denvers Early Stages: 1858-1902..........................141
The City Beautiful and New Deal Eras: 1902-1945...........142
Suburbanization: 1945-1983 ...............................162
Urban Renaissance: 1983-2015..............................171
Looking Back to Look Ahead................................178
Park Planning....................................................178
Parkland Acquisition and Park Construction................180
Park Design, Management, and Use..........................243
Park Planning in Denver: A Brief Summary..................266
Land Use Planning................................................267
Land Use Planning and Park Establishment..................268
Land Use Planning and Residential Location................281
Land Use Planning for Park Establishment and Residential
Location: A Summary.......................................304
Housing Policies and Practices
305


XI
Ethnic and Racial Discrimination in Housing: Policies and
Practices...................................................307
Affordable Housing in Denver................................317
Housing Policies and Practices: A Summary...................330
Park Planning Processes: Bringing It All Together..................332
Park Planning Processes: 1902-1945..........................334
Park Planning Processes: 1946-1988..........................338
Park Planning Processes: 1989-2015..........................344
Park Planning Processes: A Brief Comparison.................351
V. PARK SPATIAL DISTRIBUTION FINDINGS...................................354
Developing a Park Quality Index for Youth.........................354
Dates, Locations and Methods................................355
What Green Space Features Matter for Children and Teenagers?
............................................................356
Operationalizing These Findings in a Park Quality Index for Youth
(PQIY)......................................................368
Park Spatial Distribution..........................................381
Descriptive Statistics......................................382
Summary of Findings.........................................389
Analysis of Park Proximity..................................400
Analysis of Park Acreage....................................413
Analysis of Park Quality....................................433
Park Spatial Distribution: Discussion.......................472


xii
VI. MIXED RESEARCH INTEGRATION......................................478
Data Reduction and Data Display...............................482
Data Comparison...............................................483
Does the Partially Equitable Distribution of Parks Based on
Certain Variables Correspond to Equitable Processes?.........485
Does the Inequitable Distribution of Parks Based on Certain
Variables Correspond to Discriminatory or Negligent Processes?
........................................................494
Do Perceptions of Distributional Inequity and GIS-Measured
Distributional Inequity Coincide or Differ?.............501
Do Notable Policies, Plans, and Practices Developed at Different
Times Relate to Parks Established at Specific Times?....503
Do Policies, Plans, and Practices Addressing Different Parts of the
City Relate to Park Provision in Specific Parts of the City?.506
Data Comparison: A summary..............................510
Extreme Case Analysis..............................................511
Why Do Some Low-Income Areas Have Good Access to Regional
Parks?..................................................512
Why Do Some Low-Income Areas Have Good Access to Best
PQIY Parks?.............................................515
Why Do Some High-Income Areas Have Poor Access to Best
PQIY Parks?.............................................518
Why do Some Low-Income Areas Have Good Access to Parks
with Low Violent Crime Density?.........................520
Why Do Some Low-Income Ethnic Minority Areas Have Good
Access to All Parks and Parkways, with Parkways Making a
Difference?.................................................522


Xlll
Why Do Some Low-Income Neighborhoods That Are Perceived as
Park-Poor Have Decent GIS-Measured Access to Parks in Terms
of Park Proximity?.........................................529
Summary of Extreme Case Analyses...........................530
Data Consolidation................................................531
Park Provision in the West Side............................546
The Growing Role of Non-Profits for Park Distributional Equity
......................................................548
Procedural Injustice and Its Outcomes: A Summary............550
VII. IMPLICATIONS..................................................553
Implications for Park Planning..............................554
Obstacles to Park Establishment: Today and Looking Ahead ....554
Changes in Municipal Policies and Park Planning to Facilitate Park
Establishment..............................................558
Implications of the Current Spatial Distribution of Parks..566
Parameters and Thresholds for Park Quality.................569
Implications for Park Planning: A Summary..................573
Implications for Environmental Justice............................573
Implications for Environmental Justice Research............574
Implications for Environmental Justice Practice: A Park Advocacy
Framework..................................................577
Implications for Mixed Research in Planning.......................586
Limitations.......................................................596
Design Limitations
597


XIV
Limitations in Specific Methods....................600
Directions for Future Research...........................603
Continuing Studies of Spatial Distribution and Park Quality.604
In-Depth Study of Unexpected Findings..............605
A Complex Picture of Environmental Injustice.............607
REFERENCES...........................................................615
APPENDIX A...........................................................668
APPENDIX B...........................................................696
APPENDIX C...........................................................705


CHAPTER I
INTRODUCTION
Research conducted in several Western countries shows that the time children and
teenagers spend playing outdoors and in nature is decreasing (Clemens, 2004; Fyhri,
Hjorthol, Mackett, Fotel, & Kytta, 2011; Karsten, 2005; Tandy, 1999; Wridt, 2004). This
decrease is problematic, given the benefits of contact with nature for mental and physical
health among young people (McCurdy, Winterbottom, Mehta, & Roberts, 2010), for their
overall well-being (Chawla, Keena, Pevec, & Stanley, 2014; Chawla, 2014; N. M. Wells,
2014), and for their development (Burdette & Whitaker, 2004; Fjortoft & Sageie, 2000; R.
C. Moore, 1986a; N. M. Wells, 2000). The growing attention to nature-deficit disorder
described by Richard Louv (2005) led to the establishment of several programs and
policies aimed to increase childrens repeated contact with nature (see Natural Learning
Initiative, 2014; No Child Left Inside, 2007). In 2007, the World Future Society predicted
that nature-deficit disorder will increase globally as a health risk (World Future Society,
2007). In urban areas, young people mostly rely on neighborhood parks for accessing
nature, as their home range is often limited to their neighborhood (Fagerholm & Broberg,
2011; Loebach & Gilliland, 2014). Also, since the world is urbanizing at very high rates
(United Nations, 2014), and as the United States is experiencing an urban renaissance for
the first time in decades (S. G. Wilson et al., 2012), nearby parks are becoming more and
more important for access to nature.
However, a recent review showed that low-income and ethnic minority children
and teenagers have even less contact with nature than Non-Hispanic White middle- and
upper-class children and teenagers due to lack of available parks and recreational


2
opportunities in their communities (National Recreation and Park Association, 2011). A
study by Downey (2007) suggested that spatial inequalities in access to parks are often
related to residential segregation. The inequalities are particularly disturbing because
contact with nature has stronger benefits for low-income children in terms of stress relief
(N. M. Wells & Evans, 2003).
Neighborhood parks can be democratizing public spaces that help balance
structural inequities in recreation and public health for low-income ethnic minorities.
Low-income ethnic minority children and teenagers need nearby parks more than other
groups given their limited transportation options (Floyd & Johnson, 2002; Loukaitou-
Sideris & Stieglitz, 2002; Ries et al., 2008), and given their inadequate access to private
recreational opportunities (Loukaitou-Sideris & Stieglitz, 2002). Neighborhood parks
matter the most for public health among ethnic minorities, as African American and
Hispanic young people have higher rates of obesity than their Non-Hispanic White
counterparts (Ogden, Carroll, Kit, & Flegal, 2014; Wang & Beydoun, 2007). Also, low-
income young people of color tend to have lower levels of physical activity than other
groups (Floyd, Taylor, & Whitt-Glover, 2009). Therefore, when measuring access to
parks, using the equality of spatial distribution as a desired result is problematic because
equality does not take into account who needs parks the most (Boone, Buckley, Grove, &
Sister, 2009). Rather, it is most appropriate to use the concept of equity that, although
harder to measure, considers the demographic groups who are more park dependent
(Boone et al., 2009).
As my literature review will show, low-income ethnic minority groups might live
as close or closer to parks than more privileged groups, but low-income communities of


3
color have less park acreage per person and lower quality parks, depending on the city.
These inequities in access to parks for different ethnic and income groups show that in
many U.S. cities park planning, land use, and housing policies failed, intentionally or
unintentionally, to provide equal access to parks to all children and teenagers. These
inequities in recreation are part of a wider history of ethnic discrimination that ethnic
minorities have experienced in the United States, including housing, employment, and
education (Boone et al., 2009; Wolch, Wilson, & Fehrenbach, 2005).
The recent ruling of the U.S. Supreme Court on the interpretation of 1968 Fair
Housing Act case established that allegations of ethnic discrimination are not restricted to
discriminatory intents (U.S. Supreme Court, 2015). Rather, discriminatory outcomes, like
the exclusion of ethnic minorities from certain neighborhoods, are sufficient evidence to
make claims about the existence of discrimination (U.S. Supreme Court, 2015). To make
allegations of ethnic discrimination in housing, it is sufficient to show evidence about the
effect of exclusionary zoning (residential segregation) without finding evidence about
intents to exclude specific demographic groups. The 1968 Fair Housing Act adopted a
similar approach in regards to discrimination against ethnic minorities (U.S. Supreme
Court, 2015). In this dissertation, I use this approach to define evidence of discrimination.
Thus, to claim inequity and discrimination in recreation, evidence about inequitable
distribution of parks is sufficient regardless of clear discriminatory intent in policies and
practices.
Furthermore, safety concerns in low-income and ethnic minority communities
create additional barriers to accessing parks for children and teenagers (Evans, 2004;
Gomez, Johnson, Selva, & Sallis, 2004; Platt, 2012). Regardless of income level, young


4
peoples use of parks and their benefits differ depending on park design, including the
presence of different play amenities (Dyment, Bell, & Lucas, 2009; Potwarka, Kaczynski,
& Flack, 2008; Ries et al., 2008; Staempfli, 2009; Veitch, Salmon, et al., 2014), and on
park safety (McCormack, Rock, Toohey, & Hignell, 2010; Platt, 2012; Ries et al., 2008;
Stodolska, Shinew, Acevedo, & Roman, 2013).
These bodies of literature establish that planning for young peoples access to
urban nature has park planning, land use planning, housing, public health, child
development, and environmental justice implications. However, no previous study on
access to parks has integrated planning policies, planning and design practices, young
peoples perceptions and use of parks, and young peoples socio-economic and ethnic
characteristics in the investigation of the spatial distribution of urban parks across income
and ethnic groups. For example, most equity mapping literature focusing on access to
public amenities, with few exceptions (see Boone et al., 2009), did not investigate the
reasons behind the spatial inequalities in the distribution of parks. According to
Schlosberg (2004), environmental justice investigations should focus both on the spatial
distribution of public amenities and environmental hazards, and on the decision-making
processes to determine where amenities and hazards are located. Also, no previous
investigation on access to parks has integrated qualitative research on children and
teenagers use and perception of parks in a Geographic Information System (GIS) spatial
analysis. Then, very few studies have investigated access to parks differing in terms of
size, play amenities, and presence of nature, which are all features that influence
childrens and teenagers play behaviors and perceptions of parks. I argue that the
concurrent focus on whether decision-making processes, park distribution, and park


5
quality are just in relation to childrens and teenagers needs is necessary to gain a deeper
understanding of the intersections between park distribution, the policies and practices
that created such distribution, and park needs. This is the focus of my dissertation.
This dissertation also seeks to address a general gap in the discourse on
sustainable planning. According to Campbell (1996), planners should seek sustainable
solutions that balance the often competing instances of environmental quality, economic
development and social equity, which are known as the three Es of sustainability. Thus,
as parks can be considered a form of sustainable infrastructure (Byrne & Wolch, 2009;
Chiesura, 2004; Cranz & Boland, 2004), their establishment and maintenance should
reflect the three Es framework: bringing environmental and economic benefits to cities
and their people while serving the populations who need them most (equity). However,
while several studies have investigated the environmental and public health benefits of
sustainable infrastructures like parks, greenways, and bike trails (Bedimo-Rung, Mowen,
& Cohen, 2005; A. C. K. Lee & Maheswaran, 2011; Mailer et al., 2009; McCormack et
al., 2010; Potwarka et al., 2008; Sherer, 2006), less research has addressed whom these
sustainable infrastructures are really benefitting (Boone et al., 2009; Wolch et al., 2005).
In other words, not enough research has investigated the beneficiaries of sustainability
interventions like parks. Therefore, this dissertation is also motivated by the need to
highlight the key role of environmental justice in sustainability discourses and practices
(Boone, 2010), also following Agyeman and Evans' (2004) call for just sustainability
(p. 155).
To address these gaps, the general goal of this research is to understand how
planning policies and practices in Denver, Colorado affected the way children and


6
teenagers of different demographic groups can access public parks, and specifically
public parks with high quality and maintenance. To do so, I focus on the planning
policies and practices that determined childrens and teenagers access to significant play
opportunities in urban parks in relation to their socio-economic status and ethnicity.
Building on a pilot study (Rigolon & Flohr, 2014), I conduct a mixed research
investigation (Teddlie & Tashakkori, 2009) to uncover how park planning, land use
planning, and housing policies and practices in Denver have influenced the current equity
or inequity of the distribution of parks, with a specific focus on parks with high-quality
amenities for children and teenagers.
The first strand of the study, referred as Park Planning Processes, focuses on how
park planning, land use planning, and housing policies and practices in Denver have led
to todays parks distribution (analysis of procedural justice). In particular, I study whether
park planning, land use planning, or housing policies and practices have had the biggest
relative weight in influencing the distribution of parks and of high-quality parks in
relation to the location of different demographic groups. To gain a general understanding
of the equity or inequity of planning processes that influenced young peoples access to
parks and high-quality parks, I explore whether certain demographic groups were
intentionally left out from access to parks (for example, planners located parks most often
in richer neighborhoods) or whether in other circumstances planners tried to fill the gaps
in park distribution and deliberately located parks in areas of the city with higher
concentrations of low-income ethnic children and teenagers. Also, to understand
historical processes of park location in relation to demographics, I mapped the residential


location of African Americans and Hispanics in relation to park location in several
moments of Denvers history.
7
In the second strand of the study, referred as Park Spatial Distribution, I employ a
GIS spatial analysis of park distribution, with a specific focus on parks that, based on the
qualitative literature on childrens and teenagers outdoor play, include significant play
spaces for children and teenagers (analysis of distributional justice). In particular, I use
the literature on young peoples play in parks, including how young peoples play
preferences differ by ethnicity, to develop a Park Quality Index for Youth (PQIY), also
building on a park typology I previously developed (Rigolon & Flohr, 2014). My GIS
spatial analysis focuses on the relationships between socio-economic levels, ethnic
variations and percentage of population under 18 (independent variables) and access to
parks defined as park proximity, park acreage, and park quality (dependent variables,
considered one at a time).
Finally, I integrate the findings of the two strands by looking at the way park
planning, land use planning, and housing policies and practices have influenced todays
park distribution in Denver, with a specific focus on parks suitable for childrens and
teenagers play. In particular, I connect specific policies to their spatial outcomes i.e.,
park location or land uses in different areas, to uncover relationships between procedural
and distributional justice. These activities cover all constructs I mentioned above,
including planning policies, planning and design practices, young peoples perceptions
and use of parks, young peoples socio-economic status and ethnicity, and the spatial
distribution of parks, through a fully mixed research study (Leech & Onwuegbuzie, 2009),
thus addressing the main research gap I identified.


8
In this chapter, I discuss what makes Denver an interesting case study to
investigate access to urban parks for young people. Then, I briefly introduce the most
significant findings that emerged from my study, I discuss why these findings are
relevant for urban planning, and define some key terms I use in this dissertation.
Denver as a Compelling Case Study
This place-specific environmental justice study on access to parks focuses on
Denver, Colorado, which is situated at the foothills of the Rocky Mountains, in the Front
Range metropolitan area. Denver offers a compelling site to investigate access to parks
because of its demographics, including ethnic diversity and a rapidly growing population,
its highly rated park system, its history of racial and ethnic segregation, some current
documented environmental justice issues, and the possibility of extending Denvers
findings to other cities in the American West.
Denvers Population Growth and Park System
Denver is a relevant case study because it has a propensity to attract residents that
are avid parks and recreation users and is a recognized leader in sustainable planning and
livability (Godschalk, 2004). Denver is also one of the top ten fastest growing U.S. cities
(Carlyle, 2014) with an estimated 17.3 percent population increase between 2000 and
2013 (United States Census Bureau, 2013a). Denvers population density in its urban
core is also growing (Jaffe, 2013) and this will likely increase the need for high-quality
parks in Denvers urban core.
The analysis conducted by Park Score shows that Denvers park system is close to
excellent (ParkScore, 2014). However, the current park acreage (4,233 acres) does not


9
meet the National Recreation and Parks Associations standard of 10 acres of park per
1,000 inhabitants (Mertes & Hall, 1996), due to the growing population (649,495; United
States Census Bureau, 2013). The origins of the city park system date to the late 1800s
and include historic parks like City Park, Civic Center Park, designed by Frederick Law
Olmstead, Jr., and Cheesman Park (Snow, 2009; The Trust for Public Land, 2010).
Denver was also an early adopter of greenways for recreation and preservation (Cranz &
Boland, 2004; Searns, 1995). Denvers park system includes Denvers Mountain Parks,
which are open spaces in the foothills of the Rocky Mountains owned by the city of
Denver (Denver Parks and Recreation, n.d.). This further shows the city's focus on land
preservation and outdoor recreation. Also, Denvers guidelines for park accessibility state
that each resident should live within six blocks of a park and that there should not be
physical barriers between a residents home and a park, such as highways or railways
(Harnik & Simms, 2004). In terms of current park funding, Denvers mayor plans to
spend 10 million dollars in 2015 for capital improvements of parks, with a commitment
to creating equitable opportunities in every area of the city (Proctor, 2014).
Denvers Demographics
Denver is a relevant case study to investigate access to parks as an environmental
justice issue also for its demographics. Among Denvers 650,000 inhabitants in 2013,
53.6 percent identified as Non-Hispanic White alone, while the remainder 46.4 percent
includes Hispanic or Latinos alone (30.9%), African Americans alone (10.2%), Asians
alone (3.8%), American Indians (2%) and mixed ethnicities (3.1%; United States Census
Bureau, 2013). This ethnic composition shows that Denver is an ethnically diverse city
with almost an exact split between Non-Hispanic Whites and other ethnic groups. Indeed,


10
the percentage of ethnic minority people (including Hispanics, African Americans,
Asians, Native Americans and Pacific Islanders, and people of mixed ethnicity) is higher
than the Colorado average (United States Census Bureau, 2013a) and the national
average (United States Census Bureau, 2010). Denver also has a long history of African
American and Hispanic residents, dating back to its early urban history (de Baca, 1998;
Stephens, Larson, & Black American West Museum, 2008).
Ethnic diversity is even higher in the K-12 school population: Only 22.5 percent
of Denver Public Schools students are Non-Hispanic Whites, and Hispanic young people
are 57.5 percent of the student population (Denver Public Schools, 2013). Neighborhood
parks are particularly important for Denvers ethnic minority young people. Eighty
percent of Denver Public Schools students have never been to the Rocky Mountains
(Cheek, 2015), yet the mountains are less than 20 miles away from the city.
In terms of socio-economics, the median household income, $49,091, is lower
than the Colorado average, reflecting its higher percentage of ethnic minorities (United
States Census Bureau, 2013 a), with approximately 15% of households living with less
than $15,000 per year and about 11% of households making more than $100,000 per year
(United States Census Bureau, 2012). Around 72% of Denver Public Schools K-12
students qualify for free or reduced lunch (Denver Public Schools, 2013). These
variations in ethnicity and household income, combined with very high percentages of
ethnic minority and low-income children and adolescents in the K-12 student population,
make Denver an interesting case study for an environmental justice investigation.


11
Denvers History of Ethnic Residential Segregation
Denver is an appropriate city to study access to parks across income and ethnic
groups because of its history of residential segregation. From the early twentieth century
until nowadays, Denver experienced significant residential segregation between Non-
Hispanic Whites, African Americans, and Hispanics due to several land use policies,
housing practices and racial intimidation (Abbott, 1978; N. Hernandez, 2006; R. L.
Simmons & Simmons, 1995). This phenomenon was not uncommon in American cities
during the same years (Boone et al., 2009; T. H. Simmons & Simmons, 2010; Wyly &
Holloway, 2002).
In Denver, starting in the 1920s African Americans were de facto segregated in a
few neighborhoods located northeast of downtown due to racially restrictive covenants,
intimidation by the Ku Klux Klan, redlining by the Homeowner Loan Corporation, and
real estate practices (N. Hernandez, 2006; R. L. Simmons & Simmons, 1995; T. H.
Simmons & Simmons, 2010). Starting in the 1930s, Hispanics concentrated in the
northern and western parts of the city, due to low land value and to subdivisions
containing small lots; also they did not experience the same openly discriminatory
housing practices that African Americans faced (Abbott, 1978; Duran, 2011; Langegger,
2012). Ethnic and racial groups in Denver remained significantly segregated until the
current times, as my study will show.
Residential segregation matters for access to parks, as previously mentioned. If
underprivileged groups are concentrated in certain areas of a city, policy makers and non-
profit organizations can intentionally target park investment or disinvestments in such
areas (Boone et al., 2009; Downey, 2007). Also, when racial and ethnic minorities were


12
constrained in certain areas of the city, these areas were become overcrowded due
exclusion from other areas (R. L. Simmons & Simmons, 1995), thus making it harder for
cities or community groups to acquire land to establish new parks.
Denvers Environmental Justice Issues
A small but growing body of research shows that in Denver access to
environmental goods and exposure to environmental hazards are unequally distributed
across income and ethnic groups. In particular, research has established that in metro
Denver ethnic minority and low-income communities live closer to toxic hazards,
including static sources of atmospheric pollution (Harner, Warner, Pierce, & Huber,
2002; Shaikh & Loomis, 1999), and that Denvers wealthier neighborhoods have
significantly more vegetation than low-income areas (Mennis, 2006). Two recent studies
documented other spatial inequalities in Denver. A pilot study I conducted with a
colleague on a small sample of Denver neighborhoods showed that low-income
communities of color have significantly less access to parks and to parks with play
amenities than prevalently mid- and high-income Non-Hispanic White neighborhoods
(Rigolon & Flohr, 2014). Also, a paper by Nemeth and Ross (2014) showed that medical
marijuana retailers are mostly found in areas with high concentrations of low-income and
ethnic minority residents and that this distribution is mostly due to city regulations.
Also, the Piton Foundation, a Denver philanthropic organization, identified areas
of the city with high concentrations of at risk children, including risks of being in poverty,
attending low-performing schools, having parents with low educational levels, lacking
access to healthy food, and living in unsafe areas (The Piton Foundation, 2012). These
areas are concentrated in the northeastern parts of the city and have been named the


13
Childrens Corridor (The Piton Foundation, 2012). This concentration of at risk
children in certain areas, as highlighted by the advocacy of philanthropic organizations,
makes Denver an even more interesting case study for childrens access to urban parks.
Generalizability
Denver is in some ways representative of other cities in the American West,
including cities like Portland, OR, Seattle, WA, Salt Lake City, UT, Albuquerque, NM,
San Francisco, CA, and Oakland, CA, but partially differs from other cities in the East
Coast for which access to parks has been extensively studied, including Baltimore, MD
and New York, NY. The aforementioned cities located in the West are similar to Denver
for their comparable population size, although of varying ethnic composition (United
States Census Bureau, 2010), while most cities I listed above are comparable to Denver
for their historical patterns of development. The latter include a dense commercial
downtown developed early in the cities history; de facto residential segregation among
income and ethnic groups; public improvements undertook within the City Beautiful
framework, comprising the construction of parks and parkways; interstates cutting
through central parts of the city, often displacing low-income communities of color;
patterns of metropolitan growth in the post World War II period; and recent urban
renaissance trends connected to the arrival of the creative class. All these cities are also
experiencing steady and in some cases fast population growth in the last decade (United
States Census Bureau, 2013a); thus, they will face similar challenges to Denver in terms
of park provision.
However, cities located in the East Coast such as Baltimore and New York are
dissimilar from Denver for their earlier period of historical development, which led to


14
establishing the first parks within different political, social, land use and technological
contexts. Also, East Coast cities have historically been segregated along the white-black
line (Boone et ah, 2009; Romero, 2004), while Denver has been a tri-ethnic city from its
relatively early history, including Non-Hispanic Whites, African Americans, and
Hispanics, (Abbott, 1978; Romero, 2004). This adds complexity to the study of access to
parks in Denver.
Furthermore, Denver is a compelling case study for the size of its population.
Approximately half of the people residing in the worlds urban areas live in cities that are
smaller than 500,000 inhabitants (United Nations, 2014), which are comparable to
Denver in terms of population (United States Census Bureau, 2013a). Finally, the
methodology I use in this dissertation is generalizable beyond the aforementioned cities.
To summarize, the dual nature of Denver having progressive urban planning
including a good park system, though one that has not kept up with growth, and at the
same time presenting documented environmental justice and residential segregation
issues combined with its demographic profile and the possibility to generalize findings
to other cities, makes it a compelling case study to investigate the influence of park
policies and practices on childrens and teenagers access to urban parks as an
environmental justice issue.
A Complex Picture of Environmental Injustice
My analysis will show a complex picture of environmental injustice, including
cases of distributional and procedural inequities. The analysis of park planning processes
(see Chapter 4) will show that, in different historical periods of Denvers history, more
cases of discriminatory, neglectful, and color blind processes have occurred than


15
initiatives oriented to increase park equity, which are actions aimed to increase park
provision for low-income people, ethnic minorities, and young people. The policies and
practices that most significantly fostered or limited park establishment have included
park funding systems, City Beautiful visions in the early decades of the twentieth century,
a series of municipal policies on parkland dedication or cash fees for infill and greenfield
developments, and a state policy restricting Denvers annexation capacity, the
Poundstone Amendment to the Colorado Constitution. In particular, my research will
show that the Poundstone Amendment, a discriminatory policy implemented to prevent
ethnic integration in schools, has negatively affected Denvers capacity to establish large
parks beyond its affluent neighborhoods, thus indirectly affecting park equity. Also, my
analysis will highlight that young people have typically been excluded from park design
considerations, as economic reasons and the aesthetic values of adults have often
prevailed.
Findings of the geospatial analysis (see Chapter 5), expressing the spatial
outcomes of the policies and practices I studied, will show more cases of distributional
inequity than distributional equity. Among the inequities, I found that African Americans
are the most disadvantaged ethnic group, particularly in relation to park acreage and park
quality. The cases of partial equity include Hispanics and low-income groups having
good access in terms of park proximity (distance to the closest park), and to parks with
good enough levels of quality. However, high-income groups and Non-Hispanic
Whites have access to more acres of parks, to more parkways, to more parks with
excellent quality, and to more safe parks than other groups. In particular, low-income
ethnic minority young people have very low access to parks with low violent crime


16
density, or safe parks, which significantly limits their park visitation. This is the most
concerning inequity issue in Denvers park spatial distribution.
The integration between qualitative and geospatial quantitative datasets showed
some of the most interesting findings (see Chapter 6). Discriminatory and neglectful
processes can often explain inequities in park spatial distribution. Yet, inequitable
processes were not always directed at creating discrimination in recreation but park
inequity has been a consequence of discrimination in housing and education. Fiscally
conservative policies have also had significantly negative impacts on park equity. On the
other hand, partially equitable distributions of parks benefitting low-income groups and
Hispanics have been rarely the results of equity-oriented processes. Rather, class and
color blind initiatives, combined with residential relocation waves have had a stronger
role in partially equitable park spatial distributions. From the interviews with park and
land use planners, and from the analysis of city documents, I noticed that initiatives
aimed to create equal park provision across geographical areas of the city have been
significantly more common than efforts aimed to increase park provision for low-income
communities of color.
As discussed, discriminatory outcomes, evidenced by inequities in park spatial
distribution, are sufficient to make claims about discrimination in recreation in Denver,
based on the recent ruling of the U.S. Supreme Court regarding the 1968 Fair Housing
Act (U.S. Supreme Court, 2015). However, the integration between the analysis of
planning processes and their spatial outcomes gives a more comprehensive picture of
how ethnic discrimination in recreation has been perpetuated in Denver.


17
Relevance to Urban Planning
The results of my dissertation can provide insights to help planning practitioners
and scholars, community organizations, and underprivileged communities address the
environmental justice issues of access to parks that have been created by past policies and
practices. Its main merits are twofold. First, this dissertation takes a critical look at
planners' participation in children's and teenagers experiences of environmental injustice
in recreation and provides communities with tools for engaging the planning process.
Second, the dissertation proposes to bring a focus upon children's and teenagers
experiences into a field that has overlooked them. For example, the American Planning
Association has divisions focusing upon gays, lesbians, women, Latinos, and African
Americans but not children and/or teenagers (American Planning Association, n.d.-b).
More specifically, the results of this study can inform different planning
disciplines, both in terms of scholarship and practice. In particular, the analysis of park
planning processes highlights the complex procedures that led to park locations and
funding, suggesting more equitable practices centered on environmental justice. With this
in mind, Denver Parks and Recreation can use the results of this dissertation to develop
participatory suitability analyses for new park locations or for improving existing parks
without significant play amenities and natural elements. In Denver and other cities, parks
and recreation departments can use the results of the literature review of young peoples
outdoor play to update park guidelines and standards in terms of types of play amenities,
presence of natural elements, and park size. Related to this, the Park Quality Index for
Youth (PQIY), including several indicators of quality, can be used to create park
classifications that take into account young peoples preferences and needs.


18
The results of the Park Planning Process analysis also illuminate the intersections
between park, land use, and housing policies and practices, as the current distribution of
parks across income and ethnic groups are often due to residential segregation (Boone et
al., 2009). These intersections will highlight ways for park planners, land use planners,
and housing authorities to collaborate to solve these interrelated issues (see Chapter 7,
Implications). For example, investments to expand or improve affordable housing and
parks should be coordinated through interrelated suitability analyses. In terms of zoning,
the results of this study suggest the possibility of allowing higher residential densities
around parks, as parks can be seen as public resources that more residents can share. The
analysis of park planning processes also highlights economic reasons behind park
location, which raise park equity issues.
This research can benefit low-income communities of color and the community
organizations that advocate for their rights. As Bocarro and Stodolska (2013) argued that
leisure research related to social justice should generate findings that can be used to bring
about change, I further developed a framework for action. I conceptualized this
framework earlier (Rigolon & Flohr, 2014) to help low-income and ethnic minority
communities conduct park advocacy efforts to increase park provision in their
neighborhood. The framework includes a set of recommendations that can be applied in
practice (see Chapter 7, Implications). To refine my framework, I drew from the results
of my previous study (Rigolon & Flohr, 2014), on the implications of this dissertation,
and reviewed further relevant literature (Anguelovski, 2013; Arredondo et al., 2013;
Garcia, 2013; Hou, Johnson, & Lawson, 2009; Hoyt, 2013; Schilling & Logan, 2008;
Tidball & Krasny, 2007; Westphal, 2003). Also, I will disseminate my findings among a


19
series of Denvers neighborhood organizations to raise their awareness about how to
increase access to high-quality parks in their neighborhoods.
This dissertation advances the equity mapping and environmental justice
literatures in several ways. I employ a finer method to measure access to play in parks
than most previous research by using a combination of census block and parcel data.
Then, the focus on children and teenagers and the use of the new Park Quality Index for
Youth (PQIY) based on the literature on young peoples outdoor play advances previous
equity mapping research because it introduces childrens and teenagers perspective on
parks in a quantitative spatial analysis. Finally, the analysis of park planning processes
builds on Boone et al.'s (2009) analysis and enriches it by using a combination of
interviews, reviews of books and articles, and analyses of primary documents.
Finally, this study will advance the use of mixed methods research in planning
studies by combining qualitative and quantitative geospatial approaches and data in a
fully integrated mixed methods research design. My mixed research approach also will
allow me to study the relationships between constructs that were never studied together in
a systematic way, such as park distribution, parks perceptions and use, young peoples
socio-economic status (SES) and ethnicity, and planning policies and practices. This
dissertation will include a critical reflection on the methods used in this study to share the
methodological lessons I learned from this study with other planning scholars (see
Chapter 7, Implications).
Definition of Key Terms
To frame this dissertation, a few key terms and phrases used to describe spaces
and individuals need to be defined. First, I use the phrase urban green space to describe


20
all accessible green spaces in a city, including land that is owned and managed by the city
such as public parks, greenways, gardens, and other public open spaces; land that is
owned by other public institutions such as accessible schoolyards, community center
fields, and playgrounds in housing projects; and private spaces that are open to the public
through right-of-way agreements such as those owned by homeowner associations. Then,
urban parks or parks include all public green spaces designated as parks in a citys
limits. Urban parks are publicly funded, planned, designed and managed by a citys Parks
and Recreation Department.
I use access to parks to indicate the possibility to enjoy any type of park. Also, I
include in this concept play opportunities in playgrounds, sport fields, and other spaces
found in urban parks can all provide contact with natural elements, such as trees, grass,
and rocks. Indeed, Kaplan (1984) claimed that nature can be found in cities in the form of
parks, trees, plants, and water. Natural places and elements that can be experienced in
cities on a daily basis were defined as nature-at-the-doorstep (Kaplan, 1984, p. 189).
The definition of nature adopted in this dissertation, including the fact that nature can be
found in cities, is widely accepted in the planning and environmental psychology
literatures (Baur, Tynon, & Gomez, 2013; Beatley, 2011; Herzog, 1989; Sullivan, Kuo, &
Depooter, 2004). Thus, in this dissertation every reference to parks is a reference to urban
nature.
Although there are no generally accepted definitions of children, adolescents,
and youth (Ansell, 2005), I use the term child to describe an individual aged zero to
11, as 11 corresponds to the last year of elementary or primary education (The World
Bank, 2013) and on average to puberty (P. A. Lee, Guo, & Kulin, 2001). I use the term


21
adolescent or teenager to describe an individual aged 11 to 17, as 18 represents the
age at which individuals become legally adults in most Western countries (United
Nations Childrens Fund, 2011) and finish their secondary education (UNESCO Institute
for Statistics, 2005). Then, I use the phrase young people to group children and
adolescents together. The distinction between children and adolescents is critical to play,
as different age groups tend to have different play preferences (McCormack et al., 2010;
Ries et al., 2008) and different mobility licenses (Soori & Bhopal, 2002). I discuss how
childrens and teenagers play preferences differ in my review of young peoples outdoor
play and park visitation (see Chapter 5, Park Spatial Distribution Findings).
I use the term parent to refer to the biological or adoptive parent of the young
people. I use the word caregiver to refer to any guardian, related or unrelated to the
young person, including grandparents, older siblings, aunts and uncles, and other
guardians. Thus caregiver is more comprehensive than parent and describes
supervising adults of any kind.
In terms of ethnicity, I chose to focus mostly on three ethnic groups: Non-
Hispanic Whites, Hispanics, and African Americans. Denver has often been considered a
tri-ethnic city (Abbott, 1978; Romero, 2004) and Non-Hispanic Whites, Hispanics, and
African Americans still constitute the three largest ethnic groups in the city (United
States Census Bureau, 2013a). Also, I use the term Hispanic as a general term to
describe people who identify as Latino/Latina, Chicano/Chicana, Mexican descent, or
Hispanic. Although historically these terms have had different meanings and
connotations (Langegger, 2013), I use the general term Hispanic to simplify the
discussion. Similarly, the U.S. Census regroups people who identify as Latino/Latina,


22
Chicano/Chicana, Mexican descent, or Hispanic in a general category named Hispanics
(United States Census Bureau, 2013a). Also, I use capitalized terms for ethnic groups
defined by the U.S. Census (e.g. African Americans), while I use non-capitalized terms to
refer to groups of ethnicities (e.g. people of color and ethnic minorities).
In the following chapters of this dissertation, I will define my theoretical
framework, review the literature on young peoples access to parks based on the
framework, highlight gaps in the literature, and lay out my research goals and questions
(Chapter 2). Then, I will describe the research design and the specific methods I used to
investigate access to parks for young people in Denver (Chapter 3). I will present the
findings of the analysis of park planning processes in Chapter 4 and the findings of the
spatial distribution of parks in Chapter 5. In Chapter 6,1 will integrate the findings on
planning processes and park distribution. Finally, in Chapter 7,1 will discuss the
implications of my findings for park and land use planning, for environmental justice,
and for the use of mixed methods research in planning.


23
CHAPTER II
THEORETICAL FRAMEWORK, LITERATURE REVIEW, AND RESEARCH
QUESTIONS
Theoretical Framework
The problem laid out in Chapter 1 identifies the constructs this dissertation will
investigate, namely planning policies, planning and design practices, young peoples
perceptions and use of parks, young peoples socio-economic and ethnic backgrounds,
and the spatial distribution of parks. I first define these constructs in Table 1 and then
discuss their potential relationships in the next section.
Identification of Constructs
The constructs presented in Table 1 are similar to the ones that Byrne and Wolch
(2009) introduced in their conceptual model named space, race, and park use (p. 751).
In particular, their model includes the historical and cultural context of park provision,
park space, perceptions, potential users, and park use choices, comprising use
and non-use (Byrne & Wolch, 2009, p. 751). This framework does not focus
specifically on children or teenagers, but it can be useful to define the constructs included
in Table 1 better, and to explore some of the potential connections among the constructs
that influence park perceptions and use. For example, the historical context of park
provision includes aspects like spatialized ethno-racial discrimination, the ideology of
land use, the history of property development, park politics, the ideology of park
provision, and racial politics of park development (Byrne & Wolch, 2009, p. 751). In
other words, this construct comprises a series of cultural, policy, historical, and racial


24
aspects related to land use and park planning. I further discuss Byrne and Wolch's (2009)
model when introducing the potential connections among the constructs identified in
Table 1.
Table 1
Constructs Included in the Theoretical Framework
Construct Description
Planning Policies Public regulations affecting the built environment at different scales as the result of decision-making processes involving stakeholders with different visions. I focus on policies addressing land use, zoning, park location and funding, affordable housing location and funding.
Planning and Design Practices Actions undertaken within policies frameworks to transform space or change its use, sometimes by challenging the regulations imposed by policies. I focus on how plans for communities, parks, and housing were implemented.
Young Peoples Perceptions and Use of Parks The way children and teenagers perceive and use their neighborhood parks, often related to their sense of safety and to their caregivers regulatory strategies.
Young Peoples Socio- Median household income combined with assessed parcel
Economics and Ethnicity value of a young persons home (as an asset of the family)
and ethnic self-identification. I chose to focus on ethnicity,
Spatial Distribution of Parks instead of race, given Denvers large Latino population. How far different types of parks are from young peoples homes. Park spatial distribution is a spatial quantitative measure and is one of the variables influencing young peoples park visitation, as distance to parks matters for young peoples use of parks.
Planning policies. According to McConnell (2010), a policy has generally been
defined as a decision or action that public institutions undertake to distribute value. In
planning, these decisions and actions undertaken by governments are intended to improve


25
the places where people live, as the American Planning Association defined planning as
discipline intended to create more convenient, equitable, healthful, efficient, and
attractive places for present and future generations (American Planning Association,
n.d.-a). Policy making is a complex activity that generally affects the interests of a variety
of public and private actors (Innes & Booher, 2010). This problem-solving task usually
generates conflicts between stakeholders, which can be solved through lawsuits or, more
efficiently, through collaborative planning (Innes & Booher, 2010). The conflicts
occurring between stakeholders suggest the power dynamics that are inherent in policy
making. In the past, most policies were created by expert public administrators based on
data they collected, and had to go through political ratification (Innes & Booher, 2010).
In recent times, more collaborative policy making processes, involving different
stakeholders, have emerged (Innes & Booher, 2010). I discuss the impacts of policies on
equitable park distribution in Chapter 6.
In this study, I focus on the planning policies that influenced how urban parks are
distributed in relation to ethnic and income groups (Boone et al., 2009; Byrne & Wolch,
2009; Wolch et al., 2005), including: city-level policies about land use, zoning, park
location and funding, affordable housing location and funding, and park design and
maintenance standards; and state-level policies about park funding and annexations.
Byrne and Wolch (2009) considered the historical and cultural context of park
provision as a primary factor in determining park spatial distribution and the specific
features of each park (p. 715). In particular, they highlighted the history of ethnic and
racial discrimination that pervaded park planning in several U.S. cities in the twentieth
century (Byrne & Wolch, 2009). In my framework, Byrne and Wolch's (2009) historical


26
and cultural context of park provision can be conceptualized as planning policies, and as
planning and design practices.
Land use policies can reflect political and cultural ideologies and can be used to
spatially discriminate against ethnic and racial minorities, even in terms of access to
parks (Byrne & Wolch, 2009). In the United States, zoning has been the main planning
instrument to determine appropriate land uses (Maantay, 2001). Racial zoning ordinances
started developing after the Civil War, with African Americans migrating from the South
to the North and affluent whites trying to prevent African Americans from living close to
them (Bono, 2007). In Baltimore, Maryland, a zoning ordinance passed in 1910 created
segregated residential areas for Non-Hispanic Whites and African Americans, and other
racial zoning ordinances were approved in the following years in several American cities
(Bono, 2007). Some of these ordinances targeted specific blocks, others segregated entire
districts of a city (Bono, 2007). Clearly, these zoning ordinances aiming to separate
African Americans from Non-Hispanic Whites were an example of de jure segregation,
as segregation was created by law (Bono, 2007).
Zoning ordinances have also been used to segregate people by socio-economic
status (Hartnett, 1993; Liberty, 2002). Exclusionary zoning ordinances implemented in
several U.S. cities have dictated minimum lot sizes, minimum house sizes and set back
rules for residential units located in certain neighborhoods, as a means to exclude low
socio-economic groups from these neighborhoods (Hartnett, 1993; Liberty, 2002;
Maantay, 2001). Limitations or bans on multi-family housing were another means
through which exclusionary zoning limited low-socioeconomic groups opportunity to
live in certain neighborhoods (Hartnett, 1993; Liberty, 2002).


27
Zoning codes have also been used to locate locally unwanted land uses like
polluting industries in proximity to residential areas inhabited by low-income
communities of color (Harvey, 2009; Maantay, 2001, 2002). In particular, zoning
changes implemented in New York City between the 1960s and 1980s have increased
underprivileged groups exposure to polluting industries (Maantay, 2001). Therefore,
zoning ordinances in the U.S. have contributed to creating environmental injustice and
public health threats for low-income communities of color (Harvey, 2009; Maantay, 2001,
2002).
Planning and design practices. If policies create regulations and dictate actions,
design and planning practices are actions undertaken within policy frameworks by public
and private agencies to transform space or change its use at different scales of the
physical environment, sometimes by challenging the regulations imposed by policies. For
example, a park designer can work around the limitations imposed by park standards to
design a natural playground for childrens informal play. Like policies, planning practice
is influenced by values and ethical dilemmas (H. Campbell & Marshall, 1998) and is
intended to achieve good or better outcomes (Alexander, 2009). Planning practices
need to conceive the spatial realms they intend to modify as places rather than simple
physical settings, thus taking into account the socially constructed meanings attached to
physical settings (Graham & Healey, 1999). In this study, I focus on the work of planning
and design practitioners, and of Denvers residents, to establish parks, land uses, and
housing policies and practices, and to design and maintain parks.
Byrne and Wolch's (2009) historical and cultural context of park provision
includes several planning and housing practices that contributed to residential segregation


28
based on race and ethnicity, such as restrictive racial covenants and redlining by the
Home Owners Loan Corporation. Restrictive racial covenants were conditions attached to
properties that only allowed the sale or rent of these properties to certain racial and ethnic
groups (Boone et al., 2009; Gotham, 2000; Jones-Correa, 2000; Nelson, Sanchez, &
Dawkins, 2004). Starting from the 1920s, neighborhood improvement associations and
realtors encouraged Non-Hispanic White affluent property owners to establish covenants
that would prohibit selling housing units to non-whites and ethnic minorities (Boone et al.,
2009; Gotham, 2000; Jones-Correa, 2000; Nelson et al., 2004). Restrictive covenants
were aimed to maintain high property values for owners by keeping undesired
populations out of certain neighborhoods (Boone et al., 2009). According to Bono (2007),
these practices, which restricted housing options for ethnic and racial minorities, can be
considered forms of de facto discrimination. However, in this case there is a fine line
between de jure and de facto segregation because these covenants, although not created
by public policy, were legally enforced.
The practice of redlining deals with lending bias that derive from residential
security maps created by the Home Owners Loan Corporation (Greer, 2013; J.
Hernandez, 2009; Hillier, 2003a, 2003b; Squires, 2003). Starting in the 1930s, inner-city
areas with deteriorating housing stock, which were generally inhabited by low-income
African Americans, were defined as too risky for lending and bounded by red lines in the
Home Owner Loan Corporation maps (Greer, 2013; J. Hernandez, 2009; Hillier, 2003a,
2003b; R. L. Simmons & Simmons, 1995; T. H. Simmons & Simmons, 2010; Squires,
2003). African Americans living in these areas could not receive mortgages to buy or
improve their houses, which led to the quality of housing further deteriorating, and


29
discriminatory lending practices persist even today (Hillier, 2003a, 2003b; R. L.
Simmons & Simmons, 1995; Squires, 2003; Wyly & Holloway, 2002). On the other hand,
African Americans who attempted to leave the redlined areas were often the object of
intimidation and violence, including actions by the Ku Klux Klan, and of real estate
practices like racial steering (N. Hernandez, 2006; Leonard & Noel, 1990; Mohl, 1995;
Morrison, 2006; R. L. Simmons & Simmons, 1995).
Design practices and the culture in which they are based influence the aesthetic
qualities of parks and the activities that parks support. For example, Byrne and Wolch
(2009) highlighted that traditionally most parks in American cities were designed based
on English and northern European canons of beauty, which reflect picturesque ideals of
pastoral landscapes, and which are not appealing for all ethnic or racial groups.
Young peoples perceptions and use of parks. The ways young people perceive
and use their neighborhood parks and green spaces are important aspects to consider
when studying access to urban parks as an environmental justice issue (Byrne & Wolch,
2009). As discussed earlier, park design and quality influences childrens and teenagers
use of parks (Byrne & Wolch, 2009; Potwarka et al., 2008; Veitch, Salmon, et al., 2014).
Some parks might not serve the specific play and recreation needs of certain age or ethnic
groups (Byrne & Wolch, 2009). Else and perceptions are often related to childrens and
teenagers sense of safety, as neighborhood violence or the presence of gangs in parks
often limit young peoples access to parks in low-income neighborhoods (Gomez et al.,
2004; Platt, 2012). Caregivers can hold the same concerns for safety and therefore restrict
their childrens access to outdoor play in neighborhood parks (Valentine, 1997; Veitch,
Bagley, Ball, & Salmon, 2006).


30
Even when young people and their caregivers hold positive perceptions of parks,
other factors limit childrens and teenagers park visitation. Besides barriers related to
distance, which I discuss below, children and teenagers might not use parks because of
time constraints, physical abilities, and other competing recreational activities (Byrne &
Wolch, 2009). Thus, understanding young peoples perceptions and use of parks is
important as is identifying young peoples preferences for different types of parks and
young peoples and caregivers sense of safety in their neighborhood. To operationalize
young peoples perceptions and uses of parks, I create a Park Quality Index for Youth
(PQIY) based on the literature on young peoples outdoor play and park visitation.
Young peoples socio-economic and ethnic characteristics. This construct uses
demographic data from the U.S. Census and from the City and County of Denver to
operationalize childrens and teenagers socio-economic status and ethnicity. In particular,
to operationalize socio-economic status I employ a combination of median household
income from the U.S. Census, the assessed parcel land value of young peoples homes
from the City and County of Denver, and other variables on home ownership from the
U.S. Census (see Chapter 3, Research Methods). To operationalize young peoples
ethnicity, I use the familys ethnic self-identification from U.S. Census data. Further
details about these constructs are given in Chapter 3 (Research Methods). Also, I focus
on ethnicity, instead of race, given Denvers large Hispanic population (around 30%),
even larger K-12 student Hispanic population in Denver Public Schools (around 57%),
and strong cultural identification that includes different races.
Spatial distribution of parks. This construct describes where parks are located in
different parts of the city and how they serve different populations. This spatial


31
distribution determines the distance between parks and every childs and teenagers home.
In this study, access is a quantitative spatial concept related to the geographic distribution
of parks i.e., having the opportunity to enjoy settings that are open to the public and that
are within walking distance from young peoples homes. Therefore, I study access to play
in parks as an opportunity-related concept or, in other words, in terms of potential [park]
users (Byrne & Wolch, 2009, p. 751). However, I recognize that distance and parks
spatial distribution is not the only factor influencing young peoples actual access to
parks. As the following sections and the literature review show, other factors affecting
actual park visitation for young people include physical barriers like arterial roads and
railways, park design, social factors, perceptions of safety, and personal factors related to
age, gender, and culture.
The classic work of two psychologists, James Gibson and Roger Barker, and more
recent developments in environmental psychology and critical geography, help clarify the
way I define young peoples access to play in parks. James Gibson (1979) formulated the
theory of affordances, which he defined as all the possibilities of action inherent in the
objective physical environment, in relation to an actor and dependent on the actors
capabilities, including physical skills, intentions, and knowledge. For example, a tree
branch five feet high does not afford the act of climbing for an infant, but can afford the
same act for an adolescent. While the environment potentially includes an infinite
number of affordances (Kytta, 2004), actualized affordances are a subgroup of all
potential affordances that includes what an individual notices, uses, or modifies (Heft,
1989). Therefore, affordances can be conceptualized as ordered features (Greeno, 1994),
including potential, perceived, utilized, and shaped affordances (Kytta, 2002, p. 109).


32
The degree to which affordances are noticed, used or modified depends on individuals
capabilities, but also on collective and cultural norms and expectations (Reed, 1993).
Excessive distance from parks with significant play amenities is an example of
how the physical environment may or may not afford playing in nature, depending on
individuals skills and parental regulatory practices. Distance from parks matters because
children spend most of their time around home (Loebach & Gilliland, 2014) and because
children generally use active forms of transportation to access parks (Veitch, Carver, et
ah, 2014). Indeed, a study conducted in two Swedish small towns shows that children
often use playgrounds that are located close to their home (Jansson & Persson, 2010).
Therefore, the theory of affordances helped me conceptualize access from a spatial
quantitative point of view.
After defining access to play as the opportunity to reach a park, it is important to
understand which types of settings within parks can support play for young people. In
terms of physical features, the theory of affordances can also help explain in what ways
physical settings influence behaviors. For example, a very steep hill might not afford
running uphill (potential or perceived affordance) but affords rolling downhill (actualized
affordance). However, play behaviors can also be explained by social norms and
expectations. In this regard, Roger Barker (1968) proposed the theory of behavior settings,
which holds that in certain physical settings, behaviors are influenced by social
expectations and rules, which vary greatly by culture (Barth, 1998). For example,
classrooms host learning activities while sport fields host athletic games (Ding, Sallis,
Kerr, Lee, & Rosenberg, 2011). Thus, according to Barker's (1968) behavior settings,
behaviors match the environment through social rules. However, childrens perception


33
and use of space differs from adults, as they tend to play in a variety of different settings,
often in places that were not conceived as spaces for childrens play (Matthews & Limb,
1999). For example, some studies showed that standard playgrounds are most often used
by young children accompanied by adults, suggesting that adults rather than children
choose playgrounds as play settings (Matthews & Limb, 1999). In particular, adults
attempt to limit childrens play to playgrounds is a way to segregate children in certain
parts of the city for adults convenience (Hart, 2002). Therefore, behavior settings as
places for childrens and teenagers play need to be investigated from young peoples
perspectives and in relation to adults regulatory systems.
For the purpose of this research, which is focused on access to urban nature, play
amenities located in parks can be considered behavior settings where different types of
activities can occur (Cosco, Moore, & Islam, 2010). Indeed, different types of play
amenities within a park can be associated with different levels of physical activity (Cosco
et al., 2010) and types of play (Dyment & OConnell, 2013). I introduce and discuss
different types of play amenities in Chapter 3, Research Methods.
Relationships between Constructs
After having introduced these five constructs, I discuss the potential connections
among the constructs to create an encompassing framework for my dissertation, as shown
in Figure 1. The relationships among the mentioned constructs are hypothesized based on
urban planning, environmental justice, and environmental psychology literatures, which I
analyze in detail in the literature review. Byrne and Wolch's (2009) space, race, and park
use conceptual model also provided useful insights to build the framework described (p.
751).


34
The Planning, Ethnicity, Park Distribution and Quality framework (PEPDQ,
Figure 1) is rendered as a globe because it includes complex and mutual relationships
among the constructs I study; thus, a linear one-way model would not have been
appropriate. In this study I investigate through empirical data only some of the
connections described in the PEPDQ framework, including the relationships between
policies and practices and the spatial distribution of parks, mediated by young peoples
socio-economic status and ethnicity (Boone et al., 2009; Byrne & Wolch, 2009); the
possible associations between the spatial distribution of parks and young peoples socio-
economic status and ethnicity (Byrne & Wolch, 2009; Wolch, Byrne, & Newell, 2014);
and the connection between young peoples socio-economic status and ethnicity and their
perceptions and use of parks due to neighborhood safety (Gomez et al., 2004; Platt, 2012).
The PEPDQ framework contains other connections that deserve further
exploration in future research, such as the way the spatial distribution of parks, and young
peoples use and perceptions of parks can inform future planning policies and design
(Talen, 2003); or that can be investigated in future studies, for example the connections
between the spatial distribution of parks and young peoples park perceptions and use
(Kytta, 2002, 2004). Including young peoples park preferences through a park quality
index is a key part of this study because it is a means to partially address the connection
between park availability and park visitation. As Floyd, Taylor, and Whitt-Glover (2009)
suggested, it is fundamental to include park quality aspects that have been correlated with
park visitation and physical activity in parks.


35
Figure 1. The Planning, Ethnicity, Park Distribution and Quality framework. The symbol
describes connections I investigate in this study. The symbol [] describes planning
and design implications. The symbol <> depicts future directions for research.
Transformative Paradigm and Environmental Justice
To analyze the connections among planning policies, planning and design
practices, young peoples socio-economic status and ethnicity, and the spatial distribution
of parks, I employ a combination of the transformative approach to research (Mertens,
2003) and an environmental justice lens (Floyd & Johnson, 2002; Heiman, 1996;
Schlosberg, 2004). The transformative paradigm serves as a general framework for social
change, such as creating better conditions for playing in urban nature for low-income and
ethnic minority young people, as well as a paradigmatic foundation for mixed methods
research (Mertens, 2003, 2007). Environmental justice serves as a theoretical lens for the


36
specific focus of this research, which deals with distributional and procedural injustice
(Floyd & Johnson, 2002; Heiman, 1996; Schlosberg, 2004).
Transformative paradigm. The transformative paradigm addresses issues of
power by focusing on the injustices experienced by underprivileged populations like
women, ethnic and racial minorities, disabled people, the LBGTQ community and
whoever experiences oppression, and intends research as a means to undertake or inform
action (Mertens, Bledsoe, Sullivan, & Wilson, 2010; Mertens, 2003, 2007). The
transformative paradigm is rooted in participatory inquiry, which emerged as a research
paradigm in the 1960s and 1970s (Chambers, 1994), as they both conceive reality
through the perspectives of marginalized individuals and communities (Mertens, 2009).
Participatory inquiry is a research approach in which investigators employ participatory
actions to learn about communities and to help them bring about social change (Brydon-
Miller, 1997). Participatory and activist researchers believe that research should be
conducted with people and not on or about people (Heron & Reason, 1997).
The transformative paradigm developed partially because scholars and
marginalized communities expressed frustrations about the prevalent research approaches,
which failed to represent and empower such communities (Mertens et al., 2010). Also,
the transformative approach aims to increase awareness of social injustice and to include
the opinions of marginalized communities in scholarly research, which can lead to
actions oriented to social equity (Mertens et al., 2010).
The work of Donna Mertens (Mertens et al., 2010; Mertens, 2003, 2007, 2012)
helped conceptualize the ontological, epistemological, axiological, and research practice
foundations of the transformative paradigm. Regarding ontology, transformative


37
researchers believe that multiple realities exist and that social groups construct them, but
also that social, cultural, economic, political, racial, gender, age, and disability stances
influence the way these realities are constructed (Mertens, 2007). The resulting realities
can be dissimilar because participants and investigators have various degrees of privilege
and disadvantage (Mertens, 2007). In terms of epistemology, transformative researchers
believe that participants and investigators should establish a cooperative connection,
including respect for any type of difference and recognition of power relationships
(Mertens, 2007).
Regarding the practice of research, Mertens (2007) suggested that in
transformative research the participants and the investigators should collaborate in
defining the research goal and techniques. Also, transformative research techniques
should be appropriate to welcome cultural diversity, should minimize power imbalances,
and avoid discrimination (Mertens, 2007). In terms of axiology, the transformative
paradigm is based on the acknowledgment of power imbalances between groups with
different incomes, ethnicity, sexual orientations, and physical and mental abilities, age
and on the ethical consequences that stem from these differences, such as injustice,
persecution, distortion, and negligence (Mertens, Holmes, & Harris, 2009).
Mertens (2007) also conceptualized how transformative mixed methods studies
can address societal inequities and be a tool for change. The use of both qualitative and
quantitative methods can be a way to address power issues, as participants have more
than one way to express their views (Mertens, 2007). Also, when studying social justice
issues, the complementarity of qualitative and quantitative methods can provide a broader
range of ideas and solutions about the studied issues (Hesse-Biber, 2010; Mertens, 2007).


38
For example, quantitative methods can be used to identify the groups that experience
discrimination or oppression and then researchers can use qualitative methods to better
understand the groups issues (Mertens, 2007). Further, Mertens (2007, 2012) proposed
that mixed methods can be used cyclically in transformative studies, starting and ending
the research process with participatory qualitative methods, while collecting and
analyzing quantitative data between the two qualitative phases.
In this study, which focuses on low-income young people of color, the
transformative approach is linked to concepts of social justice, which is a vision of
society in which each individual is warranted equal social, political, and economic rights
and opportunities (Bell, 2007; Gibelman, 1995), and of environmental justice (Heiman,
1996; Schlosberg, 2004), which is described in the next section. Also, since in this
dissertation I did not collaborate with the participants to define research goals and
methods due to time constraints, I am only partially using the transformative paradigm.
Environmental justice. Environmental justice, encompassing environmental
racism and environmental classism, describes the idea that populations of different
ethnicity or race and socio-economic status have access to uneven environmental quality
(Schlosberg, 2004; Schweitzer & Stephenson, 2007). The concept of environmental
justice developed from the acknowledgement that low-income and ethnic minority groups
are more exposed to environmental hazards, like air pollution from traffic, incinerators
and industrial plants, solid waste landfills, and water contamination (Boone et al., 2009;
Floyd & Johnson, 2002; Heiman, 1996; Schlosberg, 2004). However, environmental
justice also addresses inequalities in terms of access to positive infrastructure and
amenities, including housing, health, food (Heiman, 1996), recreation (Floyd & Johnson,


39
2002), and transportation (Bullard, 2003; Zavestoski & Agyeman, 2014). Several U.S.
federal agencies have implemented strategies to address environmental injustice (U.S.
Department of Housing and Urban Development, 2012; U.S. Department of
Transportation, 2012; U.S. Environmental Protection Agency, 2014), showing a growing
interest in these issues. However, two recent reviews of sustainability plans in American
cities showed that, due to policy or implementation limitations, ecological issues receive
more attention than equity issues (Pearsall & Pierce, 2010; Warner, 2002). A more
detailed discussion of the intersections between sustainability and environmental justice
is reported below.
Until recently, most environmental justice literature focused on exposure to
bads, while fewer studies have been dedicated to access to goods, including nature
(Boone et al., 2009; Floyd & Johnson, 2002). Also, most basic definitions of
environmental justice generally deal with distributional aspects i.e., that goods and bads
should be equally distributed among demographic groups, without taking into account the
processes leading to spatial distributions (Boone et al., 2009).
However, Schlosberg (2004) and Young (1990) argued that the concept of
environmental justice goes beyond the distributional aspects of goods and bads.
According to Schlosberg (2004), the concept of global environmental justice must
include three components (p. 517): parity in the distribution of environmental hazards
and public services, the acknowledgment of the different and specific cultural
backgrounds of each community, and community engagement in decision making that
focuses on environmental policy (Schlosberg, 2004). In particular, Young (1990) stressed
the importance of procedural justice, describing equitable decision making processes to


40
determine the distribution of environmental amenities and hazards. The analysis of
procedural justice needs to take into account the social, political, and cultural conditions
that contribute to spatial distributions (Schlosberg, 2004; I. M. Young, 1990). This
expanded definition of environmental justice connects well with social justice. Indeed,
social justice is a vision of society in which each individual is warranted equal social,
political, and economic rights and opportunities (Bell, 2007; Gibelman, 1995).
Equality vs. equity. When discussing access to parks as an environmental justice
issue, a few scholars suggest going beyond the concept of equality to focus on equity, as
low-income young people of color need parks more than other demographic groups
(Boone et al., 2009; Wolch et al., 2005). When planning for parks in cities, an equality
approach would mean distributing parks and resources evenly across income, age, and
ethnic-racial groups (Boone et al., 2009; Wolch et al., 2005). On the other hand, planning
for equity would mean locating more parks and investments in areas with higher
concentrations of people who need parks the most, including: low-income people, people
under the age of 19 or above the age of 65, and people without access to a private vehicle
(Boone et al., 2009; Talen, 2003; Wolch et al., 2005). In other words, an equity approach
to recreation planning would ensure that the people with the greatest need and fewest
means to access recreational areas could do so (Thomas, 2010, p. 503).
Focusing on equity instead of equality is important because children and
teenagers from low-income and ethnic minority backgrounds generally lack the means
for traveling to destination parks due to transportation limitations (Castonguay & Jutras,
2009; Karsten, 2005; Loukaitou-Sideris & Sideris, 2010; Ries et al., 2008); have limited
opportunities to use private recreational opportunities, including backyards (Loukaitou-


41
Sideris & Stieglitz, 2002); generally live in denser urban environments (Loukaitou-
Sideris & Stieglitz, 2002); and take part in organized recreational activities less
frequently than other groups (Valentine & McKendrick, 1997). Given these limitations,
children in low-income densely populated areas spend more time in public space,
including parks and playgrounds, than children living in more affluent suburban areas
(Karsten, 2005; Loukaitou-Sideris & Stieglitz, 2002). Thus, Castonguay and Jutras
(2009) argued that high-quality public space is fundamental for children living in low-
income neighborhoods.
Using the equity approach to plan parks by taking into account park dependent
populations also makes sense financially (Boone et al., 2009). To prioritize park
investment in a period with shrinking budgets for public amenities, planners need to
conduct suitability analyses to identify the areas of cities and metropolitan areas where
park dependent populations live and where park service is currently low (Boone et al.,
2009; Nemeth, Faulkner, & Ross, 2012; Talen, 2003).
Environmental justice and sustainability. As I briefly explained in the
introduction, this dissertation is also driven by the necessity of stressing the fundamental
role of environmental justice in sustainability discourses and practices. Although equity is
one of the three Es of sustainability (S. Campbell, 1996), its importance in sustainability
scholarship and policies has often been overlooked (Agyeman & Evans, 2004, 2003;
Boone, 2010). In particular, Agyeman and Evans (2004) introduced the concept of just
sustainability to argue that the discussion of environmental justice needs to be located
within the sustainability framework (p. 155). Also, the idea of just sustainability
describes the fertile connections that can be established between environmental justice


42
and sustainability and acknowledges that sustainability cannot only be about
environmentalism (Agyeman & Evans, 2004, 2003).
The inadequate attention to equity in sustainability discourses and practices has
been highlighted by a few studies (Environmental Law Institute, 1999; Pearsall & Pierce,
2010; Warner, 2002). Warner's (2002) review of sustainability policies in large American
cities highlighted that very few municipalities (5 out of 33) considered environmental
justice as an integral part of sustainability. Even among the cities that included
environmental justice concerns in sustainability projects, there were notable differences
in how environmental justice was implemented (Warner, 2002). In some cities, for
example, environmental justice was only mentioned as a general goal, but no specific
actions to achieve environmentally just sustainability policies were implemented (Warner,
2002). Comparable findings emerge from a report by the Environmental Law Institute
(1999), which studied 579 applications to the Environmental Protection Agencys 1966
Sustainable Development Challenge Grant Program. Only 5 percent of the submitted
applications included equity as an objective (Environmental Law Institute, 1999). A more
recent review of sustainability plans in large U.S. cities showed that more and more cities
are including environmental justice goals in their sustainability plans, but that in practice
justice-oriented actions have less leverage than efforts aimed to solve ecological issues
(Pearsall & Pierce, 2010), possibly due to power imbalances among groups contributing
to these plans.
This theoretical and empirical literature suggests the importance of framing equity
and environmental justice within the sustainability framework, as well as the need to
highlight the key role of equity as one of the three parts of sustainability (see S. Campbell,


43
1996). This is particularly relevant because ecologically-driven policies, which stem from
bipartisan political efforts, often privilege economic development over social equity
(Checker, 2011).
To summarize, in this dissertation I adopt an environmental goods approach to
analyze the social disparities in young peoples access to play opportunities in parks,
including the reasons behind the observed disparities. To study the planning processes
and outcomes that influence access to parks for young people in Denver, I employ a
theoretical lens that combines the transformative paradigm (Mertens, 2003) with an
environmental justice lens (Schlosberg, 2004). Also, like other scholars (Boone et al.,
2009; Wolch et al., 2005), I argue that equity is a better construct than equality when
studying access to parks across demographic groups. Finally, I frame equity and
environmental justice within the sustainability framework (Agyeman & Evans, 2004;
Boone, 2010), and argue that equity should have the same importance as environmental
quality and economic development in sustainability discourses and practices.
Literature Review
In this section, I use my theoretical framework to connect several bodies of
literature that shed light on childrens and teenagers access to urban nature. In particular,
I focus on most of the connections among the constructs included in the theoretical
framework (see Figure 1) and then highlight the literature gaps and inconsistencies,
which lead to my research goals and questions.


44
Park Spatial Distribution and Demographics
Increasingly, research has studied how access to different types of public facilities
and other amenities differ by socio-economic and ethnic groups (Lobao & Saenz, 2002;
Macintyre, 2007; National Recreation and Park Association, 2011; Wolch et al., 2014)
and in particular leisure research has increased its focus on social justice (Floyd, 2014).
The body of research focusing on the spatial distribution of public amenities has been
named equity mapping (Talen, 1998) because studies generally map resources in
relation to where different demographic groups live. The results of these environmental
justice studies are often influenced by the unit of analysis and by the geographic scope of
the investigation (Baden, Noonan, & Turaga, 2007); thus, they need to be considered in
relation to the methods they use to measure access.
In particular, scholars have studied how various income, ethnic, and racial groups
have different levels of access to parks and playgrounds (Boone et al., 2009; Comber,
Brundson, & Green, 2008; Cradock et al., 2005; Dai, 2011; Ellaway, Kirk, Macintyre, &
Mutrie, 2007; Erkip, 1997; Estabrooks, Lee, & Gyurcsik, 2003; Gilliland, Holmes, Irwin,
& Tucker, 2006; Jones, Brainard, Bateman, & Lovett, 2009; Kabisch & Haase, 2014;
Loukaitou-Sideris & Stieglitz, 2002; Madsen, Radel, & Endter-Wada, 2014; Maroko,
Maantay, Sohler, Grady, & Arno, 2009; Miyake, Maroko, Grady, Maantay, & Arno,
2010; L. V. Moore, Diez Roux, Evenson, McGinn, & Brines, 2008; Nemeth et al., 2012;
Nicholls, 2001; Potwarka et al., 2008; Rigolon & Flohr, 2014; Sister, Wolch, & Wilson,
2009; Smoyer-Tomic, Hewko, & Hodgson, 2004; Talen & Anselin, 1998; Talen, 2010; A.
F. Timperio, Ball, Salmon, Roberts, & Crawford, 2007; Vaughan et al., 2013; Wen,
Zhang, Harris, Holt, & Croft, 2013; Willemse, 2013; Wolch et al., 2005; Wright Wendel,


45
Downs, & Mihelcic, 2011; Zhou & Kim, 2013), walkable and bikeable neighborhoods
(Christie et al., 2011; Crawford et al., 2008; Cutts, Darby, Boone, & Brewis, 2009;
Duncan et al., 2012; Franzini et al., 2010; Talen, 2003), green school grounds (Dyment,
2005; Stewart, Purner, & Guzman, 2013), schools with joint-use agreements for
community use of school grounds (Kanters, Bocarro, Moore, Floyd, & Carlton, 2014),
greenways (Lindsey, Maraj, & Kuan, 2001), public recreational programs (Dahmann,
Wolch, Joassart-Marcelli, Reynolds, & Jerrett, 2010), school resources (Condron &
Roscigno, 2003; Roscigno, Tomaskovic-Devey, & Crowley, 2006; Zhang & Cowen,
2009) , food provision (Larsen & Gilliland, 2008; Morland, Wing, Diez Roux, & Poole,
2002; Powell, Slater, Mirtcheva, Bao, & Chaloupka, 2007; Sharkey & Horel, 2008),
healthy and unhealthy food (Block, Scribner, & DeSalvo, 2004; Day & Pearce, 2011;
Kestens & Daniel, 2010; Pearce, Blakely, Witten, & Bartie, 2007; D. M. Smith et al.,
2010) , and street trees or tree canopy (Heynen, Perkins, & Roy, 2006; Landry &
Chakraborty, 2009; Mennis, 2006; Zhou & Kim, 2013). Both place-specific and
aggregate national evaluations have been conducted (Boone et al., 2009; National
Recreation and Park Association, 2011). In general, this body of research shows that
access to public and private amenities is an environmental justice issue because such
amenities are not equitably distributed income and ethnic groups.
A few recent reviews have summarized research addressing how the provision of
parks and recreational facilities varies by ethnic and income groups, showing mixed
findings in terms of access to parks regardless of their size, quality and amenities they
include. The reviews carried on by the National Recreation and Park Association (2011)
and by Wolch et al. (2014) show a prevalence of studies finding that low-income groups


46
and ethnic minorities are underserved in terms of access to parks and recreational
facilities. Macintyre's (2007) review challenges the assumption that low-income and
ethnic minority groups always have lower access to health-promoting amenities,
including parks. Indeed, the results of her review show that low-income communities of
color do not always have lower access to health-promoting facilities (Macintyre, 2007).
The differences between the findings of these reviews could derive from park funding
policies, on the developmental patterns of the cities that were studied, and on the methods
that were used to measure access to parks. The next section critically analyzes empirical
studies about access to parks to clarify the contradictions between these three reviews.
Access to parks. To gain a better understanding of the current knowledge about
the equity of park distribution across demographic groups in several geographic contexts,
I conducted a literature search on the most important full-text academic databases,
including Web of Science, Science Direct, Jstor, and EBSCOhost. My goal was to
identify empirical studies measuring the spatial distribution of parks or other public green
spaces in relation to socio-economic, ethnic, or racial factors (criteria for this search).
This review focused on access to parks including park proximity, park acreage, and park
quality. Given this goal, in the literature search I used the following keywords and
phrases: park equity, green space equity, park equality, green space equality,
park access, green space access, park environmental justice, green space
environmental justice, park distribution, and green space distribution.
The search yielded 38 journal entries that met the criteria. Then, I expanded this
sample by looking at the references of the 38 articles and found five additional papers
meeting the criteria. Thus, the final sample of articles I analyzed includes 43 empirical


47
papers on access to parks or green spaces (see Table A1 in Appendix A for the complete
list). These articles focused on a variety of geographical contexts, including big cities like
Los Angeles and New York City and small suburban towns in the South and Midwest.
The variability of geographies likely influenced some of the results of these studies.
To analyze these empirical articles, I developed a codebook including descriptive
codes, such as the name of the authors, the year of publication, the journal and the
geographical focus; topical codes, such as the scale of investigation, the methods use to
measure access and the specific GIS methods, the unit of analysis, the critical distance to
define access (if present), the demographic variations and the type of statistical analysis;
and analytical codes, such as the presence of statistically significant results, the findings
in terms of park proximity, park acreage, and park quality, and other detailed findings
(Richards, 2005). Some codes were close-ended, including the scale of investigation and
the specific GIS methods, and others were open-ended, including the findings of the
analyzed studies. The full codebook is reported in Table B1 in Appendix B.
Dates, locations and methods. The sampled papers span from 1997, with the
early work by Emily Talen (1997) to 2014, when a variety of papers have been published
(Kabisch & Haase, 2014; Reyes, Paez, & Morency, 2014; Rigolon & Flohr, 2014). The
search also highlighted an increase in studies on access to parks in the last five years,
which employed more and more refined GIS methods, probably due to progress in GIS
software and to improved GIS techniques. The great majority of the sampled studies was
conducted in geographic locations within the United States (29 entries), followed by
places in the United Kindgom (4 entries), Canada (3 entries), Australia (2 entries), and
other countries (South Africa, Germany, Turkey, Germany, Bolivia, Israel). Most of the


48
selected studies focused on cities as their geographical scale (28 entries), followed by
metro areas (13 entries), regions (2 entries), and countries (2 entries).
In terms of methods used to measure access to parks, the vast majority of
empirical studies employed GIS analyses (37 entries), while other methods that were
used include quantitative surveys (5 entries), qualitative methods (2 entries), or other
quantitative methods (2 entries). To classify the specific GIS method that the selected
studies employed, I used Talen's (2003) framework, which includes five categories of
techniques to measure access to facilities for various demographic groups: container,
coverage, minimum distance, travel cost, and gravity. Container describes the quantity of
amenities included within a certain spatial unit, for example a census block (Talen, 2003).
Coverage depicts the amount of amenities located within a certain distance from a source
(Talen, 2003). Minimum distance describes the distance between a source and the closest
amenity (Talen, 2003). Travel cost depicts the mean distance between a source and all
amenities (Talen, 2003). Gravity represent an indicator in which the total number of
amenities (weighted by their size) is divided by the average distance of each amenity
from the source (Talen, 2003).
Based on this classification, my analysis shows a relatively balanced distribution:
the majority of the sample studies employed the minimum distance approach (19 entries),
followed by container (16 entries), coverage (11 entries) and gravity (9 entries). A
significant number of studies employed more than one GIS method to measure
accessibility (13 entries), which shows that many authors aim to investigate access to
parks from several perspectives. Then, among the studies that employed minimum
distance, coverage and gravity, most of them used a quarter mile (12 entries) or half a


49
mile (9 entries) to define access. Other distances that were employed include one mile (5
entries), two miles (1 entry), and 300 meters (1 entriy), while 11 studies used multiple
distances. This finding is not surprising, as most authors consider a quarter and half a
mile as the maximum distances that people are willing to walk to reach a facility (Boone
et al., 2009; Nicholls, 2001).
Other methodological aspects include the unit of analysis used as the areas to be
served by parks, the demographic variations included by the authors, and the type of
statistical analysis. The most used units of analysis are census tracts (12 entries), census
block groups (10 entries), and neighborhoods (8 entries), followed by computer-
generated geometries (5 entries), census blocks (4 entries), households through surveys (3
entries), parcels (2 entries), cities (1 entry) and other units. Also, four studies employed
more than one unit of analysis. The choice of the unit of analysis involves a tradeoff
between a detailed geometric measure of access (smaller units of analysis) and detailed
demographic information (larger units of analysis). In terms of demographic variation,
the vast majority of the sampled studies focus on socio-economic status (SES) including
income, poverty and educational attainment (37 entries), while many studies focus on
ethnicity (28 entries), race (14 entries) and age differences (12 entries). A large number
of studies (33 entries) include more than one of the above variables. Finally, regarding
statistical analysis, slightly more studies employed comparative statistics (24 entries) than
correlational statistics (21 entries).
This general review of geographical focus and methods serves as important
contextual information that I use to evaluate the findings presented in the next sections.


50
Specificially, I focus on distance to the closest park (park proximity), number of parks,
park acreage and park congestion (park acreage), and park quality and safety.
Park proximity. The evidence is mixed about how access to any type of park,
expressed in terms of distance between homes and the closest park, varies by socio-
economics, ethnicity, race, and age groups. In summary, African Americans and other
racial minorities tend to live closer to parks than Non-Hispanic Whites. Among the
sampled articles, nine studies found that low-income groups live further from parks than
other groups, while seven other studies found that low-income groups live closer to parks
than other groups (see Table 2). Thus, these findings do not give a clear, unidirectional
picture about who lives closer to parks. In terms of ethnicity, two studies highlighted that
ethnic minorities like Hispanics live further from parks than other groups, while three
other studies found the opposite (see Table 2). The picture is clearer when focusing on
race, as seven studies found that African Americans and other racial minorities live closer
to parks than Non-Hispanic Whites, while only one study found the opposite (see Table
2). Finally, two studies highlighted that the percentage of people under 18 or above 65 is
positively correlated with distance from parks, showing that children, adolescents, and
seniors live further from parks than other age groups. To summarize, Table 2 shows that
in terms of distance to parks the only substantial differences are in terms of racial groups,
with racial minorities and African Americans in particular living closer to parks than
Non-Hispanic Whites.


51
Table 2
Park Proximity: Who Lives Closer to Parks?
Income
Low-income groups live closer to parks 7 entries (Boone et al., 2009; Kessel et al., 2009; Lindsey et al., 2001; Nicholls, 2001; Smoyer-Tomic et al., 2004; K. Wells, 2005; Wen et al., 2013)
Low-income groups live further from parks 9 entries (Cradock et al., 2005; Dai, 2011; Erkip, 1997; Nemeth et al., 2012; Omer & Or, 2005; Rigolon & Flohr, 2014; Veitch, Salmon, & Ball, 2007a; Wen et al., 2013; Willemse, 2013)
Ethnicity
Ethnic minorities live closer to parks 3 entries (Cutts et al., 2009; Miyake et al., 2010; Wen etal., 2013)
Ethnic minorities live further from parks 2 entries (Omer & Or, 2005; Rigolon & Flohr, 2014)
Race
Racial minorities live closer to parks 7 entries (Boone et al., 2009; Cutts et al., 2009; Lindsey et al., 2001; Nicholls, 2001; Wen et al., 2013; Wolch et al., 2005; Zhou & Kim, 2013)
Racial minorities live further from parks 1 entry (Dai, 2011)
Age
% of people under 18 and above 65 live further from parks 2 entries (Cutts et al., 2009; Nemeth et al., 2012)
After examining the general picture of how distance to parks varies by income,
ethnic and racial group, it is appropriate to look at some of the specific studies included
in Table 2, with particular attention to the connections between their methods and their
findings. This sheds light on some of the findings presented in Table 2, and leads to


52
reconsidering some of them. For example, Nicholls' (2001) article, which found that
ethnic minority and low-income people have higher access to parks than other groups, is
based on an analysis of Bryan, TX, a very low-density city with relatively few parks and
large parts of the city that are not served by parks. This lack of variation in terms of
access versus no-access to parks casts some doubts on the planning significance of the
authors findings.
Also, Smoyer-Tomic et al.'s (2004) study, which found that low-income people
live slightly closer to playgrounds than other groups, sampled only neighborhoods that
are mainly residential and urban, leaving out significant parts of the city. This sampling is
problematic, as selecting only urban residential neighborhoods usually leaves out Non-
Hispanic White and wealthy suburbs, which might have high numbers of recreational
opportunities per person. Also, neighborhoods with mixed land use (industrial or retail)
generally have high percentages of low-income ethnic minority groups. Wen et al.'s
(2013) article, which highlighted that people in poverty were more likely to live closer to
parks in urban and suburban areas, used a method to measure distance that is problematic
when thinking about the needs of low-income children and teenagers of color. Indeed,
Wen et al. (2013) calculated the weighted distance to the seven closest parks. However,
some of these parks might be too far to reach for low-income children and teenagers of
color, due to their limited mobility; thus, their findings are not very significant in terms of
childrens and teenagers access to parks.
As mentioned above, one aspect to consider when defining which geographic
areas of the city have access to parks is the distance defining access, which describes how
far people are willing to walk, bike, or drive to parks. For example, Dai (2011), in his


53
study of access to parks in Atlanta, employed several thresholds expressed in driving
times (between 10 and 30 minutes, with intervals of 5 minutes). Dai (2011) argued that
using driving times instead of walking or biking distances is appropriate for Atlanta given
the citys low density. This choice is also based on the assumption that people generally
drive to parks instead of walking or biking (Dai, 2011). By using this approach, Dai
(2011) found that African Americans, but not Asians, had significantly lower access to
parks than Non-Hispanic Whites, and that low-income areas also have disproportionately
lower access to parks than higher income areas. While using driving times can be an
appropriate choice for Atlanta, driving times do not reflect childrens and teenagers
opportunity to access park independently. To this regard, it would be interesting to assess
ethnic and income differences in access to parks for Atlanta when using a quarter mile to
define access. Indeed, in denser urban environments like New York City, Baltimore, and
Denver, many studies have used the quarter mile as the critical distance defining access
(Boone et al., 2009; Miyake et al., 2010; Rigolon & Flohr, 2014).
To summarize, the analysis of the mentioned papers (Dai, 2011; Nicholls, 2001;
Smoyer-Tomic et al., 2004; Wen et al., 2013) raises questions about the way the methods
researchers used have influenced their findings about distance to the closest park across
demographic groups, and particularly on whether low-income groups live closer to parks
than other groups. Therefore, the first part of this review shows that, in terms of distance
to parks, the only evident difference is that African Americans and other racial minorities
tend to have a shorter distance to the closest park than Non-Hispanic Whites.
Park acreage. When looking at the number of accessible parks, park acreage and
park acreage per resident or child (potential congestion), a clear picture of inequity


54
emerges. Table 3 shows that, among the sampled articles, low-income groups, ethnic
minorities, racial miniorities, and young people and seniors have access to fewer parks or
to less park acreage than other groups. In particular, 17 studies found that low-income
people have fewer parks or less park acreage, while only five studies found that low-
income groups have more parks or more park acreage (see Table 3). In terms of ethnicity,
12 studies reported ethnic minorities like Hispanics having access to fewer parks or to
less park acreage, while only one study found the opposite (see Table 3). Regarding race,
nine studies found that African Americans and other racial minorities have access to
fewer parks or to less park acreage compared to Non-Hispanic Whites, while only one
study found the opposite (see Table 3). Finally, three studies highlighted that the
percentage of people under 18 or above 65 is negatively correlated with the number of
parks or park acreage, showing that children, adolescents, and seniors have access to
smaller parks or to lower park acreage than the rest of the population.


55
Table 3
Park Acreage: Who Has More Accessible Park Acreage and Acreage per Person?
Income
Low-income groups have more facilities, or more park acreage 5 entries (Ellaway et al., 2007; Gilliland et al., 2006; Smoyer-Tomic et al., 2004; Talen, 1997; Vaughan et al., 2013)
Low-income groups have fewer facilities, or less park acreage 17 entries (Boone et al., 2009; Estabrooks et al., 2003; Gordon-Larsen, Nelson, Page, & Popkin, 2006; Jones et al., 2009; Loukaitou-Sideris & Stieglitz, 2002; Omer & Or, 2005; Reyes et al., 2014; Sister et al., 2009; Talen & Anselin, 1998; Talen, 1997, 2010; Weiss et al., 2011; Wen et al., 2013; Willemse, 2013; Wolch et al., 2005; Wright Wendel et al., 2011; Wright Wendel, Zarger, & Mihelcic, 2012)
Ethnicity
Ethnic minorities have more facilities, or more park acreage 1 entry (Maroko et al., 2009)
Ethnic minorities have fewer facilities, or less park acreage 12 entries (Comber et al., 2008; Cutts et al., 2009; Gordon-Larsen et al., 2006; Kabisch & Haase, 2014; Maroko et al., 2009; L. V. Moore et al., 2008; Omer & Or, 2005; Sister et al., 2009; Talen, 1997; Weiss et al., 2011; Wen et al., 2013; Wolch et al., 2005)
Race
Racial minorities have more facilities, or more park acreage 1 entry (Talen, 1997)
Racial minorities have fewer facilities, or less park acreage 9 entries (Boone et al., 2009; Cutts et al., 2009; Kabisch & Haase, 2014; Maroko et al., 2009; L. V. Moore et al., 2008; Sister et al., 2009; Talen & Anselin, 1998; Weiss et al., 2011; Wolch et al., 2005)
Age
% of under 18 or above 65 have fewer facilities, or less park acres 3 entries (Cutts et al., 2009; Gilliland et al., 2006; Loukaitou-Sideris & Stieglitz, 2002)


56
In particular, a few studies found that low-income and ethnic-racial minority areas
have considerably less park acreage per person and per child than other areas, which can
lead to park crowding issues (Boone et al., 2009; Sister et al., 2009; Wolch et al., 2005).
In particular, Boone et al. (2009) reported that areas of the Baltimore metropolitan area
with high percentages of park dependent people (people under 18 and above 65, people in
poverty, and people without access to a car) have significantly fewer acres per person
than other parts of the metro area, notably middle- and upper-class Non-Hispanic White
suburbs. Sister et al. (2009) reported very similar results in terms of park dependence in
the Los Angeles metropolitan area. In metropolitan Los Angeles, the most densely
populated areas, which house disproportionately high percentages of ethnic minorities
and low-income groups, have significantly fewer park acres per person than suburban
communities, particularly the San Fernando Valley, where higher percentages of Non-
Hispanic White middle- and upper-class populations live (Loukaitou-Sideris & Stieglitz,
2002; Wolch et al., 2005).
Furthermore, three studies reported that areas with higher percentages of people
under 18 have significantly fewer park acres nearby or a lower density of recreational
facilities like sport fields and recreational centers (Cutts et al., 2009; Gilliland et al.,
2006; Loukaitou-Sideris & Stieglitz, 2002). In particular, Cutts et al. (2009) highlighted
that in Phoenix, AZ the areas located in proximity to large parks included significantly
lower percentages of people under 18 years of age. Also, Gilliland et al.'s (2006) study of
London, Ontario showed that areas with higher concentrations of low-income groups had


57
significantly fewer recreational opportunities per youth than mid- and high-income
neighborhoods.
The shortage of park acreage and of recreational opportunities in low-income
communities of color with high percentages of children and teenagers is a very
significant finding in terms of childrens and teenagers access to urban nature, as park
overcrowding might prevent children and teenagers from playing in parks and deteriorate
the conditions of parks (Sister et al., 2009). These findings are disturbing in terms of
distributional equity, as they show that the populations who need parks the most, low-
income ethnic and racial minority young people, have access to fewer park acres per
person. These patterns of inequity are often related to geographical differences among
different parts of cities or metropolitan areas, which are analyzed later in this review.
In summary, while the literature shows mixed findings about which demographic
groups live closer to parks, regardles of their size and quality, my review highlights
striking inequities in terms of the number of parks and of the acreage of parks that
different socio-economic, ethnic, racial, and age groups have access to. In particular,
areas with high concentrations of low-income groups, ethnic minorities, racial miniorities,
and people under 18 or above 65 years of age, have access to significantly fewer parks or
to less park acreage than other areas.
Park quality. After assessing park proximity and park acreage, I reviewed the
sampled articles to understand whether the quality of parks varies by socio-economic,
ethnic/racial, and age groups. In this analysis, park quality includes the presence of
amenities in parks, the level of maintenance of parks, and the presence of potential
physical hazards in parks. The literature I reviewed consistently shows that park quality is


58
lower in low-income communities of color with high concentrations of children and
teenagers, who are the groups with the highest park dependence (Anderson, Jackson,
Egger, Chapman, & Rock, 2014; Crawford et al., 2008; Ellaway et al., 2007; Estabrooks
et al., 2003; Jones et al., 2009; Landry & Chakraborty, 2009; Loukaitou-Sideris &
Stieglitz, 2002; L. V. Moore et al., 2008; Rigolon & Flohr, 2014; Smoyer-Tomic et al.,
2004; Talen & Anselin, 1998; Vaughan et al., 2013; Zhou & Kim, 2013).
In particular, access to play amenities in parks and the quality of such amenities
are inequitably distributed. Park dependent populations have lower levels of access to
playgrounds and recreational opportunities than Non-Hispanic White middle- and upper-
class groups (Jones et al., 2009; L. V. Moore et al., 2008; Rigolon & Flohr, 2014; Talen
& Anselin, 1998; Vaughan et al., 2013). Most importantly, such play amenities and in
general the parks have lower quality (Ellaway et al., 2007; Loukaitou-Sideris & Stieglitz,
2002; Vaughan et al., 2013), to have lower levels of maintenance (Carlson, Brooks,
Brown, & Buchner, 2010; Smoyer-Tomic et al., 2004), to have less shading (Anderson et
al., 2014), to be more crowded (Loukaitou-Sideris & Stieglitz, 2002), and to include more
physical environment hazards (Carlson et al., 2010; Cradock et al., 2005; Suecoff, Avner,
Chou, & Crain, 1999). Also, parks in low-income and ethnic minority neighborhoods
tend to be more exposed to air pollution than parks in wealthier Non-Hispanic White (Su,
Jerrett, de Nazelle, & Wolch, 2011).
This research, conducted in the United States, Canada, England, and Australia,
including U.S. cities like New York (Suecoff et al., 1999), Boston (Cradock et al., 2005),
Kansas City (Vaughan et al., 2013), Denver (Rigolon & Flohr, 2014), and Los Angeles


59
(Loukaitou-Sideris & Stieglitz, 2002; Su et al., 2011), suggests that access to high-quality
play spaces in urban nature is an issue for low-income children and teenagers of color.
In particular, Loukaitou-Sideris and Stieglitz's (2002) study of eight parks in Los
Angeles highlights that children living in the inner city, which has a high proportion of
low-income ethnic minorities, had negative views of the quality of their playgrounds and
of their levels of maintenance. On the other hand, children living in the San Fernando
Valley, a higher income and less ethnically diverse area, expressed dissatisfaction about
the activities afforded by their parks, complaining that most playgrounds are designed for
young children (Loukaitou-Sideris & Stieglitz, 2002). Therefore, the quality and
maintenance of play settings was an issue in lower income and higher density areas.
In addition, people residing in low income and ethnic/racial minority areas have
less access to street trees than people living in other parts of the city (Crawford et al.,
2008; Franzini et al., 2010; Landry & Chakraborty, 2009; Mennis, 2006; Zhou & Kim,
2013). Disadvantaged areas of cities also have fewer amenities in public open space,
including tables, fountains, and cycling paths (Crawford et al., 2008), and parks with less
interesting aesthetic features (Vaughan et al., 2013).
Therefore, these results suggest that when studying access to outdoor play
opportunities for young people, it is important to focus on access to parks with high
quality and safe play opportunities rather than access to every park.
Other findings. The analysis of this literature highlighted some other findings that
can help shed light on how planning can help balance inequities in park access. First, a
few studies showed geographical differences in terms of park acreage between urban
cores and outskirts (e.g. Boone et al., 2009; Dai, 2011; Sister et al., 2009; Wolch et al.,


60
2005). Most of these studies found that suburban areas and the outskirts of a city include
more park acreage per person and per child than inner cities (Boone et al., 2009; Dai,
2011; Kabisch & Haase, 2014; Loukaitou-Sideris & Stieglitz, 2002; Reyes et al., 2014;
Sister et al., 2009; Talen, 2010; Wolch et al., 2005), while fewer studies found the
opposite spatial pattern (Miyake et al., 2010; Talen, 1997; Zhou & Kim, 2013). The
prevalence of studies highlighting higher park acreage in the outskirts is not surprising,
due to greater land availability in these areas compared to dense inner-cities and due to
lower population density in the suburbs. However, in most studies, geographical
differences are correlated with demographic differences, as low-income communities of
color generally live in inner-cities (fewer park acres) and Non-Hispanic White middle-
and upper-class groups are more often located in suburbs (more park acres). However,
the landscape of the country is shifting with the back-to-the-city movement, which has
led to an increased suburbanization of poverty and diversity (Kneebone & Garr, 2010).
In particular, Sister et al. (2009) found that metropolitan Los Angeles inner-city
areas, where low-income communities of color live the most, include mainly small
neighborhood parks, while suburban areas, with mid- and high-income residents,
comprise large parks like natural preserves, arboreta, and beaches. Similarly, Miyake et al.
(2010) highlighted that, in New York City, inner-city low-income areas have many small
parks that do not fit the needs of active recreation.
Second, some authors proposed possible solutions to the distributional inequities
they observed. The most common suggested solution is conducting a suitability analysis
to identify areas that lack parks and where park dependent groups live, with the goal of
prioritizing investments for the creation of new parks (Nemeth et al., 2012; Rigolon &


61
Flohr, 2014; Talen, 2003). Also, Sister et al. (2009) simulated the impact of locating new
parks based on distributional equity, showing that in certain circumstances small
interventions can bring about large improvements, at least in terms of park proximity.
Rigolon and Flohr (2014) proposed a more comprehensive framework for action for
neighborhood greening that would include low-income communities, local non-profits
and local universities.
Finally, very few of the studies I reviewed highlighted park planning processes or
tried to uncover the reasons for the observed park distributions (e.g. Boone et al., 2009;
Joassart-Marcelli, 2010; Wolch et al., 2005). I will discuss their findings in terms of
planning processes in the next section.
To summarize, the review of studies focusing on park spatial distribution in
relation to demographics showed that results are mixed in terms of distance to parks due
to methodological issues and to the broad range of geographies included in the review.
Thus, a pattern of equity or inequity is not clearly identifiable. The review also
highlighted that when focusing on the number of parks, park acreage and park congestion,
low-income ethnic communities, children, and teenagers are significantly underserved.
Finally, in terms of park quality, my literature analysis showed that low-income ethnic
minority people, children, and teenagers have access to parks with lower quality than
other groups. When combining these three measures, the inequities between low-income
children of color and other groups are clear.
Policies, Practices, and Park Spatial Distribution
A substantial body of literature has investigated the connections between urban
policy and planning and social equity, including several studies trying to demonstrate the


62
hypothesis that policies often created large inequities in urban areas (Bolotin &
Cingranelli, 1983; Gillette, 1995; Lineberry, 1977; Maantay, 2001; Massey, 1990). In a
classic paper, Rittel and Webber (1973) argued that planning problems dealing with the
distribution of value are wicked because in a pluralistic society with many publics
and contrasting interests, it is hard to agree upon what common good and equity mean.
This stance can help explain why equity issues have occurred, as groups with different
interests tend to hold dissimilar opinions about what good and bad policies are (Rittel &
Webber, 1973); thus, the groups with the strongest voice influence policies the most.
Public policy and equity. The public policy literature has shown several
examples of policies that, intentionally or unintentionally, failed to address social equity
goals. A recent book edited by Carmon and Fainstein (2013) showed the detrimental
effects on social equity of neo-liberal planning policies oriented to increasing the
competitiveness of cities, with a focus on housing, transportation, public space and
regional planning. Even policies that are driven by social justice or ecological goals can
have negative consequences for underprivileged groups. For example, Gillette (1995)
showed the failure of policies oriented to social equity in Washington, D.C., as most
housing efforts intended to improve African Americans' living conditions excluded them
from the most desirable areas of the city. Also, a study conducted in Brazil showed that
policies aimed to preserve urban green areas can have environmental justice
consequences for underprivileged populations who live in informal settlements (A. F.
Young, 2013). This suggests that sustainability practice, as previously mentioned, often
overlooks justice as one of its main components (Boone, 2010; Checker, 2011). Also, the
implementation of ecologically oriented sustainability interventions like new parks can


63
lead to disadvantaged populations like low-income and ethnic minority groups being
forced to relocate due to increased rents and property values; a phenomenon known as
ecological or environmentalgentrification (Checker, 2011; Dooling, 2009). Also, as
previously noted, a few scholars have recently argued that the inclusion of environmental
justice stances can advance sustainability research and practice (Agyeman & Evans, 2004,
2003; Boone, 2010).
Historical and funding explanations of park distribution. Relatively few
studies focused on how urban policies and planning practices have influenced the spatial
distribution of parks in urban areas (Boone et al., 2009; Joassart-Marcelli, 2010; D. H.
Koehler & Wrightson, 1987; Pincetl, 2003; J. W. Smith & Floyd, 2013; Thomas, 2010;
Wolch et al., 2005), while others have suggested possible explanations for the
distributional patterns that they observed (Jones et al., 2009; Lindsey et al., 2001; Maroko
et al., 2009; Omer & Or, 2005). Residential segregation of low-income ethnic minority
groups from other wealthier groups is an important factor in the distributional inequity of
parks, as areas where wealthier groups reside tend to receive more funding for public
amenities and less locally unwanted land uses (Boone et al., 2009; Omer & Or, 2005).
Several factors throughout the history of American cities contributed to todays
park distribution (Boone et al., 2009). The historical period during which different parts
of cities were developed plays a key role in understanding park location: Neighborhoods
developed before the widespread diffusion of the automobile tend to be denser, to be
more walkable, and to include smaller parks (Boone et al., 2009). In these older
neighborhoods and cities, creating new parks was often a very problematic endeavor due
to the high land value of densely-populated areas (Boone et al., 2009). In these contexts,


