Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Environmental justice in real estate, public services, and policy
(USC Thesis Other)
Environmental justice in real estate, public services, and policy
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Environmental Justice in Real Estate,
Public Services, and Policy
by
Madison. R. E. Swayne
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
URBAN PLANNING AND DEVELOPMENT
May 2020
ii
Acknowledgements
I had a lot of help with this.
I am thankful for this experience at USC and my advisor, Lisa Schweitzer, who always
encouraged and inspired me to do my best work. Lisa: thank you for the endless hours you
spent with me both in-person and online. Your feedback at every step of my project was critical
to bringing me to where I am today. I am also grateful for everything you taught me outside of
research itself. You have taught me how to be kinder to myself and others and have taught me
to stand up for what I believe, no matter what.
To the rest of my dissertation committee, thank you. Richard, I will miss reading and talking
with you. Our research walks and talks helped ground me and carry me forward. Marlon, thank
you for bringing me onto some of your METRANS projects. I have learned so much from you
methodologically and your ability to do it all is inspiring. Manuel, thank you for encouraging
me to pursue environmental justice questions and for showing me what it means to be an
engaged researcher. I will carry the lessons I learned in your classroom with me forever.
My life has changed drastically during my years at Price. Perhaps this is a product of being in
my early twenties. It could also be the nature of pursuing a Ph.D. Regardless, I will always look
back at this period as one of some of my highest highs and lowest lows. I lost some of my
closest loved ones and gained a few along the way.
To those who I have lost: I want to recognize both of my grandmothers. You were both always
so encouraging and asked some of the most interesting questions about my research. You were
both incredibly powerful women and I will continue to strive to make you proud. Nana, you
fought for people who didn’t have a voice and gave your time to causes you cared deeply about.
You were years ahead of the rest of the world in the fight for human rights. Grana, your
iii
attention to slow and forgotten things like tatting, antiques, preserving, and canning was a
beautiful representation of the beauty in slowness. You taught me to guard my time and focus
on the things that matter most. I am so thankful for every moment I had with you.
To the ones I have gained: Santi and Hobie, you are the best two furry creatures ever. Thank
you for keeping me company at my desk. Chris, you have been the most amazing partner
through this entire process. When we entered our respective Ph.D. programs, I did not know
where we would end up years later. Marrying you during my program was the best decision I
have made. I cannot wait for the years ahead. I am proud to have gained a husband in pursuit of
my Ph.D.
I want to thank my family for being by my side every step of the way. Mom, thank you for a
lifetime of education. Thank you for showing me how to ask questions and how to find the
knowledge I crave. Dad, thank you for encouraging me, for your patience and for your support.
To all my friends and extended family: Thank you for keeping me sane and inspiring me to go
on many, many adventures. Let’s keep at it.
Bless this mess. Hallelujah, I am done.
iv
Table of Contents
Acknowledgements .......................................................................................................................................... ii
List of Figures .................................................................................................................................................. vi
List of Tables .................................................................................................................................................. vii
Abstract ........................................................................................................................................................... viii
Introduction ....................................................................................................................................................... 1
1 Chapter 1: The Effects of Environmental Hazard on House Price Appreciation ..................... 4
1.1 Introduction ..................................................................................................................................... 4
1.2 Related Literature .......................................................................................................................... 5
1.3 Data .................................................................................................................................................... 8
1.3.1 Property Transaction Data ..................................................................................................... 8
1.3.2 Environmental Hazard Data ................................................................................................ 12
1.3.3 Sociodemographic Data ......................................................................................................... 15
1.4 Results ............................................................................................................................................ 16
1.5 Conclusions & Future Work ..................................................................................................... 20
2 Chapter 2: Park Access for the Transit-Dependent. Do communities really have access? ..... 1
2.1 Abstract ............................................................................................................................................. 1
2.2 Introduction ..................................................................................................................................... 1
2.3 Literature .......................................................................................................................................... 2
2.3.1 Measurement tools .................................................................................................................... 4
2.3.2 Distance-Based Methods .......................................................................................................... 4
2.3.3 Park Service Areas ..................................................................................................................... 5
2.4 Methods ............................................................................................................................................ 7
2.4.1 Study area .................................................................................................................................... 7
2.4.2 Data ............................................................................................................................................ 10
2.4.3 Measuring access to parks .................................................................................................... 12
2.5 Results ............................................................................................................................................ 14
3 Chapter 3: Access to Environmental Law: Incidence and distribution of California
Environmental Quality Act (CEQA) Environmental Impact Reviews and CEQA litigation ..... 26
3.1 Introduction: ................................................................................................................................. 26
v
3.2 Background: .................................................................................................................................. 27
3.2.1 What is CEQA ......................................................................................................................... 27