64
parks were sometimes established by purchasing adjacent small private lots, which was a
time consuming process requiring a strong commitment by public officials and citizens
(Boone et al., 2009). Another way to develop parks before the advent of the automobile,
and especially in the 1800s, was wealthy individuals donating land to cities, which often
led to large estates being turned into parks (Boone et al., 2009; Pincetl, 2003).
Among the studies that focused on the historical and planning processes that led
to todays park distribution in relation to income and ethnic groups, Boone et al.'s (2009)
article provides an in-depth picture of Baltimores park history. In particular, Boone et al.
(2009) conducted a historic process analysis to track the evolution of park location in
relation to where African Americans and Non-Hispanic Whites lived throughout
Baltimores history. In their study, they found that African Americans have been spatially
segregated throughout Baltimores history through the use of segregation regulations,
restrictive covenants attached to properties, and redlining by the Home Owners Loan
Corporation (Boone et al., 2009). Their research also highlighted that, around the 1930s,
politically influential neighborhood improvement associations helped steer public
investment, including investment for parks, to areas mostly populated by upper-class
Non-Hispanic Whites (Boone et al., 2009). With the departure of Non-Hispanic Whites to
the suburbs starting in the 1950s, the geography of access to parks in Baltimore
significantly changed, as African Americans started moving to previously Non-Hispanic
White neighborhoods (Boone et al., 2009). Therefore, Boone et al. (2009) claim that
African Americans in Baltimore currently live in neighborhoods with good access to
parks because these parks were established when Non-Hispanic White affluent people
resided in them; thus, African Americans inherited good access to parks from previously


65
Non-Hispanic White neighborhoods. Similarly, Maroko et al. (2009) suggested that most
parks were built in relatively early phases of most American cities and that, during the
evolution of cities, the populations living around parks could have changed significantly.
Given this process, Boone et al. (2009) argued that the current distribution of parks in
Baltimore is a case of environmental injustice, as todays park distribution is the outcome
of unjust and sometimes openly racist policies and practices.
Furthermore, Boone et al. (2009) and Pincetl (2003) suggested that in the early
1900s parks were sometimes established in impoverished neighborhoods as a form of
social engineering, as decision makers at the time thought that park access could reduce
youth misbehavior. As African Americans and Non-Hispanic Whites in Baltimore were
residentially segregated throughout the twentieth century, parks were also de facto
segregated between Non-Hispanic White and African American parks (Boone et al.,
2009). Therefore, evaluating procedural justice is a complex endeavor that requires
taking into account a variety of factors, including the mechanisms and reasons through
which parks were built, the demographics around parks when new parks were built, land
use and zoning regulations, and residential segregation policies and practices.
In an effort to conceptualize the mechanisms that determine park access in
American cities, J. W. Smith and Floyd (2013) presented two competing hypotheses that
can help explain park distribution across income and ethnic groups, the growth machine
and the central place theories (p. 87). The growth machine theory posits that, since
parks are the outcome of the political economy in cities, their distribution tends to benefit
higher-income and Non-Hispanic White people (J. W. Smith & Floyd, 2013). In other
words, this hypothesis holds that city officials intentionally locate parks in wealthy


66
neighborhoods, given the several benefits of parks (Byrne & Wolch, 2009), and locate
locally unwanted land uses like industrial or commercial developments in low-income
communities of color (J. W. Smith & Floyd, 2013). On the other hand, the central place
theory is grounded in the historical mechanisms of suburbanization and city expansion (J.
W. Smith & Floyd, 2013). This theory proposes that todays park distribution can be
explained by the historical developmental process of cities, which from their core
expanded into a first ring of suburbs, and then to farther areas (J. W. Smith & Floyd,
2013). Thus, the central place theory does not use social class or racial explanations for
park distributions across socio-economic and racial groups (J. W. Smith & Floyd, 2013).
Rather, it posits that the development cycles of cities, including centralization and
decentralization periods, influenced how different demographic groups can access parks
today (J. W. Smith & Floyd, 2013).
To examine these two theories, J. W. Smith and Floyd (2013) conducted a case
study of Raleigh, NC focusing on access to parks through variables pertaining to the two
aforementioned theories. Their findings show that the growth machine and the central
place theories can be used together to explain the pattern of park spatial distribution
across income and ethnic groups (J. W. Smith & Floyd, 2013). Indeed, in Raleigh African
Americans have less access to parks than Non-Hispanic Whites, but the inequity is not
evident for Hispanics and Asians (J. W. Smith & Floyd, 2013). Also, as the city grew,
parks and other open spaces were preserved in the urban core, showing the value
attributed to these settings (J. W. Smith & Floyd, 2013). Following the growth machine
and central place theories, Boone et al.'s (2009) findings can be seen in the following
way: The growth machine theory determined park distribution in the first decades of the


67
twentieth century, also through residential segregation; and the central place theory
explains the post-WWII white flight to the suburbs and African Americans inheriting
previously Non-Hispanic White neighborhoods served by parks.
Looking at the processes that led to land purchase for the establishment of new
parks can provide useful insights about procedural justice in access to parks. Thomas'
(2010) review of land acquisition policies from the 1960s to the early 2000s highlighted
that in the U.S. land purchases to establish new open spaces has overlooked equity issues
and missed chances to balance existing inequities between income and ethnic groups.
These missed opportunities could be due to the fact that, in the last decades of the
twentieth century, private land trusts have taken a prominent role in land acquisition to
establish new open spaces (Thomas, 2010). Also, although the goal of creating access to
recreation for every U.S. citizen was stated in several policy documents, the
implementation of most policies did not properly take into account income, ethnicity, age,
and gender differences in outdoor recreation (Thomas, 2010).
Also, when studying how policies have addressed park quality and park
maintenance, it is important to consider that in the 1970s many American inner cities
struggled with park funding (Low, Taplin, & Scheld, 2005), including Los Angeles
(Pincetl, 2003). The decline of public funding for parks was mostly due to white flight to
the suburbs, which caused a reduction in real estate tax revenues (Low et al., 2005). Due
to reduced investment, many inner-city parks experienced a drop in maintenance and
visitation (Low et al., 2005).
Park funding in Los Angeles. Metropolitan Los Angeles has been the object of
four studies that investigated specific park funding mechanisms exacerbating inequity in


68
park distributions (Joassart-Marcelli, Wolch, & Salim, 2013; Joassart-Marcelli, 2010;
Pincetl, 2003; Wolch et al., 2005). Pincetl's (2003) investigation shows that non-profit
organizations played a key part in seeking funding, creating and maintaining parks in Los
Angeles. The role of non-profits was particularly strong in the 1980s when, given the
reduction of property taxes, available funds for parks were very limited (Pincetl, 2003).
Environmental non-profits are active in park planning because they aim to improve their
citys environmental quality, and parks are perceived as factors of environmental quality
(Pincetl, 2003). Several public-private enterprises were formed, also building on the
rationale that new parks would increase property values (Pincetl, 2003). In relation to
environmental justice, Pincetl's (2003) analysis shows that non-profit organizations in
Los Angeles between the 1980s and 1990s often concentrated their efforts on large
regional parks because they were moved by environmental goals, and that they partially
overlooked the establishment of parks in inner-city Los Angeles, where low-income
communities of color lived and where several riots occurred.
Wolch et al. (2005) analyzed how two relatively recent programs to fund new
parks or park improvement through grants allocated resources across the Los Angeles
metropolitan area (Propositions K and A). Their results show that these programs created
more inequities between socio-economic and ethnic groups in terms of park acreage, as
more resources per child were invested in Non-Hispanic White high-income areas than in
areas with low-income communities of color (Wolch et al., 2005). In particular,
Proposition K favored park improvement over the establishment of new parks by
acquiring new properties, thus bringing investments to areas that were already served by
parks (Wolch et al., 2005). The majority of resources from Proposition A were allocated


69
for regional parks, which in Los Angeles are located far from the inner-city, thus far from
the areas where most low-income and ethnic minorities are concentrated (Wolch et al.,
2005). These finding parallels Pincetl's (2003) results, as they highlight the emphasis on
large parks over small neighborhood parks.
Joassart-Marcelli (2010) conducted a more comprehensive study on park funding
in different municipalities within Los Angeles, with a particular focus on park spending
per capita in different types of municipalities. The authors findings reveal that park
spending varied significantly across metropolitan Los Angeles, with large inequities
existing between wealthy and impoverished cities (Joassart-Marcelli, 2010). Although
this result is not surprising, Joassart-Marcelli (2010) also highlighted that funding coming
from state and non-profit sources was also inequitably distributed, with low levels of
investment in suburban municipalities that included high percentages of ethnic minorities
and low income people. In particular, non-profit organizations allocated on average $21
per person annually in affluent cities, while they spent only $5 per person on average in
the least wealthy cities (Joassart-Marcelli, 2010). In addition, affluent cities spent a
higher percentage of their parks and recreation budget on new parks and facilities
(26.1%), compared to low-income municipalities (12.6%; Joassart-Marcelli, 2010). Also,
when controlling for average income and overall city budget, cities with higher
percentages of children and teenagers spent more on parks and recreation (Joassart-
Marcelli, 2010). Yet, many impoverished municipalities also had higher proportions of
children and teenagers (Joassart-Marcelli, 2010). Overall, the highest spending for parks
was found in older municipalities with good economic activities, in which residential


70
areas are mixed with commercial and industrial land uses, and in which public space is
usually considered important (Joassart-Marcelli, 2010).
Joassart-Marcelli's (2010) study highlighted that non-profit organizations
contributed to creating significant differences in park spending among municipalities,
thus confirming Pincetl's (2003) results. Starting from this finding, Joassart-Marcelli et al.
(2013) conducted a study focusing specifically on how non-profit expenditure on parks
varies among different cities in Southern California. Their results show that non-profit
organizations operated more often in affluent, fiscally stronger, suburban, conservative,
and white municipalities, reproducing intra-urban differences underlying health
disparities (p. 682). Joassart-Marcelli et al. (2013) suggested that these inequalities
could derive from a variety of factors including discrimination in the distribution of
funding due to ethnicity and race; lower levels of social capital in communities with high
percentages of ethnic minorities, and especially communities with recent immigrants; and
other more important priorities for non-profit organizations in low-income areas such as
creating jobs and providing social services like drug or teenage pregnancy prevention to
underrepresented groups. Joassart-Marcelli et al. (2013) also claimed that the role of non-
profit organizations in park funding has had a growing importance, as recent policy
trends led to fiscal austerity and less public money available for parks and recreation.
To summarize, this relatively small body of literature shows that the history of
park location in relation to where demographic groups lived, combined with the analysis
of park funding in recent times, can offer some insights about the reasons for the current
level of equity in park distribution in cities and metropolitan areas. Regarding park
funding, four studies focusing on the Los Angeles metropolitan area showed the growing


71
role of non-profit organizations in allocating resources, and how non-profit organizations
further contributed to park funding inequities between wealthy and less affluent
municipalities (Joassart-Marcelli et al., 2013; Joassart-Marcelli, 2010; Pincetl, 2003).
Young Peoples Demographics and Perceptions and Use of Parks
In this section, I review the literature that analyzed how play and recreation
preferences vary by income, ethnic, and racial group; and the literature that shows how
low-income young people of color experience additional barriers to park access, besides
the lack of available parks near their residences. These two bodies of literature are related
because park designs that do not include elements to attract users from different
ethnicities might create barriers to park visitation for certain ethnic groups.
Play and recreation preferences and ethnicity. The literature reviewed by
Byrne and Wolch (2009) highlights that Non-Hispanic Whites generally look for park
settings in which they can enjoy secluded nature, whereas African Americans tend to
prefer parks where group activities can be organized, such as sports and community
events. On the other hand, Asians tend to prefer space for contemplating attractive nature,
rather than spaces for structured recreation, and Hispanics generally value more
structured park spaces including settings for group and socializing activities such as
tables, bathrooms and nearby parking (Byrne & Wolch, 2009). In particular, Gobster
(2002) found that Non-Hispanic Whites generally visit parks alone or in couples, while
ethnic and racial minorities like African Americans, Hispanics and Asians access parks in
larger groups, which shows a preference for social uses of parks. Then, Non-Hispanic
Whites access parks most often for individual active recreation such as running, biking,
and walking, while other ethnic groups tend to pursue passive recreation social activities,


72
including food-related happenings (Gobster, 2002). Also, Gobster's (2002) study shows
that within Chicagos largest park specific ethnic territories existed i.e., that different
ethnic groups took ownership of specific spaces within the park.
When analyzing park design through the cultural landscape lens, Byrne and
Wolch (2009) suggested that most parks in the United States were planned based on
English and northern European canons of beauty, which reflect ideals of pastoral
landscapes, and which might not be appealing for all ethnic groups or for people born in
other countries. For example, in Gobster's (2002) study African Americans were not
attracted by natural features as much as other ethic groups, they were more interested in
spaces for structured activities, and they were concerned about general level of
maintenance of the park.
These findings do not pertain to young people specifically, but they show general
differences among ethnic/racial groups. Also, these findings suggest that people from
various cultures and ethnicities may have different perceptions of a certain park (Byrne &
Wolch, 2009), thus highlighting the importance of park designs accommodating different
uses and cultures. Similarly, Floyd, Taylor, and Whitt-Glover (2009) suggested that
research evaluating how parks can foster physical activity in low-income ethnic groups
should include the assessment of whether parks fit the recreation preferences of the
studied group, and their perceptions of their neighborhoods and of its parks.
A few articles focused more specifically on how childrens and teenagers
outdoor play and park visitation varies by ethnicity, or on play behaviors for specific
ethnic minority young people. These studies show a few ethnic differences in play
behaviors and park visitation. Loukaitou-Sideris and Sideris (2010) found ethnic


73
differences in childrens use of sport fields and courts: Hispanic children and youth
utilized soccer fields more often than other groups, African Americans and Asian young
people used basketball courts more often than other groups, and Non-Hispanic White
young people used baseball fields more frequently than other ethnicities. Also, more
Hispanic and Non-Hispanic White young people participated in little leagues games than
African Americans and Asians (Loukaitou-Sideris & Sideris, 2010). Kaczynski, Stanis,
Hastmann, and Besenyi's (2011) study of physical activity in the parks of Kansas City,
MO highlighted significant differences in physical activity levels based on ethnicity and
gender. In particular among children, male Non-Hispanic White park visitors were more
physically active than female ethnic minority park visitors (Kaczynski et al., 2011). This
situation was reversed for teenagers, as ethnic minority female teenagers were engaged in
moderate to vigorous physical activity more often than Non-Hispanic White males
(Kaczynski et al., 2011). These findings suggest significant variations in terms of age and
ethnicity, and therefore that parks need to include a variety of settings for active and
passive recreation.
Differences in park and playground visitation among ethnic groups were also
found. Karsten (2003) highlighted that in Amsterdam the prevalent ethnic group of the
neighborhood was overrepresented in the playground. Also, Moroccan and Turkish
female children over the age of 10 visited playgrounds very rarely, possibly due to
cultural reasons (Karsten, 2003). Also, in Kansas City, MO park users from ethnic
minority groups exceeded their representation in the city as a whole (Kaczynski et al.,
2011), suggesting that ethnic minorities might have visited parks more often than Non-
Hispanic Whites. Similarly, in Los Angeles children living in the inner-city, an area with


74
more ethnic minority groups, used parks more often for outdoor recreation than children
in the San Fernando Valley, an area with a lower percentage of ethnic minorities
(Loukaitou-Sideris & Stieglitz, 2002).
In addition, some studies found that children of different ethnicities play together
in playgrounds or in organized sport activities (Blatchford, Baines, & Pellegrini, 2003;
Loukaitou-Sideris, 1995, 2003), while in other more unstructured settings and situations
play groups can be ethnically segregated (Loukaitou-Sideris, 1995). Also, sports fields
and organized sports activities like Little League games are important settings for
activities for ethnic minorities, especially for Hispanic children and teenagers
(Loukaitou-Sideris & Sideris, 2010; Perry, Saelens, & Thompson, 2011; Ries et al., 2008).
Young people living in low-income inner-city neighborhoods tend to value highly
the presence of natural settings, including parks and playgrounds (Castonguay & Jutras,
2009; Nasar & Holloman, 2013; Rudkin & Davis, 2007), probably due to a shortage of
green spaces in their neighborhoods. In particular, young people sometimes see nature as
an element of neighborhood beautification, which would encourage residents to care
more about their locale (Rudkin & Davis, 2007). Also, low-income Hispanic teenagers
living in inner-city Los Angeles looked for parks to have flexible, multiple, and
unstructured uses, including active and passive activities, while adults in the same
neighborhood want parks to have very structured and predetermined uses (Gearin &
Kahle, 2006). The availability of a large variety of play activities was also valued by
African American children living in Columbus, OH (Nasar & Holloman, 2013). The
importance of parks affording a variety of play activities is supported by many studies,
and it is discussed in the Park Quality Index for Youth (PQIY; see Chapter 5). Another


75
element that low-income inner-city ethnic minority young people perceive as important is
park maintenance and cleanliness (Gearin & Kahle, 2006; Loukaitou-Sideris & Stieglitz,
2002; Ries et al., 2008), probably due to parks in low-income ethnic minority inner city
neighborhoods having low levels of maintenance (see above discussion).
Barriers to park access for low-income young people of color. Besides lack of
available parks and amenities, low-income and ethnic minority children and teenagers
have more obstacles to play in parks and to public space use than Non-Hispanic White
middle and upper-class youth (R. Austin, 1998; Cutts et al., 2009; Ginwright & James,
2002; D. A. Harris, 2003; Neckerman et al., 2009; Platt, 2012; Spilsbury, 2005; Stodolska
et al., 2013; Valentine, 1997, 2004; Weiss et al., 2011).
First, negative neighborhood factors limit low-income and ethnically diverse
young peoples play in parks. A few studies showed that real and perceived dangers limit
low-income ethnic minority young peoples use of public space, including parks (Gomez
et al., 2004; Platt, 2012; Spilsbury, 2005; Stodolska et al., 2013). Other studies found that
low-income and ethnic/racial minority neighborhoods have higher concentrations of
crime (Cutts et al., 2009; Franzini et al., 2010; Neckerman et al., 2009; Rutledge et al.,
2003; Weiss et al., 2011), which limit childrens and teenagers park visitation.
In particular, Spilsbury's (2005) study showed that neighborhood violence has
significant effects on childrens independent mobility depending on gender and age. In
neighborhoods with high violence, girls home range was significantly lower when
travelling alone compared to travelling in groups, while no difference was found for boys
(Spilsbury, 2005). Also, Platt's (2012) investigation of a low-income neighborhood in
Milwaukee highlighted that children are afraid to access parks and other public amenities


76
due to the presence of gangs or unknown teenagers. Instead, children prefer to play in
alleys, vacant lots, and sidewalks, in which they feel safer because these settings are free
from gangs (Platt, 2012), and possibly subject to passive surveillance from residences.
Similarly, Castonguay and Jutras's (2009) study of low-income childrens place
preference in Montreal suggested that outdoor spaces around homes such as backyard
gardens or shared green spaces in public housing were the settings where children felt
safer, also due to the presence of familiar people.
Studies conducted in Chicago indicated that Hispanic teenagers limit their access
to parks due to fears of crime, such as gang violence occurring in public settings like
streets and parks (Shinew, Stodolska, Roman, & Yahner, 2013; Stodolska et al., 2013).
On the other hand, a study conducted in Los Angeles showed that in low-income areas
parks can also be perceived as relatively safe if compared to streets, especially during the
day (Loukaitou-Sideris, 1995). Also, Hispanic teenagers in Chicago consider recreational
activities that take place within institutionalized settings like school grounds and that are
supervised by adults as safer (Shinew et al., 2013; Stodolska et al., 2013). Similarly,
another study in Los Angeles highlighted that the presence of a community center near a
park made the park feel safer than the rest of the neighborhood, which children perceived
as dangerous (Loukaitou-Sideris, 2003). Therefore, the presence of well-reputed
institutions with trusted adults increases childrens and teenagers sense of safety in low-
income and ethnic minority neighborhoods.
Studies about adults sense of safety in relation to the demographic composition
of their neighborhoods are also important, as caregivers generally influence their
childrens mobility licenses. Franzini et al. (2010) highlighted that, in three cities across


77
the United States, residents of low socio-economic areas perceived their neighborhoods
as "less safe, with more physical disorder, decay, and vacant lots/houses" (p. 272). A
study conducted in St. Louis, MO showed that African Americans tend to consider their
community as more dangerous and less aesthetically pleasing than Non-Hispanic Whites
perceptions, despite the demographic characteristics of their community (Boslaugh, Luke,
Brownson, Naleid, & Kreuter, 2004). The subpar quality of the physical environment and
the higher safety concerns that low-income communities of color experience can also be
considered environmental justice issues (W. C. Taylor, Floyd, Whitt-Glover, & Brooks,
2007; W. C. Taylor, Poston, Jones, & Kraft, 2006).
When looking at reported crime data, Cutts et al. (2009) found that the percentage
of Hispanic immigrants living near parks in Phoenix, AL was positively correlated to the
frequency of crime. The percentage of people under 18 was also positively associated to
the number or crimes in proximity of parks (Cutts et al., 2009). In New York City, Weiss
et al. (2011) found that the least affluent areas of the city included more reported crime,
less pedestrian safety and more locally unwanted land uses, such as industrial areas.
Second, research shows that race, class, and age influence young peoples use of
public space regardless of crime and other social environmental factors (R. Austin, 1998;
Norris & Armstrong, 1999; Valentine, 2004). Youth of color experience issues in
navigating public space due to racial profiling, police harassment or vigilante violence (R.
Austin, 1998; C. J. Cohen, 2010; D. A. Harris, 2003; Norris & Armstrong, 1999). Also,
when looking at age groups, children in public space tend to be socially depicted as
victims and teenagers as troublemakers, leading to both groups being restricted in public
space use (Valentine, 1997, 2004).


78
In particular, Norris and Armstrong's (1999) study of British public spaces
monitored through Closed Circuit Television (CCTV) showed that black male youth were
the public space users most commonly targeted by police or vigilantes. Disturbingly, in
most cases black people were targeted not for their behaviors but for their skin color
(Norris & Armstrong, 1999). Also, according to Austin (1998) several local, state, and
federal laws in the United States created de facto barriers to recreation for blacks and
other racial and ethnic minorities, including limitations to access public and private
venues (for example, by denying the use of public venues for hip hop or reggae concerts),
heavy patrolling of recreational gatherings with high percentages of ethnic minorities,
curfews targeted to ethnic minorities neighborhoods, and lack of public transit to access
recreational sites.
When looking specifically at parks, in the past many U.S. cities, and especially
municipalities located in the South, had park systems that were racially segregated
(Boone et al., 2009; Byrne & Wolch, 2009). In particular, park systems included specific
parks that only Non-Hispanic Whites could access and other parks, which were smaller
and less maintained, that African Americans could use (Boone et al., 2009; Byrne &
Wolch, 2009). Some forms of subtle segregations still exists in parks, such as the
prohibition of playing specific games like soccer might be intended to exclude certain
park users (Martin, 2004). Also, Byrne's (2012) study of an urban national park located in
Los Angeles showed that Hispanics felt unwanted in such a park for a variety of reasons:
the majority of park users were Non-Hispanic Whites; the areas located around the park
were mostly Non-Hispanic White and upper-class; the park did not include signage in
Spanish; Hispanics were concerned about being harassed or victimized; and some


79
Hispanics reporting actual occurrences of intolerance and prejudice. Similarly, African
Americans in a Chicago urban park were concerned about being prejudged by other park
visitors, park workers, and police (Gobster, 2002).
Finally, studies of low-income youth of color show that some teenagers have one
or more full-time jobs to support their families (Ginwright & James, 2002; Williams &
Kornblum, 1985), thus time constraints limit their opportunities to be in public space. In
particular, Byrne and Wolch (2009) suggested that low-income peoples use of parks
might be limited by time constraints due to work.
To summarize, these two bodies of literature suggest that, besides lack of
available parks, low-income ethnic minority children and teenagers experience three
additional types of barriers to outdoor recreation in their neighborhood. The first derives
from negative neighborhood factors like crime and perceived dangers often associated
with parks; the second deals with issues in navigating public space due to racial profiling,
police harassment or vigilante violence; and the third involves having limited time for
outdoor recreation due to other commitments.
Park Availability and Young Peoples Use of Parks
This section reviews the scholarship that investigated whether the presence of
parks and spaces for play close to childrens and teenagers homes fostered their
visitation of such spaces. A recent review of qualitative research about park access
showed associations between the presence of a park within walking distance and repeated
access to that park (McCormack et al., 2010). Also, several recent quantitative and
qualitative empirical studies highlighted that the availability of nearby parks is a
necessary condition for childrens and teenagers use of parks (Boone-Heinonen,


80
Casanova, Richardson, & Gordon-Larsen, 2010; D. A. Cohen et al., 2014; Jansson &
Persson, 2010; Kytta, 2004; Loukaitou-Sideris & Sideris, 2010; Ries et al., 2008, 2009;
Silver, Giorgio, & Mijanovich, 2014; Slater, Fitzgibbon, & Floyd, 2013; Veitch et al.,
2006).
The connection between park proximity and young peoples visitation of parks
can be found in different geographical contexts and in cities of different sizes, including:
small towns in Finland and Belarus (Kytta, 2004), small towns in Sweden (Jansson &
Persson, 2010), Melbourne, Australia (Veitch et al., 2006), Baltimore, MD (Ries et al.,
2008), Los Angeles, CA (Loukaitou-Sideris & Sideris, 2010), and New York City, NY
(Silver et al., 2014). In particular, Loukaitou-Sideris and Stieglitz (2002) found that park
proximity mattered more for inner-city children than for children living in suburban areas,
as parks were the only opportunity for outdoor recreation in inner cities. In inner cities,
park proximity also matters due to childrens and teenagers limited means of
transportation (Loukaitou-Sideris & Sideris, 2010; Ries et al., 2008). Also, some of the
studies finding connections between park availability and park use focused on ethnic
minorities like African Americans and Hispanics (Ries et al., 2008, 2009; Silver et al.,
2014; Slater et al., 2013); thus showing the importance of park proximity for ethnic
minorities. The presence of nearby parks can also be a positive contribution to children
and teenagers active play and physical activity (Boone-Heinonen et al., 2010; D. A.
Cohen et al., 2014).
However, while the presence of parks in young peoples communities matters for
their park visitation, other factors also influence whether children and teenagers access
parks or not, including: caregivers mobility licenses, park quality, and park and


81
neighborhood safety. First, as previously discussed, caregivers fears of strangers limits
childrens licenses to play outdoor (Valentine, 1997; Veitch et al., 2006). Further, the
amenities included in parks and their level of maintenance also influence childrens and
teenagers park visitation (McCormack et al., 2010; Ries et al., 2009). In particular, Ries
et al.'s (2008) study of African American teenagers in Baltimore highlighted that
teenagers looked for parks with good maintenance and with sport facilities. Also,
teenagers tended to avoid parks that were poorly maintained and that included physical
hazards (Ries et al., 2008). Park aesthetics and the presence of age-appropriate amenities
were also connected to park visitation (Ries et al., 2009). In Chapter 5,1 review more in
depth the literature on childrens and teenagers outdoor play and distill a few factors of
park quality in relation to young peoples needs.
Also, as introduced in the previous section, studies conducted in low-income
communities of colors show that, even if parks are available, children and teenagers
might not use them because they perceive them as dangerous (Platt, 2012; Ries et al.,
2008; Slater et al., 2013; Veitch et al., 2006; Veitch, Salmon, & Ball, 2007b) or because
they perceive the neighborhood as unsafe (Carver, Timperio, & Crawford, 2008; Griffin,
Wilson, Wilcox, Buck, & Ainsworth, 2008; Jarrett, Bahar, McPherson, & Williams,
2013; Loukaitou-Sideris & Sideris, 2010; Malone, 2007; Molnar, Gortmaker, Bull, &
Buka, 2004) due to gangs or physical hazards. However, a study by Outley and Floyd
(2010) showed that African American caregivers, even when they perceived their
neighborhood as unsafe, often adopted strategies that allowed children to access parks,
including other adults supervision through affiliation networks (family and friends) and
structured recreation activities.


82
This literature shows that, although park proximity is necessary for park visitation,
especially among inner-city ethnic minority youth, other factors affect park visitation,
including park and neighborhood safety, and park quality. Thus, the connection between
park availability and park use is problematic, as the presence of nearby parks is a
necessary but not sufficient condition for park visitation.
Park Availability and Environmental Justice
To summarize, the findings of these bodies of literature are disturbing from an
environmental justice perspective, as they reveal documented inequities in how children
and teenagers of different ethnicities and social classes have opportunities to access parks
and green spaces. Although parks are not always perceived as amenities in low-income
communities of color, they at least are public spaces with the potential to serve their
communities (Boone et al., 2009). These findings also show that in most U.S. cities urban
planning failed, intentionally or unintentionally, to provide equal and equitable access to
urban parks for all ethnic and income groups. According to Talen (2010), urban parks are
the single most important category of publicly owned open space in US cities (p. 473).
Furthermore, urban parks bring economic, public health, and environmental benefits to
cities (Byrne & Wolch, 2009; Hamik & Crompton, 2014; Sherer, 2006). These benefits
are another reason why it is important to understand the spatial distribution of parks in
cities and the reasons for such distribution better. In the next section, I highlight gaps and
inconsistencies in the literature about park accessibility, which lead to my research goals
and questions.


83
Literature Gaps and Inconsistencies
Main Gap: Lack of Focus on Just Processes, Just Outcomes, and Park Quality
This literature review showed a series of gaps in relation to young peoples access
to play in parks, which I briefly outlined in the introduction. Based on my extensive
review of the literature, the main gap is that no previous study on access to parks has
systematically integrated planning policies, design and planning practices, young
peoples perceptions and use of parks, and young peoples socio-economic and ethnic
characteristics in the investigation of the spatial distribution of parks across income and
ethnic groups (see theoretical framework). The integration of these constructs is
important for two main reasons.
First, as previously discussed, David Schlosberg's (2004) definition of
environmental justice includes both just distribution of public amenities and
environmental hazards and just decision-making processes that led to such distributions.
Therefore, it is important that studies focusing on access to parks as an environmental
justice issue integrate the foci on just processes and just outcomes, intended as
todays spatial distribution of parks. Although a variety of which-came first studies
(Been & Gupta, 1997; Boone, 2002; Hurley, 1997; Pulido, 2000) explored whether
environmental bads have been intentionally located in low-income and ethnic minority
communities, only a few recent studies on access to parks (Boone et al., 2009; Pincetl,
2003; Wolch et al., 2005) started exploring the reasons behind the spatial distribution of
parks among different income and ethnic groups. Boone et al. (2009) argued that a
shortcoming of the environmental justice literature on access to environmental goods is
that, in many studies, claims about injustice and racism are made based on distributional


84
aspects only, without assessing the processes that led to the observed distributional
patterns. Based on my literature review, this is particularly true for studies focusing on
access to parks.
Second, as the literature on childrens and teenagers outdoor play shows that
park quality matters for park use (see Chapter 5), it is important to include an index of
park quality in the analysis of park distribution, which can be developed based on the
literature on childrens and teenagers park use and park preference. Therefore, studies on
access to parks as an environmental justice issue should integrate the foci on just
processes, just outcomes, and which parks really matter? According to my review,
no study on access to parks brought together these three key pieces.
Less Substantial Gaps
Besides this main fundamental gap, my literature review showed a few minor
gaps that need to be addressed. First, relatively few studies on access to parks as an
environmental justice issue included the percentage of young people in the spatial and
statistical analysis (Cutts et al., 2009; Franzini et al., 2010; Loukaitou-Sideris & Stieglitz,
2002; Talen & Anselin, 1998; Talen, 2003; Wolch et al., 2005). Not including such
percentages means assuming that children and teenagers are equally distributed in the city,
without taking into account the areas that have more demand for play spaces. Although
qualitative Geographic Information System (GIS) is a growing field (Kahila & Kytta,
2009; Kytta, Broberg, Tzoulas, & Snabb, 2013), no previous study about access to parks
has integrated qualitative research on childrens and teenagers use and perception of
parks in a GIS spatial analysis.


85
Second, a few authors highlighted that research on access to parks did not
sufficiently focus on park quality and suggested that future studies should investigate
more in detail the quality of parks and of play amenities, with a particular emphasis on
the needs of the age and cultural groups that will mostly benefit from them (Boone et al.,
2009; Eliaway et al., 2007; Kabisch & Haase, 2014; Loukaitou-Sideris & Stieglitz, 2002;
Reyes et al., 2014; Smoyer-Tomic et al., 2004; Talen, 2003; A. F. Timperio et al., 2007).
This links to the necessity to investigate which parks really matter for children and
teenagers. In particular, some authors recommended considering young peoples play
needs (Boone et al., 2009; Rigolon & Flohr, 2014), and others pointed out that
recreational preferences tend to differ by ethnic and racial groups (Kabisch & Haase,
2014; Madsen et al., 2014), also due to safety concerns in low-income ethnic
neighborhoods (A. F. Timperio et al., 2007; Weiss et al., 2011). In terms of the methods
to evaluate these differences, Loukaitou-Sideris and Stieglitz (2002) suggested that
"quantitative methods cannot sufficiently measure quality" (p. 479), thus supporting the
use of qualitative methods to analyze park quality for different age and ethnic groups.
These suggestions further reinforce the need to include young peoples perceptions and
use of parks in analyzing childrens and teenagers access to urban nature as an
environmental justice issue.
Third, the land use and zoning around parks are other aspects that have not been
taken into account by most studies, with a few exceptions (e.g. Boone et al., 2009; Cutts
et al., 2009; Talen, 2010). Land uses and zoning in the proximity of parks need to be
included because different land uses and zoning regulations are likely to generate
different park needs (Talen, 2010). For example, areas with a prevalence of multi-family


86
housing, which are generally located in low-income ethnic areas (Wolch et al., 2005),
need more parks than areas composed of single-family residences, as the latter generally
have private yards. Also, in the past planners and decision makers have used land use
policies to locate undesired land uses far away from middle and upper class Non-
Hispanic Whites (Boone et al., 2009). Thus, land use policies have unintentionally
contributed to how different socio-economic and income groups can benefit from parks
today. This is another reason why land use and zoning policies need to be taken into
account when studying access to parks.
Finally, as previously highlighted, while some studies have proposed solutions to
increase access to play in parks for young people living in underserved areas (Nemeth et
al., 2012; Rigolon & Flohr, 2014; Sister et al., 2009; Talen, 2003; Wolch et al., 2005), to
my knowledge only one article about access to parks (Rigolon & Flohr, 2014) suggested
a framework for neighborhood greening in partnership with local communities.
Methodological Variations and Inconsistencies
Besides the gaps I highlighted, this literature review showed three main
methodological variations or inconsistencies. The first methodological variation is that, in
the park access literature, accessibility and proximity to public amenities have been
measured through different instruments. Most of the studies I reviewed used a
Geographic Information System (GIS) spatial analysis to measure distances between
residences and public amenities, while fewer studies used surveys of residents, asking
them to report their distance from various public amenities (see above section). GIS
analyses may highlight more reliable results about physical spatial distributions than
surveys because they rely on objective measures of the physical environment rather than


Full Text

PAGE 1

INEQUITIES IN YOUNG PEOPLE'S ACCESS TO URBAN PARKS: AN ENVIRONMENTAL JUSTICE INVESTIGATION IN DENVER By ALESSANDRO RIGOLON B. Arch., M. Arch. University of Bologna, 2007 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Design and Planning 2015

PAGE 2

ii 2015 ALESSANDRO RIGOLON ALL RIGHTS RESERVED

PAGE 3

iii This thesis for the Doctor of Philosophy degree by Alessandro R igolon has been approved for the Design and Planning Program by Jeremy R. NÂŽmeth, Chair Louise Cha w l a, Advisor Nancy L. Leech Sharon E. Sutton Willem van Vliet July 24 2015

PAGE 4

iv Rigolon, Alessandro (Ph.D., Design and Planning) Inequities in Young P eople's Access to Urban Parks: An Environmental Justice Investigation in Denver Thesis directed by Associate Professor Jeremy R. NÂŽmeth ABSTRACT Given the benefits of repeated contact with nature for young people, inequities in park spatial distribution b etween young people of differen t socio economic status and ethnicity are a serious environmental justice issue. A growing body of literature has documented that low income young people of color have access to fewer acres of parks and to parks with lower qu ality than wealthier Non Hispanic White young people. However, very few studies have investigated the planning process es that have contributed to these inequities. Also, no previous investigation on access to parks has combined qualitative research on youn g people's use and perception of parks in a geospatial analysis of park spatial distribution. In this dissertation I integrate d the analyses of equitable decision making processes, of equitable park distributions, and of park quality in relation to young people's needs to understand how planning policies and practices contributed to park provision for different income and ethnic groups with a particular focus on parks that can be me aningful for young people Draw ing from geospatial methods, interviews wit h experts, and policy analysis, I uncovered a complex picture of environmental injustice, including inequities in the spatial distribution of parks and in the policies and practices that led to such distribution. My geospatial analysis showed that low inco me ethnic minority young people have very low access to safe parks and to parks with excellent levels of quality which might substantially limit their park visitation. Also

PAGE 5

v the Poundstone Amendment, a state policy limiting Denver's annexations from surro unding counties aimed to avoid ethnic integration in schools, combined with the lack of an impact fee for parks in infill developments has significantly hindered Denver's capacity to improve park provision, particularly for its low income ethnic minority people. The mixed research integration of qualitative and quantitative datasets showed that most cases of distributional injustice are linked to unjust processes, while very few cases of partial distributional equity are related to just processes This dis sertation has important implications for park and land use planning, including current barriers and possible solutions to increasing park equity, for environmental justice research and practice, and for the use of mixed research in planning. The form and content of this abstract are approved. I recommend its publication. Approved: Jeremy R. NÂŽmeth

PAGE 6

vi Para SofÂ’a

PAGE 7

vii ACKNOWLEDGEMENTS At the end of this four year journey, I am overwhelmed by feelings of gratitude. I am grateful for the beautiful stat e of Colorado, with its mountains, its sunsets, and its people who have welcomed me, supported me, and made me feel at home. At the University of Colorado, the Children, Youth, and Environment Center and its wonderful people have been a safe and stimulati ng harbor for a foreign student looking for an anchor and new perspectives. Willem, Louise, Tori, Mara, and other fellow students and interns have made this journey an exciting and smooth ride. Jeremy has been a recent and inspiring addition to the great t eam of mentors I was lucky to work with, and I wish our paths had crossed earlier. Sharon and Nancy have been exceptional for my growth as a researcher, from two different perspectives. Sharon welcomed me at the University of Washington in Seattle five yea rs ago when this journey started, as I walked into her room with an Italian English dictionary My fellow PhD students, particularly Travis and Mehdi, have also been incredible travel mates, and greatly helped with my research. Colorado has also gifted m e with a family and a new home. I would not have met my wife, Sofia, if I had not started a PhD program in Colorado. She is the greatest among all great gifts of life Sofia and her family, Suse, Manuel and Fabian, have become my family They have become a nother home, away from home, and helped me cope with missing my parents and friends in my native land, Italy. My parents, who are also coping with their only child living oceans away deserve the greatest gratitude. They accepted my dream of pursuing an ac ademic career and a life full of a d ventures in the United States. I hope to make the most of the next journey s that Sofia and I will share

PAGE 8

viii TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ 1 Denver as a Compelling Case Study ................................ ........................... 8 Denver's Population Growth and Park System ............................... 8 Denver's Demographics ................................ ................................ .. 9 Denver's History of Ethnic Residential Segregation ..................... 11 Denver's Environmental Justice Issues ................................ ......... 12 Generalizability ................................ ................................ ............. 13 A Complex Picture of Environmental Injustice ................................ ........ 14 Relevance to Urban Planning ................................ ................................ .... 17 Definition of Key Terms ................................ ................................ ........... 19 II. THEORETICAL FRAMEWORK, LITERATURE REVIEW, AND RESEARCH QUESTIONS ................................ ................................ ................... 23 Theoretical Framework ................................ ................................ ............. 23 Identification of Constructs ................................ ........................... 23 Relationships between Constructs ................................ ................. 33 Transformative Paradigm and Environmental Justice ................... 35 Literature Review ................................ ................................ ...................... 43 Park Spatial Distribution and Demographics ................................ 44 Policies, Practices, and Park Spatial Distribution ......................... 61

PAGE 9

ix Young People's Demographics and Perceptions and Us e of Parks ................................ ................................ ................................ ....... 71 Park Availability and Young People's Use of Parks ..................... 79 Park Availability and Environmental Justice ................................ 82 Literature Gaps and Inconsistencies ................................ .......................... 83 Main Gap: Lack of Focus on Just Processes, Just Outcomes, and Park Quality ................................ ................................ ................... 83 Less Substantial Gaps ................................ ................................ .... 84 Methodological Variations and Inconsistencies ............................ 86 Research Objectives and Q uestions ................................ ........................... 88 Research Objectives ................................ ................................ ...... 89 Research Questions ................................ ................................ ....... 90 Mixing Purposes ................................ ................................ ............ 92 III. RESEARCH METHODS ................................ ................................ ................ 94 Population and Sampling Frame ................................ ............................... 94 Park Planning Processes Analysis ................................ ................. 95 Park Spatial Distribution Analysis ................................ ................ 97 Research Design ................................ ................................ ........................ 98 Data Collection Instruments ................................ ................................ .... 101 Park Planning Processes ................................ .............................. 102 Park Spatial Distribution ................................ ............................. 106 Data Analysis ................................ ................................ ........................... 107

PAGE 10

x Park Planning Process Data Analysis ................................ .......... 107 Park Spatial Distribution Data Analysis ................................ ...... 110 Mixed Research Integration ................................ ........................ 128 Data Legitimation ................................ ................................ .................... 131 IV. PARK PLANNING PROCESS FINDI NGS ................................ ................. 135 An Introduction to Denver's History ................................ ....................... 138 Denver's Early Stages: 1858 1902 ................................ .............. 141 The City Beautiful and New Deal Eras: 1902 1945 .................... 142 Suburbanization: 1945 1983 ................................ ....................... 162 Urban Renaissance: 1983 2015 ................................ ................... 171 Looking Back to Look Ahead ................................ ..................... 178 Park Planning ................................ ................................ ........................... 178 Parkland Acquisition and Park Construction .............................. 180 Park Design, Management, and Use ................................ ............ 243 Park Planning in Denver: A Brief Summary ............................... 266 Land Use Planning ................................ ................................ .................. 267 Land Use Planning and Park Establishment ................................ 268 Land Use Planning and Residential Location ............................. 281 Land Use Planning for Park Establishment and Residential Location: A Summary ................................ ................................ 304 Housing Policies and Practices ................................ ................................ 305

PAGE 11

xi Ethnic and Racial Discrimination in Housing: Policies and Practices ................................ ................................ ....................... 307 Affordable Housing in Denver ................................ .................... 317 Housing Policies and Practices: A Summary .............................. 330 Pa rk Planning Processes: Bringing It All Together ................................ 332 Park Planning Processes: 1902 1945 ................................ ........... 334 Park Planning Processes: 1946 1988 ................................ ........... 338 Park Planning Processes: 1989 2015 ................................ ........... 344 Park Planning Processes: A Brief Comparison ........................... 351 V. PARK SPATIAL DISTRIBUTION FINDINGS ................................ ........... 354 Developing a Park Quality Index for Youth ................................ ............ 354 Dates, Locations and Methods ................................ .................... 355 What Green Space Features Matter for Children and Teenagers? ................................ ................................ ................................ ..... 356 Operationalizing These Findings in a Park Quality Index for Youth (PQIY) ................................ ................................ ......................... 368 Pa rk Spatial Distribution ................................ ................................ ......... 381 Descriptive Statistics ................................ ................................ ... 382 Summary of Findings ................................ ................................ .. 389 Analysis of Park Proximity ................................ ......................... 400 Analysis of Park Acreage ................................ ............................ 413 Analysis of Park Quality ................................ ............................. 433 Park Spatial Distribution: Discussion ................................ .......... 472

PAGE 12

xii VI. MIXED RESEARCH INTEGRATION ................................ ........................ 478 Data Reduction and Data Display ................................ ........................... 482 Data Comparison ................................ ................................ ..................... 483 Does the Partially Equitable Distribution of Parks Based on Certain Variables Correspond to Equitable Processes? ............... 4 85 Does the Inequitable Distribution of Parks Based on Certain Variables Correspond to Discriminatory or Negligent Processes? ................................ ................................ ................................ ..... 494 Do Perceptions of Distributional Inequity and GIS Measured Distributional Inequity Coincide or Differ? ................................ 501 Do Notable Policies, Plans, and Practices Developed at Different Times Relate to Parks Established at Specific Times? ................ 503 Do Policies, Plans, and Practices Addressing Different Parts of the City Relate to Park P rovision in Specific Parts of the City? ....... 506 Data Comparison: A summary ................................ .................... 510 Extreme Case Analysis ................................ ................................ ............ 511 Why Do Some Low Income Areas Have Good Access to Regional Parks? ................................ ................................ .......................... 512 Why Do Some Low Income Areas Have Good Access to Best PQIY Parks? ................................ ................................ ................ 515 Why Do Some High Income Areas Have Poor Access to Best PQIY Parks? ................................ ................................ ................ 518 Why do Some Low Income Areas Have Good A ccess to Parks with Low Violent Crime Density? ................................ .............. 520 Why Do Some Low Income Ethnic Minority Areas Have Good Access to All Parks and Parkways, with Parkways Making a Difference? ................................ ................................ .................. 522

PAGE 13

xiii Why Do Some Low Income Neig hborhoods That Are Perceived as Park Poor Have Decent GIS Measured Access to Parks in Terms of Park Proximity? ................................ ................................ ....... 529 Summary of Extreme Case Analyses ................................ .......... 530 Data Consolidation ................................ ................................ .................. 531 Park P rovision in the West Side ................................ .................. 546 The Growing Role of Non Profits for Park Distributional Equity ................................ ................................ ................................ ..... 548 Procedural Injustice and Its Outcomes: A Summary .............................. 550 VII. IMPLICATIONS ................................ ................................ ......................... 553 Implications for Park Planning ................................ ................................ 554 Obstacles to Park Establishment: Today and Looking Ahead .... 554 Changes in Municipal Policies and Park Planning to Facilitate Park Establishment ................................ ................................ .............. 558 Implications of the Current Spatial Distribution of Parks ........... 566 Parameters and Thresholds for Park Quality ............................... 569 Implications for Park Planning: A Summary .............................. 573 Implications for Environmental Justice ................................ ................... 573 Implication s for Environmental Justice Research ....................... 574 Implication s for Environmental Justice Practice: A Park Advocacy Framework ................................ ................................ ................... 577 Implications for Mixed Research in Planning ................................ ......... 586 Limitations ................................ ................................ ............................... 596 Design Limitations ................................ ................................ ...... 597

PAGE 14

xiv Limitat ions in Specific Methods ................................ ................. 600 Directions for Future Research ................................ ................................ 603 Continuing Studies of Spatial Distribution and Park Quality ..... 604 In Depth Study of Unexpected Findings ................................ ..... 605 A Complex Picture of Environmental Injustice ................................ ...... 607 REFERENCES ................................ ................................ ................................ ................ 615 APPENDIX A ................................ ................................ ................................ ................. 668 APPENDIX B ................................ ................................ ................................ .................. 696 APPENDIX C ................................ ................................ ................................ .................. 705

PAGE 15

CHAPTER I INTRODUCTION Research conducted in several Western countries shows that the time children and teenagers spend playing outdoors and in nature is decreasing (Clemens, 2004; Fyhri, Hjorthol, Mackett, Fotel, & KyttŠ, 2011; Karsten, 2005; Tandy, 1999; Wridt, 2004) This decrease is problematic, given the benefits of contact with nature for mental and physical health among young people (McCurdy, Winterbottom, Meht a, & Roberts, 2010) for their overall well being (Chawla, Keena, Pevec, & Stanley, 2014; Chawla, 2014; N. M. Wells, 2014) and for their development (Burdette & Whitaker, 2004; Fjortoft & Sageie, 2000; R. C. Moore, 1986a; N. M. Wells, 2000) The growing attention to "nature deficit disorder" described by Richard Louv (2005) le d to the establi shment of several programs and policies aimed to increase children's repeated contact with nature (see Natural Learning Initiative, 2014; No Child Left Inside, 2007) In 2007, the World Fu ture Society predicted that nature deficit disorder will increase globally as a health risk (World Future Society, 2007) In urban areas, young people mostly rely on neighborhood parks for accessing nature, as their home range is often limited to their neighborhood (Fagerholm & Broberg, 2011; Loebach & Gilliland, 2014) Also, since the world is urbanizing at very high rates (United Nations, 2014) and as the United States is experiencing a n urban renaissance for the first time in decades (S. G. Wilson et al., 2012) nearby parks are becoming more and more important for access to nature. However, a recent review showed that low income and ethnic minority childre n and teenagers have even less contact with nature than Non Hispanic W hite middle and upper class children and teenagers due to lack of available parks and recreational

PAGE 16

2 opportunities in their communities (National Recreation and Park Association, 2011) A study by Downey (2007) suggested that spatial ineq ualities in access to parks are often related to residential segregation. The inequalities are particularly disturbing because contact with nature has stronger benefits for low income children in terms of stress relief (N. M. Wells & Evans, 2003) Neighborhood parks can be democratizing public spaces that help balance structural inequities in recreation and public health for low income ethni c minorities. L ow income ethnic minority children and teenagers need nearby parks more than other groups given their limited transportation options (Floyd & Johnson, 2002; Loukaitou Sideris & Stieglitz, 2002; Ries et al., 2008) and given their inadequate access to private recreational opportunities (Loukaitou Sideris & Stieglitz, 2002) Neighborhood parks matter the most for public health among ethnic minorities, as African American and Hispanic young people have higher rates of obesity than their Non Hispanic White counterparts (Ogden, Carroll, Kit, & Flegal, 2014; Wang & Beydoun, 2007) Also, low income young people of color tend to have lower levels of physical activity than other groups (Floyd, Taylor, & Whitt Glover, 2009) Therefore, when measuring access to park s using the equality of spatial distribution as a desired result is problematic because equality does not take into account who needs parks the most (Boone, Buckley, Grove, & S ister, 2009) Rather, it is most appropriate to use the concept of equity that, although harder to measure, considers the demographic groups who are more park dependent (Boone et al., 2009) As my literature review will show, low income ethnic minority groups might live as close or closer to parks than m ore privileged group s, but low income communities of

PAGE 17

3 color have less park acreage per person and lower quality parks, depending on the city. T hese inequities in access to parks for different ethnic and income groups show that in many U.S. cities park planning land use, and h ousing policies failed, intentionally or unintentionally to provide equal access to parks to all children and teenagers These inequities in recreation are part of a wider history of ethnic discrimination that ethnic minorities have experienced in the Uni ted States, including housing, employment, and education (Boone et al., 2009; Wolch, Wilson, & Fehrenbach, 2005) The recent ruling of the U.S. Supreme Court on the interpretation of 1968 Fair Housing Act case established that allegations of ethnic discriminat ion are not restricted to discriminatory intents (U.S. Supreme Court, 2015) Rather, discrimi natory outcomes like the exclusion of ethnic minorities from certain neighborhoods, are sufficient evidence to make claims about the existence of discrimination (U.S. Supreme Court, 2015) T o make allegations of ethnic discrimination in housing it is sufficient to show evidence about the effect of exclusionary zoning (residential segregation) wi thout finding evidence about intents to exclude specific demographic groups. The 1968 Fair Housing Act adopted a similar approach in regards to discrimination against ethnic minorities (U.S. Supreme Court, 2015) In this dissertation, I use this approach to define evidence of discrimination. Thus to claim inequity and discrimination in recreation evidence about inequitable distribution of parks is sufficient regardless of clear discriminatory intent in policies and practices. Furthermore, safety concerns in low income and ethnic minority communities create additional barriers to accessing park s f or children and teenagers (Evans, 2004; G—mez, Johnson, Selva, & Sallis, 2004; Platt, 2012) Regardless of income level, young

PAGE 18

4 people's use of park s and their benefits differ depending on park design, including the presence of diffe rent play amenities (Dyment, Bell, & Lucas, 2009; Potwarka, Kaczyn ski, & Flack, 2008; Ries et al., 2008; Staempfli, 2009; Veitch, Salmon, et al., 2014) and on park safety (McCormack, Rock, Toohey, & Hignell, 201 0; Platt, 2012; Ries et al., 2008; Stodolska, Shinew, Acevedo, & Roman, 2013) These bodies of literature establish that planning for young people 's access to urban nature has park planning, land use planning, housing, public health, child development a nd environmental justice implications. However, no previous study on access to parks has integrated planning policies, planning and design practices, young people 's perceptions and use of park s, and young people 's socio economic and ethnic characteristics in the investigation of the spatial distribution of urban parks across income and ethnic groups. For example, most equity mapping literature focusing on access to public amenities, with few exceptions (see Boone et al., 2009) did not investigate the reasons behind the spatial i nequalities in the distribution of parks. According to Schlosberg (2004) environmental justice investigations should focus both on the spatial distribution of public amenities and environmental hazards, and on the decision making processes to determine where amen ities and hazards are located. Also, no previous investigation on access to parks has integrated qualitative research on children and teenager's use and perception of parks in a Geographic Information System (GIS) spatial analysis Then, very few studies h ave investigated access to park s differing in terms of size, play amenities, and presence of nature which are all features that influence children's and teenagers' play behaviors and perceptions of park s. I argue tha t the concurrent focus on whether decis ion making processes, park distribution, and park

PAGE 19

5 quality are just in relation to children's and teenagers' needs is necessary to gain a deeper understanding of the intersections between park distribution, the policies and practices that created such distr ibution, and park needs. This is the focus of my dissertation. This dissertation also seeks to address a general gap in the discourse on sustainable planning. According to Campbell (1996) planners should seek sustainable solutions that b alance the often competing instances of environmental quality, economic development and social equity, which are known as the three E's of sustainability. Thus as parks can be considered a form of sustainable infrastructure (Byrne & Wolch, 2009; Chiesura, 2004; Cranz & Boland, 2004) their establishment and maintenanc e should reflect the three E' s framework: bringing environmental and economic benefits to cities and their people while serving the populations who need them most (equity) However, w h ile several studies have investigated the environmental and public healt h benefits of sustainable infrastructures like parks, greenways, and bike trails (Bedimo Rung, Mowen, & Cohen, 2005; A. C. K. Lee & Maheswaran, 2011; Maller et al., 2009; McCo rmack et al., 2010; Potwarka et al., 2008; Sherer, 2006) less research has addressed whom these sustainable infrastructures are really benefitting (Boone et al., 2009; Wolch et al., 2005) In other words, not enough research has investigated the beneficiaries of sustainability interventions like parks Therefore, this dissertation is also motivated by the need t o highlight the key role of environmental justice in sustainability discourses and practices (Boone, 2010) also following Agyeman and Evans' (2004) call for "just sustainability" (p. 155). To address these gaps, the general goal of this research is to understand how planning policies and practices in Denver, Colorado affe cted the way children and

PAGE 20

6 teenagers of different demographic groups can access public parks, and specifically public parks with high quality and maintenance. To do so, I focus on the planning policies and practices that determined children's and teenager's access to significant play opportunities in urban park s in relation to their socio economic status and ethnicity. Building on a pilot study (Rigolon & Flohr, 2014) I conduct a m ixed research investigation (Teddlie & Tashakkori, 2009) t o uncover how park planning land use planning and housing policies and practices in Denver have influenced the current equity or inequity of the distribution of parks, with a specific focus on parks with high quality amenities for children and teenagers. The first strand of the study referred as Park Planning Processes focuses on how park planning, land use plan ning, and housing policies and practices in Denver have led to today's parks distribution (analysis of procedural justice). In particular, I study whether park planning, land use planning, or housing policies and practices have had the biggest relative wei ght in influencing the distribution of parks and of high quality parks in relation to the location of different demographic groups. To gain a general understanding of the equity or inequity of planning processes that influenced young people 's access to par ks and high quality parks, I explore whether certain demographic groups were intentionally left out from access to parks (for example, planners located parks most often in richer neighborhoods) or whether in other circumstances planners tried to fill the g aps in park distribution and deliberately located parks in areas of the city with higher concentrations of low income ethnic children and teenagers Also, to understand historical processes of park location in r elation to demographics, I map ped the residen tial

PAGE 21

7 location of African Americans and Hispanics in relation to park location in several moments of Denver's history. In t he second strand of the study referred as Park Spatial Distribution I employ a GIS spatial analysis of park distribution, with a spe cific focus on parks that, based on the qualitative literature on children's and teenagers' outdoor play, include significant play spaces for children and teenagers (analysis of distributional justice). In particular, I use the literature on young people 's play in parks, including how young people 's play preferences differ by ethnicity, to develop a Park Quality Index for Youth (PQIY) also building on a park typology I previously developed (Rigolon & Flohr, 2014) My GIS spatial analys is focuses on the relationships between socio economic levels, ethnic variations and percentage of population under 18 (independent variables) and access to parks defined as park proximity, park acrea ge and park quality (dependent variables, considered one at a time). Finally, I integrate the findings of the two strands by looking at the way park planning, land use planning, and housing policies and practices have influenced today's park distribution in Denver, with a specific focus on parks suitable for childre n's and teenagers' play. In particular, I connect specific policies to their spatial outcomes i.e., park location or land uses in different areas, to uncover relationships between procedural an d distributional justice. These activities cover all constructs I mentioned above, including planning policies, planning and design practices, young people 's perceptions and use of parks young people 's socio economic status and ethnicity and the spatial distribution of parks, through a fully mixed research study (Leech & Onwuegbuzie, 2009) thus addressing the main research gap I identifi ed.

PAGE 22

8 In this chapter, I discuss what makes Denver an interesting case study to investigate access to urban parks for young people. Then, I briefly introduce the most significant findings that emerged from my study I discuss why these findings are relevan t for urban planning, and define some key terms I use in this dissertation. Denver as a Compelling Case Study This place specific environmental justice study on access to parks focuses on Denver, Colorado, which is situated at the foothills of the Rocky M ountains, in the Front Range metropolitan area. Denver offers a compelling site to investigate access to parks because of its demographics including ethnic diversity and a rapidly growing population, its highly rated park system, its history of racial and ethnic segregation, some current documented environmental justice issues and the possibility of extending Denver's findings to other cities in the American West Denver's Population Growth and Park System Denver is a relevant case study because it has a propensity to attract residents that are avid parks and recreation users and is a recognized leader in sustainable planning and livability (Godschalk, 2004) Denver is also one of the top ten fastest growing U.S. cities (Carlyle, 2014) with an estimated 17.3 percent population increase between 2000 and 2013 (United States Census Bureau, 2013a) Denver's population density in its urban core is also growing (Jaffe, 2013) and t his will likely increase th e need for high quality parks in Denver's urban core. The analysis conducted by Park Score shows that Denver's park system is close to excellent (ParkScore, 2014) However, the current park acreage (4,233 acres) does not

PAGE 23

9 m eet the National Recreation and Parks Association's standard of 10 acres of park per 1,000 inhabitants (Mertes & Hall, 1996) due to the growing population (649,495; United States Census Bureau, 2013) The origins of the city park system date to the late 1800s and include historic parks like City P ark, Civic Center Park, designed by Frederick Law Olmstead, Jr., and Cheesman Park (Snow, 2009; The Trust for Public Land, 2010) Denver was also an e arly adopter of greenways for recreation and preservation (Cranz & Boland, 2004; Searns, 1995) Denver's park system includes Denver's Moun tain Parks, which are open spaces in the foothills of the Rocky Mountains owned by the city of Denver (Denver Parks and Recreation, n.d.) This further shows the city's focus on land preservation and outdoor recreation. Also, Denver's guidelines for park acces sibility state that each resident should live w ithin six blocks of a park and that there should not be phy sical barriers between a resident's home and a park s uch as highways or railways (Harnik & Simms, 2004) In terms of current park funding, Denver's mayor plans to spend 10 million dollars in 2015 for capital improvements of parks with a commitment to creating equitable opportunities in every area of the city (Proctor, 2014) Denver's Demographics Denver is a relevant case study to in vestigate access to parks as an environmental justice issue also for its demographics. Among Denver's 650,000 inhabita nts in 2013, 53.6 percent identified as N on Hispanic White alone, while the remainder 46.4 percent includes Hispanic or Latinos alone (30. 9%), African Americans alone (10.2%), Asians alone (3.8%), American Indians (2%) and mixed ethnicities (3.1%; United States Census Bureau, 2013) This ethnic composition shows that Denver is an ethnically diverse city with almost an exact split between N on Hispanic Whites and other ethnic group s. Indeed,

PAGE 24

10 the percentage of ethnic minority people (including Hispanics, African Americans, Asians, Native Americans and Pacific Islanders, and people of mixed ethnicity) is higher than the Colorado average (United States Census Bureau, 2013a) and the national average (United States Census Bureau, 2010) Denver also has a long history of African American and Hispanic residents, dating back to its early urban history (de Baca, 1998; Stephens, Larson, & Black American West Museum, 2008) Ethnic diversity is even higher i n the K 12 school population: O nly 22.5 percent of Denver Public Schools' studen ts are N on Hispanic Whites, and Hispanic young people are 57.5 percent of the s tudent population (Denver Public Schools, 2013) Neighborhood parks are particularly important for Denver's ethnic minority young people Eighty perce nt of Denver Public School's students have never b een to the Rocky Mountains (Cheek, 2015) yet the mountains are less than 20 miles away from the city. In terms of socio economics, the median household income, $49,091 is lower than the Colorado average reflecting its higher percentage of ethnic minorities (United States Census Bureau, 2013a) with app roximately 15% of households li ving with less than $15,000 per year and about 11% of households making more than $100,000 per year (United States Census Bureau, 2012) A round 72% of Denver Public School's K 12 students qualify for free or reduced lunch (Denver Public Schools, 2013) These variations in ethnicity and household income, combined with very high percentages of ethnic minori t y and low income children and adolescents in the K 12 student population, make Denver an interesting case study for an environmental justice investigation.

PAGE 25

11 Denver's History of Ethnic Residential Segregation Denver is an appropriate city to study access to parks across income and ethnic groups because of its history of residential segregation. From the early twentieth c entury until nowadays Denver experienced significant residential segregation between Non Hispanic Whites, African Americans, and Hispanic s due to several land use policies, housing practices and racial intimidation (Abbott, 1978; N. Hernandez, 2006; R. L. Simmons & Simmons, 1995) This phenomenon was not uncommon in American cities during the same years (Boone et al., 2009; T. H. Simmons & Simmons, 2010; Wyly & Holloway, 2002) In Denver, starting in the 1920s African Americans were de facto segregated in a few neighborhoods located northeast of d owntown due to r acially restrictive covenants, intim idation by the Ku Klux Klan, redlining by the Homeowner Loan Corporation and real estate practices (N. Hernandez, 2006; R. L. Simmons & Simmons, 1995; T. H. Simmons & Simmons, 2010) Starting in the 1930s, Hispanics concentrated in the northern and western parts of the city due to low land value and t o subdivisions c ontain ing small lots ; also they did not experience the same openly discriminatory housing practices that African Americans faced (Abbott, 1978; Dur‡n, 2011; Langegger, 2012) Ethnic and racial groups in Denver remained significantly segregated until the current times, as my study will show Residential se gregation matters for access to parks, as previously mentioned. If underprivileged groups are concentrated in certain areas of a city policy makers and non profit organization s can intentionally target park investment or disinvestments in such areas (Boone et al., 2009; Downey, 2007) Also, when racial and ethnic minorities were

PAGE 26

12 constrained in certain areas of the city, these areas were be come overcrowded due exclusion fro m other areas (R. L. Simmons & Simmons, 1995) thus making it harder for cities or community groups to acquire land to establish new parks. Denver's Environmental Justice Issues A small but growing body of research sh ows th at in Denver access to environmental goods an d exposure to environmental hazards are unequally distributed across income and ethnic groups. In particular, research has established that in metro Denver ethnic minorit y and low income communities live closer to toxic hazards, including static sources of atmospheric pollution (Harner, Warner, Pierce, & H uber, 2002; Shaikh & Loomis, 1999) and that Denve r's wealthier neighborhoods ha ve significantly more vegetation than low income areas (Mennis, 2006) T wo recent studies documented other spatial inequalities in Denver. A pilot study I conducted with a colleague on a small sample of Denver neighborhoods showed that low income communities of color have significantly less access to parks and to parks with play amenities than prevalently mid and high income Non Hispanic W hite neighborhoods (Rigolon & Flohr, 2014) Also, a paper by NÂŽmeth and Ross (2014) showed that medical marijuana retailers are mostly found in areas with high concentrati ons of low income and ethnic minority residents and that this distributi on is mostly due to city regulations. Also, the Piton Foundation a Denver philanthropic organization, identified areas of the city with high concentrations of at risk children, including risks of being in poverty, attending low performing schools, having parents with low educational levels, lacking access to healthy food, and living in unsafe areas (The Piton Foundation, 2012) These areas are concentrated in the northeastern parts of the city and have been named the

PAGE 27

13 "Children's Corridor" (The Piton Foundation, 2012) This concentration of at risk children in certain areas, as highlighted by the a dvocacy of philanthropic organizations, makes Denver an even more interesting case study for children's access to urban parks Generalizability Denver is in some ways representative of other cities in the American West, including cities like Portland, OR, Seattle, WA, Salt Lake City, UT, Albuquerque, NM, San Francisco, CA, and Oakland, CA, but partially differs from other cities in the East Coast for which access to parks has been extensively studied, including Baltimore, MD and New York, NY. The aforement ioned cities located in the West are similar to Denver for their comparable population size althou gh of varying ethnic composition (United States Census Bureau, 2010) while most cities I listed above are comparable to Denver for their historical patterns of development The latter include a dense commercial downtown developed early in the cities' history; de facto residential segre gation among income and ethnic groups; public improvement s undertook within the City Beautiful framework comprising the construction of parks and parkways; interstates cutting through central parts of the city often displacing low income communities of c olor ; patterns of metropolitan growth in the post World War II period ; and recent urban renaissance trends connected to the arrival of the creative class All these cities are also experiencing steady and in some cases fast population growth in the last de cade (United States Census Bureau, 2013a) ; thus, they will face similar challenges to Denver in te rms of park provision. However, cities located in the East Co a st such as Baltimore and New York are dissimilar from Denver for their earlier period of historical development which led to

PAGE 28

14 establishing the first parks within different political, social, lan d use and technological contexts. Also, East Coast cities have historically been segregated along the white black line (Boone et al., 2009; Romero, 2004) while Denver has been a tri ethnic city from its relatively early history, including Non Hispanic Whites, African Americans, and Hispanics, (Abbott, 1978; Romero, 2004) This adds complexity to the study of access to parks in Denve r. Furthermore, Denver is a compelling case study for the size of its population. Approximately half of the people residing in the world's urban areas live in cities that are smaller than 500,000 inhabitants (United Nations, 2014) which are comparable to Denver in terms of population (United States Census Bureau, 2013a) Finally, the methodology I use in this dissertation is generalizable beyond the aforementioned cities To summarize, the dual nature of Denver having progressive urban planning including a good park syste m though one that has not kept up with growth, and at the same time presenting documented environmental justice and residential segregation issues combined with its demographic profile and the possibility to generalize findings to other cities, makes it a compelling case study to investigate the influence of park policies and practices on children's and teenagers' access to urban parks as an environmental justice issue. A Complex Picture of Environmental Injustice My analysis will show a complex picture of environmental injustice, including cases of distributional and procedural inequities. The analysis of park planning processes (see Chapter 4) will show that, in different historical periods of Denver's history, more cases of discriminatory, neglectful, and color blind processes have occurred than

PAGE 29

15 initiatives oriented to increase park equity, which are actions aimed to increase park provision for low income people, ethnic minorities, and young people The policies and practices that most significantly fos tered or limited park establishment have included park fundi ng systems, City Beautiful visions in the early decades of the twentieth century, a series of municipal policies on parkland dedication or cash fees for infill and greenfield developments and a s tate policy restricting Denver's annexation capacity the Poundstone Amendment to the Colorado Constitution In particular, my research will show that the Poundstone Amendment, a discriminatory policy implemented to prevent ethnic integration in schools, h as negatively affected Denver's capacity to establish large parks beyond its affluent neighborhoods, thus indirectly affecting park equity. Also, my analysis will highlight that young people have typically been excluded from park design consideration s, as economic reasons and the aesthetic values of adults have often prevailed Findings of the geospatial analysis (see Chapter 5) expressing the spatial outcomes of the policies and practices I studied, will show more cases of distributional inequity than dis tributional equity. Among the inequities, I found that African Americans are the most disadvantaged ethnic group, particularly in relation to park acreage and park quality. The cases of partial equity include Hispanics and low income groups having good acc ess in terms of park proximity ( distance to the closest park), and to parks with "good enough" levels of quality However, high income groups and Non Hispanic Whites have access to more acres of parks, to more parkways, to more parks with excellent quality and to more safe parks than other groups. In particular, low income ethnic minority young people have very low access to parks with low violent crime

PAGE 30

16 density, or safe parks, which significantly limit s their park visitation. This is the most concerning in equity issue in Denver's park spatial distribution. The integration between qualitative and geospatial quantitative datasets showed some of the most interesting findings (see Chapter 6). Discriminatory and neglectful processes can often explain inequities in par k spatial distribution. Yet, i nequitable processes were not always directed at creating discrimination in recreation but park inequity has been a consequence of discrimination in housing and education. Fiscally conservative policies have also had sig nificant ly negative impact s on park equity. On the other hand, partially equitable distributions of parks benefitting low income groups and Hispanics have been rarely the results of equity oriented processes. Rather, class and color blind initiatives, comb ined with residential relocation waves have had a stronger role in partially equitable park spatial distributions. From the interviews with park and land use planners, and from the analysis of city documents, I noticed that initiatives aimed to create equa l park provision across geographical areas of the city have been significantly more common than efforts aimed to increase park provision for low income communities of color. As discussed, discriminatory outcomes, evidenced by inequities in park spatial di stribution, are sufficient to make claims about discrimination in recreation in Denver based on the recent ruling of the U.S. Supreme Court regarding the 1968 Fair Housing Act (U.S. Supreme Court, 2015) However, the integration between the analysis of planning processes and their spatial outcomes gives a more comprehensive picture of how ethnic discrimination in recreation has been perpetuated in Denver

PAGE 31

17 Relevance to Urban Planning The results of my dissertation can provide insights to help planning practitioners and scholars, community organizations, and underprivileged communities address the environmental justice issues of access to parks that have been created by past policies and practices. Its main merits are two fold First, this dissertation takes a critical look at planners' participation in children's and teenagers' experiences of envir onmental injustice in recreation and provides communities with tools for engaging the planning process. Second the dissertat ion proposes to bring a focus up on children's and teenagers' experiences into a field that has overlook ed them. For example, the Am erican Planning Association has divisions focusing upon gays, lesbians, women, Latinos, and African Americans but not children and/or teenagers (American Planning Association, n.d. b) More specifically, t he results of this study can inform different planning disciplines, both in terms of scholarship and practice In particular, the analysis of park planning processes highlight s the co mplex procedures that led to park locations and funding suggesting more equitable practices centered on environmental justice. With this in mind, Denver Parks and Recreation can use t he results of this dissertation to develop participat ory suitability ana lyses for new park locations or for improving existing parks without significant play amenities and natural elements In Denver and othe r cities, parks and recreation d epartments can use the results of t he literature review of young people 's outdoor play t o update park guidelines and standards in terms of types of play amenities presence of natural elements, and park size Related to this, the Park Quality Index for Youth (PQIY) including several indicators of quality, can be used to create park classific ations that take into account young people 's preferences and needs.

PAGE 32

18 The results of the Park Planning Process analysis also illuminate the intersections between park, land use, and housing policies and practices, as the current distribution of parks across income and ethnic groups are often due to residential segregation (Boone et al., 2009) These intersections will highlight ways for park planners, land use planners and housing authorities to collaborate to solve these interrelated issues (see Chapter 7, Implications). For example, investments to expand or improve affordable housing and parks should be coordinated through interrelated suitability analyses. In terms of zoning, the results of this study suggest the possibility of allow ing higher residential densities around parks, as parks can be seen as public resources that mo re residents can share. The analysis of park planning processes also highlight s economic reasons behind park location which raise park equity issues This research can benefit low income communities of color and the community organizations that advocate f or their rights. As Bocarro and Stodolska (2013) argued that leisure research related to social justice should generate findings that can be used to bring about change I further develop ed a framework for action I conceptualized this framework earlier (Rigolon & Flohr, 2014) to help low income and ethni c minority communities condu ct park advocacy efforts to increase park provision in their neighborhood Th e framework include s a set of recommendations that can be applied in practice ( see Chapter 7, Implications ) To refine my frame work, I drew from the results of my previous study (Rigolon & Flohr, 2014) on the implications of this dissertation, and review ed further relevant literature (Anguelovski, 2013; Arredondo et al., 2013; GarcÂ’a, 2013; Hou, Johnson, & Lawson, 2009; Hoyt, 2013; Schilling & Logan, 2008; Tidball & Krasny, 2007; Westphal, 2003) Also, I will disseminate my findings among a

PAGE 33

19 series of Denver's neighborhood organizations to raise their awareness about how to increase access to high quality parks in their neighborhoods This dissertation advance s the equity mapping and environmental justice literatures in several ways. I employ a finer method to measure access to play in parks than most previous research by using a combination of census block and parcel data. Then, the focus on children and teenagers and the use of the new Park Quali ty Index for Youth (PQIY) based on the literature on young people's outdoor play advances previous equity m apping research because it introduce s children's and teenagers' perspective on parks in a quantitative spatial analysis Finally, the analysis of par k planning processes builds on Boone et al.'s (2009) analysis and enrich es it by using a combination of interviews, reviews of books and articles, and analyses of primary documents. Finally, this study will advance the use of mixed methods research in planning studies by combining qualitative and quantitative geo spatial approaches and data in a fully integrated mixed methods research design. My mixed research approach also will allow me to study the relationships between constructs that were never studied together in a systematic wa y such as park distribution, par ks perceptions and use, young people 's socio economic status (SES) and ethnicity, and planning policies and practices. This dissertation will include a critical reflection on the methods used in this study to share the meth odological lessons I learn ed from this study with other planning scholars ( see Chapter 7, Implications ) Definition of Key Terms To frame this dissertation a few key terms and phrases used to describe spaces and individuals ne ed to be defined. First, I us e the phrase "urban green space" to describe

PAGE 34

20 all accessible green spaces in a city, including land that is owned and managed by the city such as public parks, greenways, gardens, and other public open spaces; land that is owned by other public institutions such as accessible schoolyards, community center fields, and playgrounds in housing projects; and private spaces that are open to the public through right of way agreements such as those owned by homeowner associations. Then, "urban parks" or "parks" incl ude all public green spaces designated as parks in a city's limits. Urban parks are publicly funded planned, designed and managed by a city's Parks and Recreation Department. I us e "access to parks to indicate the possibility to enjoy any type of park Also, I include in this concept play opportunities in playgrounds, sport fields, and other spaces found in urban parks can all provide contact with natural elements, such as trees, grass, and rocks. Indeed, Kaplan (1984) claimed that nature can be found in cities in the form of parks, trees, plants, and water. Natural places and elements that can be experienced in cities on a daily basis were defined as "nature at the doorstep" (Kaplan, 1984, p. 189) The definition of nature adopted in this dissertation including the fact that nature can be found in cities, is widely accepted in the planning and environmental psychology l iteratures (Baur, Tynon, & G—mez, 2013; Beatley, 2011; Herzog, 1989; Sullivan, Kuo, & Depooter, 2004) Thus, in this dissertation every reference to parks is a reference to urban nature. Although there are no generally accepted def in itions of "children ," "adolescents ," and "youth" (Ansell, 2005) I use the term "child" to describe an individual aged zero to 11, as 11 corresponds to the last year of elementary or prim ary education (The World Bank, 2013) and on average to puberty (P. A. Lee, Guo, & Kulin, 2001) I use t he term

PAGE 35

21 "adolescent or "teenager" to describe an indi vidual aged 11 to 17, as 18 represents the age at which individuals become legally adults in most Western countries (United Nations Children's Fund, 2011) and finish their secondary education (UNESCO Institute for Stat istics, 2005) Th en, I use the phrase "young people" to group children and adolescents together The distinctio n between children and adolescents is critical to play, as different age groups tend to have different play preferences (McCormack et al., 2010; Ries et al., 2008) and different mobility licens es (Soori & Bhopal, 2002) I discuss how children's and teenagers' play preferences differ in my review of young people's outdoor play and park visitation (see Chapter 5, Park Spatial Distribution Findings ) I use t he term "parent" to refer to the biological or adoptive parent of the young people. I use t he word "caregiver" to refer to any guardian, related or unrelated to the young person including grandparents, older siblings, aunts and uncles and other guar dians Thus "caregiver" is more comprehensive than "parent" and describes supervising adults of any kind. In terms of ethnicity, I chose to focus mostly on three ethnic groups: Non Hispanic Whites, Hispanics, and African Americans. Denver has often been c onsidered a tri ethnic city (Abbott, 1978; Romero, 2004) and Non Hispanic Whites, Hispanics, and African Americans still constitute the three largest ethnic groups in the city (United States Census Bureau, 2013a) Also, I use the term "Hispanic" as a general term to describe people who identify as Latino/Latina, Chicano/Chicana, M exican descent, or Hispanic. Although historically these terms have had different meanings and connotations (Langegger, 2013) I use the general term "Hispanic" to simplify the discussion. Similarly, t he U.S. Census re groups people who identify a s Latino/Latina,

PAGE 36

22 Chicano/Chicana, Mexican descent, or Hispanic in a general category named "Hispa nics" (United States Census Bureau, 2013a) Also, I use capitalized terms for ethnic groups define d by the U.S. Census (e.g. African Americans), while I use non capitalized terms to refer to groups of ethnicities (e.g. people of color and ethnic minorities). In the following chapters of this dissertation, I will define my theoretical framework, review the literature on young people's access to parks based on the framework, highlight gaps in the literature, and lay out my research goals and questions (Chapter 2). Then, I will describe the research design and the specific methods I used to investigate ac cess to parks for young people in Denver (Chapter 3). I will present the findings of the analysis of park planning processes in Chapter 4 and the findings of the spatial distribution of parks in Chapter 5. In Chapte r 6, I will integrate the findings on pla nning processes and park distribution. Finally, in Chapter 7, I will discuss the implications of my findings for park and land use planning, for environmental justice, and for the use of mixed methods research in planning.

PAGE 37

23 CHAPTER II THEORETICAL FRAMEWORK LITERATURE REVIEW, AND RESEARCH QUESTIONS Theoretical Framework T he problem laid out in Chapter 1 identifies the constructs this dissertation will investigate namely planning policies, planning and design practices, young people's perceptions and use o f parks, young people's socio economic and ethnic backgrounds, and the spatial distribution of parks. I first define these constructs in Table 1 and then discuss their potential relationships in the next section. Identification of Constructs The constructs presented in Table 1 are similar to the ones that Byrne and Wolch (2009 ) introduced in their conceptual model named "space, race, and park use" (p. 751). In particular, their model includes the "historical and cultural context of park provision," "park space," "perceptions," "potential users," and "park use choices," compris ing "use" and "non use" (Byrne & Wolch, 2009, p. 751) This fram ework does not focus specifically on children or tee nagers, but it can be useful to define the constructs included in Table 1 better and to explore some of the potential connections among the constructs that influence park perceptions and use. For example the historical context of park provision includes aspects like "spatialized ethno racial discrimination," the "ideology of land use," the "history of property development," "park politics," the "ideology of park provision," and "racial politics of park d evelopment" (Byrne & Wolch, 2009, p. 751) In other words, this construct comprises a series of cultural, policy, historical, and racial

PAGE 38

24 aspects related to lan d use and park planning. I further discuss Byrne and Wolch's (2009) model when introducing the potential connections among the constructs identified in Table 1. Table 1 Constructs Included in the Theoretical Framework Construct Description Planning Policies Public regulations affecting the built environment at different scales as the result of decision making processes involving stakeholders with different visions. I focus on policies addressing land use, zoning, park location a nd funding, affordable housing location and funding. Planning and Design Practices Actions undertaken within policies frameworks to transform space or change its use, sometimes by challenging the regulations imposed by policies. I focus on how plans for c ommunities, parks, and housing were implemented. Young People's Perceptions and Use of Parks The way children and teenagers perceive and use their neighborhood parks, often related to their sense of safety and to their caregivers' regulatory strategies. Young People's Socio Economics and Ethnicity Median household income combined with assessed parcel value of a young person's home (as an asset of the family) and ethnic self identification. I chose to focus on ethnicity, instead of race, given Denver's la rge Latino population. Spatial Distribution of Parks How far different types of parks are from young people's homes. Park spatial distribution is a spatial quantitative measure and is one of the variables influencing young people's park visitation, as dis tance to parks matters for young people's use of parks. Planning policies According to McConnell (2010) a "policy" has g enerally been defined as a decision or action that public institutions undertake to distribute value. In planning, these decisions and actions undertaken by governments are intended to improve

PAGE 39

25 the places where people live, as the American Planning Associat ion defined planning as discipline intended to create "more convenient, equitable, healthful, efficient, and attractive places for present and future generations" (American Planning Association, n.d. a) Policy making is a complex activity that generally affects the interests of a variety of public and private actors (Innes & Booher, 2010) This problem solving task usually generates conflicts between stakeholders, which can be solved through lawsuits or, more efficiently, through collaborative planning (Innes & Booher, 2010) The conflicts occurring between stakeholders suggest the power dynamics that are inherent in policy making. In the past, most policies w ere created by expert public administrators based on data they collected, and had to go through political ratification (Innes & Booher, 201 0) In recent times, more collaborative policy making processes, involving different stakeholders, have emerged (Innes & Booher, 2010) I discuss the i mpacts of polic ies on equitable park distribution in Chapter 6 In this study, I focus on the planning policies that influenced how urban parks are distributed in relation to ethnic and income groups (Boone et al., 2009; Byrne & Wolch, 2009; Wolch et al., 2005) including: city level policies about land use, zoning, park location and funding, affordable housing location and funding, and park design and maintenance standards; and state lev el policies about park funding and annexations Byrne and Wolch (2009) considered the "historical and cultural context of park provision" as a primary factor in determining park spatial distribution and the specific features of each park (p. 715). In particular, they highlighted the history of ethnic and racial discriminatio n that pervaded park planning in several U.S. cities in the twentieth c entury (Byrne & Wolch, 2009) In my framework, Byrne and Wolch's (2009) historical

PAGE 40

26 and cultural context of park prov ision can be conceptualized as planning policies, and as planning and design practices. Land use policies can reflect political and cultural ideologies and can be used to spatially discriminate against ethnic and racial minorities, even in terms of access to parks (Byrne & Wolch, 2009) In the U nited S tates zoning has been the main planning instrument to determine approp riate land uses (Maantay, 2001) Racial zoning ordinances started developing after the Civil War, with African Americans migrating from the South to th e North and affluent whites trying to prevent African Americans from living close to them (Bono, 2007) In Baltimore, Maryland, a zoning ordinance passed in 1910 created segregated residential areas for Non Hispanic Whites and African Americans, and other racial zoning ordinances were approved in the following years in several American cities (Bono, 2007) Some of these ordinances targeted specific blocks, others segregated entire districts of a city (Bono, 2007) Clearl y, these zoning ordinances aiming to separate African Americans from Non Hispanic Whites were an example of de jure segregation, as segregation was created by law (Bono, 2007) Zoning ordinances have also been used to segregate people by socio economic status (Hartnett, 1993; Liberty, 2002) E xclusi onary zoning ordinances implemented in several U.S. cities have dictated minimum lot sizes, minimum house sizes and set back rules for residential units located in certain neighborhoods, as a means to exclude low socio economic groups from these neighborho ods (Hartnett, 1993; Liberty, 2002; Maantay, 2001) Limitations or bans on multi family ho u sing were another means throu gh which exclusionary zoning limited low socioeconomic groups' opportunity to live in certain neighborhoods (Hartnett, 1993; Liberty, 2002)

PAGE 41

27 Zoning codes have also been used to locate locally unwanted land uses like polluting industries in proximity to residential areas inhabited by low income communities of color (Harvey, 2009; Maantay, 2001, 2002) In particular, zoning changes implemented in New York City between the 1960s and 1980s have incre ased underprivileged groups' exposure to polluting industries (Maantay, 2001) Therefore, zoning ordinances in the U.S. have contributed to creating en vironmental injustice and public health threats for low income communities of color (Harvey, 2009; Maantay, 2001, 2002) Planning and design practices If policies create regulations and dictate actions, design and planning practices are actions undertaken w ithin policy frameworks by public and private agencies to transform space or change its use at different scales of the physical environment, sometimes by challenging the regulations imposed by policies. For example, a park designer can work around the limi tations imposed by park standards to design a natural playground for children's informal play. Like policies, planning practice is influenced by values and ethical dilemmas (H. Campbell & Marshall, 1998) and is intended to achieve "good" or "better" outcomes (Alexander, 2009) Planning practices need to conceive the spatial realms they intend to modify as places rather than simple physical settings, thus taking into account the socially constructed meani ngs attached to physical settings (Graham & Healey, 1999) In this study, I focus on the work of planning and design practitioners, and of Denver's r esidents, to establish park s land uses and housing policies and practices, and to design and maintain parks. Byrne and Wolch's (2009) "historical and cultural context of park provision" includes several planning and housing practices that contributed to residential segregation

PAGE 42

28 based on race and ethnicity, such as restrict ive racial covenants and redlining by the Home Owners Loan Corporation. Restrictive racial covenants were conditions attached to properties that only allowed the sale or rent of these properties to certain racial and ethnic groups (Boone et al., 2009; Gotham 2000; Jones Correa, 2000; Nelson, Sanchez, & Dawkins, 2004) Starting from the 1920s, neighborhood improvement associations and realtors encouraged Non Hispanic White affluent property owners to establish covenants that would prohibit selling housing un its to non whites and ethnic minorities ( Boone et al., 2009; Gotham, 2000; Jones Correa, 2000; Nelson et al., 2004) Restrictive covenants were aimed to maintain high property values for owners by keeping undesired populations out of certain neighborhoods (Boone et al., 2009) According to Bono (2007) these practices, which restricted housing options for ethnic and racial minorities, can be considered forms of de facto discrimination. However, in this case there is a fine line between de jure and de facto segregation because these covenants, although not created by public policy were legally enforced. The practice of redlining deals with lending bias that derive from "residential security maps" created by the Home Owners L oan Corporation (Greer, 2013; J. Hernandez, 2009; Hillier, 2003a, 2003b; Squires, 2003) Starting in the 1930s, inner city areas with deteriorating housing stock, which were generally inhabited by low income African Ameri cans, were defined as too risky for lending and bounded by red lines in the Home Owner Loan Corporation maps (Greer, 2013; J. Hernandez, 2009; Hillier, 2003a, 2003b; R. L. Simmons & Simmons, 1995; T. H. Simmons & Simmons, 201 0; Squires, 2003) African Americans living in these areas could not receive mortgages to buy or improve their houses, which led to the quality of housing further deteriorating and

PAGE 43

29 discriminatory lending practices persist even today (Hilli er, 2003a, 2003b; R. L. Simmons & Simmons, 1995; Squires, 2003; Wyly & Holloway, 2002) On the other hand, African Americans who attempted to leave the redlined areas were often the object of intimidation and violence, including actions by the Ku Klux Kla n and of real estate practices like racial steering (N. Hernandez, 2006; Leonard & Noel, 1990; Mohl, 1995; Morrison, 2006; R. L. Simmons & Simmons, 1995) Design prac tices and the culture in which they are based influence the aesthetic qualities of parks and the activities that parks support. For example, Byrne and Wolch (2009) highlighted that traditionally most parks in American cities were designed based on English and northern European canons of beauty, which reflect picturesque ideal s of pastoral landscapes, and which are not appealing for all ethnic or racial groups. Young people's perceptions and use of parks The ways young people perceive and use their neighborhood parks and green spaces are important aspects to consider when stud ying access to urban parks as an environmental justice issue (Byrne & Wolch, 2009) As discussed earlier, park design and quality influences children's and teenagers' use of parks (Byrne & Wolch, 2009; Potwarka et al., 2008; Veitch, Salmon, et al., 2014) Some parks might not serve the specific play and recreation needs of cert ain age or ethnic groups (Byrne & Wolch, 2009) Use and perceptions are often related to children's and teenagers' sen se of safety, as neighborhood violence or the presence of gangs in parks often limit young people's access to parks in low income neighborhoods (G—mez et al., 2004; Platt, 2012) Caregivers can hold the same concerns for safety and therefore restrict their children's access to outdoor play in neighborhood parks (Valentine, 1997; Veitch, B agley, Ball, & Salmon, 2006)

PAGE 44

30 Even when young people and their caregivers hold positive perceptions of parks, other factors limit children's and teenagers' park visitation. Besides barriers re lated to distance, which I discuss below, children and teenage rs might not use parks because of time constraints, physical abilities, and other competing recreational activities (By rne & Wolch, 2009) Thus, understanding young people's perceptions and use of parks is important as is identifying young people's preferences for different types of parks and young people's and caregivers' sense of safety in their neighborhood. To operati onalize young people's perceptions and uses of parks, I create a Park Quality Index for Youth (PQIY) based on the literature on young people's outdoor play and park visitation. Young people's socio economic and ethnic characteristics This construct use s d emographic data from the U.S. C ensus and from the City and County of Denver to operationalize children's and teenagers' socio economic status and ethnicity. In particular, to operationalize socio economic status I employ a combination of median household i ncome from the U.S. Census, the assessed parcel land value of young people's homes from the City and County of Denver, and other variables on home ownership from the U.S. Census (see Chapter 3, Research Methods). To operationalize young people's ethnicity, I use the family's ethnic self identification from U.S. Census data. Further details about t hese co nstructs are given in Chapter 3 ( Research Methods ) Also, I focus on ethnicity, instead of race, given Denver's large Hispanic population (around 30%), even larger K 12 student Hispanic population in Denver Public Schools (around 57%), and strong cultural identification that includes different races. Spatial distribution of parks This construct describes where parks are located in different parts of the city and how they serve different populations. This spatial

PAGE 45

31 distribution determines the distance between parks and every child's and teenager's home. In this study, access is a quantitative spatial concept related to the geographic distribution of parks i.e., having the opportunity to enjoy settings that are open to the public and that are within walking distance from young people's homes. Therefore, I study access to play in parks as an opportunity related concept or, in other words, in terms of "potential [pa rk] users" (Byrne & Wolch, 2009, p. 751) However, I recognize t hat distance and parks spatial distribution is not the only factor influencing young people's actual access to parks. As the following sections and the literature review show, other factors affecting actual park visitation for young people include physical barriers like arterial roads and railways, park design, social factors, perceptions of safety, and personal factors related to age, gender, and culture. The classic work of two psychologists, James Gibson and Roger Barker, and more recent developments in environmental psychology and critical geography, help clarify the way I define young people's access to play in parks. James Gibson (19 79) formulated the theory of affordances, which he defined as all the possibilities of action inherent in the objective physical environment, in relation to an actor and dependent on the actors' capabilities, including physical skills, intentions, and kno wledge. For example, a tree branch five feet high does not afford the act of climbing for an infant, but can afford the same act for an adolescent. While the environment potentially includes an infinite number of affordances (KyttŠ, 2004) actualized affordances are a subgroup of all potential affordances that includes what an individual notices, uses or modifies (Heft, 1989) Therefore, affordances can be conceptualized as ordered features (Greeno, 1994) including "potential, perceived, utilized and shaped affordances" (KyttŠ, 2002, p. 109)

PAGE 46

32 The degree to which affordances are noticed, used or modified depen ds on individuals' capabilities, but also on collective and cultural norms and expectations (Reed, 1993) Excessive distance from parks with significant play amenities is an example of how the physical environment may or may not afford playing in nature, depend ing on individuals' skills and parental regulatory practices. Distance from parks matters because children spend most of their time around home (Loeba ch & Gilliland, 2014) and because children generally use active forms of transportation to access parks (Veitch, Carver, et al., 2014) Indeed, a study conducted in two Swedish small towns shows that children often use playgrounds that are located close to their home (Jansson & Persson, 2010) Therefore, the theory of affordances helped me conceptualize access from a spatial quantitative point of view. After defining access to play as the opportunity to reach a park, it is important to understan d which types of settings within parks can support play for young people. In terms of physical features, the theory of affordances can also help explain in what ways physical settings influence behaviors. For example, a very steep hill might not afford run ning uphill (potential or perceived affordance) but affords rolling downhill (actualized affordance). However, play behaviors can also be explained by social norms and expectations. In this regard, Roger Barker (1968) proposed the theory of behavior settings, which holds that in certain physical settings, behaviors are influenced by social ex pectations and rules, which vary greatly by culture (Barth, 1998) For example, classrooms host learning activities while sport fields host athletic game s (Ding, Sallis, Kerr, Lee, & Rosenberg, 2011) Thus, accord ing to Barker's (1968) behavior settings, behaviors match the environment through soci al rules. However, children's perception

PAGE 47

33 and use of space differs from adults', as they tend to play in a variety of different settings, often in places that were not conceived as spaces for children's play (Matthews & Limb, 1999) For example, some studies showed that standard playgrounds are most often used by young children accompanied by adults, suggesting that adults rather than children choose playgrou nds as play settings (Matthews & Limb, 1999) In particular, adults' attempt to limit children's play to playgrounds is a way to segregate children in cert ain parts of the city for adults' convenience (Hart, 2002) Ther efore, behavior settings as places for children's and teenagers' play need to be investigated from young people's perspective s and in relation to adults' regulatory systems. For the purpose of this research which is focused on access to urban nature, play amenities located in parks can be considered behavior settings where different types of activities can occur (Cosco Moore, & Islam, 2010) Indeed, different types of play amenities within a park can be associated with different levels of physical activity (Cosco et al., 2010) and types of play (Dyment & O'Connell, 2013) I introduce and discuss different types of play amenities in Chapter 3, Research Methods. Relationships between Constructs After having introduced the se five constructs, I discuss the potential connections among the constructs to create an encompassing framework for my dissertation, as shown in Fi gure 1. The relationships among the mentioned constructs are hypothesized based on urban planning, environmental justice, and environmental psychology literatures, which I analyze in detail in the literature review. Byrne and Wolch's (2009) "space, race, and park use" conceptual model also provided useful insights to build the framework described (p. 751).

PAGE 48

34 The Planning, Ethnicity, Park Distribution and Quality framework (PEPDQ, Figure 1) is rendered as a globe because it includes complex and mutual relationships among the constructs I study; thus, a linear one way model woul d not have been appropriate. In this study I investigate through empirical data only some of the connections described in the PEPDQ framework, including the relationships between policies and practices and the spatial distribution of parks, mediated by you ng people's socio economic status and ethnicity (Boone et al., 2009; Byrne & Wolch, 2009) ; the possible associations between the spatial distribution of parks and young people's socio economic status and ethnicity (Byrne & Wolch, 2009; Wolch, Byrne, & Newell, 2014) ; and the connection bet ween young people's socio economic status and ethnicity and their perceptions and use of parks due to neighborhood safety (G—mez et al., 2004; Platt, 2012) The PEPDQ framework contains other connections that deserve further exploration in future research, such as the way the spatial distribution of parks, and young people's use and perceptions of parks can inform future planning policies and d esign (Talen, 2003) ; or that c an be investigated in future studies, for exam ple the connections between the spatial distribution of parks and young people's park perceptions and use (KyttŠ, 2002, 2004) Including young people's park preferences through a park quality index is a key part of this study because it is a means to partially address the connection between park availability and park visitation. As Floyd, Taylor, and Whitt Glover (2009) suggested, it is fundamental to include park quality aspects that have been correlated with park visitation and physical activity in parks.

PAGE 49

35 Figure 1 The Plan ning, Ethnicity, Park Distribution and Quality framework. The symbol "#" describes connections I investigate in this study. The symbol "[]" describes planning and design implications. The symbol "<>" depicts future directions for research. Transformative Paradigm and Environmental Justice To analyze the connections among planning policies, planning and design practices, young people's socio economic status and ethnicity, and the spatial distribution of parks, I employ a combination of the transformative ap proach to research (Mertens, 2003) and an environmental justice lens (Floyd & Johnson, 2002; Heiman, 1996; Schlosberg, 2004) The transformative paradigm serves as a general framework for social change, such as creating better conditions for playing in urban nature for low income and ethnic minority young people, as well as a paradigmatic foundation for mixed methods research (Mertens, 2003, 2007) Environmental justice serves as a theoretical lens for the

PAGE 50

36 specific focus of this research, which deals wit h distributional and procedural injustice (Floyd & Johnson, 2002; Heiman, 1996; Schlosberg, 2004) Transformative paradi gm The transformative paradigm addresses issues of power by focusing on the injustices experienced by underprivileged populations like women, ethnic and racial minorities, disabled people, the LBGTQ community and whoever experiences oppression, and intend s research as a means to undertake or inform action (Mertens, Bledsoe, Sullivan, & Wilson, 2010; Mertens, 2003, 2007) The transformative paradigm is rooted in participatory inquiry, which emerged as a research paradigm in the 1960s and 1970s (Chambers, 1994) as they both conceive reality through the perspectives of marginalized individuals and communities (Mer tens, 2009) Participatory inquiry is a research approach in which investigators employ participatory actions to learn about communities and to help them bring about social change (Brydon Miller, 1997) Participatory and activist researchers believe that research should be conducted with people and not on or about people (Heron & Reason, 1997) The transformative paradigm developed partially because scholars and marginalized communities expressed frustrations about the prevalent research approac hes, which failed to represent and empower such communities (Mertens et al., 2010) Als o, the transformative approach aims to increase awareness of social injustice and to include the opinions of marginalized communities in scholarly research, which can lead to actions oriented to social equity (Mertens et al., 2010) The work of Donna Mertens (Mertens et al., 2010; Mertens, 2003, 2007, 2012) helped conceptualize the ontological, epistemological, axiological, and research practice foundations of the transformative paradigm. Regarding ontology, trans formative

PAGE 51

37 researchers believe that multiple realities exist and that social groups construct them, but also that social, cultural, economic, political, racial, gender, age, and disability stances influence the way these realities are constructed (Mertens, 2007) The resulting realities can be dissimilar because partici pants and investigators have various degrees of privilege and disadvantage (Mertens, 2007) In terms of epistemology, transformative researchers believe that participants and investigators should establish a cooperative connection, including respect for any type of difference and recognition of power relationships (Mertens, 2007) Regarding the practice of research, Mertens (2007) suggested that in transfo rmative research the participants and the investigators should collaborate in defining the research goal and techniques. Also, transformative research techniques should be appropriate to welcome cultural diversity, should minimize power imbalances, and avo id discrimination (Mertens, 2007) In terms of axiology, the tra nsformative paradigm is based on the acknowledgment of power imbalances between groups with different incomes, ethnicity, sexual orientations, and physical and mental abilities, age and on the ethical consequences that stem from these differences, such as injustice, persecution, distortion, and negligence (Mertens, Holmes, & Harris, 2009) Mertens (2007) also conceptualized how transformative mixed methods studies can address societal inequities and be a tool for change. The use of both qualitative an d quantitative methods can be a way to address power issues, as participants have more than one way to express their views (Mertens, 2007) Also, when studying social justice issues, the complementarity of qualitative and quantitative methods can provide a broader range of ideas and solutions about the studied issues (Hesse Biber, 2010; Mertens, 2007)

PAGE 52

38 For example, quantitative methods can be used to identify the groups that experience discrimination or oppression and then researchers can use qualitative methods to better understand the groups' issues (Mertens, 2007) Further, Mertens (2007, 2012) proposed that mixed methods can be used cyclically in transformative studies, starting and ending the research process with participatory qualitative methods, while collec ting and analyzing quantitative data between the two qualitative phases. In this study, which focuses on low income young people of color, the transformative approach is linked to concepts of social justice, which is a vision of society in which each indiv idual is warranted equal social, political, and economic rights and opportunities (Bell, 2007; Gibelman, 1995) and of environmental justice (Heiman, 1996; Schlosberg, 2004) which is described in the next section. Also, since in this dissertation I did not collaborate with the participants to define research goals and methods du e to time constraints, I am only partially using the transformative paradigm. Environmental justice Environmental justice, encompassing environmental racism and environmental classism describes the idea that populations of different ethnicity or race and socio economic status have access to uneven environmental quality (Schlosberg, 2004; Schweitzer & Stephenson, 2007) The concept of environmental justice developed from the acknowledgement that low income and ethnic minor ity groups are more exposed to environmental hazards, like air pollution from traffic, incinerators and industrial plants, solid waste landfills, and water contamination (Boone et al., 2009; Floyd & Johnson, 2002; Heiman, 1996; Schlosberg, 2004) However, environmental justice also addresses inequalities in terms of access to positive infrastructure and amenities, including housing, health, food (Heiman, 1996) recreation (Floyd & Johnson,

PAGE 53

39 2002) and transportation (Bullard, 2003; Zavestoski & Agyeman, 2014) Several U.S. federal agencies have implemented strategies to address environmental injustice (U.S. Department of Housing and Urban Development, 2012; U.S. Department of Transportation, 2012; U.S. Environmental Protection Agency, 2014) showing a growing interest in these issues. However, two recent reviews of sustainability plans in American cities showed that, due to policy or implementation limitations, ecological issues receive more attention than equity issues (Pearsall & Pierce, 2010; Warner, 2002) A more detailed disc ussion of the intersections between sustainability and environmental justice is reported below. Until recently, most environmental justice literature focused on exposure to "bads", while fewer studies have been dedicated to access to "goods", including nat ure (Bo one et al., 2009; Floyd & Johnson, 2002) Also, most basic definitions of environmental justice generally deal with distributional aspects i.e., that goods and bads should be equally distributed among demographic groups, without taking into account the pr ocesses leading to spatial distributions (Boone et al., 2009) However, Schlosbe rg (2004) and Young (1990) argued that the concept of environmental justice goes beyond the distributional aspects o f "goods" and "bads." According to Schlosberg (2004) the concept of "global environmental justice" must include three components (p. 517): parity in the distribution of environmental hazards and public services, the acknowledgment of the different and specific cu ltural backgrounds of each community, and community engagement in decision making that focuses on environmental policy (Schlosberg, 2004) In particular, Young (1990) stressed the importance of procedural justice, describing equitable decision making processes to

PAGE 54

40 determine the distribution of environmental amenities and hazards. The analysis of procedural justice needs to take into account the social, political, and cul tural conditions that contribute to spatial distributions (Schlosberg, 2004; I. M. Young, 1990) This expanded definition of environmental justice connects well with social justice. Indeed, social justice is a vision of society in which each individual is warranted equal social, political, and economic rights and opportunities (Bell, 2007; Gibelman, 1995) Equality vs. equity When discussing access to parks as an environmental justice issue, a few scholars suggest going beyond the concept of equality to focus on equity as low income young people of color need parks more than other demographic groups (Boone et al., 2009; Wolch et al., 2005) When planning for parks in cities, an equality approach would mean distributing parks and resources evenly across income, age, and ethnic racial groups (Boone et al., 2009; Wolch et al., 2005) On the other hand, planning for equity would mean locating more parks and investments in areas with higher concentrations of people who need park s the most, including: low income people, people under the age of 19 or above the age of 65, and people without access to a private vehicle (Boone et al., 2009 ; Talen, 2003; Wolch et al., 2005) In other words, an equity approach to recreation planning would ensure that the people "with the greatest need and fewest means to access recreational areas could do so" (Thoma s, 2010, p. 503) Focusing on equity instead of equality is important because childr en and teenagers from low income and ethnic minority backgrounds generally lack the means for traveling to destination parks due to transportation limitations (Castonguay & Jutras, 2009; Karsten, 2005; Loukaitou Sideris & Sideris, 2010; Ries et al., 2008) ; have limited opportunities to use private recreational opportunities, including backyards (Loukaitou

PAGE 55

41 Sideris & Stieglitz, 2002) ; generally live in denser urban environments (Loukaitou Sideris & Stieglitz, 2002) ; and take part in organized recreational ac tivities less frequently than other groups (Valentine & McKendrick, 1997) Given these limitations, children in low income densely populated areas spend more time in p ublic space, including parks and playgrounds, than children living in more affluent suburban areas (Karsten, 2005; Loukaitou Sideris & Stieglitz, 2002) Thus, Castonguay and Jutras (2009) argued that high quality public space is fundamental for children living in low income neighborhoods. Usin g the equity approach to plan parks by taking into account park dependent populations also makes sense financially (Boone et al., 2009) To prioritize park investment in a period with shrinking budgets for public amenities, planners need to conduct suitability analyses to identify the areas of cities and metropolitan areas whe re park dependent populations live and where park service is currently low (Boone et al., 2009; NÂŽmeth, Faulkner, & Ross, 2012; Talen, 2003) Environmental justice and sustainability As I briefly explained in the introduction, this dissertation is also driven by the necessity of stressing the fundamental role of environmental justice in sustainability discourses and practices. Although equity is one of the three E's of sustainability (S. Campbell, 1996) its importance in sustainability scholarship and po licies has often been overlooked (Agyeman & Evans, 2004, 2003; Boone, 2010) In particular, Agyeman and Evans (2004) introduced the concept of "just sustainability" to argue that the discussion of environmental justice needs to be located within the sustainability framework (p. 155). Also, the idea of just sustainability describes the fertile connections that can be established between environmental justice

PAGE 56

42 and sustainability and acknowledges that sustainability cannot only be abou t environmentalism (Agyeman & Evans, 2004, 2003) The inadequate attention to equity in sustainability discourses and practices has been highlighted by a few studies (Environmental Law Institute, 1999; Pearsall & Pierce, 2010; Warner, 2002) Warner's (2002) review of sustainability policies in large American cities highlighted that very few munic ipalities ( 5 out of 33) considered environmental justice as an integral part of sustainability. Even among the cities that included environmental justice concerns in sustainability projects, there were notable differences in how environmental justice was i mplemented (Warner, 2002) In some cities, for example, environmental justice was only mentioned as a general go al, but no specific actions to achieve environmentally just sustainability policies were implemented (Warner, 200 2) Comparable findings emerge from a report by the Environmental Law Institute (1999) which studied 579 appl ications to the Environmental Protection Agency's 1966 Sustainable Development Challenge Grant Program. Only 5 percent of the submitted applications included equity as an objective (Environmental Law Institute, 1999) A more recent review of sustainability plans in large U.S. cities showed that more and more cities are including environmental justice goals in their sustainability plans, but that in practice justice oriente d actions have less leverage than efforts aimed to solve ecological issues (Pearsall & Pierce, 2010) possibly due to p ower imbalances among groups contributing to these plans. This theoretical and empirical literature suggests the importance of framing equity and environmental justice within the sustainability framework, as well as the need to highlight the key role of e quity as one of the three parts of sustainability (see S. Campbell,

PAGE 57

43 1996) This is particularly relevant because ecologically driven policies, which stem from bipartisan political efforts, often privilege economic development over social equity (Checker, 2011) To summarize, in this dissertation I adopt an environmental goods approach to analyze the social disparities in young people's access to play opportunities in parks, including the reasons behind the observed disparities. To study the p lanning processes and outcomes that influence access to parks for young people in Denver, I employ a theoretical lens that combines the transformative paradigm (Mertens, 2003) with an environmental justice lens (Schlosberg, 2004) Also, like other scholars (Boone et al., 2009; Wolch et al., 2005) I argue that equity is a better construct than equality when studying access to parks across demographic groups. Finally, I frame equity and environmental justice within the sustainability framework (Agyeman & Evans, 2004; Boone, 2010) and argue that equity should have the same importance as environmental quality and economic development in sustainability discourses and practices. Literature Review In this section, I use my theoretical framework to connect several bodies of literature that shed light on children's and teena gers' access to urban nature. In particular, I focus on most of the connections among the constructs included in the theoretical framework (see Figure 1) and then highlight the literature gaps and inc onsistencies, which lead to my research goals and questi ons.

PAGE 58

44 Park Spatial Distribution and Demographics Increasingly, research has studied how access to different types of public facilities and other amenities differ by socio economic and ethnic groups (Lobao & Saenz, 2002; Macintyre, 2007; National Recreation and Park Association, 2011; Wolch et al., 2014) and in particular leisure research has increased its focus on social justice (Floyd, 2014) The body of research focusing on the spatial distribution of public amenities has been named "equity mapping" (Talen, 1998) because studies generally map resources in relation to where different demographic groups live. The results of these environmental justice studies are often influenced by the unit of analysis and by the geographic scope of the investigation (Baden, Noonan, & Turaga, 2007) ; th us, they need to be considered in relation to the methods they use to measure access. In particular, scholars hav e studied how various income, ethnic, and racial groups have different levels of access to parks and playgrounds (Boone et al., 2009; Comber, Brundson, & Green, 2008; Cradock et al., 2005; Dai, 2011; Ellaway, Kirk, Macintyre, & Mutrie, 2007; Erkip, 1997; Estabrooks, Lee, & Gyurcsik, 2003; Gi lliland, Holmes, Irwin, & Tucker, 2006; Jones, Brainard, Bateman, & Lovett, 2009; Kabisch & Haase, 2014; Loukaitou Sideris & Stieglitz, 2002; Madsen, Radel, & Endter Wada, 2014; Maroko, Maantay, Sohler, Grady, & Arno, 2009; Miyake, Maroko, Grady, Maantay, & Arno, 2010; L. V. Moore, Diez Roux, Evenson, McGinn, & Brines, 2008; NÂŽmeth et al., 2012; Nicholls, 2001; Potwarka et al., 2008; Rigolon & Flohr, 2014; Sister, Wolch, & Wilson, 2009; Smoyer Tomic, Hewko, & Hodgson, 2004; Talen & Anselin, 1998; Talen, 201 0; A. F. Timperio, Ball, Salmon, Roberts, & Crawford, 2007; Vaughan et al., 2013; Wen, Zhang, Harris, Holt, & Croft, 2013; Willemse, 2013; Wolch et al., 2005; Wright Wendel,

PAGE 59

45 Downs, & Mihelcic, 2011; Zhou & Kim, 2013) walkable and bikeable neighborhoods (Christie et al., 2011; Crawford et al., 200 8; Cutts, Darby, Boone, & Brewis, 2009; Duncan et al., 2012; Franzini et al., 2010; Talen, 2003) green school grounds (Dyment, 2005; Stewart, Purner, & Guzm‡n, 2013) schools with joint use agreements for community use of school grounds (Kanters, Bocarro, Moore, Floyd, & Carlton, 2014) greenways (Lindsey, Maraj, & Kuan, 2001) public recreational programs (Dahmann, Wolch, Joassart Marcelli, Reynolds, & Jerrett, 2010) school resources (Condron & Roscigno, 2003; Roscigno, Tomaskovic Devey, & Crowley, 2006; Zhang & Cowen, 2009) food pr ovision (Larsen & Gilliland, 2008; Morland, Wing, Diez Roux, & Poole, 2002; Powell, Slater, Mirtcheva, Bao, & Chaloupka, 2007; Sharkey & Horel, 2008) healthy an d unhealthy food (Block, Scribner, & DeSalvo, 2004; Day & Pearce, 2011; Kestens & Daniel, 2010; Pearce, Blakely, Witten, & Bartie, 2007; D. M. Smith et al., 2010) and street trees or tree canopy (Heynen, Perkins, & Roy, 2006; Landry & Chakraborty, 2009; Mennis, 2006; Zhou & Kim, 2013) Both place specific and aggregate national evaluations have been conducted (Boone et al., 2009; National Recreation and Park Association, 2011) In general, this body of research shows that access to public and private a menities is an environmental justice i ssue because such amenities are not equitably distributed income and ethnic groups A few recent reviews have summarized research addressing how the provision of parks and recreational facilities varies by ethnic and income groups, showing mixed findings in terms of access to parks regardless of their size, quality and amenities they include. The reviews carried on by the National Recreation and Park Association (2011) and by Wolch et al. (2014) show a prevalence of studies finding that low income groups

PAGE 60

46 and ethnic minorities are underserved in terms of access to parks and recreational facilities. Macintyre's (2007) review challenges the assumption that low income and ethnic minority groups always have lower access to health promoting amenities, including parks. Indeed, the results of her re view show that low income communities of color do not always have lower access to health promoting facilities (Macintyre, 2007) The differences between the findings of these reviews could derive from park funding policies, on the developmental patterns of the cities that were studied, and on the methods that were used to measure access to parks. The next section critically analyzes empirical studies about access to parks to clarify the contradictions between these three review s. Access to parks To gain a better understanding of the current knowledge about the equity of park distribution across demographic groups in several geographic contexts, I conducted a literature search on the most important full text academic databases, including Web of Science, Science Direct, Jstor, and EBSCOhost My goal was to identify empirical studies measuring the spatial distribution of parks or other public green spaces in relation to socio economic, ethnic, or racial factors (criteria for this search). This review focused on access to parks including park proximity, park acreage, and park quality. Given this goal, in the literature search I used the following keywords and phrases: "park equity," "green space equity," "park equality," "green spac e equality," "park access," "green space access," "park environmental justice," "green space environmental justice," "park distribution," and "green space distribution." The search yielded 38 journal entries that met the criteria. Then, I expanded this sa mple by looking at the references of the 38 articles and found five additional papers meeting the criteria. Thus, the final sample of articles I analyzed includes 43 empirical

PAGE 61

47 papers on access to parks or green spaces (see Table A1 in Appendix A for the co mplete list). These articles focused on a variety of geographical contexts, including big cities like Los Angeles and New York City and small suburban towns in the South and Midwest. The variability of geographies likely influenced some of the results of t hese studies. To analyze these empirical articles, I developed a codebook including descriptive codes, such as the name of the authors, the year of publication, the journal and the geographical focus; topical codes, such as the scale of investigation, the methods use to measure access and the specific GIS methods, the unit of analysis, the critical distance to define access (if present), the demographic variations and the type of statistical analysis; and analytical codes, such as the presence of statistica lly significant results, the findings in terms of park proximity, park acreage, and park quality, and other detailed findings (Richards, 2005) Some codes were close ended, including the scale of investigation and the specific GIS methods, and others were open ended, including the findings of the analyzed studies. The full codebook is reported in Table B1 in Appendix B. Dates, locations and methods The sampled papers span from 1997, with the early work by Emily Talen (1997) to 2014, when a variety of papers have been published (Kabisch & Haase, 2014; Reyes, P‡ez, & Morency, 2014; Rigolon & Flohr, 2014) The search also highlighted an increase in studies on access to parks in the last five years, which employed more and more refined GIS methods probably due to progress in GIS software and to improved GIS techniques. The great majority of the sampled studies was conducted in geographic locations within the United States (29 entries), followed by places in the United Kindgom (4 entries), Canada ( 3 entries), Australia (2 entries), and other countries (South Africa, Germany, Turkey, Germany, Bolivia, Israel). Most of the

PAGE 62

48 selected studies focused on cities as their geographical scale (28 entries), followed by metro areas (13 entries), regions (2 entr ies), and countries (2 entries). In terms of methods used to measure access to parks, the vast majority of empirical studies employed GIS analyses (37 entries), while other methods that were used include quantitative surveys (5 entries), qualitative method s (2 entries), or other quantitative methods (2 entries). To classify the specific GIS method that the selected studies employed, I used Talen's (2003) framework, which includes five categories of techniques to measure access to facilities for various demographic groups: container, coverage, minimum distance, travel cost, and gravity. Container describes the quantity of amenities included within a certain spatial unit, for example a census block (Talen, 2003) Coverage depicts the amount of amenities located within a certain distance from a source (Talen, 2003) Minimum distance describes the distance between a source and the closest amenity (Talen, 2003) Travel cost depicts the mean distance between a source and all amenities (Talen, 2003) Gravity represent an indicator in which the total number of amenities (weighted by their size) is di vided by the average distance of each amenity from the source (Talen, 2003) Based on this classification, my analysis shows a relatively balanced distribution: the majority of the sample studies employed the minimum distance approach (19 entries), followed by container (16 entries), coverage (11 entries) and gravity (9 entries). A significant number of studies employed more than one GIS method to measure accessibility (13 entries), which shows that many authors aim to investigate access to parks from several perspectives. Then, among the studies that employed minimum distance, co verage and gravity, most of them used a quarter mile (12 entries) or half a

PAGE 63

49 mile (9 entries) to define access. Other distances that were employed include one mile (5 entries), two miles (1 entry), and 300 meters (1 entriy), while 11 studies used multiple d istances. This finding is not surprising, as most authors consider a quarter and half a mile as the maximum distances that people are willing to walk to reach a facility (Boone et al., 2009; Nicholls, 2001) Other methodological aspects include the unit of analysis used as the areas to be served by parks, the demographic variations included by the authors, and the type of statistical analysis. The most used units of an alysis are census tracts (12 entries), census block groups (10 entries), and neighborhoods (8 entries), followed by computer generated geometries (5 entries), census blocks (4 entries), households through surveys (3 entries), parcels (2 entries), cities (1 entry) and other units. Also, four studies employed more than one unit of analysis. The choice of the unit of analysis involves a tradeoff between a detailed geometric measure of access (smaller units of analysis) and detailed demographic information (lar ger units of analysis). In terms of demographic variation, the vast majority of the sampled studies focus on socio economic status (SES) including income, poverty and educational attainment (37 entries), while many studies focus on ethnicity (28 entries), race (14 entries) and age differences (12 entries). A large number of studies (33 entries) include more than one of the above variables. Finally, regarding statistical analysis, slightly more studies employed comparative statistics (24 entries) than correl ational statistics (21 entries). This general review of geographical focus and methods serves as important contextual information that I use to evaluate the findings presented in the next sections.

PAGE 64

50 Specificially, I focus on distance to the closest park ( pa rk proximity ), number of parks, park acreage and park congestion ( park acreage ), and park quality and safety Park proximity The evidence is mixed about how access to any type of park, expressed in terms of distance between homes and the closest park var ies by socio economics, ethnicity, race, and age groups. In summary, African Americans and other racial minorities tend to live closer to parks than Non Hispanic Whites. Among the sampled articles, nine studies found that low income groups live further fro m parks than other groups, while seven other studies found that low incom e groups live closer to parks than other groups (see Table 2). Thus, these findings do not give a clear, unidirectional picture about who lives closer to parks In terms of ethnicity, two studies highlight ed tha t ethnic minorities like Hispanics live further from parks than other groups, while three other studies found the opposite (see Table 2). The picture is clearer when focusing on race, as seven studies found that African American s and other racial minorities live closer to parks than Non Hispanic Whites, while only one study found the opposite (see Table 2). Finally, two studies highlighted that the percentage of people under 18 or above 65 is positively correlated with distance f rom parks, showing that children, adolescents, and seniors live further from parks than other age groups. To summarize, Table 2 shows that in terms of distance to parks the only substantial differences are in terms of racial groups, with racial minorities and African Americans in particular living closer to parks than Non Hispanic Whites.

PAGE 65

51 Table 2 Park Proximity: Who Lives Closer to Parks? Income Low income groups live closer to parks 7 entries (Boone et al., 2009; Kessel et al., 2009; Lindsey et al., 2001; Nicholls, 2001; Smoyer Tomic et al., 2004; K. Wells, 2005; Wen et al., 2013) Low income groups live further from parks 9 entries (Cradock et al., 2005; Dai, 2011; Erkip, 1997; NÂŽmeth et al., 2012; Omer & Or, 2005; Rigolon & Flohr, 2014; Veitch, Salmon, & Ball, 2007a; Wen et al., 2013; Wille mse, 2013) Ethnicity Ethnic minorities live closer to parks 3 entries (Cutts et al., 2009; Miyake et al., 2010; Wen et al., 2013) Ethnic minorities live further from parks 2 entries (Omer & Or, 2005; Rigolon & Flohr, 2014) Race Racial minorities live closer to parks 7 entries (Boone et al., 2009; Cutts et al., 2009; Lindsey et al., 2001; Nicholls, 2001; Wen et al., 2013; Wolch et al., 2005; Zhou & Kim, 2013) Racial minorities live further from park s 1 entry ( Dai, 2011) Age % of people under 18 and above 65 live further from parks 2 entries (Cutts et al., 2 009; NÂŽmeth et al., 2012) After examining the general picture of how distance to parks varies by income, ethnic and racial group, it is appropriate to look at some of the specific studies included in Table 2, with particular attention to the connection s between their methods and their findings. This shed s light on some of the findings presented in Table 2, and lead s to

PAGE 66

52 reconsidering some of them. For example, Nicholls' (2001) art icle, which found that ethnic minority and low income people have higher access to parks than o ther groups, is based on an analysis of Bryan, TX, a very low density city with relatively few parks and large parts of the city that are not served by parks. This lack of variation in terms of access versus no access to parks casts some doubts on the plan ning significance of the author's findings. Also, Smoyer Tomic et al.'s (2004) study, which found that low income people live slightly closer to playgrounds than other groups, sampled only neighborhoods t hat are mainly residential and urban, leaving out significant parts of the city. This sampling is problematic, as selecting only urban residential neighborhoods usually leaves out Non Hispanic White and wealthy suburbs, which might have high numbers of rec reational opportunities per person. Also, neighborhoods with mixed land use (industrial or retail) generally have high percentages of low income ethnic minority groups. Wen et al.'s (2013) article, which highlighted that people in poverty were more likely to live closer to parks in urban and suburban areas, used a method to measure distance that is problem atic when thinking about the needs of low income children and teenagers of color. Indeed, Wen et al. (201 3) calculated the weighted distance to the seven closest parks. However, some of these parks might be too far to reach for low income children and teenagers of color, due to their limited mobility; thus, their findings are not very significant in terms of children's and teenagers' access to parks. As mentioned above, one aspect to consider when defining which geographic areas of the city have access to parks is the distance defining access, which describes how far people are willing to walk, bike, or drive to parks. For example, Dai (2011) in his

PAGE 67

53 study of access to parks in Atlanta, employed several thresholds expressed in driving times (between 10 and 30 minutes, with intervals of 5 minutes). Dai (2011) ar gued that using driving times instead of walking or biking distances is appropriate for Atlanta given the city's low density. This choice is also based on the assumption that people generally drive to parks instead of walking or biking (Dai, 2011) By using this app roach, Dai (2011) found that African Americans, but not Asians, had significantly lower access to parks than Non Hispanic Whites, and that low income areas also have disproportionately lower access to parks than higher income areas While using driving times can be an appropriate choice for Atlanta, driving times do not reflect children's and teenager's opportunity to access park independently. To this regard, it would be interesting to assess ethnic and income differences in access to parks for Atlanta when using a quarter mile to define access. Indeed, in denser urban environments like New York City, Baltimore, and Denver, many studies have used the quarter mile as the critical distance defining access (Boone et al., 2009; Miyake et al., 2010; Rigolon & Flohr, 2014) To summarize, the analysis of the mentioned papers (Dai, 2011; Nicholls, 2001; Smoyer Tomic et al., 2004; Wen et al., 2013) raises questions about t he way the methods researchers used have influenced their findings about distance to the closest park across demographic groups, and partic ularly on whether low income groups live closer to parks than other groups. Therefore, the first part of this review shows that, in terms of distance to parks, the only evident difference is that African Americans and other racia l minorities tend to have a shorter distance to the closest park than Non Hispanic Whites. Park acreage When looking at the number of accessibl e parks, park acreage and park acreage per resident or child (potential congestion), a clear picture of inequity

PAGE 68

54 emerges. Table 3 shows that, amo ng the sampled articles, low income groups, ethnic minorities, racial miniorities, and young people and seniors have access to fewer parks or to less park acreage than other groups. In particula r, 17 studies found that low income people have fewer parks or less park acreage, while only five studies found that low income groups have more parks or more park acreage ( see Table 3). In terms of ethnicity, 12 studies reported ethnic minorities like Hispanics having access to fewer parks or to less park acreage while only one study found the opposite (see Table 3). Regarding race, nine studies found that African Americans and other racial minorities have access to fewer parks or to less park acreage compared to Non Hispanic Whites while only one study found the opposite (see Table 3). Finally, three studies highlighted that the percentage of people under 18 or above 65 is negatively correlated with the number of parks or park acreage, showing that children, adolescents, and seniors have access to smaller parks or to lower park acreage than the rest of the population.

PAGE 69

55 Table 3 Park Acreage: Who Has More Accessible Park Acr eage and Acreage per Person? Income Low income groups have more facilities, or more park acreage 5 entries (Ellaway et al., 2007; Gilliland et al., 2006; Smoyer Tomic et al., 2004; Talen, 1997; Vaughan et al., 2013) Low income groups have fewer facilities, or less park acreage 17 entries (Boone et al., 2009; Estabrooks et al., 2003; Gordon Larsen, Nelson, Page, & Popkin, 2006; Jones et al., 2009; Loukaitou Sideris & Stieglitz, 2002; Omer & Or, 2005; Reyes et al., 2014; Sister et al., 2009; Talen & Anselin, 1998; Talen, 1997, 2010; Weiss et al., 2 011; Wen et al., 2013; Willemse, 2013; Wolch et al., 2005; Wright Wendel et al., 2011; Wright Wendel, Zarger, & Mihelcic, 2012) Ethnicity Ethnic minorities have more facilities, or more park acreage 1 entry (Maroko et al., 2009) Ethnic minorities have fewer facilities, or less park acreage 12 entries (Comber et a l., 2008; Cutts et al., 2009; Gordon Larsen et al., 2006; Kabisch & Haase, 2014; Maroko et al., 2009; L. V. Moore et al., 2008; Omer & Or, 2005; Sister et al., 2009; Talen, 1997; Weiss et al., 2011; Wen et al., 2013; Wolch et al., 2005) Race Racial mino rities have more facilities, or more park acreage 1 entry (Talen, 1997) Racial minorities have fewer facilities, or less park acreage 9 entries (Boone et al., 2009; Cutts et al., 2009; Kabisch & Haase, 2014; Maroko et al., 2009; L. V. Moore et al., 2008; Sister et al., 2009; Talen & Anselin, 1998; Weiss et al., 2011; Wolch et al., 2005) Age % of under 18 or above 65 have fewer facilities, or less park acres 3 entries (Cutts et al., 2009; Gilliland et al., 2006; Loukaitou Sideris & Stieglitz, 2002)

PAGE 70

56 In particular, a few studies found that low income and ethnic racial minority areas have considerably less park acrea ge per person and per child than other areas, which can lead to park crowding issues (Boone et al., 2009; Sister et al., 2009; Wolch et al., 2005) In particular, Boone et al. (2009) reported that areas of the Baltimore metropolitan area with high percentages of park dependent people (people under 18 and above 65, people in poverty, and people without access to a car) have significantly fewer acres per per son than other parts of the metro area, notably middle and upper class Non Hispanic White suburbs. Sister et al. (2009) reported very similar results in terms of park dependence in the Los Angeles metropolitan area. In metro politan Los Angeles, the most densely populated areas, which house disproportionately high percentages of ethnic minorities and low income groups, have significantly fewer park acres per person than suburban communities, particularly the San Fernando Valley, where higher percentages o f Non Hispanic White middle and upper class populations live (Loukaitou Sideris & Stieglitz, 2002; Wolch et al., 2005) Furthermore, three studies reported that a reas with higher percentages of people under 18 have significantly fewer park acres nearby or a lower density of recreational facilities like sport fields and recreational centers (Cutts et al., 2009; Gilliland et al., 2006; Loukaitou Sideris & Stieglitz, 2002) In particular, Cutts et al. (2009) highlighted that in Phoenix, AZ the areas located in proximity to large parks included significantly lower percentages of people under 18 years of age. Also, Gilliland et al.'s (2006) study of London, Ontario showed that areas with h igher concentrations of low incom e groups had

PAGE 71

57 significantly fewer recreational opportu nities per youth than mid and high income neighborhoods. The shortage of park acreage and of recreational opportunities in low income communities of color with high percentages of children and teenager s is a very significant finding in terms of children's and teenagers' access to urban nature, as park overcrowding might prevent children and teenagers from playing in parks and deteriorate the conditions of parks (Sister et al., 2009) These findings are disturbing in terms of distributional equit y, as they show that the populations who need parks the most, low income ethnic and racial minority young people, have access to fewer park acres per person. These patterns of inequity are often related to geographical differences among different parts of cities or metropolitan areas, which are analyzed later in this review. In summary, while the literature shows mixed findings about which demographic groups live closer to parks, regardles of their size and quality, my review highlights striking inequities in terms of the number of parks and of the acreage of parks that different socio economic, ethnic, racial, and age groups have access to. In particular, areas wit h high concentrations of low income groups, ethnic minorities, racial miniorities, and people under 18 or above 65 years of age, have access to significantly fewer parks or to less park acreage than other areas. Park quality After assessing park proximity and park acreage, I reviewed the sampled articles to understand whether the quality of parks varies by socio economic, ethnic/racial, and age groups. In this analysis, park quality includes the presence of amenities in parks, the level of maintenance of parks, and the presence of potential physical hazards in parks. The literature I reviewed consi stently shows that park quality is

PAGE 72

58 lower in low income communities of color with high concentrations of children and teenagers, who are the groups with the highest park dependence (Anderson, Jackson, Egger, Chapman, & Rock, 2014; Crawford et al., 2008; Ellaway e t al., 2007; Estabrooks et al., 2003; Jones et al., 2009; Landry & Chakraborty, 2009; Loukaitou Sideris & Stieglitz, 2002; L. V. Moore et al., 2008; Rigolon & Flohr, 2014; Smoyer Tomic et al., 2004; Talen & Anselin, 1998; Vaughan et al., 2013; Zhou & Kim, 2013) In particular, access to play amenities in parks and the quality of such amenities are inequitably distributed. Park dependent populations have lower levels of access to playgrounds and recreational opportunities than Non Hispanic White middle an d upper class groups (Jones et al., 2009; L. V. Moore et al., 2008; Rigolon & Flohr, 2014; Talen & Anselin, 1998; Vaughan et al., 2013) Most importantly, such play amenities and in general the parks have lower quality (Ellaway et al., 2007; Loukaitou Sideris & Stieglitz, 2002; Vaughan et al., 2013) to have lower levels of maintenance (Carlson, Brooks, Brown, & Buchner, 2010; Smoyer Tomic et al., 2004) to have less shading (Anderson et al., 2014) to be more crowded (Loukaitou Sideris & Stieglitz, 200 2) and to include more physical environment hazards (Carlson et al., 2010; Cradock et al., 2005; Suecoff, Avner, Chou, & Crain, 1999) Also, parks in low income and ethnic minority neighborhoods tend to be more exposed to air pollution than parks in wealthier Non Hispanic White (Su, Jerrett, de Nazelle, & Wolch, 2011) This research, conducted in the United States, Canada, England, and Australia, including U.S. cities like Ne w York (Suecoff et al., 1999) Bo ston (Cradock et al., 2005) Kansas City (Vaughan et al., 2013) Denver (Rigolon & Flohr, 2014) and Los Angeles

PAGE 73

59 (Loukaitou Sideris & Stieglitz, 2002; Su et al., 2011) suggests that access to high quality play spaces in urban nature is an issue for low income children and teenag ers of color. In particular, Loukaitou Sideris and Stieglitz's (2002) study of eight parks in Los Angeles highlights tha t children living in the inner city, which has a high proportion of low income ethnic minorities, had negative views of the quality of their playgrounds and of their levels of maintenance. On the other hand, children living in the San Fernando Valley, a hi gher income and less ethnically diverse area, expressed dissatisfaction about the activities afforded by their parks, complaining that most playgrounds are designed for young children (Lo ukaitou Sideris & Stieglitz, 2002) Therefore, the quality and maintenance of play settings was an issue in lower income and higher density areas. In addition, people residing in low income and ethnic/racial minority areas have less access to street tree s than people living in other parts of the city (Crawford et al., 2008; Franzini et al., 2010; Landry & Chakraborty, 2009; Mennis, 2006; Zhou & Kim, 2013) Disadvantaged areas of cities also have fewer amenities in public open space, including tables, fountains, and cycling paths (Crawford et al., 2008) and parks with less interesting aesthetic features (Vaughan et al., 2013) Therefore, these results suggest that when studying access to outdoor play opportunities for young people, it is important to focus on access to parks with high quality and safe play opportunities rather than access to every park. Other findings The a nalysis of this literature highlighted some other findings that can help shed light on how planning can help balance inequities in park access. First, a few studies showed geographical differences in terms of park acreage between urban cores and outskirts (e.g. Boone et al., 2009; Dai, 2011; Sister et al., 2009; Wolch et al.,

PAGE 74

60 2005) Most of these studies found that suburban areas and the outskirts of a city include more park acreage per person and per child than inner cities (Boone et al., 2009; Dai, 2011; Kabisch & Haase, 2014; Loukaitou Sideris & Stieglitz, 2002; Reyes et al., 2014; Sister et al., 2009; Talen, 2010; Wolch et al., 2005) while fewer studies found the opposite spatial pattern (Miyake et al., 2010; Talen, 1997; Zhou & Kim, 2013) The prevalence of studies highlightin g higher park acreage in the outskirts is not surprising, due to greater land availability in these areas compared to dense inner cities and due to lower population density in the suburbs. However, in most studies, geographical differences are correlated w ith demographic differences, as low income communities of color generally live in inner cities (fewer park acres) and Non Hispanic White middle and upper class groups are more often located in suburbs (more park acres). However, the landscape of the count ry is shifting with the back to the city movement, which has led to an increased suburbanization of poverty and diversity (Kneebone & Garr, 2010) In particular, Sister et al. (2009) found that metro politan Los Angeles inner city areas, whe re low income communities of color live the most, include mainly small neighborhood parks, while suburban areas, with mid and high income residents, comprise large parks like natural preserves, arboreta, and beaches. Similarly, Miyake et al. (2010) highlighted that in New York City, inner city low income areas have many small parks that do not fit the needs of active recreation. Second, some authors proposed possible solutions to the distributional inequities they observed. The most common suggested solution is conducting a suitability analysis to identify areas that lack parks and where park dependent groups live, with the goal of prioritizing investments for the creation of new parks (NÂŽmeth et al., 2012; Rigolon &

PAGE 75

61 Flohr, 2014; Talen, 2003) Also, Sister et al. (2009) simulated the impact of locating new parks based on distributional equity, showing that in certain circumstances small interventions can bring about large improvements, at least in terms of park proximity. Rigolon and Flohr (2014) proposed a more co mprehensive framework for action for neighborhood greening that would includ e low income communities, local non profits and local u niversities. Finally, very few of the studies I reviewed highlighted park planning processes or tried to uncover the reasons for the observed park distributions (e.g. Boone et al., 2009; Joassart Marcelli, 2010; Wolch et al., 2005) I will discuss their findings in terms of planning processes in the next section. To summarize, the review of studies focusing o n park spatial distribution in relation to demographics showed that results are mixed in terms of distance to parks due to methodological issues and to the broad range of geographies included in the review. Thus, a pattern of equity or inequity is not clea rly identifiable. The review also highlighted that when focusing on the number of parks, park acreage and park congestion, low income ethnic communities, children, and teenagers are significantly underserved. Finally, in terms of park quality, my literatur e analysis showed that low income ethnic minority people children, and teenagers have access to parks with lower quality than other groups. When combining these three measures, the inequities between low income children of color and other groups are clear Policies, Practices, and Park Spatial Distribution A substantial body of literature has investigated the connections between urban policy and planning and social equity, including several studies trying to demonstrate the

PAGE 76

62 hypothesis that policies often c reated large inequities in urban areas (Bolotin & Cingranelli, 1983; Gillette, 1995; Lineberry, 1977; Maantay 2001; Massey, 1990) In a classic paper, Rittel and Webber (1973) argued that planning problems dealing with the distribution of value are "wicked" because in a pluralistic society with many "publics" and contrasting i nterests, it is hard to agree upon what common good and equity mean. This stance can help explain why equity issues have occurred, as groups with different interests tend to hold dissimilar opinions about what good and bad policies are (Rittel & Webber, 1973) ; thus, the groups with the strongest voice influence policies the most. Public policy and equity The public policy literature has shown several examples of policies that, intentionally or unintentionally, failed to address social equity goals. A recent book edited by Carmon and Fainstein (2013) showed the detrimental effects on social equity of neo liberal planning policies oriented to increasing the competitiveness of cities, with a focus on housing, transportation, public space and regional planning. Even policies that are driven by social justice or ecological goals can have negative consequences for underprivileged groups. For example, Gillette (1995) showed the failure of policies oriented to social equity in Washington, D.C., as most housing effor ts intended to improve African Americans living conditions excluded them from the most desirable areas of the city. Also, a study conducted in Brazil showed that policies aimed to preserve urban green areas can have environmental justice consequences for underprivileged populations who live in informal settlements (A. F. Young, 2013) This suggests that sustainability practice, as previously mentioned, often overlooks justice as one of i ts main components (Boone, 2010; Checker, 2011) Also, the implementation of ecolog ically oriented sustainability interventions like new parks can

PAGE 77

63 lead to disadvantaged populations like low income and ethnic minority groups being forced to relocate due to increased rents and property values; a phenomenon known as ecological or environmen tal gentrification (Checker, 2011; Dooling, 2009) Also, as previously n oted, a few scholars have recently argued that the inclusion of environmental justice stances can advance sustainability research and practice (Agyeman & Evans, 2004, 2003; Boone, 2010) Historical and funding explanations of park distribution Relatively few studies focused on how urban policies and planning practices have influenced the spatial distribution of parks in urban areas (Boone et al., 2009; Joassart Marcelli, 2010; D. H. Koehler & Wrightson, 1987; Pincetl, 2003; J. W. Smith & Floyd, 2013; Thomas, 2010; Wolch et al., 2005) while ot hers have suggested possible explanations for the distributional patterns that they observed (Jones et al., 2009; Lindsey et al., 2001; Maroko et al., 2009; Omer & Or, 2005) Residential segregation of low income ethnic minority groups from other wealthier groups is an important factor in the distributional inequity of parks, as areas where wealthier groups reside tend to receive more funding for public amenities and less locally unwanted land uses (Boone et al., 2009; Omer & Or, 2005) Several factors throughout the history of American cities contributed to today's park distribution (Boone et al., 2009) The historical period during which different parts of cities were developed plays a key role in understanding park location: Neighborhoods developed before the widespread diffusion of the automobile tend to be denser, to be more walkable, and to include smaller parks (Boone et al., 2009) In these older neighborhoods and cities, creating new parks was often a very problematic endeavor due to the high land value of densely populated areas (Boone et al., 2009) In these contexts,

PAGE 78

64 parks were sometimes established by purchasing adjacent small private lots, which was a time consu ming process requiring a strong commitment by public officials and citizens (Boone et al., 2009) Another way to develop parks before the advent of the automobile, and especially in the 1800s, was wealthy individuals donating land to cities, which often led to large estates being turned into parks (Boone et al., 2009; Pincetl, 2003) Among the studies that focused on the historical and planning processes that led to today's park distribution in relation to income and ethnic groups, Boone et al.'s (2009) article provides an in depth picture of Baltimore's park history. In particular, Boone et al. (2009) conducted a historic process analysis to track the evolution of park location in relation to where African Americans and Non Hi spanic Whites lived throughout Baltimore's history. In their study, they found that African Americans have been spatially segregated throughout Baltimore's history through the use of segregation regulations, restrictive covenants attached to properties, an d redlining by the Home Owners Loan Corporation (Boone et al., 2009) Th eir research also highlighted that, around the 1930s, politically influential neighborhood improvement associations helped steer public investment, including investment for parks, to areas mostly populated by upper class Non Hispanic Whites (Boone et al., 2009) With the departure of Non Hispanic Whites to the suburbs starting in the 1950s, the geography of access to parks in Baltimore significantly changed, as African Americans started moving to previously Non Hispanic White neighborhoods (Boone et al., 2009) Therefore, Boone et al. (2009) claim that African Americans in Baltimore currently live in neighborhoods with good access to parks because these parks were established when Non Hispanic White affluent people resided in them; thus, African Americans inherited good access to parks from previously

PAGE 79

65 Non Hispanic White neighborhoods. Similarly, Maroko et al. (2009) suggested that most parks were built in relatively early phases of most American cities and that, during the evolution of cities, the populations living around parks could have changed significantly. Given this process, Boone et al. (2009) argued tha t the current distribution of parks in Baltimore is a case of environmental injustice, as today's park distribution is the outcome of unjust and sometimes openly racist policies and practices. Furthermore, Boone et al. (2009) and Pincetl (20 03) suggest ed that in the early 1900s parks were sometimes established in impoverished neighborhoods as a form of social engineering, as decision makers at the time thought that park access could reduce youth misbehavior. As African Americans and Non Hisp anic Whites in Baltimore were residentially segregated throughout the twentieth c entury, parks were also de facto segregated between Non Hispanic White and African American parks (Boone et al., 2009) Therefore, evaluating procedural justice is a complex endeavor that requires taking into account a variety of factors, includin g the mechanisms and reasons through which parks were built, the demographics around parks when new parks were built, land use and zoning regulations, and residential segregation policies and practices. In an effort to conceptualize the mechanisms that det ermine park access in American cities, J. W. Smith and Floyd (2013) presented two competing hypotheses that can help explain park distribution across income and ethnic groups, the "growth machine" and the "central place" theories (p. 87). The growth machine theory posits that, since par ks are the outcome of the political economy in cities, their distribution tends to benefit higher income and Non Hispanic White people (J. W. Smith & Floyd, 2013) In other words, this hypothesis holds that city officials intentionally locate parks in wealthy

PAGE 80

66 neighborhoods, given the several benefits of parks (Byrne & Wolch, 2009) and locate locally unwanted land uses like industrial or commercial developments in low income communities of color (J. W. Smith & Floyd, 2013) On the other hand, the central place theory is grounded in the historical mechanisms of suburbanization and city expansion (J. W. Smith & Floyd, 201 3) This theory proposes that today's park distribution can be explained by the historical developmental process of cities, which from their core expand ed into a first ring of suburbs, and then to farther areas (J. W. Smith & Floyd, 2013) Thus, the central place theory does not use social class or racial explanations for park distributi ons across socio economic and racial groups (J. W. Smith & Floyd, 2013) Rather, it posits that the development cycles of cities, including centralization and decentralization periods, influenced how different demographic groups can access parks today (J. W. Smith & Floyd, 2013) To examine these two theories, J. W. Smith and Floyd (2013) conducted a case study of Raleigh, NC focusing on access to parks through variables pertaining to the two aforemen tioned theories. Their findings show that the growth machine and the central place theories can be used together to explain the pattern of park spatial distribution across income and ethnic groups (J. W. Smith & Floyd, 2013) Indeed, in Raleigh African Americans have less access to parks than Non Hispanic Whites, but the inequity is not e vident for Hispanics and Asians (J. W. Smith & Floyd, 2013) Also, as the city grew, parks and other open spaces were preserved in the urban core, showing the value attributed to these settings (J. W. Smith & Floyd, 2013) Following the growth machine and central place theories, Boone et al.'s (2009) findings can be seen in the following way: The growth machine theory determined park distribution in the first decades of the

PAGE 81

67 twentieth c entury, als o through residential segregation; and the central place theory explains the post WWII white flight to the suburbs and African Americans inheriting previously Non Hispanic White neighborhoods served by parks. Looking at the processes that led to land purc hase for the establishment of new parks can provide useful insights about procedural justice in access to parks. Thomas' (2010) review of land acquisition policies from the 1960s to the early 2000s highlighted that in th e U.S. land purchase s to establish new open spaces has overlooked equity issues and missed chances to balance existing inequities between income and ethnic groups. These missed opportunities could be due to the fact tha t, in the last decades of the twentie th century, private land trusts have taken a prominent role in land acquisition to establish new open spaces (Thomas, 2010) Also, although the goal of creating access to recreation for every U.S. citizen was stated in several policy documents, the implementa tion of most policies did not properly take into account income, ethnicity, age, and gender differences in outdoor recreation (Thomas, 2010) Also, when studying how policies have addressed park quality and park maintenance, it is important to consider that i n the 1970s many American inner cities struggled with park funding (Low, Taplin, & Scheld, 2005) including Los Ange les (Pincetl, 2003) The decline of public funding for parks was mostly due to white flight to the suburbs, which caused a reduction in real estate tax revenues (Low et al., 2005) Due to reduced investment, many inner city parks experienced a drop in maintenance and visitation (Low et al., 20 05) Park funding in Los Angeles Metropolitan Los Angeles has been the object of four studies that investigated specific park funding mechanisms exacerbating inequity in

PAGE 82

68 park distributions (Joassart Marcelli, Wolch, & Salim, 2013; J oassart Marcelli, 2010; Pincetl, 2003; Wolch et al., 2005) Pincetl's (2003) investigation shows that non profit organizations played a key part in seeking funding, creating and maintaining parks in Los Angele s. The role of non profits was particularly strong in the 1980s when, given the reduction of property taxes, available funds for parks were very limited (Pincetl, 2003) Environmental non profits are active in park planning because they aim to improve their city's environmental quality, and parks are perceived as factors of environmental quality (Pincetl, 2003) Several public private enterprises were formed, also building on the rationale that new parks would increase property values (Pincetl 2003) In relation to environmental justice, Pincetl's (2003) analysis shows that non profit organizations in Los Angeles between the 1980s and 1990s often concentrated their efforts on large regional parks because they were moved by environmental goals, and that they partially overlooked the establishment of parks in inner city Los Angeles, where low income communities of color lived and where several riots occurred Wolch et al. (2005) analyzed how two relatively recent programs to fund new parks or park improvement through grants allocated resources across the Los Angeles metro politan area (Propositions K an d A). Their results show that these programs created more inequities between socio economic and ethnic groups in terms of park acreage, as more resources per child were invested in Non Hispanic White high income areas than in areas with low income communit ies of color (Wolch et al., 2005) In particular, Proposition K favored park improvement over the establishment of new parks by acquiring new properties thus bringing investments to areas that were a lready served by parks (Wolch et al., 2005) The majority of resources from Proposition A were allocated

PAGE 83

69 for regional parks, which in Los Angeles are located far from the inn er city, thus far from the areas where most low income a nd ethnic minorities are concentrated (Wolch et al., 2005) The s e finding parallels Pincetl's (2003) results, as they highli ght the emphasis on large parks over small neighborhood parks. Joassart Marcelli (2010) conducted a more comprehensive study on park funding in different municipalities within Los Angeles, with a par ticular focus on park spending per capita in different types of municipalities. The author's findings reveal that park spending varied significantly across metro politan Los Angeles, with large inequities existing between wealthy and impoverished cities (Joassart Marcelli, 2010) Although this result is not surprising, Joassart Marcelli (2010) also highlighted that funding coming from state and non profit sources was also inequitably distributed, with low levels of investment in suburban municipalities that include d high percentages of ethnic minorities and low income people. In particular, non profit organizations allocated on average $21 per person annually in affluent cities, while they spent o nly $5 per person on average in the least wealthy cities (Joassart Marcelli, 2010) In addition, affluent cities spen t a higher percentage of their parks and recreation budget on new parks and facilities (26.1%), compared to low income municipalitie s (12.6%; Joassart Marcelli, 2010) Also, when controlling for average income and overall city budget, cities with higher percentages of children and teenagers spent more on parks and recreation (Joassart Marcelli, 2010) Yet, many impoverished municipalities also had higher proportions of children and teenagers (Joassart Marcelli, 2010) Overall, the highest spending for parks was found in older municipalities with good economic activiti es, in which residential

PAGE 84

70 areas are mixed with commercial and industrial land uses, and in which public space is usually considered important (Joassart Marcelli, 2010) Joassart Marcelli's (20 10) study highlighted that non profit organizations contribute d to creating significant differences in park spending among municipalities, thus confirming Pincetl's (2003) results. Starting from this finding, Joassart Marcelli et al. (2013) conducted a study focusing specifically on how non profit expenditure on parks varies among different cities in Southern California. Their re sults show that non profit organizations operate d more often in "affluent, fiscally stronger, suburban, conservative, and white municipalities, reproducing intra urban differences underlying health disparities" (p. 682). Joassart Marcelli et al. (2013) suggested that these inequalities could derive from a variety of factors including discrimination in the distribution of funding due to ethnicity and race; lower levels of soci al capital in communities with high percentages of ethnic minorities, and especially communities with recent immigrants; and other more important priorities for non profit organizations in low income areas such as creating jobs and providing social service s like drug or teenage pregnancy prevention to underrepresented groups. Joassart Marcelli et al. (2013) also claimed that the role of non profit organizations in park fundin g has had a growing importance, as recent policy trends led to fiscal austerity and less public money available for parks and recreation. To summarize, this relatively small body of literature shows that the history of park location in relation to where d emographic groups lived, combined with the analysis of park funding in recent times, can offer some insights about the reasons for the current level of equity in park distribution in cities and metropolitan areas. Regarding park funding, four studies focus ing on the Los Angeles metro politan area showed the growing

PAGE 85

71 role of non profit organizations in allocating resources, and how non profit organizations further contributed to park funding inequities between wealthy and less affluent municipalities (Joassart Marcelli et al., 2013; Joas sart Marcelli, 2010; Pincetl, 2003) Young People's Demographics and Perceptions and Use of Parks In this section, I review the literature that analyzed how play and recreation preferences vary by income, ethnic, and racial group; and the literature that shows how low income young people of color experience additional barriers to park access, besides the lack of available parks near their residences. These two bodies of literature are related because park designs that do not include elements t o attract us ers from different ethnicities might create barriers to park visitation for certain ethnic groups. Play and recreation preferences and ethnicity The literature reviewed by Byrne and Wolch (2009) highlights that Non Hispanic Whites generally look for park settings in which they can enjoy secluded nature, whereas African Ameri cans tend to prefer parks where group activities can be organized, such as sports and community events. On the other hand, Asians tend to prefer space for contemplating attractive nature, rather than spaces for st ructured recreation, and Hispanics generall y value more structured park spaces including settings for group and socializing activities such as tables, bathrooms and nearby parking (Byrne & Wolch, 2009) In particular, Gobster (2002) found that Non Hispanic Whites generally visit parks alone or in couples, while ethnic and racial minorities like African Americans, Hispanics and Asians access parks in larger groups, which shows a preference fo r social uses of parks. Then, Non Hispanic Whites access parks most often for individual active recreation such as running, biking, and walking, while other ethnic groups tend to pursue passive recreation social activities,

PAGE 86

72 including food related happening s (Gobster, 2002) Also, Gobster's (2002) study shows that within Chicago's largest park spec ific ethnic territories existed i.e., that different ethnic groups took ownership of specific spaces within the park. When analyzing park design through the cultural landscape lens, Byrne and Wolch (2009) suggested that most parks in the U nited S tates were planned based on English and northern European canons of beauty, whic h reflect ideals of pastoral landscapes, and which might not be appealing for all ethnic groups or for people born in other countries. For example, in Gobster's (2002) study African Americans were not attracted by natural features as much as other ethic groups, they were more inter ested in spaces for structured activities, and they were concerned about general level of maintenance of the park. These findings do not pertain to young people specifically, but they show general differences among ethnic/racial groups. Also, these finding s suggest that people from various cultures and ethnicities may have different perceptions of a certain park (Byrne & W olch, 2009) thus highlighting the importance of park designs accommodating different uses and cultures. Similarly, Floyd, Taylor, and Whitt Glover (2009) suggested that research evaluating how parks can foster physical activity in low income ethnic groups should include the assessment of whether parks fit the recreation preferences of the studied group, and their perceptions of their neighborhoods and of its parks. A few articles focused more s pecifically on how children's and teenagers' outdoor play and park visitation varies by ethnicity, or on play behaviors for specific ethnic minority young people. These studies show a few ethnic differences in play behaviors and park visitation. Loukaitou Sideris and Sideris (2010) found ethnic

PAGE 87

73 differences in children's use of sport fields and courts: Hispanic children and youth utilized soccer fields more often than other groups, African Americans and Asian young people used basketbal l courts more often than other groups, and N on Hispanic W hite young people used baseball fields more frequently than other ethnicities. Also, more Hispanic and Non Hispanic White young people participated in little leagues games than African Americans and Asians (Loukaitou Sideris & Sideris, 201 0) Kaczynski, Stanis, Hastmann, and Besenyi's (2011) study of physical activity in the parks of Kansas City, MO highlighted significant differences in physical activity levels based on ethnicity and gender. In particular among children male Non Hispanic White park visitors were more physically active than female ethnic minority park visitors (Kaczynski et al., 2011) This situation was reversed for teenagers, as ethnic minority female tee nagers were engaged in moderate to vigorous physical activity more often than Non Hispanic White males (Kaczynski et al., 2011) These findings suggest significant variations in terms of age and ethnicity, and therefore that parks need to include a variety of settin gs for active and passive recreation. Differences in park and playground visitation among ethnic groups were also found. Karsten (2003) highlighted that in Amsterdam the prevalent eth nic group of the neighborhood was overrepresented in the playground. Also, Moroccan and Turkish female child ren over the age of 10 visited playgrounds very rarely, possibly due to cultural reasons (Karsten, 2003) Also, in Kansas City, MO park users from ethnic minority groups exceeded their representation in the city as a whole (Kaczynski et al., 2011) suggesting that ethnic minor ities might have visited parks more often than Non Hispanic Whites. Similarly, in Los Angeles children living in the inner city, an area with

PAGE 88

74 more ethnic minority gro ups, use d parks more often for outdoor recreation than children in the San Fernando Valley, an area with a lower percentage of ethnic minorities (Loukaitou Sideris & Stieglitz, 2002) In addition, some studies found that children of different ethnicities play together in playgrounds or in organized sport activities (Blatchford, Baines, & Pellegrini, 2003; Loukaitou Sideris, 1995, 2003) while in other more unstructured settings and si tuations play groups can be ethnically segregated (Loukaitou Sideris, 1995) Also, sports fields and organized sports activities like L ittle L eague games are important settings for activities for ethnic minorities, especially for Hispanic chil dren and teenagers (Loukaitou Sideris & Sideris, 2010; Perry, Saelens, & Thompson, 2011; Ries et al., 2008) Young people living in low income inner city neighborh oods tend to value highly the presence of natural settings, including parks and playgrounds (Castonguay & Jutras, 2009; Nasar & Holloman, 2013; Rudkin & Davis, 2007) probably due to a shortage of green spaces in their neighborhoods. In particular, young people sometimes see nature as an element of neighborhood beautification, which would encourage residents to care more about their locale (Rudkin & Davis, 2007) Also, low income Hispanic teenagers living in inner city Los Angeles looked for parks to have flexible, multiple, and unstructured uses, including active and passive activities, while adults in the same neighborhood want parks to have very structured and predetermined uses (Gearin & Kahle, 2006) The availability of a large varie ty of play activities was also valued by African American children living in Columbus, OH (Nasar & Holloman, 2013) The importance of parks affording a variety of play activities is supported by ma ny studies, and it is discussed in the Park Quality Index for Youth ( PQIY; see Chapter 5) Another

PAGE 89

75 element that low income inner city ethnic minority young people perceive as important is park maintenance and cleanliness (Gearin & Kahle, 2006; Loukaitou Sideris & Stieglitz, 2002; Ries et al., 2008) probably due to parks in low income ethnic minority inner city neighborhoods having low levels of maintenance (see above discussion). Bar riers to park access for low income young people of color Besides lack of available parks and amenities, low income and ethnic minority children and teenagers have more obstacles to play in parks and to public space use than Non Hispanic White middle and upper class youth (R. Austin, 1998; Cutts et al., 2009; Ginwright & James, 2002; D. A. Harris, 2003; Neckerman et al., 2009; Platt, 2012; Spilsbury, 2005; Stodo lska et al., 2013; Valentine, 1997, 2004; Weiss et al., 2011) First, negative neighborhood factors limit low income and ethnically diverse young people's play in parks. A few studies showed that real and perceived dangers limit low income ethnic minority young people's use of public space, including parks (G—mez et al., 2004; Platt, 2012; Spilsbury, 2005; Stodolska et al., 2013) Other studies found that low income and ethnic/racial minority neighborhoods have higher concentrations of crime (Cutts et al., 2009; Franzini et al., 2010; Neckerman et al., 2009; Rutledge et al., 2003; Weiss et al., 2011) which limit children's and teenagers' park visitation. In particular, Spilsbury's (2005) study showed that neighborhood violence has significant effects on children's independent mobility depending on gender and age. In neighborhoods with high violenc e, girls' home range was significantly lower when travelling alone compared to travelling in groups, while no difference was found for boys (Spilsbury, 2005) Also, Platt's (2012) investigation of a low income neighborhood in Milwaukee highlighted that children are afraid to access parks and other public ame nities

PAGE 90

76 due to the presence of gangs or unknown teenagers. Instead, children prefer to play in alleys, vacant lots and sidewalks, in which they feel safer because these settings are free from gangs (Platt, 2012) and possibly subject to passive surveillance from residences. Similarly, Castonguay and Jutras's (2009) study of low income children's place preference in Montreal suggested that outdoor spaces around homes such as back yard gardens or shared green spaces in public housing were the settings where children felt safer, also due to the presence of familiar people. Studies conducted in Chicago indicated that Hispanic tee nagers limit their access to parks due to fears of crime, such as gang violence occurring in public settings like streets and parks (Shinew, Stodolska, Roman, & Yahner, 2013; Stodolska et al., 2013) On the other hand, a study conducted in Los Angeles showed that in low income areas parks can also be perceived as relatively safe if compared to streets, especially during the day (Loukaitou Sideris, 1995) Also, Hispanic teenagers in Chicago consider recreational activities that take place within institutionalized settings like school grounds and that are supervised by adults as safer (Shinew et al., 2013; Stodolska et al., 2013) Similarly, another study in Los Angeles highlighted that the presence of a community center near a park made the park feel safer than the rest of the neighborhood, which children perceived as dangerous (Loukaitou Sideris, 2003) Therefore, the presence of well reputed institutions with trusted adults increases children's and teena gers' sense of safety in low income and ethnic minority neighborhoods. Studies about adults' sense of safety in relation to the demographic composition of their neighborhoods are also important, as caregivers generally influence their children's mobility licenses. Franzini et al. (2010) highlighted that, in three cities across

PAGE 91

77 the U nited S tates residents of low socio economic areas perceived their neighborhood s as "less safe, with more physical disorder, decay, and vacant lots/houses" (p. 272). A study conducted in St. Louis, MO showed that African Americans tend to consider their community as more dangerous and less aesthetically pleasing than Non Hispanic Whi te 's perceptions, despite the demographic characteristics of their community (Boslaugh, Luke, Brownson, Naleid, & Kreuter, 2004) The subpar quality of the physical environment and the higher safety concerns that low income communities of color experience can also be considered environmental justice issues (W. C. Taylor, Floyd, Whitt Glover, & Brooks, 2007; W. C. Taylor, Poston, Jones, & Kraft, 2006) When looking at reported crime data, Cutts et al. (2009) found that the percentage of Hispanic immigrants l iving near parks in Phoenix, AZ was positi vely correlated to the frequency of crime. The percentage of people under 18 was also positively associated to the number or crimes i n proximity of parks (Cutts et al., 2009) In New York City, Weiss et al. (2011) found that the least affluent areas of the city included more reported crime, less pedestri an safety and more locally unwanted land uses, such as industrial areas. Second, research shows that race, class, and age influence young people's use of public space regardless of crime and other social environmental factors (R. Austin, 1998; Norris & Armstrong, 1999; Valentine, 2004) Youth of color experience issues in navig ating public space due to racial profiling, police harassment or vigilante violence (R. Austin, 1998; C. J. Cohen, 2010; D. A. Harris, 2003; Norris & Armstrong, 1999) Also when looking at age groups, children in public space tend to be socially depicted as victims and teenagers as troublemakers, leading to both groups being restricted in public space use (Valentine, 1997, 2004)

PAGE 92

78 In particular, Norris and Armstrong's (1999) study of B ritish public spaces monitored through Closed Circuit Television (CCTV) showed that black male youth were the public space users most commonly targeted by police or vigilantes. Disturbingly, in most cases black people were targeted not for their behaviors but for their skin color (Norris & Armstrong, 1999) Also, according to Austin (1998) several local, state, and federal laws in the United States created de facto barriers to recreation for blacks and other racial and ethnic minorities, including limitations to access public and privat e venues (for example, by denying the use of public venues for hip hop or reggae concerts), heavy patrolling of recreational gatherings with high percentages of ethnic minorities, curfews targeted to ethnic minorities' neighborhoods, and lack of public tra nsit to access recreational sites. When looking specifically at parks, in the past many U.S. cities, and especially municipalities located in the South, had park systems that were racially segregated (Boone et al., 2009; By rne & Wolch, 2009) In particular, park systems included specific parks that only Non Hispanic Whites could access and other parks, which were smaller and less maintained, that African Americans could use (Boone et al., 200 9; Byrne & Wolch, 2009) S ome forms of subtle segregations still exists in parks, such as the prohibition of playing specific games like soccer might be intended to exclude certain park users (Martin, 2004) Also, Byrne's (2012) study of an urban national park located in Los Angeles showed that Hispanics fel t unwanted in such a park for a variety of reasons: the majority of park users were Non Hispanic Whites; the areas located around the park were mostly Non Hispanic White and upper class; the park did not include signage in Spanish; Hispanics were concerned about being harassed or victimized; and some

PAGE 93

79 Hispanics reporting actual occurrences of intolerance and prejudice. Similarly, African Americans in a Chicago urban park were concerned about being prejudg ed by other park visitors, park workers, and police (Gobster, 2002) Finally, studies of low income youth of color show t hat some teenagers have one or more full time jobs to support their families (Ginwright & James, 2002; Williams & Kornblum, 1985) thus time constraints limit their opportunities to be in public space. In particular, Byrne and Wolch (2009) suggested that low income people's use of parks might be limited by time constraints due to work. To summarize, these two bodies of literature suggest that, besides lack of available parks, low income ethnic minority children and teenagers experience three additional types of barriers to o utdoor recreation in their neighborhood. The first derives from negative neighborhood factors like crime and perceived dangers often associated with parks; the second deals with issues in navigating public space due to racial profiling, police harassment o r vigilante violence; and the third involve s having limited time for outdoor recreation due to other commitments. Park Availability and Young People's Use of Parks This section reviews the scholarship that investigated whether the presence of parks and spa ces for play close to children's and teenager's homes fostered their visitation of such spaces. A recent review of qualitative research about park access showed associations between the presence of a park within walking distance and repeated access to that park (McCormack et al., 2010) Also, several recent quant itative and qualitative empirical studies highlighted that the availability of nearby parks is a necessary condition for children's and teenagers' use of parks (Boone Heinonen,

PAGE 94

80 Casanova, Richardson, & Gordon Larsen 2010; D. A. Cohen et al., 2014; Jansson & Persson, 2010; KyttŠ, 2004; Loukaitou Sideris & Sideris, 2010; Ries et al., 2008, 2009; Silver, Giorgio, & Mijanovich, 2014; Slater, Fitzgibbon, & Floyd, 2013; Veitch et al., 2006) The connection between park proximity and young people's visitation of parks can be found in different geographical contexts and in cities of different sizes, including: small towns in Finland and Belarus (KyttŠ, 2004) small towns in Sweden (Jansson & Persson, 2010) Melbourne, Australia (Veitch et al., 2006) Baltimore, MD (Ries et al., 2008) Los Angeles, CA (Loukaitou Sideris & Sideris, 2010) and New York City, NY (Silver et al., 2014) In particular, Loukaitou Sideris and Stieglitz (2002) fou nd that park proximity mattered more for inner city children than for children living in suburban areas, as parks were the only opportunity for outdoor recreation in inner cities. In inner cities, park proximity also matters due to children's and teenagers limited means of transportation (Loukaitou Sideris & Sideris, 2010; Ries et al., 2008) Also, some of the studies finding connections between park availability and park use focused on ethnic minorities like African American s and Hispanics (Ries et al., 2008 2009; Silver et al., 2014; Slater et al., 2013) ; thus showing the importance of park proximity for ethnic minorities. The presence of nearby parks can also be a positive contribution to children and teenagers' active play and physical activity (Boone Heinonen et al., 2010; D. A. Cohen et al., 2014) However, while the presence of parks in young people's communities matters for their park visitation, other factors also inf luence whether children and teenagers access parks or not, including: caregiver's mobility licenses, park quality, and park and

PAGE 95

81 neighborhood safety. First, as previously discussed, caregivers' fears of strangers limit s children's licenses to play outdoor (Valentine, 1997; Veitch et al., 2006) Further, the amenities included in parks and their level of maintenance also influence children's and teenagers' park visitation (McCormack et al., 2010; Ries et al., 2009) In particular, Ries et al.'s (2008) study of African American teenagers in Baltimore highlighted that teenagers loo ked for parks with good maintenance and with sport facilities. Also, teenagers tended to avoid parks that were poorly maintained and that included physical hazards (Ries et al., 2008) Park aesthetics and the prese nce of age appropriate amenities were also connected to park visitation (Ries et al., 2009) In Chapter 5, I review more in depth the litera ture on children's and teenager s outdoor play and distill a few factors of park quality in relation to young people's needs. Also as introduced in the previous section, studies conducted in low income communities of colors show that, even if parks are available, children and teenagers might not use them because they perceive them as dangerous (Platt, 2012; Ries et al., 2008; Slater et al., 2013; Veitch et al., 2006; Veitch, Salmon, & Ball, 2007b) or because they perceive the neighborhood as unsafe (Carver, Timperio, & Crawford, 2008; Griff in, Wilson, Wilcox, Buck, & Ainsworth, 2008; Jarrett, Bahar, McPherson, & Williams, 2013; Loukaitou Sideris & Sideris, 2010; Malone, 2007; Molnar, Gortmaker, Bull, & Buka, 2004) due to gangs or physical hazards. However, a study by Outley and Floyd (2010) showed that African American caregivers, even when they percei ved their neighborhood as unsafe, often adopted strategies that allowed children to access parks, including other adults' supervision through affiliation networks (family and friends) and structured recreation activities.

PAGE 96

82 This literature shows that, altho ugh park proximity is necessary for park visitation, especially among inner city ethnic minority youth, other factors affect park visitation, including park and neighborhood safety, and park quality. Thus, the connection between park availability and park use is problematic, as the presence of nearby parks is a necessary but not sufficient condition for park visitation. Park Availability and Environmental Justice To summarize, the findings of these bodies of literature are disturbing from an environmental j ustice perspective, as the y reveal documented inequities in how children and teenagers of different ethnicities and social classes have opportunities to access parks and green spaces. Although parks are not always perceived as amenities in low income commu nities of color, they at least are public spaces with the potential to serve their communities (Boone et al., 2009) These findings also show that in most U.S. cities urban planning failed, intentionally or unintentionally, to provide equal and equitable access to urban parks for all ethnic and income groups. According to Talen (2010) urban parks are "the single most important category of publicly owne d open space in US cities" (p. 473). Furthermore, urban parks bring economic, public health, and environmental benefits to cities (Byrne & Wolch, 2009; Harnik & Crompton, 2014; Sherer, 2006) These benefits are another reason why it is important to understand the spatial distribution of parks in cities and the reasons for such distribution better In the next section, I highlight gaps and inconsistencies in the literature about park accessibility, which lead to my research goals and questions.

PAGE 97

83 Literature Gaps and Inconsistencies Main Gap: La ck of Focus on Just Processes, Just Outcomes, and Park Quality This literature review showed a series of gaps in relation to young people's access to play in parks, which I briefly outlined in the introduction. Based on my extensive review of the literatu re, the main gap is that no previous study on access to parks has systematically integrated planning policies, design and planning practices, young people's perceptions and use of parks, and young people's socio economic and ethnic characteristics in the i nvestigation of the spatial distribution of parks across income and ethnic groups (see theoretical framework). The integration of these constructs is important for two main reasons. First, as previously discussed, David Schlosberg's (2004) definition of environ mental justice includes both just distribution of public amenities and environmental hazards and just decision making processes that led to such distributions. Therefore, it is important that studies focusing on access to parks as an environmental justice issue integrate the foci on "just processes" and "just outcomes," intended as today's spatial distribution of parks. Although a variety of "which came first" studies (Been & Gupta, 1997; Boone, 2002; Hurley, 1997; Pulido, 2000) explored whether environmental bads have been intentionally located in low income and ethnic minority communities, only a f ew recent studies on access to parks (Boone et al., 2009; Pincetl, 2003; Wolch et al., 2005) started exploring the reasons behind the spatial distribution of parks among different income and ethnic groups. Boone et al. (2009) argued that a shortcoming of the environmental justice literature on access to environmental goods is that, in many studies, claims abo ut injustice and racism are made based on distributional

PAGE 98

84 aspects only, without assessing the processes that led to the observed distributional patterns. Based on my literature review, this is particularly true for studies focusing on access to parks. Secon d, as the literature on children's and teenager's outdoor play shows that park quality matters for park use (see Chapter 5), it is important to include an index of park quality in the analysis of park distribution, which can be developed based on the liter ature on children's and teenagers' park use and park preference. Therefore, studies on access to parks as an environmental justice issue should integrate the foci on "just processes," "just outcomes," and "which parks really matter?" According to my review no study on access to parks brought together these three key pieces. Less Substantial Gaps Besides th is main fundamental gap, my literature review showed a few minor gaps that need to be addressed. First, relatively few studies on access to parks as an environmental justice issue included the percentage of young people in the spatial and statistical analysis (Cutts et al., 2009; Franzini et al., 2010; Loukaitou Sideris & Stieglitz, 2002; Talen & Anselin, 1998 ; Talen, 2003; Wolch et al., 2005) Not including such percentages means assuming that children and teenagers are equally distributed in the city, without taking into account the areas that have more demand for play spaces. Although qualitative Geographic Information System (GIS) is a growing field (Kahila & KyttŠ, 2009; KyttŠ, Broberg, Tzoulas, & Snabb, 2013) no previous study about access to parks has integrated qualitative research on children's and teenagers' use and perc eption of parks in a GIS spatial analysis.

PAGE 99

85 Second, a few authors highlighted that research on access to parks did not sufficiently focus on park quality and suggested that future studies should investigate more in detail the quality of parks and of play am enities, with a particular emphasis on the needs of the age and cultural groups that will mostly benefit from them (Boone et a l., 2009; Ellaway et al., 2007; Kabisch & Haase, 2014; Loukaitou Sideris & Stieglitz, 2002; Reyes et al., 2014; Smoyer Tomic et al., 2004; Talen, 2003; A. F. Timperio et al., 2007) This links to the necessity to investigate "which parks really matter" fo r children and teenagers. In particular, some authors recommended considering young people's play needs (Boone et al., 2009; Rigolon & Flohr, 2014) and others pointed out that recreational preferences tend to differ by ethnic and racial groups (Kabisch & Haase, 2014 ; Madsen et al., 2014) also due to safety concerns in low income ethnic neighborhoods (A. F. Timperio et al., 2007; Weiss et al., 2011) In terms of the methods to evaluate these differences, Loukaitou Sideris and Stieglitz (2002) suggested that "quantitative methods cannot sufficiently measure quality" (p. 479), thus supporting the use of qualitative methods to analyze park qua lity for different age and ethnic groups. These suggestions further reinforce the need to include young people's perceptions and use of parks in analyzing children's and teenagers' access to urban nature as an environmental justice issue. Third, the land u se and zoning around parks are other aspects that have not been taken into account by most studies, with a few exceptions (e.g. Boone et al., 2009; Cutts et al., 2009; Talen, 2010) Lan d uses and zoning in the proximity of parks need to be included because different land uses and zoning regulations are likely to generate different park needs (Talen, 2010) For example, areas with a prevalence of multi family

PAGE 100

86 housing, which are generally located in low income ethnic areas (Wolch et al., 2005) ne ed more parks than areas composed of single family residences, as the latter generally have private yards Also, in the past planners and decision makers have used land use policies to locate undesired land uses far away from middle and upper class Non His panic Whites (Boone et al., 2009) Thus, land use policies have unintent ionally contributed to how different socio economic and income groups can benefit from parks today. This is another reason why land use and zoning policies need to be taken into account when studying access to parks. Finally, as previously highlighted, whi le some studies have proposed solutions to increase access to play in parks for young people living in underserved areas (NÂŽmeth et al., 2012; Rigolon & Flohr, 2014; Sister et al., 2009; Talen, 2003; Wolch et al., 2005) to my knowledge only one article about access to parks (Rigolon & Flohr, 2014) suggested a framework for neighborhood greening in partnership with local communities. Methodological Variations and Inconsistencies Besides the gaps I highlighted, this literature review showed three main methodological variations or inconsistencies. The first methodological variation is that, in the pa rk access literature, accessibility and proximity to public amenities have been measured through different instruments. Most of the studies I reviewed used a Geographic Information System (GIS) spatial analysis to measure distances between residences and p ublic amenities, while fewer studies used surveys of residents, asking them to report their distance from various public amenities (see above section). GIS analyses may highlight more reliable results about physical spatial distributions than surveys becau se they rely on objective measures of the physical environment rather than

PAGE 101

87 on self reporting. On the other hand, surveys can capture perceived distances from parks and other perceived barriers to park visitations related to both the physical and social env ironment (e.g. Franzini et al., 2010; Veitch et al., 2007a) The choice between these two methods depends on the goal of the study and on the available means (GIS data are generally publ icly available secondary data, while survey data are generally primary data). Second, my literature analysis showed that some variability exists even in the studies using GIS. In particular, I discussed above the different GIS methods that were used to def ine access, including container, coverage, minimum distance, travel cost, and gravity (Talen, 2003) When examining the results of equity mapping studies, it is important to check that the methods to define access are appropriate to what the researchers are trying to measure. In particular, while approaches such as container and c overage are better suited to study park acreage, the minimum distance approach is more appropriate to measure park proximity. Third, variations were also found in the size of the unit of analysis (parcel, census block, census block group or neighborhood). Indeed, Talen (2003) and Baden et al. (2007) suggested that the unit of analysis can be a factor influencing the results of equity mapping investigations. Thus, it is important to be aware that, in case of studies that employ large u nits of analysis like census tracts or neighborhoods, there are significant variations within them that the findings cannot describe. Also, as previously noted, choosing a unit of analysis requires a tradeoff between a detailed geometric measure of access (smaller units of analysis) and detailed demographic information (larger units of analysis).

PAGE 102

88 Fourth, my review showed inconsistencies in the distance that various researchers used to define which census blocks, census groups, or census tracts have access t o parks. While the quarter mile is the most used measure (Boone et al., 2009; Rigolon & Flohr, 2014; Wolch et al., 2005) there are significant variations within the sample I analyzed. When studying access to parks for children and teenagers, it is important to use measures that are based on the literature on young people's independent mobility. Indeed, children's home range is very limited and rarely exceeds a mile (Herman, Heins, & Cohen, 1987; Loebach & Gilliland, 2014) Also, as it was previously mentioned, Miyake et al. (2010) appropriately suggest that researchers need to consider the size and urban form of cities when choosing the distance that defines access. For example, there is an evident difference between densely populated cities like New York City, where people mostly walk or take transit to destinations, and dispersed car dependent cities like Los Angeles (Miyake et al., 2010) In this dissertation, I address the main substantial gap and the lesser ones through my general research approach which includes the focus on planning processes, on park spatial distribution, and on park quality, and through my specific research methods. In the methods section I explicitly discuss how I address the variations and inconsistencies that the literature showed about the measurement of park access. Research Objectives and Questions Given the literature review and the gaps I highlighted, this study seeks to uncover the potential connections between planning policies, planning and design practices, park spat ial distribution across demographic groups, and young people's perceptions and use

PAGE 103

89 of parks. In particular, it builds on a pilot study I conducted with a colleague (Rigolon & Flohr 2014) by expanding its scope and sample and by refining its methods. Research Objectives The overall research objective of this study is to understand how planning policies and practices in Denver, CO affect the way young people of different demographi c groups access public parks, and specifically public parks with high quality amenities that are important for young people's park visitation. A study stemming from this goal can make significant contributions to planning scholarship and practice, which I highlighted previously. The general research objective can be detailed in three more specific goals pertaining to young people's perceptions and use of parks, park planning processes, and park spatial distribution. First, I aim to uncover what are the park features that children and teenagers of different ethnicities perceive as positive, that lead s to park visitation. This also allow s me to assess the quality of parks based on young people's preferences and to develop recommendations for park planning in D enver. Second, I aim to analyze the spatial distribution of parks, expressed as park proximity, park acreage, and park quality, across socio economic, age, and ethnic groups in the city of Denver. By doing this, I also aim to advance the way previous plann ing studies have measured the distribution of public amenities by using a finer and more comprehensive measure of access, as my analysis includes park proximity, park acreage and park quality. Third, I intend to explore how park planning, land use plannin g, housing policies, and design practices contributed to today's spatial distribution of parks in Denver, with a particular focus on high quality parks for children and teenagers. In depth qualitative

PAGE 104

90 investigations about the historical and political proce sses leading to environmental equity or inequity are a necessary complement to quantitative studies about spatial distributions of environmental goods and bads (Boone et al., 2009; Downey, 2007; Mennis & Jordan, 2005; Saha & Mohai, 2005) By understanding the reasons behind observed distributional equities and inequities, planners can hopefully develop better policies and practices to create more equitable cities in the future. To this regard, I also aim to further develop a park advocacy framework that I conceptualized in th e pilot study (Rigolon & Flohr, 2014) to help low income and ethnic minority communities take action to increase the quantity and quality of their parks. This dissertation also in cludes a methodological goal. In this study, I aim to test the use of a mixed methods research approach in a planning study focusing on environmental justice issues. In particular, I seek to uncover how well a quantitative Geographic Information System spa tial analysis focusing on distributional equity and a qualitative analysis of park planning process can be integrated in a cohesive mixed methods research study. Also, this study intends to exemplify how a qualitative analysis of park planning process can complement a quantitative GIS spatial analysis by provid ing deeper insights on the complex mechanisms of park distribution, thus generating more significant findings for the practice of planning for park provision. Research Questions Based on the literatur e gaps and research objective, this concurrent mixed research study (Leech & Onwuegbuzie, 2009) includes an overarching mixed research qu estion and several more detailed research questions for each strand of the study.

PAGE 105

91 The overarching mixed research question is: In Denver, CO, what policies and practices have resulted in the current spatial distribution and quality of public parks f or young people of different incomes and ethnicities? As this general research question includes park planning processes, park spatial distribution, and park quality I ask three sub questions that are related to these three constructs. The first sub question, whic h focuses on park planning processes, addresses the different planning policies and practices that have contributed to the location of parks: Historically, what park planning, land use planning and/or public housing policies led to decisions for locating p ublic parks, and especially parks that are suitable for children's and teenagers' play? This question can help uncover "which came first" (Boone et al., 2009) between parks and people i.e., whether park availability influenced the settlement of wealthy populations around parks, or vice versa, whether the presence of wealthy po pulations attracted more park investment in certain neighborhoods. Also, this first sub question will shed light on which policies and practices throughout Denver's history have had the strongest influence on today's park distribution. The second sub quest ion focuses on park spatial distribution across demographic groups: Currently, are public parks equitably distributed across socio economic, ethnic, and age groups? This sub question can help uncover whether some areas of Denver are underserved in terms of parks, and whether certain demographic groups currently benefit from parks more than other groups. Finally, the third sub question, which covers park quality, focuses on differences among different neighborhoods: Currently, are public parks in low income ethnic minority neighborhoods equitably designed and maintained in comparison to public parks in other parts of the city?

PAGE 106

92 The overarching mixed research question and the three sub questions cover all the constructs included in the PEPDQ f ramework (see Fig ure 1). Mixing Purposes Mixed research studies require a rationale for mixing qualitative and quantitative data (Cresw ell, 2009; Teddlie & Tashakkori, 2009) Mixing purposes identify the reasons for employing both qualitative and quantitative approaches and data (Collins, Onwuegbuzie, & Sutton, 2006) This study employs some o f the mixing purposes theorized by Greene, Caracelli, and Graham (1989) including complementarity, development, and expansion. Complementarity C omplementarity deals with taking advantage of the complementary strengths between qualitative and quantitative research, while minimizing the shortcomings of both (Greene et al., 1989) The methods I use to study the different constructs included in my theoretical framework are complementary. The qualitative review of young people's outdoor play and park visitation helps me uncover children's and teenagers' play patterns and preferences, and to develop a park quality index. The quantitative spatial an alysis allows me to assess the distribution of parks in relation to demographic groups in the whole city of Denver. Finally, the qualitative policy analysis helps me understand the reasons behind the observed spatial distribution of parks. By using the mos t appropriate methodology in each phase, I can maximize the strengths and minimize the non overlapping weaknesses of qualitative and quantitative research.

PAGE 107

93 Development Development involves employing the findings from one method (for example, qualitative) to design the techniques employed in the next phase of the study (for example, quantitative; Greene et al., 1989) In my study, each phase builds on the previous one in terms of findings and methods. For example, the park quality score that I employ in the GIS spatial analysis is built on the review of the literature on young people's outdoor play and park visitation. Expansion Expansion describes the use of more than one method to broaden the scope of an investigation (Greene et al., 1989) In this dissertation, I employ qualitative research to study park planning processes and young people's perceptions and uses of parks, thus, expanding the scope of most quantitative only equity mapping studies of access to parks, which focus only on park spatial distribution across income and ethnic groups.

PAGE 108

94 CHAPTER III RESEARCH METHODS In this study, I employ an mixed methods research design (Leech & Onwuegbuzie, 2009) to uncover how park planning, land use planning, and housing policies and practices in Denver have influenced the current equity or inequity of the distribution of parks, with a sp ecific focus on parks with high quality amenities for children and teenagers. This study includes two main strands, focusing on Park Planning Processes and Park Spatial Distribution In the rest of this dissertation, these two phrases are used to identify the two main strands. The Park Planning Processes analysis includes the qualitative data analysis of secondary sources that can document Denver's park, land use, housing, and ethnic history, and interviews of professionals and scholars who are knowledgeab le of Denver's past and current park and land use planning practices. The Park Spatial Distribution analysis comprises a qualitative review of literature on young people's outdoor play to create an index of park quality; and a GIS spatial analysis of the c urrent spatial distribution of parks in Denver. Population and Sampling Frame Each strand of this study has its specific population and sampling frame. As a general sampling framework, I employ concurrent parallel mixed research sampling (Onwuegbuzie & Collins, 2007) The sampling frame is concurrent because the sampling for the Park Planning Processes and Park Spatial Distribution strands occurs at the same time (Onwuegbuzie & Collins, 2007) The sampling frame is parallel because the

PAGE 109

95 populations of these two strands do not have overlaps (Onwuegbuzie & Colli ns, 2007) Also, given the spatial nature of this study, much of the research does not involve interactions with participants. Park Planning Processes Analysis For the Park Planning Processes analysis, the population includes experts on Denver's planning and ethnic history, and secondary data providing information on Denver's planning and ethnic history. In particular, experts include professionals and scholars who are knowledgeable about Denver's parks, land use, public housing, and urban history, compri sing: present and former Parks and Recreation staff, Community Planning and Development staff, and U.S. Department of Housing and Urban Development Colorado staff ; and members of non profit organizations working on green space in Denver, and historians and scholars with a specific knowledge of Denver's urban history. Secondary data on Denver's planning and urban history comprise books, journal articles, reports, policy documents, plans and maps that cover aspects related to park planning, land use planning, housing, and ethnic history. To sample experts and secondary data, I used a combination of multilevel purposive sampling and snowball sampling (Onwuegbuzie & Leech, 2007a) First, to coll ect books, journal articles, reports, policies, plans and maps, I searched several academic full text databases and local libraries using keywords related to Denver's history, parks, land use, housing, and ethnic groups (see data collection instruments sec tion below for details). To recruit experts I used a snowball approach, starting from the authors of the books and articles I reviewed for the historians and from a few connections at Denver Parks and Recreation and Denver Community Planning and

PAGE 110

96 Developmen t. I stopped recruiting new experts once I reach ed data saturation (Sandelowski, 1995) The final sample of experts I interviewed included 15 planners and scholars comprising three females and 12 males The predominance of male experts could reflect gender im balances in the planning profession The estimated approximate age of these experts varied from the ear ly thirties to the late seventies, with most professionals likely included in the 40 65 age group Probably due to time availability, the response rate for experts who are not currently working for city departments was higher. To gain insight on Denver's h istory, I interviewed three scholars who have studied the city. These three male scholars shared their expertise on Denver's ethnic history, including residential segregation, the City Beautiful movement in Denver, public space and recreation. I also condu cted interviews with two current park planners one female and one male, and four former park planners at Denver Parks and Recreation including two females and two males Their roles at Parks and Recreation included drafting citywide park ma ster plans de signing specific parks, preparing maintenance and capital improvement budgets, and managing the department. Interviewing former park planners (1980s 2000s ) was particularly useful to gain a historical perspective. Two of the former park planners have manag ed Denver Parks and Recreation, thus their insight was very important To gain insights on land use, I interviewed three professionals: a current planner in the Community Planning and Development D epartment ; a practicing land use planner who has been forme rly working at the department; and a practicing land use planner and scholar who is knowledgeable on Denver's planning history. The expertise of these three

PAGE 111

97 male professionals encompassed citywide land use policies like zoning annexation policies and impa ct fees, and specific cases in which these policies were applied to establish parks. As for Parks and Recreation, interviewing planners who worked at Community Planning and Development in two different phases provided valuable historical depth. Finally, I conducted interviews with three male professionals working for housing authorities and non profit organizations. I interviewed a member of the U.S. Department of Housing in Denver (hence, HUD Colorado) who works on distributing federal funding for afford able housing in the Rocky Mountain region. I interviewed a member of Great Outdoors Colorado (GOCO) a non profit organization funding public open space acquisition and improvement in Colorado This expert works, among other tasks, on distributing grants f or open space to local municipalities in the state. Finally, I interviewed a staff member of Denver Urban Gardens ( DUG), a non profit organization focusing on community gardens in metropolitan Denver This professional facilitates the establishment of comm unity gardens in metropolitan Denver and has collaborated with Denver Parks and Recreation when gardens were located in parks. Park Spatial Distribution Analysis For the Park Spatial Distribution analysis, the population is composed of geospatial data for the GIS spatial analysis of today's park spatial distribution and of literature on young people's outdoor play and park visitation. Geospatial data includes GIS geodatabases and shapefiles covering the whole city of Denver, except areas included in the De nver International Airport census tract for the large size of the area (around 27,000 acres) and its very low population (around 1,166 persons; United States

PAGE 112

98 Census Bureau, 2010) These data cover demographic data from the 2010 U.S. Census (census blocks and census block groups), assessed land value and total value at the par cel level, the geometry and the features of parks and parkways, the street network, land use data, and crime data (City and County of Denver, 2014b) To select literature on young people's outdoor play and park visitation I searched several full text academic databases for social science disciplines with keywords focusing on children, teenagers, parks, play, and ethnic racial variations (see data collection instruments section below for details). Research Design This study employs an equal status concurrent fully mixed design (Lee ch & Onwuegbuzie, 2009) to uncover how park planning, land use planning and housing policies and practices in Denver have influenced the current equity or inequity of the distribution of parks, with a specific focus on parks with high quality amenities fo r children and teenagers. Figure 2 describes the two main strands of the study that deal with Park Planning Processes (analysis of procedural justice) and Park Spatial Distribution (analysis of distributional justice). Park quality aspects are mostly cover ed in the Park Spatial Distribution strand. The Park Planning Processes strand, a qualitative explanatory research design (C. Marshall & Rossman, 2006) focuses on how park planning, land use planning, and housing policies and practices in Denver have led to to day's parks distribution across socioeconomic and ethnic groups (analysis of procedural justice). I chose to use a qualitative explanatory design (C. Marshall & Rossman, 2006) because I aim to uncover which policies have contributed to shaping a phenomenon i.e., today's park spatial distribution across demographic groups. The data collection and analysis methods, which

PAGE 113

99 include expert interviews and the review of books, articles, and policy documents (see Figure 2), are described in the next sections. The second st rand of the study, a quantitative dominant sequential fully mixed design (Leech & Onwuegbuzie, 2009), is centered on a GIS spatial analysis of park distribution, with a specific focus on parks that, based on a qualitative review of the literature on young people's outdoor play, include significant play spaces for children and teenagers (analysis of distributional justice). The design of this strand is quantitative dominant because it relies mostly on quantitative data; it is sequential because qualitative a nd quantitative data are collected at different times; and it is fully mixed because integration occurs at several stages (Leech & Onwuegbuzie, 2009). In particular, I used the literature on young people's play in parks, including how young people's play preferences differ by ethnicity to develo p the Park Quality Index for Youth (PQIY) Through my GIS spatial analysis, I studied the relationships between socio economic levels, ethnic variations and percentage of population under 18 (independent variables) and access to parks defined as park proximity, park acreage, and park quality (dependent variables, considered one at a time). Finally, I integrated the findings of the two strands by looking at the way park planning, land use planning, and housing policie s and practices have influenced today's park spatial distribution in Denver, with a specific focus on parks suitable for children's and teenagers' play (see Figure 2). In particular, I connected specific policies to their spatial outcomes i.e., park locati on, land use, or housing types in different areas, to uncover relationships between procedural and distributional justice. Table 4 describes which research methods I used to answer the research questions of this study.

PAGE 114

100 Figure 2 Research design and metho ds

PAGE 115

101 Table 4 Connections among Research Questions and Research Methods Research Questions Research Methods In Denver, CO, what policies and practices have resulted in the current spatial distribution and quality of public parks f or young people of different incomes and ethnicity? All methods Historically, what park planning, land use planning and/or public housing policies led to decisions for locating public parks, and especially parks that are suitable for children's and teenagers' play? Secondary data on Denver's urban history; interviews with experts; literature on young people's outdoor play Currently, are public parks equitably distributed across socio economic, ethnic, and age groups? GIS spatial analysis Currently, are public parks in low income e thnic minority neighborhoods equitably designed and maintained in comparison to public parks in other parts of the city? GIS spatial analysis; literature on children's outdoor play Data Collection Instruments This study relies mostly on secondary data, although I also collected some primary data. Secondary data includes: geospatial data (GIS), literature on children's and teenagers' outdoor play and park visitation, books and articles about Denver's urban history, and policy documents including city ordi nances, plans, and maps about park s land use s and housing in Denver. Interviews with experts on Denver's parks, land use, housing and urban history are the only primary data collection instrument. The different data collection instruments pertaining to t he two main strands of the research are described below.

PAGE 116

102 Park Planning Processes Methods for collecting historical secondary data and primary interview data are described in this section Secondary data on Denver's urban and planning history To collect books, journal articles, reports, policy documents, plans and maps covering Denver's urban history, I searched several academic full text databases ( Web of Science, Science Direct, Jstor, and EBSCOhost) and local libraries (Denver Public Library, Universi ty of Colorado Boulder Library, University of Colorado Denver Library, and University of Denver Library) using the following keywords combined with "Denver:" "urban history," "park system," "parks," "land use," "zoning," "housing," "residential segregation ," "African American," "Hispanic," "Latino," and "low income;" and other keywords deriving from some of the in progress findings, such as: "redlining," "racially restrictive covenants," "exclusionary zoning," "school segregation," "northeast Denver "north west Denver and others. Interviews with experts Expert interviews allowed me to gain a good understanding of how, now and in the past, different city departments plan for park provision; of the mechanisms involved in gathering funding for parks ; and of h ow land use planning and housing policies and practices influenced the residential locations of different income and ethnic groups Also, interviewing Denver's historians allow ed me to gain general knowledge about the history of different ethnic and racial groups in the city of Denver. Besides the intrinsic value of the data I collected through interviews with experts, the interview data also helped me select additional secondary data like policy documents, plans, and maps.

PAGE 117

103 After obtaining Institutional Rev iew Board approval at the University of Colorado Denver (protocol number 14 2365 ), I contacted professionals and scholars who are knowledgeable about Denver's park planning, land use planning, housing and ethnic history. I asked their availability for pers onal interviews about park planning, land use planning, housing policies and practices, or Denver's ethnic history, depending on their expertise. When contacting experts (email or phone), I explained my research goals and how their involvement would contri bute to my dissertation without sharing the specific questions After the experts informally agreed to the interview, I set up a time depending on their availability. I conducted all interviews personally in the expert s' offices or other settings dependi ng on the participants' preferences. The interviews lasted between 30 minutes and one hour and were audio recorded. At the beginning of the interview, I explained more in detail the goals and the design of my research, how my research can benefit scholarsh ip and practice, and the specific contribution that each participant can make. After this introduction, I asked each participant if he/she agreed to take part to the study or if he/she wanted to opt out. If he/she agreed to proceeding, I asked him/her to s ign the Informed Consent form (see Form C1 in Appendix C) and started the semi structured interview (B. L. Berg, 2001) Depending on the expert ise of each participant, I asked questions related to Denver's history, park planning, land use planning, housing policies, or the role of non profit organizations. Tables 5 and 6 include the wider topics o f the q uestions, while Table C3 in Appendix C comp rise the specific questions.

PAGE 118

104 Table 5 Question Topics for Planning Professionals Question Topic Experts The mechanisms to acquire parkland (now and in the past) The barriers to acquiring parkland (now and in the past) Types of park funding and their dis tribution The role of non profit organizations in park establishment Inclusion or exclusion of young people's play needs in parks Participation in park planning (including young people) Park design and maintenance Parks and Rec Current and past sources of funding for affordable housing in Denver Criteria to prioritize the choice of subsidized housing sites (including location) Barriers to establishing affordable housing in Denver Mechanisms to achieve income integration in housing HUD Colorado Land use pl anning instruments to acquire parkland or cash for parks (now and in the past) Impact of small area plans on park establishment (now and in the past) Land use barriers to park establishment (now and in the past) Land use instruments and other mechanisms th at have led to residential segregation (now and in the past) Participation in land use planning (including young people) Community Planning

PAGE 119

105 Table 6 Question Topics for Scholars Question Experts The goals of the City Beautiful movement in Denver Park funding during the City Beautiful era Aesthetic values in parks during the City Beautiful era Political structure during the City Beautiful era Historian, City Beautiful Mechanisms of residential segregation among income and ethnic groups Effects of r esidential segregation on ethnic minorities Types of discrimination against ethnic minority groups Historic residential locations of ethnic groups in Denver Historian, ethnic groups Cases of discrimination in recreation (location, design, and regulations) Public space and ethnic minority culture Historian, recreation and public space Building on the questions presented in Table C3 in Appendix C I created a specific interview protocol for each expert I interviewed. After the interview, I transcribed the audio recording and sent the appropriate transcription to each participant to give him/her the possibility to review it and to request changes. After participants agreed to a final version of the interview transcription, I asked them to sign the Final Con sent Form (see Form C2 in Appendix C) for the use of the interview's content for my dissertation and future scholarly publications, while keeping their identity anonymous in any report. Also, the recordings and recording transcripts are kept anonymous, wit hout any reference to the subjects' identity. Finally, I deleted any email correspondence I had with the subjects to protect their identity in case of any possible data spillage.