3.2.2 CEQA Litigation ..................................................................................................................... 31
3.3 Data: ................................................................................................................................................ 34
3.4 Findings ......................................................................................................................................... 37
3.4.1 Nature and Extent of CEQA Litigation ............................................................................ 37
3.4.2 CEQA Litigation & Time to Development ...................................................................... 39
3.4.3 Access to Environmental Law ............................................................................................. 40
3.5 Conclusion: .................................................................................................................................... 43
4 References: ............................................................................................................................................... 47
vi
List of Figures
Figure 1 – Price Index Comparison .......................................................................................................... 10
Figure 2 – Property Transactions by Year ............................................................................................. 12
Figure 3 – Mean RSEI Scores by Year .................................................................................................... 15
Figure 4 – Property Transactions Above RSEI Thresholds by Year .............................................. 18
Figure 5 – Map of Census Blocks by RSEI Score ................................................................................. 19
Figure 6 – Parks and Public Transit in Los Angeles County ............................................................. 11
Figure 7 – Representative Isochrones ...................................................................................................... 13
Figure 8 – Access to Parks Using Transit – 15 minutes ..................................................................... 16
Figure 9 - Access to Parks Using Transit – 30 minutes ...................................................................... 19
Figure 10 - Overview of CEQA Process .................................................................................................. 30
Figure 11 - EIRs by Project Type ............................................................................................................. 35
Figure 12 - Distribution of EIR Project Types ...................................................................................... 38
Figure 13 - EIR and CEQA Petition Locations ..................................................................................... 41
vii
List of Tables
Table 1 - Property Sales Summary Statistics ......................................................................................... 11
Table 2 - Repeat Sales Models .................................................................................................................... 22
Table 3 - Demographic Characteristics of Los Angeles County and the United States ................. 8
Table 4 - Los Angeles County Transit Providers ................................................................................. 11
Table 5 - Count of tracts with total number of parks accessible ........................................................ 14
Table 6 - Census tracts descriptives t-test for equality of means ...................................................... 17
Table 7 - Census tract descriptives t-test for equality of means ........................................................ 18
Table 8 - Linear Regression Results of Park Access ............................................................................. 20
Table 9 - Park Quality Index ...................................................................................................................... 21
Table 10 - Parks by Quality Score ............................................................................................................. 22
Table 11 - Linear Regression Results of Park Quality ......................................................................... 23
Table 12- EIRs by Year ............................................................................................................................... 34
Table 13 - EIR Cording Schema ................................................................................................................ 35
Table 14 - CEQA Petition Database Summary ...................................................................................... 36
Table 15 - Litigation by Project Type ...................................................................................................... 39
Table 16 - EIR Counts ................................................................................................................................. 42
Table 17 - Linear Regression of EIR Projects ........................................................................................ 43
viii
Abstract
In this dissertation, I examine how real estate, public service provision, and policies are
informed by and exacerbated by long-standing and well-documented patterns of environmental
injustice. Environmental justice has gained increasing attention, forcing scholars to improve
their measurement techniques and pushing policymakers to develop solutions to rectify the
disproportionate burden of environmental harm. At this point, many of the measurement issues
have been settled, but large questions regarding the impacts environmental justice patterns
have on multiple facets of life. Environmental justice is more than an issues of exposure to
toxics. Real estate markets, the built environment, accessibility, public service provision, and
policymaking are all interrelated with environmental justice. This dissertation uses mixed
methods, including big data, automated computing methods, and primary data sources, to begin
to understand the broad, urban impacts of environmental injustice.
In my first essay, I examine how spatial patterns of environmental contamination relate
to long-term impacts on local housing markets and wealth-building opportunities for people
living in communities of environmental justice concern. In this paper, I rely on repeat-sale
house price data and the US EPA’s Risk Screening Environmental Indicator data to understand
whether environmental contamination lowers home sale prices. Environmental contamination
may change homeowners’ appreciation rates and additionally harm home asset values in these
neighborhoods—a potential “double whammy” for lower-income homeowners and homeowners
of color. My study aims to understand and document long-term dynamics in house price
appreciation. The results suggest that houses in more contaminated communities appreciate
about one percent per year slower than their less contaminated counterparts, offering a lower
rate of return for homeowners who may already be at a disadvantage.
ix
My second essay examines how well served Los Angeles County residents are by transit
and its connections to local parks. Drawing from a new, computer coded tool I developed to
measure access via public transit, I document access to parks for every neighborhood in Los
Angeles. Results show that low-income communities of color have access to a higher quantity
of parks with mixed results in terms of differential access to quality parks. This paper offers an
improvement over previous studies that relied on distance-based methods and park service area
boundaries to calculate park access by using real travel time via transit as the measure for
access.