PAGE 120

106 Park Spatial Distribution The analysis of distributional justice, focusing o n the spatial distribution of parks in Denver, relies mostly on a GIS spatial analysis. Also, it includes a qualitative review of the literature on young people's outdoor play and park visitation to incorporate aspects of park quality in the GIS spatial an alysis. GIS data collection Geographic Information System shapefiles and geodatabases were collected via the U.S. Census Bureau (U nited States Census Bureau, 2010) the City and County of Denver (City and County of Denver, 2014b) and the Denver Regional Council of Governments (DRCOG, 2015) This dataset includes the geometry of spatial elements such as census blocks, lots, buildings, parks, and roads; also, it includes attributes attached to geometries, such as demographic data relative to census blocks, land value relative to parcels, which I used to create a proxy variable for income, and amenities that can be found in parks (playground, sport fields, lakes, picnic areas, etc.). All these data are publicly accessible through the U.S. Census, City and County of Denver, and DRCOG websites, and they are anonymous and aggregated. Literature on young people's outdoor play To understand what determines children's and teenagers' outdoor play in parks and their preferences for outdoor play spac es, I collected literature on children's and teenagers' outdoor play published in the last 30 years. In particular, this review is int ended to investigate what are the park features that children and teenagers prefer or where they play more often, which are related to the reasons that contribute to children and teenagers' park visitation, to their longer engagement in play activities in parks, and to more physically active play activities. To do so, I searched several full text academic databases for social science

PAGE 121

107 disciplines with the following expressions: ("children" AND "outdoor play"), ("children" AND "play" AND "parks), ("children" AND "playground"), ["children" AND ("schoolyard" OR "school ground" OR "play yard")], ("children" AND "nature"), and ["children" AND "outdoor play" AND ("ethnic" OR "race" OR "racial")]. Data Analysis As a general framework for mixed research data analy sis, I employed an equal status, concurrent, multi type mixed analysis (Onwuegbuzie, Sl ate, Leech, & Collins, 2007) The mixed analysis is equal status because qualitative and quantitative techniques have the same weight; it is concurrent because I conduct qualitative and quantitative analyses at the same time; and is multi type because I a nalyze both qualitative and quantitative data (Onwuegbuzie et al., 2007) This overall strategy is the blueprint for the data analysis of th is study and guides the integration of the results of the two strands (see Figure 2). However, each of the two strands of the study is based on a specific dat a analysis strategy, given the particular focus and data of each strand. Park Planning Process Da ta Analysis For the park planning process analysis, I used a combination of theory driven coding (Miles & Huberman, 1994) and constant comparison analysis (Corbin & Strauss, 2007; Glaser & Strauss, 1967) In other words, I analyzed my qualitative da ta using both predetermined and emerging codes. Through this analysis I tried to uncover aspects of procedural justice or injustice in park planning. To do so, I searched for topics related to parks, land use, housing, race, ethnicity, park funding, and at titudes towards children and youth.

PAGE 122

108 Analysis of secondary data on Denver's urban history First, I critically analyzed a few books and articles about Denver's urban history to collect information about the planning and establishment of the city's major pa rks, about the different phases of the city's urbanization and land use policies, about the history of different racial and ethnic groups within the city, and about the provision of affordable housing through the city's history. This factual information pr ovided some context for the policy documents and interview data. While I read these documents, I took notes about facts and opinions that shed light on planning processes related to park, land use, and housing planning in relation to income and ethnic grou ps. To analyze my notes, I used a combination of theory driven coding (Miles & Huberman, 1994) and constant comparison analysis (Corbin & Strauss, 2007; Glaser & Strauss, 1967) My analysis included three steps (Corbin & Strauss, 2007; Glaser & Strauss, 1967) : First, I divided the narrative data into small chunks, such as sentences or groups of them; second I assigned a code to each chunk i.e., a label abstracting the content of the chunk; as said above, I used both code s deriving from the literature about park planning processes (deductive) and codes emerging from my notes (inductive); third, I clustered matching codes into themes, which are categories with a higher level of abstraction. Analysis of interview transcript s I analyzed the interview transcripts through the same combination of theory driven coding (Miles & Huberman, 1994) and constant comparison analysis (Corbin & Strauss, 2007; Glaser & Strauss, 1967) as described above. These techniques are appropri ate for the interview data because they allowed me to develop general themes that help answer my research questions, and because their

PAGE 123

109 procedures made it easier to leave an audit trail and to replicate the analyses (Leech & Onwuegbuzie, 2008) While I coded the two qualitative datasets (secondary data and interview transcripts), I establ ished connections between them. The theory driven codes for the analysis of both datasets were organized in four categories, including Denver's urban history, park planning, land use planning, and housing. The complete list of theory driven codes for both datasets can be found in Table B2 in Appendix B. In Chapter 4 (Park Planning Processes Findings), I present the findings from the two datasets jointly. Analysis of policy documents Based on the review of books and articles and on the analysis of intervie ws, I analyzed a small sample of policy documents about park, land use, and housing planning in Denver (see above sections for selection criteria). In this analysis, I attempted to relate geographically the establishment of new parks to the location of dif ferent ethnic and income groups throughout Denver's history. In particular, I looked at which ethnic groups lived in the areas where parks were implemented during several phases of Denver's history and I also analyzed which types of residential land uses w ere allowed around parks of different size and quality. Credibility and dependability To increase credibility for all quantitative data analyses (Onwuegbuz ie & Leech, 2007b) I compared my emerging findings to previous research about park planning and funding (i.e. Joassart Marcelli, 2010; Wolch et al., 2 005) and shared with academic advisors. Also, during interviews, I asked many probing questions to get detailed descriptions of park planning, land use planning, housing policies and practices, and Denver's ethnic history. To increase dependability, I le ft an audit trail of the qualitative data analysis so that the analysis can be reproduced (Onwuegbuzie & Leech, 2007b) The audit trail

PAGE 124

110 includes all chunks, codes, and themes of the qualitative data analysis process and can be found in Table B3 in Appendix B. Also, a colleague coded around 5 percent of the interview transcript data to assess intercoder reliability (Lombard, Snyder Duch, & Bracken, 2002) Percent agreement was above .80 for all codes related to park planning, land use planning, and housing, is considered a n acceptable rate (Lombard et al., 2002) Park Spatial Distribution Data Analysis For the park spatial distribution analysis, I employ ed a quantitative dominant, sequential, multi type mixed analysis (Onwuegbuzie et al., 20 07) The analysis is quantitative dominant because qualitative analyses and data have the highest relative weight on the findings (GIS spatial analysis); and it is sequential because data analysis includes several consecutive steps (see Figure 2; Onwuegbuzie et al., 2007) The core of this analysis is the geospatial analysis of park distributi on in relation to the location of different income, ethnic, and age groups in Denver (see Figure 2). Young people's outdoor play literature analysis First, I conducted an in depth qualitative literature analysis of empirical and review articles on childr en's and teenagers' outdoor play. In particular, I sought the features of park and green space design that are related to increase park visitation for young people, and to positive parks perceptions for young people. Also, I analyzed studies that investiga ted ethnic racial and income differences in young people's play behaviors and preferences. To analyze this literature, I developed a codebook including descriptive codes, such as the name of the authors, the year of publication, the journal, book, or u niv ersity for dissertations, and the geographical focus; topical codes, such as the scale of investigation, the methods used to measure children's and teenagers' play preferences or

PAGE 125

111 observed play behaviors, whether the entry focused on ethnic, income and gend er differences; and analytical codes, such as the green space features that children and teenagers preferred, the green space features that were observed to be more often used by children and teenagers, ethnic and gender differences (when appropriate), and other detailed findings (Richards, 2005) The full codebook is reported in Table B4 in Appendix B. To analyze the open ended codes on young people's observed play behavior s, young people's play preferences, and ethnic differences in play, I used a combination of theory driven coding (Miles & Huberman, 1994) and constant comparison analysis (Corbin & Strauss, 2007; Glaser & Strauss, 1967) as described for the analysis of the interviews transcripts. In particular, I employed the number of codes contributing to each theme to weigh the importance of each theme. Then, I used the results that emerged from this literature analysis to develop a Park Quality Index for Youth ( PQIY) which took into account park features that according to the literature matter for children's and teenagers' outdoor play and park visitation. GIS spatial analysis To measure how access to parks varies across different ethnic, income, and age group s, I conducted a Geographic Information System (GIS) spatial analysis. I used three parameters to define access to parks, including park proximity, park acreage, and park quality. These three parameters are defined in Table 7. The concurrent focus on park proximity, park acreage, and park quality provides a more comprehensive picture of access to parks than most other equity mapping studies. Also, this concurrent focus reflects the approach I used to review the equity mapping literature (see Chapter 2).

PAGE 126

112 Tab le 7 Parameters to Define Access to Parks Parameter Description Park proximity Distance to any park, regardless of size and quality (measured as DAS see below) Distance to any park or parkway (DAS) Park acreage Acres of parks within 1/4 mile of a cen sus block; (Acres of parks within 1/4 mile of a census block) / population of that 1/4 buffer; (Acres of parks within 1/4 mile of a census block) / population under 18 years of age of that buffer; and the same three variables considering acreage of parks a nd parkways Park quality Distance to parks (measured as DAS see below) with different levels of quality based on: PQIY, including five categories; parks without locally unwanted land uses (LULUS) nearby; parks with low violent crime density; and regiona l parks Analysis of park proximity To analyze access as park proximity, I measured accessibility to any park regardless of size and quality and to any park and parkway through a weighted spatial network analysis service area function in a GIS (ESRI's A rcGIS version 10.1; Wade & Sommer, 2006) In this analysis, I focus ed both on parks alone and on parks and parkways combined, as parkways have been a key element of Denver's p ark system since the early decades of the twentie th century, when Denver adopted a City Beautiful approach to guide its urban development (Etter, 1986) Denver's parkways are publicly owned, are run by Denver Parks and Recreation, and are often use d for active recreation like jogging and cycling and for passive recreation (Etter, 1986) To measure park proximity, I employed a modified version of the "minimum distance" approach introduced by Talen (2003) Talen's (2003) method created a

PAGE 127

113 dichotomous variable of access or no access to public amenities, while introduci ng potential errors because of the size and shape of certain census block groups. Building on my pilot study (Rigolon & Flohr, 2014) I advanced Talen's (2003) method by using a combination of census blocks and parce ls geometries to measure, within each census block, the parcels' average distance to the closest park, and the closest park or parkway. Only census blocks with a population different than zero were selected; thus, I excluded census blocks that have only in dustrial, commercial, or public uses. Smaller parcel geometry creates a more refined picture of access than census blocks or census block groups. Parcels' distances from parks were calculated in increments of 100 meters to make the network analysis doable. I calculated a census block's distance average score (DAS) based on the distance scores of single parcels, as shown in Equation (1). ! ! !" !"#$%&' !"#$"% !"#$%&'( ! !" ! !"#$% ! ! !"#$%&'( !"#$% ! ! ! ! ! !" !"#$%&' !"#$"% !"#$%&'( ! !" ! !"#$% ! !" !"#$%& !"#$% ! ! ! !"# ! ! !" !"#$%&' !" ! !"#$% ! ! ! ! ! ! ! ! ! !"#$%$ !"#$% !"#$%&'( !"#$!%# !"#$% ! !"# The parcels' distance score represents the distance interval from a park access point within which the parcel lies. For example, a parcel's di stance score is "1" if it is within 100 meters from a park access point, and "2" if i t is between 100 and 200 meters from an access point. Therefore, in the equation the phrases "dis tance 1" and "distance 2" stand for increments of 100 meters, for example 100 and 200 meters. The census block distance average score (DAS) represents measure of distance between census blocks and

PAGE 128

114 park access points, which approximates the average distance from parks of parcels within a census block. To fill Equation 1, I condu cted a series of steps through the software ESRI ArcMap 10.1. First, I ran a network analysis (Wade & Sommer, 2006) with the ArcMap "service area" command (A. Mitchell, 2012) using increments of 100 meters as distances from facilities. In this GIS network analysis, facilities represent the access points to the Denver's parks and parkways. Access po ints are located at the perimeter of parks and parkways in correspondence to street intersections and paths (Comber et al., 2008) Access points better represent the real distance of parks and parkways from parcels than the centroids of such spaces (Comber et al., 2008) Indeed, if choosing centroids, the shape of parks and parkways may strongly influence the level of acces s. The service areas calculated for the different distances generated segments along the city streets. Second, using the command "select by location" (A. Mitchell, 2012) I assigned distance scores representing distances to the different parcels. Third, I created a new attribute field for parcels and assign ed a constant value of "1", which I used later in the process as a count for the parcel. Fourth, I conducted a "spatial join" (Wade & Sommer, 2006) from parcels to census blocks. When doing the spatial join, I included the option "sum" for attribute field. F or each block ArcMap automatically summe d the distance scores (step 2) and the parcel counts (step 3). Fifth, I added a new attribute field to census blocks and calculated it as a ratio between the distance score and the parcel count. This field represented the parcels' distance average score to parks, and to parks and parkways, in each census block (DAS). Figure 3 shows how I assigned distance scores to parcels within census blocks and examples of census blocks distance average scores.

PAGE 129

115 Figure 3 The method to calculate the parcels' average di stance to parks for each census block I chose to use this complex way of measuring accessibility for a variety of reasons. The first group of reasons deals with the spatial scale at which demographic data are available. The U.S. Census releases most of the demographic data I am interested in (ethnicity, age groups, percent of owner vs. renter occupied, and percent of vacant housing units ) at the census block level. However, household income is released at the census group level, which is a very large spatia l unit to measure access to parks. The mean census block group size in Denver is 205.44 acres, while the mean size of the census block included in the analysis is 6.28 acres (calculated from DRCOG data) On the other hand, assessed land value is available at the parcel level for Denver (DRCOG, 2015) Assessed land value, which expresses an asset, has been used as a proxy for household income i n several studies focusing on the spatial distribution of environmental

PAGE 130

116 amenities across socio economic groups (e.g. Block, Scribner, & DeSalvo, 2004; Larson, Story, & Nelson, 2009; Mennis, 2006; Rigolon & Flohr, 2014) To obtain a measure of household income at the census block level, I integrated househo ld income at the census block group level and assessed land value at the parcel level (for residential parcels only). First, I transferred the median household income value from each census block group to the census blocks included within that census block group. Second, I calculated the average assessed land value of parcels with residential land use for each census block and for each census block group through the ArcGIS command "Spatial Join." Third, for each census block I calculated a coefficient (CB S pec) by dividing the census block's average assessed land value by the census block group's a verage assessed land value. The CB Spec coefficient is a proxy for the census blocks' income variation within the census block group in which they are contained F inally, I multiplied the CB Spec coefficient by the median household income and obtained an adjusted median household income at the census block level (see Figure 4). Figure 4. How median household income was transferred to census blocks

PAGE 131

117 The second gro up of reasons deals with the desire to define a finer measure for access to parks than measures used in previous studies. As noted earlier, most studies using the minimum distance approach (Talen, 2003) have defined access in a dichotomous way i.e., spatial units have access to public amenities if they are within a certain distanc e from them and have no access if they are outside such distance In this study, I also used a ratio measure for access, represented by the census block DAS which describes quite accurately young people's distance from parks. Walkability analysis I also calculated a child and teenager appropriate walkability index for the city streets, which I employed in the GIS network analysis for park proximity and park quality. Including a walkability index is a means to express how different streets can create barr iers or opportunities to walking for children and teenagers; and thus, how they influence their access to parks. This walkability scale is based on three street design aspects: the speed limit, as a proxy for traffic (Krizek, 2003; Napier, Brown, Werner & Gallimore, 2011; Pont, Wadley, Ziviani, & Khan, 2013; Rosenberg et al., 2009) the presence of tree canopy, as a proxy for street trees (Krizek, 2003; Larsen et al., 2009) and the presence of a sidewalk (Baran et al., 2013; Kerr et al., 2006; Krizek, 2003; Pont et al., 2013; Rosenberg et al., 2009) Speed limit, representing traffic, is an appropriate variable for a child and teenager specific walkability scale because research consistently shows that traffic danger is one of the main reasons why parents limit their children's mobility (Napier et al., 2011; Pont et al., 2013; Trapp et al., 2011; Veitch et al., 2006) Also, street trees can foster young people's wa lking behaviors because they are important elements for shading, particularly in Denver, which has an average of 300 sunny days per year (City and County of Denver n.d.) and for neighborhood aesthetics

PAGE 132

118 (Larsen et al., 2009) Finally, the presence of sidewalks can reduce parents' and young people's traffic concerns (Pont, Ziviani, Wadley, & Abbott, 2011) I assigned scores to different street segments using a combination of scores related to speed limit, presence of tree canopy, and presence of sidewalks. Lower scores represent very walkable streets, while higher scores represent streets with low walkability. Traffic level scores were as signed to different street segments using coefficients related to speed limits, as shown in Table 8. Street segments are parts of streets between two intersections. Scores for tree canopy were calculated as follows: Each street segment received a score of 1 if it had at least some tree cover along it, and a score of 1.1 if it had no trees along it. Finally, I used the presence of sidewalk as follows: Each street segment received a score of 1 if it had a sidewalk, and a score of 1.1 if it had no sidewalk. I assigned scores to each street segment and then multiplied t he scores for the three aspects to create a single walkability score. Thus, I created street segment s' adjusted lengths by multiplyi ng the walkability score of each street segment by its length (Wade & Sommer, 2006) Then, I used adjusted street segments lengths in the network analysis to better represent access to parks, and to parks and parkways, by including a child and teenager appropriate walkability measure Table 8 Speed Limit Walkability Scores Speed (MPH) 15 25 30 35 45 65 Score 0.85 0.92 1 1.05 1.15 1.35

PAGE 133

119 The output of the GIS spatial analysis for park proximity included a dataset comprising a series of demographic variables that describe socio econom ic status, ethnicity and age, a series of DAS approximating census blocks distances from parks, and from parks and parkways, and other control variables, which I define in the statistical analysis section. Analysis of park acreage To measure access as par k acreage, I calculated a series of parameters that represent acres of parks, and acres of parks and parkways, within a quarter of a census block, also including acreage measures divided by population within the quarter mile buffer, and acreage measures di vided by the population under 18 years of age within the quarter mile buffer (see Table 7). The last two measures can give an idea of potential park congestion, as exemplified by Boone et al. (2009) A quarter mile has been used in several other studies about access to parks as the c ritical distance that people are likely to walk to reach a park (Boone et al., 2009; Wolch et al., 2005) To measure park acreage, and park and parkway acreage, within a quarter mile of a census block, I created a quarter mile buffer around each census block. This basically means offsetting each census b lock's perimeter by a quarter mile. Then, using the spatial join command in ArcMap, I attributed to each census block the total acreage of parks and of parks and parkways that are within a quarter mile of each census block. Finally, for each census block I calculated park acreage divided by the total population within the quarter mile buffer, and divided by the population under 18 years of age within the quarter mile buffer. These additional variables describing park acreage (see Table 7) were added to the census blocks shapefile, which already included DAS, demographic variables, and control variables.

PAGE 134

120 Analysis of park quality The analysis of park quality includes two groups of sub analyses: The first is based on a Park Quality Index for Youth (PQIY) that I developed from the literature review on young people's outdoor play and park visitation, including a variety of categories (see Chapter 5 ); and the second is based on other parameters describing park quality as lack of locally unwanted land uses (LULUS) low violent crime density, and park types based on Denver's Parks and Recreation classification. Given the specificities of parks (playgrounds, sport fields, water elements, etc.), this analysis applies only to parks and not to parkways. For the analysi s of park quality based on the Park Quality Index for Youth (PQIY), I developed five parameters of quality (see Table 9). The first parameter is based on the overall PQIY score. Similarly to NÂŽmeth, Faulkner, and Ross (2012) I classified the par ks in the top quartile of the PQIY as "high quality," the parks in the two mid quartiles as "medium quality," and the parks in the lowest quartile as "low quality." The parks in the top quartile of the PQIY can be considered as being "good enough," or as h aving an acceptable level of quality for young people park visitation. The PQIY also includes four themes or categories, which describe features that are important for children's and teenagers' use of parks: diversity of structure play amenities; presence of natural elements; park size; and park maintenance. For the first two categories, I developed three specific quality parameters (see Table 9 and, for more details, see Chapter 5, Park Spatial Distribution Findings). The last parameter is based on the ove rall PQIY score and includes parks in the top five percent of the PQIY. This parameter, labeled as best PQIY parks, is aimed to identify the parks that, based on the PQIY, have excellent levels of quality in Denver.

PAGE 135

121 Table 9 Parameters of Park Quality base d on the Park Quality Index for Youth (PQIY) PQIY Parameter Description High quality PQIY parks Parks in the top quartile of the overall PQIY score "good enough" Parks with high quality PQIY play Parks in the top quartile of the structured play diversi ty sub score Parks with high quality PQIY nature Parks in the top quartile of the presence of natural elements sub score Parks with high quality PQIY play and nature Parks in the top quartile of the structured play diversity sub score and in the top quar tile of the presence of natural elements sub score Best PQIY parks Parks in the top 5% of the overall PQIY score "excellent" Table 10 Parameters of Park Quality not based on the Park Quality Index for Youth (PQIY) Parameter Description Parks with no LULUS Parks without locally unwanted land uses (LULUS) within 100 meters of their perimeter Parks with low violent crime density Parks in the lower tercile of violent crime density Regional parks Parks classified as "regional" by Denver's Parks and Recr eation department The analysis of park quality includes three additional quality parameters that are not based on the PQIY (see Table 10). The use of these additional parameters of quality is a means to include aspects that were not represented in the PQ IY. The first non PQIY

PAGE 136

122 parameter of quality is the absence of locally unwanted land uses (LULUS) like industries, freeways and railways in the proximity of parks (within 100 meters, which is approximately the dimension of a Denver's block). Studying the pa rks without LULUS in their proximity is important because, as Weiss et al. (2011) noted, LULUS can act as a deterrent to park visitation due to air and noise pollution. The second non PQIY parameter of park quality deals with violent crime density. Research consistently shows that the fear of violent crime strongly discourages park visitation among children and teenagers, and particularly among ethnic minorit y young people (G—mez et al., 2004; Platt, 2012; Shinew et al., 2013; Spilsbury, 2005; Stodolska et al., 2013) To include violent crime density in the analysis, I collected data on georeferenced violent crime from the City and County of Denver's GIS portal (City and County of Denver, 2015) The crime data released by the City and County of Denver spans the last five years, is updated every week and derives f rom records of the Denver Police Department (City and County of Denver, 2015) To operationalize violent crime, I used the Federal B ureau of Investigation's definition of violent crime, which includes the following crimes: "murder and nonnegligent manslaughter, forcible rape, robbery, and aggravated assault" (Federal Bureau of Investigation, 2010) To calculate violent crime density, I conducted the following s teps: First, I created a buffer of 100 meters (1 block) around each park; second I divided the number of violent crimes occurred in the last five years within the park and its 100 meter buffer by the surface of park and its 100 meter buffer. Violent crimes occurring in close proximity to parks can be a deterrent for park visitation, thus they were included in the analysis. Also, I chose to focus on violent crime density rather than on the

PAGE 137

123 number of crimes because large parks can give the opportunity to chil dren and teenagers to avoid gangs and crime (Slater et al., 2013) while small parks with frequent violent crime are likely to discourage park use. After calculating the violent crime density for each park, I classified parks in three categories: "low crime density if the score was in the lowest tercile; "mid crime density" if the score was in the middle tercile; and "high crime density" if the score was in the highest tercile. In my analysis, I focused on parks with low crime density (see Table 10). The third non PQIY parameter of park quality describes parks that Denver's Parks and Recreation Department classified as "regional parks" (Denver Parks and Recr eation, 2008) Regional parks are parks larger than 80 acres that provide the highest diversity of recreational opportunities in Denver and that receive the highest investment in terms of maintenance (Denver Parks and Recreation, 2008) For these reasons, these pa rks can be considered as having high levels of quality and their location in relation to socio economic, ethnic, and age groups should b e studied. To include elements of park quality in my GIS analysis, I repeated the same steps included in the park proximity analysis, but focusing only on the parks comprised in each of the eight quality parameters (see Tables 9 and 10). In other words, I calculated census block's DAS to the parks described in Tables 9 and 10. Statistical analysis To measure whether access to parks is equitable for different income, ethnic and age groups in Denver, I conducted two main groups of statistical tests: The fir st was intended to test whether demographic variables describing park dependency could predict access to parks in terms of park proximity and park quality. Logistic regression was chosen as an appropriate test for this goal (Leech, Barrett, & Morgan,

PAGE 138

124 2011) The second group of tests was aimed to assess differences in access to parks in terms o f park proximity, park acreage, and park quality among census blocks of different income levels and ethnic composition; and to test differences in demographic variables between census blocks with access and without access to different types of parks, with access being defined based on a quarter mile threshold (see details in the logistic regressions section). As none of these variables was normally distributed, I conducted non parametric median comparisons in IBM SPSS 22 (Kraska Miller, 2013) and other non parametric tests like the Mann Whitney U test and Kruskal Wallis test (Leech et al., 2011) Details about these t ests are reported below. In Chapter 5 (Park Spatial Distribution Findings), I present the statistical results thematically i.e., organized by par k proximity, park acreage, and park quality. Logistic regressions A series of logistic regressions were conducted to test whether vari ables describing socio economic status ethnicity, and age could predict access to parks in terms of park proximity and p ark quality. I chose to conduct logistic regression because of the statistical distribution of the demographic and park access variables described below. A s neither the demographic nor the park access variables were normally distributed, multiple regressio n could not be conducted (Leech et al., 2011) Also, as the relations between demographic and park access variables were not monotonic, Spearman's rank order correlation could not be used (Leech et al., 2011) Therefore, I chose to conduct logistic regressions by transforming the park proximity and park quality variables, which were ratio variables expressing distances from census blocks to parks, into dichotomous variables. In other words, for logistic regressions I chose to define access in terms o f park proximity and park quality in a

PAGE 139

125 dichotomous way: If a census block's average distance to the closest park was below a quarter mile, I classified that census block as having access to parks; it not I classified that census block as not having access to parks. Ten dichotomous outcome variables were created, including access to all parks and parkways, access to all parks, and access to parks with eight different quality parameters, as described in Tables 9 and 10. I chose to employ a quarter mile as a threshold to define access because, as I previously discussed, some of the most relevant studies on access to parks as an environmental justice issue used a quarter mile as the maximum distance that people in the United States might be willing to walk to r each a park (Boone et al., 2009; Talen, 2003; Weiss et al., 2011; Wolch et al., 2005) In these logistic regressions, the predictor vari ables are a series of demographic variables describing the degree to which populations are park dependent: adjusted median household income at the census block level, as described above; percentage of Non Hispanic Whites; percentage of Hispanics; percentag e of African Americans; percentage of people under 18; percentage of owner occupied units; and percentage of vacant housing units. Thus, I used logistic regression to test whether the demographic predictor variables reported above can predict the log odds of a census block being within a quarter mile from all parks and parkways, all parks, and parks with one of the eight quality parameters (see Tables 9 and 10). In particular, to exclude the variance associated with census blocks' acreage and population de nsity, or in other words to control for these two variables, I conducted hierarchical logistic regression (Leech, Barrett, & Morgan, 2014) I first entered in the logistic regression equation the census blocks' acreage and population density (step 1),

PAGE 140

126 and then I entered the variables describing socio economic status, ethnicity, and age (step 2 ). I chose to control for census blocks' acreage and population density because another study found that population density could account for some of the variance in access to parks (Boone et al., 2009) and because, given the method I used to measure a census block's average distance to parks (DAS), the size of census blocks could influence their average distance to parks. Assumptions for logistic regression, including the linear relation between the independent variables and the logit of the dependent variable (Leech et al., 2011) were tested and met. In Chapter 5, I report statistical significance levels, odd ratios and pseudo R square values describing the percentage of variance that was accou nted for. Testing differences among groups I also conducted two series of tests to compare medians among groups. First, I tested differences in park proximity, park acreage, and park quality among census blocks of different income levels and ethnic compo sition. Second, I tested differences in demographic variables between census blocks with access and without access to different types of parks, based on the quarter mile threshold. The lack of normality in the statistical distributions of the park proximit y, park acreage, park quality, and demographic variables restricted the choice of the statistical tests I could run. For example, I had to exclude independent sample t tests and one way ANOVA tests comparing means, as my data violated the assumption of nor mality (Leech et al., 2011) Therefore, I conducted non parametric median comparisons in IBM SPSS 22 (Kraska Miller, 2013) and non parametric tests comparing mean ranks like the Mann Whitney U test and Kruskal Wallis H test (Leech et al., 2011) I chose to conduct non parametric median comparisons, besides the more commonly used

PAGE 141

127 Mann Whitney U test and Kruskal Wallis H test, becau se the first allowed me to compare medians rather than mean ranks, which are harder to interpret. To compare medians and mean ranks of park access variables (park proximity, park acreage, and park quality) across income and ethnic groups, I first created groupings of census blocks based on income and ethnicity. For income, I classified each census tract by dividing the median household income in the tree terciles, so that each income group (low, medium, and high) had the same number of cases. For ethnicit y, I classified each c ensus tract based on the percentages of Non Hispanic Whites, Hispanics, and African Americans, which historically have the three main ethnic groups of Denver (Romero, 2013) To describe concentrations of ethnic groups, I used thr ee thresholds: 25 percent, 50 percent and 75 percent of population of each ethnic group within census tracts. Thus, for each ethnic group, I classified census blocks as having 25 to 50 percent of that group, 50 to 75 percent of that group, and 75 to 100 of that group, the latter describing the highest residential segregation. A similar approach to study ethnic groups has been used in relevant studies on this topic (Boone et al., 2009; Wolch et al., 2005) Through non parametric median comparisons, Mann Whitney U test s and Kruskal Wallis H test s, I asses sed the null hypothesis that the medians and mean ranks of park access variables did not differ for the three income groups, and for the various ethnic groups as described above. Table 7 reports all the park access variables, broken down by park proximity, park acreage and park quality, for which I compared medians and mean ranks. Assumptions for the non parametric medians comparison, Mann Whitney U Test, and Kruskall Wallis H Test were tested and met (Kraska Miller, 2013; Leech et al., 2011)

PAGE 142

128 The second group of comparison tests evaluates d ifferences in demographic variables, including socio economics, ethnicity, and age, between census blocks with access and without access to all parks and parkways, all parks, and the parks meeting the quality parameters described in Tables 9 and 10. As exp lained for the logistic regression, census blocks with access and no access are census blocks within and beyond a quarter mile from parks. Thus, through non parametric median comparisons, Mann Whitney U test, and Kruskal Wallis H test, I tested the null hy pothesis that the medians and mean ranks of demographic variables (socio economics, ethnicity, and age) did not differ for census blocks with access and no access to different types of parks. In Chapter 5, I report the median values and the level of statis tical significance between such medians, and the effect sizes of the mean ranks differences based on the Mann Whitney U Test results. Although I cannot make claims about statistically significant differences among mean values, I also report differences in the mean values and highlight the planning implications of differences in median and mean values, based on the statistical distribution of the park access variables. Chapter 5 also includes maps that help highlight which parts of the city are park rich and park poor in relation to where different demographic groups are concentrated. Mixed Research Integration During the data analysis process I integrated the findings of the qualitative and quantitative strands by analyzing how park planning, land use plan ning, and housing policies and practices have influenced today's park distribution in Denver in relation to where different demographic groups live, with a specific focus on high quality parks suitable for young people's play (see Figure 2). To do so, I so ught correspondences

PAGE 143

129 between specific policies and their spatial outcomes i.e., park location, zoning, or housing types in different areas, to uncover connections between procedural and distributional justice. The experts' interviews and policy document an alysis shed light on the current park distribution patterns that emerged from the GIS spatial analysis. In particular, I looked for geographical, temporal and equity related connections between Park Planning Processes and Park Spatial Distribution To in tegrate the qualitative findings of the Park Planning Processes analysis strand and the mainly quantitative findings of the Park Spatial Distribution analysis strand, I employed a series of techniques conceptualized by a few mixed research theorists (Bazeley, 2009; Caracelli & Greene, 1993; Onwuegbuzie & Teddlie, 2003) Integration is at the core of mixed rese arch studies ; as Bazeley (2009) argued "[a]ll mixed methods studies, by definition, attempt some form of integration" (p. 203). Integration of qualitative and quantitative approaches can be defined as the reciprocally beneficial connection between the two approaches, so that the results are more relevant than the sum of the two strands (Woolley, 2009) In my mixed research data analysis, I used a series of emergent techniques for data integration (Onwuegbuzie & Combs, 2010) including: data reduction and data display (Onwuegbuzie & Teddlie, 2003) ; data comparison (Onwuegbuzie & Teddlie, 2003) ; extreme case analysis (Caracelli & Greene, 1993) and data consolidation (Caracelli & Greene, 1993) Data red uction and data display involve summarizing qualitative and quantitative data through graphic representations like tables and figures (Onwuegbuzie & Teddlie, 2003) I visually represented the main findings of the two strands (park planning process and park spatial distribution) through maps, diagrams and tabl es. These graphic

PAGE 144

130 representations helped me highlight the main findings of each strand and facilitated the process of integration. In data comparison researchers compare data from various datasets (Onwuegbuzie & Teddlie, 2003) Based on the data display I created, I compared the findings from the two strands of the study, looking for specific geographic and tem poral correspondences, and for connections between equitable and inequitable processes and spatial outcomes. For example, I compared specific findings about park planning processes to the maps depicting levels of access to parks in different parts of the c ity. Also, I compared the geographic locations of parks established in a certain period with the planning instruments that the city adopted in the same period. In extreme case analysis, researchers detect extreme cases by analyzing one dataset and further examine them by investigating the other dataset, possibly collecting additional data (Caracelli & Greene, 1993) In my analy sis, extreme cases were areas of the city that did not follow the general pattern of equity or inequity in terms of access to parks that was prevalent in Denver (quantitative dominant Park Spatial Distribution analysis). To find explanations for extreme ca ses, I looked at the findings of Park Planning Process analysis (qualitative dominant). For example, a low income ethnic minority area could have good access to high quality parks because of the planning efforts undertaken by local non profit organizations Therefore, in this research extreme case analysis (Caracelli & Greene, 1993) was used as an expansion of data comparison (Onwuegbuzie & Teddlie, 2003) for some particular cases. Data consolidation (Caracelli & Greene, 1993) is the process of simultaneously examining the two datasets and merging them t o generate new or modified variables or categories, expressed in either qualitative or quantitative ways (Onw uegbuzie & Teddlie,

PAGE 145

131 2003) In my mixed methods data analysis, I used the description of park planning processes (analysis of procedural justice) to interpret the results of the logistic regression and of the comparison tests (analysis of distributional ju stice). Through this process, I developed conceptual propositions based on the combination of the two datasets (Caracelli & Gr eene, 1993) For example, some of these propositions have the following form: "Area A has less access to high quality parks than Area B because of X, Y, and Z In these propositions, Areas A and B are city neighborhoods that quantitatively differ in term s of access to high quality parks (with quality defined in relation to young people's outdoor play), and X, Y, and Z are explanations of quantitative differences, which emerged from the qualitative data. Thus, this process took advantage of the complementa ry strengths of quantitative and quantitative approaches (Greene et al., 1989) Tables summarizing these propositions and explaining how they were derived from the t wo data sets are included in Chapter 6 ( Mixed Research Integration ) Data Legitimati on Data validation, or data legitimation, deals with the quality of the data that has been collected, including their "trustworthiness, credibility, dependability, legitimation, validity, plausibility, applicability, consistency, neutrality, reliability, o bjectivity, confirmability, and/or transferability" (Collins, Onwuegbuzie, & Sutton, 2006, p. 72) In particular, legitimation evaluates the quality of meta inferences which are the conclu sions drawn from mixing qualitative and quantitative data and approaches (Onwuegbuzie & Johnson, 2006; Teddlie & Tashakkori, 2003) In mixed methods research, the word legitimation is more commonly used because it goes beyond the

PAGE 146

132 distinction between quantitative validity and qualitative trustworthiness (Onwuegbuzie & Johnson, 2006) Qualitative and quantitative data legitimation has been addressed in the previous sections. For mixed research legitimation, I employed some of the strategies described by Onwuegbuzie and Johnson (2006) Addressing mixed research legitimation is necessary to demonstrate the quality of a study's mixed research findings (Collins, Onwuegbuzie, & Johnson, 2012) Weakness minimization validation (Onwuegbuzie & Johnson, 2006) was tackled when integrating the results of the two strands of the study and interpreting the combined results. The Park Spatial Distribution analysis highlighted connections between parameters of access to parks and demographic variables, but very few plausible inferences could be made about the reasons for such associations. The results of the Park Planning Process a nalysis helped clarify some of these reasons. Conversion legitimation (Onwuegbuzie & Johnson, 2006) was addressed by counting codes in the analysis of the young people's outdoor play literature and by assessing how many code count s contribute d to each theme for the park quality score. In relation to paradigmatic mixing legitimation, I approached epistemological, ontological, axiological, methodological, and rhetorical issues that arose from combining qualitative and quantitative me thods as continua (Onwuegbuzie & Johnson, 2006) Also, using the transformative paradigm as a lens, I acknowledge that different realities and conceptions of knowledge exist depending on age, gender, ethnicity, culture social clas s, and disabilities (Mertens, 2007) In the transformative parad igm tradition, the qualitative and quantitative methodologies, and their constructivist and post positivist foundations, are

PAGE 147

133 seen as legitimate methodologies and worldviews that can be integrated to increase the understanding of socially complex realities and bring about social change (Mertens, 2007) Commensurability legitimation is the "extent to which the meta inferences made reflect a mixed worldview based on the cognitive process of Gestalt switching and integration" (Onwuegbuzie & Johnson, 2006, p. 57) In other words, this legitimation type addresses the extent to which researchers acknowledge the necessity of integrating different paradigms in mixed research (Onwuegbuzie & Johnson, 2006) The use of a transformative lens to study the reasons of today's park spatial distribution across income, ethnic and age groups, combining parts of the post positivist and constructivist paradigms, is a way to address commensurability legitimation. Political legitimati on addresses the degree to which readers of mixed research studies consider the meta inferences stemming from both the qualitative and qualitative phases legitimate, including whether power relationships are adequately represented (Onwuegbuzie & Johnson, 2006) To this regard, this study includes the perspectives about Park Planning Processes of different actors, including professionals working for the city of Denver, professionals not currently working for Denver, and scholars who are independent from Denver. In the p ark p lanning p rocesses f indings, I highlight how the perspectives on park planning, land use planning, and housing policies and practices vary among several professionals and scholars. Finally, I employed all applicabl e quantitative, qualitative and mixing legitimation strategies to address multiple validities legitimation This legitimation type deals with the degree to which quantitative validity and reliability and qualitative

PAGE 148

13 4 trustworthiness are successfully used (Onwuegbuzie & Johnson, 2006) In particular, the quantitative secondary data I used was collected through valid and reliable instruments, including the U.S. Census and American Community Survey. The secondary qualitative data I em ployed, including written sources on Denver's history and the literature on young people's outdoor play, was collected from well reputed academic sources. Also, for qualitative data analysis, audit trails (Lincoln & Guba, 1985) and intercoder reliability (Lombard et al., 2002) were employed, as explained in the prev ious sections. In the next three chapters, I present the findings of the analysis of Park Planning Processes (Chapter 4), of the analysis of Park Spatial Distribution (Chapter 5), and of the integration between processes and spatial distribution (Chapter 6 ).

PAGE 149

135 CHAPTER IV PARK PLANNING PROCESS FINDINGS Through the analysis of park planning processes, I aimed to uncover how the intersections between park planning, land use planning, and housing policies and practices help explain the processes that led to De nver's current and past spatial distribution of parks. Also, I studied the processes that contributed to the current and past residential locations of that city's different income and ethnic groups. Within the analysis of park planning processes, I identif ied four main areas of focus, which are reported in Figure 5: Denver urban history, park planning, land use planning, and housing policies and practices. This chapter is organized based on these four areas. First, I present a brief history of Denver's urba n development, which includes a timeline of major historical facts, and specific sections on the history of Denver's African Americans and Hispanics, which have been the city's main ethnic minority groups. This introduction to Denver's history serves as co ntextual information for the specific findings regarding park planning, land use planning, and housing. Second, I present an in depth analysis of how, currently and in the past, park planning has influenced park establishment and park design. This include s describing the mechanisms that the city has been using to acquire land and to develop it into parks (park establishment), and the analysis of the different values and actors that have determined the physical layout of parks and the activities that are al lowed in them (park design and management). Third, the analysis of land use planning history depicts processes related to park establishment and residential location. In terms of park establishment, I discuss the processes through which land use planning i nstruments have favored or limited the

PAGE 150

136 acquisition of land or the development of fees for parks. Also, I describe the way land use, including zoning and subdivision regulations, has contributed to the residential location of different income and ethnic gro ups. Fourth, I analyze the influence of housing policies and practices on residential segregation by income and ethnicity. In particular, I discuss a series of policies and practices that were used to limit the areas of Denver where African Americans could live, as well as a brief history of affordable housing in Denver. Finally, I explain how park planning, land use planning, and housing policies and practices have interacted in influencing park distribution across income and ethnic groups in Denver. Fig ure 5 Areas of focus for the park planning process analysis

PAGE 151

137 To understand the way park planning, land use planning, and ho u sing have contributed to park spatial distribution in relation to the residential location of different demographic groups, I rely on a series of primary a nd secondary data sources as described in Chapter 3. Table 11 summarizes the types of data I employed in each of the four areas of focus in this chapter The specific primary and secondary sources I analyzed, along with an audit tra il describing the qualitative data analysis process, can be found in Tables A2 A3 in Appendix A, and Table B3 in Appendix B. Audit trails are important to make the qualitative data analysis process more transparent (Lincoln & Guba, 1985) Table 11 Sources of Data for the Different Areas of Focus of the Park Planning Proc ess Analysis Area of Focus Sources Denver Urban History Books on Denver's history; journal articles; GIS data (urban expansion); research reports; books on Hispanics and African Americans in Colorado; historic magazine and newspaper articles; U.S. Census data (data about ethnicity over time); expert interviews Park Planning Books on Denver's parks; journal articles; research reports; GIS data (timeline of Denver's parks); expert interviews; historic magazine and newspaper articles; primary documents (pa rk maps and park plans) Land Use Planning Books on Denver's land use and zoning; journal articles; research reports; GIS data for current and past land uses; expert interviews; historic magazine and newspaper articles; primary documents (historic and cur rent land use plans). Housing Policies and Practices Books on Denver's housing; journal articles; research reports; GIS data for current affordable housing; expert interviews; historic magazine and newspaper articles; primary documents (historic and curre nt housing policies).

PAGE 152

138 The analysis of park planning processes highlighted more cases of discriminatory, exclusionary and class and color blind processes than process es o riented to increasing park equity. The latter include initiatives aime d to improve park provision for low income people, ethnic minorities, and young people The preponderance of processes led by exclusionary and economic goals in park planning, land use planning, and housing, raises serious concerns in terms of procedural injustice. In addition, economic reasons and the aesthetic values of Non Hispanic White adults have typically prevailed in park design, thus excluding the views of young people and of ethnic minority groups. An Introduction to Denver's History Denver's history spans mor e than 150 years, starting from its foundation in 1858 as a mining camp (Baker, 2004; Leonard & Noel, 1990) A thriving 21st century metropolis in the American West, Denver has gained its status for its advantageous location, near sources of water and natural resources, and for a ser ies of federal investments around World War II (Baker, 2004; Leonard & Noel, 1990) Like many cities in the United States, Denver has a history of discrimination against its ethnic minorities, particularly African Americans and Hispanics (Leonard & Noel, 1990; Romero, 2004) A brief introduction to the different phases of its history can help understand the context in which different park planning processes were undertaken. Figure 6 provides a timeline of Denver's major historical facts and of its planning and urban development history. The description of Denver's urban history is organized in a few cohesive historical periods, bounded by important historical events and characterized by significant cultural shifts. The first phase covers the period spanning from the city's

PAGE 153

139 foundation as a mining camp to the establishment of the City and County of Denver in 1902. The second phase includes the City Be autiful and New Deal eras, spanning from the early 1900s to World War II. During this period, M ayor Speer, M ayor Stapleton and their park planners created a strong system of parks and parkways. In this phase Denver also became ethnically diverse, with Afr ican Americans and Hispanics being the visible minorities in the city. During the third phase, which covers the period between World War II and 1983, Denver experienced a significant suburban growth and a reduced attention to its public sphere. This phase saw Denver losing population to its suburbs. The fourth phase, which spans from 1983 (the year when M ayor Pe–a was elected) to today, is characterized by a renewed interest in the city, as expressed by significant investment in public infrastructure, and b y economic and population growth.

PAGE 154

140 Figure 6 A timeline of Denver's history, with a particular focus on planning

PAGE 155

141 Denver's Early Stages: 1858 1902 Situated at the foothills of the Rocky Mountains by the South Platte River, Denver was founded as a camp for gold mining in 1858 (Baker, 2004; Leonard & Noel, 1990) Prior to white settlers, Arapahoe tribes had temporarily occupied the area were Denver would be founded (Leonard & Noel, 1990) The first settlement originally included three separate cities: Denver, Auraria, and Highland (Leonard & Noel, 1990) As the gold placers near Denver were exhausted, the city started providing supplies for the booming mining activities in the Rocky Mountains (Leonard & Noel, 1990) The silver boom occurring in the mountains in the 1870s boosted the economic and population growth of Denver, which also benefitted from the construction of a railway line linking Denver to the national railway network (Baker, 2004; Leonard & Noel, 1990) From the late 1860s until the 1880s, the city experienced its first construction boom, including several new subdivisions, numerous public buildings a few streetcar lines, and the first parks (Baker, 2004; Leonard & Noel, 1990) In pa rticular, Curtis Park, Denver's oldest park, was established in 1868 thanks to a land donation from developer Samuel Curtis (Leonard & Noel, 1990; Mauck, 2001) while City Park, the city's largest park, was established in 1881, as the city purchased land from the s tate of Colorado (Noel & Norgren, 1987) Denve r's economic growth translated into a significant population increase, as the city housed more than 100,000 inhabitants in 1890; at that time, the only Western city with a larger population was San Francisco (Leonard & Noel, 1990) The economic and population growth, however, came to a halt in 1893, when a crash in the silver mining industry caused an economic downturn (Leonard & Noel, 1990) Due to lack of public

PAGE 156

142 funding, Denver's first comprehensive plan for parks and parkways, the Evans Rollandet plan (1894), did not have a significant impact on park establishment (Noel & Norgren, 1987) Even during a pe riod of economic recession, Denver did not stop annexing nearby land, as the town of Highland became part of Denver in 1896 (Wiberg, 1995) In an effort to simplify administrations, the City and County of Denver was established in 1902, when the two were joined (Baker, 2004; Leonard & Noel, 1990) The City Beautiful and New Deal Eras: 1902 1945 The City Beautiful ideals inspired the work of a few mayors, park planners, and urban planners in the first four decades of the twentie th century (Noel & Norgren, 1987) The City Beautiful was an architectural and planning movement that advocated for the beautification of cities through parks, parkways and civic buildings, with the goal of inspiring positive moral values in the resid ents of cities (Noel & Norgren, 1987) Mayor Robert Speer, who served for two and a half terms between 1904 and 1918, was the biggest advocate and actuator of City Beautiful ideas in Denver (Etter, 1986; Murray, 2002; Noel & Norgren, 1987) Inspired by Burnham's Chicago World Exposition, M ayor Speer implemented a city wide vision of beautification including parks and parkways, a new civic center, mountain parks, and several other important public buildings (Murray, 2002; Noel & Norgren, 1987) To materialize his vision, M ayor Speer hired nationally renowned planners to create a comprehensive park and parkway plan: The Robinson Kessler plan, completed in 1907, influenced Denver's urban development throughout the City Beautiful era and still contributes to today's image of Denver (Etter, 1986; Noel & Norgren, 1987) Through his work M ayor Speer managed to double Denver's park acreage during his three terms

PAGE 157

143 (Leona rd & Noel, 1990) Embedded in M ayor Speer's and his planners' vision was the idea that parks and parkways should serve all parts of the city (Etter, 1986; Noel & Norgren, 1987) thus achieving a geographic equality in park distribution. The concept of geographic equality is further discussed in the park plann ing section of this chapter. Mayor Benjamin Stapleton and his planners, including Saco DeBoer and George Cranmer, continued Speer's work in the 1920s and 1930s, following the lines of the Robinson Kessler plan (Noel & Norgren, 1987) During M ayor Stapleton's terms, the city also create d its first comprehensive zoning ordinance (1925), and its first comprehensive plan in 1929 (Noel & Norgren, 1987; E. A. Taylor, 1986) As the City Beautiful era shaped Denver's historic neighborhoods, N oel and Norgren (1987) consider the years between Speer's mayorship and World War II as a golden age for Denver's public space. However, while in the City Beautiful era Denver succeeded in improving the physical image of the city, it failed to create a co hesive society where everyone could live in harmony under shared morals. The years between 1900 and the 1920s saw an increase of ethnic diversity within the city along with a growing system of ethnic discrimination (Leonard & Noel, 1990; Mauck, 2001; R. L. Simmons & Simmons, 1995) By the 1920s, the Ku Klux Klan controlled Denver's and Colorado's government (Goldberg, 1980, 1981; Mauck, 2001; Noel & Norgren, 1987; Stephens et al., 2008) Mayor Stapleton and Colorado governor Clarence Marley were both members of the Klan (Leonard & Noel, 1990; Stephens et al., 2008) According to Goldberg (1980) the Klan managed to gain significant political weight in Denver because dominant Non Hispanic White protestant

PAGE 158

144 groups considered the Klan as a strong a nchor to contrast the increase of ethnic diversity, including African Americans, Hispanics, Catholics, and Jews. From the 1910s to the 1950s, a de facto system of ethnic discrimination was in place in Denver, targeting mostly African Americans and to some degree the Japanese and Hispanics (Abbott, 1978; Leonard & Noel, 1990; Newsum, 2012) By the 1910s, ethnic minorities faced discrimination in public recreation (parks, po ols, tennis courts and golf courses), in private businesses (restaurants, shops, and theaters), and in housing (Abbott, 1978; Leonard & Noel, 1990; Mauck, 2001 ; Newsum, 2012) For example, in the 1910s African Americans and the Japanese were excluded from the beaches in the city's public parks (Newsum, 2012) Also, most theate rs in the city forced African Americans to sit in the balcony (King, 2011; Leonard & Noel, 1990; R. L. Simmons & Simmons, 1995) According to several historical accounts, episodes of ethnic discrimination in public recreation, private businesses, and housing lasted until at least the 1950s (Leonard & Noel, 1990; T. H. Simmons & Simmons, 20 10) and its effects persist today The specific forms of ethnic discrimination in housing are described in the housing policies and practices section of this chapter. Ethnic diversity and residential segregation During the City Beautiful and New Deal er as, Denver's population started to become more ethnically diverse, with an increased presence of African Americans, Hispanics, Irish, Italians, and Jews (Abbott, 1978; Goldberg, 1980; Leonard & Noel, 1990) In this section, I present the history of the two ethnic minorities that have had the most important role in Denver's history : African Americans and Hispanics (Romero, 2003) The focus on these two ethnic minorit y groups also refl ects Denver being a tri ethnic city i.e., a city in which the ethnic dynamics do not

PAGE 159

145 occur along a dichotomous white black line but involve complex interactions between Non Hispanic Whites, African Americans, and Hispanics (Romero, 2003) While Denver's Hispanic population was about three times as large as Denver's African American population in 2013 (United States Census Bureau, 2013b) African Americans started moving to the Denver earlier than Hispanics, around the late 1870s and early 1880s (Abbott, 1978; Junne, Ofoaku, Corman, & Reinsvold, 2011; R. L. Simmons & Simmons, 1995) Hispanic Coloradoans were initially rural workers and their presence in Denver in the 1920s was still very small (Abbott, 1978) On the other hand, around three fifths of Colorado's African American population already lived in Denver in the 1920s (Abbott, 1978) While both ethnic minorities have historically been in the lower income groups among Denver's residents, African Americans were better off than Hispanics in th e 1940s, as they on average had higher incomes, more home ownership, and less infant mortality (Abbott, 1978) This slight advantage can be attributed to African Americans having a longer established history in the city than Hispanics (Abbott, 1978) For simplicity purposes, the presentation of African Amer ican and Hispanic history in Denver will span across the four main phases of Denver history that I introduced at the beginning of this chapter. Also, in this discussion, I do not cover specific aspects of park provision, which are discussed in the park pla nning section. African Americans in Denver The history of African Americans in Denver is often associated with Five Points, a neighborhood northeast of downtown that has been considered Denver's Harlem (Mauck, 20 01; R. L. Simmons & Simmons, 1995) African American history in Denver has also strong connotations of racial discrimination,

PAGE 160

146 spanning housing, education, employment and recreation (Leonard & Noel, 1990; Romero, 2004; Stephens et al., 2008) Before Five Poin ts became Denver's Harlem, African Americans in the 1880s and 1890s were not concentrated or segregated in a specific area of the city and many African American Denverites were part of the middle class (Junne et al., 2011; Mauck, 2001; R. L. Simmons & Simmons, 1995) Five Points: Denver's Harlem Several factors contributed to Five Points becoming the heart of the African American community around the late 1890s and early 1990s, including: the proximity to factories and the railway, which provided workplaces for several African Americans (Mauck, 2001; R. L. Simmons & Simmons, 1995) ; the availability of fairly inexpensive housing stock, as the Non Hispanic White affluent left the area due to the construction of the railway and industries (Mau ck, 2001; R. L. Simmons & Simmons, 1995) ; and consequently the establishment of churches and private schools for the growing African American community (R. L. Simmons & Simmons, 1995) The mechanisms that led to the forma tion of Five Points as an African American enclave in Denver can be seen in many other U.S. cities (R. L. Simmons & Simmons, 1995) : As cities expand, the establishment of industrial areas and other locally unwanted land use s caused property values of some highly regarded residential areas to decrease; this led to low income people moving to formerly desirable residential areas located next to unwanted land use s In the Curtis Park area, a subdivision that is part of the larg er Five Points neighborhood, large mansions were transformed into boarding houses after Non Hispanic White wealthy Denverites left the area, and African Americans started to move to the area (Leonard & Noel, 1990; Mauck, 2001; R. L. Simmons & Simmons, 1995)

PAGE 161

147 This shows interesting intersection between land use patterns, hou sing, ethnic history, and parks As the concentration of African Americans in Five Points increased and their population started to grow (Mauck, 2001; R. L. Simmons & Si mmons, 1995) Non Hispanic White middle class neighborhoods located nearby Five Points started establishing racially restrictive covenants to prevent the growing African American population from expanding to other neighborhoods (Leonard & Noel, 1990; Mauck, 2001) In 1916, the Af rican American community managed to stop a proposed law that would have established de jure residential segregation by race in Denver (R. L. Simmons & Simmons, 1995) Segregation ordinances were in place in other American c ities at that time, including Baltimore (Boone et al., 2009) and the nat ion's capital Although a segregat ion ordinance was never established in Denver, racially restrictive covenants, redlining, intimidation actions, and real estate practices contributed to African Americans being confined in Five Points and some nearby areas until at least the 1950s (Leonard & Noel, 1990; Mauck, 2001; Newsum, 2012; R. L. Simmons & Simmons, 1995) And it was not until ratification of the 1968 U.S. Fair Housing Act that they had a legal tool to contest discriminatory practices. Although confined in Five Points, Denver's African American population continued to grow in the 1920s, with African American families migrating from the South and the Northeast to seek a better life in the American West (R. L. Simmons & Simmons, 1995; Snow, 2009) These migration s contributed to a further increase of Five Points' population and to African Americans being increasingly segregated residentially and by school (Abbott, 1978; R. L. Simmons & Simmons, 1995) In the 1920s, around 90

PAGE 162

148 percent of African American children attended five elementary schools located around Five Points, and very few African American teenagers made it to high school (Leonard & Noel, 1990) Although nearby neighborhood improvement associations were trying to keep blacks confined in Five Points through racially restrictive covenants (Leonard & Noel, 1990) African Americans obtained a small victory in terms of residential choi ce s and managed to settle in the western portion of Whittier, an elegant streetcar suburb just east of Five Points (Leonard & Noel, 1990; Snow, 2009) Seeking more housing than Five Points could provide, some African Americans were allowed, by covenant and custom, to live in fairly restricted area of Whittier located west of Race Street (Leonard & Noel, 1990; Raughton, 2000) as shown in Figure 7. The "color line," an invisible line that de facto separated the areas where African Americans were allowed to live from areas th at were off limits for them, was moved a few times in the 1920s, fluctuating between Race Street and High Street (Leonard & Noel, 1990; Raughton, 2000) African Americans who crossed the color line were targeted by violence (Leonard & Noel, 1990; Raughton, 2000)

PAGE 163

149 Figure 7. Residential locations of African Americans in the twentie th century. Elaborated based on maps and data from (Abbott, 1978; Leonard & Noel, 1990; Newsum, 2012; R. L. Si mmons & Simmons, 1995) Between the 1920s and the 1940s, Welton Street between 22nd and 29nd Streets became Five Points' main street and the center of African American life in Denver (Mauck, 2001; R. L. Simmons & Simmons, 1995; Stephens et al., 2008) As African Americans were segregated in Five Points, Welton Street became a service provider for the g rowing African American community, with jobs, educational opportunities, and entertainment (Mauck, 2 001; Stephens et al., 2008) While African Americans were still mostly concentrated in Five Points and parts of Whittier in the 1940s (King, 2011) the post World War II period saw a few demographic and residential location change s (Leonard & Noel, 1990) With Denve r's African American population doubling between 1940 and 1950 ( then at 2 percent of

PAGE 164

150 Denver's population) and with general economic conditions improving in the city, African Americans managed to move the color line further to the east, around York Street (Leonard & Noel, 1990) Thus, neighbo rhoods like Whittier (entire), Cole, and City Park West were places where African Americans could live (see Figure 7). In the mid 1950s, a few wealthier African Americans move d east of York Street (Leonard & Noel, 1990) to the neighborhoods that today are known as Clayton and Skyland (see F igure 7). However, the majority of African Americans were still concentrated in Five Points, Whittier, City Park West, and Cole (Leonard & Noel, 1990) The expansion of African American's geographical domain eastbound, outside of the initial ghetto in Five Points, did not happen for equity concerns arising in Denver's population, but it was the effect of real estate practices that, by design, took advantage of Non Hispanic White homeowners' fears of ethnic integration (Cole, 2014) To start the process of blockbusting realtors sold or rented homes to African American families in traditionally Non Hispanic White areas, in order to create panic among Non Hispanic White homeowners (Cole, 2014) Capitalizing on their fear of African Americans lowering property values of nearby homes, realtors recommended that Non Hispanic White homeowners s ell their houses at very low price s (Cole, 2014) Then, realtors rented or sold the same houses to African Americans at increased costs (Cole, 2014) African Americans to Park Hill If Five Points is considered the center of the early African American history in Denver (Mauck, 2001) Northeast Park Hill, located east of City Park, is the other neighborhood that is most commonly associated with African Americans in Denver, and which still currently hosts a predominantly African American community (Goldstein, 2012) The beginning of African Americans' history in Park Hill

PAGE 165

151 happened by design due to blockbusting practices In 1949, City Council approved, among outraged protests, the Cavalier subdivision, a small subdivision located in Park H ill at Dahlia Street and East 35th Avenue that was advertised as a residential area exclusively for African Americans (Leonard & Noel, 1990; Newsum, 2012; T. H. Simmons & Simmons, 2010) Some opponents to this subdivision argued that opening Park Hill to African Americans would lower the neighborhood's property values (Newsum, 2012) The Cavalier subdivision fostered another case of panic selling. As African Americans started living east of Colorado Boulevard in Northeast Park Hill (see Figure 7), many realtors suggested that Non Hispanic White homeowners in North and Northeast Park Hill to sell their properties (Cole, 2014; Newsum, 2012) Besides the effect of the Cavalier subdivision, North and Northeast Park Hill were becoming less desirable places to live for Non Hispa nic White middle class families in the 1950s due to increased traffic at the Stapleton A irport, located next to these neighborhoods, and for the relatively small size of their housing stock (Cole, 2014) While North and Northeast Park Hill included quite modest small house s and an increasing African American population in the 1950s, the southern areas of Park Hill, today named South Park Hill, were, and still are, among the most prestigious neighborhoods of Denver (Cole, 2014; Leonard & Noel, 1990) With African Americans moving to the northern park of Park Hill and becoming the larger ethnic group in the area by 1970 (Schmidt & Lee, 1978) several efforts were made to create a partially racially integrated community (Dorward, 2010; Leonard & Noel, 1990; Newsum, 2012; Woods, 1998) Local churches and the Park Hill Action Committee (PHAC) attempt ed to build bridges between the newcomer African Americans in the northern part of the

PAGE 166

152 neighborhood, and the old timer Non Hispanic Whites living in the sout h side, and to reassure the latter who were considering leaving the neighborhood (Dorward, 2010; Leonard & Noel, 1990; Newsum, 2012; Woods, 1998) Although these attempts to facilitate racia l integration were unprecedented in Denver in the 1950s and 1960s (Leonard & Noel, 1990) and reflected ethos of the national Civil Rights Movement the goal of preventing massive white flight to the suburbs was also motivated by economic reasons, as the expansion of African Americans to Sout h Park Hill would have meant a significant decrease in property values (Cole, 2014) By the 1970s, the work of the Park Hill Action Committee contributed to turn the Gr eater Park Hill area into a partially integrated community (Leonard & Noel, 1990) Integration, however, was not perfect, as the neighborhood has been divided in three main areas, from the south to the extreme north (Cole, 2014; Leonard & Noel, 1990; Woods, 1998) The northernmost part, today named Northeast Park Hill, has been almost completely African American and low income, while the southernmost part, today called South Park Hill, has been predominantly Non Hispanic White and upper class (Cole, 2014; Leonard & Noel, 1990; Woods, 1998) The only partially integrated part has been the neighborhood between the two extremities, today named North Park Hill (Cole, 2014) While in the 1960s African Americans were allowed to live in North and North east Park Hill, north of East 26th Avenue, there were few other places in the Denver metro area where rea ltors would se ll homes to them (Leonard & Noel, 199 0) There fore, Park Hill north of East 26th Avenue had become just another ghetto, as real estate agents allowed African Americans to live in Park Hill to find relief from the dense

PAGE 167

153 living conditions of Five Points, but did not allow African Americans to freely decide where to live (Leonard & Noel, 1990) African Americans from the 1970s to today Between 1960 and 1970, African Americans grew at a higher rate than Non Hispanic Whites (Massey, 1983) and constituted approximately 8 percent of Denver's population in 1970 (Logan, Xu, & Stults, 2014) With the expansion from Five Points to Park Hill, by 1970 both central and northeast Denver were predominantly African American (Abbott, 1977; Bacon, 1975; Leonard & Noel, 1990) In the 1970s, many African Americans moved to Montbello, a new subdivision located in far northeast Denver, wh ich welcomed every ethnic group and included moderately low priced houses (Dwyer, 1993; Leonard & Noel, 1990) In the same years, more African Americans were able to move to suburban communities outside Denver's city limits (Leonard & Noel, 1990) However, they were still concentrated in mainly African American neighborhoods (Leonard & Noel, 1990) By the 1980s, historically African American neighborhoods like Five Points, Whittier, City Park West, and Cole still had percentages of African American population above 40 percen t, although the presence of H ispanics was increasing in these areas (T. A. Clark, 1985) In 1980 and 1990, African Americans were still predominant ly located in northeast Denver, although they were less residentially segregated than in 1970 (Dwyer, 1993) In 1990, 57 percent of the African American populatio n of Denver lived in neighborhoods in which the prevalent ethnic group was African American (Dwyer, 1993) Montbello and Green Valley Ranch became areas with high concentrations of African Americans in the far northeast (Dwyer, 1993) In other words, the far east part of the city was fairly open to African Americans. Howev er, these areas were located far from

PAGE 168

154 downtown and from other large employment centers, near large infrastructures like interstate highways, and near industrial areas (see Figure 7). Therefore, they were not the most desirable areas for Non Hispanic White middle class families. The last available demographic data from the American Community Survey shows that African Americans constituted 10.2 percent of Denver's population in 2013 (United States Census Bureau, 2013b) African Americans are still concentrated in northeast Denver, although some historically African American neighborhoods like Five Points, Whittier, and Cole have experienced an in crease in Hispanic population around the 1990s (Logan et al., 2014) and a strong w ave of gentrification from the second half of the 2000s (J. Koehler, 2014) Figure 8 shows the current residential locations of African Americans in Denver. Figure 8 Spatial distribution of African Americans in 2013 (U.S. Census data)

PAGE 169

155 Hispanics in Denver Hispanics, initially identified by demographers as Mexican Americans, were originally rural workers in the Colorado countryside and t heir presence in Denver was still limited in the 1920s (Abbott, 1978) While many accounts report a strong Mexican component in the Hispanic population of Colorado (Abbott, 1978; Langegger, 2013) Hispanics have be come a very complex ethnic group in terms of geographic origin, race, and self identification. Spanish surnames were often used by local and national demographers to identify Spanish speaking people as a distinct e thnic group (Abbott, 1977, 1978) According to two scholars I interviewed the history of Hispanics in Denver does not have the same strong connotations of discrimination as its African American history. Another historian I interviewed claimed that "around the 19 30s, Latinos called themselves Spanish Americans as a means to be identified as of white race and limit discrimination." This further reinforces the need of studying the history of Hispanics and African Americans in Denver separately, in line with the tri ethnic character of Denver (Romero, 2003) Hispanic population growth, 1940s In the 1920s, Hispanics started living in Denver during the winter months, as they were off from working on the beet fields (T. H. Simmons & Simmons, 2010) Specifically, most Hispanics found their temporary winter homes in the western section of Five Points (T. H. Simmons & Simmons, 2010 ) From the early 1920s to 1940, the Hispanic population of Denver increased from 2,000 to 12,000 people (Abbott, 1978) which constituted approximately 4 percent of Denver's population in 1940 (Dur‡n, 2011) During the 1940s, Hispanics were mostly concentrated in a few census tracts along the South Platte River northeast and southwest of downtown (Dur‡n, 2011; R. L.

PAGE 170

156 Simmons & Simmons, 1995) as shown in Figure 9. In particular, Hispanics mostly occupied the neighborhoods that today are named Auraria, Sun Valley, La Alma/Lincoln Park and portions of the Five Points neighborhood (Dur‡n, 2011; R. L. Simmons & Sim mons, 1995) Almost half of Denver's Hispanics in the 1940s lived in the Five Points neighborhood (R. L. Simmons & Simmons, 1995) It is important to note that today's Five Points statistical neighborhood is much larger th an what historically was identified as Five Points, which is the oddly shaped intersection between Welton Street, Washington Street, East 27th Avenue, and 26th Street (R. L. Simmons & Simmons, 1995) A study on the status o f ethnic minorities in Denver in the 1940s reported that Hispanics lived in very poor housing conditions during these years (Dur‡n, 2011) Figure 9 Residential locations of Hispanics in the twentie th century. Elaborated based on maps an d data from (Abbott, 1978; Bacon, 1975; Diaz, 2005; Dur‡n, 2011; Dwyer, 1993; Goldstein, 2011; Leonard & Noel, 1990; R. L. Simmons & Simmons, 1995)

PAGE 171

157 According to one historian I interviewed, the reasons that led to residential segregation for Hispanics in Denver were different than the ones that led to the formation of African American ghettos. In particular, the historian suggested that "Hispanics were interested in creating closely knitted communities where residents spoke Spanish and where their kids could find Spanish speaking s chools." This suggest s that, as an ethnic minority, Hispanics have generally different goals in terms of residential location than African Americans, who aimed to be more integrated with the rest of society. Denver's Hispanic communities flourished around a few Roman Catholic churches built in the 1920s and 1930s in Auraria and Highland, including St. Cajetan's and Our Lady of Guadalupe (Leonard & Noel, 1990) These local churches contributed to the development of a sense of community among Denver's Hispanic population (Leonard & Noel, 1990) Whatever the reasons for residentia l segregation, my analysis of U S census data and of evidence reported in a series of studies showed that Hispanics have been increasingly concentrating in west and north Denver, with their residential segregation from other ethnic groups increasing in th e last four decades (see Figure 9). Hispanics in Northwestern Denver. Besides inhabiting the areas along the South Platte River corridor, Hispanics started concentrating in northwester Denver in the late 1940s and early 1950s (Goldstein, 2011; Leonard & Noel, 1990; Wiberg, 1995) mostly in the neighborhoods that today are named Highland and Sunnyside (see Figure 9). The expansion of Hispanic s to northwest Denver can be linked to a sign ificant increase in Denver's Hispanic population between 1940 and 1950. In that decade, Denver's Hispanics almost doubled, constituting approximately 10 percent of the city's population in 1950 (Dur‡n, 2011; Leonard & Noel, 1990) Northwest Denver had experienced

PAGE 172

158 several cycles of ethnic minorities inhabitin g the area, and then leaving when reaching better economic status, including Irish and Italian s (Goldstein, 2011; Lan gegger, 2012) The Chicano Movement and Denver's Crusade for Justice To understand the evolution of Hispanic communities in Denver in the 1960s and 1970s, it is important to introduce the national Chicano Movement which laid the foundation for Denver's Crusade for Justice (Langegger, 2013) T he Chicano Movement of the 1960s (Langegger, 2013; Vigil, 1999) was grounded in two main ideas (Langegger, 2013) The first deals with the Hispanic pop ulations of the southwestern United States and northwestern Mexico being colonized people, which saw the border moving across them (Langegger, 2013) Chicanos fel t attachment to their land, which they refer red to as Aztl‡n, and one of the goals of the Chicano movement wa s the reacquisition of their land (Langegger, 2013) Second, the Chicano Movement wa s grounded in the wider Civil Rights movement that developed in the United States in the 1960s, which claimed that the work of ethnic minorities has only benefitted the Non Hispanic White affluent prevailing groups (Langegger, 2013) The Crusade for Justice, which contributed to creating a strong sense of ethnic identification among Denver's Chicano youth, developed within the ethos of the Chicano Movement. Led by former professional boxer Rodolfo "Corky" Gonzales, it was a cultural and political movement in Denver that aimed to achieve social justice for Hispanics, or Chicanos, as the members of the Crusade called themselves (Vigil, 1999) The Crusade sparked from police brutality and racism affecting Hispanics in Colorado in this era (Vigil, 1999) These pushes for social and environmental justice are important in

PAGE 173

159 understand ing park planning and management in Denver's Hispanic neighborhoods from the 1960s to today (see park planning section). Residential location 19 60s 1970s Denver's Hispanic population continued to grow between 1960 and 1970, with an increase rate higher than 50 percent (Bacon, 1975; Massey, 1983) By 1970, Hispanics constituted 17 percent of Denver's population (Bacon, 1975) The patterns of residential segregation between Hispanics, African Americans, and Non Hispanic Whites in the 1960s and 1970s show some interesting trends. While until 1960 Hispanics and African Americans were q uite integrated in some low income neighborhoods, between 1960 and 1970 they separate d with ethnic segregation between the two groups increasing (Bacon, 1975) Also, in the 1960s Hispanics took over low income residential areas along the South Platte River in northeast Denver that successful African Americans and Non Hispanic Whites had vacated (Abbott, 1978) Another interesting trend is that Hispanics and Non Hispanic Whites were relatively more integrated than African Americans and Non Hispanic Whites (Abbott, 1978) Hispanics were generally more accepted than African Americans when they started living in Non Hispanic White working class neighborhoods in northern and western Denver (Abbott, 1978) According to Abbott (1978) this is partially due to Hispanics slowly assimilating to the dominant Non Hispanic White culture. Hispanics in the 1960s and 1970s in Colorado can be considered as second generation immigrants, and they were not a minority group with a long history of racial discrimination like African Americans (Abbott, 1978) However, as immigration from Latin American countries has continued, new wa ves of Hispanics kept moving to Colorado (Immigration Policy Center,

PAGE 174

160 2013) thus contr ibuting to a very multifaceted ethnic group in terms of language, national origin, and income (United States Census Bureau, 2013b) Although the residential segregation index between Hispanics and Non Hispanic Whites was not as high as the one between African Americans and Non Hispanic Whites ( Abbott, 1978) in the 1960s and 1970s Hispanics were strongly concentrated in northern and western Denver (see Figure 9), which have been known as the North Side and West Side (Abbott, 1978; Bacon, 1975; Diaz, 2005) The area known as West Side includes a series of neighborhoods that today are named Vill a Park, Barnum, West Barnum, Valverde, Athmar Park, Ruby Hill, Westwood, and Mar Lee, while the North Side includes Sloan's Lake, Highland, West Highland, Sunnyside, Globeville, Chaffee Park, Jefferson Park, Regis, and Berkeley (see Figure 9). In 1970 the areas in which Hispanics were the majority of the population included Highland, West Highland, Sun Valley, La Alma/Lincoln Park, and West Colfax (Dwyer, 1993) In particular, the areas with higher concentrations of Hispanics in the North and West Sides, which have been considered Denver's barrios, have retain ed their ethnic character and have been the places where the most recent Hispanic immigrants and the poorest Hispanics live (Diaz, 2005) The West Side, regardless of ethnicity, was one of the poorest areas of Denver in the 1970s (T. A. Clark, 1985) Residential location 1980s 1990s In the 1980s and 1990s, the trends observed in the previous decades continued, with Hispanics becoming further spatially concentrated in the N orth and West Sides, with an increase in northeast Denver in the 1990s (Dwyer, 1993) In the same period, the index of residential segregation for Hispanics signi ficantly increased (Dwyer, 1993) Starting in the 1980s, some areas of the West Side have

PAGE 175

161 includ ed the highest percentages of Hispanics in Denver, followed by Hig hland and Sunnyside in northwestern Denver (Diaz, 2005; Dwyer, 1993) In particular, the number of neighborhoods in which Hispanics were the majority of the population increased between 1980 and 1990, encompassing in 1990 a large part of the city in the North and West Sides, including Highland, Sunnyside, Berkeley, La Alma/Lincoln Park, Barnum, Villa Park, and Westwood (Dwyer, 1993) Besides being concentrated in the North and West Sides, Hispanics significantly expanded to northeast Denver, as they constituted the largest ethnic group of Five Points i n 1990 (R. L. Simmons & Simmons, 1995) Finally, Denver's Hispanic population rose from approximately 17 percent of the city population in 1970 to 22.5 percent in 1990 (Dwyer, 1993) thus nearing one quarter of Denver's population. Residential location 2000s 2015s. The most recent data from the American Community Survey shows that Hispanics comprised 30.9 percent of Denver's popula tion in 2013 (United States Census Bureau, 2013b) As Figure 10 shows, Hispanics are still concentrated in the West Side and in some areas o f northwestern and northeastern Denver. However, some neighborhoods in which Hispanics were the majority in the early 2000s like Highland, Sunnyside, Baker, and La Alma/Lincoln Park have experienced a strong wave of gentrification starting in the second ha lf of the 2000s (J. Koehler, 2014; Langegger, 2013) In particular, while Highland was still predominantly Hispanic until the early 2000s, in the second half of the 2000s the area has become one of the trendiest parts of Denver, including expensive condomi niums and gourmet restaurants (Langegger, 2012, 2013)

PAGE 176

162 Figur e 10 Spatial distribution of Hispanics in 2013 (U.S. Census data) Suburbanization: 1945 1983 The third macro phase of Denver's history covers the period between World War II and 1983, when Federico Pe–a became Denver's first Hispanic m ayor. During this p hase, Denver lost significant weight in its metro area, as its suburbs experienced a substantial population and geographic growth (Abbott, 1977, 1978) Also, after public space had been the focus of several mayors during the City Be autiful and New Deal eras, the period after World War II saw the city reducing its fun ding for pu blic space and focusing its efforts on automobile driven horizontal growth (Noel & Norgren, 1987) Mayor Newton and Denver Commission on Human Relations The first significant political fact of the post World War II era in Denver was the election of M ayor J. Quigg Newton in 1947 (Baker, 2004; Dur‡n, 2011; Leonard & Noel, 1990) The

PAGE 177

163 election of Newton meant a political shift from Ku Klux Klan driven politics to policies and actions more oriented to social equity (Dur‡n, 2011) Concerned with the conditions of Denver's ethnic minorities, mayor Newton established the Denver Commission on Human Relations (DCHR) with the goal to study the relations between ethnic groups in Denver ( Abbott, 1978; Baker, 2004; Dur‡n, 2011; Leonard & Noel, 1990; Romero, 2003) The study commissioned by Newton found large evidence of ethnic discrimination against African Americans, Hispanics, Japanese, and Jews (Abbott, 1978; Dur‡n, 2011; Leonard & Noel, 1990; Romero, 2003) In particular, DCHR found that job discrimination, including public employment (Leonard & Noel, 1990) was the start of most problems for ethnic minorities (Dur‡n, 2011) As ethnic minorities earned less than Non Hispanic Whites, they were confined to very poor areas of the city, which were the ones with low land value (Dur‡n, 2011) These areas soon becam e overcrowded and homes and public buildings quickly deteriorated (Dur‡n, 2011) The DCHR report called out Non Hispanic White Denverites for establishing "a barbed wire of prejudice" against any ethnic minority (Romero, 2003, p. 849) The report also found that, besides job discrimination and res idential segregation, ethnic minorities also experienced discrimination in private businesses (Abbott, 1978; Leonard & Noel, 1990) "health care, recreation, schooling, [and] law enforcement" (Abbott, 1978, p. 255) Besides the establishment of the DCHR, Mayor Newton undertook other initiatives oriented to social equity includi ng the construction of a variety of public housing projects across the city (Baker, 2004) and municipal ordinances that prohibited ethnic discriminations in swimming pools and other recreational facilities (Mauck, 2001) However, in the late 1940s, Denver still expressed evident signs of ethic discrimination.

PAGE 178

164 A poll by the Denver Post dating 1947 showed that the vast majority of Denver's residents were in favor of resident ial segregation: 72 percent of respondents believed that African Americans did not have the right to live wherever they wanted, and this percentage was around 61 62 percent for Hispanics and Asians (Newsum, 2012) Several other episodes of widespread ethnic discrimination we re documented throughout the 1950s and 1960s (Dur‡n, 2011; Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010) Suburban development As discussed, in the 1940s the City and County of Denver started losing population to its suburbs (Abbott, 1977; Kurtz, 1957) Denver's Planning Office director Maxine Kurtz expressed concerns about this rate of suburbanization and named this phenomenon "suburbanitis" (Kurtz, 1957, p. 58) Maxine Kurtz and the Denver Planning Office at large were concerned about suburbanization because they feared that white flight to the suburbs could cause significant decreases in tax revenues for Denver (Cole, 2014) These concerns pa rtially drove the drafting of a new zoning code (1956), which promoted a suburban oriented vision of Denver to appeal Non Hispanic White middle and upper class families who were moving to the suburbs (Cole, 2014) More details about this code are provided in the land use planning section. The predominance of s uburban development and car oriented environments also meant a decline in Denver's care for its pedestrian orien ted public space, including parks. As most other American cities, after World War II Denver emphasized private spaces and partially neglected its central city and its City Beautiful tradition (Noel & Norgren, 1987) According to Helen Arndt, a member of the planning board in the late 1 950s and 1960s, after WWII Denver lost its focus on beauty and let the very fast economic development

PAGE 179

165 dictate its future, losing its identity (Leonard & Noel, 1990) This perceived disinterest in public space is particularly striking given the city's City Beautiful tradition (Leonard & Noel, 1990) Economic development: Oil an d gas industry Denver's proximity to the Rocky Mountains attracted several oil and gas companies that opened branches in the city after World War II (Baker, 2004) When oil and gas prices increased in the 1970s, th ese companies flourished and drove a construction boom in downtown Denver (Baker, 2004; Leonard & Noel, 1990; Murray, 2002) The De nver Urban Renewal Authority (DURA) led several urban renewal projects downtown, including the new higher education Auraria campus (Denver Urban Renewal Authority, 2008; Murray, 2002) By the late 1970s and early 1980s, several skyscrapers were built in downto wn Denver ; thanks to the expansion of the oil and gas industry, Denver's skyline drastically changed (Baker, 2004; Leonard & Noel, 1 990; Murray, 2002) School segregation and the Keyes case During this period of economic growth and significant urban development, Denver was still experiencing harsh conflicts among ethnic groups, which dealt with school segregation issues. Given the pa tterns of residential segregation and the school attendance boundaries, Denver Public Schools were strongly segregated by ethnicity in the 1950s (Romero, 2004) With the Brown v Board of Education decision in 1954, the pressure was on to desegregate Denver schools (Romero, 2004) The 1950s and the 1960s saw fights between segregationists (mostly No n Hispanic White middle and upper class people) and integrationists (ethnic minorities) over school attendance boundaries and school policies (Romero, 2004) The history of school segregation and desegregation between the 1950s and the 1970s

PAGE 180

166 deserves particular attention because it shows that Denver was still a strongly divided city along color lines, with Non Hispanic Whites fiercely battling to maintain ethnic segregation, and for its indirect consequences on park establishment. With demographic migrations happening in northeast Denver in the 1950s and early 1960s, Denver Public Schools redrew school attendance boundaries every year to keep schools racially segregated, or to create new segregated schools (Leonard & Noel, 1990; Romero, 2004) The case of Barrett Elementary School is a clear example of how school boundaries were drawn to perpetuate racial segregation. Barrett, located in the Clayton neighborhood, wa s established in 1960 with the goal of enrolling African American children living in northeast Denver (Connery, 2013; Leonard & Noel, 1990; Romero, 2004) The attendance boundaries of Barrett Elementary extended east of Colorado Boulevard, an arterial street with heavy traffic, into North and Northeast Park Hill, where Park Hill's African American population was increasing (Connery, 2013; Leonard & Noel, 1990) In other words, Barre tt Elementary and its attendance lines had the effect of drawing away Park Hill's African American children from Park Hill Elementary, which after 1960 became almost completely Non Hispanic White (Connery, 2013; Leonard & Noel, 1990; Newsum, 2012; Romero, 2004) According to one historian I interview, the practice of redrawing school boundaries reinforced what real estate agents and lending companies were already doing to promote racial segregation," and "was done in direct respon se to migration through the city of African Americans, especially in northeast Denver." The same historian noted that the attendance lines for the two main high school in east Denver managed to segregate by design African American s and Non Hispanic Whites, with the first being assigned to a

PAGE 181

167 technical school (Manual High School) and the latter being assigned to a college preparatory school (East High School). The yearly manipulation of attendance boundaries ended in 1965, with Rache l Noel, an African American woman, becoming part of Denver Public Schools' Board of Education (Connery, 2013; Romero, 2013) In the late 1960s and 1970s the tensions between separationists and integrationists became even stronger (Moran, 1992; Romero, 2004) Integrationists advocated for school busing that would have transported et hnic minority students to schools located in Non Hispanic White and more affluent neighborhoods, thus providing ethnic minority students with a better education (Romero, 2004) In the 1960s, meetings to discuss school attendance boundary changes drew large crowds of parents, with Non Hispanic White mid dle and upper class parents often advocating against boundary changes that would have integrated schools, and succeeding (Moran, 1992) The first attempts of school busing, in the mid 1960s, fueled additional tensions (Romero, 2004) The Keyes case marked a turning point in the battle between separationists and integrationists. In 1969, Wilfred Keyes and a few other African American parents filed a law suit agains t Denver Public Schools, accusing the district of having designed and perpetuated a system of ethnic segregation in Denver's schools, especially in northeast Denver (Abbott, 1978; Hermon, 2004; Romero, 2004) Significant evidence that Denver's school district was inte ntionally creating segregation was presented in the Keyes v. School District No. 1 particulary the use of ethnically biased school boundaries (Connery, 2013) In 1973, the U.S. Supreme Court ordered Denver Public Schools to desegregate its schools through court mandated school busing (Abbott, 1978; Connery,

PAGE 182

168 2013; Hermon, 2004; Romero, 2004) In 1974, Denver Public Schools started a city wide program of school busing (Abbott, 1978; Connery, 2013; Hermon, 2004; Romero, 2004) The U.S. Supreme Court's decision on the Keyes case sparked a strong reaction from some Denver's Non Hispanic Whites, including acts of violence and new segregation oriented legislation. After the school busing program started, several episodes of violence occurred, like the bombing of school buses and constant threats to pro busing school officials (Leonard & Noel, 1990; Romero, 2004; Stephens et al., 2008) On the legislative side, in 1974 the Co lorado Republican Party convinced voters to approve two changes to the Colorado Constitution that would have stopped school busing to spread into Denver's Non Hospanic White suburbs (Hermon, 2004) The first change was the Poundstone Amendment to the Colora do Constitution, which has strongly limited Denver's capacity to annex land from its neighboring counties, while the second was language change in the Colorado Constitution that prohibited using the goal of ethnic balance as a rationale to assign students to different schools (Hermon, 2004) The Poundstone Amendment deserves a specific discussion, as it has inflenced Denver's urban development during the last 40 years. The Poundstone Amendment (1974) During a period of rapid suburban growth, Denver and its school district were annexing land very actively between 1950 and 1970 (Romero, 2013) In particular, the city grew from around 67 square miles in 1950 to 111 square miles in 1970 (Romero, 2013) Suburban municipalities and counties located aroun d Denver fought back and tried to prevent Denver from annexing their land as they were concerned about Denver's ethnic diversity s preading to their areas and about the racial tensions taking place in the 1960s and 1970s (Romero, 2013) Court mandated

PAGE 183

169 school busing following the Keyes case further contributed to the fears of suburbanites living in metropolitan Denver (Hermon, 2004) In 1974, residents of Greenwood Village, a home rule muncipality located south of Denver, and Republican State Representa tive Freda Poundstone proposed an amendment to the Colorado State Constiutution that would have strongly limited Denver's ability to annex land from surrounding municipalities, as well as Denver Public Schools' ability to bus students to suburban schools (Leonard & Noel, 1990; Romero, 2013) Colorado voters approved th e Poundstone Amendment with a large majority (Leonard & Noel, 1990; Romero, 2003) Under the Poundstone Amendment, Denver could annex land from surrounding counties only if t he latter voted in favor of annexation (Romero, 2013) According to the sources reported in Romero (2013) and in Leonard and Noel (1990) the Poundstone Amendment passed very easily because of th e suburbanites' fears of busing. Suburban residents wanted the ethnic and racial diversity to remain within Denver, and not to expand into their communities (Leonard & Noel, 1990; Romero, 2013) The Poundstone amendment was a strong tool to maintain socio economic and ethnic segregation between inner city Denver and it s outlying area s especially school segregation (Leonard & Noel, 1990; Romero, 2013) The Poundstone Amendment made Denver a landlocked city. After 1974, Denver only managed to annex significant amounts of land o n one occasion, when a portion of Adams County was added to Denver to build the new Denver International Air port in the 1980s (Johnston, 2011) The lack of annexation also had consequences on Denver's capacity to acquire land for public uses. The Poundstone Amendment has made it harder

PAGE 184

170 for the city of Denver to acquire parkland, as the city has had to fight against o ther interests when acquiring land because the land within the city limits is a finite resource (E. A. Taylor, 1986) When land is a finite resource, parks a re also competing against other public uses that need land, including schools, streets, and other utilities (E. A. Taylor, 1986) The implications of the Pou ndstone Amendment on Denver's capactity to acquire new parkland are further discussed in the land use planning section. Further suburbanization and white flight (1970s) As a partial effect of the Poundstone Amendment and of the annexation and de annexati on battles, the rate of white flight to the suburbs further increased in the 1970s (Chemerinsky, 2003) Many Non Hispanic Whites moved to Denver's suburbs to avoid school desegragation, as ordered by the Keyes court ruling (Chemerinsky, 2003) As a results, suburbs became increasingly Non Hispanic White and the percentage of ethnic minorities within Denver further rose (Chemerinsky, 2003) As the borders between school districts followed the boundaries between municipal ities, white flight contributed to creating ethnically segregated school districts, w ith high ethnic diversity in poorer Denver Public Schools and mostly Non Hispanic White students in more affluent suburban school districts (Chemerinsky, 2003) As most other American cities in the post World War II era, Denver's suburban growth was stimulated by a series of federal policies and interventions, including freeway construction and homeowner loans for suburban housing, and by local land use policies that favored suburban sprawl (Romero, 2013) In particular, the construction of two major Interstate freeways, I 25 and I 70, made it possible for suburbanites to live thei r

PAGE 185

171 American Dream by owning a single family home in Denver's suburbs, while having a relatively fast way to reach their office in downtown Denver (Romero, 2013) Unsurprisignly, white flight fueled by racial fears and federal policies oriented to suburbanization contribu ted to Denver's population loss to the suburbs. While in 1950 most people in the Denver metropolitan area lived in the city of Denver (74 percent), the situation was reversed in 1970, when the city of Denver retained only 42 percent of the metro politan are a's population (Romero, 2013) By 1975, Denver's population dropped to only 35 percent of the metropolitan area (Abbott, 1978) In the late 1970s, d uring a time of strong suburbanization, the city of Denver undetook a couple of long term planning efforts to guide its future development. The first was the 1978 Comperehensive Plan, which during a time of racial tensions and white flight, was oriented to social equity (Palmisano, 2014) In particular, the plan explicitly stated that everyone should have the same right to housing and to public s ervices, regardless of their demographics (Denver Planning Office, 1978) In the same period, the city started efforts to revitalize the South Platte River corri dor, including the creation of new parks and of a regional trail system (Leonard & Noel, 1990) These two long term planning initiatives anticipated a paradigm shift happening in the 1980s, which involved a renewed attention to Denver's public space and its central neighborhoods. Urban Renai ssance: 1983 2015 The fourth phase of Denver's history, which covers the years between 1983 and today, is characterized by a renaissance of the city of Denver, including significant public investment in public transportation and public buildings (Goetz, 2013; Johnston, 2011; Leonard & Noel, 1990; Murray, 2002) and substantial economic and population growth

PAGE 186

172 since 2000 (Leeds School of Business, 2015; United States Census Bureau, 2013b) While an urban renaissance has been observed in many other American cities in the last decade (Leinberger, 2009) Denver's start of its urban resurgence can be traced back to the 1980s and to the wo rk of mayor Federico Pe–a's administration. Mayor Federico Pe–a Elected for his first term in 1983, Federico Pe–a was Denver's first Hispanic and ethnic minority mayor (Kaufmann, 2003; Leonard & Noel, 1990) With gas prices decreasing in the mid 1980s, Denver experienced a significant economic downturn, which caused a substantial loss of j obs, high vacancy rates in downtown's office buildings, and population loss (T. A. Clark, 1985; Krumholz & Clavel, 1994; Leonard & Noel, 1990; Murray, 2002) During a period of economic recession, Pe–a's administration started a program of public investment to help the economy restart, including the Denver International Airpor t, the Colorado Convention Center, a new baseball stadium, and other public projects including schools and parks (Leonard & Noel, 1990) Although some opposed this plan of public investment, Denver voters approved several bond issues to fund a variety of public projects during the Pe–a admin istration (Leonard & Noel, 1990) P e–a's public investment programs were also aimed to balance social equity issues and to bring a message of inclusion to Denver's ethnic minorit y population (Kaufmann, 2003; Leonard & Noel, 1990) Besides the landmark projects he undertook (new airport and baseball stadium), Pe–a's efforts to improve schools, parks, and streets speak about hi s intention to improve the everyday lives of every Denverite (Krumholz & Clavel, 1994) Also, the Pe–a administration established structured forms of participatory

PAGE 187

173 planning with the city's neighborhood organizations, which shows Pe–a's efforts to make decisions with a bottom up approach (Krumholz & Clavel, 1994; Leonard & Noel, 1990) The 1980s depression provided the city with an occasion to plan its future and th ink about quality of life (Leonard & Noel, 1990) As during the Speer and New Deal eras, which both follow ed economic downturns, Denver took advantage of the stagnant economy to implement big projects to impro ve the city (Leonard & Noel, 1990) M ayor Pe–a committed to parks and public works during his administration, linking his work to M ayor Speer's City Beautiful vision (Leonard & Noel, 1990) To implement his park vision, Pe–a hired Carolyn and Don Etter, two lawyers with an expertise in Colorado's cultural landscape, as co mana gers of the Parks and Recreation department, and convinced Denver's voters to approve a 59 million dollars bond to improve the city's parks and parkways (Leonard & Noel, 1990) During the 1980s, also due to limited possibilities to expand by annexing land, the city undertook extensive effort s to revitalize and preserve its older neighborhoods (Leonard & Noel, 1990) Historic Denver, Inc. has initiated many preservation actions starting in the 1970s, including the Curtis Park neighborhood and Larimer Square in downtown Denver (Leonard & Noel, 1990) Mayor Wellington Webb Federico Pe–a decided not to run for his t hird mandate in 1991 (Cooper, 2012; Kaufmann, 2003) Wellington Webb, an African American Democrat, achieve d a come from behind win in the 1991 mayoral elections (Cooper, 2012; Hermon, 2004; Kaufmann, 2003; Romero, 2013) Webb, Denver's first African American mayor, won against all odds by conducting a low budget campaign that included walking through all Denver's neighborhoods (Cooper, 2012; Hermon, 2004 ;

PAGE 188

174 Kaufmann, 2003) Webb appealed to both African American and Hispanic voters and tried to deemphasize race in his campaign (Cooper, 2012; Hermon, 2004; Kaufmann, 2003) Wellington Webb remained in office between 1991 and 2003, reaching the limit of three mandates (Cooper, 2012; Hermon, 2004; Kaufmann, 2003) Social equity was part of M ayor Webb's political agenda (Noel, 2008; Stephens et al., 2008) During his three terms, Webb focused his efforts on four main areas: e stablishing and improving parks and open space; improving public safety; fostering economic development; and improving children's quality of life (Steph ens et al., 2008; Webb & Brovsky, 2007) According to a few former park planners I interviewed M ayor Webb was another champion of parks, continuing the tradition of strong mayors building parks in Denver. In addition, Webb conducted several efforts to im prove the living conditions of Denver's homeless and helped establish public institutions that honored Denver's African American history (Noel, 2008; Stephens et al., 2008) Economic growth and downtown revitalization In the 1990s, economic diversification led Colorado, and Denver in particular, to economic growth (Murray, 2002) Most of Denver's econo mic downturns were related to Denver's economy being overly based on few sectors, like oil and gas (Murray, 2002) This economic growth contributed to Denver's efforts to revitalize its downtown, and in particular the formerly industrial area known as Lower Downtown (J. Koehler, 2014; Murray, 2002) In a few years, parts of downtown became upscale residential areas for young professionals who have been attracted by city living (J. Koehler, 20 14) Large infill developments: Stapleton and Lowry Two New Urbanist developments, located on two former airports decommissioned in the 1990s, dominated

PAGE 189

175 Denver's planning efforts in the 1990s and its construction projects in the 2000s: Stapleton and Lowr y (Godschalk, 2004; Leccese & McCormick, 2003; Murray, 2002) The two large brownfield sites became available when the old Stapleton International Airport closed due to the opening of the new Denver International Airport and when the Lowry Field Air Force Base clos ed in the mid 1990s (Godschalk, 2004; Murray, 2002) Both Lowry and Stapleton were redeveloped following some of the principles of New Urbanism, including an emphasis on open space (Leccese & McCormick, 2003; Leccese, 2005; Piatkowski & Marshall, 2014) Stapleton and Lowry contr ibute d to significantly increasing Denver's park acreage (Leccese & McCormick, 2003; Leccese, 2005) The details of Stapleton's and Lowry's park planning and land use planning are discussed in the following sections of this cha pter. Expansion of public transportation Along with large infill developments, two large transportation projects drove urban development between the 1990s and the 2010s: the T REX and the FasTracks projects (Goetz, 2013; Johnston, 2011; Murray, 2002) While the T REX project, developed between the 1900s and the 2000s, including both a commuter rail and highway widening, FasTracks, starting in the 2000s an d still ongoing, has been focusing on building a light rail network throughout metropolitan Denver (Goetz, 2013; Johnston, 2011; Murray, 2002) Fa sTracks has been fostering the construction of several transit oriented dev elopments (TODs) in metropolitan Denver (Johnston, 2011) including Stapleton. Comprehensive Planning, 2000 2010 In the decade between 2000 and 2010, the Community Planning and Development and Parks and Recreation departments undertook a series of comprehensi ve planning efforts to create cohesive framework s for the

PAGE 190

176 development of the city during a time of rapid population growth. Community Planning and Development created a Comprehensive Plan in 2000 and an integrated land use and transportation plan, called Blueprint Denver, in 2002 (Godschalk, 2004) Blueprint Denver "divides the city into areas of stability' (established residential neighborhoods) and areas of change' (vacant and deteriorated infill site)," thus providing a frame work for future development (Godschalk, 2004, p. 10) In addition, a new form based zoning code was introduced in 2010, replacing the 1956 zoning code that, after 50 years of amendments, had become extremely hard to use (Bowen, 2011; Talen, 2013) Along with land use plans, in the 2000s Denver also developed comprehensive park plans, filling a gap in park planning that had existed since Saco DeBoer's park and parkway plan in 1929. Denver's Game Plan, adopted in 2003, is a comprehensive framework for futu re park planning in Denver, identifying priorities in terms of park needs, including park equity (Denver Parks and Recreation, 2003) The Denver Play Area Master Plan, adopted in 2008, focuses specifically on play spaces for children within parks, schools, and other public sites (Denver Parks and Recreation, 2008) The details of these land use and park plans, as well as their impact on park establishment, are discussed in the following sections of this chapter. Recent gentrification trends With a good and diversified economy, a constant population growth since 2000, and the Poundstone Amendment limiting Denver's possibility to expand, the city has experience d significant gentrification in some historically ethnic minority neighborhoods (J. Koehler, 2014; Langegger, 2013) with ethnic minority populations probably relocating to some of Denver's suburbs (Kneebone & Garr, 2010) Gentrification has been particularly strong in northwest Denver, mainly in

PAGE 191

177 Highland (Langegger, 2013) and in northeast Denver, primarily in Five Points, Whittier and North City Park (J. Koehler, 2014) Northeast Denver has been increasingly inhabited by young highly educated professionals, which Richard Florida described as the creative class in his writings (J. Koehler, 2014) Th e Non Hispanic White affluent gentrifiers are attracted by northwest and northeast Denver's proximity to downtown and b y their historic character (J. Koehler, 2014; Langegger, 2013) The 2000 Comprehensive Plan and Blueprint Denver, which expressed the goal of attracti ng the creative class to Denver and classified many low income neighborhoods as "areas of change" (Palmisano, 2014) likely contributed to gentrification in northwest and northeast Denver. With ethnic minorities being displaced by gentrification pro cesses, some of Denver's suburbs have become more and more ethnically diverse in the last ten years (Romero, 2013) Also, nationwide, racially integrated communities are now more prevalent in the suburbs than in central cities (Romero, 2013) Thi s shows that the traditional split between central city, hosting high percentages of ethnic minorities, and suburbs, hosting mostly Non Hispanic Whites, needs to be reevaluated (Romero, 2013) With an estimated population growth of 17.4 percent betwee n 2000 and 2013 (United States Census Bureau, 2013b) Denver is among the top ten fastest growing cities in the United States (Carlyle, 2014) reaching the highest population of its history (649,495; United States Census Bureau, 2013) Denver's ethnic diversity has also been growing, compared to the last decades of the twentie th century. Among Denver's 650,000 residents in 2013, 53.6 percent identify as Non Hispanic White alone, with the reminder 46.4 percent including Hispanics alone (30.9%), African Americans alone (10.2%), Asians alone (3.8%), American Indians (2%) and mixed ethnicities (3.1%;

PAGE 192

178 United States Census Bureau, 2013) Finally, Denver's urban renaissance seems far from being over, with its population density in its urban core significantly growing (Jaffe, 2013) and with its residential construction boom, especially around downtown, not giving any sings of decrease (Schroeppel, 2014) Looking Back to Look Ahead This long overview of Denver's history serves as a comprehensive his torical context for discussi ng how park planning, land use planning, and housing policies and practices have driven park location and park quality in relation to the location of different demographic groups. In particular, this overview showed the evolutio n of Denver as a tri ethnic city, including significant episodes of ethnic discrimination, the role of a few strong mayors in directing urban development, including parks It showed the strong imprint that the City Beautiful era has given to Denver's urban design, a s well as the complex connections between different economic periods, visions of society, and urban development. The current economic and population growth that Denver is experiencing creates interesting opportunities but it also raises equity an d financial issues for Denver's park system, which are discussed in the following sections. Park Planning The analysis of park planning in Denver spans from the establishment of the earliest parks in the 1860s to the processes that lead to parkland acquis ition and park construction today, including the current barriers to park establishment and improvement. This analysis also covers the evolution of park design in Denver, with a specific focus on how parks could have been more respon sive to young people's recreational needs. To study these processes in depth, I mostly relied on books, journal articles, and historical

PAGE 193

179 park plans for the period covering Denver's earliest phase until the 1970s; and on interviews with park planners, land use planners, and histo rians for the period between the 1980s and today. Also, it is important to introduce who were the actors in my analysis of "park planning" in Denver. These actors include individuals and institutions that have contributed to establishing and/or designing p arks in Denver, comprising: Denver's Parks and Recreation department, and its predecessors; developers, or other individuals and institutions donating land for parks; consultants working on park master plans and park designs; non profit organizations provi ding funding and support; City Council members and residents advocating for park budget s ; and any individuals and groups participating in park planning. However this section on park planning does not cover the contribution of land use planning to parkland acquisition, which is discussed next To organize the discussion of park planning, I use a few general themes that have emerged from my qualitative data analysis, which combined theory driven coding (Miles & Huberman, 1994) and constant comparison analysis (Corbin & Strauss, 2007; Glaser & Strauss, 1967) These themes shown in Tables 12 and 13 can help clarify what drove park planning in Denver during different phases of its history. The themes include the economic value of park s the preponderance of geographic equality over social equity in prioritizing park location, the evolution of park funding, the limited role of park master plan s the growing attention on privately owned parks, the current focus on park reprogramming, and the increasingly po sitive role of non profit organizations. Most of these themes highlight procedural inequity in park planning. For example, the use of parks as drivers of economic development has led to overlooking equity goals in

PAGE 194

180 recreation and to establish ing privately o wned parks to benefit a few wealthy New Urbanist enclaves. Also, initiatives to serve all geographical areas of the city equally are based on the assumption that every Denverite has the same nee d for parks, which clashes with low income ethnic minority you ng people needing parks the most. Parkland Acquisition and Park Construction Parkland acquisition involves the city purchasing land or obtaining land from developers in subdivision processes, while park construction deals with the actual physical modifica tion of an undeveloped site into a park. Clearly, park construction cannot take place until the city has acquired land. O n some occasions, Denver Parks and Recreation has developed parks for active use several years after the parkland had been acquired, du e to funding limitations. This is still the case for a few parcels of undeveloped open space in Denver (City and County of Denver, 2014b) This section provides an in depth analysis of parkland acquisition and park construction. Before starting th is discussion, it is important to note that there have been two main types of situations for parkland acquisition in Denver's history: parkland acquisitions in urban expansions or large infill planned unit developments and parkland acquisition in established neighborhoods. As emerged from my interviews, the opportunities and barriers to parkland acquisition are completely different in these two situations. In case of a large infill development like Stapleton, park establishment is relatively easy, as suggested by a current park planner inte rviewee : The bulk of new park establishment is happening in Stapleton. It makes sense because a new community was created there. That was a great example of park planning, but that doesn't happen often. It's easier when you start from scratch with a compre hensive master plan, but it's much more complicated when you have to intervene in existing neighborhoods.

PAGE 195

181 Urban expansions can be assimilated to large infill developments in terms of park establishment, if there is a good availability of land and a compreh ensive vision for the development expressed in a master plan. In this case, the land for parks can be obtained from a developer due to annexation regulations or a contract between city and developer (see land use planning section). On the other hand, to ac quire parkland in an established neighborhood, the city needs to purchase a property that is on the market. Also, land in established neighborhoods is often more expensive and harder to obtain than in urban expansions and large infill developments. Another former park planner agreed that "these are two completely different situations, in which you need to use very different strategies." Table 12 presents the themes and the codes describing the main aspects that have been driving park planning decisions in Denver in terms of park establishment with funding, equity or discriminatory goals, and the intrinsic and economic value attributed to parks being the most recurring aspects P ark funding is key to acquiring parkland and to establish ing new parks. Most of Denver Parks and Recreation's recent efforts have dealt with attempts to make up for a structural lack of publicly sourced park funding which has had a negative impact on park equity Therefore, while park funding is one of the themes included in Table 1 2 and is the focus of a specific discussion, funding issues and oppo rtunities are part of most themes. Also, while one of the themes explicitly mentions equity (park geographic equality vs. park equity), I discuss the implications of the mechanisms describ ed in each theme on procedural equity or inequity, and connect the latter to the equity of p ark spatial distribution in Chapter 6 (Mixed Research Integration).

PAGE 196

182 Table 12 Themes and Codes from the Analysis of Park Planning: Park Establishment Theme Descrip tion Codes The economic value of parks Parks have a significant economic value for different actors: Parks help developers sell homes; parks increase property values for homeowners; parks are a good investment for the city in terms of increased property t ax revenues and urban marketing ( tourism ) Developers donating land in subdivisions or asking for parks: "Parks sell homes" Good parks increase property values and taxes Good parks improve the image of the city tourism The prevalence of p ark geogra phic equality over p ark equity Planning for parks serving every part of the city regardless of need (geographic equality) vs. Planning for parks serving areas with highest need i.e. low income ethnic minority children (equity) Geographic equality has been significantly prevalent City Beautiful and Playground Movement: geographic equality "Class and race blind" park planners today Call for equity in some park plans The role of non profit organizations The evolution of park funding The availability of funding for parks (city's economy and political will), and how it has been distributed between operations, capital improvement, and parkland acquisition, has strongly influenced park planning in Denver's history Park districts (1900s 1950s) : inequiti es Economic conditions and political will Types of public funding for parks Barriers to acquiring new parkland The limited role of park master plans Park master plans, in their various forms, have provided frameworks and priorities for establishing parks, but have had a limited role in parkland acquisition Early park plans as Denver's urban design "Park plans do not build parks" Funding and political will needed to implement plans

PAGE 197

183 Table 12 Continued Theme Description Codes The establishment of p rivately owned parks and parkways Due to limited public park funding, several privately owned parks and parkways have been established in large infill developments (Stapleton and Lowry) Trunk system: large parks are city owned, small parks are HOA ow ned Private funding means better park quality Parks and Rec's strategy to leverag e partnerships and to reprogram parks Given funding limitations, Parks and Rec's strategy is to leverage partnerships to acquire new parkland and to focus most of its fundi ng on park improvement (already public land) Partners are non profits, other city departments, and state agencies Today, focus on park improvement more than acquiring new parkland as cost of land is very high The role of non profit organizations: lan d acquisition and park improvement In recent years, due to public funding limitations, Parks and Rec has significantly relied on non profits (mostly, GOCO, TPL and TGF) for funds to acquire land and to improve parks. Non profits are making a significant im pact, even on park equity Funding through competitive grants, not guaranteed TPL and GOCO focus on equity (park need) GOCO has funded more park improvement than parkland acquisition The economic value of parks The positive role of parks in cities has been historically conceptualized in terms of providing an oasis to escape the city, a democratic space for all citizens, a space that could promote public health (Cranz, 1982) More recen tly, a focus on the economic value that urban park system s can provide to cities and its residents has emerged (Harnik & Crompton, 2014) My analysis of park planning in Denver showed that many decisions on th e locations of parks have been made with their economic benefits in mind, especially as they relate to property values of the residential

PAGE 198

184 real estate located close to parks. The way the economic value of parks has shaped park planning in Denver also has ne gative consequences on park equity, as park planning driven by the goal of increasing property values has led to inequities in park spatial distribution. "Parks sell homes." The idea that "parks sell homes," as expressed by two park planner interviewees was clear from the earliest phase of Denver's history. Some of Denver's oldest parks, including Curtis Park and Fuller Park, were established through land donations from developers who used parks to advertise their subdivisions (Leonard & Noel, 1990; E. A. Taylor, 1986) As two experts I interviewed suggested, parks were seen as an amenity and brought economic value to a subdivision, especially in Denver's arid climate and semi desert landscape. A former park planner interviewe e in particular, pointed out : If you look at historical subdivisions of Denver, their plat map generally shows a mythical park. Some of these parks were actually built and others were not. Developers put parks in the plat maps to attract potential buyers, as a form of advertisement for their subdivision. The case of Curtis Park, Denver's first park, is particularly emblematic of the intersection between parks as an economic boost and equity. Established in 1868 through a land donation from Samuel Curtis, t he developer of the Curtis Park subdivision (Leonard & Noel, 1990) it was intended to create an amenity in the new subdivision "so that the lots would be more attractive for buyers," as expressed by a historian interviewee In the early 1880s, the Curtis Park neighborhood was one of Denver' s most desirable neighborhoods, hosting some of Denver's wealthiest citizens (Leonard & Noel, 1990; R. L. Simmons & Simmons, 1995)

PAGE 199

185 As the railway and some industries were established close to Curtis Park, and with the establishment of Capitol Hill as the newest hip residential neighborhood in the mid 1880s, affluent Denverites left Curtis Park for Capitol Hill (Leonard & Noel, 1990; R. L. Simmons & Simmons, 1995) This led to the creation of boarding homes and to the Five Point area and Curtis Park becoming the African American ghetto in the 1900s. Therefore, Curtis Park was a "hand me down" park from Non Hispanic White wealthy Denverites to low and middle class African American Denverites. T he park was established when the neighborhood was Non Hispanic White and affluent to increase property values and then African American s benefitted from the park after the first residents left the area Therefore, Curtis Park was not established with equity goals, to serve the growing and racially segregated African American community. Similar cases of "hand me down" parks were found in B altimore (Boone et al., 2009) and elsewhere Finally, the case of Curtis Park, as expressed by a former park planner I interviewed, shows the importance of considering the demographic changes that neighborhoods experience: When studying parks and equity, it is important to consider that neighborhoods have life cycles. Parks ar e built and are still there 80 or 100 years later. The demographics of the people who live around that park can change a lot over time. Developers who in the late 1800s and early 1900s aimed to increase their profit by selling park proximity as an amenity were not always eager to donate part of their subdivision to establish parks. In some cases, as suggested by Dorward (2010) they advocated for public park investmen t in their subdivision. In particular, Dorward (2010) reported that the developers of the exclusive Downington subdivision, located in today's South Park Hill, stron gly advocated with the city to obtain parks and parkways in their subdivision. By the early 1900s, plans for several parkways in Downington were underway, as we re

PAGE 200

186 the construction of large mansions (Dorwar d, 2010) This case shows how the business community and the local government collaborated in creating public green space to increase property values for the wealthy The use of parks to market a new development is also clear in Stapleton and Lowry, the two New Urbanist communities built in the 2000s. As several current and former park planners I interviewed suggested, in Stapleton and Lowry developers used the high provision of park s and their high quality as one of their main selling points. In particul ar, most park planner interviewees agreed that, thanks to several privately owned parks and parkways, Stapleton has much better park provision and maintenance than other parts of the city. This raises serious issues of park equity, as neighborhoods without wealthy homeowner associations cannot achieve the same park standards found in Stapleton (see section on privately owned parks in this chapter). A nother way Stapleton's developer has used parks to market Stapleton is by stressing the environmental benefit s of parks to attract eco conscious residents (Duffy, Binder, & Skrentny, 2010) The increase in property values that is linked to parks also benefits public revenues in terms of higher property taxes (The Trust for Public Land, 2010) A study conducted by The Trust for Public Land (2010) on the economic value of Denver's park system estimated that, city wide, parks contribut ed to increasing property tax revenues by approximately 4.1 million dollars per year (2008 data). Park planners during t he City Beautiful era were aware of this mechanism. Indeed, during that time, parks and parkways were often built before residential neighborhoods (Noel & Norgren, 1987) Thus, while the city made the first investments, "parkways contributed to establishing the

PAGE 201

187 status of some neighborh oods," according to a historian interviewee thus attracting wealthy residents and boosting property tax revenues. According to several former park planners I interviewed Denver has continued to focus on the potential economic revenues deriving from parks when planning the parks along the central South Platte River corridor. In particular, while in the 1980s Parks and Recreation was planning a large regional park near downtown to serve the historically park poor North and West Sides developer pressures an d the city's economic interests prevailed. "What happened is that the new parks along the Platte, like Commons Park, just created a front yard for a bunch of expensive condos," as a forme r park planner interview explained As I show in Chapter 5, some area s northwest and west of downtown are indeed park poor and would have benefitted from a large regional park northwest of downtown. The use of parks as generators of revenue is well expressed by a former Parks and Recreation manager in a public interview (Fi scher & Whitney, 2012) : By investing a few million dollars into those parks and transforming them from industrial areas, true redfields, into some of the greatest parks within the city, we've seen more than a 10 billion dollar investment throughout the So uth Platte corridor. The Greenway Foundation, a non profit organization that has led the efforts to revitalize the South Platte River, has further reinforced this focus on economic development through parks for the South Platte River corridor (The Greenway Foundation, n.d.) as shown by the slogan "green equals green" (Fischer & Whitney, 2012) These sources documenting the various values that con tributed to the establishment of parks along the central South Platte River corridor suggest that economic growth (fostering private

PAGE 202

188 investments and tax revenues) have prevailed over social equity (creating a regional park for the low income areas located northwest and w est of downtown). Parks for urban marketing Denver has also successfully leveraged the economic value of its parks by marketing itself as a place where visitors can enjoy great public space, including parks, in a very pleasant climate (The Trust for Public Land, 2010) The use of parks and other public spaces for urban marketing purposes can also be traced back to the City Beautiful era. Denver's city beautification efforts, led by M ayor Speer and focusing mostly on the city's parks and parkways, were aimed to establish a prestigious city in the American West and to make Denver a more desirable place to live (Noel & Norgren, 1987) Today, Denver undoubtedly benefits from the parks and parkways that were esta blished during the City Beautiful era (The Trust for Public Land, 2010) The study conducted by The Trust for Public Land (2010) on the economic benefits of Denver's park system valued the contribution of parks t o tourism generated tax revenues as approximately 3 million dollars annually (2008 da ta). The prevalence of p ark geographic equality over park equity As introduced in Chapter 2, the study of access to parks as an environmental justice issue can be undertaken with an equality lens or with an equity lens (Boone et al., 2009; Wolch et al., 2005) The goal of achieving an equal geographical spatial distribution of parks means allocating parks and resources evenly throughout the city, thus serving all demographic groups equally (Boone et al., 2009; Wolch et al., 2005) On the other hand, the goal of achieving an equitable spatial distribution of parks means prioritizing parks and recreation in areas of the city with high percentages of park dependent people, including: low income people, people under the age of 19 or above the age of 65, and people

PAGE 203

189 without access to a private vehicle (Boone et al., 2009; Talen, 2003; Wolch et al., 2005) As previously discussed, this dissertation uses the lens of equity, as low income ethnic minority people experience a series of additional barriers to recreation when compared to other demographic groups (Castonguay & Jutras, 2009; Karsten, 2005; Loukaitou Sideris & Sideris, 2010; Loukaitou Sideris & Stieglitz, 2002; Ries et al., 2008) My anal ysis of park planning throughout Denver's history shows that park planners have most often focused on achieving geographic equality in the spatial distribution of parks than on achieving park equity. This raises park equity concerns, as the predominance of class and color blind initiatives has often led to overlooking the recreational needs of low income ethnic minority young people. Also, the distinction between geographic equality and equity runs along a fine line, as some park planning documents have use d the word "equity" when they were actually describing "geographic equality," thus making it hard to interpret the actual intentions of these documents. This section critically evaluates how different park visions, park master plans, and actual interventio ns to establish parks have addressed geographic equality and equity goals. This analysis is important because it can help clarify to what degree actors involved in park planning aimed to improve park provision for park dependent people, thus helping evalua te procedural justice. Park geographic equality during the City Beautiful era The first efforts to achieve geographic equality in park distribution occurred during the City Beautiful era. Mayor Speer, championing a City Beautiful vision, pushed for creat ing parks in every part of Denver (Le onard & Noel, 1990) As a historian interviewee who extensively studied the City Beautiful era expressed, "the idea of the City Beautiful was about

PAGE 204

190 bringing green spaces to all neighborhoods. Speer was well intentioned and had a real wish to reach to the poor." While there is a reference to park dependent people ("wish to reach to the poor"), the idea of distributing recreational opportunities across the city with the goal of geographic equality is prevalent ("brining green spaces to all neighborhoods"). Public playgrounds were part of M ayor Speer's City Beautiful vision of park geographic equality. Historical accounts from the Denver Municipal Facts (City and County of Denver, 1912) and GIS data from the City and County of Denver (City and County of Denver, 2014b) show that playgrounds in the 1910s served most parts of the city, following a pattern of geographic equality. In o ther words, these source s suggest that playgrounds served both low income neighborhoods (Curtis Park, Sunnyside, and Globeville) and high income areas (Washington Park, City Park, and Montclair; City of Denver, 1912) Anna Louise Johnson, a public school teacher and leader of the Playground Movement in Denver, c onvinced M ayor Speer to build playgrounds across the city (City and County of Denver, 1912; Noel & Norgren, 1987) The Robinson Kessler park and parkway plan (1907) contributed to bringi ng parks to all parts of the city (Etter, 1986; Noel & N orgren, 1987) thus reinforcing a vision of geographic equality. This "windmill" plan included three parkway systems that radiated to the east, south, and north of downtown (Etter, 1986) As Noel and Norgren (1987) clearly expres sed, the windmill plan attempted to serve all Denver's neighborhoods (see Figure 11): But the Kessler Plan brought scattered parks and parkways into a single, city wide master plan for the first time. All parts of the city would benefit, and all parts of t he city would share in the cost of the Kessler Plan, which voters approved (p. 16).

PAGE 205

191 Figure 11 The Robinson Kessler Plan (1907). Image credit: Denver Public Library, Western History Map Collections. When discussing differences between park geographic equ ality and park equity, it is also important to discuss who "the poor" during Speer's mayorship were. A quote from Leonard and Noel (1990) can help clarify this point: "Denver's parks and beaches, its libraries and swimming pools, its Museum of Natural History and its zoo, were free and open to the public although that public did not consistently include African Americans" (p. 148). The lack of African Americans in Denver's parks and recreation during the

PAGE 206

192 1900s and 1910s can be attributed to tw o main reasons: Denver's African American population during that period was very low, reaching only two percent of the city's population in 1920 (Goldberg, 1980) ; and exclusion ary practices prevailed in public parks and pools (Abbott, 1978; Leonard & Noel, 1990) Thus, as most of Denver's populatio n at the time was composed of Non Hispanic Whites (Goldberg, 1980) and as e thnic discrimination in recreation targeted ethnic minorities, park planning under M ayor Speer aimed to achieve partial geographic equality in park provision Saco DeBoer's park and parkway master plan (1929), building on the Robinson Kessler plan, was ano ther attempt to establish parks and playgrounds in every part of the city (Etter, 1986; E. A. Taylor, 1986) According to Etter (1986) t he plan intended to "make the parks more available to people of all ages, of all groups, and of all needs" (p. 83). Ta ylor's (1986) account of Saco DeBoer's writing further clarifies the goal of park geographic equality: "The plan prepared by the Denver Planning Commission [the 1929 park and parkway plan] tries to extend the park system so park service will be within rea ch of every citizen. It tries to extend the playground service in the same way. [] (DeBoer 1930, 6)" (p. 44). Also in the 1920s and 1930s, the Ku Klux Klan dominated Denver's politics, with Benjamin Stapleton representing the Klan in the mayor 's office (Goldberg, 1980) Thus, in a period of increasing discrimination against ethn ic minorities, it would be unlikely that park planning, an effort driven by the city government, was intended to improve park provision for low income communities of color. After discussing what could have been the goals of park planners during the Speer and Stapleton era, it is important to analyze the overlaps between park geographic

PAGE 207

193 equality and park equity. Park planning efforts conducted with the goal of park geographic equality can have positive effects on park provision for low income communities of color, thus contributing to park equity. In particular, a few park planning interventions under M ayor Stapleton, which were motivated by park geographic equality goals, unintentionally m a de a positive contribution to park equity. Saco DeBoer and George Cr anmer, between the 1920s and 1930s, expanded Denver's park system along the South Platte River corridor and along the gulches located in West Denver (Noel & Norgren, 1987) When low income Hispanics started moving in the areas along the river and gulches in the 1960s (Diaz, 2005; Dwyer, 1993) they found an urban environment with decent park provision, thanks to the work of DeBoer and Cranmer. In other words, parkland alon g the South Platte River and the West Denver gulches was acquired in the framework of park geographic equality visions, but it unintentionally advanced equity with low income Hispanics moving to the area. Evaluating the City Beautiful park vision in Denve r The above discussion of park planning during the City Beautiful in Denver showed a preponderance of park geographic equality goals, with some unintended positive outcomes for park equity. T he processes and the ideology that led to these results highligh t some procedural injustices A few authors reported that Robert Speer and George Cranmer were considered political "bosses" by their contemporaries i.e., strong personalities who, without seeking consensus, implemented their vision for park establishment with authority (Noel & Norgren, 1987; E. A. Taylor, 1986) Thus an interesting aspect of Denver's park planning during the City Beautiful/New Deal era s is that a few personalities, Speer and Cranmer, dominated the decis ions on parks, often using methods that people at the time

PAGE 208

194 considered as debatable (Noel & Norgren, 1987; E. A. Taylor, 1986) This shows that, although their work was guided by good intentions, the park planning process happened mostly behind closed doors (E. A. Taylor, 1986) Given the exclusivity of the process, this can be considered an early example of procedural injust ice in park planning. T he purpose that culturally and politically dominant groups attributed to parks also influenced park planning during the City Beautiful As Etter (1986) argued, the cultural values driving the City Beautiful era hel d that a "good society" can be achieved in a properly laid out city. They would also include the idea that beauty is of value and "pays better than any other commodity"" (p. 17). This quote suggests that, as happened in other Ame rican cities (Boone et al., 2009; Pipkin, 2005) park planning during the City Be autiful era in Den ver had embedded social engineering goals. In other words, parks were seen as a means to create a "better" society and to reduce the risks of civil unrest, especially among low income young people (Boone et al., 2009; Pipkin, 2005) Following the ideas of landscape architect Frederick Law Olmsted, the idyllic landscape of parks would help lift the moral s of a city (Pipkin, 2005) By several accounts, the City Beautiful movement in Denver conceived parks as a public and democratic space, open to everyone (Etter, 1986; Noel & Norgren, 1987) The idea of parks as democratic spaces conflicts with the social engineering purposes of parks, as the latter pos tulates the acceptance of a unified set of values, or morals (Pipkin, 2005) while the first implies that different opinions and values can coexist in public space (Geenens & Tinnevelt, 2009; Parkinson, 2013) Also, while according to experts I interviewed "many parks established in that er a [City Beautiful] had a spirit of inclusion" and "parks were the only place where everyone could go," the widespread episodes of

PAGE 209

195 ethnic discrimination in parks directed towards African Americans, the Japanese, and Hispanics (Abbott, 1978; Leonard & Noel, 1990) show a problematic picture of park equity during the City Beautiful era in Denver. It is important to note, however, that Denver was not the only American city perpetuating segregation and discrimination in recreation at the time and using parks as a tool of Americanization (Boone et al., 2009; Kirk, 2005; T. Marshall, 1967) Another criticism that can be lodged against Denver's City Beautiful movement is a tendency to elitism in recreation. While Denver was quite successful in creating some recreational opportunities in every neighborhood, they work of George Cranmer, the manager of Improvemen t and Parks in the 1930s and 1940s, also focused on elitist forms of recreation such as golf courses, mountain parks, and ski resorts (Noel & Norgren, 1987) While M ayor Speer initially developed the i dea of creating a mountain park system in Denver (Leonard & Noel, 1990) much of the land acquisition for mountain parks happened during M ayor Stapleton's administration and thanks to Cranmer's work (Noel & Norgren, 1987) Cranmer managed to acquire land and establish the Red Rocks Amphitheater, a venue for concerts and movies in the Rocky Mountains foothills, and Winter Park Ski Resort, a publicly owned ski resort in the Rocky Mountains (Noel & Norgren, 1987) The Denver Planning Commission in 1929 warned that Mountain Parks "cannot possibly replace the city parks, which provide open air for crowded areas, recreational places in home districts, and contribute much to the beauty of th e city" (Noel & Norgren, 1987, p. 170) According to a Denver historian I interviewed, while M ayor Speer h ad a genuine interest in distributing recreation equally "people like Cranmer can be considered elitists," as most mountain parks "were out of reach for most poor people."

PAGE 210

196 Current efforts for park geographic equality A few recent sources, including the Denver Play Area Master Plan (Denver Parks and Recreation, 2008) and my interviews with land use planners, show that Denver's approach in park planning can still be considered mostly oriented to geographic equality. The current and past efforts concerned with park equit y are discussed in the following section. The Denver Play Area Master Plan (Denver Parks and Recreation, 2008) a master plan for play areas with in Denver's parks and school grounds, is oriented to distributing play areas equally to every part of the city, as it states: "one of the essential goals of System of Play is ensuring equitable geographic distribution of play areas throughout Denver" (p. 1 05). Based on the literature on access to parks as an environmental justice issue, what the Denver Play Area Master Plan calls "equitable" is instead a goal of geographic equality, as the use of the word "geographic" s hows The concept of geographic equity through which equity is based on geography (serving all the city) and not based on who needs play areas the most, is expressed in two other sections of the plan (see on p. 13 and p. 31). Also, while the plan includes a spatial analysis to determine areas that are underserved in terms of play areas by looking at the number of play areas divided by the number of people under 18 years of age (Denver Parks and Recreation, 2008) the spatial analysis does not take into account ethnic minorities and people in poverty in the analysis of play area needs. Therefore, the Denver Play Area Master Plan makes a step towards equity by including the number of you ng people in its spatial analysis but it does not consider ethnic minorities and low income groups as park dependent people. Indeed, an in depth analysis of the plan shows no words related to

PAGE 211

197 ethnicity (except regarding ethnic integration in play), race, s ocial class, income, or poverty. Similarly, some land use planners I interviewed perceive park planning in Denver as oriented to park geographic equality. As one explained, Denver's Parks and Recreation "feels that it is very important that parks are sprea d throughout the city. [...] The decision of investing money to create this park [Mayfair Park] was purely a geographic consideration, as the area lacked parks." Mayfair Park is located in a fairly affluent and Non Hispanic White neighborhood, thus park in vestment was prioritized based on geographic needs. A nother land use planner interviewee further expressed the focus on park geographic equality: "From my experience, park planners are pretty class and race blind. They want to put parks in the parts of the city that do not have enough parks regardless of who lives there." While this evidence shows that park plann ing in Denver has been mainly oriented to park geographic equality, with a "class and race blind" approac h, other evidence suggests that few equity oriented park planning efforts have occurred in Denver. Past and current efforts for park equity. My analysis of Denver's park planning highlighted three main cases of equity oriented park planning, from the 1960s to today: The establishment of recreation centers and small parks along the South Platte River in the 1960s and 1970s paralleling such efforts nationally ; the park equity analysis and priorities expressed by Denver's Game Plan (Denver Parks and Recreation, 2003) ; and the recent work of non profit organizations to build parks in park poor low income ethnic minority neighborhoods possibly reflecting the current national intere st in health disparities in such neighborhoods

PAGE 212

198 While other initiatives oriented to park equity could have been implemented before the 1960s, the data I collected shows that the first e xplicitly equity oriented initiative in Denver park planning is the use of federal money in the 1960s and in the 1970s to establish recreation centers with small parks around them in low income communities of color. According to a former park planner at Parks and Recreation, "recreation centers were built mostly along the [So uth Platte] river due to federal investment in the 1960s, during a time of civil unrest." In particular, Goldstein (2011) reported th at the Rude Recreation Center, located in the Sun Valley neighborhood, was built thanks to funding from the War on Poverty program. The War on Poverty program was established in 1964, as Congress approved President Lyndon B. Johnson's proposal, with the go al of fostering community development in impoverished urban areas (W. J. Wilson & Aponte, 1985) After Congress approved the federal program, Denver established its local War on Poverty organizat ion (later renamed Denver Opportunity, Inc.) which, amon g other initiatives, built a series of health centers in the city's lowest income neighborhoods (Cowen, 1968) Ba sed on Denver's GIS data, the city built 17 recreation centers between the 1960s and 1970s in low income ethnic minority neighborhoods (City and County of Denver, 2014b) More archival research would be necessary to identify the contribution of federal funding to each specific rec reation center. The second equity oriented initiative I have encountered in my analysis is Denver's Game Plan, a city wide park master plan developed in 2003 (Denver Parks and Recreation, 2003) To prioritize park investment, Parks and Recreation conducted a GIS spatial analysis that overlays areas that are underserved by parks, and areas that are more park dependent (Denver Parks and Recreation, 2003) The Game Plan's definition of park

PAGE 213

199 dependency includes the number of people under 18 years of age, poverty data, an d income data (Denver Parks and Recreation, 2003) Therefore, the Game Plan 's approach to prioritize investments is more oriented to park equity than to the one used in the Denver Play Area Master Plan The Game Plan also acknowledged that, based on a survey, ethnic minorities expressed more dissatisfaction about park provision and park maintenance than Non Hispanic Whites (Denver Parks and Recreation, 2003) This shows openness to discussing park inequity concerns, including issues related to social class and ethnic ity. The third and more recent example of equity oriented park planning is the work of two non profit organizations, Great Outdoors Colorado (GOCO) and the Trust for Public Land (TPL), to establish pocket parks in low income ethnic minority neighborhoods t hat were underserved by parks. Several park and land use planners I interviewed acknowledged the fundamental contribution that these two organizations provided in terms of funding (mostly for park improvement) and in terms of administrative support (mostly to acquire land qui ckly). The specific efforts of non profit organization s for park equity are discussed later in this chapter. To summarize, these three cases of equity oriented park planning openly acknowledged, through investments and priority setting, that low income and ethnic minority people have an additional need for parks and recreation than Non Hispanic Whites middle and upper class people. Therefore, these examples go beyond the "class and race" blindness some planners discussed in regard to pa rk geographic equality. Perceived unbalances in park distribution. E vidence I collected from secondary data and from my interviews shows that some Denver's residents and park planners have

PAGE 214

200 perceived some areas of the city, and some demographic groups, as b eing underserved by parks. In terms of geographic distribution, interview data and other sources (Goldstein, 2011; La ngegger, 2012) point to a picture of park provision divided along a southwest to northeast line: Planners, residents, and scholars have perceived East and South Denver as park rich and West and North Denver as park poor. These geographic unbalances were p erceived by some residents of Highland, in northwest Denver, since the early 1900s (Goldstein, 2011) and continued through the 1980s until today (Langegger, 2012) In terms of ethnicity, the Denver Parks and Recreation 2001 survey reported in the Game Plan shows that African Americans were the ethnic group expressing the highest dissati sfaction with park provision and park maintenance (Denver Parks and Recreation, 2003) This dissatisfaction is concerning as African Americans, according to the survey, relied on public parks for recreation more than Non Hispanic Whites (Denver Parks and Recreation, 2003) Perceived unbalances and inequities also dealt with park maintenance. In particular, Hispanics living in northwest and west Denver lamented poor conditions of parks and pools in their neighborhoods, when compared to parks in other wealthier areas o f the city (Langegger, 2012; Leonard & Noel, 1990) In the late 1960s young Chicanos, who were part of the Crusade for Justice, protested the poor maintenance of the pool in La Alma/Lincoln Park, a low income Hispanic neighborhood at the time, by occupying the Eisenhower pool, located in an affluent area in southeastern Denver (Leonard & Noel, 1990) Also, some Hispanic residents of the North Side interviewed by Langegger (2012) perceived that, between the 1970s and the 1990s, parks located in Highland and Sunnyside had a lower maintenance standards than parks located in south and east

PAGE 215

201 Denver. Two experts I interviewed reinforced these percep tions of inequity in park maintenance, and reported that residents still perceive inequities today. In Chapter 5, I present the results of the GIS measured distributional equity and inequity of parks based o n different parameters, and in Chapter 6, I discu ss the connections between perceived and GIS measured equities and inequities. The evolution of park funding Besides park plans, park visions, and the political will to establish parks, park funding is the most obvious necessary condition for a city to de velop a good park system. My analysis of park planning in Denver show ed that the amount of funding available for parks has changed throughout Denver's history, depending on the city's econom ic health, on the value that different mayors and administrations attributed to parks, on external funding and on the citizens' willingness to be taxed to support public services In summary, park funding in Denver reached a peak during the City Beautiful era, when mayors and the city's elites supported park investment and the parks and public works departments were joined in the Department of Improvements and Parks. With the creation of separate Parks and Recreation and Public Works Departments in 1956, and with a diminishing interest in Denver's public space due to sub urbanization, park funding significantly decreased. The 1980s and 1990s saw a renewed interest in parks through the work of the Pe–a and Webb administrations, but park funding did not increase significantly, except for a few bond issues. Currently, Denver, like many other U.S. cities, is struggling with park funding and with developing creative funding strategies. The current structural deficiencies in funding to establish new parks suggest that Denver might not consider increasing its parkland as a priorit y. In this section, I draw a brief history of park funding in Denver

PAGE 216

202 Park funding in Denver's early stages As mentioned earlier, many of Denver's first parks were established through land donations, with developers using parks to market their subdivision s (Leonard & Noel, 1990; E. A. Taylor, 1986) In the 1880s and 1890s, the first conflicts over public park funding occurred. The Sopris Lee Downing plan (1878) and the Evans Rollandet plan (1894) well exemplify these conflicts Denver's first park plan, although not comprehensive, was the Sopris Lee Downing plan, known as the "hourglass plan," developed by M ayor Sopris and others (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) The plan proposed the creation of two large city parks east of downtown (City Park) and north west of downtown (Sloan's Lake), connected by a large boulevard (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) O nly Ci ty Park was implemented from this plan (Etter, 1986; Noel & Norgren, 1987; E. A. Taylor, 1986) Although land was not expensive at the tim e, after several debates and disagreements, City Council approved purchasing the land for City Park but not for Sloan's Lake, explaining that "money would be better spent on sewers, streets, lights, and police and fire protection" (Taylor, 1986, p. 23) The failure to complete the Sopris Lee Downing plan is one of the first examples of parks competing with other pu blic needs for funding, and losing. The Evans Rollandet plan was Denver's first comprehensive park and parkway plan, proposed in 1894 by John and William Evans, who owned the city streetcar system (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987) The Evans Rollandet plan was very ambitious and complex, with a series of parks spread throughout the city and parkways con necting these parks and a greenbelt in the periphery of the city (Etter, 1986) As for the Sopris Lee Downing plan, decision makers and the local press

PAGE 217

203 considered the plan too grand and costly because the funding would have had to cover parkland a cquisition, park construction, and park maintenance (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987) As a result, only Park Avenue and improvements to City Park were actually built (Leonard & Noel, 1990) While the Evans Rollandet plan followed the 1893 economic recession by only one year (Leonard & Noel, 1990) and thus concerns about funding were justified, some of the arguments against the plan pointed to the lack of ne ed for parks (Noel & Norgren, 1987) which spea ks to the value that decision makers and the local press attributed to parks in the 1890s. Park funding during the City Beautiful era Available funding for parks significantly increased during the City Beautiful era and during Speer's administration in p articular. As a result of a paradigm shift in the value attributed to parks, M ayor Speer, not without opposition, launched a city wide program of park construction by significantly increasing the budget for parks (Noel & Norgren, 1987; E. A. Taylor, 1986) Funding for parkland acquisition was large, as M ayor Speer managed to double Denver's park acreage during his terms (Leonard & Noel, 1990; Noel & Norgren, 1987) Among other parkland acquisitions, the city created Berkeley Park and Washington Park by purchasing several small parcels of land from various property owne rs in the 1890s and 1900s (ASLA Colorado, 2007; Goldstein, 2011; Wiberg, 1995) which was a fairly common mechanism at the time (Boone et al., 2009) According to the park planners I interviewed, this mechanism would be extremely costly today, due to increased property values in Denver and to limited funding. Also, landscape architect Saco DeBoer reported that money for park construction was not an issue under M ayor Speer, as construction crews were of ten increased to speed up park establishment (E. A. Taylor, 1986)

PAGE 218

204 While many criticized Speer for his park investments, which were considered in some cases as a waste of taxpayers' money, Speer succeeded in his park funding and construction efforts because the local business community and power elite backed his work (Noel & Norgren, 1987) Similar conditions of large park spending, with the criticisms that were associated with them, were found during M ayor Stapleton's terms, when George Cranmer managed to acquire significant amounts of parkland (Noel & Norgren, 1987) Mayor Speer's and Stapleton's successes shows that the political will and the power of certain decisi on makers has made a positive difference for park f unding, especially when Denver's elites have shared their visions Park districts as mechanism for park funding The way park funding was distributed across the city between the 1900s and the 1950s deserves particular attention, as it raises some questions of equity. After the City and County of Denver was established in 1902, the city organized park planning, funding, and management in four distinct park districts: the East Denver Park District, the Highland Park District, the South Denver Park District, a nd the Montclair Park District (City and County of Denver, 1913; Goldstein, 2011; Noel & Norgren, 1987) Each of these four park districts had a park commissioner, who was part of Denver's Board of Park Commissioners (City and County of Denver, 1913) The four park districts are highlighted in Figure 12. According to historians and park planners I interviewed, the division of the city in to four park districts had significantly negative implicatio ns for park equity. As the city collected funding for park establishment and park improvement based on property taxes, the districts with the highest property values, where the wealthiest groups lived, generated more park funding. As a result, the wealthie st park districts obtained more

PAGE 219

205 parks and parkways. As one historian I interviewed explained, "East Denver, South Denver, and Montclair, during Mayor Speer's administration, were wealthy and raised a lot of tax money for park establishment and improvement. This led to perceived inequities in park distribution. As early as the 1910s, residents of north and west Denver were dissatisfied with park provision in their neighborhood s when compared to the parks and parkways located in south and east Denver (Noel & Norgren, 1987) According to two park planners I interviewed, the general pattern of inequity between south and east Denver (more parks), and west and north Denver (fewer parks) can partially be attributed to the funding system based on the four park districts.

PAGE 220

206 Figure 12 The four park districts highlighted in the Robinson Kessler Plan (1907). Image credit: Denver Public Library, Western History Map Collections (edited). Economic conditions and park funding During the City Beautiful era, New Deal era, and later in other periods of Denver's history, the health of the city, state and national economies have influenced the amount of available park funding. In some cases, economic recessions have limited park spending due to concerns about the city's finances, as in the case of the Eva ns Rollandet plan (Leonard & Noel, 1990; Noel & Norgren, 1987) Also, as the Denver was facing a lo cal economic recession in the mid 1980s, lack

PAGE 221

207 of park funding forced Denver Parks and Recreation to focus on long term planning for Stapleton, Lowry and the Gateway area rather than on establishing new parks, according to two former park planners I intervi ewed. However, in other cases park planners took advantage of economic downturns to acquire parkland and establish new parks. Park planners during the City Beautiful and New Deal periods, including Saco DeBoer and George Cranmer, capitalized on the 1893 a nd 1929 recessions (Leonard & Noel, 1990; Noel & Norgren, 1987) These two recessions meant lower p roperty values and decreased competition from developers aiming to acquire land, as well as unemployed workers who could provide inexpensive labor (Leonard & Noel, 1990; Noel & Norgren, 1987) George Cranmer was particularly effective in capitalizing on low land prices and cheap labor costs to establish new parks (Leonard & Noel, 1990; Noel & Norgren, 1987) It is important to note that Cranmer also benefitted from a strong support by the city govern ment and by the economic elites of the city (Leonard & Noel, 1990; Noel & Norgren, 1987) Denver Pa rks and Recreation D epartment (1956) According to park interviewees the establishment of the Parks and Recreation department in 1956 led to decreased funding for parks in Denver. Until 1956, the Department of Improvements and Parks encompassed park plann ing and public infrastructure planning, which contributed to a cohesive vision for different parts of the city's infrastructure (E. A. Taylor, 1986) In this context, urban planners, landscape architects and civil engineers worked side by side in the same department (E. A. Taylor, 1986) contributing to the succe ss of the City Beautiful movement in Denver.

PAGE 222

208 In 1956, the Department of Improvements and Parks was divided into the Department of Public Works and the Department of Parks and Recreation (E. A. Taylor, 1986) According to a former planner at Parks and Recreation, "after this split, Parks and Recreation started competing for the budget, resulting in less money available for parks. Parks were seen as the soft, l ess necessary part of the city. Public works and planning became predominant." Thus, the split between Public Works and Parks and Recreation put the latter in a subordinate position in terms of political weight and, consequently, budget. Reduced funding fo r parks meant lower standards of maintenance in Denver's parks and funding issues with Denver's Mountain Parks system (Leonard & Noel, 1990; E. A. Taylor, 1986) Denver lost approximately 40 percent of its street trees between the 1960s and 1970s (Leonard & Noel, 1990) and by the 1980s Denver Parks and Recreation had serious concerns about the level of maintenance of Denver's parks (E. A. Taylor, 1986) Similar pat terns of reduced public spending for parks were observed in several other U.S. cities in the 1960s and 1970s, mostly due to white flight to the suburbs and federal disinvestment in cities which caused a reduction in real estate tax revenues (Low et al., 2005) This reduced importance of parks in terms of city vision and funding reflects the paradigm shift from a period valuing urbanity and public spac e (City Beautiful) to a phase supporting suburbanization and private space. Forms of public park funding: 1980s today Interviews with park planners, land use planners and historians provided a very rich description of public funding for parks from the 19 80s until today. The types of available funding and their purposes have been fairly consistent from the 1980s, during Pe–a's administration, to the current times. Denver Parks and Recreation has had three main types of funding: operating budget,

PAGE 223

209 capital im provement budget and bond revenue Table 13 includes the brief descriptions of each type of public funding. Of these three sources of funding, operating and capital improvement budgets are emitted on a yearly basi s, while bond issue money is generally ava ilable every ten years. Table 13 Types of Public Funding for Parks in Denver Funding Types Description/purposes Recurrence Operating budget Parks and recreation operations, department operations, and normal maintenance Yearly Capital improvement budge t Large deferred maintenance and small improvements (parks and rec centers) Yearly Bond issues Parkland acquisition, park or rec center construction, and large park or rec center improvements Approximately every 10 years The operating budget covers the normal operation of Denver's park system, recreation centers and administrative functions, as well as its scheduled maintenance. In other words, the operating budget allows Denver's park system to function every year, but does not include any funding to i mprove or create new parks. The specific budget for city park operations in 2014 was approximately 54 million dollars (Denver Parks and Recr eation, 2013) while total spending for parks amounted to 67 million (The Trust for Public Land, 2015a) The last number translates to $104 per resident, which is significantly low er than Seattle, WA ($298), Portland, OR ($141), Denver's neighbor Aurora, CO ($131), but higher than Los Angeles, CA ($82), Baltimore, MD ($57), and of approximately 60 percent of the 100 most populated cities in the U nited S tates (The Trust

PAGE 224

210 for Public Land, 2015a) Thus, Denver's park spending per person, although not among t he highest, is relatively large, mostly due to high operating costs. Capital improvement funding is another part of the yearly Parks and Recreation budget and is aimed to fund major deferred maintenance projects ("for example remaking the roof of a rec center") and small improvement projects. The two sub types of funds are called "capital maintenance and repair" funds and "capital upgrades and expansion" funds (Denver Parks and Recreation, 2003) The capital improvement money available for De nver Parks and Recreation was approximately 10 million dollars in 2014 (Denver Parks and Recreation, 2013; Robson, Wohlgenant, Harmon, & Gripne, 2014) which is significantly lower than the operating budget. This reflects what happens citywide As a land use planner I interviewed explained, the vast majority of the city budget is for operating the city (police, firemen, etc.). The budget for capital improvement is not that much." Therefore, capital improvement money is very rarely, if ever used to acquire new parkland. Like every other type of municipal funding, capital improvement funds are distributed by the city to differen t departments. Several present and former park planners of Parks and Recreation pointed out that parks have to compete against other city departments to obtain capital improvement money and that often other departments, like Public Works and Community Plan ning and Development, have more leverage in obtaining these funds. As a current park planner at Parks and Recreation explained, "the budget is an annual effort. Each department puts together the budget requests for that year." Thus, parks have to compete f or budget against other infrastructure needs. Denver

PAGE 225

211 uses a system of priorities to distribute capital improvement money among departments, according to a land use planner interviewee : The first [priority] level deals with public health issues we want to make sure that people don't get seriously injured or die due to something broken or dangerous; second, we spend money on things that next year would be too expensive to fix; then there are other levels, which could include money for parks. Also, several e xperts I interviewed pointed out that the capital improvement budget is very political. In particular, while Parks and Recreation has been using data driven strategies to identify the priorities for capital improvement money (GIS analysis and several surve ys), the city budget involves political negotiations and decisions are ultimately made by City Council. A park planner at Parks and Recreation well described that the department's evidence based rational planning model can provide useful information, but t hat rational planning does not always inform political choices: Recently, we tried to generate data through the GIS department to identify where the priorities for the budget are (based on the conditions of different sites). We can see things at a very sm all scale, for example a sidewalk that needs repair. GIS analysis can really help identify priorities for funding. However, in the end budget becomes a political decision. Another factor that makes capital improvement funding very political is that part of these funds, citywide, are split between Denver's 11 City Council voting districts. Given this system, further political negotiations occur between the representatives of each district and within each district. In particular, a land use planner intervie wee explained a recent case of capital improvement budget prioritization for District 10: Recently the district with Cheesman Park [District 10] spent a lot of money to recreate the historic trails within the park. It is probable that the decision to spend the money in this park came from public meetings in which the voice of historic preservation people was very strong. Who speaks the loudest has the last word. This shows that a portion of the budget is political.

PAGE 226

212 This example also shows how political powe r and advocacy ("who speaks the loudest") can significantly drive decisions on capital improvement funding. This interview passage also suggests that public meetings within voting districts can strongly influence budget decisions, at least if representativ es take into account the requests of their constituencies. Once City Council makes the annual budget decisions, capital improvement funding cannot be moved between projects. Indeed, "the budget includes some line items that describe specific projects for w hich money is needed." The capital improvement budget also comprises discretionary funds, which Parks and Recreation uses to leverage partnerships and match external grants. Discretionary funds can be a significant portion of the yearly capital improvement budget. For example, the 2014 Parks and Recreation budget included 3.5 million dollars for discretionary funds, which are part of the 10 million dollars for capital improvement projects (Denver Parks and Recreation, 2013) Funding from bond issues, which is available approximately every 10 years, has been covering parkland acquisition, construction of parks and recreation ce nters, and large improvements to parks and recreation center s Like capital improvement funding, bond issues specify how funds need to be spent project by project. According to many experts I interviewed, bond issues are the major source of public funding through which Par ks and Recreation can acquire new parkland. As a land use planner I interviewed explained, in the past 25 years Denver's voters approve d three major bond issues that provided park funding : the 1989 bond issue proposed by M ayor Pe–a; the 1998 bond issue pro posed by M ayor Webb; and the 2007 Better Denver program promoted by M ayor Hickenlooper. For example, t he 1989 bond issue included 59 million dollars for

PAGE 227

213 parks (Leonard & Noel, 1990) B ond issues have positively impacted the establishment of parks and recreation centers by providing funding f or land acquisition and construction. As bond issues need to be approved by voters, they are, like capital improvement funds, very political. Bond issues require significant lobbying and funding to campaign for them, as well as gaining "support from well c onnected Denver resid ents," as expressed by a park planner interviewee formerly at Parks and Recreation. The case of the 1989 bond to improve Speer Boulevard clearly shows how issues of power, which have led to inequities in park spending, have been embedd ed in bond issues. As a former park planner interviewee for Parks and Recreation explained: People in the Country Club neighborhood were initially reluctant to support a bond issue [in 1989] to improve and widen Speer Boulevard on the edge of their neighbo rhood because of traffic and other concerns. But they shifted to supporting the Speer Boulevard and 1st Avenue bond after the bond issue included the construction of large walls on their property to counter sound and traffic issues. These beautiful decorat ed walls with nice entrance gates are built on parkland with public money. Country Club is a very affluent neighborhood with a private golf course, thus the voices of its residents have been very strong in the 1989 bond issue campaign. When analyzing this episode through a park equity lens, the funds to build Country Club's walls and entrance gates could have been more appropriately used to improve parks located in low income ethnic minority neighborhoods. Another former member of Parks and Recreation point ed out that, in the debate for the 1989 bond issue, "everyone was thinking about their park People would only think about their neighborhood." It took a significant effort by Parks and Recreation to "instill the idea that every park is your park ." This sh ows how neighborhood interests can strongly influence which parks are funded in bond issues.

PAGE 228

214 Another peculiar feature of bond issues is their focus on regional parks. As a park planner at Parks and Recreation pointed out, "historically, Denver's residents have recognized the importance of regional parks. In the past, most of the bond money was focused on regional parks, which were the most used parks." The focus on regional parks is further reinforced by higher maintenance investments, as shown in the Park s and Recreation's park maintenance standards (Denver Parks and Recreation, 2007) Several current and former Parks and Recreation staff members indica ted tha t, in Denver, regional parks receive the largest share of park funding and, for this reason, they are the objects of constant park planning efforts. To summarize, among the three types of municipal funding for parks, only bond issues provide a good source of funding for parkland acquisition. However, funding from bond issues is available approximately every ten years, which significantly limits Denver's capacity to acquire new parkland and to address inequity issues in park provision Several park pl anner interviewees lamented a chronic lack of funding to purchase land for parks in Denver in the last 30 years, as the city has "no dedicated funding source for park land acquisition" (Robson et al., 2014, p. 3) This is clea rly an issue in a landlocked city with high property values, as an effect of the Poundstone Amendment. Another issue for establishing new parks is that "land acquisition must compete internally against all capital improvements and the mountain park system" (Robson et al., 2014, p. 3) Parks and Recreation has recently expressed concerns about park provision (Fischer & Whitney, 2012; Robson et al., 2014) With a growing population, and with the new form based zoning code allowing high densities in areas that are already underserve d

PAGE 229

215 by parks, the park system is facing greater pressure, as the acres of parks available for each resident will likely decrease without the establishment of new large parks (Fischer & Whitney, 2012; Robson et al., 2014) However, even if funding w ere available to acquire new parkland, or if land use policies r equired developers to donate parkland to the city in infill developments, the cost of maintaining parks in the long term would partially deter Parks and Recreation from acquiring parkland. Maintenance barriers to acquiring new parkland. Several park plan ners and land use planners I interviewed pointed out that one of the major barriers to acquiring new par kland is the need to maintain these parks in the long run As reported in the Game Plan, the issue of adding new acres of parks to the existing system i s that the operating budget needs to be updated to include the maintenance costs of the new parks (Denver Parks and Recreation, 2003) In particular, the Game Plan reports that "the cost of operations will grow dramatically with close to 1,500 acres of new parkland at Stapleton, Green Valley Ranch, Lowry, and other new neighborhoods" (Denver Parks and Recreation, 2003, p. 112) These concerns were well grounded. As a planner interviewee at the Commun ity Planning and Development department suggest ed the city didn't receive enough additional funding to maintain the additional parkland that was created in Lowry and Stapleton." This issue is particularly problematic in Denver because of the high cost o f park maintenance. According to the experts I interviewed, park maintenance in Denver is exceptionally high due to its very dry climate, with very little rain, and for the year round high park usage. As a park planner at Parks and Recreation explained, "I t's amazing to see how much money goes to maintenance. Sport fields require a lot of maintenance,

PAGE 230

216 especially in our dry climate. They require a lot of irrigation." One design solution that can lower park maintenance expenses is the shift from the water nee dy Kentucky bluegrass, typical of Denver's City Beautiful parks, to natural landscapes with native prairie. This strategy was suggested in the Game Plan (Denver Parks and Recreation, 2003) and implemented quite largely in Stapleton. However, park planners reported that the turf conversion to native prairie has upset some historic preservationists and residents who are attached to the id ea of a perfect green lawn. I discuss this issue more in depth in the section on park design, maintenance and use. Given high park maintenance costs, the Game Plan's main financial priority is to maintain the existing park system, followed by increasing se rvices, and by banking land for the future (Denver Parks and Recreation, 2003) The park planners I interviewed reported that park ma intenance is still Parks and Recreation's main priority in 2015. As a former Parks and Recreation's staff member explained, these funding limitations have been hindering the department's park equity efforts: "Parks and Recreation does want to add acreage f or equity but the department is always cautious (a s are politicians) because the city then has to take care of it." Another maintenance issue that limits the city's park equity efforts is that small parks are inefficient to maintain. Small parks can play a n important role for park equity because, as Sister, Wolch, and Wilson (2009) have shown, adding a small park in dense park poor areas can significantly improve the residents' access to parks in terms of proximity and park acreage. Several in formants have pointed out that Denver is reluctant to buy land for small parks or to accept small par ks from developers because small parks are inefficient to maintain in terms of logistics and because regional parks already absorb

PAGE 231

217 a large share of park ma intenance costs. In terms of maintenance efficiency, a land use planner formerly at Community Planning and Development explained: They [Parks and Recreation] generally will acquire parks if they are large enough, something around 5 acres. Maintaining smal l parks is expensive and inefficient because the park maintenance crew would have to drive around town and stop in several places, rather than optimizing their work by having to reach only a few large parks. Although they are inefficient to maintain, Parks and Recreation as been acquiring small parks, under the condition that "pocket parks are surrounded by residential land uses, better if with fairly high densities" and that they are "successfully placed to serve needs." In particular, Parks and Recreation has recently acquired a few small parks in areas with low income ethnic minority areas that were park poor including Cuatro Vientos Park in the Westwood neighborhood and New Freedom Park in the East Colfax neighborhood, thanks to the support of a few non profit organizations. The establishment of these two pocket parks is discussed in the section covering the role of non profit organizations. Impossibility to set aside funds for parks Besides a structural shortage of funding for parkland acquisition and for large park improvements, Denver Parks and Recreation has lacked instruments to set aside money for parks during periods of economic growth. As a former staff member interviewee of Parks and Recreation explained: [In the late 1980s,] we wanted to creat e an endowment. When the city is moving, when things go well, it makes sense to set aside some money. We wanted to create an endowment for the overall park system, through which we could put aside money during good economic times. That money could have the n been used during economic downturns. Colorado State law did not allow Parks and Recreation to create an endowment during the 1980s. As Denver's Game Plan highlighted the Taxpayer Bill of Rights (TABOR), an amendment to the Colorado Constitution approved in 1992 (James & Wallis, 2004)

PAGE 232

218 has created additional barriers to the amount of money that Denver can set aside for its parks (Denver Parks and Recreation, 2003) Indeed, TABOR "limits the revenue growth from taxes that the City is allowed to retain" (Denver Parks and Recreation, 2003, p. 109) Thus, TABOR constrains the degree to which park funding can increase during times of econom ic growth because Denver, as any other Colorado local government, has to redistribute the increased revenues to taxpayers (Denver Park s and Recreation, 2003) On the other hand, during economic recession s budgets are often significantly cut (Denver Parks and Recreat ion, 2003) A s a former park planner explained, The city doesn't have the ability to set aside money every year to build a fund to improve parks or buy land. They can't save money year to year. The city needs to go to other partners like the Trust for Pub lic Land to get some funds. The case of TABOR shows that fiscally conservative policies have had significantly negative impacts on Denver's capacity to acquire parkland and to attend to park inequity issues, which contributes to the general picture of proc edural injustice that emerged in the analysis of park planning. To summarize, this detailed section on park funding has shown that, unsurprisingly, the availability of funding and how funding i s to be distributed has dominated the discourse on park planni ng in Denver. My analysis also showed that park funding is inherently political, as shown by capital improvement funding and bond issues. The amount of available park funding besides being linked to a city 's or state's economic health also reflects the v alue that city decision makers, the power elites, and vot ers attribute to parks, as highlighted by M ayor Speer. Finally, but not less importantly, the available funds for parks and how these funds are distributed as a reflecti on of political power can have significant consequences for park equity.

PAGE 233

219 The limited role of park master plans Another interesting theme that emerged from the analysis of Denver's park planning is the limited contribution of park master plans to actual park construction. In particular park master plans, in their various forms and in different historic periods, have provided Denver with frameworks for park establishment, including prioritizing certain areas of the city and specific types of parks. However, park master plans have had on ly a small role in acquiring parkland and building parks because, as discussed funding and political will have been key to park establishment. In other words, as a former park planner interviewee explained, "park plans don't build parks" because, like all city plans, they have "no legal enforcement, as policy is reflected only in budget." This section describes how park master plans have contributed to Denver's park system by defining visions, priorities, and strategies. Since its foundation, Denver has de veloped seven master plans for parks, six of which were comprehensive. Table 14 lists all seven master plans and describes their impact on Denver's park planning. The first master plan was developed very early in Denver's history, in 1878, while the latest plan dates to 2008. Also, the temporal density of park master plan has varied significantly. While four master plans were drafted between 1878 and 1929, there w as a gap of more than 50 years between the 1929 DeBoer plan and the 1986 plan commissioned (but not adopted) by M ayor Pe–a. After DeBoer's 1929 plan, the next master plan that City Council adopted was the Game Plan in 2003, with a gap of almost 75 years, reflect ing a disinterest in parks.

PAGE 234

220 Table 14 Master Plans for Parks in Denver Name Year Featur es Effects Sopris Lee Downing plan 1878 Two large parks (City Parks and Sloan's Lake) connected by a parkway Only City Park was built Evans Rollandet plan 1894 A comprehensive plan for parks and parkways, with a greenbelt and water Park Avenue and improv ements to City Park Robinson Kessler plan 1907 City wide park and parkway system, reaching every neighborhood: a "windmill" plan A blueprint for Denver's urban design, and large impact on park establishment DeBoer plan 1929 Comprehensive plan, including park, parkways and roads. Goal of brining parks to everyone A blueprint for Denver's urban development, including land use, building on 1907 plan Mayor Pe–a master plan (not adopted) 1986 A comprehensive plan to improve the park and parkway system Not ad opted: Due to limited participation, the plan did not address needs Game Plan 2003 A comprehensive plan to maintain and improve the park and parkway system New park provision standards; priorities for park investment; creative strategies for parkland acqu isition and funding Denver Play Area Master Plan 2008 A master plan for play areas within parks and accessible school grounds New play area standards; priorities for play area investment; creative strategies for funding Sources: (D enver Parks and Recreation, 2003, 2008; Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) and interviews with former park planners When analyzing the features of these plans, it is easy to identify a clear demarcation in terms of scope and content between the first four master plans and the last

PAGE 235

221 two master plans. The Sopris Lee Downing, Evans Rollandet, Robinson Kessler, and DeBoer plans are all physical plans that clearly delineate specific parks for different parts of the cit y (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) The content of these plans is expressed, in large part or exclusiv ely, through maps that describe the geographic location of existing and proposed parks and parkways (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) On the other hand, the Game Plan and the Denve r Play Area Master Plan are not physical plans and do not commit to establishing specific parks or play areas in Denver (Denver Parks and Recre ation, 2003, 2008) Rather, they define standards for park and play area provisions, they prioritize future park and play area investment s based on spatial analyses and surveys on park and play area needs, and they propose creative strategies to acquire parkland and to build play areas (Denver Parks and Recreation, 2003, 2008) Their content is expressed through reports, which include maps rep resenting spatial analyses but not committing to specific new parks (Denver Parks and Recreation, 2003, 2008) This reflects a shift of the pl anning profession from physical planning to policy and strategic planning. The available data on the 1986 master plan (Taylor, 1986, and one interview) did not allow me to determine whether it is a physical or priority setting plan. Besides the differences in scope and content, the master plans' effects on park establishment have also varied. As previou sly discussed, the implementation of both the Sopris Lee Downing plan and the Evans Rollandet plan clas hed against funding limitations (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) The first park master plan that significantly influenced park establishment in Denver, and that argu ably had the biggest impact on the city park system, was the

PAGE 236

222 Robinson Kessler plan (Etter, 1986; Noel & Norgren, 1987) The reasons why this plan succeeded are grounded in M ayor Speer's adherence to the City Beautiful ideals. Indeed, parks were integral to M ayor Speer's vision of a beautified Denver and conse quently he generously funded parkland acquisition and park construction (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) Similarly, DeBoer developed and implemented his 1929 master plan during a period in which the mayoral administration valued parks and supported park spending (Noel & Norgren, 1987) The succes s of the Robinson Kessler and DeBoer plans in guiding Denver's park establishment and urban development can be attributed, on one side, to the political and monetary support they enjoyed and, on the other side, to their physical planning scope. Indeed, the y placed on the map specific parks, parkways, roadways and public buildings that changed the image of the city (Etter, 1986; Noel & Norgren, 1987) These plans acted more like urban design plans (Robinson Kessler) and as comprehensive plans (DeBoer) than traditional park plans (Etter, 1986; Noel & Norgren, 1987) as their frameworks helped direct the growth of a city that was still largely undeveloped. After more than 50 years without a park master plan, M ayor Pe–a in the 1980s commissioned a new comprehensive vision for Denver's parks (E. A. Taylor, 1986) According to several sources cited by Taylor (1986) Parks and Recreation unt il the mid 1980s had been managing Denver's parks without a cohesive framework. The 1986 master plan was an attempt to create a new framework for Denver's park system (E. A. Taylor, 1986) However, according to two former Parks and Recreation's staff members, the plan was created without participation and it did not incorporate the recreational

PAGE 237

223 preferences of different demographic groups, and therefore it was never adopted by City Council. Thus, Denver lacked an adopted comprehensive master plan until 2003. The Game Plan, a spinoff from Denver's Comprehensive Plan 2000, is a general framework that prioritizes where park resources should be allocated based on a rigorous analysis of park need (Denver Parks and Recreation, 2003) as explain ed in the park geographic equality vs. park equity sect ion of this chapter. According to a park planner interviewee formerly working at Parks and Recreation, the Game Plan's impact on park establishment was very limited. Besides park plans, "policy, which means providing funding, and a strong leadership are al so necessary to establish new parks." During a period of limited park funding, Parks and Recreation needed to use words carefully in the Game Plan. According to a former Parks and Recreation staff member interviewee : The language used in plans also matters You don't want to create public expectations about specific parks you are not sure you can deliver. You can't put certain words in a plan even if the plan is not binding. You have to be careful not to pinpoint things, also to avoid spreading the voice th at the city is interested in a particular plot of land, as landowners could raise the price. This demarcates a strong difference between the Game Plan, and its companion plan Denver Play Area Master Plan, and the physical plans developed during the City Be autiful and New Deal eras. However, the strategies that the Game Plan adopted are appropriate for its historical period, as the 2000s saw a park funding shortage and a very competitive real estate market, which are opposite conditions to the ones when the Robinson Kessler and DeBoer plans were developed. Given funding limitation an d Denver's landlocked condition the Game Plan suggests a series of creative strategies to acquire parkland in established neighborhoods (Denver Parks and Recreation, 2003) These strategies include reprogramming public

PAGE 238

224 properties like schoolyars, utility corridors and ditches to create new accessible parks, and acquiring new large parcels of land, like industrial properties and large private educational campuses (Denver Parks and Recreatio n, 2003) The suggestion to acquire large parcel s of land is appropriate because establishing parks by acquiring small parcels piecemeal, as it was done in the past (Boone et al., 2009) would currently be almost impossible However, other people could be interested in these properties, according to a former park planner inte rviewee hence the Game Plan's strategy not to name them: The Game Plan suggested keeping an eye on these properties [large parcels]. The city would have to jump on these properties quickly if they became available, otherwise developers would acquire them. The problem is that yearly budgets must be used up and there is no easy financial or political tool (or political will) that enables the city to move quickly for acquisition. This statement shows that Denver's competitive real estate market not only means high prices clashing with Parks and Recreations' low budget for parkland acquisition, but it also means that desirable properties sell quickly, which clashes with the city being unable to act rapidly in the private market, given its bureaucratic limitatio ns. To summarize, master plans in Denver have shifted from physical plans providing a comprehensive vision for parks and urban design during the City Beautiful and New Deal eras to priority setting master plans in the 2000s that act as general frameworks but do not commit to establishing any specific parks. This shift coincided with, and derived from changes in available park funding, political will, land availability and diversification of special interest groups In terms of park equity, this shift mea ns that, currently, park master plans can only address park equity issues by setting priorities and suggesting strategies, but they can not initiate actions to solve inequities issues.

PAGE 239

225 Privately owned parks A recent trend of park planning in Denver is the establishment of privately owned parks and parkways in large Planned Unit Developments (PUDs) like Stapleton and Lowry. In particular, Stapleton's and Lowry's parks are organized based on what park planners call a "trunk system." According to a former Park s and Recreation staff member interviewee trunk systems include "large parks, paid by city tax dollars, and local smaller parks paid by developers and HOA [Home Owner Associations] fees." The trunk system allows Stapleton and Lowry to have excellent park provision in terms of park proximity, park acreage, and park quality, which developers have used to market these two projects. For this reason, the establishment of privately owned parks raises serious park equity issues. Figure 13 shows Stapleton's park s ystem, including large public parks and small privately owned parks and parkways. Figure 13 Stapleton's trunk system. Map developed using data from City and County of Denver (2014) and S tapleton MCA (n.d.)

PAGE 240

226 When asked why large PUDs adopted trunk systems, the park planners and land use planners I interviewed suggested three reasons. First, the trunk system allows developers and HOAs to keep control of the rules for using privately owned parks. As a planner interviewee formerly working at Community Planning and Development explained, "in some case they [developers and HOAs] want their parks to be for the residents and not for unwanted uses or visitors. For example, Forest City [the develop er] might be able to prevent homeless people from being in Stapleton's smaller parks." Thus, the possibility of controlling what are considered "appropriate public behaviors" (Langegger, 2012, p. 24) have motivated HOAs to keep ownership of these small parks. Second, the trunk system allows the city to save money on park maintenance, as HOAs maintain the smaller parks. Since small parks are inefficient to maintain, the city is keen on leaving these small parks to developers or HOAs. Third, the large amount of funding obtained by combining public and private money allowed the trunk systems in Stapleton and Lowry to achie ve higher park standards than the rest of the city in terms of park proximity, park acreage, and park quality. According to a Parks and Recreation park planner interviewee the small parks, which are paid by HOA money, make a difference in terms of park ac reage and park quality. The density and level of design of Stapleton's pocket parks are above standards. That probably would have been too much to maintain for the city. Some of these custom shade structures and playgrounds in Stapleton's parks are harder and more expensive to maintain. In other words, this interview passage shows that in Stapleton and Lowry HOA fees contribute to higher park standards through privately owned parks and parkways. This system raises equity concerns, especially if repeated ac ross the city in large PUDs. Indeed,

PAGE 241

227 residents of large infill developments with wealthy HOAs can decide to tax themselves to achieve better parks provision. "For example, South Stapleton alone has already four swimming pools that are owned by the HOA, whi le many Denver neighborhoods lack pools," as a former park planner interviewee explained. On the other hand, older and denser neighborhoods, often inhabited by low income ethnic minorities, do not have the financial capacity and the land availability to ac hieve the same provision. According to a park planner formerly at Parks and Recreation: This means that today more parks are buil t where the development money is i.e., where developers build large projects like Stapleton and Lowry. This almost means going back to 130 years when parks were donated by land developers. The developers' marketing and generous provision of parkland and amenities has led to high expectation from these residents (sometimes referred to city staff as "Entitleton.") The privatization of parks in Denver can be linked to a more general trend of public space privatization in the United States, including shopping malls, corporate plazas, and open space, which is described by a growing literature (e.g. Banerjee, 2001; Kohn, 2004; Langholz, Lassoie, Lee, & Chapman, 2000; NÂŽmeth, 2009, 2012) In particular, Kohn (2004) poin ted out that it is not uncommon for New Urbanist communities, like Stapleton and Lowry, to include privately owned parks and other open spaces. Kohn (2004) also questioned whether the same high provision of public space that New Urbanist developments provide is also available for those living in less expensive neighboring communities. Her question is legitimate because the neighborhoods that b order Stapleton and Lowry in Denver have a much lower l evel of park provision lower household incomes and highe r percentages of ethnic minorities than these two communities

PAGE 242

228 Another issue of privately owned parks is that such parks cannot be dedicated n or designated. Park dedication and park designation protect parks from being sold or converted to other uses, with designation being the strongest protection (Denver Parks and Recreation, 2009) When a park is designated as such, the city cannot sell it, swap it, or transform it to other uses unless Denver's voters approve this change (Denver Parks and Recreation, 2009) However, Denver's public parks and open spaces are not automatically designated. As some interviewees explained, Denverites have expressed concerns about park designation after the city sw apped part of Hentzell Park, located in southeast Denver, with Denver Public School s in 20 12. In particular, a former park planner interviewee suggested that "the Hentzel l controversy put some pressure on Parks and Recreation to designate new parks" and th at it reminded decision makers that "Denver citizens want their parks to be dedicated." As they are not owned by the city, it is clear that Stapleton's and Lowry's privately owned parks can neither be designated nor dedicated. To summarize, although privat ely owned parks provide excellent park provision in two Denver neighborhoods, and in Stapleton specifically, this form of park cannot be considered a solution to improve Denver's park system. Besides the inequity issues that privately owned parks raise, in cluding inequitable park provision and control of "appropriate public behaviors," privately owned parks do not match Denverites' preference for designated parkland. Parks and Recreation's current philosophy: Partnerships and re programming The extreme di fficulty in acquiring parkland, which is due to low park funding, the high cost of land in Denver, and to the lack of land use instruments to obtain

PAGE 243

229 land from developers (see land use section), led Denver Parks and Recreation to develop a lternative approac hes to improving park provision in the city. Parks and Recreation's philosophy includes two main strategies: leveraging partnerships to acquire parkland, and partially to improve parks; and reprogramming existing parks to increase their quality. The first strategy was well expressed by the words of a park planner interviewee at Parks and Recreation: "Our broad philosophy to establish new parks or to improve existing parks is leveraging partnerships with other departments and organizations." In particular, "to get land we [Parks and Recreation] strategize partnerships with other departments or non profit organizations." Partners include "other city departments, with the Trust for Public Land, with Colorado Open Lands, Great Outdoors Colorado, and the Colorad o Health Foundation, among others." The advantage of collaborating with other departments, like Community Planning and Development and Public Works, is being part of "larger projects that generally have a bigger budget, for example transportation projects. For instance Parks and Recreation is currently involved in the North Denver Cornerstone Collaborative (NDCC), an "organizational umbrella" that Denver has established to coordinate the planning efforts along the I 70 corridor in northeast Denver, inclu ding Public Works, Community Planning and Development, and the Colorado Department of Transportation. A park planner interviewee suggested that "maybe some new parkland could come out of this initiative." This example shows that Parks and Recreation is app ropriately capitalizing on partnerships with other city and state institutions, like the ones focusing on transportation, that have significantly higher yearly budgets. Partnerships with non profits are discu ssed in the next section.

PAGE 244

230 The second strategy is to reprogram existing parks in order to improve the quality of the parkland that is already owned by the city. The idea of reprogramming is an example of pragmat ic thinking that stems from recognizing the difficulty in acquiring parkland: "It is cost proh ibitive to buy new parkland, so in most recent cases we [Parks and Recreation] reprogrammed some of the parkland we already own," a s a Parks and Recreation park planner interviewee explained Another planner interviewee reinforced this concept: "many of ou r [Parks and Recreation's] efforts center on adding amenities to existing parks," which shows a commitment to improving the quality of existing parks. As an effect of this efforts, Denver Parks and Recreation's website lists several completed and ongoing p rojects of park improvement that were undertaken in the last 10 years (Denver Parks and Recreation, 2015a, 2015b) Among them, a park planner I interviewed point ed out the initiatives aimed to add amenities in Denver's greenways: We also got funding from the Greenway Foundation. We reprogramed underutilized park areas along greenways through these funds. An example of that is Paco Sanchez Park. That was an existin g park along the Lakewood Gulch Greenway, to which we will bring new features and new programs. Therefore, these efforts are focused on maximizing the potential of every publicly owned park and open space in Denver by implementing new amenities that can br ing new park visitors to areas that were previously underused. As I discuss in Chapter 6, these interventions have had a positive impact on the quality of parks located in low income Hispanic neighborhoods. To summarize, these two strategies that Denver P arks and Recreation is currently using are part of a pragmatic philosophy that acknowledges the current economic and political contexts and attempts to capitalize on every small opportunity to improve the

PAGE 245

231 city park system, including getting involved in lar ger planning projects. This means that, while Parks and Recreation ha s not completely give n up on acquiring new parkland, the department recognizes that park funding is currently better spent in park reprogramming projects, given Denver's high real estate prices. The role of non profit organizations In the last ten years, given limited public funding for large park improvements and parkland acquisition, Parks and Recreation has significantly relied on funding and services provided by local and national n on profit organizations to purchase land and to reprogram existing parks. As my informants pointed out, the non profit organizations that have provided the most support to Parks and Recreation have been Great Outdoors Colorado (GOCO), the Trust for Public Land (TPL), and the Greenway Foundation (TGF). Besides funding, non profit organizations have provided Denver with administrative services by directly purchasing the land for some of the most recent parks, thus speeding up the process of parkland acquisiti on. By every account, non profit organizations are currently h a v ing a significant impact on improv ing park provision in Denver, making up for a structural lack of funding for parkland acquisition and large park improvements Non profit organizations have b een particularly important in balancing some park inequity issues by helping the city establish new parks in park poor low income ethnic minority neighborhoods. Before analyzing in detail the role of non profit organizations in improving park provision, i t is important to introduce the scope and the mission of each of the three organizations I mentioned. Great Outdoors Colorado (GOCO) was established in 1992 thanks to an amendment to the Colorado Constitution (Great Outdoors Colorado, n.d. b) GOCO employs revenues fr om the Colorado Lottery to fund projects aimed to "help

PAGE 246

232 preserve, protect, enhance, and manage the state's wildlife, park, river, trail, and open space he ritage" (Great Outdoors Colorado, 2015, p. 6) GOCO aims to distribute funding approximately equally among th ese four categories: "wildlife, outdoor recreation, open space, and local governments" (Great Outdoors Colora do, 2015, p. 20) As a current GOCO staff member interviewee suggested, GOCO has traditionally focused its funding on acquiring land for open space, and has invested less money on urban parks for active use. Recently, GOCO has increased its focus on acqui ring parkland in cities: "We are looking for opportunities to invest in urban open spaces. The problem is that properties are much more expensive in urban areas than in rural areas." GOCO distributes different types of grants that are assigned on a competi tive basis (see discussion below for details). The Trust for Public Land (TPL) is a national non profit organization created in 1972 that focuses on establishing parks and preserving public land with the goal of fostering healthier and more livable cities in the long term (The Trust for Public Land, n.d.) With more than 30 offices located in several U.S. cities, including Denver, TPL helps "communities raise funds, conduct research and p lanning, acquire and protect land, and design and renovate parks, playgrounds, trails, and gardens" (The Trust for Public Land, n.d.) Thus, TPL does not directly provide funding to loca l governments, but it helps them fundraise. Also, TPL's mission is significantly more oriented to creating open space in cities than Great Outdoors Colorado. However, TPL Colorado's web page shows more projects in rural counties than in urban areas (The Trust for Public Land, 2015b) The Greenway Foundation (TGF) is a Denver based non profit organization that, starting from the 1970s, has focused on revitalizing the South Platte River and its tributaries by helping the city acquire land along waterways and transform it for public

PAGE 247

233 use (The Greenway Foundation, n.d.) The Greenway Foundation's budget derives from several sources, including grants, fundraising, and public money, and its specific composition varies year by year (The Greenway Foundation, 2010, 2 011) How non profit organizations distribute funding The se three non profit organizations have been using different mechanisms to distribute funding for parks. In brief, GOCO distributes funding from the Colorado Lottery among Colorado local governments based on competitive grants that are ranked based on a set of criteria (Great Outdoors Colorado, 2015) TPL does not directly provide funding but has helped Denver raise grants from other public institutions and non profit organizations (The Trust for Pub lic Land, 2012b) TGF distributes funding for several open space projects along waterways located in metropolitan Denver, including land acquisition, park or trail construction, and several outdoor education programs (The Greenway Foundation, n.d.) A search in The Greenway Foundation's website did not show competitive grant programs to distribute funding as for GOCO. The GOCO grants deserve a specific discussio n. GOCO has a biannual grant program that comprises grants for seven different types of projects, including parks and outdoor recreation, open space, trails, and school yards (Great Outdoors Colorado, n.d. a) Any Colorado local government can apply to these grants (Great Outdoors Colorado, n.d. a) As a curr ent GOCO staff member interviewee explained, each grant category has its own criteria, which GOCO uses to score grant applications: We score the applications based on some criteria that are included in our call for grants. We have some peer reviewers from across the state who score the applications. These are competitive applications, so the grants go to the applications that get the highest score.

PAGE 248

234 While criteria vary based on the grant category, the score system that GOCO uses to distribute funding is cent ered on the recreational needs and recreational preferences that communities have, and on how the project would serve recreational needs that are currently unmet. As a current GOCO staff member interviewee explained: A part of the score includes the descr iption of the community profile, which comprises the demographics, the recreational preferences of the community, and accessibility to these parks. Our score system also puts a large emphasis on the recreational needs that the community has. We want applic ants to explain how the project is going to serve a community that is in need of recreation. In particular, applications need to include evidence of recreational need through different means, including "recreation master plan identifying underserved areas, mapping of recreational resources, and planning processes that show that new parks are needed by the population." GOCO strongly values a bottom up approach and requires community engagement in both grant writing and planning the projects that receive awar ds. When asked whether GOCO grants specifically include equity or equality criteria, the GOCO staff member I interviewed responded the following: In our strategic plan we have identified the need to serve children and their families. We don't have somethin g specific on equity in our criteria, but if the application addresses equity issues, the peer reviewers generally give a high score on recreational need. That was the case of Cuatro Vientos Park in Westwood. Thus, not having criteria that explicitly menti on park equity does not mean that, in practice, equity is not considered as an important part of how grants are distributed. Also, the focus on "recreational needs" can be clearly related to equity, as low income communities with no or limited access to pr ivate recreation have higher needs of public recreation.

PAGE 249

235 Another important aspect of the GOCO grant system is the competitive nature of these grants. As explained by a GOCO staff member interviewee "every city and local authority in the state can compete for GOCO grants. So it's not guaranteed money you can count on." This means that, although Denver has been successful in obtaining GOCO grants on a yearly basis, GOGO grants are not a long term secure solution to balance a structural lack of funding for p arkland acquisition and for large park improvement. Denver Parks and Recreation (2008) in the Denver Play Area Master Plan expressed concerns about the city overreliance on one time grants, like GOCO's. In particular the Denver Play Area Master Plan (Denver Parks and Recreation, 2008) suggested that: One time grants and donations from foundations may be able to supplement a play area to fund a special feature or program but they cannot supply the capital required to build and maint ain the entire play area system. A long term sustainable funding source is required. (p. 136). This quote shows that Parks and Recreation, in a park master plan adopted by City Council, recognizes that grants and donations are not sustainable funding sourc es and recommends that the city should study how to establish a more consistent funding system for parks. The "sustainable funding" solutions that the Game Plan suggests include government funding (gas taxes, property taxes, or impact fees for developers), or the establishment of a non profit organization aimed to constantly raise money for Denver's park and parkway system (Denver Parks and Recreati on, 2008) One final aspect to consider when evaluating the GOCO grant system is that GOCO requires local governments to match the grant s they issue. To match grants and donations, as briefly mentioned, Parks and Recreation's capital improvement budget i ncludes "discretionary funds" that the department uses "to leverage partnerships and

PAGE 250

236 match grants." Thus, Parks and Recreation needs to plan in advance to have these funds available when GOCO grants are awarded. Specific impacts of non profit organization s Most park planners and land use planners I interviewed recognized the eminent importance of GOCO, TPL, and the Greenway Foundation to directly fund parkland acquisition, obtain funds for parks, and providing support in the process of parkland acquisitio n. As a land use planner interviewee suggested: "Great Outdoors Colorado is a huge source of money for Parks and Rec to acquire land and to build parks." A park planner interviewee at Parks and Recreat ion specifically highlighted that non profit organizati ons have been key to purchasing land for new parks: "The recent large park acquisitions have been done through the Trust for Public Land or Great Outdoors Colorado." As Parks and Recreation has been very successful in getting GOCO grants, a strong partners hip between GOCO and P arks and Recreation is emerging : "We [Parks and Recreation] got involved in their strategic planning, so there are good synergies. We need them as much as they need us." Although some informants pointed out the role of GOCO for parkla nd acquisition, the analysis of the organization's records showed that GOCO, since 1992, has contributed to funding parkland acquisition for four Denver parks: Montbello Open Space ELK Education Center, River North Park, Cuatro Vientos Park, and Camp Rolla ndet Open Space (Great Outdoors Colorado, 2014) Thus, GOCO has more frequently funded park reprogramming projects than new parks as explained by a staff member intervie wee : "Park improvement is something we've been involved in more often. To give you a general number, in Denver we've awarded around $22 million for park improvement, and

PAGE 251

237 around $5 for land acquisition." This connects to the Parks and Recreation's emphasis on park reprogramming, rather than on land acquisition. Among the park improvement projects, some "involved the creation of new parks on land that was already owned by the city. An example of that is the land along the South Platte River." This shows that, in some occasions, the city owns parcels of undeveloped land but lacks the funds to transform that land into parks. When asked if there are any specific park improvement projects that GOCO funds more often than others, the GOCO staff member I interviewed responded: We try to prioritize applications for facilities that serve diverse ages and diverse uses. Lastly, we started prioritizing projects that focus on youth and unstructured play spaces. This has been a trend statewide recently, so we followed that. It is more rare that we fund athletic fields, if they are the only project component. However, if a community demands a baseball field and shows a real need for it, we'll fund the baseball field. Community demand has been the ground line for what we fund. This interview passage highlights a few ideas. First, GOCO seeks to fund projects that can serve a variety of users within a community, not just a specific group. Second, the focus on funding nature based play derives from a growing request in the state f or nature play experiences. Third, GOCO recognizes that recreational needs vary by community and therefore tries not to impose a specific type of recreation through its grants, for example nature based play. This is important in terms of park equity, as se veral studies show differences in recreational preferences among different ethnic groups (e.g. Byrne & Wolch, 2009; Gobster, 2002; Loukaitou Sideris & Sideris, 2010) The Trust of Public Land's contribution to parkland acquisition and park improvement in Denver has also been very important, according to my informants. In particular, TPL has contributed to five major projects in the last decade, including

PAGE 252

238 parkland acquisitions and construc tion of the Montbello Open Space ELK Education Center, Cuatro Vientos Park, and Interstate Trucking Park (along the South Platt River); the establishment of the New Freedom Park (on city owned land); and improvements to Mestizo /Curtis Park (The Trust for Public Land, 2015b) Thus, TPL and GOCO collaborated in providing funding for acquiring the land of the Montbello Open Space ELK Education Center and of Cuatro Vientos Par k. In particular, the Trust for Public Land has made a key contribution for acquiring the land for the Cuatro Vientos Park in terms of administrative support. According to several informants one of the biggest barriers that Denver has faced whe n attempti ng to acquire land is the inability to act quickly enough when properties are available on the market. TPL has helped the city overcome this barrier by directly purchasing the land for Cuatro Vientos Park, as explained by a Parks and Recreation park planne r: To establish Cuatro Vientos Park, TPL bought the land for us [Parks and Recreation]. TPL can buy the land in a much more flexible and fast way than the city. TPL is a non profit. For the city it would take a longer process, which involves more bureaucra cy. We have to get these land acquisitions approved by City Council if we're spending public money. TPL has a much quicker way to operate in the market. This is key. This model is promising for future parkland acquisition. If this model was implemented on a large scale in the future, TPL and other non profit organizations could become even more important partners of Parks and Recreation. The positive impact of The Greenway Foundation (TGF) on establishing parks and regional trails along Denver's waterways also deserves attention. While park planners Saco DeBoer and George Cranmer during the City Beautiful and New Deal eras had made efforts to acquire land along the city's river, creeks and gulches (Noel &

PAGE 253

239 Norgren, 1987) the South Platte River and its tributaries in the 1960s and early 1970s were extremely polluted and mostly surrounded by industrial land uses (Leonard & Noel, 1990) Moved by the desire to reclaim Denver's waterways, The Greenway Foundation led a series of revitalization efforts and convinced the City and County of Denver of the importance of its waterway s (The Greenway Foundation, n.d.) In particular, in the last 40 years TGF has contributed to creating a large network of trails long Denver's waterways and to establishing more than 20 parks and open spaces, with the goals of ecological restoration, outdoor recreation, and economic growth (The Greenway Foundation, n.d. ) Impacts on park equity In the last decade, the work of Great Outdoors Colorado and of the Trust for Public Land has had a very positive impact on establishing new parks and improving existing parks in low income ethnic minority communities that were p reviously underserved by parks, thus contributing to park equity. As a former park planner interviewee explained: "In recent years, non profit organizations like GOCO and the Trust for Public Land have really helped balance equity issues in parks because o f their philosophical support and their additional funding." A park planner interviewee currently working at Parks and Recreation added: "GOCO's goals align with us [Parks and Recreation] in seeking to increase amenities in underserved neighborhoods." Acc ording to my informants, recent examples of collaborations between GOCO, TPL, and Parks and Recreation that were particularly oriented to solving park equity issues are Cuatro Vientos Park, New Freedom Park, and the Montbello Open Space ELK Education Cente r. In particular, Cuatro Vientos Park and New Freedom Park were established in two neighborhoods, Westwo od and East Colfax, which had very low park provision. The Montbello Open Space ELK Education Center seeks to cover the outdoor

PAGE 254

240 education and nature exp erience needs of a community, Montbello, which has a high percentage of low income ethnic minority young people. The story of Cuatro Vientos Park is particularly telling for the complexity of actions and for the multiple organizations that were necessary t o establish a park in a consolidated low income Hispanic neighborhood in West Denver. Besides Parks and Recreation, GOCO, and TPL, this project involved LiveWell Colorado, a non profit organization focusing on healthy living (LiveWell Colorado, 2013) and other neighborhood organizations. Cuatro Vientos Park is located in Westwood, which is one of the most impoverished area s of Denver, with a high percentage of Hispanics. The mechanism to create this park, besides TPL directly purchasing the land, involved a few complex steps, as described by a land use planner interviewee : To create this park, they intervened on a block fil led with very poorly maintained mobile homes, along Alameda [Avenue]. LiveWell Colorado worked with the people living in that block and found a better place to live for each of them. After that, they managed to establish the park. The Westwood Park [Cuatro Vientos Park] is an example of a non profit community development organization creating parks. Thus, creating Cuatro Vientos Park required relocating people living in extreme poverty. The work of a strong non profit organization, like LiveWell Colorado, a nd the will of the local community to establish a new park, made residential relocation possible. Also, this example shows that, even in a low income neighborhood with low property values, it is extremely hard to source land fo r parks, due to Denver's land locked condition This project also exemplifies TPL's park equity efforts, as expressed on TPL's website: "This collaboration with Denver Parks and Recreation is part of an ongoing partnership to

PAGE 255

241 create and improve parks in underserved, low income neighbor hoods throughout the city" (The Trust for Public Land, 2014) The establishment of New Freedom Park also deserves a brief discussion. Locat ed in the East Colfax neighborhood, the park serves an area with high population density and with a growing refugee population. Parks and Recreation identified half a block of undeveloped open space as a potential site for a new park. As for Cuatro Vientos Park, the project required the collaboration among several organizations, including: Denver Parks and Recreation, GOCO, TPL, the Colorado Health Foundation, and Denver Urban Gardens. As a Denver Urban Garden interviewee explained: The neighborhood and the Parks and Recreation department recognized the need of a new park there. We [Denver Urban Gardens] partnered with the Trust for Public Land and GOCO to create a new park. Everyone recognized the need to create a new park for a growing population of recent refugees living in that area. Thus, the idea of establishing a new park derived from the acknowledgement of an unmet need in terms of park provision, and from the recognition of the low income refugee community living in the area as a park dependent commu nity. T he park is "a cherished common ground" that brings "cultures together, as immigrant families put down roots in their new country" (The Trust fo r Public Land, 2012a) Finally, besides efforts to establish new parks in park poor low income ethnic minority neighborhoods, park and land uses planner interviewees mentioned that the Trust for Public Land has recently conducted some studies in metropo litan Denver to identify areas that are park deprived and park dependent As a GOCO staff member interviewee explained:

PAGE 256

242 Recently, we've helped fund a TPL's spatial park analysis of the whole metro area, including seven counties. The study was led by TPL. T his study focused on availability of open spaces, parks and trails in relation to schools and communities. These studies show that TPL and GOCO are aware of the larger park equity picture and that they do not only focus on solving specific issues, case by case. The impacts of these non profit organizations on park spatial di stribution are discussed in Chapter 6 Parkland acquisition and park establishment: A summary This section presented an in depth analysis of the mechanisms that have determined how and where parkland has been acquired and how parks have been established. As summarized in Table 12, parkland acquisition and park establishment involve interactions among economic interests, visions of equality or equity, evol ving public funding for parks, pa rk master plans, and recent trends related to limited public funding for parks, including: privately owned parks, Parks and Recreation's pragmatic approach to improving park provision, and the growing role of non profit organizations. The section also high lighted how changes in Denver's political system, in the value that the city elites attribute to parks, in the city's econom y and in the physical fabric of the city contributed to shifts in the availability of public funding for parks and in the city's ca pacity to establish new park s and to address inequities in park provision S ignificant changes in these factors have occurred from the City Beautiful era, considered as the golden age of Denver's park s to the post World War II era, which showed a diminish ed interest in parks and public space. Also, while during the most recent phase of Denver's history (1983 today) the local government and Denver's residents have shown a renewed interest in parks park establishment has been quite limited, with the excepti on of a few infill development s due to very limited annexations and to funding limitations.

PAGE 257

243 From an environmental justice perspective, this analysis show ed a complex picture of procedural inequity. Several episodes of open procedural injustice clearly em erged from the analysis, like the park district funding system, the use of publicly funded parks as a means to in crease property values, power imbalances in debates over bond issues, the recent reliance on privately owned parks, and others. Cases of proced ural injustice have been substantially more numerous than cases of procedural justice. The latter have mostly included federal or philanthropic interventions, such as funding efforts to establish recreation centers in low income ethnic minority areas and t he current work of non profit organizations. In Chapter 6, I compare these cases with quantitative findings describing distributional injustice and partial justice. However, this analysis also showed that given the complexity of park planning processes, it is difficult to conduct an analysis of procedural justice, which according to Schlosberg (2004) needs to be investigated in environmental justice studies. In other words, when analyzing more than 160 years of park planning it is hard to identify all processes that have been just, or moved by with equity oriented goals, and all process that have been unjust, or moved by the intention to discriminate or neglect underprivileged groups. Park Design, Management, and Use This section analyzes design and management p ractices of parks throughout Denver's history, as well as the way design and management have influenced park use, with a particular focus on low income ethnic minority young people Borrowing from Byrne and Wolch's (2009) Space, Race and Park Use conceptual model, in this analysis park design describes the physical definiti on of park space, including which amenities are

PAGE 258

244 included and how they are arranged in space. As previously mentioned, park design can embody specific cultural views of recreation, as historically most parks in American cities were designed based on norther n European canons of beauty (Byrne & Wolch, 2009) Park management, on the other hand, describes how official city rul es and de facto regulations have determined the "appropriate public behaviors" in parks (Langegger, 2012, p. 24) and in some cases, the appropriate users including different ethnic groups. The analysis of park design management, and use highlighted concepts that also emerged in the study of park establishment (see Table 15) In particular, the way parks are designed and managed has been influenced by the economic value that residents attribute to parks, available funding for parks, power, equity or exclusionary goals, different visions of recreation, and park ownership. In other words, as any public space, parks embod y conflicts between different economic interests, values, and visions, as well as power unbalances between different groups of society (Loukaitou Sideris & Ehrenfeucht, 2009; D. Mitchell, 1995, 2003; NÂŽmeth, 2006) My analysis of park planning will show that, i n Denver, young people and ethnic minority groups have typically been excluded from park design considerations, with a few exceptions. Table 15 presents the major themes and codes that can illuminate what has influenced park planning decisions in Denver in terms of park design and management. The three main themes include the environmental conflict over parks, with a specific focus on conflicts between competing visions of recreation, ethnic discriminations in park design and management, and park safety as a result of park patrolling and park design (see Table 15). The rest of this section is organized based on these three themes.