My third and final essay investigates claims about the California Environmental Quality
Act (CEQA) – California’s hallmark environmental law – and its interaction with development.
Even though CEQA statute guides so much of the development in California, little systematic
research documenting its interactions with projects through the EIR process and through
litigation exists. Results show that CEQA oversight varies according to project type but that
residential projects aimed at helping to solve California’s affordability crisis are seldom subject
to CEQA litigation.
This research makes significant contributions to urban planning scholarship. First, I
identify the potential long-term impacts on real estate markets in association with differential
patterns of environmental contamination. I also create a new tool for measuring access via
transit and use these data to identify opportunities to increase equity via public service parks
provision. And finally, I contribute to California policy conversations on CEQA by
documenting CEQA’s interactions with development and communities.
1
Introduction
Environmental justice holds that all people, regardless of their identity, should have the
same protections from environmental harm and access to participation in environmental
decision-making and amenities (US EPA, 2017). Environmental justice has influenced advocacy
efforts, academic scholarship, and broader definitions of environmental quality (Schweitzer &
Stephenson, 2007). Since its inception in the 1980s, environmental justice researchers have
attempted to measure and understand how environmental hazards such as hazardous waste
facilities, are distributed across space and time, affecting vulnerable communities in ways that
violate theories of justice. As recognized by Schweitzer and Stephenson (2007), few studies have
questioned how the established patterns of environmental hazards permeate multiple facets of
urban life beyond exposure to harmful health effects. This dissertation takes previous
environmental justice studies documenting harm to be the starting point for understanding
urban systems and pushes environmental justice scholarship to consider interactions with land
use and policy – two central tenets of urban planning.
The term environmental justice was coined in 1987 to describe the findings of the United
Church of Christ’s Commission for Racial Justice report on the hazardous waste sites and their
proximity to minority communities. Other studies around the same time also found that low-
income, minority communities were more likely to be burdened by nearby noxious land uses
(United States General Accounting Office, 1983; Bullard, 1983). Following these early reports,
environmental justice research exploring the characteristics of residents living near
environmentally hazardous facilities flourished. However, early national studies employed rough
geospatial methodological approaches, relying on zip code and larger spatial units of analysis.
Criticism of these studies soon recognized issues of the modifiable areal unit problem (MAUP)
where analyses at different geographic scales yield different results (Anderton et al. 1994; Cutter
et al., 1996; Sheppard et al., 1992; Mennis, 2002).
2
Over time, methods for understanding environmental distributions have become more
sophisticated in recognition of the subjective, social constructions which underlay map making
(Maantay, 2002). Unit-hazard coincidence models (Been, 1995; Been & Gupta, 1997) are now less
often used. Instead, distance-based methods that measure the precise locations of environmental
hazards and their distances to nearby residents are specified (Mohai and Saha, 2006). Exposure
indices designed to measure the actual amount of pollutions residents are exposed to have also
been employed to better understand the health effects of environmental inequality (Bowen et al.,
1995; Boer et al., 1997; Pastor et al., 2005). Research has fairly consistently found income and
race to be associated with environmental hazard; neighborhoods with larger minority
populations and lower incomes have more significant environmental burdens (Ash and Fetter,
2004; Been, 2004; Hamilton, 1995; Sadd et al., 1999; Mohai and Bryant, 1992; Morello-Frosh et
al., 2001; Mennis and Jordan, 2005; Pastor et al., 2002). While both race and income are
repeatedly found to be important, income does tend to be a stronger predictor of exposure to
harm.
Substantial evidence has also been developed to discount theories of minority move-in
(Oakes, Anderton, & Anderson, 1996; Been & Gupta, 1997; Pastor et al., 2001). The siting of
existing hazardous land uses and location of neighborhoods has occurred through political and
economic processes over time, but disproportionate siting matters more than minority move-in
to hazardous areas (Pastor et al., 2001). Distributional environmental inequity is a result of
discriminatory siting decisions.