PAGE 259

245 Table 15 Themes and Codes from the Analysis of Park Planning: Park Design and Management Theme Description Codes Parks as cont ested public spaces With population growth, and different interests over parks, a series of competing interests battle to define what should be included in Denver's parks. Children are rarely taken into account in these conflicts. Power has played a role i n park design and in defining the "appropriate uses" that are allowed Park purposes include: aesthetic/passive recreation; active sport use; social use; productive garden; and ecological parks Territoriality over parks: "insiders" vs. "outsiders" Eth nic discrimination in park design and management Ethnic minorities have been discriminated against in park design and management: From forms of open discrimination like exclusion from swimming pools and other sport facilities (until 1960s), to more subtle forms of discrimination dealing with defining "appropriate uses" (until 2000s) Ethnic discriminations in defining "appropriate use" De facto park segregation and discrimination in 1910s 1960s D iscrimination against Hispanics in NW Denver, 1970s 2000s Safety in parks: Patrolling and park design Patrolling and law enforcement in parks have diminished in the last 30 40 years, with different levels of patrolling today depending on the park (regional parks receive more resources). A few aspects of park de sign can also influence park safety Denver parks had park police and a maintenance person on site until 1980s Today, park rangers and park maintenance crews, but not in every park Gardens, group sports, and social gatherings provide "eyes on the park Parks as contested public spaces Denver's parks, l ike parks in many other U.S. cities, have been the object of environmental conflicts about their design and regulations throughout the city's history. Environmental conflicts over public space occur w hen multiple parties attempt to attribute competing uses to the same space, thus requiring a

PAGE 260

246 negotiation among parties (Loukaitou Sideris & Ehrenfeucht, 2009) Embedded in environmental conflicts is the idea that public space is a scarce good, thus the willingness of different actors to influence its design and regulations (Loukaitou Sideris & Ehrenfeucht, 2009) Power im balances and eco nomic interests often drive the resolution of environmental conflicts (S. Campbell, 1996) Recently, with Denver's population rapidly growing and different demographic groups demanding a variety of park amenities, the pressure on Denver's parks has significantly increased. P opu lation growth and the city's incapacity to add new significant parkland to dense neighborhoods has led to parks becoming highly contested space s with several park purposes and visions, pushed by different groups, competing to find room in Denver's parks. Figure 14 shows the purposes that, based on my analysis of park planning are currently conflicting for physical space in Denver's parks and specific conflicts occurring among them which involve significant trade offs The park purposes included in Figure 14 emphasize park aesthet ics and historic preservation vision s active group recreation, food production, social gatherings, and the ecological preservation. Unfortunately, young people and ethnic minority groups are rarely taken into account in the disco urse about these conflicts and in their resolution, due to their limited power in decision making.

PAGE 261

247 Figure 14 Environmental conflict on park design and management The idea of parks as visual amenities, which emphasize the ir aesthetic value, originated wi th Denver's most historic parks, which were developed before and during the City Beautiful era. This park purpose emphasizes passive recreation and is grounded in the idea of parks as idyllic pastoral landscapes, with large lawns surrounded by tree canopy, which Frederick Law Olmsted introduced to the United States (Byrne & Wolch,

PAGE 262

248 2009) For example, City Park was initial ly designed in the 1880s based on Olmsted's pastoral visions (E. A. Taylor, 1986) and in the 1910s his son, Frederick Law Olmsted Jr., was involved in the re design of the park (Noel & Norg ren, 1987) Denver's City Beautiful park designs canons included a strong presence of water through lakes and ponds, an unusual and soothing elements in a dry climate, large meadows covered with bluegrass, and extensive views of the Rocky Mountains as a b ackdrop (Etter, 1986; Leonard & Noel, 1990; Noel & Norgren, 1987) As noted in Chapter 2, this pastoral park vision that pri vileges the contemplation of an idyllic natural landscape mostly adheres to White European canons, and might not reflect the recreational preferences of ethnic minorit y people (Byrne & Wolch, 2009) Also, with the exception of the Playground Movement (Noel & Norgren, 1987) the discourse surrounding City Beautiful idyllic parks very rarely has included young people. The second general park purpose th at emerged in my analysis is active group recreation. Spaces and activities that fit this purpose are sport fields, sport courts, and swimming pools. As reported by Langegger (2012) and by some of my informants, Denver parks are heavily used for group sports, including organized sport activities like little leagues and adult leagues. Park space for soccer games is particularly needed (Langegger, 2012) S ports fields and organized sports activities like L ittle L eague games are particularly valued by ethnic minorit y people especially by His panic children and teenagers (Loukaitou Sideris & Sideris, 2010; Perry et al., 2011; Ries et al., 2008) The third park purpose is conceiving parks as productive landscapes that include gardens for food pr oduction. With Denver increasing its focus on locally grown food, as shown by the growth of community gardening (Denver Urban Gardens, 2012) parks can

PAGE 263

249 provide significant amount of land for food production. According to a Denver Urban Garden staff member interviewee Parks and Recreation is increasingly interested in establishing community gardens in parks "for two main reasons: activating under utilized areas of parks, and the stewardship that gardens promote." Community gardens located in parks can also "help bring people from different socio economic groups together," as shown by a community garden in St apleton's Greenway Park. The fourth park purpose that emerged from my analysis is programming parks as spaces for social gatherings, including small group parties with families and friends and large festivals and events. According to several experts I inte rviewed in Denver, as in many other U.S. cities (Byrne & Wolch, 2009) Hispanics tend to highly value spaces for soci al gathering and family parties in parks. This preference might derive from the public spaces that can be found in Latin American cities. According to Langegger (2013) Mexican plazas, often referred as z —calos, are a hybrid between European parks and squares and are flexible spaces for gathering, bei ng able to host a variety of functions: parties, community meetings, play space for children, and market. Hispanics can very rarely find these types of places in the United States (Langegger, 2013) This idea of public space for recreation is very different from Non Hispanic White's views of parks as places to escape from the city and experi ence nature (Langegger, 2013) Finally, the fifth general purpose that emerged from my analysis is parks as places prioritizing ecology and native vegetation. This conception can also have economic advantages in terms of lowered park maintenance costs. According to park planners and a Denver Urban Gardens member I interviewed, a recent tende ncy in park design in Denver (since 2000s) has been the partial shift from water needy Kentucky bluegrass, typical of

PAGE 264

250 Denver's City Beautiful parks, to natural landscapes with native prairie. Park planners explained that the advantage of natural landscapes it that they require significantly less water than bluegrass. Also, these landscapes work well for stormwater management, and help create habitat for other native species. This strategy was suggested in the Game Plan (Denver Parks and Recreation, 2003) and implemented quite largely in Stapleton. Parks and Recreation is working on a city wide "program for turf conversion, to turn bluegra ss into native prairie grass," according to a park planner. Fina lly, natural landscapes along waterways can also create opportunities for unstructured natural play for children (Derr & Lance, 2012; Rigolon, 2013) although no informant openly discussed this specific point. Specific conflicts on park purposes One of the most significant conflicts between park purp oses that emerged in my analysis of park planning is the conflict between park aesthetics and large sport activities. As several informants pointed out, heavy sport use contributes to grass deterioration, and people living around parks complain about. In p articular, a Denver Urban Garden member interviewee explained: The clash of interests over parks is a real one. There are the heavy users, like soccer teams and other sports (both kids and adults). There is the aesthetic piece. Sunny flat fields are requir ed for sport fields, and are also ideal for gardens. Once the fields get overused, and the grass starts to turn yellow, I imag in e the neighbors calling the park department to tell them about the status of the grass. This passage highlights one of the reas ons of the conflict between park aesthetics and heavy sport use. Residents living around parks and parkways, especially in wealthy areas, tend to be strongly attached to the image of a green lawn in front of their house. Throughout Denver's history, there are several accounts of Non Hispanic White wealthy residents living in front of parks fighting against changes to green idyllic meadows, as shown in City Park (Noel & Norgren, 1987) Cheesman Park (E. A. Taylor, 1986)

PAGE 265

251 Stapleton, and several parkways (experts' interviews). This shows that some Denver residents have conceived parks as an extension of their front yard, on which they feel entitled to have a louder voice than other citizens wh o do not live close to "their" park. This case of territoriality over park s is also linked to the ir economic value, as a well maintained green meadow in the park across the street is significantly more appealing to realtors and buyers than an overused socc er field without grass. The location of large sport complexes within parks has caused Not In My Back Yard (NIMBY) oppositions. In particular, according to a park planner interviewee formerly working at Parks and Recreation, residents of Lowry have opposed without succeeding, the construction of a series of sport fields in Lowry Sport Complex Park: Neighbors often oppose larger, more active recreational facilities near them they can have bright night lighting, bring in traffic, and people beyond the neig hborhood. For example, we hurried to plan and start building the Lowry sports complex before there were immediate neighbors to ward off the opposition and NIMBY attitude that we anticipated. Another case of NIMBY opposition involving the conflict between p ark aesthetic and historic preservation, on one side, and active recreation, on the other side, is the City Loop/Re Imagine Play project, a large playground for multiple ages that was previously discussed. This project, initially planned for City Park, was moved to Paco Sanchez Park, a smaller and less historic park in the West Side, after strong opposition from residents living around City Park. In particular, project opponents were concerned about City Loop being a regional recreational attraction that wo uld have brought increased traffic volumes to the area and that would have endangered City Park's historic character, according to a few current and former Parks and Recreation park planners. Coverage from the local

PAGE 266

252 press confirms this battle to keep City Park as a quiet, neighbor hood park, which the park neighbors perceive as their park (Booth, 2014; Prendergast, 2013) The conflict between park aesthetics and group sport activities can influence young people's opportunities for play in parks. Indeed, chi ldren and teenagers, and especially low income ethnic minority children, are some of the most frequent participants to organized sport activities in public parks, according to a survey reported in the Game Plan (Denver Parks and Recreation, 2003) Also, using adults' aesthetic preference s to design parks might limit children's attachment to parks, as children tend to look for flexible o pen spaces they can transform (Chawla et al., 2 014) rather than rigid clean cut landscapes. Similarly, adopting a landscape design that adheres exclusively to the European tradition can be considered as an imposition by ethnic minorit y group s, like Hispanics and African Americans. A second signific ant conflict is occurring between park aesthetics and community gardens. While Parks and Recreation is increasingly interested in establishing community gardens in underused areas of parks, gardens are relatively rare in Denver's parks, as shown by Denver Urban Gardens's (2015) active garden list. A Denver Urban Gardens interviewee gave two reasons why gardens are not commonly located in Denve r's parks: gardens look messy and not aesthetically pleasing to some, and parks have a lot of pressure from multiple groups. In particular: The first reason is that, in the past and still now to some degree, there is the perception that gardens can be a bi t of an eyesore. They can look a little messy to non garden participants. We try to make gardens more aesthetically appealing. [] The second reason is that in Denver parks are overrun by multiple interests. Neighbors want a pristine place, sport teams wan t parks to play their leagues, people want to use parks for festivals and parties There are too many users for a single space. It's like the Colorado River, there's not enough water for everyone.

PAGE 267

253 Most parks tend to be un programmed, they can host multiple uses in the same space (playing Frisbee, soccer, a picnic). A garden might become a fixed piece that some people might not like, because they could see the garden as benefitting only the gardeners and not the whole community. This interview passage shows, on one hand, that community gardens can create a visual disturbance to people valuing neat park aesthetics. Also, the passage highlights the larger conflict over park space in Denver, with "multiple interests" competing for physical space. As community ga rdens are spaces with a specific purpose and limited access, som e people perceive them as places benefitting only certain groups. The implementation of natural landscapes with native prairie grass in Denver's parks can also conflict with the historic pre servation and aesthetic values. As a few park planners pointed out, historic preservationists and advocates of the City Beautiful aesthetics are against turf conversion to native landscapes. The use of native grass highlights conflicts between environmenta l goals and aesthetic values, as a former Parks and Recreation staff interviewee explained: "It's an emerging ethic that is meaningful to many people but still perceived by some as scruffy land." In particular, as another former park planner interviewee po inted out: "There's a perception issue with natural landscapes. Some people think that natural prairie landscapes here are ugly. People often want blue grass in parks to match their front lawn." This interview passage shows, once again, that some people pe rceive parks and parkways as an extension of their front yard. Other park planner interviewees have reported that some Denverites are strongly attached to City Beautiful aesthetics, which limit s the implementation of natural landscapes. A Parks and Recreat ion planner interviewee explained that "the historical nature of Denver's parks and parkways also influences the aesthetics that we implement

PAGE 268

254 in parks today." Also, according to a former park planner interviewee "to some people, Denver's City Beautiful la ndscapes are sacred." These two interview passages show that the City Beautiful heritage significantly limit s the type s of amenities that can be included in Denver's historical parks with negative consequences for young people's play To summarize, while other environmental conflicts are currently occurring over Denver's parks (see Figure 14), most of my informants stressed the relevance of conflicts between aesthetic and historic preservationist values, on one side, and a series of other park purposes th at, according to preservationists, endanger City Beautiful aesthetics. These conflicts between park aesthetics and other more active uses also highlight issues of power and territoriality, as people living close to some of Denver's most historical parks an d parkways seem to perceive these green spaces as an extension of their front yard. Also, the strong emphasis on City Beautiful aesthetics that is still influencing park design in Denver might limit young people's play opportunities in parks, as shown by C ity Loop/Re Imagine Play and other projects. In addition, the City Beautiful aesthetic canons do not fit all Denver's ethnic groups. While Denver's historical parks were established at a time when Denver was predominantly Non Hispanic White ( before the 192 0s), Denver's demographics are very different today, with ethnic minorities comprising almost 50 percent of Denver's population (United State s Census Bureau, 2013b) Therefore, my analysis of park planning processes shows that Denver's parks should better respond to different ethnic groups' views of recreation. Territoriality over parks. Environmental conflicts over Denver's parks also involve issues of territoriality between the resident s of a neighborhood, referred to as

PAGE 269

255 "insiders," and individuals and groups from other areas visiting the neighborhood park, referred to as "outsiders." Territoriality over public space describes individuals' or groups' efforts to gain control over a specific spatial realm, thus informally determining who can use that public space and what behaviors are allowed in that public space (Karrholm, 2007) The spatial realm that becomes the object of informal control is referred to as territory (Karrholm, 2007) Territoriality has o ften involved conflicts among people of different ethnicit ies claiming the same space (Peters, Elands, & Buijs, 2010) In particular, accounts from several experts I interviewed highlight th at in recent and current times, Denver parks have experienced conflicts over park use between Non Hispanic White affluent groups and low income ethnic minority groups. According to a former Parks and Recreation staff interviewee wealthy Non Hispanic White residents of Lowry and Curtis Park have complained about outsiders using their parks : Some Lowry playgrounds had early conflicts between "outside" families using the playgrounds. People don't articulate it but from an observer, there are racial, economic implications. [] When I was working on Curtis Park, the white newcomers were not happy about Spanish speaking adults coming to the park on Sundays and playing robust soccer, eating, and drinking there. This shows the conflict over the use of public space who belongs and who doesn't. T hese issues of territoriality further show a narrow view of parks as something that the neighborhood "owns," not that the city as a whole owns. The examples of Lowry and Curtis park also s how that territoriality often runs a long ethnic and social class lines, and that episodes of informal discrimination in parks are still occurring in Denver. Non Hispanic White's territorial control over parks in Denver has also been enhanced by design. For example, Stapleton, a middle and upper class predominantly Non Hispanic White neighborhood was intentionally disconnected from the low income

PAGE 270

256 ethnic minority neighborhoods that surround it through large arterial roads (Duffy et al., 2010) Therefore, this lack of physical connection makes it harder for "outsiders" to enjoy Stapleton's parks that, as discussed, have very high standards. Furthermore, creating a privately owned park is an even stronger way to assert territoriality. A s homeowners associations in Stapleton and L owry own the neighborhoods' pocket parks, they can define park designs and management rules that fit the vision of the demographically homogeneous groups who live in these neighborhoods thus keeping unwanted users a way from privately owned parks. Ethnic d iscrimination in park design and management Several park design and management practices in Denver's distant and recent past have, intentionally and unintentionally, discriminated against ethnic minorities, particularly against African Americans and Hispa nics. Discriminatory practices in park design and management can be classified in two main phases: A first phase of open ethnic discrimination spanning from the 1900s until the 1960s; and a second phase of more subtle ethnic discrimination spanning from th e 1970s until the 2000s. Open discrimination: 1900s 1960s Several scholars report ed episodes of de facto discrimination in park use for African Americans, in larger measure, Hispanics, and other ethnic minorities until the 1960s (Abbott, 1978; Leonard & Noel, 1990; Newsum, 2012; R. L. Simmons & Simmons, 1995) Many historians and park planners I interviewed also pointed out Denver's history of discrimination in recreation Discrimination, often expressed through segregation practices, applied mostly to swimming pools and beaches, but also to other public recreation facilities like tennis courts and golf courses (Abbott, 1978; Leonard & Noel, 1990; Newsum, 2012; R. L. Simmons & Simmons, 1995)

PAGE 271

257 African Americans were particularly targeted by ethnic discrimination in recreation during this phase (Abbott, 1978; Leonard & Noel, 1990; Newsum, 2012; R. L. Simmons & Simmons, 1995) as they were Denver's largest ethnic minority until the 1950s. According to Abbott (1978) the "Denver Parks Department forbad them [African Americans] the use of either Washington Park or Berkeley Park swimming pools, while allowing them Wednesdays at the downtown pool on Curtis Street (p. 254). Although other accounts describe forms of de facto segregation instead of de jure segregation (Leonard & Noel, 1990) the effect was still the exclusion of African Americans from several public recreational facilities. In some cases, African Americans protested against segregation in recreation. In 1932, a group of African Americans entered Washington Park, which was unofficially off limits for them, and swam in one of the park's lakes (Leonard & Noel, 1990) A large group of whites attempted to chase the group of African Americ ans away from the water, armed with rudimentary weapons (Leonard & Noel, 1990) Instead of pr o secuting white violence, the police arrested several African Americans (Leonard & Noel, 1990) This episode shows that, at a time when the Ku Klux Klan was ruling Denver's politics, public authorities not only tolerated but also contr ibuted to enforcing discrimination in recreation. Subtle discrimination: 1970s 2000s Between the 1970s and the 2000s, more subtle forms of discrimination in park design and management occurred against Hispanics living in northwest Denver (Goldstein, 2011; Langegger, 2012, 2013; Leonard & Noel, 1990) In this analysis, I borrow extensiv ely from the work of Langegger (2012, 2013) as well as from an interview with a historian. One form of discrimination dealt with park management. According to a historian interviewee the Non Hispanic White

PAGE 272

258 city elite perpetuated discriminations through park management b y defining and enforcing "normal" uses of public space that reflected their culture, but that significantly differ from Hispanics' use of public space. Thus defining what are the "normal," or appropriate uses is a subtle way of discriminating against who does not fit in this definition of "normal." Another subtle r form of discrimination involved park design. In some cases, Denver Parks and Recreation made physical changes to parks layouts that, intentionally or unintentionally, discouraged Hispanics' use o f these parks (Goldstein, 2011; Langegger, 2012, 2013; Leonard & Noel, 1990) The cases of Columbus Park, St. Patrick Park, and Berkeley Park in northwestern Denver show examples of park design and management practices that openly or subtly discriminated against Hispanics. La Raza Park The history of Columbus Park or La Raza Park, as t he local Hispanic population used to call it, can be considered as a case of open discrimination against Hispanics in parks. In the 1970s, Columb us/La Raza Park, located in Sunnyside became the center of a growing Hispanic community by providing a place f or events, informal gatherings, and community identity (Goldstein, 2011; Langegger, 2013) In particular, Hispanic youth from the Crusa de for Justice took control of La Raza Park and its pool and used it for their cultural and political activities (Langegger, 2012, 2013) Local residents, in particular, heavily used the park's pool during the summer mont hs (Langegger, 2012, 2013) Denver's police tried to oppose th is informal appropriation of public space and started enforcing the 11 P.M. curfew that applied to all Denver's parks (Goldstein, 2011; Langegger, 2012, 2013; Leonard & Noel, 1990) The tension between the police and the local Hispanic population exploded in the summer of 1981 when a riot occurred at La Raza Park (Goldstein, 2011; Langegger, 2012,

PAGE 273

259 2013; Leonard & Noel, 1990) While the local Hispanic population was celebrating the anniversary of Chicano youth taking over the park, the police urged people to leave the park because no formal permit to hold a large gathering had been secured (Goldstein, 2011; Langegger, 2012, 2013) As the celebration had been held for several years the local Hispanics initially refused to leave the park, and the police started to charge them, thus contributing to the riot (Goldstein, 2011; Langegger, 2012, 2013) Riot squads were called to disperse the Hispanic residents from the park (Goldstein, 2011; Langegger, 2012, 2013; Leonard & Noel, 1990) After the riot, Denver Parks and Recreation used the damages created during the riot as a rationale to shut down the pool by filling it with concrete, and to introduce parking regulations around the park that made it harder for people to access the park (Langegger, 2012, 2013) According to Langegger (2013) closing down the pool contributed to less and less people using the park, and to the local Hispani c population getting disenfranchised from the park and the neighborhood. In the case of La Raza Park, Denver Police and Parks and Recreation made management and design decisions that were clearly against Hispanics, and intended to prevent them from using the park the way they were used to. As a Denver's historian I interviewed suggested: In the case of the La Raza Park riot, the police were looking for a fight. It was a party. The idea of a normal use of a park in the Latino world and white world just cl ashed. [After the riot,] filling La Raza Park's pool delegitimized the use of the park for Latinos. The city came in and said "we're condemning this pool." This is complete exclusion. For the pool, the Latinos were basically shut out. The city came in and said "you're done with this park now." This interview passage shows that, due to divergent views of recreation between Hispanics and the dominant Non Hispanic White city elite, Hispanics were openly

PAGE 274

260 excluded from La Raza Park through eliminati on of the par k amenity they valued the most. Also, a fter the park was altered, racial profiling against Hispanic youth became more and more common (Langegger, 2013) S ome of the local residents interviewed by Langegger (2013) suggested that the increase of gang activity in the 1980s was partially due to Hispanic youth's reduced access to outdoor recreation at La Raza Park. St. Patri ck Park L ocated in the Highland neighborhood, St. Patrick Park provides another example of recent discrimination against Hispanics in parks, although not as open as in the case of La Raza Park St. Patrick Park was established in 1984 through a community led effort, whic h included the acquisition of a parking lot located in front of St. Patrick's church (Langegger, 2012) During the 1980s and 1990s, St. Patrick Park was a successful pocket park, with local Hispanic children and their families using the park frequently (Langegger, 2012) In 1999, to comply with the Americans with Disabilities Ac t (ADA) standards, Parks and Recreation redesigned the park, including its playground (Langegger, 2012) These design changes contributed to lower park use, as child ren tended to perceive the playground as less interesting (Langegger, 2012) After these changes, and with the Highland neighborhood gentrifying, a once active z—cal o like park lost its Hispanic character and users (Langegger, 2012) In the case of St. Patrick Park, the playground renovation to fulfill ADA regulations and other small design changes were not implemented openly to hurt the Hispanic community. However, these changes did have a negative effect on these communities (Langegger, 20 12) According to Langegger (2012) due to the lack of obvious discriminatory intentions, the Hispanics' reduced use of St. Patrick Park can be considered an "unintended, collateral consequence" of the design changes that were

PAGE 275

261 implemented (p. 248). However, the local Hispanic population was not involved in the redesign of the park (Langegger, 2012) If they had been, these negative consequences might not have occurred. Berkeley Park The case of Berkeley Park, located in the Berkeley neighborhood, can be considered as an example of relat ively open discrimination against Hispanic youth, involving both park management and park design. Berkeley Park until the early 2000s included a few soccer fields that were mostly used by Hispanic youth and adults (Langegger, 2012) The soccer fields in Berkeley Park were problematic because local youth used them year round, thus making it impossible for grass to grow there (Langegger, 2012) Parks and Recreation expressed concerns about the state of the grass in the early 2000s (Langegger, 2012) T o prevent people from using the soccer fields year round, Parks and Recreation established movable goal posts instead of fixed posts, and moved the posts away from the field during winter months (Langegger, 2012) Thus, Parks and Recreation prioritized aesthetic goals over group sports, with the effect of reducing Hispanic youth's opportunities to play soccer. In the early 2000s, after the movable goal posts were installed, Denver Parks and Recreation invested some money in Berkeley Park, but not to improve the fields where local Hispanic youth used to play (Langegger, 20 12) Rather, Parks and Recreation decided to spend the available funds to create an off leash dog park and eliminated the soccer fields (Langegger, 2012) This chan ge in park design is somehow surprising, as dog parks in Denver, including the one implemented in Berkeley Park, did not stem from a strong local requests for such spaces (Langegger, 2012)

PAGE 276

26 2 Establishing a dog park and taking away soccer fields from local Hispanic youth meant robbing them of a public space that they felt as theirs, that they used frequently (Langegger, 2012) According to Langegger (2012) city officials are responsible for this deci sion and used power to implement this shift in park design without asking the local long time residents their opinion about the soccer fields. Not every staff member of Denver Parks and Recreation agreed with this decision. As a park planner interviewee fo rmerly working at Parks and Recreation explained: Public opinion will always be divided over to what parks should be used for. For me, it is sometimes hard to justify the creation of dog parks, even though they are a legitimate use of parkland, when there are several areas of the city without a playground. Thus, it is unclear why dog parks were prioritized over spaces where young people could play. In the case of Berkeley Park, the low income Hispanic youth who used the soccer fields lost one of their few p ublic recreational opportunities. This is problematic for another reason. The long term residents of northwestern Denver that Langegger (2012) interviewed reported that the elimination of these soccer fields could have contributed to increasing Hispanic youth's gang membership in Berkeley According to Langegger (2012) Denver authorities did not understand that, by taking away recreational opportunities to northwest Denver's Hispanic youth, the city failed to provide activities that can be an alter native to gang membership. Thus, these changes implemented in Berkeley Park are problematic because they disadvantaged the weakest group, low income Hispanic young people. Also, this case raises serious equity issues because the design changes were decide d and implemented

PAGE 277

263 with a top down approach and not based on needs that local residents had expressed and because of their impact on low income Hispanic young people's play opportunities To summarize, the analysis of the design and management changes mad e in the three parks located in northwest Denver show ed subtle forms of discrimination. E thnic discrimination in recreation w as still evident in the early 2000s. In all three cases, interventions by Parks and Recreation disadvantaged Hispanics, and especia lly Hispanic youth. In two cases, La Raza/Columbus Park and Berkeley Park, the design and management changes were clearly made against Hispanic youth. In the case of St. Patrick Park, discriminatory intentions are not evident, although according to Langegger (2012) the city failed to get Hispanic residents involved in the redesign of the park. Park safety The analysis of park planning showed that a few park management and design practices can influence the level of active and passive supervision in Denver's parks and, consequently, the perception of s afety. Park management has contributed to the active supervision of parks thro ugh park patrolling, including park maintenance crews, a dedicated park police, and park rangers. P ark design has contributed to park passive supervision by providing spaces that encourage park users to access parks constantly, thus providing "eyes on the park." After the 1980s, cuts to Park s and Recreation's budget and im balances in park patrolling funding have led to diminished patrolling in some of Denver's parks, with negative consequences for public safety in parks. Park management and safety Park man agement practices in Denver have included formal and informal park patrolling. A few former park planner interviewees suggested that Denver parks, until the 1980s, had higher levels of patrolling then today. In particular, a former park planner interviewee explained that:

PAGE 278

264 Originally and until the 1980s, there was a park maintenance person who was on site for each park. That person was someone who everybody knew, who you could trust to watch your kids. Until around 1986, there was also a park police. Now we have park rangers. This interview passage suggests that the presence of a local park maintenance person increased people's sense of safety in park s as caregivers could feel comfortable letting their children play in parks. The presence of a specific park police shows that, until the 1980s, the city put more emphasis on an d resources in to providing active park patrolling. According to a former Parks and Recreation staff member, the reduced spending in active patrolling can be linked to 10 percent cuts to Pa rks and Recreation's operating budget the 1980s, due to the local economic recession. Today, park patrolling through police and maintenance crews has significantly reduced but not completely disappeared. From a few informal observations of Denver's parks, I noted that Washington Park and City Park include police pavilions, which increase park patrolling. Also, dedicated park maintenance facilities can be found in Washington Park, City Park, Sloan's Lake, Berkeley Park, Rocky Mountain Lake Park, Central Par k, and Ruby Hill Park, which are all large regional parks. This shows that the largest and most important parks still include some permanent form of patrolling. Currently, besides the maintenance and police pavilions I mentioned, Parks and Recreation is ru nning a park ranger program that is aimed to "maximize public safety, protect park resources, and provide service to park visitors" (Denver Parks and Recreation, 2006) Started as a pilot program in 2002 (Denver Parks and Recreation, 2003) and f ully established in 2006 (Prendergast, 2014) the park rangers program was intended to respond, among other needs, to an increased demand for park safety (Denver

PAGE 279

265 Parks and Recreation, 2003) Park rangers patrol Denver's parks i n vehicles and often collaborate with Denver's Police Departmen t in law enforcement in parks (Denver Parks and Recreation, 2006) The program has grown from including only two full time park rangers in 2006 to seven in 2014 and to an annual budget around $700,000 in the same year (Prendergast, 2014) While the park ranger program has grown, anecdotal evidence reported in Prendergast (2014) suggests that park rangers might not patrol the park s that, due to higher crime rates, need them the most. For example, in 2014 one seventh of the park ranger budget was spent to patrol Washington Park during the summer, especially to enforce alcohol laws for large parties on weekends (Prendergast, 2014) Washington Park is one of Denver's most histo ric and prestigious parks, and mostly Non Hispanic White affluent people live around the park This disproportion ate concentration of resources i n Washington Park caused protests from neighborhood organization s in other parts of Denver that have parks with biggest safety problems (Prendergast, 2014) This brief analysis of park patrolli ng in Denver shows that there has been a decrease in park patrolling from the 1980s to today, due to cuts to Parks and Recreation's operating budget. Also, although some parks still includ e permanent maintenance crews and maintenance facilities and a park ranger program has been established, the current distribution of these patrolling resources raises important equity issues as Denver's largest and most prestigious parks have received the highest share. Park design and safety The analysis of Denver's park planning also highlighted that a few park design features can encourage repeated and prolonged park use, thus increasing passive supervision through park users' "eyes on the park." In pa rticular, the

PAGE 280

266 experts I interviewed reported that community gardens, spaces for group sports, and spaces for social gathering have been linked to increased levels of passive supervision in parks (see Figure 14). In particular, according to a Denver Urban G ardens interviewee community gardens in parks, besides increasing park visitation due to garden maintenance needs, can increase gardeners' place attachment to parks: When a garden is in the park, the place making capacity of gardens might transfer to the rest of the park. Neighbors who work in gardens feel empowered because they feel ownership over the space. The place stewardship fostered by gardens can increase passive supervision and sense of safety. Parks and Recreation is aware of this benefit that c ommunity gardens can provide and is attempting to establish more gardens in Denver 's parks. Also, a few park planners pointed out that sport fields, especially those with organized activities and spaces for social gathering like picnic areas and watering holes can provide additional "eyes on the park" by increasing park visitation based on specific schedules and traditions. For example, a park planner interviewee formerly working at Parks and Recreation remembered successful implementations of picnic areas that were constantly used by Hispanic and African American families during weekends. Thus when attempting to solve environmental conflict over parks, park planners should consider that community gardens, spaces for group sports, and spaces for social gat hering can increase park passive supervision and pe ople's sense of safety. Park Planning in Denver: A Brief Summary This extensive analysis of park planning in Denver focused on two main areas: parkland acquisition and park establishment, which deal with w here parks are located in the city (see Table 12) and park design, management and use, which describe what

PAGE 281

267 amenities can be found in parks and what behaviors are considered as appropriate in them (see Table 15) These analyses showed a complex landscape o f procedural inequity, with low income ethnic minorities and young people typically excluded from park planning. The two analyses also showed that some larger concepts inherently permeate planning for both park establishment and park design and management, including: the economic value that residents attribute to parks, the intrinsic value that residents and elites attributed to parks, available funding for parks, power, equity or exclusionary goals, different visions of recreation, and park ownership. In t he final section of this chapter, I discuss how these larger concepts and values of park planning have interacted with land use planning and housing policies and practices in creating park provision for various demographic groups during different periods o f Denver's history. Land Use Planning The analysis of Denver's land use planning history can contribute to a better understanding of the processes that led to today's park spatial distribution in relation to the location of different demographic groups. T h is analysis based on interviews and secondary data, depicts processes related to park establishment and residential location. In terms of park establishment, I discuss the processes through which land use planning instruments have favored or limited land acquisition for parks and park funding through "cash in lieu of land" policies. I n some cases, cities give developers the possibility to pay an impact fee for parks instead of donating land when creating a new subdivision or an infill development (Baden & Coursey, 1998) Also, I describe the way la nd use, including zoning and subdivision regulations, contributed to the residential location of different income and ethnic groups throughout Denver's history.

PAGE 282

268 Land Use Planning and Park Establishment In t his section I discuss how land use planning has f ostered or limited parkland acquisition and park establishment in Denver through land donation or cash in lieu of land funding. Some of the policies that are part of this discussion were created by City Council and mayors but have significant impacts on th e degree to which land use planning can facilitate parkland acquisition. Thus, this analysis further highlights the political imprint on parkland acquisition and park funding, as well as economic mechanisms driving land use decisions and municipal regulati ons. The practical contribution of land use planning to park establishment and park funding in Denver has significantly differe d based on the geographic context: annexations including subdivision processes and infill developments. Table 16 describes the ma in land use planning instruments that Denver has used to acquire parkland or collect cash in lieu of land in these two geographical contexts. Denver has been much more effective in acquiring parkland or obtaining cash in lieu of land fees in annexations th an in infill developments. Also, as the Poundstone Amendment has limit ed Denver's capacity to annex land from its surrounding counties since 1974, and as the city lacks a park impact fee for infill development, the contribution of land use planning to park land acquisition and park funding after 1974 has been very limited, with a few exceptions. In addition, other citywide land use planning instruments, like zoning ordinances and comprehensive plans, did not have a significant impact on parkland acquisition or park funding. The limited cont ribution of land use planning to acquiring land or cash for parks has had negative consequences for park equity, as it has restricted Denver's ability to establish large parks in park poor low income ethnic minority areas.

PAGE 283

269 Table 16 The Land Use Planning Contribution on Parkland Acquisition in Two Geographic Contexts Annexation: Subdivision Process Instrument or Lack Of Years Effect Land donation requirement (8 percent) or cash equivalent 1940s 1957 Significant annexations to the City and County of Denver, including residential areas with parks Land donation requirement (8 percent) and $2,000 fee per acre of residential land 1957 1961 Extremely limited annexation to the City and County of Denver, with no residential land a nnexed Land donation requirement (8 to 12 percent) or cash equivalent 1961 1974 Significant annexations to the City and County of Denver until 1974 (Poundstone Amendment). Several parks were created Land donation requirements based on projected populat ion needs 1988 The Gateway DIA area was annexed in 1988, and several parks in far northeast Denver were created Infill Developments Instrument or Lack Of Years Effect Lack of impact fee for parks in infill developments 1980s (or earlier) 2015 No pa rk funding from infill development, added pressure to existing parks due to increased population, and equity issues Contracts between Denver and developers for PUDs with publicly owned land (Stapleton and Lowry) 1990s 2000s Significant acreage of public p arks and of privately owned parks created in Stapleton and Lowry (trunk system) General Development Plans (GDPs) require 10 percent of open space, which does not mean publicly owned parks 1980s (or earlier) 2015 Mixed used infill developments above 2 3 acres include some open space (plazas, green spaces, and trails), which can be privately owned

PAGE 284

270 Annexations and park establishment The subdivision process that occurred when Denver annexed land from neighboring counties is one of the main ways that De nver acquire d land for parks, according to land use planners I interviewed. City wide and area specific subdivision regulations for greenfield developments have included clauses that required developers to transfer a certain percentage of land for public u se to the city, including land for parks, schools, police and fire stations, or to pay a cash in lieu of land fee. According to a planner interviewee formerly working at Commu nity Planning and Development developers generally prefer to donate land in gree nfield developments, as they have significant amounts of available land. The requirements in terms of land or cash have changed a few times, with various effects on the amount of land that was annexed and on the park acreage that was established. The analy sis of the different annexation requirements is mostly based on the publications by Maxine Kurtz (1957, 1962) who held leadership positions at Denver's planning office from the late 1940s until the 1960s, and on interviews with land use planners and park planners. Annexation policies in the 1950s 1960s The first annexation policy I found evidence of was in place between the 1940s and approximately 1957, as explained in Kurtz (1957) This policy required developers of land that was ann exed to the City and County of Denver to donate eight percent of the annexed land to Denver for public use, including parks, schools and other uses, or cash in lieu of land based on land value (American Planning Association, 1958; Kurtz, 1957) Under this regulation, significant amounts of land were annexed to Denver (approximately 2,700 acres in 4 years), as the regulation appealed to developers for its lack of annexation fees (Kurtz, 1957)

PAGE 285

271 In 1957, Denver made changes to the annexation policy: Besides requiring land donation or cash in lieu of land (eight percent of the annexed land), the city decided charge developers a fee of $2 ,000 for each acre of residential land that was annexed (Kurtz, 1957) This regulation had the effect of almost stopping land annexation between 1957 and 1961, as only 91 acres were annexed to Denver, which included no residential land uses (Kurtz, 1962) Given the failure of the policy implemented in 1957, Denver decided to change the land and cash requirement in a new annexation policy developed in 1961 (Kurtz, 1962) After conducting a few studies, the Denver Planning Office decided to eliminate the $2,000 annexation fee and to require developers to transfer eight to 12 percent of the annexed land to Denver for public use, or cash in lieu of land based on land value (Kurtz, 1962) This new policy, which only increased the amount of required land compared to the policy in place until 1957, had a positive impact on annexation, as mo re than 3,800 acres were annexed to Denver between 1962 and 1963 (Kurtz, 1962) T hanks to these policies, annexations to the City and County of Denver between the 1940s and 1974 (Poundstone Amendment) led to significant parkland acquisitions for Denver, as shown in Figure 15. As a park planner interviewee formerly working at Parks and Recreation explained, "when a new neighborhood [subdivision] was built, generally a neighborhood park was included." In the post World War II years until 1974, land annexations were mostly made in the southeastern, southwestern, and northeaster parts of th e city (see Figure 15). Therefore, these parts of Denver have relatively good park provision.

PAGE 286

272 Figure 15 Annexation to the City and County of Denver and park establishment The Poundstone Amendment and Gateway Denver's significant land annexations and park establishment came to a halt in 1974, when Colorado voters approved the Poundstone Amendment (Leonard & Noel, 1990; Romero, 2013; E. A. Taylor, 1986) According to several land use planners I interviewed, the Poundstone Amendment was detrimental to parkland acquisition because it stopped the use of annexation policies, which had the most effective instrument that Denver used to obtain parks. Besides halting the acquisition of "free" parkland through annexations, the Poundstone Amendment made Denver's land a finite resource, thus increasing land values and making it harder for the city to compete against private interests when acquiring land (E. A. Taylor, 1986) Between 1974 and 2015, only a portion of Adams

PAGE 287

273 County was annexed to Denver in 1988, including the Gateway area and the land for Denver International Airport (Johnston, 2011) A land use planner interviewee explained th at for the Gateway area (far nort heast Denver), the city created specific subdivision regulations to obtain parkl and through annexation : The way we [Denver] and most Colorado communities obtain new park space was by writing a requirement into the subdivision regulations. When we wrote the Gateway subdivision regulations, we said that each developer must contribute X acres of parkland and Y acres of school land, in relation to the number of residents and students we expected to live there. It was an index based on the expected populatio n need. This is the process that was generally used for raw land, for greenfield development. Thus, the specific annexation regulation developed for Gateway in the 1980s calculated the amount of public land for parks and schools that developers needed to d onate based on the expected population, and not as a simple portion of the annexed land. While a formula was in place, "the property owner likes to negotiate," according to the land use planner interviewee mentioned above. "They [property owners, developer s] have lots of land available and little money, they wanted to negotiate. The developer is often willing to give up more parkland if the city is willing to negotiate more flexibility in housing." This willing ness to donate more land for public can be rela ted to the availability of large amounts of land in greenfield developments. Infill developments: barriers to park establishment Given the Poundstone Amendment, annexations to the City and County of Denver will be very limited in the future, as they have been in the last 40 years. Therefore, it is important to focus on what are the instruments, or better, what instruments the city is lacking to obtain parkland or cash in lieu of land in infill developments. As I mentioned earlier, most experts I interviewe d pointed out that for Denver it has been extremely difficult, if not impossible,

PAGE 288

274 to obtain land for public parks or funding for park improvement from developers in infill development projects. The exceptions to this general trend are Stapleton and Lowry, which are special cases because they are built on land that was previously publicly owned. Lack of an impact fee for parks The lack of an impact fee for parks in infill development projects is currently one of the biggest reasons of the shortage of fundi ng for park establishment and park improvement in Denver, according to several land use and park planners I interviewed. Impact fees are charges that cities or counties impose on developments to collect funding for public services intended to benefit these developments (Nelson, 1988) According to the Denver Municipal Code (City and County of Denver, 2000b) with the exception of an impact fee applied in the Gateway area in 2000, Denver lacks an impact fee for parks. The impac t fee in the Gateway area was intended to pay for c apital i mprovement p rojects, including parks (City and County of Denver, 2000b) Thus developers building infill developments of any size are not legally required to contribute monetarily for park maintenance or improvement. Many other cities in the United States have required developers to pay impact fees for parks (Baden & Coursey, 1998; Houtman, 2012; Nelson, 1988) including municipalities in metropolitan Denver like Aurora (City of Aurora, 2015) and Littleton (City of Littleton, 2013) According to a land use planner interviewee formerly at Community Planning and Development, the impact fees system makes sense in small infill developments: For 8 townhomes, it wouldn't make sense for the city to ask for open sp ace because the amount of land obtained would be too small to be useful. As a result, cities often use a cash in lieu of land approach for smaller infill redevelopment projects. Developers generally pay development impact fees, which can be spent for park construction in the neighborhood.

PAGE 289

275 Many land use and park planners I interviewed identified the lack of an impact fee for parks as one of the major issues for lack of park funding in Denver. As this issue emerged from the first interviews, I decided not to ask specific questions about impact fees until the end of the interview to see how many experts would mention this issue: Seven of the ten land use and park planners I interviewed pointed at the lack of an impact fee for parks as one of the major issues fo r Denver's park system today. In particular, the lack of impact fees affects Parks and Recreation's capacity to acquire new parkland and, in general, its financial power. A park planner interviewee currently working at Parks and Recreation indicated that the department is a strong advocate of impact fees for parks, but that policy decisions concerning impact fees are generally made at a higher level: At Parks and Rec, we strongly advocate for an impact fee. We make the case for it every time we can. I know that the Community Planning and Development department is also in favor of an impact fee. Decisions about impact fees are made at a much higher level. The mayor needs to be on board. I can tell you that Parks and Rec really advocates for it. One of the ba rriers to the implementation of a n impact fee for parks is that city officials might be concerned about impact fees slowing down development. As a planner interviewee at Community Planning and Development explained: "It's critical to have an impact fee, bu t for developers it's already expensive to come to Denver and develop. It's mostly about brownfield development at the moment, so it's expensive." In other words, the lack of an impact fee for parks can be linked to Denver's landlocked condition, as infill developments are more expensive than greenfield developments. However, a few park planners argued that, given the current economic and population growth, an impact fee would be reasonable and appropriate. "With today's

PAGE 290

276 booming economy, you would expect t hat the city could add such a fee," a former Parks and Recreation staff interviewee argued. In addition, "although Denver population is booming, the city doesn't seem to be concerned about a lack of impact fee, or at least they're not taking direct action to introduce one." This quote highlights another important aspect: The lack of an impact fee or land donation in infill developments has serious consequences on park maintenance. According to a park planner interviewee : The lack of an impact fee or of land requirement for infill developments has an impact on existing parks. The new population living and working in the infill development creates a pressure on neighboring parks. That means that Parks and Rec needs to invest more on these parks for park mainte nance, and with no budget increase, that can become a problem. The lack of an impact fee for parks also has significantly negative consequences on park equity, as only neighborhoods with affluent homeowner associations and privately owned parks can afford to pay for good park maintenance. Also, limited public funding for parks strongly limits Parks and Recreation's capacity to address park inequity issues. According to a park planner formerly working at Parks and Recreation: Only places like Stapleton and L owry, with strong HOAs, get parks. And these parks are generally on private lands managed and owned by their HOAs. Also, some poor neighborhoods (Globeville, Swansea) have land that would be available for parks, but that land is contaminated by heavy indus trial use, and it would need to be cleaned. To summarize, the lack of an impact fee for parks in infill developments can be considered one of the biggest issues for park planning in Denver today. Specifically, the lack of impact fees for park s creates fund ing issues for Parks and Recreation that have repercussions in terms of parkland acquisition, park establishment, park improvement, park maintenance, and park equity. With the construction boom currently occurring in Denver, the city is missing a great opp ortunity to obtain consistent funding for parks.

PAGE 291

277 Large PUDs on public land: Contractual agreements For large Planned Unit Developments (PUDs) that are locate d on previously public land (city or federal ly owned), parks and open space are included in a priv ate contract between the City and County of Denver and the developer. This has been the case of Stapleton and Lowry, which were developed on land that was previously owned by the city (Stapleton Airport) and by the U.S. Army (Lowry Base). In other wo rds, S tapleton and Lowry are an exception to the general policy of not requiring parkland or cash in lieu of land in infill developments because they were developed on land that was previously publicly owned. As a park planner interviewee formerly working at Par ks and Recreation explained: There are exceptions to these mechanisms, like in Stapleton and Lowry, in which large parts of land were transferred to the city for parks. These projects were built on land that was previously municipal and federal land, so pu blic institutions had the power to negotiate with the developer. In these cases, as a land use planner interviewee previously at Community Planning and Development pointed out, "there's a negotiation between city and developer that leads to a contract. Wi th large developments like Stapleton and Lowry, you have a private contract." Public land ownership gave Denver a position of power in negotiating with the developer the establishment of public parks. In Stapleton and Lowry, negotiations led to significant acreage of public parks and of privately owned parks However, other factors have also contributed to Stapleton's and Lowry's significant park acreage, including: strong advocacy from the Stapleton Foundation, a non profit which led the efforts to create a blueprint for Stapleton, the Green Book (Duffy et al., 2010) ; urban design vision s based on New Urbanism and therefore valuing parks (Leccese & McCormick, 2003; Leccese, 2005) ; and developers intending to use

PAGE 292

278 parks to market these new neighborhoods, as explained by Duffy et al. (2010) and by some of the experts I interviewed. Overall, Stapleton and Lowry are positive exceptions to the general failur e to acquire parkland in infill development through land use instruments. Their exceptionality mostly derives from public land ownership prior to development, as well as from other factors. If other public land will become available due to the dismissal of large public facilities, the model used in Stapleton and Lowry could lead to significant park acreage added to Denver's park system. However, establishing parks, including privately owned parks, in relatively upscale New Urbanist development exacerbates p ark inequity issues The section on housing policies and practices elucidates the partial failure of affordable housing in Stapleton, which contributed to Stapleton being a relatively homogeneous middle and upper class neighborhood. GDPs open space requir ement General Development Plans (GDPs) are implemented for infill developments larger than 2 3 acres and including mixed uses, and mostly focus on the layout of streets, open space and public utilities (Denver Community P lanning and Development, 2005) Denver's policy for General Development Plans (GDPs) requires developers to set aside 10 percent of the net land surface for publicly accessible open space (Denver Community Planning and De velopment, 2005) This open space is not necessarily publicly owned, as the policy states that "open space may not be dedicated to or maintained by the City and County of Denver" (Denver Community Planning and Development, 2005, p. 30) As a few land use planners I interviewed reported developers and HOAs often maintain ownership over these open spaces. In this case, the policy for GDPs sta tes that

PAGE 293

279 there needs to be an organization which is able to maintain this open space in the long term (Denver Community Planning and Development, 2005) Also, developers have the option to create different types of open s pace, including "parks, plazas, or trails," according to a planner interviewee at Community Planning and Development Thus Denver's policy does not guarantee that in each GDP the city will acquire dedicated parkland, or that the developer will provide pri vately owned parks for public use. Several GDPs that Denver is currently working on do not include parks, but rather provide open space through privately owned plazas and wider sidewalks (Denver Community Planning and Development, 2015) The South Sloan's Lake and the 9th and Colorado GDPs well exempli fy this trend (Denver Community Planning and Development, 2012, 2014) To summarize, the policy for Denver's GDPs is not an effective instrument to acquire new dedicated parkla nd in infill developments, although it can help established privately owned plazas and trails. Zoning and comprehensive plans T he role of zoning ordinances and comprehensive plan s in parkland acquisition or obtaining park funding has been rather limited, if at all identifiable. According to a planner interviewee at Community Planning and Development, the department "can't create an impact fee for parks through zoning" because, as previously mentioned, decisions about impact fees are made at a higher level including City Council and the mayor. The planner also pointed out that zoning codes in Denver have not generally included requirements for public parkland. However, the recent form based zoning code (2010) comprises some special zoning districts that in clude requirements for privately owned open space, like plazas and small pocket parks. According to a planner interviewee at Community Planning and Development:

PAGE 294

280 We can create some setback rules and we can require the space resulting from setbacks as open s pace. This was done recently in the Cherry Creek Zoning District. This requires portions of the site to be turned into little pocket parks or plazas. [] It's a very urban, very compact form of development. That open space will be privately owned but gener ally accessible. Thus, more and more land use planning instrument s are requiring, or encouraging, the establishment of privately owned open spaces, including the GDP policy and a few special zoning district s This is in line with Parks and Recreation's gen eral reluctance to acquire small pocket parks, and with developers and HOAs aiming to keep control over privately owned open space by defining their own regulations (see park planning section). The diffusion of publicly owned parks and open spaces has nega tive consequences on park equity in terms of park location and park management, as previously discussed. Comprehensive plans, as park master plans, are general frameworks that define directions of growth and priorities, but they do not act as policies beca use they do not include funding to implement specific interventions. Thus, it is difficult to evaluate the impact of comprehensive plans on park establishment. However, a difference can be identified in the general approach to park equity between the 1978 Comprehensive Plan and the two most recent comprehensive planning instruments, the 2000 Comprehensive Plan and Blueprint Denver (2002). According to Palmisano (2014) while the 1978 Comprehensive Plan was centered on equity, the 2000 Comprehensive Plan Blueprint Denver duo focused on economic growth by attracting the creative class to Denver. In particula r, the 1978 Comprehensive Plan called for "quality public facilities appropriately located and equitably distributed to all segments of the population" (Denver Planning Office, 1978, p. 57) The plan's section on public facilities identifie d parks as key amenities for Denver's neighborhoods, and epecially for low income neighborhoods

PAGE 295

281 (Denver Planning Office, 1978) On the other hand, the 2000 Comprehensive Plan and Blue Print Denver did not include this specific focus on equitable distribution of public facilit ies, bur rather focused on incentivizing private investment in some strategic areas of the city (City and County of Denver, 2000a, 2002) Alt hough different visions are clear, it is hard to recognize the specific impacts that the 1978 Comprehensive Plan and that the 2000 Comprehensive Plan Blue Print Denver duo have had on park establishment, or in particular on equity oriented parkland acquisi tion and park improvement. To summarize, this analysis of land use planning's contribution to parkland acquisition and park funding showed that, unsurprisingly, Denver has been significantly more successful in acquiring parkland or obtaining funding for pa rks in annexations than in infill developments. However, the Poundstone Amendment has virtually stopp ed Denver's capacity to acquire parkland through annexations since 1974, with the exception of the Gateway DIA area. Also, the lack of an impact fee for pa rks in infill developments is one of the main reasons for the structural sho rtage of park funding. Thus with a few exceptions (Stapleton and Lowry), the impact of land use planning on acquiring parkland and collecting park funding after 1974 has been very limited. Land Use Planning and Residential Location While it is important to understand how land use planning contributed to the location of parks and to how the city has collected funding for parks, it is also crucial to analyze whether land use planning has contributed to the residential location of different income and ethnic groups in Denver, with a particular focus on locations that have good and poor park provision. As residential land value influences where different income groups can afford to liv e it is important to study how land use planning influenced land

PAGE 296

282 values in different parts of Denver. My analysis showed that distance to locally unwanted land uses (LULUs) and housing types have significantly contributed to residential land values in Denv er. This section describes how municipal land use planning instruments like zoning ordinances and developer driven subdivision regulations have influenced which housing types have been allowed in different parts of the city and, in part, the location of LU LUs. The analysis of land use planning's influence on the residential location of different demographic groups mostly highlighted discrimination based on economic grounds. S ubdivision or zoning regulations that limit housing types to large single family h omes on oversized lots thus excluding multi family housing types, have been used as means to exclude low socio economic groups from certain areas of cities (Hartnett, 1993; Liberty, 2002; Maantay, 2001; Pendall, 2000; Ritzdorf, 1985) Therefore forms of exclusionary land use, such as exclusionary zoning, have been successfu l in their goal of creating residential segregation by income within cities and metropolitan areas (Maantay, 2001) While exclusionary land use policie s have used economics as a means of discriminating, a few scholars have suggested that, in many cases, income has been used as a proxy for race or ethnicity (Cole, 2014; Nelson et al., 2004; Pendall, 2000) What drove land value in Denver Before analyzing the different land use instruments that have influenced the location of different housing types and LULUs, it is important to briefly discuss what are the factors that have been driving land values in Denver. As a scholar interviewee who has studied planning in Denver sugge sted, "land value is the primary mechanism of residential segregation in the city." Although openly racist housing practices have existed (see housing section), many of the experts I

PAGE 297

283 interviewed pointed out that land value has been one of the main drivers of the residential location of different socio economic groups in Denver, and of their segregation. The interviews also highlighted that the main drivers of residential land value Denver have been distance to locally unwanted land uses (LULUs) and housing types. Other factors have included topographic elevation, access to clean water, the proximity to the most prestigious parks and parkways, and the ethnic composition of the neighborhood. As previously discussed, the construction of the Denver Pacific Railw ay in northeast Denver combined with the establishment of industrial activities in the area in the late 1880s contributed to lowering the value of residential properties and to affluent Non Hispanic Whites leaving the neighborhood for areas located in east and south Denver (R. L. Simmons & Simmons, 1995) Thus, the establishment of the railway and the associated industries is one of the biggest reasons why north Denver has traditionally been a less desirable place to live th an east and south Denver. The impact of LULUs on property values and, consequently, on residential segregation based on economics has be en documented for several other U.S. cities (Been & Gupta, 1997; Been, 1994; Greenberg & Hughes, 1992; Schively, 2007) In terms of housing type multi family housing units have been often considered as a nuisance in affluent neighborhoods (Cole, 2014; Leonard & Noel, 1990; Park Hill Action Committee, 1962) In particular, multi family housing units have been seen as a factor detracting from the family friendly character of neighborhoods until at least the 1980s, as th e more transient population of renters, including students, singles and non traditional families, did not always fit in the mainstream vision of the middle class American dream (Cole, 2014) Thus, several exclusionary land use policies and practices

PAGE 298

284 in Denver have attempted and managed to separate single family housing from multi family housing (see Table 17). The topographical elevations of certain neighborhoods contr ibuted to their status. As a scholar interviewee who has studied planning in Denver suggested, "being on higher ground is advantageous for flooding reasons, for views, but also because areas on higher grounds have less pollution." This is ref lected on the names of some wealthy neighborhoods, which "have hill in their name: Park Hill, Capitol Hill, and Hilltop." Another factor that has contributed to land value in Denver is access to clean water. A Denver historian interviewee pointed out that "water plannin g policy also contributed to segregation in Denver. The newest, best water infrastructure was built in places that were unaffordable and inaccessible for communities of color." Also, the proximity to the city's most prestigious parks and parkways can be co nnected to increased property values. As another Denver historian interviewee pointed out, Denver's historic parkways in particular transmit the idea of a "respectful, prosperous neighborhood." Finally, homogeneous Non Hispanic White neighborhoods have bee n considered as more desirable than neighborhoods with ethnic minorities, thus fostering higher property values (Cole, 2014; Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010) Exclusionary land use planning instruments In t his section I discuss how land use planning instruments have contributed to the establishment of different housing types in several neighborhoods, with a particular focus on exclusionary instruments that were used to maintain single family homogeneity. While I did not find any specific evidence of land use planning instruments that openly aimed to exclude low income or ethnic minority people from living in proximity to parks, some exclusionary land use policies

PAGE 299

285 and practices implemented in Denver prescribed large singl e family housing units around Denver's most historical parks and parkways, thus excluding low income people from living around these green spaces Table 17 summarizes the land use instruments that have contributed to excluding certain housing types, and by consequence certain demographics, from some of Denver's neighborhoods, comprising developer driven subdivision regulations and zoning ordinances. Unsurprisingly, the instruments described in Table 17 were mostly implemented in some of Denver's wealthiest and least ethnically diverse neighborhoods and have contributed to establishing their status. Exclusionary subdivision regulations Before the institution of the first zoning code in 1925, some private developers put in place a series of exclusionary subdi vision regulations to help establish the status of new subdivisions by excluding low income ethnic minority people These exclusionary land use instruments were mostly implemented in streetcar suburbs like Montclair, Park Hill, City Park West, South Denver Highlands, Berkeley, and Garden Park (Dorward, 2010; Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010) In particular, the earliest example of these regulations were found in the 1890s in Park Hill (Leonard & Noel, 1990) and similar regulations were still implemented in the 1940s in Garden Park (T. H. Simmons & Simmons, 2010) Most of these regulations prescribed minimum lot size, minimum house size, minimum construction cost, minimum setbacks, and aesthetic rules (Dorward, 2010; Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010)

PAGE 300

286 Table 17 Ex clusionary Land Use Planning Instruments in Subdivisions and Zoning Ordinances Subdivisions: Private Regulations Instrument Years Description Subdivision regulations requiring large expensive houses on large lots 1890s 1940s Regulations prescribing minim um lot size, minimum house size, minimum construction cost size, setback rules, and rules regulating aesthetics Subdivision regulations excluding multi family housing and commercial activities 1890s 1910s Regulations excluding any type of multi family ho using and vice ordinances excluding commercial activities to create "family friendly" neighborhoods Zoning Ordinances Instrument Years Description Residence A district in 1925 zoning code 1925 1956 Zoning district prescribing exclusively single family h ousing with high minimum lot sizes R 0 district in 1956 zoning code the "living in sin" ordinance 1956 1989 (2010) Zoning district prescribing exclusively single family housing with high minimum lot sizes and prohibiting unrelated people from living tog ether For example, the Park Hill Syndicate, which purchased a large part of the Park Hill land before Park Hill was incorporated in Denver, established multiple exclusionary land use rules (Dorward, 2010 ) Houses had to be built on two parcel lots, they had to have a certain setback, and they had to have a minimum construction cost, depending on the area (Dorward, 2010; Leonard & Noel, 1990) In particular, developers used the wider lots, 50 feet instead of the standard 25 feet, to promote the vision of an exclusive

PAGE 301

287 neighborhood with large mansions where only middle and upper class families could afford to live (Leonard & Noel, 1990) Besides prescribing a series of minimum standards, some of the most upscale subdivisions also excluded housing types and s pecific land uses that were considered as a nuisance (Leonard & Noel, 1990) In particular, subdivision regulations prohibited multi family housing units and any commercial or industrial activity (Leonard & Noel, 1990) These regulations excluding unwanted land uses, and excluding unwanted people, were an attempt to create idy llic neighborhoods without the temptations and the dangers of dowtown, where children and families could be physically and morally safe (Leonard & Noel, 1990) These early versions of American suburbia successfully attracted middle and upper class residents (Dorward, 2010; Leonard & Noel, 1990) Many of the exclusive subdivisions and neighborhoods that included these exclusionary land use regulations had good provision of parks and parkways when they were established, as shown in Figure 16. However, I have not found any evidence suggesting that the exclusion ary land use regulations were implemented to keep low income people away from parks in these subdivisions. On the other hand, developers likely aimed to capitalize on the good provision of parks and attract wealthy residents to stabilize neighborhood s (see park planning section). Regardless, these regulations can be considered discriminatory for recreation based on the recent ruling of the U.S. Supreme Court on housing (U.S. Supreme Court, 2015) as they had the effect of excluding low income ethnic minority people from some of Denver's park rich areas. Figure 16 shows that, among the seven neighbo rhoods mentioned above, five of them had good access to parks or parkways (Park Hill, City Park West, Montclair, South Denver, and Berkeley),

PAGE 302

288 while only two do not ( Highland and Garden Park). Although other neighborhoods not mentioned in the sources I coll ected could have included exclusionary subdivision regulations, the differences observed in Figure 16 show that such regulations were often established in areas around Denver's most prestigious parks, including City Park, Was hington Park and Berkeley Park. Figure 16 Areas that employed exclusionary subdivision regulations and their parks. Data source: (Dorward, 2010; Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010) ! The analysis of exclusionary subdivision regulations highlighted that, before forms of city driven exclu sionary zoning were implemented, developers created early visi ons of suburbia by excluding unwanted land use s and prescribing housing standard that only the middle and upper class could afford. As private developers, rather than the city, established thes e regulations, the search for profit has most likely driven the choice

PAGE 303

289 to implement them. A scholar interviewee who has studied Denver's planning pointed out that exclusionary subdivision regulations have had a significant impact in creating neighborhoods segregated by income: In my view, it's more the subdivision regulations than the zoning ordinances that determined economic segregation." Overall, while these sudvision regulations were mostly utilized between the 1890s and the 1910s (Dorward, 2010; Leonard & Noel, 1990) their implementation is still influencing the land values and the demographics in some of Denver's most historic neighborhoods. Exclusionary zoning The analysis of Denver's zoning ordinances from 1925 to 2010 showed that two zoning ordinances comprised some districts with exclusionary purposes: the Residence A district in the 1925 ordinance and the R 0 in the 1956 ordinance (see Table 17). To analyze Denver's examples of exclusionary zoning I rely largely on the historical evidence reported in Cole (2014) and on the 1925 and 1956 zoning ordinances. Exclusionary zoning districts are not peculiar to Denver ; they have been implemented in several other U.S. metropolitan areas, and particularly in suburban areas (Fischel, 2004; Hartnett, 1993; Maantay, 2001; Ritzdorf, 1985) As previously mentioned, exclusionary zoning policies have concurrently banned multi family housing, commercial and industrial land uses, and have pr escribed expensive minimum standards for single family housing, the only form of housing allowed in certain districts (Hartnett, 1993; Liberty, 2002; Maantay, 2001; Pendall, 2000; Ritzdorf, 1985) In Denver, exclusionary zoning districts were in place between 1925 until 2010, when the new form based zoning code was approved, although one of the most exclusionary clauses of the R 0 district from the 1956 code was dropped in 1989 (Cole, 2014) Also, based on my analysis, Denver's exclusionar y zoning districts have limited

PAGE 304

290 the capacity of low income people to live in proximity to some of Denver's most historic and prestigious parks and parkways. Residence A district, 1925 zoning code Denver's first zoning code, which was in full effect in 192 5, organized the city into industrial, commercial, and residential districts, with restrictions applied to various districts (Cole, 2014; Elliott, 2008; Noel & Norgren, 1987; R. L. S immons & Simmons, 1995) This first code was relatively simple, as it included only 13 different districts, five of which were residential (Elliott, 2008) Figure 17 shows the 1925 zoning map, with amendments approved in 1929. The 1925 zoning map located commercial and business districts in the downtown area and along a few corridors to the south (South Broadway), to the east (East Colfax) and to the west (West Colfax, West 32nd and West 38th). In addition, industrial areas were mostly concentrated along the South Platte River, from south to north, and in the northeastern part of the city. Historical evidence reported in Cole (2014) shows that the 1925 zoning ordinance intentionally attempted to establish a city of single family homes and to limit the development of multi family housing. The two planners who mostly contributed to drafting the 1925 code, Robert Whitten and Saco DeBoer, both expressed a positive bias towards single family housing and warned about the negative effects of multi family housing on cities in their writing, and especially of large apartment complexes (Cole, 2014) The two planners pointed out that multi family housing would have contributed to lowering the property values of single family homes, would have had negative public health consequences by limiting the amount of light and air in dwelling units, and would have had negative moral impacts on Non Hispanic White s (Cole, 2014) On the other hand, Whitten and DeBoer explained that single family homes would have been ideal for

PAGE 305

291 children, as they could provide green space and natural light, and would have been a better fit with the idea of Western living that at tracted people to Denver (Cole, 2014) Figure 17 Denver's 1925 zoning map (amended in 1929). Image credit: Denver Public Library, Western History Map Collections. Th e business community and property owners also supported the emphasis on single family housing (Cole, 2014) T hese two groups shared the vision of using zoning to protec t single family residential areas from being affected by undesirable land uses,

PAGE 306

292 with the goal of preserving property values (Cole, 2014) A 1926 article in the Denver M unicipal Facts shows that more than 95 percent of the applications for residential building permits between 1925 and 1926 were for single family homes (City and County of Denver, 1926) In particular, areas zoned as "Residence A" were aimed to the "stabilization of property values through its protection against commercial buildings" (City and County of Denver, 1926) The Residence A district prescribed single family homes as the only form of housing and banned any commercial and industrial land use (Cole, 2014) Also, the 1925 zoning map (Figure 17) shows that the Residence A district requi red a minimum lot size of 6,000 square feet prescribed a height limit of 35 feet. As a comparison Residence B requi red a minimum lot size of 3,000. According to Cole (2014) this emphasis on single family housing expressed mostly through the implementation of the Residence A district had the effect of excluding low income ethnic minorities from certain neighborhoods. The 1925 zoning o rdinance zoned predominantly Non Hispanic White neighborhoods as single family residential, while it zoned ethnic minority areas as mostly multi family housing, with commercial and industrial zones in close proximity (Cole, 2014) A similar approach was used in Los Angeles' first zoning code, established in 1904 (Wolch et al., 2005) Also, Cole (2014) pointed out that allowing multi family housing only in ethnic minorities areas was aimed to limit the areas where African Americans and other mino rities could afford to live. Although no words about race or ethnicity were included in the ordinance (Cole, 2014) this economic segregati on had racial and et hnic impl ications, especially during a time when the Ku Klux Klan dominated Denver's politics.

PAGE 307

293 Cole (2014) also argued that the Residence A d istrict was used as an economic proxy for racial exclusion in Park Hill. According to Cole (2014) American zoning, while officially intended to define appropriate land use, has often been used to determine the demographic groups who were appropriate to different parts of cities. Thus, "the line between unwanted land use' and unwanted people' in American zoning law and practice was a lways thin, but often glaringly absent in Park Hill" (Cole, 2014, p. 90) Excluding stores and other commercial activities i n Park Hill was a means to avoid unfamiliar and undesirable people in the neighborhood (Cole, 2014) The analysis of the spatial arrangement of Residence A districts sh ows some interesting patterns around Denver's largest and most prestigious parks and its historical parkways. The 1925 zoning map (amended in 1929) reported in Figure 17 shows that Denver's first zoning ordinance favored single family housing, and Residenc e A in particular, around the largest parks and around most parkways. T he three largest and most prestigious parks in the 1920s, City Park, Cheesman Park, and Washington P ark are all mostly surrounded by Residence A districts (see Figure 18). Also, Figure 18 displays significant variations between the first blocks around these three major parks, zoned as Residence A, and blocks located farther away from these parks, zoned as Residence B, C, and D. This variation shows that larger lots, with larger houses, w ere prescribed in the very proximity of these three parks, no doubt enhancing the property value s Also, the analysis of Figure 17 shows that most parkways established during Denver's City Beautiful era were surrounded almost exclusively by Residence A dis tricts. The predominance of Residence A districts around parkways aligns with the City

PAGE 308

294 Beautiful vision aiming to create wide boulevards that integrate trees, lawns, and large mansions. Figure 19 depicts the Residence A district around East 7th Avenue Park way. Figure 18 Zoning districts around City Park, Cheesman Park and Washington Park. Edited from Denver's 1925 zoning map (amended in 1929). Image credit: Denver Public Library, Western History Map Collections.

PAGE 309

295 Figure 19 Zoning districts around Eas t 7th Avenue Parkway. Edited from Denver's 1925 zoning map (amended in 1929). Image credit: Denver Public Library, Western History Map Collections. Overall, the analysis of the zoning districts located around all parks and playgrounds included in the 1925 zoning map (Figure 17) shows that, of the 31 parks included in the zoning map, 19 parks were mostly surrounded by Residence A and B districts (single family, the most restrictive), 6 parks were mostly surrounded by multi family housing (Residence C, D, and E), and 6 parks were mostly surrounded b y non residential districts, including commercial, business and industrial land uses Table 18 lists the parks that were included in each of the three categories introduced above (single family residential, multi fa mily residential, and non residential). M ost of the areas surrounding Denver's largest parks, including but not limited to the parks displayed in Figure 18, were zoned as Residence A and Residence B (see Table 18) These exclusionary zoning districts, comb ined with the discriminatory housing policies and practices that are discussed in the next section, contributed to stabilizing the demographics of the neighborhoods around Denver's largest parks and historical parkways, which have mostly included Non Hispa nic White middle and upper class people (see Chapter 5, Park Spatial Distribution Findings).

PAGE 310

296 Table 18 Predominant Zoning Districts Around Denver's Parks in 1929 Parks mostly surrounded by single family residential (Residence A and B) Parks mostly surrou nded by multi family residential (Residence C, D, and E) Parks mostly surrounded by non residential districts (business, commercial, industrial) City Park Cheesman Park Washington Park Sloan's Lake Berkeley Park Rocky Mountain Lake Park Barnum Park Platt Park City of Kunming Park Alamo Placita Park Observatory Park Mountain View Park Montclair Park Chaffee Park Argo Park Russell Square Park City of Nairobi Park City of Axum Park Dunham Park Highland Park Lincoln Park Sunken Gardens McDonough Park Curtis Pa rk Fuller Park Jefferson Park Overland Park Civic Center Park Columbus Park Lawson Park Elyria Park R 0 district, 1956 zoning code The 1956 zoning ordinance was drafted in a historical period in which Denver, although expanding its population, was signi ficantly losing its predominance in the metropolitan area due to suburbanization (Cole, 2014;

PAGE 311

297 Kurtz, 1957) In this context, the 1956 zoning ordinance was an attempt to make Denver more attractive to people who wanted to move to the suburbs i.e., middle and upper class Non Hispanic White f amilies (Cole, 2014) In other words, the 1956 zoning code, like the 1925 ordinance, sought to make Denver's older neighborhoods more suburban by increasing the weight of single family housing in the city (Cole, 2014) Thus, the zoning ordinance was motivated by the goal of keeping middle and upper class Non Hispanic White families f rom moving to the suburbs, and consequently to maintain jobs and tax revenues in Denver. The 1956 ordinance, which was in place until 2010, succeeded in making Denver more suburban by strongly reducing mixed uses and completely breaking from its more urban past (Halbur & Laetz, 2010; Onaran, 2009) The 1956 zoning ordinance was created following a rational planning model, including an economic survey to estimate the cit y's needs in terms of residential, commercial, and industrial land (Cole, 2014; Kurtz, 1957) According to Cole (2014) the economic survey developed to estimate land needs contributed to creating an idealized vision of the future Denver residents, a vision that mostly included young families with good education, good jobs and who preferred to live in single family homes. Thus, the 1956 zoning ordinance aimed to make De nver a better place for the growing American middle class by planning for a city of single family homes (Cole, 2014) However, one of the issue s that some of Denver's old neighborhoods were facing in the 1950s was the de facto transformation of single family residences into multi family housing (Cole, 2014) The roots of this phenome non can be found in a housing crisis that occurred in Denver during World War II, as a many people were moving to the city given federal investment in Denver during the war (Cole, 2014) Due to a housing shortage

PAGE 312

298 many homeowners of single family units rented their basements or created a boarding house with different units in their property to bring in extra revenue (Cole, 2014) According to the Denver Planning Office, this de facto multiplication of multi family housing in Denver's older neighborhoods did not resonate well with the wishes of young middle and upper class families (Cole, 2014) Therefore, to welcome the idealized middle class families to Denver's older neighborhoods, Denver Planning Office employed two complementary sets of strategi es in the 1956 zoning ordinance (Cole, 2014) First, the zoning ordinance defined families as people related by "blood, marriage, or adoption" (Cole, 2014, p. 4) which was later called the "living in sin" ordinance (Associated Press, 1989) Thus, only people related by blood, marriage or adoption could live in areas zoned as R 0, which were the most restrictive, and a special permit was needed for unrelated peopl e sharing housing units in R 1 districts (Cole, 2014) This strategy was clearly intended to exclude ce rtain types of people, such as renters and extended families, fro m Denver's older neighborhoods (Cole, 2014) The second set of strategies involved establishing much stricter rules to virtually ban multi family housing and other unwa nted land uses in neighborhoods zoned as single family housing (Cole, 2014) The 1956 zoning ordinance established five residential districts, from R 0 to R 4 (City and County of Denver, 1956) The R 0 and R 1 district could include exclusively single family houses, with R 0 being the most restrictive by prohibiting unrelated people from living together (City and County of Denver, 1956) The R 2, R 3 and R 4 districts could include multi family housing, with the R 4 district allowing the highest residential densities (City and County of Denver, 1956)

PAGE 313

299 The R 0 district deserves a specific attention. According to Cole (2014) the R 0 district was intended to create a residential environment that could be suitable for families with children. The 1956 zoning code describes R 0 districts as peaceful suburban landscapes without the nuisance of commercial activities (City and County of Denver, 1956) Parks, schools, and churches were the only other land uses allowed in R 0 districts (City and County of Denver, 1956 ) This suggests that p arks, schools and churches were seen as beneficial for children's development, but also as places that would not have brought "outsiders" to the neighborhood. Banning unrelated people from living together in areas zoned as R 0 was a nother strong way to keep undesired people out of these neighborhoods, including roommates, boarders, extended families, and unmarried couples, and specifically same sex couples (Cole, 2014) By excluding undesired people and activities, the Denver Planning Office hoped to recreate a suburban idyll that could attract middle and upper class families (Cole, 2014) In the 1950s and 1960s, unwanted people included ethnic minorities. Cole (2014) argued that the imple mentation of R 0 zoning districts was a subtle form of racism. Affluent Non Hispanic White people used zoning to further residentially segregate ethnic minorities through economic means by making living in R 0 districts very expensive (Cole, 2014) For example, the Park Hill Action Committee (PHAC), a neighborhood organization of Park Hill, created a private task force to detect violations to the R 0 ordinance (Cole, 2014) PHAC, in its zoning violations reports, overly targeted African Americans and Hispanic residents of Park Hill (Cole, 2014) Violation reports were filed more than once even after evidence was shown against the report (Cole, 2014) A Hispanic fa mily with eight children was told that they did not belong in Park Hill, given

PAGE 314

300 their large number (Cole, 2014) As a historian interviewee explained: "The R 0 district shows that some neighborhoods used zoning to prevent communities of color who didn't have the same housing practices, like sharing homes with unrelated people and with the extended family, from living close to them." Although officially introduced to prese rve property values in single family neighborhoods, the regulations of the R 0 district show signs of classism and racism (Cole, 2014) Another interesting aspect of th e R 0 district is that communities had to petition to the city to have their neighborhoods, or part of it, zoned as R 0 (Cole, 2014) When the new 1956 zoning ordinance was approved, the standard district for neighborhoods with high percentages of single family homes was R 1 (Cole, 2014) From the late 1950s to the 1960s, several affl uent and predominantly Non Hispanic White communities requested and obtained rezoning to R 0 (Cole, 2014) This means that the desire of excluding individuals and group s that were considered as undesirable came from the residents of certain areas, not from planners. However, planners created an ordinance that made exclusion lawful. The communities that required and obtained R 0 zoning included large parts of "Crestmoor Park, Belcaro, Denver Country Club, Park Hill, Congress Park, Capitol Hill, Speer, Cherry Creek, Hale, and Cory Merrill" (Cole 2014, p. 5) Figure 20 shows the areas that were zoned as R 0 in 1989, comprising their parks and parkways. Cole (2014) argued th at many of the neighborhoods that obtained rezoning to R 0 had a history of ethnic discrimination, including Crestmoor Park, Country Club, and Belcaro Park.

PAGE 315

301 Figure 20 Areas zoned as R 0 district until 2010, with parks and parkways. Map developed using data from City and County of Denver (2014) The "living in sin" ordinance was lifted in 1989, when City Council passed the repeal of the clause prohibiting unrelated people from living together in R 0 districts (Associated Press, 1989; Cole, 2014) The repeal was approved by a very tight marg in, with seven council members voting in favor and six voting against the repeal (Associated Press, 1989; Cole, 2014) This very contested vote involved, on one side, people arguing that zoning should not discriminate based on relationship status, and on the other side people claiming that R 0 zoning had contributed to cons olidate family values and neighborhood morals (Associated Press, 1989; Cole, 2014) The analysis of the spatial distribution of R 0 districts in Denver shows that R 0 zoning was requested and obtained in areas that are quite well served by parks and

PAGE 316

302 parkways (see Figure 20). Several areas around City Park, Cheesman Park, Cr estmoor Park, and many prestigious parkways were zoned as R 0 until 2010, with the "living in sin" ordinance being lifted in 1989. As for the Residence A districts in the 1925 zoning ordinance, there is no evidence that R 0 districts were created with the explicit purpose of excluding low income ethnic minorities from parks and parkways. As Cole (2014) pointed out based on her extensiv e research, R 0 districts were established to protect property values by keeping undesired people out of affluent neighborhoods. Regardless of the intent the exclusionary outcome s of these zoning districts in terms of recreation and housing are sufficient to make claims about discrimination, per the recent ruling of the U.S. Supreme Court (U.S. Su preme Court, 2015) To summarize, the introduction of R 0 districts and of their "living in sin" clause in the 1956 zoning ordinance was a key component of Denver Planning Office's efforts to create an idyllic suburban vision of Denver (Cole, 2014) The general goal of this vision was to limit middle and upper class white flight to the suburbs, which had caused significant economic losses for the city (Cole, 2014) Also, the specific objectives or the regulations attached to the R 0 district included: stopping the informal transformation of single family homes into de facto multi fam ily housing; separating traditional from non traditional families; and creating quiet residential areas without the disturbance of commercial activities (Cole, 2014) W hile the ordinance did not specify racial or ethnic exclusion its application had the effect of further excluding low income ethnic minorities from affluent neighborhoods (Cole, 2014) In terms of access to parks, the analysis of the spatial distribution of R 0 districts highlighted that most R 0 districts had good provision s of parks and parkways.

PAGE 317

303 2010 form based zoning code In 2010, Denver adopted a new form based z oning code that substitutes the 1956 zoning ordinance (Denver Community Planning and Development, 2010) A large number of ame ndments to the 1956 ordinance made the ordinance too complicated to use, hence the need of a new code (City and County of Denver, 2002; Elliott, 2008; Halbur & Laetz, 201 0) The new form based code employs New Urbanist concepts as a framework, including the definition of various neighborhood contexts that differ by function and physical environment pattern (Hirt, 2013) For each neighborhoo d concept, the code conceptualizes further types of homogeneous urban forms that constitute the specific districts (Hirt, 2013) According to a scholar interviewee who studied Denver's planning history, the new form based zo ning code did not intentionally comprise exclusionary policies. However, as the form based code is grounded on the concept of neighborhood context, it would not introduce building forms that are extraneous to the neighborhood character: The only possible w ay that the form based code might propagate separation by income is this way. Say that you have a neighborhood composed by large homes. The form based code would have rezoned the neighborhood as single family homes on large lots, so that the new constructi ons would fit with the neighborhood character. Think about neighborhoods like Hilltop or Bonnie Brae. They are beautiful neighborhoods. The new zoning code would probably prohibit multi family housing or duplexes. You couldn't put up a four unit building. Therefore, while the form base code does not introduce new forms of exclusionary zoning, it also does not challenge the status quo of established single family neighborhoods, as it aims to consolidate different neighborhood contexts. However, at this point of Denver's history, it might be unreasonable and unfeasible to introduce significant changes to Denver's most historical neighborhoods by allowing multi family housing units or commercial activities.

PAGE 318

304 Land Use Planning for Park Establishment and Residenti al Location: A Summary This extensive analysis of Denver's land use pl anning centered on two main areas : the contribution of land use planning to park establishment through parkland acquisition and generation of park funding, and the cont ribution of land u se planning to the residential locations of different demographic groups in Denver. The analysis of land use planning instruments' contribution to park establishment highlighted that until 1974, when the Poundstone Amendment was approved, Denver ha d succes sfully used annexation regulations to acquire significant amounts of parkland through annexations. After 1974, Denver has struggled to acquire land for parks or park funding as most developments, with one exception, occurred in infill sites, and as the cit y has lacked an impact fee for parks. Table 16 describes the main land use instruments that the city has used to acquire parkland or park funding, as well as the instruments that the city could have used for these purposes. Land use planning has also cont ributed to the residential location of different ethnic and income groups in Denver by influencing land values in various parts of the city. Land value has been mostly linked to different housing types and to the distance to locally unwanted land uses. Tab le 17 summarizes the privately driven and municipal exclusionary land use instruments that have been used in Denver to limit the land uses, and the people, that were considered appropriate in different parts of Denver. The interactions between land use pl anning for park establishment and exclusionary land use policies and practices have had a negative effect on park equity. Indeed, on one side land use planning has failed to create instruments to acquire parkland in infill developments, including in low in come densely populated neighborhoods, and to

PAGE 319

305 obtain funding through impact fees to purchase new park land or improve existing parks. On the other side, exclusionary land use policies and practices have contrib uted to excluding low income ethnic minorities f rom the areas surrounding Denver's largest parks and historic parkways. In the final section of this chapter, I discuss how land use planning policies and practices have cooperated or clashed with park planning and housing policies and practices in providi ng parks and parkways for different demographic groups during various phases of Denver's history. Housing Policies and Practices In t his section I analyze how housing policies and practices have contributed to the residential location of different ethnic and income groups in Denv er, with a particular focus discriminatory housing policies and practices against ethnic minorit y groups and on the implementation of different forms of affordable housing. T his analysis based on interview and secondary data, foc uses on housing policies and practices that have been responsible for the residential location of low income ethnic minorities in Denver. Ethnic and racial discrimination in housing policies and practices in Denver and several other U.S. cities, has put ethnic minorit y populations at significant disadvantage in h ousing markets by limiting the geographic areas where they were allowed to buy or rent a home by limiting the ability to borrow money, by asking them to pay higher prices than Non Hispanic Whites and, as a consequence, by forcing ethnic minorities to accept subpar housing conditions that do not meet their necessities (Gotham, 2000; Massey, 1990; Wyly & Holloway, 2002; Yinger, 1997) These policies and practices, combined with the exclusionary land use mechanisms and the effect of land values, have contributed to the residential segregation of ethnic and racial minorities in American

PAGE 320

306 cities (W. A. V. Clark, 1986; Gotham, 2 000; Massey, 1990; Wyly & Holloway, 2002) including Denver. Housing discrimination has historically targeted African Americans more than other ethnic and racial minorities (Gotham, 2000; M assey, 1990) As a consequence, African Americans have generally experienced higher rates of residential segregation than other ethnic minorities (Gotham, 2000; Massey, 1990) In the sectio n on ethnic and racial housing discrimination, I analyze the various policies and practices that have contributed to segregating African Americans and Hispanics. At the end of this analysis, I discuss the consequences of housing discrimination for African Americans' and Hispanics' residential locations in Denver, with a focus on their proximity to parks. The U.S. Department of Housing and Urban Development (HUD) has defined affordable housing as "housing for which the occupant(s) is/are paying no more than 30 percent of his or her income for gross housing costs, including utilities" (M. Schwartz & Wilso n, 2008; U.S. Department of Housing and Urban Development, n.d. a) In Colorado, HUD provides public support for affordable housing mostly by funding subsidized apartments and by funding public housing authorities, includi ng the Denver Housing Authority (U.S. Department of Housing and Urban Development, 2015) In the section covering af fordable housing, I present a brief history of Denver Housing Authority, discuss the recent work by HUD in Colorado to create affordable housing opportunities, and present a recent Denver municipal policy to promote affordable housing. I conclude the secti on on affordable housing by discussing the geographic concentrations of affordable housing in Denver in relation to the location of parks and parkways.

PAGE 321

307 Ethnic and Racial Discrimination in Housing : Policies and Practices In Denver, discriminatory housing p olicies and practices targeted against ethnic and racial minorities have included racially restrictive covenants (Dorward, 2010; Leonard & Noel, 1990; Mauck, 2001; Romero, 2013; R. L. Simmons & Simmons, 1995) racial intimidation in housing (N. Hernandez, 2006; Leonard & Noel, 1990; Raughton, 2000; R. L. Simmons & Simmons, 1995) and lending and real estate discrimination, including redlining (Cole, 2014; Langegger, 2012; Leonard & Noel, 1990; Newsum, 2012; R. L. Simmons & Simmons, 1995; T. H. Simmons & Simmons, 2010) Episodes of housing discrimination against African Americans, His panics, and other ethnic minorities were most evident between the 1910s and the early 1960s in Denver (Leonard & Noel, 1990; Newsum, 2012; R. L. Simmons & Simmons, 1995; T. H. Simmons & Simmons, 2010) although evidence of discriminatory prac tices was still found in the 1980s (Palm, 1985) and in the 2010s (Denver Metro Fair Housing Center, 2014) As in many other American cities, housing discrimination in Denver was mostly directed against Afric an Americans (Abbott, 1978; Mauck, 2001; R. L. Simmons & Simmons, 1995) Significantly fewer episodes of housing discrimination against Hispani cs or other ethnic minorities were reported in historic sources (e.g. Langegger, 2012) perhaps because African Americ ans were the largest ethnic minority in Denv er between the 1910s and the 194 0s (Abbott, 1978; Leonard & Noel, 1990) when most discriminatory episodes took place. Also, although some authors have considered exclusionary zon ing as a form of discriminatory housing (e.g. Bono, 2007; Hartnett, 1993) exclusionary zoning is in fact a land use policy (Pendall, 2000; Ritzdorf, 1985) thus it was discussed in the section on land use policies and practices.

PAGE 322

308 Racially restrictive covenants Starting from the early 1900s, racially restrictive covenants attached to residential properties in certain areas of Denver have prevented homes from be ing sold or rented to non Caucasian individuals and families (Dorward, 2010; Leonard & Noel, 1990; Mauck, 2001; Romero, 2011, 2013; T. H. Simmons & Simmons, 2010) According to a Denver historian I interviewed, "the most obvious mechanism [to create residential segregation] was using racially restrictive covenants." As explained in the interview passage below and as reported by Mauck (2001) racially restrictive covenants were initially mostly targeted against African Americans: In the 1910s, the covenants were mostly directed to keeping African Americans ou t of these subdivisions. Some covenants made an exception for service help. [...] One of the covenants to a 1910 Jefferson County development said: "Only persons of the Caucasian race should use or occupy a lot." These were covenants attached to the proper ty. Like in other American cities (Boone et al., 2009; Gotham, 2000) the implementation of racially restrictive covenants in Denver partially derived from the fear, enhanced by real estate agents, that a multi ethnic neighborhood would have lowered Non Hispanic White homeowners' property values (Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010) The belief that the presence of people of color causes property values to d ecrease in a neighborhood has been widespread in the United States (de Souza Briggs, Darden, & Aidala, 1999; Gotham, 2000; D. R. Harris, 1999; Pulido, 2000) including Denver (Newsum, 2012; T. H. Simmons & Simmons, 2010) A s the Ku Klux Klan was very influential in Denver in the 1920s (Goldberg, 1981; Leonard & Noel, 1990) racism against African Americans and other racial and ethnic minorities could have been another driver of racially restrictive covenants.

PAGE 323

309 In Denve r, racially restrictive covenants were attached to properties through two main mechanisms. First, neighborhood improvement associations of existing neighborhoods encouraged property owners to establish racially restrictive covenants limiting sales to Non H ispanic Whites (R. L. Simmons & Simmons, 1995) For example, neighborhoods surrounding Five Points, where African Americans were concentrat ed in the 1900s and 1910s, established racially restrictive covenants to prevent the ghetto from spread ing outside Five Points (Leonard & Noel, 1990; Mauck, 2001; R. L. Simmons & Simmons, 1995) Similarly, neighborhood improvement associations in Baltimore urged Non Hispanic White residents to establish racially restrictive covenants after the city's segregation ordinance was revoked (Boone et al., 2009) Second, some developers incl uded racially restrictive covenants in the plat maps of their new subdivisions (Dorward, 2010; Romero, 2013) In the 1900s, Downington, one of the subdivisions of Park Hill, included racially restrictive covenants limiting home owne rs and occupants to Non Hispanic Whites only (Dorward, 2010) Also, historical evidence presented by Romero (2011, 2013) showed that some of Denver's largest developers used racially restrictive covenants from the 1920s until the late 1940s. As explained by a Denver historian I intervie wed, racially restrictive covenants established by developers "were mostly used in three locales: South Denver, around Bonny Brae and Observatory Park (1920s); Crestmoor Park (1920s); and Harvey Park (1947 1948). Some of these covenants contained religious prohibitions, mostly directed to the Jews." Therefore, some of Denver's neighborhoods included only Non Hispanic White Christians by design. As ethnic diversity was considered as a threat to property values

PAGE 324

310 (Newsum, 2012; T. H. Simmons & Simmons, 2010) homebuilders used ethnic and religious homogeneity as a selling point for their subdivisions. Regardless of how they were established, between the 1900s and the 1940s raciall y restrictive covenants were implemented in the following neighborhoods: Park Hill (Dorward, 2010) Whittier and City Park West, around Five Points (Leonard & Noel, 1990; Mauck, 2001; R. L. Simmons & Simmons, 1995) and "McCulloc h [Whittier neighborhood], Clayton, Berger and Ashley [Clayton neighborhood], Crestmoor, Bonnie Brae, Chaffee Park, Illiff's University Additions, Regis Heights, and Clingers Gardens" (Romero, 2011, p. 13) Thus, racially restrictive co venants, which in the 1940s targeted mostly Afric an Americans and Hispanics (T. H. Simmons & Simmons, 2010) focused on northeast Denver (Whittier, City Park West, Clayton McCulloch, and Park Hill), east central Denver (Crestmoor and Bonnie Brae), south Denver (I l liff's Uni versity Addition), and northwestern Denver (Chaffee Park and Regis Heights). In the 1930 s the public policy of Colorado as expressed by its Supreme Court, recognized that people could prefer to live close to people who were Non Hispanic White instead of African American or Hispanic, thus legitimizing racially restrictive covenants (Romero, 2011) Racially restrictive covenants in Colorado were still in effect in the 1940s and the 1950s, although in the 1948 the U.S. Supreme Court made their enforcement illegal (Leonard & Noel, 1990; Romero, 2011) In the late 1940s and 1950s, developers increasingly attached racially restrictive covenants to subdivisions in Denver's suburbs (Romero, 2013) The application of racia lly restrictive covenants finally came to an end in 1957, when the Colorado Supreme Court made them unenforceable (Romero, 2013) The end of racially restrictive covenants meant that some

PAGE 325

311 African Americans could start moving to Denver's suburbs in the late 1950s and early 1960s (Mauck, 2001) To summarize, racially restrictive covenants were initially established in the areas around Five Points to prevent African Americans from living outside the Five Points ghetto in the 1910s and 1920s. Other neighborhoods in east central, south, and northwest Denver, also adopted these covenants, in part to exclude Jews (south Denver), and in part as a general protection from future expansions of ethnic minorities. Racial intimidation in housing While racially restrictive covenants were enforced by Colorado State policy, other forms of housing discrimination in Denver were non governmental racial intimidation practices directed against African Americans around Five Points (N. Hernandez, 2006; Leonard & Noel, 1990; Raughton, 2000; R. L. Simmons & Simmons, 1995) While in the 1920s African Americans were gradually allowed to live in parts of Whittier, a "color line" determined the specific areas to which they were restricted (N. Hernandez, 2006; Leonard & Noel, 1990; Raughton, 2000; R. L. Simmons & Simmons, 199 5) Color lines were not legal restrictions, but they were practices that were used by realtors and residents until the 1960s (N. Hernandez, 2006) African Americans who challenged the geogra phic restrictions imposed by color lines in Whittier wer e targeted by threats and by significant acts of violence (N. Hernandez, 2006; Leonard & Noel, 1990; Raughton, 2000; R. L. Simmons & Simmons, 1995) A few middle class African Americans who dared to cross the color line were met by crowds of angry Non Hispanic Whites protesting outside their homes and by bom bs placed in their front yard (N. Hernandez, 2006; Leonard & Noel, 1990; Raughton, 2000; R. L. Simmons & Simmons, 1995) The Ku Klux Klan, which was very powerful in the

PAGE 326

312 1 920s and 1930s, helped enforce these informal laws (N. Hernandez, 2006) Similar acts of violence were undertaken against Africa n Americans who dared to cross color line s in Baltimore (Boone et al., 2009 ) and anyplace where de facto segregation was practiced Thus, while racial intimidation in housing was a fairly geographically delimited phenomenon in Denver, it further limited African Americans' residential locations, thus contributi ng to overcrowding the Five Point area. Also, the acts of violence undertaken to enforce the unwritten laws of the "color line" speak about the level of racial hatred against African Americans in Denver during the 1920s and 1930s Lending and real estate discrimination A t hird form of ethic and racial discrimination in housing involved lending practices supported by federal agencies, like the Home Owner's Loan Corporation, and unwritten laws among realtors (Cole, 2014; Langegger, 2012; Leonard & Noel, 1990; Newsum, 2012; R. L. Simmons & Simmons, 1995; T. H. Simmons & Simmons, 2010) Discriminatory lending practices contributed to African Americans a nd other ethnic minorities' inability to obtain a mortgage to buy or improve a property in areas that were not considered as a safe investment (Newsum, 2012; T. H. Simmons & Simmons, 2010) Discriminatory real estate practices like racial steering further limited African American's residential mobility in Denver (Cole, 2014; Leonard & Noel, 1990; Newsum, 2012 ; T. H. Simmons & Simmons, 2010) Lending discrimination: redlining Lending bias due to redlining significantly affected African Americans in Denver in the 1930s and 1940s (T. H. Simmons & Simmons, 2010) In the 1930s, the H ome Owner's Loan Corporation (HOLC), a federal agency, developed residential security maps for more than 200 cities in the United States to classify areas with different investment risks for banks and homeowners including

PAGE 327

313 "redlined areas considered as to o risky for lending (Greer, 2013; J. Hernandez, 2009; Hillier, 2003a, 2003b; Squires, 2003) In Denver, redlining of the Five Points and Whittier neighborhoods c ontributed to African Americans inability to obtain mortgag es to build or repair homes, as banks would not lend money in redlined areas with depreciated building stocks (T. H. Simmons & Simmons, 2010) Similarly, r edlining overly targeted African American areas in Baltimore (Boone et al., 2009) and throughout the country (Greer, 2013) Also Denver's African Americans could not leave Five Points and Whittier due to racially restrictive covenants, racial intimidation, and racial steering (Leonard & Noel, 1990) Some areas in northwestern Denver, particularly in the Highland neighborhood, have also experienced redlining and bank disinvestment in the 1940s and 1950s (Diaz, 2005; Dur‡n, 2011; Langegger, 2012) According to a Denver historian interviewee the Highland neighborhood became redlined because of "a combination of small lots and overcrowding." The availability of small hom es built on lots smaller than Denver's standards attracted different low income ethnic minorities in various historical periods (Irish, Italians, and Hispanics). With the minorit y population growing, The Highland became very crowded with people, many peopl e living in these tiny little homes. When many people live in a house there is much more wear and tear. The neighborhood starts decaying; it becomes less and less like a place where banks and insurance companies are comfortable lending money. [] It just b ecame an area where it just wasn't wise to invest. Redlining was still affecting the capacity of African Americans to obtain mortgages in the 1950s (Newsum, 2012; T. H. Simmo ns & Simmons, 2010) When an African American reporter for the Denver Post, a local newspaper, tried to buy a house in a n

PAGE 328

314 African American neighborhood of Denver in 1951, he could not find a bank that would lend him money, as these neighborhoods were redl ined (Newsum, 2012) Real estate discrimination In the 1950s, besides lending discrimination fostered by federal agencies, African Americans were experiencing discrimina tion from realtors including racial steering (Newsum, 2012; T. H. Simmons & Simmons, 2010) R eal estate agents would not show African Americans hom es that were located outs ide the redlined areas, which coincided with Denver's African American neighborhoods (Newsum, 2012; T. H. Simmons & Simmons, 2010) The report by the Denver Commission on Hu man Relations, required by M ayor Newton in the 1940s, blamed realtor practices and Non Hispanic White neighborhoods that refused to welcome ethnic minority people for African Americans' and Hispanics' extreme residential segregation, which combined with re dlining determined their abysmal living conditions (Newsum, 2012) According to a report of the Urban League in the early 1950s, gentlemen agreements between realtors a nd written regulations requiring realtors to only s ell homes to African Americans in Denver's ghetto were the strongest form of discrimination and affected residential segregation the most (Newsum, 2012) These gentlemen agreements were formalized in the official regulations of the national association of real estate agents, as reported by Newsum (2012) : "A REALTOR should never be instrumental in introducing into a neighborhood a character of property or occupancy, members of any race or nationality, or any individual whose presence will clearly be det rimental to property values in the neighborhood." Article 34 of the National Code of Ethics of the National Association of Real Estate Boards in effect from 1924 1950. (p. 77). This article shows that preserving property values was used as the official rat ionale for discriminating against African Americans and other ethnic minorities. Thus, as for

PAGE 329

315 exclusionary zoning, the goal of preserving property values had negative effects on ethnic minorities' housing choices. In the late 1950s and early 1960s, a serie s of state and federal laws, including the 1959 Fair Housing Practices Act, contributed to significantly limiting real estate discrimination (Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010) Colorado was among the first states to pass laws against discrimination in private housing practice s, including residential location and lending (T. H. Simmons & Simmons, 2010) The 1959 Fair Housing Practices Act banned discriminations based on race, ethnicity, religion, and national origin (T. H. Simmons & Simmons, 2010) According to T. H. Simmons and Simmons (2010) these anti discrimination laws, which expanded ethnic minorities' residential choice s in De nver, contributed to increasing w hite flight in the 1960s. Despite these anti discrimination laws, episodes of housing discrimination against ethnic minorities have been accounted for in the 1980s and 2010s (Denver Metro Fair Housing Center, 2 014; Palm, 1985) In particular, Palm's (1985) study highlighted that in the 1980s, real estate practices in Denver were still organized along ethnic lines, as ethnic minority realtors worked mainly in areas where their ethnicity was over represented. Also Denver Metro Fair Housing Center's (2014) recent study showed that discrimination in rental housing in Denver is still evident for people of color. In particular, rental agents and managers are still attempting to dissuade ethnic minority people from renting in areas and apartment complexes that include predominantly Non Hispanic Whites (Denver Metro Fair Housing Center, 2014) To summarize, redlining and real estate discrimination contributed to furth er expanding the discrimination in housing that ethnic minorities, and in particular African

PAGE 330

316 Americans, were experiencing through racially restrictive covenants and racial intimidation. African Americans' housing conditions were further worsened because, w hile they were not able to live outside the Five Points ghetto until the 1950s, they also could not secure mortgages to improve their homes due to redlining. Housing discrimination and residential location Discriminatory policies and practices in housing have significantly restricted the areas where African Americans have been allowed to live, at least until the 1960s (see Figure 7). Until the 1960s, African Americans have been mostly concentrated between Five Points, Whittier, some areas north of City Par k, and N ortheast Park Hill (see Figure 7). By all accounts, Five Points and partially Whittier until the 1950s had very high population densities and deteriorating housing stock, which contributed to African Americans' subpar living conditions in these are as (Mauck, 2001; Newsum, 2012; R. L. Simmons & Simmons, 1995) Middle class African Americans who could move northeast of Five Points in the 1950s and 1960s fo und lower population densities and better living conditions (Leonard & Noel, 1990) African Americans' residential segregation in a few neighborhoods of northeast Denver, which was fostered by discriminatory housing policies and practices and exclusionary land use policies had consequences for their access to parks, at least until the 1960s. Five Points included only Curtis Park, which was established by a developer when the neighborhood was Non Hispanic White and affluent (Leonard & Noel, 1990; Mauck, 2001) and a few other smaller playgrounds (see Figure 7). The n eighborhood's overcrowded living conditions made it difficult for the city to acquire parkland and establish new parks. Furthermore, until the 1950s African Americans and Hispanics were not allowed to live in a few areas with good provision of parks and pa rkways, including

PAGE 331

317 South Park Hill, City Park West, Crestmoor, and O bservatory Park, because of racially restrictive covenants attached to properties. Housing discrimination has not had the same extremely negative effects on Hispanics' residential segregati on, although lending practices have contributed to worsening their living condition in the Highland neighborhood. Highland has been fairly underserved by parks since the 1900s (Goldstein, 2011) and high population densities contributed to poor park provision until today My analysis of housing in Denver did not uncover discriminatory housing practices in other areas that have historically being inhabited by Hispanics, includi ng the West Side neighborhoods. The specific implications of ethnic discrimination in housing on park provision for African Americans and Hispanics are discussed in Chapter 6. Affordable Housing in Denver This section presents an introduction to affordabl e housing in Denver, with a specific focus on Denver Housing Authority, the latest efforts by HUD to create affordable housing options in Denver, and Denver's Inclusionary Housing Ordinance (IHO). This analysis uncovered that the location of affordable hou sing, including public and subsidized housing, has reinforced the patterns of residential segregation between social classes that were fostered by exclusionary land use and housing discrimination In Colorado the U.S. Department of Housing and Urban Devel opment funds public housing authorities and subsidized apartments (U.S. Department o f Housing and Urban Development, 2015) As a HUD Colorado interviewee explained, HUD acts as a facilitator for the provision of affordable housing today. In other words, "HUD does not build anything" but distributes funding to a variety of "state and city housing authorities."