This literature has led to a few significant substantive conclusions. Many of the early
methodological issues in the literature have been settled. Researchers now rely on distance-based
hazard measures and neighborhood-level geographies in recognition that proximity to hazards
is our best measure of environmental exposure, and the modifiable area unit problem necessitates
careful thought when deciding on a geographic unit of analysis. Using these methods, researchers
find that environmental disparities across populations are not fully explained by land use
3
patterns, market mechanisms, or population density. Instead, race and class are more consistently
associated with environmental disparity, with minority racial status being a stronger predictor
than low-income status of exposure to environmental hazards.
With this established baseline in the environmental justice literature, we are missing an
understanding of how these well-studied patterns of environmental disparity impact the
everyday lives of people living in neighborhoods with degraded environments. The research
presented here seeks to extend our understanding of environmental justice from documenting
environmental patterns to tracing their effects throughout our urban systems. In the chapters
that follow, I do not ask what causes environmental injustice or seek to measure the distributions
of particular environmental hazards. I aim to provide empirical research on the outcomes of
environmentally unjust patterns.
4
1 Chapter 1: The Effects of Environmental Hazard on House
Price Appreciation
1.1 Introduction
Homeownership is critical to wealth accumulation in the United States and a
quintessential part of the “American Dream.” Studies of homeownership in the United States
have justified homeownership as a worthy economic investment – giving homeowners a stable
asset and protection from rising rents. Homeownership also provides social and emotional
benefits which may include health, civic engagement, better child outcomes, and life satisfaction
(Di, et al., 2003; Bostic & Lee, 2009). Nearly 70 percent of United States households are occupied
by owners, but homeownership rates and experiences are not consistent across demographic
lines. While there have been new efforts to promote homeownership among lower income and
racial minority Americans, homeownership rates among black and Latino Americans are still far
lower than that of white Americans (Bostic & Martin, 2005). These are also the same residents
who environmental justice research have found to be at higher risk of being disproportionately
exposed to environmental risks. While the environmental justice literature seeks to understand
how race and income influence the distribution of environmental hazards, this research must be
done in consideration of housing markets which both reflect and drive the reproduction of
inequality by race and income.
This paper investigates the relationship between environmental hazards and home price
appreciation. I seek to understand differential opportunities for wealth-building through
homeownership between neighborhoods of varying environmental quality. Environmental
hazards, among many other negative outcomes for health and human flourishing, can alter home
values and wealth creation inter-generationally and across different groups. Neighborhood
nuisances can influence homeowners’ perceptions about the resale value of their homes, and thus
discourage their on-going investments in home maintenance. In so doing, homeowners’
decisions can compound the effects of the hazards on home prices.
5
This paper analyzes these questions in residential neighborhoods across Los Angeles
County. I find that countywide, home value in neighborhoods with above average levels of
environmental hazard appreciate about one percent per year slower than homes with average and
lower than average levels of environmental contamination.
This paper builds on three existing strands of research investigating: a) housing wealth
and the housing wealth gap; b) hedonic impacts of amenities and disamenities on housing price
levels; and c) the unjust distributions of environmental harm across poorer, minority
neighborhoods. I investigate the relative appreciation rates between houses in census blocks with
varying environmental contamination. Through repeat sales, I compare property transactions to
the previous transaction at that address to understand how house prices in areas of differential
environmental toxicity change over time. By using the repeat sales methodology (Bailey et al.,
1963), I control for other house type structural characteristics that hedonic methods may not
capture. This paper relies on the United States Environmental Protection Agency’s (US EPA)
Risk Screening Environmental Indicators (RSEI) data which is the only systematic database of
environmental releases that allows for relative comparisons of environmental harm and that is
standardized across time (Ash & Fetter, 2004). Using RSEI data allows me to confidently track
the relationship between house price appreciation rates and environmental toxicity. And because
RSEI is a nationwide dataset, I provide an analysis of Los Angeles County, the most populous
county in the United States with more than 10 million residents. No other environmental dataset
allows for such broadscale, longitudinal approaches.