PAGE 332

318 In particular, "Denver is one of these local authorities that we [HUD] fund, it is one of the recipients that we call entitlement cities ." However, a former Community Planning and Development land use planner I interviewed, pointed o ut that Denver does not have a long standing tradition of public housing, which reflects funding choices: "In the US, and Denver especially, there is not a culture for social housing." One of the ways HUD supports affordable housing is by providing funding for subsidized apartments (U.S. Department of Housing and Urban Development, 2015) Subsidized apartment providers, including developers and property owners, are chosen based on competitive grants, according to a HUD Colorado staff member I interviewed: During slower economic times, developers and property owners make some multi family housing available for affordable housing. There are different types of affordable housing that can be included in this program, such as housing for elderly and housing for people with disabilities. Developers and property owners who want to get into sectio n 8 contracts need to participate to competitive applications. If they win, the y have a guaranteed income from HUD for a relatively long time. The location of the apartments, including their proximity to amenities like public transportation and parks, is n ot one of the key aspects of the competitive grants for subsidized rental units However, recently the introduction to these competitive grants has started to acknowledge the importance of proximity to amenities : Things like access to transit, access to am enities, and access to fresh food are not at the front end of the application process, but I think they should be. Recently, we have made some progress in this. In the preamble to the NOFA (Notice of Funding Availability), we have a general statement on ac cess to amenities. This means that, currently, parks and recreation are not significantly taken into account when deciding which apartment complexes receive funding for subsidized housing. Although parks and recreation are not central to HUD's mission (U.S. Department of

PAGE 333

319 Housing and Urban Development, n.d. b) the inclusion of proximity criteria in these grants could positively impact the quality of life of people living in subdsidized housing. Denver Housing Authority The Denver Housing Authority (DHA) was founded in 1938 following the framework defined by the Federal Housing Act (Goldstein, 2011; Leonard & Noel, 1990) In the 1940s, DHA started building a series of public housing projects in the northern and western parts of the city (Goldstein, 2011; Leonard & Noel, 1990) Among them, DHA built three housing projects in 1940 that were separated by ethnicity: the Lincoln Park project, dedicated to Non Hispanic Whites, in La Alma/Lincoln Park; the Platte Valley Project, for African Americans, in Five Points; and Las Casitas Project, for Hispanics, in Su n Valley (Goldstein, 2011) One year later, DHA was accused of discriminating against Hispanics, as Las Casitas Project was built with very low construction standards (T. H. Simmons & Simmons, 2010) Besides the designed ethnic segregation and the accusation of discrimination it is important to note all three housing sites are located in relatively low income and undesirable areas of the city, ne xt to infrastructure like railways and highways and industries The proximity between public housing projects and noxious land uses is not uncommon in American cities (Bullard, 1993) and Denver has followed this trend. The population growth of ethnic minority groups in the 1950s put pressure on Denver Housing Authority (T. H. Simmons & Simmons, 2010) DHA acknowledge d that Denver ne eded to build six times more new public housing units than the number it was building in 1961 (T. H. Simmons & Simmons, 2010) This s hows that, as discriminatory housing practices did not allow many housing choices for ethnic minorities, public housing had to absorb a large demand of housing by African Americans and Hispanics.

PAGE 334

320 Las Casitas Sun Valley project The history of public housing in Denver's Sun Valley neighborhood shows some typical traits of public housing in the Uni ted States: concentrated public housing projects in areas with low land value and residents with low socio economic s tatus ; land use planning allowing LULUs near these projects; and attempts to bring recreational opportunities near these projects (Bullard, 1993; Goering, Kamely, & Richardson, 1997; Massey & Kanaiaupuni, 1993; A. F. Schwartz, 2010) After building Las Casitas Project in the Sun Valley neighborhood in 1941, DHA started concentrating other public housing projects next to Las Casitas (Goldstein, 2011) Two other large housing projects were built in the area in the 1950s, which contributed to most single family homeowners gradually leaving Sun Valley in the following decades (Goldstein, 2011) After the two additional housing projects were built which increased the concentration of poverty in Sun Valley, the new 1956 zoning code allowed industrial land uses next to the neighbo rhood's housing projects (Goldstein, 2011) In the following decades, other undesirable land uses like the Interstate 25, large sport stadiums, and big parking lots were esta blished in Sun Valley (Goldstein, 2011) As a partial relief, the city built a recreation and sport center in Sun Valley through federal funds from the War on Poverty program in 1968 (Goldstein, 2011) DHA had been advocating for increased recreational opportunities in the area (Goldstein, 2011) From concentrated housing projects to dispersed single family homes According to Leonard and Noel (1990) DHA's approach to distributing public housing across the city changed between 1965 and 1976. While initia lly DHA concentrate d multi story buildings in a few sites, after 1976 to follow federal regulations DHA attempted to disperse public housing in various areas of the city by creating and refurbishing several

PAGE 335

321 single family homes (Leonard & Noel, 1990) However, in 1990 public housing sites w ere still disproportionately located in north, west, and central Denver (Leonard & Noel, 1990) which corresponded to the areas with the highest concentrations of industrial land uses. Current spatial distribution of affordable housing units Data from Denver's GIS catalogue (City and County of Denver, 2014b) on the location of affordable housing units (including public and subsidized housing) shows that affordable housing is still mostly concentrated in north, west and central Denver, with a few exceptions (see Figure 21). In particular, the highest concentration o f affordable housing units is located in Five Points, northeast of downtown (darkest shade of blue in Figure 21) Other areas with very high concentrations of affor dable housing units include La Alma/Lincoln Park, Jefferson Park, Sun Valley, and Ruby Hill, all located in Denver's West Side. Recently established affordable housing units are located in Stapleton and Lowry (shaded areas in east Denver in Figure 21).

PAGE 336

322 Figure 21 Density of affordable housing units in Denver. Map developed using data from City and County of Denver (2014) With the exception of Stapleton and Lowry and a few areas in northwestern Denver, most of the neighborhoods with the highest concentrations of affordable housing sites in Denver are currently loc ated in proximity to large transportation infrastructure, including interstate freeways and railways, and to industrial land uses. This confirms what had been observed for the public housing sites located in Sun Valley (Goldstein, 2011) Furthermore, Figure 21 shows that many of the neighborhoods with high concentrations of affordable housing sites have relatively low park provisions, with Stapleton and Lowry being the main ex ceptions. Barriers to establishing new public housing sites today Acquiring land for new public housing projects is the biggest obstacle that public housing authorities in Denver

PAGE 337

323 are currently facing when attempting to expand their housing stock, accordin g to a HUD Colorado staff interviewee As a consequence, most HUD funding is used to improve existing housing sites in Denver. This shows an interesting parallel with Parks and Recreation's struggles to acquire new parkland, and further highlights the Poun dstone Amendment's extremely negative impact on Denver's capability to source land for public uses. As the HUD Colorado staff member I interviewed explained, "this is one of the challenges of our land locked city." In particular: A lot of money will be spe nt where affordable housing, and mostly public housing sites already are. Acquiring land can be a very big expense, so generally a big chunk of money goes to improving or expanding housing sites that already have public housing. Given limited budgets, mone y to acquire new land is not always available. For example, some of the HOPE VI money went to La Alma/Lincoln Park, which already had a public housing site. Now they rebranded the site as Mariposa. Mariposa will have a mix of incomes, including some rental and some housing for sale. This means that, unless the economic and political system undergo a complete shift, public housing in Denver will most likely remain in the areas where it is currently located. Therefore, as the HUD Colorado staff member I inter viewed suggested, the goal of desegregating income groups in Denver can partially be accomplished by introducing controlled forms of gentrification in areas with large housing projects. A few HOPE VI projects in Denver, including the Mariposa and Park Aven ue projects, have included homes for sale targeted to middle class families. Given the lack of land, this seems the most pragmatic solution to create mixed income neighborhoods, according to the HUD Colorado staff member I interviewed. Part of this strateg y has been providing the tenants that were displaced from HOPE VI sites with housing vouchers to live in some of the city's subsidized housing sites (Cloud & Roll, 2011; Rusk, 2003) Thus, the HOPE VI program in Denver has contributed to desegregating economic groups by bringing

PAGE 338

324 middle class people to some housing projects and by relocating some low income families to subsidized housing sites, which have been generally located in neighborhoods with a slightly lower concentration of poverty than public housing sites (Rusk, 2003) To summarize, this brief historical review of the Denver Housing Authority shows that, l ike public housing authorities in many other American cities, DHA has struggled to establish public housing projects in desirable areas of the city and has mostly located projects in areas with low land values. Also, many of the areas with high concentrati ons of affordable housing units are fairly park poor which raises issues of park equity. Finally, given Denver's high cost of land, which contributes to DHA's inability to purchase new land for public housing, recent efforts to create mixed income neighbo rhoods have introduced controlled gentrifi cation in public housing projects. HUD recent programs An interview with a staff member of HUD Colorado provided very valuable information on some recent initiatives that HUD is undertaking nationally to increase affordable housing options and to improve neighborhoods at large. Some of these initiatives have or could have a positive impact in Denver including the Sustainable Community Initiatives (Marsh, 2014) and the Choice Neighborhoods program (U.S. Department of Housing and Urban Development, 2014) Both programs focus on improving neighborhoods as a whole, thus acknowledging that low income communities living in public housing have a range of needs that go beyond housing. The Sustainable Community Initiative (SCI) program distributes grants for long term planning, including housing, according to the HUD Colorado staff member I interviewed: "These grants give opportunitie s to local governments for long term planning efforts and to promote collaborations across institutional lines. Thus, SCI

PAGE 339

325 acknowledge s that fostering partnerships between public agencies is key to achieving better all around communities. The City and Coun ty of Denver obtained a SCI grant in 2011 to improve areas located in the West Side, and branded the initiative as Denver Livability Partnership (City and County of Denver, 2014a) As the HUD Colorado staff member I interviewed explained: Denver was a recipient of one of these grants. Denver's Planning Department was the most active department in this grant and the funds went mostly to support the area s around the West Line. These were historically low income areas that, thanks to the SCI grant, were targeted by a series of efforts, including: land use, small businesses stimuli, affordable housing, access to fresh food, better access to resources in gen eral. The Choice Neighborhood program, an evolution of HOPE VI, also aims to improve neighborhoods at large, starting from housing, according to the HUD Colorado staff member I interviewed. As with SCI, the Choice neighborhood program distributes grants to communities that include poorly maintained public or HUD subsidized housing and that aim to positively transform their neighborhoods through grassroots efforts (U.S. Department of Housing and Urban Development, 2014) As the HUD Colorado staff member I interviewed reported th e Choice Neighborhood program can contribute to improving one of Denver's most distressed neighborhood: The Choice Neighborhood program could be applied to the improvement of the Sun Valley neighborhood in Denver. The Choice Neighborhood program is intende d to have a positive neighborhood impact. It doesn't target only housing, but its goals are improving neighborhoods at large including schools, recreation, and transit. To summarize, the Sustainable Community Initiatives and Choice Neighborhoods programs a re promising for social equity because they focus on overall neighborhood improvement rather than only housing, thus attempting to make neighborhoods with high

PAGE 340

326 concentrations of poverty better places to live. These programs are housing centered but they ac knowledge the need of good parks, good schools and good transit. In particular, these programs can provide funding to improve or build parks, or other public green open spaces, in low income neighborhoods, thus contributing to addressing the recreational needs of park dependent populations. Denver Inclusionary Housing Ordinance (IHO) In 2002, Denver established an Inclusionary Zoning Ordinance (IHO) to create housing units that are affordable for moderate income households (Denver Office of Economic Development, 2011; Rusk, 2003) In particular, the IHO requires developers creating new complexes with 30 or more residential units for sale to include "Moderately Priced Dwelling Units" (MPDUs) that are affordable for families whose income is lower than 80 percent of the Denver area medium income (AMI; Denver Office of Economic De velopment, 2011; Rusk, 2003) The main issue of the Inclusionary Zoning Ordinance, according to Rusk (2003) and to a land use planner I interviewed, is th at it does not create truly mixed income developments. While the IHO is a step towards housing affordability in Denver, it is limited to households who buy properties (Rusk, 2003) Also, using 80 percent of Denver AMI of as a threshold is a way to target middle class instead of low income households (Rusk, 2003) According to a land use planner interviewee there have been two main reasons why the IHO has been drafted in its current form, thus not contributing to serving low income households. First, rental units are not included in the IHO because "in Colorado, we are prohibited to have rent control. Under the Inclusionary Housing Ordinance (IHO) in Denver, everything applies to condos." Second, the choice of using

PAGE 341

327 80 percent of AMI as a threshold derives from financial considerations, as lower thresholds would result in public subsidies : The other problem is that, at the lower leve l, if you want to build dwellings for families that make below 50 percent of the AMI, you almost always need public money. To fund these low income units through the sale of market price units, you would have to raise the price of the market price units to o much. The developer would find ways to get out of this law (i.e. building 29 units, one less than the 30 that triggers the Inclusionary Housing Requirements). The general focus of IHOs on creating units affordable to households in the 50 100 percent AMI range is not based on a desire to exclude poorer people from the benefit of this law, but based on the practical economics of a forced cross subsidy between market rate and subsidized units. This interview passage shows that the lack of public money which indicates a lack of public will, have prevented Denver's capacity to use the IHO to create for sale housing for low income people Also, Denver's officials were aware that, with a lower threshold of AMI developers would have identified mechanisms to oper a te outside the IHO. Partial failure of affordable housing in Stapleton The IHO's failure to provide affordable housing options for low income people is also evident in Stapleton (Duffy et al., 2010) Members of affordable housing organizat ions interviewed by Duffy et al. (2010) claimed that the IHO's income threshold (80 percent of AMI) contributed to Stapleton not being an actual mixed income community. Also, Forest City, the develo per, only agreed to set aside 1.6 percent of the total number of housing units for non profit housing organizations that serve households earning less than 50 percent of AMI (Duffy et al., 2010) A park planner interviewee formerly working at Parks and Recreation shared the perception of Stapleton's affordable housing shortcomings: "Stapleton has failed to provide enough affordable housing. I know the economic recession probably didn't help."

PAGE 342

328 Besides failing to provide enough affordable housin g units and to make some of them actually affordable for low income people, housing for different income levels are segregated in Stapleton. Affordable housing advocates interviewed by Duffy et al. ( 2010) argued that large one million dollar mansions are clustered on one side of the development, while the che apest rental units are concentrated in another area, which is partially isolated in Stapleton The analysis of Forest City's (2014) map of affordable housing sites in Stapleton partially confirms the former claim, as affordable housing sites are mostly located in the southern side of the development, in contact with low income ethnic minority existing neighborhoods. A park planner interviewee formerly working at Parks and Recreation pointed out that large mansions in Stapleton (and Lowry) were str ategically clustered around parks and at the higher elevations to enjoy mountain views: In Stapleton and Lowry, the big houses ended up being around the parks, and in the highest points of the developments. They were able to place mansions in points that c ould capture the views of the mountains. To me that's a failure of policy. On the other hand, two other experts I interviewed, including a scholar who studied Denver's planning and a Denver Urban Gardens member, argued that in some areas Stapleton achieved a good integration between housing types, which separates it from the standard developments in metropolitan Denver. In particular, the scholar explained that, although se gregation between housing types and incomes still exists, it is not as evident as in other neighborhoods: There [in Stapleton], you have large single family homes and side by side you have houses that look similar but are four plexes. That was something they did well in Stapleton and it was intentional. However, there is still segregation in Stapleton between large homes and starters single family small homes, but such separation is not as bad as in other suburban developments.

PAGE 343

329 Overall, even the more positive views abou t housing in Stapleton acknowledged that segregation within the neighbor hood is occurring. The partial failure of affordable housing provision and distribution in Stapleton can be link ed to the loose interpretation that Forest City made of the Green Book, which was the framework for Stapleton developed by the city in 1995 (Duffy et al., 2010) In particular, according to Duffy et al. (2010) Forest City adhered only to the parts of the Green Book that drove up their profits, including smaller lots, many par ks, and front porches aligned to the streets. Also, as the social equity goal included in the Green Book conflicted with Forest City's economic goals, Forest City reduced the amounts of mixed income housing sites that were included in the Green Book (Duffy et al., 2010) According to some critical voices reported in Duffy et al. (2010) the city lost its position of power and therefore the Green Book lost its relevance once Forest City took control of the project. Affordable housing and residential location My analysis shows that the policies and interventions aimed to increase affordable housing options in Denver have mainly contributed to reinforcing the patterns of residential segreg ation between socio economic groups, which have often reflected ethnic segregation. In particular, as Figure 21 shows, affordable housing units are disproportionately located in a few neighborhoods that have been historically inhabited by ethnic minorities (for example, compare Figure 21 with Figures 7 10), with a few exceptions. The choice of locating several public housing projects in these areas has derived from the areas' low land values, combined with DHA's low budgets. Similarly, property owners who t urned their properties into subsidized housing sites would have avoided doing such if their properties were located in desirable, highly sought after neighborhoods. Furthermore, the creation of three

PAGE 344

330 ethnically segregated housing projects in the 1940s (Goldstein, 2011) partially contributed to enhancing the patterns of residential segregation that Denver was already experiencing (Abbott, 1978; Dur‡n, 2011; R. L. Simmons & Simmons, 1995) My analysis also highlighted that the current high cost of land in Denver makes it unlikely for the geography of public housing to significantly change within the city. However, two recent HOPE VI projects have introduced middle class homeowners in low income areas (Cloud & Roll, 2011; Rusk, 2003) thus contributing to controlled g entrification in these areas T hese projects have also relocated previous tenants to areas with lower concentrations of poverty (Rusk, 2003) However, the controlled gentrification occurring i n the Mariposa and Park Avenue HOPE VI projects should be contextualized in wider market driven gentrification processes that are occurring in the La Alma Lincoln Park neighborhood (Robinson, 2008) and the Five Points neighb orhood (J. Koehler, 2014) where the two housing projects are located. Indeed, it is unlikely that middle class families would be attrac ted by for sale multi family housing units located in low income ethnic minority areas without the prospect of gentrification. The location of affordable housing units also has partial implications on low i ncome people's access to parks, as most areas with high densities of affordable housing sites are quite underserved by parks (see Figure 21). Housing Policies and Practices: A Summary This overview of housing policies and practices in Denver covered two large areas: ethnic and racial discrimination in hou sing and affordable housing. These two analyses aimed to uncover how housing policies and practices in Denver have contributed to today's residential location of different income and ethnic groups. Overall, affordable

PAGE 345

331 housing, which became a visible presen ce in Denver in the 1940s, has further enhanced patters of ethnic and socio economic segregation that were established through ethnic discriminat i o n in housing and through exclusionary land use interventions E thnic and racial discrimination in housing has been strongly limiting the residential choices of Denver's African Americans, while Hispanics have not experience d the same level of discrimination. Racially restrictive covenants were the first and the strong est instrument that Non Hispanic Whites used t o confine African Americans in a few neighborhoods of northeast Denver starting in the 1900s. When African Americans started to gain some room in east Denver in the 1920s, they were met with violence as racial intimidation further contribu ted to confining their geographies In the next decades, redlining contributed to financial disinvestment in Five Points and in a few Hispanic areas while gentlemen agreements among realtors kept limiting, de facto, the areas of the city where African Americans were allow ed to live. Fair housing laws passed in the late 1950s and early 1960s strongly limited the applicability of these discriminatory practices, thus expanding, at least officially, the housing options for residential minorities in Denver. A ffordable housing p olicies and interventions in Denver have contributed to consolidating the residential locations of low income ethnic minorities that ethnic discrimination i n housing had helped establish Low land values in low income communities of color and limited fundi ng in public housing have both contributed to siting public housing projects between the 1940s and the 1960ds in impoverished ethnic minority areas (see Figure 21 and Figures 7 10). Subsidized housing criteria that have not included proximity to amenities and currently hi gh land values have consolidate d the affordable housing location s that were already evident in the 1960s.

PAGE 346

332 While two recent large infill developments, Stapleton and Lowry, have introduced affordable housing in areas of the city that histori cally did not include them, financial limitations in the Inclusionary Housing Ordinance combined with development pressures have reduced the impact of low income oriented affordable housing in these two d evelopments. As Stapleton has very high park provisi on, the partial failure of affordable housing for low income people in the neighborhood is a missed opportunity to address citywide park equity issues (see Chapter 6). In the next section, I discuss the intersections between housing policies and land use p lanning in influencing the residential location of different demographic groups. I also analyze how residential locations of different income and ethnic groups have influenced, and have been influenced by, the establishment of public parks and parkways. Pa rk Planning Processes: Bringing It All Together In this section, I summarize the findings of the park planning processes analysis by establishing connections between park planning, land use planning, and housing policies and practices. This encompassing an alysis is aimed to uncover the processes that, during different phases of Denver's history, have led to establishing parks and to park location in relation to the geographic distribution of different demographic groups. While the analysis of park planning and part of the analys is of land use planning covered procedures that have been used to establ ish parks in Denver, the analyse s of housing policies and practices and of exclusionary land use instruments have focused on processes contributing to the residen tial location of different demographic groups. To create connections among the three previous analyses (park planning, land use planning, and housing policies and practices), I related their findings by themes, by

PAGE 347

333 geographical areas, and by historical per iods. Connections by theme involved, for example, comparing visions of equity or inequity in park planning and housing, or comparing funding mechanisms in park planning and land use planning. Connections by geography involved associating park planning, lan d use planning and housing policies and interventions that pertained to specific geographical areas within Denver, in order to understand their contributions to park establishment and to which demographic groups have lived in different parts of the city. F inally, connections by date included associating park planning, land use planning, and housing policies and interventions that occurred during different phases of Denver's history. This analysis, in particular, has contributed the main findings of this Cha pter which are represented in Figures 22 24. In this section, I present the most significant findings for park planning processes organized in three different historical periods: the first period spanning from 1902 to 1945, covering the City Beautiful and New Deal eras between the establishment of the City and County of Denver (1902) and World War II (see Figure 22); the second period spanning from 1946 until 1988, covering Denver's suburbanization phase until the last major land annexation to Denver (see Figure 23); and the third period spanning from 1989 and 2015, covering the most recent years of Denver's history, which saw a renewed interest in the central city and its public space (see Figure 24). These three periods mainly reflect the last three phase s of Denver's history, which I defined earlier in this chapter, with small changes in thresholds due to specific land use and park planning events. I chose to present the most significant findings based on these th ree periods because these periods were sub stantial ly dif ferent in terms of: political structures, availability of park funding, physical structure of the city, values attributed to parks,

PAGE 348

334 attitudes towards social equity and/or discrimination, the state of the national and local economies, and the specific policies and interv entions that derived from the above factors. Park Planning Processes: 1902 1945 The period covering the four decades between 1902 and 1945 was characterized, as previously described, by City Beautiful ideals in park planning an d urban design and by the work of two strong mayors, Robert Speer and Benjamin Stapleton, and of their park planners (Etter, 1986; Noel & Norgren, 1987) Figure 22 summarizes the processes that led to park establishment and partially to the residential location of low income ethnic minorities during this perio d. In particular, Figure 22 shows that a few political, planning, economic, and funding factors led to significant parkland acquisition and park establishment, and to a partial geographic equality in the distribution of park provision. City Beautiful ideal s, implemented mostly by M ayor Speer and his planners, advanced the concept that parks and parkways should serve all parts of the city (Etter, 1986; Noel & Norgren, 1987) However, due to ethnic discrimination in recreation, parks during the City Beautiful era did not serve all people in Denver. Thus, the City Beautiful movement in Denver achieved a partial geographic equality in park provision T he City Beautiful vision led to the creation of park master plans, including the Robinson Kessler plan (1907), which aimed to establish good provision of parks and par kways in every area of Denver (Etter, 1986; Noel & Norgr en, 1987)

PAGE 349

335 Figure 22 Park planning process in the City Beautiful and New Dear eras (1902 1945) The years between 1902 and 1945 were also characterized by a strong mayoral system, with a significant amount of power concentrated in the hands of M ayors S peer and Stapleton (Leonard & Noel 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) In particular, Robert Speer and George Cranmer (the manager of parks during Stapleton's administration), were strong political and planning actors who, without looking for the approval of their contempora ries, implemented their park visions with authority (Noel & Norgren, 1987; E. A. Taylor, 1986) Thus, during the City Beautiful and New Deal eras, Denver's park planning was dominated by a few actors, Speer and Cranmer, who

PAGE 350

336 frequently employed methods that their contemporaries considered as questionable (Noel & Norgren, 1987; E. A. Taylor, 1986) Mayors Speer's and Stapleton's strong political commitment to park establishment, combined with a partial lack of democracy at that time, contributed to the availability of significant park funding to acquire land and build parks and parkways (Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 1986) However, the park funding mechanism based on four park districts had n egative implications for park equity, as the wealthiest park districts, located mostly in south and east Denver, obtained more funding for parks and parkways (see park planning section). Additional factors that contributed to the establishment o f significa nt park acreage during this period have been the availability of land, including Denver's possibility to annex nearby towns, economic recessions that made land and labor cheap, and the willingness of developers to donate land for parks to increase the valu e of their subdivisions (see Figure 22). In particular, the two recessions of 1893 and 1929 facilitated parkland acquisition because of reduced competition from private interests (Leonard & Noel, 1990; Noel & Norgren, 1987) All the factors included in Figure 22 contributed to Denver's parkland doubling during Speer's administration and further increasi ng during Cranmer's years (Leonard & Noel, 1990; Noel & Norgren, 1987) However, the City Beautiful and New Deal eras were also characterized by a citywide system of ethnic discrimination in housing and public recreation tar geting mostly African Americans (Abbott, 1978; Cole, 2014; Leonard & Noel, 1990; Newsum, 2012) In years that saw the Ku Klux Klan dominating Colorado's politics (Goldberg, 1980) formal and informal systems of ethnic discrimination in housing were established,

PAGE 351

337 which forced African Americans to live in the Five Points g hetto (Leonard & Noel, 199 0; Mauck, 2001; R. L. Simmons & Simmons, 1995) Exclusionary land use p olicies and practices, comprising exclusionary zoning included in the 1925 ordinance and subdivision regulations, further limited the areas of the city where low income people and ethn ic minorities were allowed or could afford to live (Cole, 2014; Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010) Also, an informal system of discrimination limited the public recreation options that ethnic minorities had access to (Abbott, 1978; Leonard & Noel, 1990; Newsum, 2012; R. L. Simmons & Simmons, 1995) Therefore, while Denver's mayors and park planners between 1902 and 1945 implemented the City Beautiful vision of equality in recreation by establishing parks and parkways in most areas of the city, ethnic discrimination in housing and recreation and exclusionary land use interventions show that recreation was not really for everyone during th e 1902 1945 period. This evidence shows that park equality was not completely achieved during the City Beautiful and New Deal eras as African Americans were excluded from most city parks First, ethnic minorities faced manifest limitations when using publ ic recreation facilities (e.g. pools for whites only). Second, housing discrimination contributed to confining African A mericans to Five Points and part s of Whittier, which have been park poor areas. Third, the park district funding mechanism contributed t o establishing the city's largest parks and the highest number of parkways in east central and south central Denver, which have historically been Denver's wealthiest areas due to higher land values and subdivisions targeted to the city's elites. Fourth, ra cially restrictive covenants and discriminatory real estate practices prohibited ethnic minorities from living in park rich

PAGE 352

338 east and south Denver. Fifth, the implementation of the Residence A district around Denver's largest parks and most parkways through the 1925 zoning code made it very expensive to live near such green spaces, thus indirectly excluding low income people. This evidence shows that, although Speer and Cranmer successfully established parks in most parts of the city, ethnicity and socio eco nomic status mattered in determining a person's ability to enjoy public recreation in Denver between 1902 and 1945. In particular, large parks with high qua lity facilities located in Non Hispanic White neighborhoods were off limits for African Americans, a s shown by the 1932 protest in Washington Park. Park Planning Processes: 1946 1988 In the years after World War II metropolitan Denver experienced significant suburban growth, with Denver's suburbs increasing their population and geographic extension (Abbott, 1977, 1978) After World War II and until 1974, Denver had been aggr essively expanding its geographic boundaries by annexing land from its surrounding counties (Kurtz, 1957, 1962; Romero, 2013) Also, after a historical period in which decision makers considered park s as integral to the city, the period following World War II saw Denver reducing its planning efforts to create urbanity through meaningful public space and instead focusing on promoting a suburban environment within the city (Cole, 2014; Noel & Norgren, 1987) Significant cultural an d policy changes also occurred during this period, leading to further conflict between ethnic groups (Romero, 2013) The Civil Rights Movement and Denver's Crusade for Justice contributed to improving the living conditions of Denver's ethnic minoritie s.

PAGE 353

339 Figure 23 summarizes the processes that led to park establishment and partially to the residential location of low income ethnic minorities during this period. In particular, Figure 23 shows that parkland acquisition and residential locations have signi ficantly differed based on geographical location: Denver's older neighborhoods, hence referred as inner city, and Denver's outskirts, which included subdivisions annexed from other counties. Denver's inner city was almost completely developed by the 1940s, including several elegant streetcar suburbs built along the streetcar network (Leonard & Noel, 1990) Figure 15 (see land use planning section) shows that, by the 1950s, the City and County of Denver already included approximately two thirds of its current territory. In the inner city, fede ral investments targeted to improving low income neighborhoods contributed to establishing new recreation centers and small parks (see Figure 23). On the other hand, annexations in southeast, southwest, and northeast Denver have had a significantly positiv e impact on parkland acquisition until 1974, when the Poundstone Amendment was added to the Colorado Constitution.

PAGE 354

340 Figure 23 Park planning process in the post World War II period (1946 1988) Park planning processes in the outskirts: 1946 1988 Between the 1950s and the 1960s Denver Planning Office developed a few annexation policies that required developers annexing subdivisions to the City and County of Denver to contribute to

PAGE 355

341 public services by donating land for public use or by paying a cash in lie u of land fee (American Planning Association, 1958; Kurtz, 1957, 1962) Under this policy, large amounts of land were annexed until 1974 (Kurtz, 1957, 1962; Romero, 2013) a nd, as a consequence, significant park acreage was created in Denver's outskirts (see Figure 15). During the se decades of easy annexatio n, developers prefer red transferring land than paying fees, as land was generally available in large quantity and had re latively low value (Kurtz, 1962) Parkland acquisition through land annexation ca me to an abrupt halt in 1974 with the Poundstone Amendment to the Colorado Constitution. With neighboring counties having to approve any annexation to the City and County of Denver, annexations virtually stopped (Leonard & Noel, 1990; Romero, 2013; E. A. Taylor, 1986) As annexation policies had been the most effective tool that Denver had used to acquire parkland, the Poundstone amendment greatly limited Denver's capacity to establish new parks. After 1974, Denver annexed only a part of Adams County in 1 988, including the land for Denver International Airport and the Gateway area (Johnston, 2011) Until the early 1960s, Denver's African American and Hispanic residents were mostly concentrated in inner city areas located northeast, north, and west of downtown d ue to housing discrimination (Abbott, 1978; Diaz, 2005; Dur‡n, 2011; Leonard & Noel, 1990; Newsum, 2012; R. L. Simmons & Simmons, 1995) thus not benefitting from th e parks created in Denver's outskirts through annexation policies. A series of fair housing laws passed in the late 1950s and early 1960s partially opened some new subdivisions in Denver's outskirts to ethnic minorities, including African Americans (Leonard & Noel, 1990; T. H. Simmons & Simmons, 2010) For example, in the 1970s many African

PAGE 356

342 Americans moved to Montbello, a new subdivision located in far northeast Denver, which welcomed every ethnic group and included moderately low priced houses (Dwyer, 1993; Leonard & Noe l, 1990) By the 1980s, Green Valley Ranch, located east of Montbello, also included significant percentages of ethnic minorities (Dwyer, 1993) Therefore, while parkland was established through annexation policies in three of the four quadrants of the city (southeast, southwest and northeast) between the 1950s and 1988, ethnic minorities were mostly confined to a few areas of the inner city until the 1960s when civil rights legislation began to take effect Thus, the parkland established through this mechanism has served three quarters of Denver's outskirts but has not benefitted all Denver's ethnic groups in the same way. For these reasons, I conceptualized par k planning processes in Denver's outskirts between 1946 and 1988 as achieving partial geographic equality in park provision (see Figure 23). Park equity was not part of this process, as I have not found evidence of specific efforts to allow low income ethn ic minorities to live in park rich subdivisions in Denver's outskirts. Park planning processes in the inner city: 1946 1988 Park planning processes in Denver's inner city had been very different from the ones occurring in its outskirts after World War II. The reduction in park funding following the creation of the Parks and Recreation department in 1956 (see park planning section) contributed to lower park maintenance standards and to slowing down the establishment of new parks (Leonard & Noel, 1990; Noel & Norgren, 1987; E. A. Taylor, 198 6) As mentioned, this decrease in park funding is related to the lower value attributed t o parks after World War II, compared to the City Beautiful and New Deal eras.

PAGE 357

343 During a time of disin vestment in the public realm, of overall suburbanization (Leonard & Noel, 1990; Noel & Norgren, 1987) and of growing ethnic tensions in Denver (Moran, 1992; Romero, 2004, 2013) the Crusade for Justice emerged in the city as a cultural and political movement advocating for the rights of Chicanos, in the larger framework of the Civil Rights Movement (Langegger, 2013; Vigil, 1999) According to some experts I interviewed, these local and national movements advocating for social equity contributed to increasing the public's awareness of social issues in inner cities and of federal initiatives aimed to alleviate these issues. Among these initiatives, the War on Poverty program con tributed to funding several recreation centers and small parks located in low income ethnic minority areas along the South Platte River corridor in the 1960s and 1970s. According to my informants, these recreation centers and small parks significantly impr oved park provision in previously underserved areas. Although federal funding all owed the city to undertake equity oriented park planning actions, a few land use planning policies and housing practices still limited the geographies of ethn ic minorities in Denver's inner city. Discriminato ry real estate practices were strongly restricting African A merican's housing options in the inner city (Leonard & Noel, 1990; Newsum, 2012; T. H. Simmons & Simmons, 2010) In addition, the R 0 district in th e 1956 zoning code de facto excluded low income groups from some areas of the city, and some evidence shows that in Park Hill economic exclusion has been used as a proxy for ethnic exclusion (Cole, 2014) As many of the areas zoned as R 0 were in close proximity to Denver's largest parks and most historical parkways (see Figure 20), the R 0 district contributed to maintaining Non Hispanic White wealthy residents in areas around these green spaces. Finally, between the 1950s and 1960s public

PAGE 358

344 housing projects were mostly located in areas with low land values and inhabited by low income ethnic minorities (north, west, and central Denver), due to limited funding available for the Denver Housing Authority which reflect a lack of political will to invest in public housing Thus, public housing after World War II has not contributed to desegregating income groups in Denver. To summarize, federal initiatives like the War on Pover ty program made a positive impact on park equity in inner city Denver in the 1960s and 1970s. However, local land use policies and housing interventions still limited low income ethnic minorities' residential choices in the city, including restrictions to living in some areas near large parks and historic parkways. Thus, I conceptualized park planning processes in Denver's inner city between 1946 and 1988 as achieving partial equity (see Figure 23). Indeed, low income ethnic minorities benefitted from equit y oriented interventions to create recreation centers and small parks in their neighborhoods, but they were still excluded from areas with large parks and historic parkways. Park Planning Processes: 1989 2015 The period covering three and a half decades be tween 1989 and 2015 is characterized by a renewed interest in Denver's urbanity, after a histo rical phase of suburbanization (Goetz, 2013; Johnston, 2011; Leonard & Noel, 1990; Murray, 2002) Also, since 1988 Denver has bee n a landlocked city with no annexations (Johnston, 2011) which has significantly limited the city's capacity to acquire parkland. During this period, M ayors Pe–a and Webb undertook a series of large infrastructure projects including the Denver International Ai rport, the Colorado Convention Center, a new baseball stadium, schools, parks, large publicly drive n infill developments, and transportation projects

PAGE 359

345 (Goetz, 2013; Johnston, 2011; Leccese & McCormick, 2003; Leonard & Noel, 1990; Murray, 2002) These projects contributed to make Denver an attractive city for the creative class (J. Koehler, 2014) Also, starting in 2000, the city has experience d significant economic and population growths (Leeds School of Busin ess, 2015; United States Census Bureau, 2013b) Figure 24 shows the very complex process that has been leading to park establishment, park improvement and partially to the residential location of low income ethnic minorities in the last 35 years. Given th e lack of annexations since 1988, all new parks established during this period are located in infill sites. Thus, besides the parks created in Stapleton and Lowry, two large infill projects located on previously public land, Denver has only acquired parkla nd by purchasing properties that were on the real estate market. In this situation, the availability of funding has been key, as Parks and Recreation has had to compete against other private and public interests in sourcing and acquiring land. Thus, the co mplex process described in Figure 24 includes a variety of steps, sources of park funding, and actors that intervene in collecting and distributing park funding. Also, the description of this process is richer than previous two because I had the chance to interview professionals that are currently part of it. Overall, Figure 24 describes a process that, given park funding limitations and policies that do not require developers to contribute to the provision of public services like parks, raises serious park equity issues. Indeed, neighborhoods with strong HOAs like Stapleton and Lowry have been able to achieve higher park standards by creating privately owned parks. The reliance on privately owned parks and public space, which derives from funding issues, is likely going to create more park inequities in Denver.

PAGE 360

346 Figure 24 Park planning process in recent years (1989 2015)

PAGE 361

347 The Poundstone Amendment has dictated the political and geographical context of the process described in Figure 24, as all the mechanism s described in this process occur in infill develop ments. In particular, since 1988 the Poundstone Amendment has strongly limited Denver's capacity to acquire parkland in two ways: It has made it impossible for the city to acquire parkland through annexati on, and it has driven up property values within the city as land within the City and County of Denver has become a limited resource. Given these limitations, the way park funding is collected and distributed has become the main driver of the park planning process (see Figure 24). Funding for parks is fairly limited in Denver due to a variety of factors, including the absence of policies requiring developers to contribute to the provision of public services like parks in infill developments. Many of the lan d use planners and park planners I interviewed identified the lack of an impact fee for parks as the strongest limitation to Denver's capacity to collect funding for parks which is needed to address inequities in park provision Also, most of the Parks a nd Recreation's budget covers operation costs, with significantly less funding dedicated to capital improvement projects like reprogramming existing parks. The operating budget constitutes a very large share of Parks and Recreation's budget because park ma intenance is very expensive in Denver, due to the city's dry climate and Denverites' frequent use of parks. Also, Parks and Recreation's budget does not include yearly funds that are dedicated to parkland acquisition and park construction, thus the departm ent has had to rely on money from bond issues, which is available approximately every ten years, and from funding from donations or other public agencies.

PAGE 362

348 In this context, non profit organizations have helped fill some of the gaps of Denver's structural la ck of funding for parks. The contribution of non profit organizations has been significant to solving park equity issues in some low income ethnic minority neighborhoods, but so far they have not had a citywide impact in establishing new parks. In particul ar, Great Outdoors Colorado, probably the non profit organization that has the biggest contribution in terms of park funding in the last 20 years, has mostly supported park improvement projects and only contributed to parkland acquisition in four occasions in Denver. Another important aspect of the process described in Figure 24 is the political nature of park funding. Significant political negotiations occur when defining the items that are included in bond issues, which have shown power unbalances leading to funding inequities. The yearly capital improvement budget for parks and other department s is also very political. Several park planners pointed out that, while park plans and other analyses conducted by Parks and Recreation set funding priorities based on data, many budget decisions reflect the political will expressed by City Council more than data. Regardless of how decisions on park funding distribution are made, another factor that limits parkland acquisition in Denver is the financial commitment in terms of park maintenance that every new park involves. In some cases, Parks and Recreation has been hesitant to acquire new parks because the city did not commit to increasing the yearly budget for park maintenance. This issue ties back to the very high park maintenance costs in Denver, which Parks and Recreation has tried to lower by converting some areas covered with bluegrass into less water demanding natural landscapes.

PAGE 363

349 Given all these park funding limitations, and given Denver's landlocked nature, mo st of the parkland that has been established between 1989 and 2015 is located in Stapleton and Lowry, two large infill projects located on land that was previously publicly owned. The peculiarity of Stapleton and Lowry's parks is that they include both lar ge public parks and more diffused privately owned pocket parks, based on what park planners call a "trunk system." Denver was able to establish large public parks in these neighborhoods because they are developed on land that was previously city owned (Sta pleton) and previously federally owned (Lowry). Public land ownership gave the city significant negotiation power with developers. However, developers were also interested in establishing a significant amount of parkland as a means to attract prospective h omebuyers to these two developments. Well aware that "parks sell homes," developers in Stapleton and L owry not only agreed to dedicate part of the land to large public parks, but also decided to include in the developments a significant amount of privately owned parks and parkways. In Stapleton and Lowry, HOA fees pay for the ownership and maintenance of the private park system. Thus, HOA fees contribute to better park provision through privately owned parks and parkways. Privately owned parks raise equity concerns, especially if repeated across the city in other large infill developments. Indeed, residents of developments with wealthy HOAs can decide to increase HOA fees to achieve higher park provision. On the other hand, older and denser neighborhoods, fr equently populated by low income communities of color, do not have the financial capability and the possibility to source land to achieve high park provision through privately owned parks.

PAGE 364

350 Concurrently, although some affordable housing units have been est ablished in Stapleton and Lowry, the provision and distribution of affordable housing in Stapleton has not succeeded in creating an actual mixed income community (Duffy et al., 2010) Thus Stapleton's parks are mostly for the enjoyment of a middle and upper class predominantly Non Hispanic White population. This set of policies does not incorporate equity goals and does not lead to equitable park pro vision, as the neighborhoods that can afford to pay for privatel y owned public space can enjo y high park provision, while these areas do not include sufficient affordable housing unit s to integrate income groups (see Figure 24). Finally, the process described in Figure 24 shows that land use planning has failed to create instruments to establish p arks in low income dense neighborhoods, or to obtain funding for parks from infill developments. For example, the city policy for General Development Plans (GDPs) requires developers to set aside ten percent of the net development surface for open space, b ut developers can and often maintain ownership of these open spaces and they rarely develop them as parks. However, decisions about impact fees for parks or about any other instruments to obtain public land from developers in infill projects are made by ci ty officials in higher positions of power than departments' managers, including City Council and mayors. P opulation growth has meant growing pressure on Denver's parks, with an increasingly diverse range of users demanding space for various activities, and with augmented maintenance costs deriving from higher park visitation rates. Given the above analysis, the current park planning processes in Denver generate significant park equity concerns (see Figure 24). The positive note is the support that

PAGE 365

351 non prof it organizations are providing for park establishment and park improvement in park poor low income ethnic minority neighborhoods. The negative notes include the structural lack of park funding, also due to municipal policies not requiring developers to con tribute through impact fees, the impossibility to annex land and to establish parks through annexations, and the city's tendency to delegate part of park and open space provision to developers and HOAs Although none of these negative notes are explicitly intended to create park inequities their neglect of low income ethnic minority people's need for recreation and their focus on economic growth are having negative impacts on park equity. In this regard, these policies and practices can be considered as di scriminatory for recreation (U.S. Supreme Court, 2015) Park Planning Processes: A Brief Com parison The park planning processes described in Figures 22 24 highlight significant differences between the three periods in terms of political structures, availability of park funding, physical structure of the city, values attributed to parks, attitudes toward social equity and/or discrimination, the state of the national and local economies, and the specific policies and interventions that derived from all of the above factors. Political structures and city leaders' will to commit to establishing parks have largely changed between the 1902 1945 and the 1946 1988 period. While in the first period political power was concentrated in the hands of a few key personalities who strongly valued parks, in the second period the political system became more plural and, in general, city elites did not attribute the same importance to parks (Noel & Norgren, 1987) In the period between 1989 and 2015, Denver's political system and decision making became increasingly plural, with M ayor Pe–a encouraging citizens' participation

PAGE 366

352 in Denver's politics a nd administration (Leonard & Noel, 19 90) Parks gained a renewed interest under Pe–a's and Webb's administrations (Leonard & Noel, 1990; Noel, 2008) and today parks are still inherently valued by city leaders and Denverites. Differences in park fu nding are even more evident. While large amounts of money were available for parks in the 1902 1945 period, park funding was significantly lower in the two following periods. This reflects, for the 1946 1988 period, a reduced interest in parks and public s pace in general. The limited amount of money that the city dedicates to parkland acquisition and park improvement today is related to the lack of an impact fee for parks, as previously discussed. The variations in the city's physical structure are obvious, as Denver developed from a mining camp to a Western metropolis, with less and less land becoming available for parks. As extensively discussed, the Poundstone Amendment was a turning point in Denver's geography. The attitudes of c ity elites, planners and Denverites towards social equity and ethnic minorities have also significantly changed across the three studied periods. The period between 1902 and 1945 was characterized by numerous practices and policies discriminating against Denver's ethnic minorities and particularly against African Americans. The goals of achieving geographic equality in park distribution during the City Beautiful and New Deal eras had some positive consequences on park equity, but various practices of discrimination in housing and recreation limited ethnic minorities' residential and recreation options i n Denver. The 1946 1988 period wa s characterized by ambivalent attitudes towards ethnic minorities and low income people. On one side, housing discriminatory practices were in place until the 1960s and exclusionary zoning played a significant role in restricting low income people's residential choices. On the

PAGE 367

353 other side, during the 1960s social equity movement like the Crusade for Justice fueled federal initiatives that funded recreat ion centers and small parks in park poor low income ethnic minority neighborhoods. The 1989 2015 period has seen a stronger attention to park equity issues as shown by park master plans developed by Parks and Recreation and by the work of non profit organi zations. However, funding limitations and free market oriented municipal policies have limited Parks and Recreation's capacity to acquire parkland to solve park equity issues. This extensive analysis of park planning processes, including an introduction to Denver's history, highlighted the processes that have led to park establishment, park improvement, and park design in different phases of Denver's history. Also, this analysis uncovered some of the mechanisms that contributed to the residential locations of different ethni c and income groups in various periods of Denver's history. In terms of park planning, visions of park geographic equality, park equity, and explicit and subtle ethnic discrimination in recreation have characterized different periods and, in some cases, overlapped. The next chapter (Chapter 5) focuses on park spatial distribution s across income, ethnic and age groups, while the following one (Chapter 6) integrates the qualitative and quantitative datasets by establishing connections betwee n the policies and practices described in this chapter and the spatial distribution of parks.

PAGE 368

354 CHAPTER V PARK SPATIAL DISTRIBUTION FINDINGS In this chapter, I present the findings of a GIS spatial analysis of park distribution, which includes a focus on par ks that, based on a qualitative review of the literature on young people's outdoor play, comprise significant play spaces for children and teenagers. To analyze distributional justice, I studied the relationships between socio economic levels, ethnic varia tions and percentage of population under 18 (independent variables) and access to parks defined as park proximity, park acreage, and park quality (dependent variables, considered one at a time). In particular, I used the literature on young people's play i n parks, including how children's and teenagers' play preferences differ by ethnicity to develop a Park Quality Index for Youth (PQIY). Before presenting the results of the GIS analysis of distributional justice, I outline the literature review process tha t led to elaborating the PQIY. Developing a Park Quality Index for Youth To understand what determines young people's use of parks and their preferences for outdoor play spaces, I collected literature about children's and teenagers' outdoor play published in the last 30 years. In particular, this review is intended to investigate what are the park features that children and teenagers prefer or where they play more often, which is related to the reasons that contribute to park visitation, to longer engageme nt in play activities in parks, and to more physically active play To do so, I searched several full text academic databases for social science disciplines ( Web of Science, Science Direct, Jstor, and EBSCOhost) with the following expressions: ("children" AND "outdoor play"), ("teenagers" AND "outdoor play"), ("children" AND

PAGE 369

355 "play" AND "parks"), ("teenagers" AND "parks"), ("children" AND "playground"), ["children" AND ("schoolyard" OR "school ground" OR "play yard")], ("children" AND "nature"), and ["childr en" AND "outdoor play" AND ("ethnic" OR "race" OR "racial")]. The search yielded 61 entries (including journal articles, books, book chapters and doctoral dissertations) that focused on children's or teenagers' outdoor play in green spaces. Then, I expand ed this sample by looking at the references of the 61 entries and found 18 additional entries focusing on outdoor play. Thus, the final sample of scholarly sources I analyzed includes 79 empirical and review pieces on children's and teenager's outdoor play (see Table A4 in Appendix A for the complete list). To analyze these entries, I developed a codebook including descriptive codes, such as the name of the authors, the year of publication, the journal, book, or University for dissertations, and the geograp hical focus; topical codes, such as the scale of investigation, the methods used to measure children and teenager's play preferences or observed play behaviors, whether the entry focused on ethnic, income and gender differences; and analytical codes, such as the green space features that children and teenagers preferred, the green space features that were observed to be more often used by children and teenagers, ethnic and gender differences (when appropriate), and other detailed findings (Richards, 2005) The full codebook is reported in Table B4 in Appendix B. Dates, Locations and Methods The sampled entries span from 1984, with the early work by Francis (1984) to 2014 and 2015, when a variety of articles and book chapters have been published or are in press (Czalczynska Podolska, 2014; Jansson, Gunnarsson, MÂŒrtensson, & Andersson, 2014; Rigolon, Derr, & Chawla, 2015) The search also highlighted an increase of studies

PAGE 370

356 on children's outdoor play in the last five years, with studies being p ublished in the fields of public health, design and planning, environmental pyschology, and geography. The great majority of the sampled entries was conducted in goeographic locations within the United States (36 entries), followed by places in continental Northern Europe (15 entries, the vast majority of which in Scandinavian countries), the United Kingdom (9 entries), Canada (8 entries), Australia (8 entries), and other countries (Turkey and Martinique). Most of the selected studies focused on schoolyards (38 entries), followed by parks and playgrounds (26 entries), neighborhoods (11 entries), and housing sites (2 entries). In terms of goals most entries centered on observing play behaviors (56 entries), followed by studies investigating play preferences (44 entries), with 24 studies focusing both on observed play behaviors and play preferences. To study young people's play behaviors and preferences, the sampled entries used methods like behavior mapping, accelerometers, GPS tracking and survey s (for obse rved behaviors), interviews, focus groups, and other qualitative visual or engaging methods (for play preferences). In terms of demographic differences, 64 studies did not focus on ethnic or income differences, while 15 included ethnic/income variations or focused on specific ethnic minorities; and 36 studies measured gender differences, while 43 studies did not. What Green Space Features Matter for Children and Teenagers? The review of the 79 selected entries showed that the green space features that matte r for children and teenagers can be categorized in five main themes: the diversity of play settings, the presence of nature, the availability of enclosed spaces, park size, and park maintenance and cleanliness (see Table 19). Although some of the findings reported in Table 19 derive from studies conducted on school grounds and housing sites, these

PAGE 371

357 findings can also be applied to parks because they focus on specific spatial features (e.g. trees, sport fields, and play equipment) that can also be found in par ks and public playgrounds. Also, some of the studies focusing on schoolyards centered on observed play behaviors after school hours (Colabianchi, Maslow, & Swayampakala, 2011) in which schoolyards are unsupervised and used like parks or public playgrounds. Also, the five themes presented in Table 19 have some overlaps, which will be discussed after presenting the five th emes in detail. Table 19 Attractive Park Features for Children and Teenagers General Park Features Why Are These Features Attractive? Weight* Diversity of play settings Age preferences Gender preferences Ethnic preferences Diversity of opportunities 33% Presence of nature Diversity of play settings Loose parts Nature's aesthetics 30% Intimacy/enclosed spaces Prospect and refuge (safety) Place ownership (dens) Clarity of space Shading 22% Park size More things to do Possibility of avoiding gangs 7% Par k maintenance and cleanliness Safety 7% Related to the number of times these features are mentioned in the reviewed literature

PAGE 372

358 Diversity of play settings The findings of the reviewed studies show that multiple and diverse play settings, including a va riety of play amenities with different characteristics, are important for children's and teenagers' play for several reasons: providing a variety of physical challenges in relation to age and physical abilities (Czalczynska Podolska, 2014; Derr & Lance, 2012; Dyment et al., 2009; Loukaitou Sideris & Stieglitz, 2002; McCormack et al., 2010; Refshauge, Stigsdotter, & Petersen, 2013; Ries et al., 2008; Veitch et al., 2006) ; responding to di fferent preferences related to gender (Castonguay & Jutras, 2010; Fjrtoft, Kristoffersen, & Sageie, 2009; Karsten, 2003; R. C. Moore & Cosco, 2007) and to ethnicity (Loukaitou Sideris & Sideris, 2010; Loukaitou Sideris, 1995; Sallis, Mckenzie, Elder, Broyles, & Nader, 1997) ; and helping keep a prolon ged interest in the park due to multiple play choices (Ca stonguay & Jutras, 2009; Cosco, 2006; Jansson, 2008; McCormack et al., 2010) Outdoor spaces with several and diverse play amenities were connected to higher rates of use (Castonguay & Jutras, 2010; Colabianchi et al., 2011; R. C. Moore & Cosco, 2007) higher levels o f physical activity (Dyment & Bell, 2007) and to different levels of physical activity, including vigorous and moderate physical activity (Refshauge et al., 2013) In particular, play settings th at were often mentioned in relation to diversity of challenges and play opportunities include sport fields and courts (Castonguay & Jutras, 2009; Fjrtoft et al., 2009; Loukaitou Sideris & Stieglitz, 2002; Ries et al., 2008) fixed play e quipment (Fjrtoft et al., 2009; Jansson & Persson, 2010; Loukaitou Sideris & Stieglitz, 2002) and natural elements and trees (Cosco, 2006; Jansson & Persson, 2010; Refshauge et al., 2013; Stanley, 2011) which offer opportunities of manipulation (Czalczynska Podolska, 2014; Malone & Tranter, 2003; Tunstall, Tapsell, & House,

PAGE 373

359 2004) In terms of age, children, and espe cially infants, are more often seen playing with fixed play equipment in playgrounds (Loukaitou Sideris & Stieglitz, 2002) while adolescents seem to play more often in sport fields and courts (Loukaitou Sideris & Stieglitz, 2002; Ries et al., 2008) In some cases, children and caregivers complained about the fact that playgrounds are designed only for early ch ildhood users and do not offer appropriate play opportunities for middle childhood (Jansson & Persson, 2010; Jansson, 2008; Veitch et al., 2006) In terms of gender, males tend to prefer spaces for ac tive games (Baran et al., 2013; Fjrtoft et al., 2009; Waller, 2010) and sports (Castonguay & Jutras, 2010; Slater et al., 2013) with a particular focus on soccer (Karsten, 2 003) and other competitive games (Holmes, 2006; Slater et al., 2013) Females generally prefer quiet, observing and collecting activities (R. C. Moore & Cosco, 2007; Waller, 2010) which can be supported by small groves (Fjrtoft et al., 2009) as well as a variety of social activities (Holmes, 2006; Slater et al., 2013; A. Timperio et al., 2008) and were more often observed in playgrounds than boys (Castonguay & Jutras, 2010; Dyment et al., 2009; Karsten, 2003) In particular, Karsten (2003) pointed out that the playground is a very gendered setting where there is a physical separation between the spaces where boys and girls play. Although some ethnic differences in observed play behaviors and play preferences exist, they are not as evident as they are for gender. Among the 15 studies that measured ethnic differences, 10 found notable d ifferences and five did not. For example, Loukaitou Sideris and Sideris (2010) found that Hispanic children and youth utilized soccer fields more often than other groups, African Americans and Asian young people used basketball fields more often than other groups, and Non Hispanic W hite

PAGE 374

360 young people used baseball fields more frequently than other ethnicities. In addition, some studies found that children of different ethnicities play together in playgrounds or in organized sport activi ties (Blatchford et al., 2003; Loukaitou S ideris, 1995, 2003) while in other situations play groups can be ethnically segregated (Loukaitou Sideris, 1995) These differences suggest that, when play activities are organized and structured, young people from different ethnicities and cultures are more likely to play together than when play is self guided. Also, sports fields and organized sports activities like L ittle L eague games seem to really matter for ethnic minorities, especially for Hispanic children and teenagers (Loukaitou Sideris & Sideris, 2010; Perry et al., 2011; Ries et al., 2008) To summarize, the diversity of play amenities is the feature that matters the most for children and teenagers, as various play settings can res pond to different preferences and abilities related to age, gender, and ethnicity, and as diverse play amenities can sustain children and youth's interest in play for a longer time. Presence of nature This extensive review of the literature shows that chi ldren greatly appreciate nature as a space for play (Castonguay & Jutras, 2009; Jansson, 2008; Loukaitou Sideris, 2003; Ozdemir & Yilmaz, 2008) that they tend to prefer nature to a rtificial environments (Nasar & Holloman, 2013; Nedovic & Morrissey, 2013; Stanley, 2011; Titma n, 1994) and that their engagement while playing in nature lasts longer than in other settings (Derr & Lance, 2012; Herrington & Studtmann, 1998; Luchs & Fikus, 2013; A. F. Taylor, Wiley, Kuo, & Sullivan, 1998) The reasons for children's connection to nature as a space for play are several: Children like the multiple affordances that nature can offer (Dyment & Bell, 2007; Dyment & Reid, 2011; Jansson et al., 2014; Samborski, 2010; Stanley, 2011; Titman,

PAGE 375

361 1994; Tunstall et al., 2004; Ward Thompson, 2007) ; children like to play with the loose parts that nature includes like sticks, ro cks and leaves, which provide children with opportunities for manipulation and customization (Cosco, 2006; Czalczynska Podolska, 2014; Dyment & Bell, 2007; R. C. Moore, 1986b; Nedovic & Morrissey, 2013; Rigolon et al., 2015) ; and children like nature's aesthetic qualities, as nature's many details and ever changing qualities throughout the day and the seasons can lead to memorable experiences (McCormack et al., 2010; R. C. Moore, 1986b; Rigolon et al., 2015; Tunstall et al., 2004) The multiple affordances that nature can provide deal with the flexibility of use that natural elements can affo rd, if compared to the static and predetermined uses of playground equipment (Dyment & Reid, 2011) For example, forests and green spaces along rivers can offer a plethora of opportunities of adventurous explorations (Stanley, 2011; Tunstall et al., 2004; Ward Thompson, 2007) Since nature af fords a variety of play activities, it can provide opportunities for a range of physical activity levels, including vigorous (Cosco, 2006; Dyment & Bell, 2007; Nedovic & Morrissey, 2013) and moderate physical activity (Dyment et al., 2009; Dyment & Reid, 2011) In particular, Dyment et al. (2009) found that green areas in schoolyards are important for children who do not like to play sports or other competitive games. The naturalized part of this schoolyard included a "large grassed area with trees, rocks, tree stumps and sandpits," as well as "native trees and shrubs" Dyment et al., 2 009, p. 264) Thus, these unstructured natural play amenities can provide additional opportunities for play for children who are not interested in playing games with rules.

PAGE 376

362 Also, since nature can provide gradual physical challenges, it can foster children 's motor development during early to middle childhood (Fjortoft & Sageie, 2000) The loose parts that children can find in natural environments, including sticks, rocks, water, sand, and leaves, can provide children with opportunities to manipulate and customize the environment (Burke, 2005; Czalczynska Podolska, 2014; Malone & Tranter, 2003) By manipulating loose parts around bushes a nd trees, children can build special places that they can call their own and that they can feel attached to, like dens (Rasmussen, 2004; Sobe l, 1993; Titman, 1994) The building games that children play together to build dens and other spaces can also create opportunities for socialization, as children work together for a common goal (Nedovic & Morrissey, 2013; Rasmussen, 2004; Stanley, 2011) My review also showed that children's preference for natural play settings can be found across the board, in a variety of urban settings. These settings include school grounds (Dyment & Reid, 2011; Malone & Tranter, 2003; Samborski, 2010; Stanley, 2011; Titman, 1994) parks, playgrounds and greenways (Jansson, 2008; R. C. Moore & Cosco, 2007; Tunstall et al., 2004; Ward Thompson, 2007) green spaces loc ated in low income neighborhoods (Castonguay & Jutras, 2009) and green spaces located in housing e states (Castonguay & Ju tras, 2009; A. F. Taylor et al., 1998) However, two studies fou nd that older children, around 6th grade, preferred more structured play spaces like playgrounds (Jansso n et al., 2014; Stanley, 2011) as they perceived nature as too messy and complex (Jansson et al., 2014) or as they were looking for places for larger group activities (Stanley, 2011) Furthermore, all the literature mentioned in this section about

PAGE 377

363 nature play focused on children. Therefore, there is less evidence showing teenagers' attraction and interest in nature as a play setting. T o summarize, the presence of nature is very important for children's play for the multiple affordances that nature provides, for the loose parts that nature offers, and for nature's aesthetic quality, which can all lead to memorable experiences in nature f or children. Intimacy and enclosed spaces Another important aspect that matters for children's use and positive experiences in outdoor spaces, as well for their caregivers' comfort, is a moderate to high enclosure of play settings, which can be provided b y trees, shrubs, and built elements (Czalczynska Podolska, 2014; Dyment & Bell, 2007; Heerwagen & Gregory, 20 08) This enclosure, which is defined here as intimacy, is experienced positively by children for a variety of reasons. First, in semi enclosed spaces, children can experience what Appleton (1975) defined as "prospect and refuge," leading to a sense of safety (Heerwagen & Gregory, 2008; Kirkby, 1989; Kylin, 2003) Second, semi enclosed spaces can also provide children with a sense of place ownership, as shown by dens (Jansson, 2008; Kirkby, 1989; Rasmussen, 2004; Titman, 1994) Third, trees and shrubs creating a boundary around a play setting can increase its spatial clarity (Czalczynska Podolska, 2014) which is also appreciated by caregivers who can better control their children's play (Dyment & Bell, 2007) Finally, if the enclosure is created by trees, intimacy provides additional benefits in terms of shading (Derr & Lance, 2012; Dyment & Bell, 2007; McCormack et al., 2010) and the play opportuniti es provided by trees (Jansson & Persson, 2010; Jansson, 2008; Refshauge e t al., 2013)

PAGE 378

364 The theory of prospect and refuge, introduced by Jay Appleton (1975) postulates that people in outdoor spaces look for a balance between seeing others (pro spect) and not being seen by others (refuge). In particular, prospect is a feature of places in which people can observe others, while refuge is found in settings where people can retreat and find their privacy ( Appleton, 1975) Children look for this balance because they generally seek certain levels of privacy from adults (M. Berg & Medrich, 1980) while adults aim to supervise their children when they play (Veitch et al., 2006) as expressed by Roger Hart's (1978) model of children's exploration of the physical world. In addition, the creation and use of dens is related to prospect and refuge, as dens are refuges where children can see peop le and things around them without being seen (Kirkby, 1989; Kylin, 2003) thus feeling safe. In dens, c hildren can also find privacy (Titman, 1994) and a sense of identity (Rasmussen, 2004) Children perceive areas enclosed by vegetation as outdoor rooms, as places they feel attached to, and that provide opportunities to social ize (Herrington & Studtmann, 1998) Trees, shrubs and built structures located around play settings like playgrounds can also create a physical boundary that increases the spatial clarity of such play settings (Czalczynska Podolska, 2014) Caregivers intending to supervise their children at play particularly value these visual and physical boundaries because the latter facilitate their supervising role (Dyment & Bell, 200 7) Children also value these boundaries, as the boundaries could represent the limits or what Roger Hart (1978) defined as "predictable environments ," which are comfort zones where risk is almost absent (p. 388). However, Hart (1978) observed that childr en seek out manageable risks. Predictable environments are related to what Hart (1978) called "attachment behavior," which leads children to

PAGE 379

365 being more confident in their explorations if they know they have a safe haven to return to, in this case represent ed by the "predictable environments" (p. 388). Therefore, the sense of clarity provided by physical and visual boundaries can foster behaviors related to both children's attachment to their caregiver and to children's willingness to explore new spaces and to seek privacy. When the enclosure is created by trees, intimacy provides additional benefits in terms of shading, which is valued both by children and adults (Derr & Lance, 2012; Dyment & Bell, 2007; McCormack et al., 2010) and it is partic ularly relevant in a place like Denver which has around 300 sunny days per year (City and County of Denver, n.d.) Shading, when appropriately placed around play amenities, is important for children's and caregivers' use of parks (Derr & Lance, 2012; McCormack et al., 2010) and can have a positive effect on physical activity levels (Dyment & Bell, 2007) Also, shading created by trees near rest spaces was correlated to children's use of schoolyards after school hours (Colabianchi et al., 2011) Therefore, shaded intimate areas can be important for active and passive activities and for socialization. Trees creating enclosure around play settings like playgrounds can also provide additional play opportunities for children (Jansson & Persson, 2010; Jansson, 2008; Refshauge et al., 2013) In particular, when trees were placed around a playground, children we re observed engaging in unstructured play activities like climbing trees and hiding behind them (Jansson & Persson, 2010; Jansson, 2008) and expressed their appreciation for the presence of trees (Jansson & Persson, 2010) In summary, semi enclosed play settings leading to a sense of intimacy can have a series of benefits in terms of psychological well being and sense of safety, which can be

PAGE 380

366 explained through Appleton's (1975) theory of prospect and refuge and Hart's (1978) theory of children's exploration of the physical world ; and b enefits in terms of physical comfort due to shading and increased play opportunities. Park size Several studies showed that park or playground size matters for children's and teenager's park utilization (Baran et al., 2013; Castonguay & Jutras, 2010; Jansson & Persson, 2010; Loukaitou Sideris & Sideris, 2010; Slater et al., 2013) These findings were particularly significant for parks located in low income ethnic minority inner city neighborhoods ( Castonguay & Jutras, 2010; Loukaitou Sideris & Sideris, 2010; Slater et al., 2013) due to the shortage of green space in these neighborhoods (National Recreation and Park Association, 2011) Larger parks provide teenagers with opportunities to avoid the spaces where gangs gath er, while in smaller parks the contact with gangs is more likely (Slater et al., 2013) Besides safety benefits, large parks and playgrounds often include more play settings than small green spaces. In a study conducted in Sweden, larger public playgrounds offered more play opportunities for children than small playgrounds (Jansson & Persson, 2010) Finally, park size matters more for teenagers than for children (Baran et al., 2013) as teenagers probably look for spaces for sports and larger group activities. Park maintenance and cleanliness Oth er features that positively influence children's and teenager's park utilization are park maintenance and cleanliness (Gearin & Kahle, 2006; Karsten, 2003; Loukaitou Sideris & Sideris, 2010; Loukaitou Sideris & Stieglitz, 200 2; McCormack et al., 2010; Ries et al., 2008) Park maintenance and cleanliness can be related to children's and teenagers' preference for safe and aesthetically pleasing environments. As in the case of park size, well maintained and

PAGE 381

367 clean parks are parti cularly important for park use in low income ethnic minority inner city areas (Gearin & Kahle, 2006; Loukaitou Sideris & Stieglitz, 2002; Ries et al., 2008) This is due to the fact that parks in low income ethnic minority inner city neighborhoods have lower levels of maintenance and more physical hazards than parks in other areas of cities (Carlson et al., 2010; Cradock et al., 2005; Smoyer Tomic et al., 2004) Playground maintenanc e matters for playground utilization more for girls than for boys (Karsten, 2003) Overlaps between park features Four of the five park features that matter for children's and teenagers' use of parks have significant overlaps that need to be discussed. Figure 25 shows the intersections among the diversity of play settings, the presence of nature, enclosed spaces, and park size. In particular, the diversity of play settings has significant overlaps with all the other three features, which confirms that having multiple and different play amenities is a key feature for park utiliza tion. The intersection between the diversity of play settings and the presence of nature is expressed through the fact that nature includes a variety of interesting affordances for children's play. Trees placed around playgrounds creating both a sense of e nclosure and opportunities for unstructured play express the overlap between the diversity of play settings and enclosure/intimacy. The diversity of play settings and park size have overlaps because larger parks are more likely to include several play oppo rtunities. Finally, the intersection between the presence of nature and intimacy/enclosure can be found in dens and vegetated rooms, which are natural child constructed spaces where children can experience a sense of refuge.

PAGE 382

368 Figure 25. Overlaps among the park features that matter for children and teenagers Operationalizing These Findings in a Park Quality Index for Youth (PQIY) To include the findings of the literature review on young people's outdoor play in my GIS spatial analysis of park distribution, I created a Park Quality Index (PQIY) based on the findings of this review. In order to build an operational PQIY, I combined the results of the literature review with a series of pragmatic considerations dealing with the available GIS data from the City and County of Denver, and with the approaches used in other indexes developed to classify public spaces (Kaczynski, Stanis, & Besenyi, 2012; NÂŽmeth & Schmidt, 2007) Table 20 shows the PQIY, which is organized in four main categories: structured play diversity, nature, size, and maintenance. Each category includes several variables, for a total of 19 v ariables (see Table 20). Compared to Table 19, which shows five attractive park features for children and teenagers, I condensed the categories "presence of nature" and "intimacy/enclosed space" into "nature" (see Table 20). This choice derives from signif icant overlaps in the literature between codes about nature and codes about enclosure (e.g. dens and the importance of vegetation).

PAGE 383

369 Table 20 Operationalized Findings A Park Quality Index for You ng People (PQIY) Categories Variables Score Structured p lay diversity Playground number 0: no playground 1: 1 playgrounds 2: 2 or more playgrounds Playground surface 0: below median playground surface 1: above median playground surface Sport fields and courts (soccer, baseball, football, rugby, and lacros se) 0: no sport field 1: sport field of 1 type 2: sport field of 2 or more types Sport courts (basketball, tennis, skateboard, and handball) 0: no sport court 1: sport court of 1 type 2: sport courts of 2 of more types Biking and walking paths 0: n o biking/walking paths 1: biking/walking paths are included Supporting facilities (picnic areas, benches, and bathrooms/water fountain) 0: no picnic area/benches or bathroom 1: picnic area/benches or bathroom 2: picnic area/benches and bathroom Public swimming pool 0: no swimming pool 1: swimming pool Organized sport activities (after school programs or rec centers) 0: no organized sport activities 1: organized sport activities Nature Water elements (lakes, streams, or fountains) 0: no water element s 1: water elements Tree coverage 0: tree coverage park acreage ratio in the lower tercile 1: tree coverage park acreage ratio in the middle tercile 2: tree coverage park acreage ratio in the upper tercile

PAGE 384

370 Table 20 Continued Categories Variables Score Nature Tree cover type: deciduous, evergreen 0: prevalence of one type (percentage of prevalent tree cover type above 75 %) 1: balanced of deciduous and evergreen (percentage of prevalent tree cover type below 75 %) Natural areas (per Denver GIS park la yer) 0: no natural areas 1: natural areas Medium to high enclosure of behavior settings: enclosed for more than 50% of perimeter 0: no behavior settings are enclosed 1: one behavior setting is enclosed 2: two or more behavior settings are enclosed Gardens 0: no gardens 1: gardens Flower beds 0: no flower beds 1: flower beds are included Mountain views 0: no mountain views 1: mountain views Size Park acreage 0: park acreage in the lower quartile 1: park acreage in the middle lower quartile 2: pa rk acreage in the middle upper quartile 3: park acreage in the upper quartile Maintenance Level of maintenance per Denver Parks and Recreation standard 0: "Yellow" level, low maintenance 1: "Blue" level, medium maintenance 2: "Green" level, high mainten ance Year of last playground renovation 0: before the median year of last renovation 1: after the median year of last renovation

PAGE 385

371 The City and County of Denver's GIS shapefiles describing parks and their amenities are very detailed and allow ed me to aut omate most of the variables described in Table 20. Assigning the scores included in Table 20 through automated GIS processes is very important, as Denver includes approximately 300 urban parks, and digitizing 19 variables for 300 parks would be very time c onsuming Also, the automated assignment of quality scores to the 19 variables is another merit of the PQIY, as the PQIY can be easily calculated for other cities that have similar GIS data sets to Denver's. To define the values of the specific scores for the 19 variables included in Table 20, I chose to follow an approach similar to the one outlined by NÂŽmeth and Schmidt (2007) For the most significant variables included in each category, I allowed a variation between 0 and 2, with "0" being the lowest quality, "1" being average quality, and "2" being the highest quality. This was intended to capture more nuances in Denver's parks and, for example, to differentiate between parks with no playground (score of 0), with one playground (score of 1), and with two or more playgrounds (score of 2). An exception to this rule is the score for park size, which can vary between 0 and 3 to better capture differences in terms of park size across the city. Structured play diversity The diversity of structured play opportunities is one of the major themes that emerged from the literature review. As previously noted, diverse play opportunities are important to accommodate different play preferences related to gender, age, ethnicity, as well as to personal likings. All the variables included in this category (number of playgrounds, playground surface, sport fields, sport courts, supporting facilities, biking and walking paths, swimming pools, and organized sport activities) are available from the Denver's GIS data sets. The choice of including

PAGE 386

372 supporting facilities like picnic areas, benches, restrooms and drinking fountains reflects the need of spaces for social interactions, particularly felt by teenagers (Gearin & Kahle, 2006) by girls (McCormack et al., 2010; Slater et al., 2013) and by caregivers supervising their children at play (R. C. Moore & Cosco, 2007) In addition, several studies showed ethnic minorities like Hispanics, African Americans and Asians particularly appreciate and use spaces for social gathering like picnic areas (Gobster, 2002; R. C. Moore & Cosco, 2007) Also, assigning higher scores to parks that have multiple types of sport fields (for example baseball and soccer fields) and multiple types of sport courts (for e xample basketball and volleyball courts) means that parks that can host several sport activities have a higher quality, as young people can find more diverse play opportunities. With the same goal I included playground number and playground surface as pro xy variables for the number of fixed play equipment features that young people can find in parks. Finally, I included organized sport activities, expressed in the PQIY through afterschool clubs located in parks and through the presence of recreation center s, because these group activities are particularly important for ethnic minority young people, especially Hispanics (Loukaitou Sideris & Sideris, 2010; Perry et al., 2011; Ries et al., 2008) Nature The p resence of nature is another key theme highlighted by the literature review on young people's outdoor play. As mentioned above, the category Nature included in Table 20 comprises the presence of natural elements and the presence of enclosed, intimate behav ior settings. As shown in the literature review, natural spaces can foster unstructured and unguided forms of play, which can appeal young people who seek quiet and unregulated play, including playing with loose parts (Czalczynska Podolska,

PAGE 387

373 2014; Dyment et al., 2009; Dyment & Bell, 2007; Nedovic & Morrissey, 2013; Refshauge et al., 2 013) To operationalize water elements like streams and lakes, I used GIS shapefiles provided by the City and County of Denver. Tree coverage was expressed by the ratio between the surface of trees within a park and the total surface of that park, excludi ng water bodies (where trees cannot be planted). This ratio identifies the degree to which a park is rich in trees. The tree cover type, comprising deciduous and evergreen trees, was included as a means to describe vegetation diversity, which provides a va riety of visual stimuli that is appreciated by children (McCormack et al., 2010; R. C. Moore, 1986b) Natural areas a re included in Denver's GIS park shapefile and they describe zones that the Parks and Recreation Department intentionally leaves natural, comprising wildlife habitat and native vegetation. The variable expressing the presence of behavior settings with med ium and high levels of enclosure was operationalized as ranging between 0 and 2. The behavior settings that I considered in this analysis included playgrounds, sport courts, and sport fields. The enclosure variable was assigned a value of 0 if vegetation w as enclosing behavior settings for less than 50 percent of their perimeters; a value of 1 if enclosure was higher than 50 percent in playgrounds or sport facilities; and a value of 2 if enclosure was higher than 50 percent in both playgrounds and sport fac ilities. The tree canopy shapefile that Denver developed through remote sensing allowed me to automate the computation of trees enclosing behavior settings. The location of gardens, another variable included in the Nature category, can be retrieved from t he Denver Urban Garden's website (Denver Urban Gardens, 2015) Finally, mountain views were included as an aesthetic element connecting the city with

PAGE 388

374 its natural mountain backdrop, and because t he literature review showed that young people value the aesthetic s of nature. The presence of flower beds, which is comprised in Denver's GIS park layer, was also included in the PQIY for the visual, tactile, and olfactory quality that flowers add to parks To operationalize mountain views in the PQIY, I used two proxy variables: the inclusion of "Scenic Views" in Denver's GIS park layer, and the presence of a "View Planes" ordinance applied to city parks (City and County of Denver, 2013) The View Planes ordinance is intended to protect mountain views in some of Denver's parks. Size The simplest variable to compute was park acreage, as each park surface was easily calculated through ArcMap's command "Calculate Geometry." As mentioned earlier, park size matter s for young people because large parks generally afford more activities and because in large parks children and teenagers can avoi d people they perceive as dangerous, such as gangs (Slater et al., 2013) To capture enough variation in park acreage, the score for this variable ranges between 0 and 3, with 0 representing the lowest quartile for Denver's parks (smallest parks), and 3 representin g the highest quartile (largest parks). Maintenance To operationalize park maintenance, I chose to rely on the Parks and Recreation Department's maintenance standards (Denver Parks and Recreation, 2007) and on the year when the last playground renovation was performed. The Parks and Recreation maintenance standards describe three levels of maintenance based on the importance of parks and on their level of use. The highest level of maintenance is the "Green" level, which is applied to regional parks that, based on the Parks and Recreation data, have high rates of use (Denver Parks and Recreation, 2007) Parks with "Green"

PAGE 389

375 maintenance levels were assigned a score of 2. The medium level of maintenance is described as "Blue," which deals with neighborhood parks that have medium level of use (Denver Parks and Recreation, 2007) In the PQIY, "Blue" maintenance parks have a score of 1. Finally, the lowest level of maintenance, defined as "Yellow," describes areas with wildlife and native vegetation (Denver Parks and Recreation, 2007) Parks with "Yellow" maintenance levels were assigned a score of 0. The maintenance levels defined by Denver's Parks and Recreation were used in the PQIY because they describe levels of investment in terms of park maintenance. Also, given the large dataset (300 parks), surveying all Denver's parks to assess their maintenance and cleanliness through site visits would have been extremely time consuming. The schedule of cleaning and maintenance would also have affected the results of a field survey focused on maintenance. The year when the last playground renovation was performed expresses a level of investment in amenities for structured play activities. This variable, which is included in Denver's GIS playground layer, c an be an indicator of more recent investment and better playground maintenance. In the PQIY, I assigned a scor e of 0 to this variable if the last renovation was done before the median year of the last renovation across all parks, and a score of 1 if the last renovation was done after the median year. Although parks can get a score of 1 for this variable only if th ey include a playground, I chose to include this variable in the maintenance category because the choice of building a playground in a park also shows a commitment to park maintenance. Score assignment The way I assigned scores partially adds weighs to th e different categories. As Table 20 shows, most of the variables related to Structured play

PAGE 390

376 diversity can vary between 0 and 2, and the category includes eight variables. While Nature also includes eight variables, most of the variables' scores range betwe en 0 and 1; Size includes one variable, while Maintenance includes two. I argue that the higher weigh of Structured play diversity to PQIY is a means of taking into account preferences for different play spaces depending on gender, age, and ethnicity. Sum ming up the highest scores for each variable in the PQIY (see Table 20), the total possible score would be 28, including: twelve points from the Structured play diversity category; ten points from the Nature category; and three points each from the Size an d Maintenance categories. The four quartiles of the PQIY were used to classify parks in three classes of quality. Parks with a PQIY in the upper quartile of the distribution (upper 25 percent) were classified as high quality parks; parks with PQIYs in the two middle quartiles as mid quality park (from lower 25 percent to upper 25 percent); and parks with PQIYs in the lower quartile as low quality parks (lower 25 percent). A similar approach was used in a study focusing on access to parks in Portland, OR (NÂŽmeth et al., 2012) Besides the overall PQIY, I also classified parks based on their partial PQIY scores for the Structured play diversity and Nature categories. For each of these two categories, I classified parks as high quality if their partial PQIY score for its category was in the upper quartiles of the distribution (upper 25 percent), as discussed in Chapter 3 (Research Methods). Thus, I created a dichotomous variable for Structured play diversity quality, and a dichotomous variable for Nature quality. Looking at Structured play diversity and Nature sep arately is a means to identify areas of the city that are specifically deficient in parks with a variety of settings for structured play, or that are deficient in

PAGE 391

377 parks with natural elements. Other parameters of park quality based on the PQIY can be found in Table 9 in Chapter 3. Reliability of GIS data To double check the reliability of the City of Denver's GIS data, I examined the aerial photos and Google Street View images of 10 percent of Denver's city parks. In particular, I selected a random sample o f 30 Denver's parks and double checked the reliability of the information related to most of the variables included in the PQIY. The only variables that I did not control were park acreage, which is calculated through the park geometry, and park maintenanc e, which is defined by the Denver's Parks and Recreation department. For all other variables, I used aerial photos of Google Street View to double check that the information reported in the Denver's GIS park layer was accurate. For example, I checked the p resence of playgrounds, sport fields, gardens, lakes and streams, and mountain views. Aerial images have been used in other research measuring the quality of parks (B. T. Taylor et al., 2011) As Denver's GIS shapefil es are updated constantly, it was not surprising to find great correspondence between the GIS data and the aerial and Google Street View images. The reliability analysis of the PQIYs comparing scores from the Denver's GIS data with scores from the analysis of areal photos for the random sample of 30 parks yielded a Cronbach's (1951) al pha coefficient of 0.967, which can be considered satisfactory when comparing two different instruments on the same subjects, in this case parks (Leech et al., 2011) The items that included the most evident variations, although small, were the presence of biking and walking path s, which are hard to distinguish from sidewalks in aerial pictures; the presence of natural areas, wh ich was hard to discern from

PAGE 392

378 simply undeveloped portions of parks; and the presence of flower beds, which are hard to detect in aerial pictures if pictures are taken during the winter season. Validation of the PQIY The PQIY is a youth specific instrument that assesses the quality of a park based on the literature I reviewed in this chapter. In an attempt to validate the PQIY by evaluating its concurrent validity, I looked for other instruments that have been used to measure the quality of parks in Denver. Concurrent validity can be assessed by comparing scores from a test with scores from a second instrument that is considered as valid (Mislevy & Rupp, 2010) Due to the lack of specific instruments evaluating park quality for children and youth t hat scored Denver's parks, I decided to rely on the scores that Denver's park users attributed to parks on the Yelp website (Yelp.com, 2015) Although not a scientific instrument, the Yelp scores reflect the perceptions of quality of some Denverites who have used, at least once, one of Denver's public parks. Table 21 reports the scores that were accessed on Yelp.com on April 20, 2015. Not all Denver's parks were ranked, so the parks included in Table 21 are the ones tha t received at least 10 reviews. Parks with less than 10 reviews were not included because of the variability that can be associated to a very low number of reviews. Also, two of the parks in Table 21 were excluded from the concurrent validity analysis beca use, when reading the comments to the parks, reviewers ranked only a small portion of the park (Berkeley Park) or because reviewers misunderstood a park for the whole park system along the South Platte River (Confluence Park).

PAGE 393

379 Table 21 Comparison between Yelp Scores and PQIY Scores for a Few Denver's Parks Park Yelp Score Number of Reviews PQIY Score Washington Park 4.5 153 26 Cheesman Park 4 103 17 City Park 4 86 25 Sloan's Lake Park 4 52 23 Civic Center Park 3 32 10 Berkeley Park 3.5* 32 24 Confl uence Park 4.5^ 27 7 Fuller Dog Park 3.5 28 10 Barnum Park 3.5 11 22 City of Cuernavaca Park 3.5 10 11 James A. Bible Park 4 10 17 *Most reviews are about the dog park. ^Most reviewers confuse it with the whole South Platte River park system In order to assess differences between the Yelp and the PQYI scores, after excluding the two parks mentioned above, I transformed the Yelp score, initially a 0 5 scale, into a 0 28 scale, like the PQYI. Table 22 compares the Yelp adjusted score, calculated by mult iplying the Yelp scores in Table 21 by 5.4, which represents the ratio between 28 and 5, and the PQYI score. The mean value for the Yelp adjusted score was 20.4, while the mean PQYI score was 17.88.

PAGE 394

380 Table 22 Comparison between Yelp Adjusted Scores and PQ IY Scores Park Yelp Adjusted Score PQIY Score Washington Park 24.3 26 Cheesman Park 21.6 17 City Park 21.6 25 Sloan's Lake Park 21.6 23 Civic Center Park 16.2 10 Fuller Dog Park 18.9 10 Barnum Park 18.9 22 City of Cuernavaca Park 18.9 11 James A. Bible Park 21.6 17 To test the null hypothesis that the means of the scores calculated based on the two instruments did not differ significantly, I conducted a dependent samples t test (Leech et al., 2011) Assumptions for dependent samples t tests were assessed and met (Leech et al., 2011) The t test showed that the PQIY and Yelp scores do not differ significantly, p = .1623, thus confirming the null hypothesis. In other words, this result shows that the scores calculated for the nine parks inclu ded in Table 22 do not significantly differ based on which instrument is used (PQIY or Yelp). Therefore, the t test confirms the concurrent validity of the PQIY when comparing it with the scores reported on Yelp.com.

PAGE 395

381 Park Spatial Distribution In this sect ion, I present the results of my Geographic Information System (GIS) and statistical analyses of park distribution. In particular, I discuss how park access defined as park proximity, park acreage, and park quality varies by income, ethnicity, and the perc entage of people under 18 years of age. As described in Chapter 3 (Research Methods), the findings I present below derive from logistic regressions and from a series of non parametric tests comparing medians and mean ranks. This analysis of distributional justice is intended to uncover whether the demographic groups who need parks the most, including low income groups, ethnic minorities, and young people, have equal or better access to parks than Non Hispanic White middle and upper class groups. My analys is of park spatial distribution showed more case s of distributional inequity than cases of partial distributional equity. The results on park proximity vary between distance to all parks and parkways combined and distance to all parks only. The analysis of park acreage shows patterns of inequity, especially when considering park acreage per 1000 people and per 1000 people under 18. Finally, th e analysis of park quality highlight s a surprising picture of partial equity when considering some variables describ ing the top quartile of Denver's parks based on the PQIY. However, significant inequities in park spatial distribution were found for parks in the top five percent of the PQIY, parks with low violent crime density, and regional parks highlighting a substa ntial picture of environmental injustice for young people's park visitation The next sections describe patterns of inequity and partial equity in park distribution, including maps displaying the parts of the city that are well served or deficient in terms of park proximity, park acreage, and park quality.

PAGE 396

382 Descriptive Statistics The City and C ounty of Denver includes 4,233 acres of parks, and 5,531 acres of parks and parkways combined (City and County of Denver, 2014b) These acres translate into 6.51 acres of parks per thousand po pulation, and 8.52 acres of parks and parkways per thousand population (City and County of Denver, 2014b) which do not satisfy the National Recreation and Park Association Standard of 10 acres of parkland per 1,000 residents (Mertes & Hall, 1996) and are significantly below the 12.5 median value calculated for the 100 most populated U.S. cities (The Trust for Public Land, 2015a) Denver has a slightly lower park acreage per thousand population than Baltimore, M D, which sat at 7.6 in 2009 (Boone et al., 2009) and significantly lowe r than Portland, OR with a 23.7 value in 2015 (The Trust for Public Land, 2015a) However, Denver a higher value than Los Angeles, which included 4.2 acres per 1,000 people in 2005 (Wolch et al., 2005) Table 23 summarizes Denver's population and park acreage descriptive statistics. Table 23 Population and Parkland Statistics in Denver, CO Category Total Population, 2013 649,4 95 Population under 18 years of age, 2013 128,766 Population density (people per square mile) 4,357 Parkland (acres) 4,233 Parkland as a percentage of total area 4.43 Park acres per 1,000 residents 6.51 Park acres per 1,000 residents under 18 years o f age 32.87

PAGE 397

383 Denver's 4,233 acres of parks are distributed between 298 parks, including publicly and privately owned parks, and excluding golf courses and cemeteries (City and County of Denver, 2014b) Denver also includes 87 registered parkways that create additional 1,298 acres of parkland (City and Cou nty of Denver, 2014b) Finally, Denver's 649,495 residents in 2013 were distributed between 8,094 census blocks that included at least one resident. Denver Parks and Recreation divides the city's parks in four maintenance districts, which roughly corresp ond to four large quadrants of the city (City and County of Denver, 2014b) Figure 26 shows the parks included in the four maintenance districts and Table 24 summarizes their park acreage and number of parks. The east district, that actually spans the southeast quadrant of the cit y, has the highest park acreage, with substantial differences compared to the rest of the city (see Table 24). The northeast district has the lowest amount of parks, and has recently benefitted from the parkland addition in Stapleton and Lowry. The northwe st district also has low park acreage and the acreage is distributed in the highest number of parks (92). This means that Northwest Denver has many small parks with a low average acreage (around 11 acres) and lack s large parks, with the exception of Berkel ey Park, Rocky Mountain Lake Park, and Sloan's Lake in the extreme northwest corner of the city (see Figure 26). The southwest district has the second highest amount of park acreage, which is boosted by a few large parks, including Washington Park, Ruby Hi ll Park, and Bear Creek Park.

PAGE 398

384 Table 24 Park Acreage and Number of Parks in the Four Maintenance Districts East Northeast Northwest Southwest Acreage 1646.96 1058.34 1075.24 1323.44 Number of parks 58 55 92 74 Figure 26 The four park maintenance districts, including privately owned parks Before presenting the findings on park spatial distribution, it is useful to discuss the residential locations of income and ethnic groups in Denver. Figures 27 30 represent the spatial distribution of the three i ncome groups (three terciles), of African Americans, Hispanics, and Non Hispanic Whites, based on census block and census block group data.

PAGE 399

385 In particular, Figure 27 shows that the highest concentrations of low income groups are located in west Denver, foll owed by areas in central Denver and northeast Denver. The low income ter cile represents yearly incomes between $4,568 and $43,967 the mid income tercile includes values between $ 43,968 and $71,550, and the high income tercile comprises incomes between $71 ,551 and $353,433. While university students mostly inhabit the low income areas of central Denver, west and northeast Denver include significant percentages of African Americans and Hispanics (see Figures 28 and 29). Some pockets of low income population are also found in far northeast Denver, mostly in the Montbello neighborhood (see Figure 27). Figure 27 Spatial distribution of the three income groups in Denver The highest concentration of high income groups is found in east central and south central Denver, including historic neighborhoods like South Park Hill, Montclair,

PAGE 400

386 Hil l top, Congress Park, Cherry Creek, Country Club, and Washington Park (see Figure 27). Other areas including high income groups are located in northwest, southeast and southwest De nver (see Figure 27). Mid income groups are quite evenly distributed throughout the city, with only a few areas in west and far northeast Denver having significant concentrations of mid income census blocks (see Figure 27). Overall, while some patterns of income segregation can be found in Figure 27, the city also includes areas where three income groups are rather integrated, comprising currently gentrifying areas northeast of downtown (Five Points, Whittier, Clayton, and Skyland), and areas northwest of d owntown (Highland, Sunnyside, and Sloan's Lake). Figure 28 Spatial distribution of African Americans in Denver Denver's Harlem, the area located in Five Points where African Americans were segregated until the 1950s, is no longer on the map (see Figure 28). Denver's African

PAGE 401

387 Americans are currently concentrated in areas located north and northeast of City Parks, including Skyland, North Park Hill, and Northeast Park Hill, and in far northeast, comprising Montbello and Gateway Green Valley Ranch (see Figur e 28). Other concentrations of African Americans, with less significant percentages, can be found in historically African American Five Points, Whittier, and Cole, and in East Colfax (see Figure 28). Although Denver's African American population is signifi cantly less segregated than in the past, African Americans are still substantially concentrated in the northeastern quadrant of the city (see Figure 28). However, their presence on the map is rather limited, as they comprise only 10 percent of Denver's pop ulation. Figure 29 Spatial distribution of Hispanics in Denver Denver's Hispanics show signs of significant residential segregation, as evidenced in Figure 29. Also, Hispanics are very visible on the map, as they comprise

PAGE 402

388 approximately 30 percent of Den ver's population. The macro area with the highest concentration of Hispanics is west Denver, often referred to as West Side (see Figure 29). This area incorporates Sun Valley, West Colfax, Villa Park, Barnum, Barnum West, Valverde, Westwood, Athmar Park, R uby Hill, and parts of Mar Lee. These areas also include substantial aggregations of low income groups, as displayed in Figure 27. Historically Hispanic areas of northwest Denver, namely Sunnyside, Chaffee Park and Globeville, still include significant con centration of Hispanics (see Figure 29). Areas in northeast and far northeast Denver, comprising Five Points, Cole, Elyria Swansea and Montbello, also comprise large percentages of Hispanic population (see Figure 29). Concentrations of Hispanics can very r arely be found in east central, south central, east, and south Denver (see Figure 29). Figure 30 Spatial distribution of Non Hispanic Whites in Denver

PAGE 403

389 Non Hispanic Whites, Denver's largest ethnic group with approximately 53 percent of the total populat ion, are mostly concentrated in areas of east central, south central, east, south, and southeast Denver (see Figure 30). Other parts of the city including high concentrations of Non Hispanic Whites are far southwest Denver and parts of northwest Denver (se e Figure 30). It is interesting to note that very few Non Hispanic Whites live in the West Side and in northeast Denver, while some degree of ethnic residential integration can be found in the gentrifying areas in northwest and near northeast Denver that I previously mentioned (see Figure 30). Overall, ethnic residential segregation is still very evident from Figures 28 30. Summary of Findings My geospatial analysis highlighted more cases of distributional i njustice than cases of partial justice in park pro vision Inequities in park spatial distribution mostly deal with park acreage, access to safe parks, access to regional parks, and access to parks with excellent quality. Cases of inequity and partial equity are summarized in Tables 25, 26, and 27, which d isplay, respectively, the most significant findings in terms of park proximity, park acreage, and park quality. African Americans are the ethnic minority group that is most underserved by parks in Denver, particularly in terms of park acreage and park qual ity (see Tables 26 and 27). Hispanics have slightly better access to parks than other ethnic groups in terms of park proximity when considering parks only (see Table 25). However, Non Hispanic Whites and African Americans live in closer proximity to parks and parkways combined than Hispanics (see Table 25). Yet, African Americans have access to parkways with very poor quality, which cannot be used for active or passive recreation and that can be considered as a danger for children's safety.

PAGE 404

39 0 Similarly, low i ncome groups have better access to parks than other income groups in terms of park proximity when including only parks, but worse access when including all parks and parkways (see Table 25). Hispanics and low income groups have also good access to parks wh en considering some PQIY variables describing the top quartile of Denver's parks which I defined as parks with "good enough" quality (see Table 27). However, Non Hispanic White s mid and high income groups have significantly better access to the parks in the top five percent of the PQIY, which I described as "excellent" quality, to parks with low violent crime density, and to regional parks than low income ethnic minorities (see Table 27). These are the most substantial differences, which raise serious equ ity concerns for low income young people of color. Also, it is important to note that most of the findings reported in Tables 25, 26, and 27 describe differences among income and ethnic groups in terms of distances to parks and park acreage or differences in demographics between areas with access to certain parks and areas without access (defined based on the quarter mile threshold). As explained in Chapter 3 (Research Methods), I only ran statistical tests comparing mean ranks and medians because none of t he variables describing access to parks were normally distributed. Results from the logistic regression, although often statistically significant, never displayed moderate or large effect sizes.

PAGE 405

391 Table 25 Summary of Statistical Findings on Access to Parks : Park Proximity Variable Findings Distance to any park or parkway High income groups have the lowest median distance to any park or parkway and they are overrepresented in areas within a quarter mile of any park or parkway African Americans have the low est median distance to any park or parkway, followed by Non Hispanic Whites and Hispanics African Americans, however, have access to parkways with very poor quality Distance to any park Low income groups have the lowest median distance to any park and ar e overrepresented in areas within a quarter mile of a park Hispanics have the lowest median distance to any park, followed by Non Hispanic Whites and African Americans Note: All differences are statistically significant The findings displayed in Table 25 show pattern of partial equity for distance to any park. Low income census blocks have a lower median distance to the closest park (328 meters) than mid income (386 meters ) and high income census blocks ( 393 meters ), with relatively small differences. Also predominantly Hispanic census blocks (above 75 percent of population ) have a shorter median distance to the closest park (314 meters) than mainly Non Hispanic White (391 meters ) and mainly African Americans census blocks (500 meters) The analysis of par k proximity also highlights partial inequities for distance to any park or parkway. High income census blocks have a slightly shorter median distance to the closest park or parkway (240 me ters) than mid income (300 meters ) and low income census blocks (280 meters) Although African Americans have a slightly lower median distance to the closest park or parkway than other ethnic groups the parkways they have access to are of very poor quality

PAGE 406

392 Table 26 Summary of Statistical Findings on Access to Parks: Park Acreage Variable Findings Acres of parks and parkways within 1/4 mile of a census block High income groups have the highest median acreage of parks and parkways African Americans have the highest median acreage of parks and parkways, followed by Non His panic Whites and Hispanics (low quality parkways in African American areas) Acres of parks and parkways within 1/4 mile of a census block / 1,000 people High income groups have the highest median acreage of parks and parkways per 1,000 people African Amer icans have the highest median acreage of parks and parkways per 1,000 people, followed by Non Hispanic Whites and Hispanics Acres of parks and parkways within 1/4 mile of a census block) / 1,000 people under 18 High income groups have the highest median a creage of parks and parkways per 1,000 people under 18 Non Hispanic Whites have the highest median acreage of parks and parkways per 1,000 people under 18, followed by African Americans and Hispanics Acres of parks within 1/4 mile of a census block Low i ncome groups have the highest median acreage of parks but high income groups have the highest mean acreage Hispanics have the highest median acreage of parks but Non Hispanic Whites have the highest mean acreage Acres of parks within 1/4 mile of a census block / 1,000 people Low income groups have the highest median acreage of parks per 1,000 people but high income groups have the highest mean acreage per 1,000 people Hispanics have the highest median acreage of parks per 1,000 people but African Americans have the highest mean acreage per 1,000 people Acres of parks within 1/4 mile of a census block) / 1,000 people under 18 No statistically significant differences among median, but Non Hispanic Whites have the highest mean acreage per 1,000 people under 18 (no statistical test) Non Hispanic Whites have the highest median acreage of parks per 1,000 people under 18, followed by Hispanics and African Americans Note: All differences between medians are statistically significant, unless otherwise noted

PAGE 407

393 The fi ndings on park acreage, summarized in Table 26, show a pattern of inequity, with several nuances. Inequities are more evident when considering acreage per thousand people and per thousand people under 18. Park acreage per 1,000 people and park acreage per 1,000 people under 18 can be considered as proxy variables to evaluate possible park congestion issues (Boone et al., 2009; Si ster et al., 2009; Wolch et al., 2005) Also, the findings on park acreage highlight interesting differences between the groups with the highest median values and the groups with the highest mean values (see Table 26). The meaning of these differences in t erms of park provision is discussed in detail in the section on park acreage. Park acreage variables for all parks and parkways substantially advantage high income groups over other income groups and African Americans and Non Hispanic Whites over Hispanic s (see Table 26 ; note that parkways in African American areas are of very low quality ). For these variables, differences in medians and means generally had the same directions. High income census blocks have a higher median acreage of parks and parkways pe r 1,000 people (17.11) than census blocks with other income groups (10.17 for mid income and 9.21 for low income) with quite large differences Census blocks with high concentrations of African Americans (above 75 percent) have a larger median acreage of parks and parkways per 1,000 people (22.89) than census blocks with concentrations of other ethnic groups (14.03 for Non Hispanic Whites and 6.56 for Hispanics). Similar patterns are observed for acreage of parks and parkways per 1,000 people under 18, wit h high income groups and Non Hispanic Whites having higher values than other income groups and ethnic minorities. Hispanics are particularly disadvantaged when focusing on the variables of park and parkway acreage.

PAGE 408

394 The situation is slightly different when considering only the acreage of parks. For park acreage per 1,000 people, low income groups have the highest median value (3.46), b ut high income groups have the largest mean value (19.01). Also, while mainly Hispanics census blocks have the highest median value of park acreage per 1,000 people (4.11), census blocks with high concentrations of African Americans have the highest mean value (18.31). The differences between the groups with the highest median and mean values signify that more low income and His panic census blocks have access to average levels of park acreage, while more high income and Non Hispanic Whit e census blocks have better access to excellent park acreage, including values higher than 10 acres per 1,000 people. However, the relatively low median values cause concerns over park overcrowding When considering park acr eage per 1,000 people under 18, high income group s have a higher mean value ( 101.04 ) than low income groups ( 56.04 ) Also, predominantly Non Hispanic Whi te census blocks have th e highest median and mean values of park acreage per 1,000 youth (18.26 and 101.11). The findings on park quality, which are displayed in Table 27, show patterns of inequity and of partial equity depending on the variable. In particular, Hispanics and low income groups tend to have better acces to parks in the top quartile of the PQIY (see Table 27). However, ethnic minorities and low income groups have lower access to parks in the top five percent of the PQIY, to parks with low violent crime density, and to regional parks, than Non Hispanic Whites and mid and high income groups (see Table 27). African Americans have very low access to the high quality parks described in Table 27, regardless of the variable.

PAGE 409

395 Table 27 Summary of Statistical Findings on Ac cess to Parks: Park Quality Variable Findings Distance to high quality PQIY parks (top quartile of PQIY) "good enough" quality Low income groups have the lowest median distance to high quality PQIY parks and are overrepresented in areas within a quarter mile of these parks Hispanics have the shortest median distance to high quality PQIY parks similar median distance for African Americans and Non Hispanic Whites Distance to parks with high quality PQIY play Low income groups have the lowest median dis tance to parks with high quality PQIY play and are overrepresented in areas within a quarter mile of these parks Hispanics have the shortest median distance to parks with high quality PQIY play similar median distance for African Americans and Non Hispan ic Whites Distance to parks with high quality PQIY nature Mid income groups have the lowest median distance to parks with high quality PQIY nature Hispanics have the shortest median distance to parks with high quality PQIY nature, followed by Non Hispani c Whites and African Americans, who are very underserved by these parks Distance to parks with high quality PQIY play and nature Mid income groups have the lowest median distance to parks with high quality PQIY play and nature and are overrepresented in a reas within a quarter mile of these parks Hispanics have the shortest median distance to parks with high quality PQIY play and nature, followed by Non Hispanic Whites and African Americans, who are very underserved by these parks Distance to best PQIY par ks (top 5% of PQIY) "excellent" quality High income groups have the lowest median distance to best PQIY parks and are overrepresented in areas within a quarter mile of these parks Non Hispanic Whites have the lowest median distance to best PQIY parks, fo llowed by Hispanics and African Americans Note: All differences between medians are statistically significant, unless otherwise noted

PAGE 410

396 Table 27 Continued Variable Findings Distance to parks without LULUs High income and low income groups have slightly be tter access to parks with no LULUs than mid income groups Non Hispanic Whites and Hispanics have shorter median distances to parks without LULUs than African Americans Distance to parks with low violent crime density High income groups have the lowest me dian distance to parks with low violent crime density Non Hispanic Whites have the lowest median distance to parks with low violent crime density, followed by African Americans and Hispanics Distance to regional parks Areas with access to regional parks h ave higher median household income than areas with no access Non Hispanic Whites have the lowest median distance to regional parks, followed by African Americans and Hispanics Note: All differences between medians are statistically significant, unless oth erwise noted These findings reported in Table 27 show that, while Hispanics and low income groups have better access to the parks that, based on the PQIY, were defined as having "good enough" quality they have poor access to the parks with "excellent" qua lity and to regional parks Also, low income ethnic minority communities lack parks with low violent crime density, which significantly limit s children's and teenagers' park visitation in these communities. This is the most substantial environmental justic e issue in terms of park distribution that emerged from my analysis. The spatial distribution of high quality PQIY parks slightly advantages low income groups and Hispanics over other income and ethnic groups, with relatively low differences (see Table 27) For example, low income census blocks have a shorter

PAGE 411

397 median distance to the closest high quality PQIY park (600 meters) than mid income (666 meters) and high income census blocks (764 meters). The findings for parks with high quality play based on PQIY s tandards are very similar. Parks with high quality nature based on the PQIY show a slightly different pattern, with mid income groups having the shortest median distance (670 meters), followed by low income groups (693 meters) and high income groups (860 m eters). Census blocks with high concentrations of Hispanics have the shortest median distance to parks with high quality PQIY nature (557 meters), followed by Non Hispanic Whites (788 meters) and African Americans (1612 meters). These findings show African American neighborhoods have a very low provision of parks with high quality PQIY nature. The findings for parks with high quality play and nature based on the PQIY showed a pattern of partial inequity Mid income census blocks have a lower median distance to parks with high quality PQIY play and nature (800 meters) than low income census blocks (914 meters) and high income census blocks (1111 meters). Census blocks with the highest concentrations of Hispanics have the lowest median distance to parks with h igh quality PQIY play and nature (640 meters), followed by Non Hispanic Whites (972 meters) and African Americans (2000 meters). This further shows African American's disadvantage in terms of high quality parks. S ignificant inequities emerged when studying access to best PQIY parks i.e., the parks in the top five percent of the PQIY scale. High income census blocks have the shortest median distance to best PQIY parks (1483 meters), followed by mid income census blocks (1595 meters) and low income census blo cks (1931 meters). Census blocks within a quarter mile of a best PQIY park have significantly higher median household

PAGE 412

398 income and percentage of Non Hispanic Whites than census blocks located beyond a quarter mile. Also, predominantly Non Hispanic White cens us blocks have a shorter median distance to best PQIY parks (1471 meters) than census blocks with strong concentrations of Hispanics (2022 meters) and African Americans (2500 meters). Overall, the main findings of the analysis of park quality based on the PQIY highlight that none of the shortest median distances to high quality parks is below a quarter mile (400 meters), which is the widely accepted threshold used to categorize areas with access and with no access to parks These findings also show that, w hen studying park quality based on an index, researchers need to be careful in defining threshold s for quality and be aware of results obtained by using different thresholds. Table 27 also includes findings on park quality not based on PQIY variables whic h show a strong pattern of inequity T he analysis of access to parks without locally unwanted land uses (LULUs) showed ver y small differences for income and ethnicity. The analysis of access to parks with low violent crime density yielded some of the bigge st inequities. High income census blocks have the shortest median distance to the closest park with low violent crime density (613 meters), followed by mid income census blocks (1054 meters) and low income census blocks (1562 meters). Thus, half of the you ng people living in low income census blocks has to walk at least a mile to reach a park with low violent crime density. Census blocks within a quarter mile of parks with low violent crime density have substantially higher median household income and perce ntages of Non Hispanic Whites than census blocks beyond a quarter mile. Also, predominantly Non Hispanic White census blocks have a shorter median distance to parks with low violent crime density (640 meters) than census blocks with strong

PAGE 413

399 concentrations o f African Americans (1358 meter s) and Hispanics (2500 meters), who are particularly disadvantaged. T he differences between median distances are the highest among all variables included in Tables 25 27. Finally, the analysis of access to regional parks show ed that census blocks within a quarter mile of regional parks include households with significantly larger median income s and higher percentage of Non Hispanic Whites than areas beyond a quarter mile. Also, Non Hispanic Whites live closer to regional parks than ethnic minorities. P redominantly Non Hispanic White census blocks have a shorter median distance t o regional parks (2061 meters) than mainly African American census blocks (2500 meters) and predominantly Hispanic census blocks (3250 meters). These et hnic differences are similar to the ones observed for best PQIY parks. To summarize, the analysis of park spatial distribution in Denver showed a multifaceted landscape of environmental injustice, with some cases of partial justice. African Americans are s ignificantly disadvantaged based on most variables, while Hispanics and low income groups have good access to parks in terms of park proximity (when considering parks only) and based on some park quality variables. The biggest inequities that low income gr oups and all ethnic minorities experience include access to parks with low violent crime density, access to parks with excellent quality, access to regional parks, and some variables describing park acreage. The next sections include the detailed analyses of park proximity, park acreage, and park quality, comprising maps that identify are as of the city with good and poor park provision based on different variables.

PAGE 414

400 Analysis of Park Proximity The analysis of access to parks as park proximity involved study ing census blocks' distances to the closest park and to the closest park or parkway. Park proximity expresses the simplest aspect of park provision, which describes the minimum distance that people living in different parts of a city have to cover to reach the closest park (Talen, 2003) In this analysis any park and parkway have the same value, regardless of their size and quality Although park proximity is a simple measure, it has uncover ed which parts of Denver are beyond a quarter mile of any park, or of any park or parkway, thus helping identify areas that need any new green space. Also, park proximity is an important variable for young people's access to urban nature, given their limited spatial mobility (Fagerholm & Broberg, 2011; Loebach & Gilliland, 2014) For this analysis, I conducted three types of statistical tests. I ran a logistic regression to test whether a series of demographic variables could predic t the log odds of a census block being within a quarter mile of a park, and of a park or parkway. Also, I ran non parametric median comparisons and mean rank comparisons to assess differences in demographic variables between census blocks within and beyond a quarter mile of any park, and of any park or parkway. Finally, I conducted non parametric median comparisons and mean rank comparisons to assess differences in census blocks' distances to parks, and to parks or parkways, between income and ethnic g roups (see Chapter 3 ). Park proximity: all parks and parkways The analysis of park proximity including parks and parkways combined showed patterns of partial distributional inequity. In particular, high income groups benefit from parks and parkways more than m id and

PAGE 415

401 low income groups, while African Americans and Non Hispanic Whites have better access to parks and parkways in terms of park proximity than Hispanics. Before running logistic regressions, I conducted multicollinearity tests to check whether high co rrelations occurred among predictor variables (Leech & Onwuegbuzie, 2011) In particular, I calculated the collinear ity tolerance and the VIF coefficient in IBM SPSS 22 between the predictor variables, including: adjusted median household income, percentage of Non Hispanic Whites, percentage of Hispanics, percentage of African Americans, percentage of people under 18, p ercentage of owner occupied residential units, percentage of vacant residential units. The collinearity coefficients showed that multicollinearity existed between percentage of Non Hispanic Whites (tolerance = .067, VIF = 15.023) and percentage of Hispanic s (tolerance = .084, VIF = 11.973). Tolerances lower than 0.20 and 0.10, and VIF higher than 5 or 10 can be considered indicators of multicollinearity issues (O'Brien, 2007) The multicollinearity between these two variables is evident, as Non Hispanic Whites and Hispanics are the two major ethnic groups in Denver, thus an increase in the percentage of one is most likely related to a decrease in the percentage of the o ther. Given this multicollinearity problem, I excluded the percentage of Non Hispanic Whites from all logistic regressions. Also, while I conducted hierarchical logistic regression (Leech et al., 2014) for all variables describing park proximity and park quality, the inclusion of census blocks' acreage and population density in the regression model did not change substantially s tatistical significance levels, logistic regression coefficients, and pseudo R squares. Therefore, when presenting the results I omit the coefficient for census blocks' acreage and population density.

PAGE 416

402 Logistic regression. I conducted a logistic regression to test whether the six predictor variables, adjusted median income, percentage of Hispanics, percentage of African Americans, percentage of people under 18, percentage of owner occupied housing, and percentage of vacant housing units, significantly predic ted whether or not a census block was within a quarter mile from any park or parkway. Assumptions for logistic regression were tested and met. When all six predictor variables are included in the model, they can significantly predict whether or not a censu s block is within a quarter mile from any park or parkway, 2 = 172.21, df = 6, N = 8094, p < 0.001. Table 28 shows that the odds of having a park or parkway within a quarter mile are increasingly larger as median household income, percentage of African Am erican, and percentage of people under 18 increase, and as percentage of owner occupied housing decreases. Table 28 Logistic Regression Results for Park Proximity, All Parks and Parkways B SE Odds ratio p Adjusted median household income .007 .001 1.007 .000 Percentage of Hispanics .001 .001 .999 .214 Percentage of African Americans .010 .002 1.011 .000 Percentage of people under 18 .007 .003 1.007 .005 Percentage of owner occupied housing units .007 .001 .993 .000 Percentage of vacant housing uni ts .006 .004 1.006 .090

PAGE 417

403 The odds of a census block having a park or parkway within a quarter mile grow by 1.007 for each increase of $1, 000 in median household income; they grow by 1.011 for each increase of one percent in p ercentage of African Americans ; and grow by 1.007 for each increase of one percent in percentage of people under 18. T he predictive power of this logistic regression model is fairly low, as expressed by low pseudo R square values (Cox & Snell R Square = .021, Nagelkerke R Square = .03) Differences in demographics between access no access Independent sample non parametric median tests showed few statistically significant differences in demographic variables between census blocks with access and with no access to any park or parkway (s ee Table 29) Census blocks with access to any park or parkway have a higher median household income ($55,167) than census blocks without access ($53,333), although this difference is not statistically significant (p = 0.059). Census blocks with access hav e a significantly lower median percentage of Hispanics (16.28%) than census blocks without access (21.96%), p = .000. Table 29 Medians of Demographic Variables for Census Blocks With Access and No Access to Any Park or Parkway Income % Non Hisp. Whites % Hispanics % African American % under 18 Access $ 55,167 63.41% 16.28%** 1.26%** 21.67%* No access $ 53,333 62.07% 21.96%** 0%** 20.43%* *: Statistically significant at the .01 level. **: Statistically significant at the .001 level.

PAGE 418

404 Differences in dist ances between demographic groups High income census blocks have a significantly lower median distance to any park or parkway (240 meters) than low income census blocks (280 meters) and mid income census blocks (300 meters), p < .01. Although statistically different, these variations are relatively small, thus low income groups are not a position of substantial disadvantage compared to other income groups A ll medians are lower than a quarter mile (400 meters), thus the combination of parks and parkways pro vides good park service in terms of park proximity in Denver (see Figure 31). Table 30 displays differences in median distances among census blocks with different percentages of ethnic minorities. Among the census blocks with the highest ethnic segregation (75 100 percent of one ethnic group), African Americans have the shortest median distance to any park or parkway (143 meters), followed by Non Hispanic Whites and Hispanics (265 meters and 274 meters), p < .001. Differences in median distances between Non Hispanic Whites and Hispanics are not statistically significant. Table 30 Median Distances to the Closest Park or Parkway for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 286 m 300 m 265 m Hispani cs 300 m 275 m 274 m African Americans 250 m* 200 m** 143 m** *: Statistically significant at the .01 level. **: Statistically significant at the .001 level. To summarize, the analysis of park proximity when considering access to any park or parkway show s a partial picture of inequity, as higher incomes were connected to

PAGE 419

405 higher access in two of the three tests I conducted. Also, Hispanics are slightly disadvantage d when considering access to any parks or parkway. African Americans are the ethnic group who appears to be more advantaged in terms of proximity to any park or parkway. This finding partially reflects patterns of park spatial distribution in Baltimore (Boone et al., 2009) and Los Angeles (Wolch et al., 2005) Figure 31 Service area (a quarter mile) for all p arks and parkways The good levels of access to any park or parkway that areas with high concentrations of African Americans experience can be related to the presence of a few parkways ( Martin Luther King Jr. Boulevard, Monaco Parkway, and Colorado Boulevar d) and of several pocket parks in the neighborhoods of northeastern Denver where most African Americans are concentrated (see Figure 32). However, as further

PAGE 420

406 discussed in Chapter 6 (Mixed Research Integration), Martin Luther King Jr. Boulevard and Colorado B oulevard are hardly usable as green space s due to the limited width of their medians. Thus, African American's good geographical access to any park or parkway does not translate in actual provision of enjoyable green space. Figure 32 Northeast Denver 's parks and parkways, and percentages of African Americans Also, high income census groups' good access to all parks and parkway can be partially explained by the high density of parkways that can be found in affluent areas of east central Denver (see Fig ure 33). These parkways were established during the City Beautiful and New Deal era and have contributed to the high status of the areas located around them. Most of these parkways, as explained in Chapter 6, are wide enough to be used for jogging and picn ics and include old growth trees casting significant shade.

PAGE 421

407 Figure 33 East central Denver's parks and parkways, and income groups Overall, studying access to parks and parkways combined was important because, together, they constitute Denver's urban par k system. Both parks and parkways are located on publicly owned land and operated by Denver Parks and Recreation, thus they both represent Denver's public investment in public green spaces in the city. Park proximity: all parks The analysis of park proxim ity focusing on parks showed a pattern of partial equity, although some inequities related to African Americans have emerged. In particular, low income groups have better access to parks in terms of proximity than mid and low income groups, while Hispanic s benefit more from park proximity than Non Hispanic Whites and African Americans. Logistic regression. I conducted a logistic regression to test whether the six predictor variables included in Table 31 significantly predicted whether or not a census

PAGE 422

408 block was within a quarter mile from any park. Assumptions for logistic regression were tested and met. When all six predictor variables are included in the model, they can significantly predict whether or not a census block is within a quarter mile from any pa rk, 2 = 128.18, df = 6, N = 8094, p < 0.001. Table 31 shows the odds ratios, which suggest that the odds of having a park within a quarter mile are increasingly larger as percentage of Hispanics, and percentage of people under 18 increase, and as percenta ge of African Americans and percentage of owner occupied housing decrease. Table 31 Logistic Regression Results for Park Proximity, All Parks B SE Odds ratio p Adjusted median household income .000 .001 1.000 .997 Percentage of Hispanics .003 .001 1.00 3 .002 Percentage of African Americans .004 .002 .996 .013 Percentage of people under 18 .011 .002 1.011 .000 Percentage of owner occupied housing units .005 .001 .995 .000 Percentage of vacant housing units .002 .003 1.002 .514 The odds of a censu s block having a park withi n a quarter mile rise by 1.003 for each increase of one perc ent in percentage of Hispanics; they rise by 1.011 for each increase of one percent i n percentage of people under 18; and they decrease by .996 for each increase of one percent in percentage of African Americans. As for the logistic

PAGE 423

409 regression for access to all parks and parkways, the pseudo R square values are also very low (Cox & Snell R Square = .016, Nagelkerke R Square = .021). Differences in demographics between acc ess no access Independent sample non parametric median tests showed that census blocks with access to all parks have a significantly lower median household income ($52,380) than census blocks without access ($57,604), p < .001. Census blocks with access h ave a significantly lower median percentage of Non Hispanic Whites (58.89%) than census blocks without access (66.67%) p < .001. Census blocks with access have a significantly higher median percentage of Hispanics (21.39%) than census blocks without access (14.91%), p < .001. Other differences are reported in Table 32. Table 32 Medians of Demographic Variables for Census Blocks With Access and No Access to Any Park Income % Non Hisp. Whites % Hispanics % African American % under 18 Access $ 52,380** 58. 89%** 21.39%** 0.29% 22.73%** No access $ 57,604** 66.67%** 14.91%** 1.12% 20%** **: Statistically significant at the .001 level. Differences in distances between demographic groups Low income census blocks have a significantly lower median distance to any park (328 meters) than mid income census blocks (385 meters) and high income census blocks (393 meters), p < .001. Although statistically different, these variations are relatively small, as for proximity to any park or parkways. Also all medians are lower than a quarter mile (400 meters), thus

PAGE 424

410 more than a half of Denver's census blocks is within walking distance from a park (see Figure 34). Table 33 displays differences in median distances among census blocks with different percentages of ethnic minor ities. Among the census blocks with the highest ethnic segregation, Hispanics have the shortest median distance to any park (314 meters), followed by Non Hispanic Whites (391 meters) and African Americans (500 meters), p < .001. Differences in median dista nces for the 50 75 and 25 50 groups are not statistically significant. This shows that, when excluding the areas with the highest concentrations of ethnic groups, the median distance to parks of ethnically mixed census blocks does not vary by their specifi c ethnic composition. Table 33 Median Distances to the Closest Park for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 358 m 400 m 391 m** Hispanics 379 m 381 m 314 m** African Americans 362 m 410 m 500 m** **: Statistically significant at the .001 level. Thus, the analysis of park proximity when considering only parks show ed a partial ly equitable distribution. This finding reflects results of studies conducted in Baltimore (Boone et al., 2009) and Los Angeles (Wolc h et al., 2005) which also found that low income people have good access to parks in terms of park proximity Also, Hispanics are more represented than other ethnic groups in areas within a quarter mile of a park African Americans are partially at disad vantage, as predominantly African

PAGE 425

411 Americans census blocks have a median distan ce to the closest park higher than the quarter mile threshold. Figure 34 Service area (a quarter mile) for all parks The good levels of access to any park that areas with hig h concentrations of Hispanics enjoy can be explained by the presence of several gulch parks and greenways in Denver's West Side, which includes several predominantly Hispanic neighborhoods (see Figure s 34 35). Parks with linear, elongated shapes create lar ger service areas than compact parks with the same surface, as they have a longer perimeter (Rigolon, 2013) In partic ular, Lakewood Gulch, Weir Gulch, and Sanderson Gulch contribute to the West Si de's good access to parks as park proximity. Thus, establishing linear parks along waterways can be a good strategy to increase access to parks in terms of park proximity.

PAGE 426

412 Howev er, linear parks might n ot provide large park acreage or amenities contributing to quality, depending on their width and on their design. Figure 35 West Denver's parks, and percentages of Hispanics To summarize, the analysis of park proximity consideri ng parks and parkways combined and parks alone shows that parkways make a substantial difference in determining which income and ethnic groups live in closer proximity to green spaces. Indeed, while the analysis of proximity focusing on parks only showed a pattern of partial equity, the analysis of proximity including parks and parkways showed a picture of inequity. As explained, parkways are disproportionately located in high income Non Hispanic White and African American areas, and are of significantly h igher quality in Non Hispanic White wealthy areas. The comparison between Figures 31 and 34 shows parkways' important contribution in creating service areas.

PAGE 427

413 In relation to young people's play opportunities in Denver, parks are clearly more important than parkways, as the latter do not include play amenities and are generally valued as visual amenities. Thus, the pattern of inequity emerged in the analysis of proximity including parks and parkways is not particularly concerning for young people's play. Als o, the results of these analyses showed other positive signs for children and teenagers, as areas within a quarter mile of parks, and of parks and parkways, have significantly higher percentages of people under 18 than areas beyond a q uarter mile. However, the low level of access that predominantly African American census blocks experience in terms of proximity to parks raises park equity concerns. Analysis of Park Acreage The analysis of park acreage involved studying acreage of parks and parkways, and of parks, within a quarter mile of census blocks, including measures of potential park congestion. These measure were obtained by dividing the acreage within a quarter mile by 1,000 residents and by 1,000 residents under 18, similarly to several other studies (Boone et al., 2009; Sister et al., 2009; Wolch et al., 2005) T hese measures can help illuminate which areas of the city ha ve overcrowded parks, due to a combination of lack of park acreage and high population densities. Thus, park acreage variables including measures of possible park congestion, are important for young people's access to urban nature. For this analysis, I co nducted non parametric median comparisons and mean rank comparisons to assess differences between income and ethnic groups in the following variables: census block's acreage of parks and parkways within a quarter mile; acreage of parks and parkways within a quarter mile per 1,000 residents; acreage of parks and

PAGE 428

414 parkways within a quarter mile per 1,000 residents under 18; and the three same variables considering parks only (see Chapter 3). Acreage of parks and parkways Differences in acreage of park s and pa rkway s by income and ethnic groups show similarities to the differences that emerged in the analysis of park proximity for parks and parkways. High income groups benefit from park and parkway acreage more than mid and low income groups for all three varia bles (acreage, acreage by 1,000 residents, and acreage by 1,000 residents under 18). Also, African Americans and Non Hispanic Whites have higher acreage of parks and parkways than Hispanics for all three variables. Thus, inequities in park and park and par kway acreage exist, as low income groups and Hispanics have access to significant lower acreage than mid and high income groups and Non Hispanic Whites. These findings reflect general trends in the literature on access to parks (see literature review in C hapter 2). Acres of parks and parkways Independent sample non parametric median tests showed that high income census blocks have a significantly higher median acreage of parks and parkways within a quarter mile (42.95 acres) than mid income census blocks (29.05) and low income census blocks (28.79), p < .001. The medians of mid income and low income census blocks do not differ significantly. Also, census blocks with high concentrations of African Americans have a significantly higher acreage of parks and parkways (57.98) than predominantly Non Hispanic White (39.24) and Hispanic census blocks (19.45), p < .001 (see Table 34). Also, the median acreage of parks and parkways increase as percentages of African American and Non Hisp anic White population increas e, with statistically significant differences, p < 0.01.

PAGE 429

415 Table 34 Median Acreage of Parks and Parkways for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 29.09 31.59 39.24** Hispanics 27.31 29.24 19. 45** African Americans 29.26 37.20** 57.98** **: Statistically significant at the .001 level. The Z coefficients calculated from Mann Whitney U tests allow estimating some of the effect sizes of the differences. For differences of park and parkway acreag e between low income and high income groups, r = 0.12, which according to Cohen (1988) is a small effect size. Differ ences between census blocks with 25 50 percent and 75 100 percent of Non Hispanic Whites yielded an even smaller effect size, r = 0.06. Figure 36 displays how the acreage of parks and parkways varies geographically in Denver. East central Denver, south cen tral Denver, parts of west Denver, and southeast Denver have the highest concentrations of census blocks with access to substantial acreage of parks and parkways. Large parks like City Park, Washington Park, Sloan's Lake, and the park established in Staple ton and Lowry strongly contribute to high park and parkway acreage. Neighborhoods with several parkways also have high acreage.

PAGE 430

416 Figure 36. Spatial distribution of acres of parks and parkways Acres of parks and parkways per 1,000 people This variable is a more significant parameter than the previous one because it takes into account the size of the population that is acres of parks and parkways serve Independent sample non parametric median tests showed that high income census blocks have a significantly higher median acreage of parks and parkways within a quarter mile per 1,000 residents (17.11) than mid income census blocks (10.17) and low income census blocks (9.21), p < .001. The medians of mid income and low income census blocks do not differ signifi cantly. Also, census blocks with high concentrations of African Americans have a significantly higher acreage of parks and parkways per 1,000 residents (28.80) than predominantly Non Hispanic White (14.03) and Hispanic census blocks (6.56), p < .001 (see T able 35). However, as discussed, parkways located in African Americans neighborhoods have very

PAGE 431

417 low quality and cannot be used for active or passive recreation. Census blocks with the highest percentages of Non Hispanic Whites have a significantly higher ac reage of parks and parkways per 1,000 residents than census blocks with lower percentages (25 50 and 50 75), p < .001. The significantly lower median acreage per 1,000 residents in predominantly Hispanic census blocks (6.56) raises equity concerns. Effect sizes of differences, calculated through the Z coefficients of the Mann Whitney U Test are also low, based on Cohen (1 988) For differences of park and parkway acreage per 1,000 residents between low income and high income groups, r = 0.15. Table 35 Median Acreage of Parks and Parkways per 1,000 Residents for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 11.85 11.04 14.03** Hispanics 9.59 10.63 6.56** African Americans 10.53 14.07** 22.80** **: Statistically significant at the .001 level. Figure 37 shows geographic differences in acreage of parks and parkways pe r 1,000 residents in Denver. M ore than half of Denver's census blocks have access to less than 6.78 acres of parks and parkways than 1,000 residents (see Figure 37). Areas in east central Denver and in southwest Denver have particularly high values of acre age per 1,000 residents. The areas with the lowest values include portions of west and northwest Denver as well as the neighborhoods around downtown (see Figure 37).

PAGE 432

418 Figure 37. Spatial distribution of acres of parks and parkways per 1,000 residents Acres of parks and parkways per 1,000 youth This variable incorporates the number of people under 18 in the analysis, thus it is very significant for young people's access to urban nature. Independent sample non parametric median tests showed that high income census blocks have a significantly higher median acreage of parks and parkways within a quarter mile per 1,000 youth (87.09) than mid income census blocks (46.77) and low income census blocks (45.01), p < .001. The medians of mid income and low income grou ps do not differ significantly. Census blocks with high concentrations of Non Hispanic Whites have a significantly higher acreage of parks and parkways per 1,000 youth (93.37) than predominantly African American (76.90) and Hispanic census blocks (21.13), p < .001 (see Table 36). Thus, when considering acreage per 1,000 youth, Non Hispanic Whites have higher median values than African Americans, possibly

PAGE 433

419 because more people under 18 live in areas with high concentrations of African Americans. Table 36 also shows that the median acreage of parks and parkways per 1,000 youth increases as percentages of Non Hispanic White and Afric an American population increase, with statistically significant differences p < .001. The substantially lower median acreage per 1, 000 youth in highly Hispanic census blocks (21.13) shows that Hispanic young people living in these areas experience more crowded parks and parkways than children and teenagers in other areas of the city Table 36 Median Acreage of Parks and Parkways per 1 ,000 Residents Under 18 for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 47.31** 61.96** 93.37** Hispanics 38.39 34.33** 21.13** African Americans 36.11 53.32** 76.90** **: Statistically signific ant at the .001 level. Effect sizes of differences, calculated through the Z coefficients of the Mann Whitney U Test are slightly higher than the ones calculated for the previous variables. For differences of park and parkway acreage per 1,000 youth betwee n low income and high income groups, r = 0.16, a small effect size (J. Cohen, 1988) Differences between census blocks with 75 100 percent of Non Hispanic Whites and with 75 100 percent of Hispanic s yielded a medium effect size, as r = 0.41. This substantially higher effect size raises park equity concerns.

PAGE 434

420 Figure 38. Spatial distribution of acres of parks and parkways per 1,000 youth Figure 38 shows geog raphic variations in acreage of parks and parkways per 1,000 youth in Denver. The eastern side of Denver has higher values of acreage per 1,000 youth than the western side, with a few exceptions in northeast Denver. Also, some areas around downtown have hi gh values of acreage per 1,000 youth because young professionals without children mostl y inhabit these areas. The neighborhoods with the lowest values include portions of west and northwest Denver as well as near northeast Denver (see Figure 38). Acreage o f parks The analysis focusing on acreage of parks yielded different findings than the analysis of park and parkway acreage combined. Also, the findings on park acreage highlight interesting variations between the groups with the highest medians

PAGE 435

421 values and the groups with the highest mean values for all three variables (acres of parks, acres of parks per 1,000 residents, and acres of parks per 1,000 youth). In particular, low income groups and areas with high concentrations of Hispanics have the highest med ian values for acres of parks and acres of parks per 1,000 residents, while high income groups and predominantly Non Hispanic White census blocks have highest mean values for all three variables. As mentioned, I could test whether these differences were st atistically significant only for medians and not for means, as these variables were not normally distributed. The variations in the sign of the differences between medians and means signify that low income and Hispanic areas include more census block group s with average, fairly good park acreage values. On the other hand, census blocks with high income households and high concentrations of Non Hispanic Whites have more variations in the park acreage values, including more zero values but also a larger numbe r of very high values, and a higher mathematical average. In the next sections, I further discuss these variations by looking at the frequency distributions of park acreage variables for different income and ethnic groups. Given these complex variations, it is more complicated to evaluate patterns of equity and inequity when studying ethnic and income differences in acreage of parks. However, regardless of whether medians or means are considered, African Americans are disadvantaged compared to Hispanics an d Non Hispanic Whites. Acres of parks Low income groups have the highest median acreage of parks but high income groups have the highest mean acreage. Also, Hispanics have the highest median acreage of parks but Non Hispanic Whites have the highest mean a creage.

PAGE 436

422 Independent sample non parametric median tests showed that low income census blocks have a significantly higher median acreage of parks within a quarter mile (10.89 acres) than high income census blocks (6.90 acres), p < .001. Mid income census blo cks also have a significantly higher median acreage of parks within a quarter mile (9.98 acres) than high income census blocks (6.90 acres), p < .001. The medians of low income and mid income census groups do not differ significantly. Table 37 shows that f or mean values the sign of the difference is reversed, with high income groups having the highest mean values. Table 37 Median and Mean Acreage of Parks by Income Group Low income Mid income High income Median Acres 10.89 9.98 6.90** Mean Acres 28.9 31 .3 38.03 **: Statistically significant at the .001 level. The frequency distributions of park acreage for low income and high income census blocks can help illuminate the reasons of the discrepancies between differences in medians and means (see Figure 39 ). The distribution for the high income group includes more cases with zero values than the distri bution for the low income group However, the first is much flatter and also includes more large values than the second (see Figure 39). This is reinforced by the standard deviation of the two distributions, which is 49.36 for the low income group and 67.37 for the high income group. In terms of park planning, these differences mean that the low income group has less variation in park acreage, while the high in come group has more zero values but also more high values of park

PAGE 437

423 acreage (above 100, for example, see Figure 39). The frequency distribution of the mid income group is similar to the one of the low income group. Figure 39 Frequency distributions of acr es of parks for low income and high income When evaluating park equity, these divergences between differences in medians and means show that low income people have better "average" park provision in terms park acreage, while high income census blocks inclu de more cases without acreage and more cases with very high acreage. Also, the high income census blocks' higher mean values for acres of parks shows that, overall, more park acreage is within walking distance from high income census blocks than from low i ncome census blocks. In terms of ethnicity, census blocks with the highest concentration of Hispanics have a significantly higher median value for acres of parks within a quarter mile (12.21 acres) than predominantly Non Hispanic White (6.95 acres) and Afr ican American census blocks (3.95 acres), p < .001. The difference in the median values between census

PAGE 438

424 blocks with the highest concentrations of Non Hispanic Whites and African Americans is also statistically significant, p < .001. However, Table 38 shows that predominantly Non Hispanic White census blocks have the highest mean values of acres of parks (35.60) among census blocks with high concentrations of ethnic groups. These discrepancies can also be explained by different frequency distributions, as for income. Effect sizes calculated for differences in mean ranks were very small. Table 38 Median and Mean Acreage of Parks for Census Blocks with Different Percentages of Ethnic Groups Medians 25 50 % 50 75% 75 100% Non Hispanic Whites 11.73 7.01 6.95** H ispanics 10.20 13.76** 12.21** African Americans 10.03 8.11 3.95** Means 25 50 % 50 75% 75 100% Non Hispanic Whites 32.67 36.21 35.60 Hispanics 30.11 32.45 24.57 African Americans 25.69 26.72 9.06 **: Statistically significant at the .001 level. Figu re 40 shows how park acreage varies geographically in the city. Severa l areas of Denver have access to zero acres of parks, and these areas are geographically distributed quite evenly. Zones with high values of park acreage are concentrated around the city 's largest parks, but also in some neighborhoods located in far southeast, far southwest, far west, far east, and far northeast Denver.

PAGE 439

425 Figure 40 Spatial distribution of acres of parks Acres of parks per 1,000 residents Low income groups have the high est median acreage of parks within a quarter mile per 1,000 residents but high income groups have the highest mean acreage per 1,000 residents. Hispanics have the highest median acreage of parks within a quarter mile per 1,000 residents but African America ns have the highest mean acreage per 1,000 residents. In particular, i ndependent sample non parametric median tests showed that low income census blocks have a significantly higher median acreage of parks within a quarter mile per 1,000 residents (3.46 acr es) than high income census blocks (2.84 acres), p < .05, and than mid income census groups (3.12), p < .05. Mid income census blocks also have a significantly higher median value than high income census blocks, p < .05.

PAGE 440

426 Table 39 shows that for mean value s the sign of the difference is completely reversed, with high income groups having the highest means with substantial numerical differences. This variation can be explained, as for acres of parks, by differences in frequency distributions. Similarly, the low income group includes more average cases of park acres per 1,000 residents, while the high income group has more cases of zero park acres per 1,000 residents but also more very high values. Table 39 Median and Mean Acreage of Parks Per 1,000 Residents by Income Group Low income Mid income High income Median Acres 3.46* 3.12* 2.84* Mean Acres 10.65 13.15 19.01 *: Statistically significant at the .05 level. When studying differences based on ethnicity, census blocks with the highest concentration of Hispanics have a significantly higher median value for acres of parks within a quarter mile per 1,000 residents (4.11 acres) than predominantly Non Hispanic White (2.61 acres) and African American census blocks (1.15 acres), p < .001. The difference in th e median values between census blocks with the highest concentrations of Non Hispanic Whites and African Americans is also statistically significant, p < .001 (see Table 40). The very low median value for census blocks with the largest percentage of Africa n Americans raises equity concerns, as 50 percent of these blocks have less than 1.15 acres of park acreage per 1,000 residents, which is about one ninth of the national standard of 10 acres per 1,000 residents (Mertes & Hall, 1996)

PAGE 441

427 Table 40 Median and Mean Acreage of Parks per 1,000 Residents for Census Blocks with Different Percentages of Ethnic Groups Medians 25 50 % 50 75% 75 100% Non Hispanic Whites 4.05** 2.27 2.61** Hispanics 3.37 4.90** 4.11** African Americans 3.17 2.65 1.15** Means 25 50 % 50 75 % 75 100% Non Hispanic Whites 13.99 16.72 16.44 Hispanics 13.77 11.51 8.64 African Americans 10.03 14.04 18.31 **: Statistically significant at the .001 level. However, Table 40 shows that predominantly African American census blocks have the highest m ean values of acres of parks per 1,000 residents (18.31) among census blocks with high concentrations of ethnic groups. Highly Non Hispanic Whites census blocks also include a high mean value (16.44), while highly Hispanic census blocks have the lowest mea n value (8.64). These discrepancies can also be explained by different frequency distributions, as for income. Thus, highly Hispanic census blocks have better average provision of park acres per 1,000 residents, while highly African American and highly Non Hispanic White census blocks include a larger number of cases with zero values and with very high values. The Z coefficients of the Mann Whitney U Test yielded very small effect sizes for mean ranks.

PAGE 442

428 Figure 41 Spatial distribution of acres of parks per 1,000 residents Figure 41 represents the geographical variation of park acreage per 1,000 residents in Denver, which has some similarities to the patterns displayed in Figure 40. Several areas of the city have access to zero acres or very few acres of pa rks per 1,000 residents, and these areas are geographically distributed quite evenly. Neighborhoods with high values of park acreage per 1,000 residents can be found near the city's largest parks, but also in areas located in east Denver, including Staplet on and Lowry, in far southeast, far southwest, far west, and far northeast Denver. Overall, when excluding the three main large historical parks, City Park, Cheesman Park, and Washington Park, the central city has notably less park acreage per 1,000 people than most areas located near the city boundary (see Figure 41). This pattern has emerged in several other studies on access to parks (e.g. Boone et al., 2009; Dai, 2011; Sister et al., 2009; Wolch et al., 2005)

PAGE 443

429 Acres of parks per 1,000 youth The analysis of how acres of parks vary by income groups showed no statistica lly significant differences among medians, but high income groups have the largest mean acreage of parks per 1,000 people under 18. Also, Non Hispanic Whites have the highest median acreage of parks per 1,000 people under 18, followed by Hispanics and Afri can Americans. Table 41 reports variations between income groups of median and mean values of park acreage within a quarter mile per 1,000 youth. While median values have very small variations there are substantial differences in mean values, with the hig h income census blocks having almost twice the value than low income census blocks (although no statistical test was conducted; see Table 41). The differences between the findings for medians and means can once again be explained by variations in f requency distributions between low income and high income groups. Table 41 Median and Mean Acreage of Parks Per 1,000 Residents Under 18 by Income Group Low income Mid income High income Median acres per 1000 youth 15.37 15.69 15.70 Mean acres per 1000 youth 56 .04 74.11 101.04 In terms of ethnicity, census blocks with the highest concentration of Non Hispanic Whites have a significantly higher median value for acres of parks within a quarter mile per 1,000 youth (18.26) than predominantly Hispanic (14.04) and African American census blocks (3.88 acres), p < .001 (see Table 42). The difference in the

PAGE 444

430 median values between census blocks with the highest concentrations of Hispanics and African Americans is also statistically significant, p < .001. The effect size for the difference in mean ranks between predominantly Non Hispanic White areas and predominantly Hispanic areas, calculated through the Z coefficient of the Mann Whitney U Test is small, r = 0.08. However, the difference between highly Non Hispanic White and African American census blocks yielded a medium effect size, r = 0.42. This shows that young people living in the areas with the highest conc entrations of African Americans experience park overcrowding issues. Table 42 Median and Mean Acreage of Parks per 1,000 Residents Under 18 for Census Blocks with Different Percentages of Ethnic Groups Medians 25 50 % 50 75% 75 100% Non Hispanic Whites 17.58** 15.31** 18.26** Hispanics 14.11** 17.08** 14.04** African Americans 11.93** 9.70** 3.88** Means 25 50 % 50 75% 75 100% Non Hispanic Whites 65.88 97.69 101.11 Hispanics 68.49 43.32 29.14 African Americans 43.77 70.83 10.61 **: Statistically significant at the .001 level. Table 42 also shows that, for park acreage per 1,000 youth, the pattern of differen ce between median and mean value does not occur for ethnicity. Indeed, predominantly Non Hispanic White census blocks also have the highest mean value of

PAGE 445

431 park acres within a quarter mile per 1,000 youth, with very substantial numerical differences compared to census blocks with the highest concentrations of ethnic minorities. This can derive from significantly more people under 18 living in ethnic minority areas. The differences in means reported in Table 42, although not calculated through statistical test s, raise serious equity concerns for young people living in areas including more than 75 percent of African Americans or Hispanics. The geographic distribution of park acres per 1,000 people under 18 is represented in Figure 42. Some areas of central, east central and near northwest Denver that have very low park acreage per 1,000 residents show higher values for park acreage per 1,000 youth, as they include fairly low percentages of people under 18. These areas are predominantly Non Hispanic White and high income, which help s e xplain the different results for median acreage per 1,000 residents and for median acreage per 1,000 youth. The pattern of difference between the central city and the areas near the city limits that emerged in the analysis of park acr eage per 1,000 people (see Figure 41), can also be found, but to a lesser degree, for park acreage per 1,000 youth (see Figure 42).

PAGE 446

432 Figure 42 Spatial distribution of acres of parks per 1,000 youth To summarize, the analysis of park acreage when consider ing parks only showed patterns of partial equity if focusing on medians but signs of substantial inequity if focusing on means. In particular, median values advantage Hispanics and low income people for acres of parks and acres of parks per 1,000 residents while they advantage Non Hispanic Whites for acres of parks per 1,000 residents under 18. Mean values are substantially higher for high income groups and Non Hispanic Whites than for low income groups and Hispanics for all three park acreage variables. A frican Americans are the most disadvantaged ethnic group when looking both at median and mean values. The higher median values for low income groups and Hispanics, compared to high income groups and Non Hispanic Whites, are a good sign in terms of park eq uity

PAGE 447

433 because fewer low income and predominantly Hispanic census blocks have access to zero acres of parks than other groups However, more Non Hispanic White and high income census blocks enjoy high amounts of park acreage. In particular, when excluding th e zero values, high income census blocks have a significantly higher median acreage of parks within a quarter mile per 1,000 people (9.85 acres) than low incom e census blocks (5.41 acres), p < .001. Also, a larger number of high income census blocks (925) have access to more 10 acres of parks per 1,000 residents (the national standard) than low income census blocks (783). Finally, when considering census blocks with very high park acreage per 1,000 residents, substantially more high income census blocks (13 0) enjoy 100 or more acres of parks per 1,000 residents than low income census blocks (41). Thus, while patterns of inequity were clear from the analysis of differences among medians when studying acreage of parks and parkways, a more complex picture of in equity emerged from the analysis of acreage of parks. The extensive review of the literature presented in Chapter 2 of this dissertation highlighted that, in most U.S. cities, ethnic minorities and low income groups are disadvantaged in terms of park acrea ge, including park acreage per 1,000 residents and per 1,000 youth. My findings for Denver generally confirm the results of most other studies, especially when parks and parkways are considered together. Analysis of Park Quality The analysis of park quali ty involved studying census blocks' proximity to parks that, based on the PQIY and other parameters of quality, are ranked in the highest brackets of quality scores. This analysis involved selecting some parks based on some of their features and then study ing access as proximity for the selected parks only. The study

PAGE 448

434 of park quality is a key component of this dissertation because it incorporates aspects of parks that matter to young people's park visitation through the Park Quality Index for Youth (PQIY). T his analysis is also important because it takes into account violent crime density in and around parks, which can also significantly influence young people's park visitation (G—mez et al., 2004; Platt, 2012; Spilsbury, 2005; Stodolska et al., 2013) The specific park quality variables I fo cused on, as well as the main findings, are summarized in Table 27. In terms of park planning and park management, the findings on park quality can mainly help understand which areas of the city do not have access to high quality parks for young people's p lay and to parks with low levels of violent crime density. These findings can help drive future investment that aims specifically to increase young people's opportunities for safe contact with nature in Denver's parks. For this analysis, I conducted the sa me three types of statistical tests I ran to study park proximity. I conducted logistic regressions to test whether a series of demographic variables could predict the log odds of a census block being within a quarter mile from parks with different quality parameters. Also, I ran non parametric median comparisons and mean rank comparisons to assess differences in demographic variables between census blocks within and beyond a quarter mile from parks with various quality parameters. Finally, I conducted non parametric median comparisons and mean rank comparisons to assess differences in census blocks' distances to parks with different parameters of quality, between income and ethnic groups (see Chapter 3 for details). High quality PQIY parks The analysis of park quality including parks in the top quartile of the PQIY, hence referred to as high quality PQIY parks, showed a

PAGE 449

435 partially equitable distribution L ow income groups and Hispanics benefit more from parks with "good enough" quality than other demographic groups I nequity issue s arise when considering census blocks with the highest concentrations of African Americ ans. However, differences among income and ethnic groups are relatively small. Logistic regression. The six predictor variables included in Table 43 can significantly predict whether or not a census block is within a quarter mile from a high quality PQIY park, 2 = 136.81, df = 6, N = 8094, p < 0.001. Table 43 shows the odds ratios, which suggest that the odds of having a high quality PQIY park wit hin a quarter mile are increasingly larger as percentage of Hispanics and percentage of people under 18 increase, and as percentage of African Americans decrease. The predictive power of this model is very low (Cox & Snell R Square = .017, Nagelkerke R Squ are = .024). Table 43 Logistic Regression Results for High Quality PQIY Parks B SE Odds ratio p Adjusted median household income .000 .001 1.000 .596 Percentage of Hispanics .006 .001 1.006 .000 Percentage of African Americans .007 .002 .993 .000 Pe rcentage of people under 18 .010 .002 1.010 .000 Percentage of owner occupied housing units .001 .001 .999 .324 Percentage of vacant housing units .002 .003 .998 .602

PAGE 450

436 Differences in demographics between access no access Independent sample non parame tric median tests showed statistically significant differences for all demographic variables included in Table 44. Census blocks with access to high quality PQIY parks have a lower median household income ($51,558) than census blocks without access ($56,23 8), p < .001. Census blocks with access have a lower median percentage of Non Hispanic Whites (58.89%) than census blocks without access (66.67%), p < .001. Census blocks with access have a higher median percentage of Hispanics (26.42%) than census blocks without access (15.6%), p < .001. Other differences are reported in Table 44. The effect sizes between mean ranks calculated through the Mann Whitney U Test are all low based on Cohen (1988) with the highest value for differences in percentage of Hispanics, r = 0.10. Table 44 Medians of Demographic Variables for Census Blocks With Access and No Access to High Quality PQ IY Parks Income % Non Hisp. Whites % Hispanics % African American % under 18 Access $ 51,558** 54.35%** 26.42**% 0*% 23.53%** No access $ 56,238** 66.67%** 15.60**% 1.18*% 20.63%** *: Statistically significant at the .01 level. **: Statistically signi ficant at the .001 level. Differences in distances between demographic groups Confirming the results of the previous two tests, low income census blocks have a significantly lower median distance to high quality PQIY parks (600 meters) than mid income cen sus blocks (666 meters), p < .01, and high income census blocks (764 meters), p < .001. The difference in

PAGE 451

437 median distance between mid and high income census blocks is also statistically significant, p < .001. Although statistically different, the variatio n between low and mid income groups is fairly small, thus low income groups are not in position of substantial advantage compared to mid income groups (r = 0.04). Table 45 displays differences in median distances among census blocks with different percen tages of ethnic minorities. Among the census blocks with the highest ethnic segregation (75 100 percent of one ethnic group), Hispanics have the shortest median distance to the closest high quality PQIY park (525 meters), followed by African Americans and Non Hispanic Whites (739 meters and 740 meters), p < .001. All effect sizes calculated through the Mann Whi tney U Test are small based on Cohen's (1988) classification. Table 45 Median Distances to the Closest High Quality PQIY Park for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 645 m 723 m 740 m Hispanics 68 7 m 546 m 525 m** African Americans 764 m 698 m 739 m *: Statistically significant at the .01 level. **: Statistically significant at the .001 level. This analysis shows that all medians distances to high quality PQIY parks are higher than a quarter mil e (400 meters), which highlights that parks with "good enough" quality are rarely within walking distance of young people, regardless of their household income or ethnicity. Figure 43 shows the quarter mile service areas created by high

PAGE 452

438 quality PQIY parks in Denver. The relatively good provision that highly Hispanic areas enjoy can be explained by the presence of several high quality PQI Y parks in Denver's West Side, which has significant concentrations of Hispanics (see Figure 43). As previously mentioned, these areas include several gulch parks that, as they comprise several sport facilities, playgrounds and natural elements, were classified in the highest quartile of the PQIY. Also, many high income Non Hispanic White areas lo cated in east central and sou th central Den ver are outside the service areas of high quality PQIY parks, which contributes to high income and highly Non Hispanic White census blocks' relatively low access to this type of parks (see Figure 43). Figure 43 Service area (a quarter mile ) for high quality PQIY parks

PAGE 453

439 Parks with high quality PQIY play The analysis of park quality including parks in the top quartile of the structured play diversity category of the PQIY, hence referred to as parks high quality PQIY play, also highlighted a p icture of partial equity. Low income groups and Hispanics are in a position of advantage with regards to parks with high quality PQIY play over high income groups, African Americans, and Non Hispanic Whites. Inequity concerns only derive from predominantly African American areas' low provision of parks with high quality PQIY play. However, as for high quality PQI Y parks, differences among income and ethnic groups are relatively small. Logistic regression The six predictor variables included in Table 46 can significantly predict whether or not a census block is within a quarter mile from a park with high quality PQIY play, 2 = 151.03, df = 6, N = 8094, p < 0.001. Table 46 shows the odds ratios, which suggest that the odds of having a park with high quality PQIY play within a quarter mile are increasingly larger as percentage of Hispanics and percentage of people under 18 increase, and as percentage of African Americans decrease. Interestingly, the odds ratio for percentage of people under 18 is the highest, which is a positive sign in terms of play opportunities for young people. However, low pseudo R square values show the low predictive power of this model (Cox & Snell R Square = .018, Nagelkerke R Square = .026).

PAGE 454

440 Table 46 Logistic Regression Results f or Parks with High Quality PQIY Play B SE Odds ratio p Adjusted median household income .000 .001 1.000 .732 Percentage of Hispanics .006 .001 1.006 .000 Percentage of African Americans .005 .002 .995 .003 Percentage of people under 18 .013 .002 1.01 3 .000 Percentage of owner occupied housing units .000 .001 1.000 .845 Percentage of vacant housing units .001 .003 1.001 .793 Differences in demographics between access no access Independent sample non parametric median tests showed that census block s with access to parks with high quality PQIY play have a significantly lower median household income ($52,425) than census blocks without access ($55,865), p < .001. Census blocks with access have a significantly lower median percentage of Non Hispanic Wh ites (55.08%) than census blocks without access (66.67%), p < .001. Census blocks with access have a higher median percentage of Hispanics (25.36%) than census blocks without access (15.28%), p < .001. Other differences are reported in Table 47. The effe ct sizes between mean ranks calculated all low based on Cohen (1988) with the highest value for differences in perce ntage of Hispanics, r = 0.11.

PAGE 455

441 Table 47 Medians of Demographic Variables for Census Blocks With Access and No Access to Parks with High Quality PQIY Play Income % Non Hisp. Whites % Hispanics % African American % under 18 Access $ 52,425** 55.08%** 25.3 6%** 0%* 23.73%** No access $ 55,856** 66.67%** 15.28%** 1.18%* 20.31%** *: Statistically significant at the .05 level. **: Statistically significant at the .001 level. Differences in distances between demographic groups This analysis substantially con firms the results of the previous two tests, as low income census blocks have a significantly lower median distance to parks with high quality PQIY play (543 meters) than mid income census blocks (576 meters), p < .05, and high income census blocks (650 me ters), p < .001. The difference in median distance between mid and high income census blocks is also statistically significant, p < .001. Although statistically different, the variation between median distances of low and mid income group is fairly small thus low income groups are not in position of substantial advantage compared to mid income groups (r = 0.03). When studying ethnic variations, predo minantly Hispanic census blocks have a significantly lower median distance to the closest park with high quality PQIY play (471 meters), followed by highly African American (652 meters) and highly Non Hispanic White census blocks (657 meters), p < .001 (see Table 48). Also, the median distance does not substantially vary for census blocks with different perce ntages of Hispanics (see

PAGE 456

442 Table 48). All effect sizes calculated through the Mann Whitney tests are small based on Co hen's (1988) classification. Table 48 Median Distances to the Closest Park with High Quality PQIY Play for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 583 m** 601 m** 657 m Hispanics 463 m** 47 6 m** 471 m** African Americans 643 m** 560 m** 652 m **: Statistically significant at the .001 level. Overall, all three groups of tests substantially confirm that low income groups are slightly better served by parks with high quality PQIY play than m id and high income groups, and that Hispanics are at advantage over Non Hispanic Whites and African Americans. Also, the median distances for parks with high quality PQIY play are generally shorter than the ones for high quality PQIY parks, which suggests that parks with high quality PQIY play are distributed more efficiently in Denver, regardless of income and ethnicit y. In terms of park design, these findings mean that amenities like playgrounds, different types of sport facilities and supporting featur es like tables and benches are distributed quite evenly in Denver.

PAGE 457

443 Figure 44 Service area (a quarter mile) for parks with high quality PQIY play Figure 44 shows the geographical distribution of park s with high quality PQIY play, which is relatively sim ilar to the geographical patt erns of high quality PQIY parks. In particular, west Denver is also quite well served and many areas of east central and sou th central Denver have relatively low provisions. Although many parks are included in both categories, there are some notable exceptions For example, Cheesman Park, which is classified as a high quality PQIY park, is not in the upper quartile of the structured play diversity category of the PQIY, as it does not include significant structured play amenities On the other hand, other smaller parks that are intended to serve their neighborhoods and young people living in them include significant play amenities (see differences between Figures 43 and 44).

PAGE 458

444 Parks with high quality PQIY nature The analysis of pa rk quality including parks in the top quartile of the nature category of the PQIY, hence referred to as parks with high quality PQIY nature, also highlighted a picture of partial equity, following the patterns of the previous two types of parks. Low income groups and Hispanics benefit more from these types of parks than high income groups and African Americans and Non Hispanic Whites. However, African Americans have particularly low access to parks with high quality PQIY nature, which raises serious equity concerns African American young people's contact with significant natural elements in parks For measures of access to these parks, differences and effect sizes are slightly larger than for the previous two types of parks. Logistic regression The six pred ictor variables included in the model (see Table 49) can significantly predict whether or not a census block is within a quarter mile from a park with high quality PQIY nature, 2 = 215.77, df = 6, N = 8094, p < 0.001. Table 49 shows the odds ratios, which suggest that the odds of having a park with high quality nature within a quarter mile of a census block are increasingly larger as percent age of Hispanics increase, and as median household income and percentage of African Americans decrease. Percentage of African Americans is the stronger predictor variable. Howeve r, the predictive power of this model, although slightly higher than the ones for the two previous types of parks, is still very low (Cox & Snell R Square = .018, Nagelkerke R Square = .026).

PAGE 459

445 Table 49 Logistic Regression Results for Parks with High Quality PQIY Nature B SE Odds ratio p Adjusted median household income .003 .001 .997 .000 Percentage of Hispanics .006 .001 1.006 .000 Percentage of African Americans .020 .002 .980 .000 Percentage of people under 18 .005 .003 1.005 .074 Percentage of owner occupied housing units .000 .001 1.000 .767 Percentage of vacant housin g units .006 .004 .994 .121 Differences in demographics between access no access Independent sample non parametric median tests showed statistically significant differences for all demographic variables included in Table 50. Census blocks with access h ave a significantly lower median household income ($52,019) than census blocks without access ($55,774), p < .001. Census blocks with access have a lower median percentage of Non Hispanic Whites (56.72%) than census blocks without access (65%), p < .001. C ensus blocks with access have a higher median percentage of Hispanics (26.19%) than census blocks without access (15.96%), p < .001. Other differences are reported in Table 50. The effect sizes between mean ranks calculated through the Z coefficient from t he Mann Whitney U Test are all low based on Cohen (1988) with the highest value for differences in percentage of His panics, r = 0.10.

PAGE 460

446 Table 50 Medians of Demographic Variables for Census Blocks With Access and No Access to P arks with High Quality PQIY Nature Income % Non Hisp. Whites % Hispanics % African American % under 18 Access $ 52,019** 56.72%** 26.19%** 0%** 2 2.58%* No access $ 55,774** 65%** 15.96%** 1.36%** 21.05%* *: Statistically significant at the .01 level. **: Statistically significant at the .001 level. Differences in distances between demographic groups This analysis mostly confirms the results of t he first two sets of tests, but with some differences. Mid income and low income census blocks have significantly lower median distances to the closest park with high quality PQIY nature (670 meters and 693 meters) than high income census blocks (860 meter s), p < .001. Although the median value for mid income groups is lower than the value for low income groups, their difference is not statistically significant. Also, the difference between median distances of low and high income groups, while statisticall y significant, is relatively small, thus low income groups are not in position of substantial advantage compared to high income groups (r = 0.09). Tests evaluating ethnic differences show that African Americans have a substantial disadvantage (see Table 51 ). Predominantly Hispanic census blocks have a significantly lower median distance to the closest park with high quality PQIY nature (577 meters) than highly Non Hispanic White census blocks (788 meters) and highly African American census blocks (1612 mete rs), p < .001. The difference in median values between predominantly Non Hispanic White and predominantly African

PAGE 461

447 American census blocks is also statistically significant, p < .001. Census block s with the highest concentration of African Americans have an extremely high median distance to parks with high quality nature. The effect size for difference in mean ranks between highly Non Hispanic White and highly African American areas is higher than in previous cases, r = 0.22, but still a small effect size acc ording to Cohen's (1988) Table 51 Median Distances to the Closest Park with High Quality PQIY Nature for Census Bl ocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 669 m* 746 m** 788 m** Hispanics 733 m* 581 m** 557 m** African Americans 948 m** 1139 m** 1612 m** **: Statistically significant at the .001 level. *: Statist ically significant at the .01 level. Figure 45 depicts the geographic distribution of parks high quality PQIY nature and their service areas. The median distances to parks with high quality PQIY nature are higher than in the two previous variables of park quality. This shows that parks with high quality PQIY nature are more concentrated in a few areas of the city. West Denver and parts of southeast Denver have the highest densities of parks with high quality PQIY nature (see Figure 45 ) The higher levels o f access that highly Hispanic and low income census blocks enjoy can be linked to the presence of gulch parks in West Denver, which include meaningful natural features like water and tree canopy (see Figure 45). On the other hand, African American areas in northeast Denver include few parks with high quality PQIY nature, which helps explain African Americans' substantial disadvantage

PAGE 462

448 Figure 45 Service area (a quarter mile) for parks with high quality PQIY nature Parks with high quality PQIY play and nat ure The analysis of park quality including parks in the top quartile of both the structured play diversity and nature categories of the PQIY, hence referred to as parks with high quality PQIY play and nature, highlighted a pattern of inequity with a subs tantial disadvantage experienced by African Americans and partially by low income groups. In particular, mid income groups and Hispanics have better access to these types of parks than other demographic groups The very low level of service in areas with h igh concentrations of African Americans raises serious park equity concerns, and builds on other issues emerged in the previous analyses of park quality. Since parks included in this category are less numerous than the ones in previous categories, differen ces and effect sizes are slightly larger.

PAGE 463

449 Logistic regression The six predictor variables included in the model (see Table 52) can significantly predict whether or not a census block is within a quarter mile from a park with high quality PQIY nature, 2 = 215.77, df = 6, N = 8094, p < 0.001. Table 52 shows the odds ratios, which suggest that the odds of having a park with high quality PQIY play and nature within a quarter mile of a census block are increasingly larger as percentage of Hispanics and percent age of people under 18 increase, and as median household income and percentage of African Americans decrease. Percentage of A frican Americans is the strongest predictor variable, as t he odds of having access to a park with high quality PQIY play and nature decrease by .969 for each increase of one percent of African Americans. The predictive power of this model, although still very low, is higher than in previous analyses ( Cox & Snell R Square = .035, Nagelkerke R Square = .055 ) Table 52 Logistic Regressio n Results for Parks with High Quality PQIY Play and Nature B SE Odds ratio p Adjusted median household income .002 .001 .998 .011 Percentage of Hispanics .006 .001 1.006 .000 Percentage of African Americans .031 .003 .969 .000 Percentage of people u nder 18 .010 .003 1.010 .001 Percentage of owner occupied housing units .001 .001 1.001 .258 Percentage of vacant housing units .006 .004 .994 .115

PAGE 464

450 Differences in demographics between access no access Independent sample non parametric median tests yi elded statistically significant differences for all demographic variables included in Table 53 and generally confirm the results of the logistic regression. Census blocks with access have a significantly lower median household income ($52,539) than census blocks without access ($55,619), p < .001. Census blocks with access have a significantly lower median percentage of Non Hispanic Whites (56.67%) than census blocks without access (64.29%), p < .001. Census blocks with access have significantly a higher me dian percentage of Hispanics (29.03%) than census blocks without access (16.33%), p < .001. Other differences are reported in Table 53. The effect sizes between mean ranks are all low based on Cohen (1988) with the highest value for differences in percentage of African Americans, r = 0.12. Table 53 Medians of Demographic Variables for Census Blocks With Access and No Ac cess to Parks with High Quality PQIY Play and Nature Income % Non Hisp. Whites % Hispanics % African American % under 18 Access $52,539** 56.67%** 29.03%** 0%** 23.08** No access $55,619** 64.29%** 16.33%** 1.37%** 20.97** **: Statistically significan t at the .001 level. Differences in distances between demographic groups These differences show a slightly different pattern in terms of income and more alarming findings for areas with high concentrations of African Americans. Mid income census blocks ha ve a significantly lower distance to the closest park with high quality PQIY play and nature

PAGE 465

451 (800 meters) than low income (914 meters) and high income census blocks (1111 meters), p < .001. While it is statistically different, the variation between median distances of low and high income group is relatively small, thus low income groups are not in position of substantial advantage compared to high income groups (r = 0.08) which is a condition of inequity Differences based on ethnicity further reinforce A frican American's disadvantage (see Table 54). Highly Hispanic census blocks have a significantly lower median distance to the closest park with high quality PQIY play and nature (640 meters) than highly Non Hispanic White census blocks (972 meters) and hi ghly African American census blocks (2000 meters), p < .001. The diff erence in medians between highly Non Hispanic White and highly African American census blocks is also statistically significant, p < .001. Census block high densities of African Americans including all percentages above 50 percent, have particularly high median distance to parks with high quality PQIY play and nature. The effect size for differences in mean ranks between highly Non Hispanic White and highly African American areas is small according to Cohen's (1988) r = 0.18. Table 54 Median Distances to the Closest Park with High Quality PQIY Play a nd Nature for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 900 m** 1000 m** 972 m** Hispanics 1051 m** 700 m** 640 m** African Americans 1386 m** 1502 m** 2000 m** **: Statistically significant a t the .001 level.

PAGE 466

452 The geographic distribution of parks high quality PQIY nature and their service areas, displayed in Figure 46, repeats some of the patterns emerged in the analysis of parks with high quality PQIY nature. Similarly, West Denver and some a reas in southeast Denver have the highest concentrations of parks with high quality PQIY play and nature (see Figure 46). Most parks included in the West Side are classified as high quality PQIY play and nature, thus providing good services to Hispanics an d low income groups who are numerous in the area (see Figure 46). Also, areas of northeast Denver with high percentages of African Americans have low provision of parks with high quality PQIY play and nature, which clarifies African American's inequity iss ues. Several areas in east central and south central Denver, which are mostly inhabited by high income Non Hispanic Whites, are outside the service areas of these types of parks. Overall, African Amer ican's very low provision of parks with high quality PQI Y play and nature, the slightly better provision that mid income groups experience compared to low income groups, and Hispanics' good levels of access display a complex picture of inequity and partial equity for this type of parks.

PAGE 467

453 Figure 46 Service are a (a quarter mile) for parks with high quality PQIY play, nature Best PQIY parks The analysis park quality including parks in the top five percent of the PQIY, hence referred to as best PQIY parks, showed a picture of clear inequity. In particular, high i ncome groups and Non Hispanic Whites benefit from best PQIY parks more than other income groups, Hispanics, and African Americans. The 15 parks that are in the top five percent of the PQIY represent the city's excellence in park quality for young people ba sed on the PQIY, thus the low provision that low income groups, Hispanics, and African Americans experience is a significant case of distributional environmental injustice. For this analysis, differences in median values are substantially higher than for p revious parameters of park quality.

PAGE 468

454 Logistic regression The six predictor variables included in the model (see Table 55) can significantly predict whether or not a census block is within a quarter mile from a best PQIY park, 2 = 207.78, df = 6, N = 8094, p < 0.001. Table 55 shows the odds ratios, which suggest that the odds of having a best PQIY park within a quarter mile of a census block are increasingly larger as percentage of people under 18 increases, and as percentage of Hispanics and percentage of African Americans decrease. Percentage of African Americans is the stronger predictor variable, as the odds of having a best PQIY park within a quarter mile decrease by 0.959 for each increase of one percent of African Americans. However, increasing percen tage s of young people are related to higher odds of a census block having access to a best PQIY park. The predictive power of this model is still very low, as Cox & Snell R Square = .025, Nagelkerke R Square = .056. Table 55 Logistic Regression Results fo r Best PQIY Parks B SE Odds ratio p Adjusted median household income .002 .001 1.002 .107 Percentage of Hispanics .013 .002 .987 .000 Percentage of African Americans .042 .005 .959 .000 Percentage of people under 18 .014 .004 1.014 .000 Percentage of owner occupied housing units .000 .002 1.000 .831 Percentage of vacant housing units .004 .005 .996 .504

PAGE 469

455 Differences in demographics between access no access Independent sample non parametric median tests yielded statistically significant differenc es for all demographic variables included in Table 56 and give a clearer picture of inequity than the logistic regression. Census blocks with access to a best PQIY park have a significantly higher median household income ($67,247) than census blocks withou t access ($53,667), p < .001. Census blocks with access have a significantly higher median percentage of Non Hispanic whites (76.15%) than census blocks without access (60.44%), p < .001. Census blocks with access have a significantly lower median percenta ge of Hispanics (12%) than census blocks without access (18.42%), p < .001. Other differences are reported in Table 56. The effect sizes between mean ranks are all low based on Cohen (1988) The highest value was found for differences in percentage of Non Hispanic Whites, r = 0.11. Although these effect sizes are low, the differences in median income and in percentage o f Non Hispanic Whites are notably higher than the ones found in previous analyses of park proximity, park acreage and park quality. Table 56 Medians of Demographic Variables for Census Blocks With Access and No Access to Best PQIY Parks Income % Non Hisp Whites % Hispanics % African American % under 18 Access $ 67,247** 76.15%** 12%** 0%** 19.61%* No access $ 53,667** 60.44%** 18.42%** 1.14%** 21.45%* **: Statistically significant at the .001 level. *: Statistically significant at the .01 level.

PAGE 470

456 Diffe rences in distances between demographic groups The differences in medians confirm the findings of the first to sets of tests. High income census blocks have a significantly lower distance to the closest best PQIY park (1483 meters) than mid income census groups (1595 meters), p < .01, and then low income census blocks (1931 meters), p < .001. The median distances of low and mid income census blocks also differ significantly, p < .001. The effect size for differences in mean ranks between low and high inc ome groups is still low, r = 0.12. Differences based on concentrations of ethnic groups highlight Non Hispanic Whites' position of advantage (see Table 57). Highly Non Hispanic White census blocks have a significantly shorter median distance to the closes t best PQIY park (1471 meters) than predominantly Hispanic census blocks (2022 meters) and predominantly African American census blocks (2500 meters), p < .001. The difference in medians between highly Hispanic and highly African American census blocks is also statistically significant, p < .001. Census blocks with some concentrations of African Americans, including all percentages above 25 percent, have particularly large median distance to best PQIY parks. The effect size for difference s in mean ranks bet ween highly Non Hispanic White and highly African American areas is medium according to Cohen's (1988) r = 0.25.

PAGE 471

457 Table 57 Median Distances to the Closest Best PQIY Park for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 1516 m** 1400 m** 1471 m** Hispanics 1775 m** 1800 m** 2022 m** African Americans 2695 m** 2403 m** 2500 m** **: Statistically significant at the .001 level. Figure 47 displays the geographic location of best PQIY parks, including their quarter mile service areas. Several areas of Denver are completely or partially underserved by parks with excellent" quality including extreme northeast, southwest and parts of the West Side (see Figure 47). In particular, the low levels of access to best PQIY parks that Hispanics and low income groups experience can be related to the partial lack of these ty pes of parks in the West Side. Interestingly, the West Side has relatively good provision of parks with high quality based on other PQIY parameters (see previous analyses). Thus, when analyzing park quality, it is important to take into account the thresho ld that is consider ed as an acceptable standard of quality. In this study, I choose two general thresholds: the upper quartile, or 25 percent, which describes a good enough" standard of quality, and the upper 5 percent, which depicts "excellent" quality, based on my literature analysis of young people's outdoor play.

PAGE 472

458 Figure 47 Service area (a quarter mile) for best PQIY parks To summarize, the analysis of park quality based on the Park Quality Index for Youth (PQIY) showed a picture of partial equity fo r low income groups and Hispanics, as well as substantial inequity issues for African Americans. The only park quality parameter that displayed a complete picture of inequity was access to best PQI parks, with high income Non Hispanic Whites disproportiona tely benefitting from these parks. Thus, different patterns of inequity and partial equity emerged when considering different thresholds in the PQIY, including the top 25 percent and the top 5 percent, with African Americans being significantly disadvantag ed in both situations. The relatively high provision of parks with good quality standards included (top 25 percent) that low income groups and Hispanics enjoy is encouraging in terms of park equity and is partially related

PAGE 473

459 to park planning efforts to creat e recreation opportunities for some park dependent groups (see Chapter 6 ). The higher provision of parks with excellent quality standards (top 5 percent) that high income people and Non Hispanic Whites benefit from raises equity concerns. Indeed, these par ks include Denver's most prestigious and visited parks, like City Park, Washington Park and Cheesman Parks, which include a variety of city renowned amenities and receive most of the city's funding for maintenance (see Chapter 6 ). Finally, the analysis of park quality based on the PQIY also showed that, besides the case of best PQIY parks, areas within a quarter mile of high quality parks have higher concentrations of people under 18 than areas beyond a quarter mile. Although differences are relatively smal l, this is an encouraging sign for young people's access to urban nature. Parks with no LULUs The analysis of park quality including parks without locally unwanted land uses within 100 meters, hence referred to as parks with no LULUs, showed very small si gnificant differences in terms of ethnicity and income. As low income ethnic minority young people do not have significantly better access to parks with no LULUS than other groups, this can be considered a case of distributional inequity In particular, No n Hispanic Whites and Hispanics have slightly better access to parks with no LULUs than African Americans. Also, low income and high income groups have marginally higher access to parks with no LULUs than mid income groups. The logistic regression and the difference in medians between census blocks with access and no access showed no statistically significant results. Differences in distances between demographic groups Independent sample non parametric median tests yielded statistically significant differe nces in terms of income and ethnicity. Low income and high income census blocks have a significantly lower

PAGE 474

460 median distance to the closest parks with no LULUs (both 400 meters) than mid income groups (445 meters), p < .01. Also, census blocks with the highe st concentrations of Hispanics and Non Hispanic Whites have significantly lower median distances to the closest park with no LULUS (408 meters and 411 meters) than highly African American census blocks (530 meters), p < .001. Table 58 shows other small dif ferences among census blocks with different percentages of ethnic groups. Table 58 Median Distances to the Closest Park with No LULUs for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 433 m 454 m* 41 1 m Hispanics 437 m 411 m 408 m African Americans 389 m* 426 m 530 m** **: Statistically significant at the .001 level. *: Statistically significant at the .01 level. Figure 48 depicts the spatial distribution of parks with no LULUs and the quarter mil e service area they generate. Some parks in north and west Denver are in proximity to several locally unwanted land uses like industrial areas, highways and railways. LULUs are predominantly located near lo w income ethnic minority areas, and they contribut e to changing the results of park proximity for all parks Indeed when considering only parks with no LULUs, the advantage that low income groups and Hispanics experienced over high income groups and Non Hispanic Whites in terms of park proximity to all p arks has disappeared Overall, the analysis of parks with no LULUs has shown the smallest

PAGE 475

461 differences among income and ethnic groups among all variables of park proximity, park acreage and park quality. Figure 48 Service area (a quarter mile) for parks with no LULUs Parks with low violent crime density The spatial distribution of parks with low violent crime density shows patterns of striking inequity. High income groups and Non Hispanic Whites have substantially better access to parks with low violent crime density than ethnic minorities and low income groups. Hispanics are the ethnic group with the lowest access to safe parks As previously discussed, the rates of violent crime density in parks are mostly influenced by park management and patrolling, b y their neighborhood's social environment, partially by park design, and to a certain extent by park size. Thus, the inequities described in this section can mostly be linked to park management (see

PAGE 476

462 Chapter 6). Differences and effect sizes for park with lo w crime density are the highest of all analyses of park spatial distribution. Logistic regression The six predictor variables included in the model (see Table 59) can significantly predict whether or not a census block is within a quarter mile from a park with low violent crime density, 2 = 1004.03, df = 6, N = 8094, p < 0.001. Table 59 shows the odds ratios, which suggest that the odds of having a park with low violent crime density within a quarter mile of a census block are increasingly larger as media n household income and percentage of people under 18 increases, and as percentage of Hispanics and percentage of African Americans decrease. Percentage of African Americans and percentage of Hispanics are the strongest predictor variables. The odds of havi ng a park with low violent crime density within a quarter mile decrease by 0.971 for each increase of one percent of Hispanics, and by 0.972 for each increase of one percent of African Americans. The pseudo R squares are the highest that have been found in these analyses, although they are still relatively low, as Cox & Snell R Square = .117, and Nagelkerke R Square = 0.188. Thus, the logistic regression results depict a larger picture of inequity in terms of ethnicity than in terms of income.

PAGE 477

463 Table 59 Lo gistic Regression Results for Parks with Low Violent Crime Density B SE Odds ratio p Adjusted median household income .007 .001 1.007 .000 Percentage of Hispanics .029 .002 .971 .000 Percentage of African Americans .029 .003 .972 .000 Percentage of people under 18 .019 .003 1.019 .000 Percentage of owner occupied housing units .000 .001 1.000 .723 Percentage of vacant housing units .003 .004 .997 .428 Differences in demographics between access no access Independent sample non parametric median tests yielded statistically significant differences for all demographic variables included in Table 60 and depict stronger distributional inequities than the logistic regression. Census blocks with access to parks with low violent crime density have a sign ificantly higher median household income ($79,269) than census blocks without access ($51,307), p < .001. Census blocks with access have a significantly higher median percentage of Non Hispanic whites (82.14%) than census blocks without access (53.33%), p < .001. Census blocks with access have a significantly lower median percentage of Hispanics (7.26%) than census blocks without access (23.29%), p < .001. Other differences are reported in Table 60. The effect sizes between mean ranks are among the highest in the analysis of park spatial distribution and are all medium, based

PAGE 478

464 on Cohen (1988) The highest values were found for differences in median household income, r = 0.27, in percentage of Non Hispanic Whites, r = 0.28, and percentage of Hispanics, r = 0.27. Table 60 Medians of Demographic Variables for Census Blocks With Access and No Access to Parks with Low Violent Cr ime Density Income % Non Hisp. Whites % Hispanics % African American % under 18 Access $ 79,269** 82.14%** 7.26%** 0%** 21.74%** No access $ 51,307** 53.33%** 23.29%** 1.36%** 20%** **: Statistically significant at the .001 level. Differences in dist ances between demographic groups These differences in median distances reinforce the findings of the first to sets of tests. High income census blocks have a significantly lower distance to the closest park with low violent crime density (613 meters) than mid income census groups (1054 meters), p < .001, and then low income census blocks (1562 meters), p < .001. The median distances of low and mid income census blocks also differ significantly, p < .001. Thus, young people living in half of Denver's low i ncome census blocks have to walk almost a mile to reach a park with low violent crime density. The effect size for differences in mean ranks between low and high income groups is medium, per Cohen (1988) r = 0.49. Differences based on concentrations of ethnic groups further reinforce Non Hispanic Whites' good access to parks with low violent crime density (see Table 61 ). Highly Non Hispanic White census blocks have a significantly shorter median distance to

PAGE 479

465 the closest park with low violent crime density (640 meters) than predominantly African American census blocks (1358 meters) and predominantly Hispanic census blocks (2500 meters), p < .001. The difference in median values between highly Hispanic and highly African American census blocks is also statistically significant, p < .001. Census blocks with a Hispanic population higher than 50 percent have particularly large median distance parks with low violent crime density, highlighting serious crime issues in Hispanic neighborhoods. Non Hispanic Whites have a significant position of advantage even in census blocks with ethnic concentrations between 50 and 75 percent (see Table 61). The effect size differences in mean ranks between predominantly Non Hispanic White and Hispanic census blocks is medium based on Cohen (1988) r = 0.48. Table 61 Median Distances to the Closest Park with Low Violent Crime Density for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 1500 m** 861 m** 640 m** Hispanics 1250 m 2177 m** 2500 m** African Americans 1332 m 1422 m** 1358 m** **: Statistically significant at the .001 level. Figure 49 depicts the spatial distribution of parks with low violent crime density in Denver. The West Side, which has high concentrations of low income Hispanics, lacks parks with low violent crime density. This helps explain Hispanics' and low income groups' significant disadvantage in terms of access to these type of parks. Other areas of the city that lack parks with low vi olent crime density are large parts of northeast Denver,

PAGE 480

466 some parts of northwest Denver, and far northeast Denver. Neighborhoods located in east central Denver, south central Denver, and southwest Denver, which are mostly inhabited by high and mid income Non Hispanic Whites, have good access to parks with low violent crime density. Overall, the spatial distribution of parks with low violent crime den sity raises the biggest concern in terms of park equity, due to substantial differences between high income Non Hispanic Whites and low income ethnic minorities. Figure 49 Service area (a quarter mile) for parks with low violent crime density Regional parks The geographic distribution of regional parks, which are large parks that Denver Parks and Recreation considers as the most important because they serve the city as a whole, shows some patterns of inequity. High income groups and Non Hispanic Whites benefit from regional parks more than ethnic minorities and low income

PAGE 481

467 groups. In particular, Hispanics are slightly more at a disadvantage than African Americans. Inequities in terms of regional parks are particularly concerning because regional parks receive the major share of park funding (see Chapter 6). The differences and effect sizes for regional parks ar e among the highest of the analysis of park spatial distribution. Logistic regression The six predictor variables included in the model (see Table 62) can significantly predict whether or not a census block is within a quarter mile from a regional park, 2 = 123.41, df = 6, N = 8094, p < 0.001. Table 62 shows the odds ratios, which suggest that the odds of having a regional park within a quarter mile of a census block are increasingly larger as median household income increases, and as percentage of Hispan ics, percentage of African Americans and percentage of owner occupied housing decrease. Percentage of African Americans and percentage of Hispanics are the strongest predictor variables. The odds of having a regional park within a quarter mile decrease by 0.988 for each increase of one percent of Hispanics and decrease by 0.977 for each increase of one percent of African Americans. The pseudo R squares are very low, Cox & Snell R Square = .015, and Nagelkerke R Square = .041, which show the low predictive p ower of the model.

PAGE 482

468 Table 62 Logistic Regression Results for Regional Parks B SE Odds ratio p Adjusted median household income .005 .001 1.005 .000 Percentage of Hispanics .012 .002 .988 .000 Percentage of African Americans .023 .004 .977 .000 Perc entage of people under 18 .003 .005 1.003 .564 Percentage of owner occupied housing units .007 .002 .993 .000 Percentage of vacant housing units .001 .006 .999 .811 Differences in demographics between access no access Independent sample non parametr ic median tests showed statistically significant differences for all demographic variables included in Table 63 and highlight more evident inequities than the logistic regression. Census blocks with access to regional parks have a significantly higher medi an household income ($70,737) than census blocks without access ($53,707), p < .001. Census blocks with access have a significantly higher median percentage of Non Hispanic whites (80%) than census blocks without access (61.15%), p < .001. Census blocks wi th access have a significantly lower median percentage of Hispanics (9.65%) than census blocks without access (18.5%), p < .001. Other differences are reported in Table 63. The effect sizes between mean ranks are low, per Cohen (1988) and the highest value was found for percentage of Non Hispanic Whites, r = .10.

PAGE 483

469 Table 63 Medians of Demographic Variables for Census Bloc ks With Access and No Access to Regional Parks Income % Non Hisp. Whites % Hispanics % African American % under 18 Access $ 70,737** 80%** 9.56%** 0%* 17.39%** No access $ 53,707** 61.15%** 18.5%** 1.06%* 21.57%** **: Statistically significant at the .001 level. *: Statistically significant at the .01 level. Differences in distances between demographic groups These differences partially confirm the findings from the first two sets of tests. In terms of income, there are no statistically significant di fferences among median distances to regional parks for low income census blocks (2500 meters), mid income census blocks (2500 meters), and high income census blocks (2415 meters). These distances are among the highest found in the analysis of park spatial distribution, as only 10 parks are classified as regional in Denver. When looking at ethnic differences, predominantly Non Hispanic White census blocks have a significantly lower median distance to the closest regional park (2061 meters) than mainly Africa n American (2500 meters) and mainly Hispanic census blocks (3250 meters), p < .001 (see Table 64) The difference between median distances of mainly African American and mainly Hispanic census blocks is also statistically significant, p < .001. All e ffect sizes for differences between mean ranks are low.

PAGE 484

470 Table 64 Median Distances to the Closest Regional Park for Census Blocks with Different Percentages of Ethnic Groups 25 50 % 50 75% 75 100% Non Hispanic Whites 2500 m 2319 m* 2061 m** Hispanics 2500 m 2500 m 3250 m** African Americans 2500 m 2500 m 2500 m** **: Statistically significant at the .001 level. *: Statistically significant at the .01 level. Figure 50 represents the geographic location of Denver's 10 regional parks and their quarter mile service areas. Regional parks are partially concentrated in east central and south central Denver, with Denver's three most historical parks, City Park, Cheesman Park, and Washington Park (see Figure 50). This helps explain high income Non Hispanic Wh ites' relatively good provision of regional parks. Also, the West Side has low access to regional parks, with the exception of Ruby Hill Park and partially Sloan's Lake to the north. Southeast Denver is the only quadrant of the city that does not include r egional parks. Most regional parks were established by the 1920s when land was cheaper and available, with the exceptions of Bear Creek Park, Central Park in Stapleton, and Parkfield Lake Park. In terms of park equity, the biggest inequities emerged from t he analysis of the median differences in demographic variables between census blocks with access and no access to regional parks (see Table 63).

PAGE 485

471 Figure 50 Service area (a quarter mile) for regional parks To summarize, the analysis of park quality not b ased on PQIY parameters, showed a substantial picture of inequity especially when focusing on parks with low violent crime density and regional parks. Areas within a quarter mile of parks with low violent crime density have significantly higher percentage of people under 18 than areas beyond a quarter mile, which is a small encouraging sign in the strong pattern of inequity emerged for the spatial distribution of this type of parks However, the situation is reversed for regional parks, as their quarter mil e service areas include lower percentages of people under 18 than areas located farther away from these parks.

PAGE 486

472 Park Spatial Distribution: Discussion The analysis of park spatial distribution covered 16 variables describing park access in terms of park pr oximity, park acreage, and park quality. The results of this complex analysis showed a multilayered landscape of environmental injustice including more cases of distributional inequity than cases of partial equity In extreme summary, the analysis of park proximity showed mixed results, while the results on park acreage and park quality highlighted several inequities In particular, African Americans are the ethnic minority who experience inequities most constantly across the analyses of proximity, acreage and quality. The findings on park proximity when considering parks only, which show that low income groups and Hispanics have slightly better access to parks than other groups, mostly reflect the results of previous investigations (see literature review in Chapter 2 ) Similar pictures of partial equity in terms of park proximity were found in Baltimore (Boone et al., 2009) Los Angeles (Wolch et al., 2005) New York City (Miyake et al., 2010) and Phoenix (Cutts et al., 2009) My findings on proximity to parks p artially contradict the results of my pilot study (Rigolon & Flohr, 2014) in which I found that low income groups and ethnic minorities had worse access to parks in terms of proximity than high income groups and Non Hispanic Whites. However, my pilot study focused o n 12 neighborhoods only (Rigolon & Flohr, 2014) while this dissertation focuses on all Denver, thus the different results can derive from specific conditions in the sample of the pilot study. Also, the findings of this dissertation show tha t in Denver African Americans are significantly disadvantage d in terms of park proximi ty when focusing on parks only, compared to other ethnic groups.

PAGE 487

473 On the other hand, when considering proximit y to parks and parkways combined, a pattern of inequity emerged, as low income people are underserved compared to high income groups while African Americans are better served than Non Hispanic Whites and Hispanics. However, some of the parkways that contri bute to African Americans' good access are not usable green spaces because they are arterial roads with a very small median (see Chapter 6). Also, the inclusion of parkways in the analysis has the effect of significantly changing the demographic groups tha t have better access in terms of park proximity. Among the literature I reviewed, I have not found any studies that focused on parks and parkways together. The analysis of park acreage shows a picture of inequity, particularly when focusing on park acreag e per 1000 people and per 1000 people under 18, which are parameters used to describe potential park congestion. The analysis of park and parkway acreage showed that high income groups are advantaged over low income groups, and that Non Hispanic Whites and African Americans have better acreage than Hispanics. However, as just mentioned, the parkways located in the mainly African American neighborhoods, which contribute to high acreage, are often thin medians that cannot be used for passive or active recreat ion. The analysis of acreage focusing only on parks shows another pattern of inequity. W hile low income groups and Hispanics have a better average service, more high income and Non Hispanic White census block have access to high park acreage, including ac cess to 10 park acres per 1,000 residents or more, which is a national standard describing a good level of service (Mertes & Hall, 1996) Also, Non Hispanic Whites and high income people have substantially better provision of acres of parks per 1,000 people

PAGE 488

474 un der 18 than low income people and ethnic minorities. These findings reflect the results of previous studies on park acreage (e.g. Boone et al., 2009; Dai, 2011; Sister et al., 2009; Wolch et al., 2005) although the inequities are not as striking as the ones found in Los Angeles (Sister et al., 2009) and Baltimore (Boone et al., 2009) However, these studies were conducted at the metropolitan scale, including suburban areas with larg er parks, lower population densities, and high concentrations of mid and high income Non Hispanic Whites, while my research focused only on the City and County of Denver. The analysis of park quality highlighted an unexpected picture of partial equity wh en focusing on some variables that depict parks in the top quartile of the PQIY and of some of its categories. Yet, substantial inequity patterns emerged from the spatial distribution of parks in the top five percent of the PQIY, parks with low violent cri me density, and regional parks. Most previous equity mapping research found that low income ethnic minorities are disadvantaged in terms of park quality, as shown by my extensive review of the literature (e.g. Ellaway et al., 2007; Loukaitou Sideris & Stieglitz, 2002; Vaughan et al., 2013) However, none of these studies had used a youth specific index to measure park quality, but they ha ve focused, for example, on the presence of playgrounds (Vaughan et al., 2013) on maintenance levels (Smoyer Tomic et al., 2004) and on the presence of environmental hazards (Carlson et al., 2010) Still, low income groups and Hispanics' high access to parks with good enough" standard of quality (top 25 percent of PQIY or its categories) is one of the most surprising findings of the analysis of park spatial distribution. This finding also contradicts the results of my pilot study of Denver (Rigolon & Flohr, 2014) because the pilot study focused only on 12 neighborhoods. However, areas with high percentages of African Americans do not enjoy

PAGE 489

475 high provisions of parks in the top quartile of the PQIY, further reinforcing their position of disa dvantage emerged from the other analyses. When studying access to parks in the top 5 percent of the PQIY parks with low violent crime density, and regional parks, the picture of partial equity is reversed to patterns of strong inequity that advantage hig h income groups and Non Hispanic Whites. To my knowledge, no other investigations ha ve focused on access to a city's best parks or to a city's regional parks, thus I cannot compare my findings to previous research. Several studies have analyzed crime rates in neighborhoods with different demographics and found that low income and ethnic/racial minority neighborhoods generally include higher concentrations of crime (Cutts et al., 2009; Franzini et al., 2010; Neckerman et al., 2009; Rutledge et al., 2003; Weiss et al., 2011) My findings strongly confirm this trend, as the highest differences a nd effect sizes emerged for parks with low violent crime density. These findings are particularly disturbing in terms of environmental justice because violent crime within an d around parks substantially limit s low income ethnic minority young people 's use of parks in Denver Studies that focused on Hispanic and African American low income children and teenagers in other U.S. cities found that fear of violence in neighborhood s and parks strongly limits young people's park visitation and home range (G—mez et al., 2004; Platt, 2012; Spilsbury, 2005; Stodolska et al., 2013) The inequitable distribution of best PQIY parks and regional parks raises concerns in terms of park investment and the potential of repeated park visitation. B est PQIY parks and regional parks are the parks that have receive d the city's largest investments in terms of amenities, maintenance and park planning efforts, and include Denver's most prestigious and historic parks like City Park, Washington Park, and

PAGE 490

476 Cheesman Park. As a result of higher investments, these "excellent" parks include a large variety of play amenities, natural elements and areas with different character s that can foster repeated visitation because of the numerous affordances and experiences they can provide T he presence of multiple and diverse affordance s prolong s young people's interest in parks due to numerous play choices (Castonguay & Jutras, 2009; Cosco, 2006; Jansson, 2008; McCormack et al., 2010) Parks including a variety of play affordan