1.2 Related Literature
Aspirations of the “American dream” connect home ownership and building housing
wealth over time. Homeownership provides housing stability and significant wealth from home
equity (Kim, 2000). However, the American dream of homeownership and its promises of wealth
generation leave much to be desired for black and Hispanic households whose homeownership
rates lag 20 or more percentage points behind their white counterparts (Kim, 2000; (Bostic &
6
Martin, 2005) and where black homeowners have significant gaps in housing equity generation
when compared their white home owning counterparts (Parcel, 1982). Among homeowners,
there is an estimated “racial premium” of an all-white neighborhood of about 3 percent per year
compared to all-minority neighborhoods. Kim suggests that racial minorities are less likely to
own homes because the homes they own appreciate slower and do less to build household wealth
over the long-term. While it is unknown if minority potential homebuyers have good information
about their expected financial returns from homeownership, there is a well-documented wealth
accumulation gap between white and minority homeowners (Kim, 2000; Parcel, 1982) making it
harder for minority families to afford the down payment on a house. Racial wealth gap aside, in
a 1983 cross-sectional analysis of white and minority wealth-constrained households, households
without enough accumulated wealth to meet standard down payment and closing cost
requirements, whites still own at a systematically higher rate than minorities (Gyourko,
Lienneman, and Wachter, 1998). While homeownership is just one of the ways to close the racial
wealth gap, it is the most important driver of the gaps that exist (Shapiro et. Al, 2013) as
homeowners hold almost all of the nation’s wealth (Di, 2005).
This idea that house prices in more toxic environments should appreciate slower than
houses in less toxic environments is a purely economic argument. However, there are
distributional and racial effects built into this argument because of what we know about the
unequal spatial distributions of minority residents and environmental hazards. Environmental
disamenity effects should be apparent and reveal themselves in the form of slower home price
appreciation rates.
Hedonic estimation studies aim to understand people’s willingness to pay to avoid
environmental disamenities such as poor air quality (Kiel and McClain, 1995; Zabel & Kiel, 2000),
noxious land uses like Superfund sites (Gayer, Hamilton & Viscusi, 2000), waste facilities (Hite
et al., 2001; Kohlhase, 1991), and poor water quality (Palmquist et al., 1997). Boyle and Kiel
(2001) reviewed these hedonic studies of environmental hazard and home prices and found that
7
air quality studies typically find that signs on coefficients for air pollution are not usually
statistically significant, and the signs on the coefficient for air quality depend on other included
variables. Water quality studies typically find that the signs on the coefficients are correct and
are generally statistically significant. Studies investigating noxious land uses find the most
promising results with signs on coefficients being as expected and results being statistically
significant. However, as information about the noxious sites changes, the impact of noxious sites
on nearby house prices changes. This finding suggests that longitudinal studies across
neighborhoods can provide important insights on hazards and markets that point-in-time
estimates cannot.
Janet Kohlhase (1991) analyzed the impact of Superfund sites on nearby housing values.
Using longitudinal data on individual houses in Houston, she tracked how new information from
the EPA on the Superfund site or its cleanup impacts buyers’ willingness to pay. Her results
suggest houses within 6 miles of a newly declared Superfund site suffer from lower prices but
houses outside the 6-mile radius are unaffected. She also finds that publicly available information
about these sites is incomplete. Individuals’ perceived risk may be different from the risks
measured in available data sources. If so, marginal prices may not significantly correlate with the
toxic site’s toxicity — likely due to inconsistent risk perceptions and incomplete or ineffective
public information. Kiel and McClain (1995) find similar results in Massachusetts where house
price appreciation rates are affected by the construction and operation of a local waste incinerator
however results are limited to the effects of one locally undesirable land use (LULU) instead of
the entire collection of LULUs in the area.
In addition to hedonic studies, scholars have examined the distribution of these
environmental hazards across communities with different socioeconomic and demographic
characteristics. Overwhelmingly, this literature reveals that community of low income and
minority racial status face higher levels of environmental contamination (Boer et al., 1997; Pastor
et al., 2005; Hird, 1993). Minority communities are home to a disproportionate share of toxic
8
facilities with racism being a stronger predictor of exposure to harm than class and the locational
decisions made by firms, government actors, and non-governmental actors in the process of
siting these facilities is not random. Unequal siting decisions are the result of racist,
discriminatory acts (Pulido, 1996). Critics of this argument have posited that these unjust
distributions of toxics are not the result of racialized siting, but instead the result of minorities
moving into more toxic environments due to lower rents. However in the case of Los Angeles
County, this is not true (Pastor et al., 2002).
Using repeat sales methodology, I investigate eleven years of home sales data in Los
Angeles County to understand how local racial composition and environmental hazards impact
home price appreciation rates. The repeat sales methodology employed here allows us to control
for individual housing unit characteristics, to assess the appreciation rates of specific houses over
a long period of time. For this research problem, repeat sales is preferable to hedonic methods
where price appreciation is the result of the revealed preferences of buyers and less so, the actual
regional price differences in home with the same structural characteristics. This paper is the first
of its kind aimed at understanding how these unevenly distributed environmental hazards and
continued patterns of racial segregation may potentially be impacting adjacent housing markets
in the long run. With our existing understanding of the stalled home price appreciation rates for
houses owned by minority residents in minority neighborhoods and hedonic studies on the
impacts of nuisance on house prices, we would expect to see slower appreciation rates for houses
located in more toxic environments where minority families are more likely to be homeowners.
However, results from this study are limited where appreciation rates seem to stall only in the
absolute most contaminated neighborhoods.
1.3 Data
1.3.1 Property Transaction Data
For this analysis, I used real housing transaction data from Los Angeles County,
California. I obtained these transaction records from CoreLogic (formerly DataQuick) which
9
collects yearly data on property transactions across the United States. The original data file
included over 3 million transactions of 1.3 million properties across the county from 1999 to
2010. Some transaction records were incomplete, lacking identifiable address or price
information. I removed these transactions from the record. Each property in the dataset has a
unique property identification number, as does each transaction. The transaction record includes
information on property location, price paid, loan type, loan amount, and characteristics of both
the house and the underlying land. I geocoded each property in the dataset so that other
geospatial data, including environmental hazard data, could be added. Census block geographies
from 2010 were linked to each property.
Using the unique property identification numbers, I identified repeat sales of individual
homes from 1 January 1999 to 31 December 2010. A total of 422,264 properties failed to qualify
as repeat sales (i.e. only appeared once) during the time period and were dropped from the dataset.
The complete, cleaned repeat sales transaction record tracks the 329,247 total properties that
sold two or more times over the study period.
To ensure repeat transaction data representativeness, I created a price index to compare
these Los Angeles County repeat sales transactions to two national home price indices —
CaseShiller and the Federal Housing Finance Agency’s House Price Index (FHFA). This
comparison appears in Figure 1 below.
10
Figure 1 – Price Index Comparison
Upon initial comparison to these indices, the Corelogic all repeat housing sales data
exhibited more severe price changes between years and overstated the market crash in 2007.
While CoreLogic provides limited data about the individual property buyers and sellers, it is
possible to retroactively estimate which transactions in the record relied on nontraditional
loans or were not arms-length transactions. Corelogic tracks loan type (fixed-interest versus
variable-interest) and loan amount in their record. This data structure allowed for trimming of
the data to exclude variable-interest rate loans and loans of more than $417,000 (the
Freddie/Fannie maximum loan amount). Removing loans that fall into either of these
categories helps proxy for some of the anomalous transactions associated with the subprime
lending crisis between 2007 and 2010 that contributed to the Great Recession. After trimming
to remove non-traditional loans, the Corelogic repeat sales mean price index more closely
mirrors the national indices. Table 1 contains summary statistics for all sales, all repeat sales,
and all repeat sales with fixed interest loans under the Freddie/Fannie maximum.
-60
-50
-40
-30
-20
-10
0
10
20
30
40
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Mean Home Price Percent Change
Price Index Comparison
Case Shiller
FHFA
Corelogic - All repeat housing sales
Corelogic - All conventional repeat housing sales*
*Note: Conventional repeat housing sales include only sales with fixed
interest loans and loan amounts under the Frannie/Freddie Mac maximum
11
Table 1 - Property Sales Summary Statistics
Table 1.
Property Sales Summary Statistics
All Sales All repeat housing sales All conventional repeat
housing sales*
Variable Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.
Price $448,789.80 $1,702,458.00 $429,980.70 $1,558,502.00 $266,954.50 $711,925.30
Bathrooms 2.163 1.030 2.137 0.994 2.004 0.824
Bedrooms 2.957 1.086 2.935 1.060 2.841 0.972
Ln
distance
to city
center
2.715 0.598 2.730 0.614 2.745 0.622
Lot size
(sq. feet)
47,050 1,095,333 47,166 951,612 47,411 852,803
Number of
stories
0.672 0.607 0.645 0.600 0.612 0.577
Total
properties
751,511 329,247 233,767
*Note: Conventional repeat housing sales include only sales with fixed interest loans and loan amounts under
the Frannie/Freddie Mac maximum
To further evaluate the representativeness of the transaction record, I compared the
distribution of transactions across years for both the trimmed and untrimmed transaction
datasets. In the trimmed dataset, the distribution of total transactions per year changed such
that fewer conventional housing sales occurred in the three-year period from 2004-2007
leading up to the market crash. This distribution is consistent with what we know about home
mortgage lending during those years (Goodman and Mance, 2011; Gertler and Gilchrist,
2018). Gertler and Gilchrist (2018) estimate that 30 percent of newly issued mortgages were
issued with variable interest rates during 2005 and 2006. These mortgages carried high default
rates. Removing them from the repeat sales database helps insure that properties and property
owners under comparison have similar characteristics.
12
Figure 2 – Property Transactions by Year
1.3.2 Environmental Hazard Data
To measure environmental hazard at each property, I merge United States
Environmental Protection Agency (EPA) Risk-Screening Environmental Indicators (RSEI)
data from 1999 to 2010 to each property transaction in the record. The RSEI data are derived
from the Toxic Release Inventory (TRI) which is an annual dataset collected by the EPA in
which industrial polluters are required to self-report quantity and composition of their
environmental releases into the environment via air, water, or disposal (EPA, 2018a). The 1986
Emergency Planning and Community Right-to-Know Act, which established the TRI
Program, designates which polluters are required to report releases. Required reporters include
operations in specific industry sectors (manufacturing, mining, electric power generation,
hazardous waste treatment and disposal, and solvent recovery) who have ten or more full-time
employees (U.S. EPA. 2018). Federal facilities are also mandated reporters.
Researchers have used TRI data to investigate community exposures to toxics but
relying on TRI data alone does not provide the most comprehensive picture of the potential for
human harm from exposure to toxics.
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
0
2
4
6
8
10
12
14
16
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Number of Total Sales
Percent of Total Sales
Transaction Year
Total Transactions by Year
All repeat housing sales
All conventional repeat housing sales*
*Note: Conventional repeat housing sales include only sales with fixed interest loans and loan amounts under the
13
TRI data only includes the total tonnage of chemical releases and does not account for
differential toxicity between chemicals. In consideration of the potential for human harm, it
does not only matter how much hazardous material is released, but also the composition of each
release. Chemicals vary widely in their pound-for-pound toxicity (Collins et. al, 2016). RSEI
uses TRI releases, combined with information on chemical fate and transport, and toxicity, to
estimate a pollutant concentration value. This modeling also considers exposure to toxics via
inhalation, ingestion, and direct skin contact.
Unlike many other pollution databases, RSEI calculations on the fate-and-transport of
chemicals moving through the environment do not rely on politically defined geographic
boundaries like county or city lines. Instead, geography is delineated into a network of cells
which cover the entire country and toxicity estimates are produced for equally sized one-
kilometer-square “grid cells” (Ash et al., 2009). Higher health impacts are attributed to grid
cells with exposure to more toxic chemicals. This grid cell approach, when compared to other
approaches relying on unequally sized geographic units, more accurately measures
environmental impacts across space and reduces potential for the modifiable areal unit problem
to introduce bias into the data on environmental harm (Mennis, 2002). With RSEI, the model
for harm does not include a priori assumptions of “community,” allowing researchers and other
data users to manipulate and aggregate RSEI grid cells into various geographies, as applicable
to the research question at hand. I aggregated the RSEI data used in this paper to the census
block level with boundaries held consistent to the 2010 census block geography. This is the
smallest geography for which demographic data were available. The use of RSEI allows for
analysis of exposure to toxics, rather than just proximity to toxics and provides us with the
best possible large-scale dataset on environmental toxics.
Like all datasets, RSEI of course has its limits (Political Economy Research Institute,
2019). RSEI relies on self-reported TRI on tonnage of releases. Firms can report a range of
pounds released for releases under 1,000 pounds and there may be instances of over- or under-
14
reporting. Facility information can also be skewed by inadequate information on stack heights,
and facility location. And, like all human health and environmental risk assessment work,
toxicity for some chemical groups may be missing or chemicals may be grouped according to
presumed toxicity similarities in instances where toxicity measures are missing. The TRI was
not design to capture mobile sources of air pollution such as automobiles traffic and small
facilities like auto body shops and dry cleaners are excluded (Ash et al., 2009). While data that
includes mobile and smaller pollution sources would offer a more complete picture of what is
released into the environment, RSEI is the most robust dataset currently available.
In order to check the robustness of the RSEI dataset across the study period and within
the Los Angeles County study area, RSEI yearly means were compared at the national, state,
and county level. RSEI data includes two values: a toxicity-weighted concentration-based
hazard score and a land-area weighted concentration-based score. For understanding the
potential for human harm due to nearby releases, the toxicity-weighted concentration-based
hazard score is best. This score does not include the number of nearby people in its calculation,
making the identification of high-risk, low-population areas possible. It is also important to
understand that RSEI scores can only be compared to other RSEI scores. These data are not
comparable to other toxic indices. A RSEI score of ten in one block represents a risk that is ten
times higher than the risk in a block with a score of one. For this reason, there is no specific
RSEI value or threshold to indicate that a location is objectively hazardous. Each location must
be evaluated in comparison with others.
Annual means for all blocks in the given geography are presented in the tables below.
While the RSEI program began prior to 1999, RSEI data prior to 1999 was removed from the
set due to changes in the facilities which were required to report as part of the TRI in that year.
Only including data from 1999 to 2010 ensures that facilities and chemicals under
consideration are held constant across time.
15
Figure 3 – Mean RSEI Scores by Year
Comparing annual RSEI means reveals that on average, Los Angeles County and California
have lower RSEI scores than the United States as a whole. However, the County and State have
much larger standard deviations with more severe changes from year to year. While being
lower than the national average is nothing to brag about, it is important to understand that
estimates of environmental hazards’ impacts on home price appreciation in the Los Angeles
County study area may in fact be underestimations of the national effect.
1.3.3 Sociodemographic Data
Property transactions and RSEI data were combined with US Census Bureau American
Community Survey (ACS) 5-Year Estimates from 2010. Unfortunately, ACS data products did
not exist prior to 1999 so yearly estimates of demographic data in the housing transaction
panel were not possible. ACS data are reported at the census tract level and each transaction in
the record has been merged to its appropriate tract-level demographic information. ACS
variables are used as controls in the estimated models.
0
2,000
4,000
6,000
8,000
10,000
12,000
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Mean RSEI Toxicity-Weighted Concentration-Based Hazard Score
United States - Mean California - Mean Los Angeles County - Mean
Standard Deviation:
United States - 2,780
California - 3,331
Los Angeles County - 8,933
16
1.4 Results
This study applies repeat sales regression to measure the impact of environmental hazard
on residential house price appreciation. By relying on repeat sales, the models control for
variation in housing unit characteristics and allows for a controlled experiment on the impacts
of hazard. I estimate the following model [1] using the basic repeat-sales equation:
ln#
𝑃′
!"
𝑃
!"
&=𝛽
#
+*𝐷
!$
%
$&'
+ *𝛽
'
%
$&'
𝑅𝑆𝐸𝐼
("
+𝑒
!"")
where 𝑃′
!"
/𝑃
!"
is the change in ln-price at property i at time t. 𝐷
!#
is a dummy variable set to -1
at time of initial sale t, +1 at subsequent sale at time t’, and 0 otherwise. 𝛽
$
is the coefficient on
the RSEI value of block b at time t and 𝑒
!""%
is the error term. This model includes all of the
repeat sales properties and effects are estimated across the entire county. Results of this model,
shown in the attached tables (Model 1), suggest RSEI values for some years are statistically
significant with small negative coefficients indicating a small negative impact on appreciation
rates of houses sold during that year. However, we would expect coefficients to be negative and
statistically significant across all years of data and this does not turn out to be true. We cannot
confidently identify a significant difference in the appreciation rate of housing prices for homes
in places with differential environmental hazard. In this most basic model, results are
inconsistent with all of the theory we have on housing, wealth, and the effects of environmental
hazard. I expand my model to include additional variables which may affect house prices.
Model 2, includes a dummy variable for whether or not a house falls within the Los
Angeles City boundaries and a measure of the percent of manufacturing jobs in a house’s census
tract. This measure is designed to capture if homebuyers are trading higher pollution for access
to manufacturing jobs. Because RSEI data are generated by facilities who report to the TRI, we
would expect that areas with greater pollution levels, also have more manufacturing jobs
available. These jobs could potentially be an asset to job seekers willing to trade a steady
income for higher levels of pollution. Results suggest that being within the city of Los Angeles
has a small, positive, statistically significant impact on house price appreciation, with no
17
discernable correlation between manufacturing jobs and appreciation rates. This model also
considers a house’s location in relation to the ocean – a natural amenity in Los Angeles. Los
Angeles geography is unique due to the fact that the area is bounded by mountains to the east
and ocean to the west. Generally, house prices increase the closer you get to the ocean with
beachfront properties fetching some of the highest prices in the region. The results indicate this
effect to be true and statistically significant.
Even with each of the additional specifications in Model 2, the results on toxicity are
inconsistent with theory. One potential explanation for this is that Los Angeles County is too
large a geography to consider, with many census blocks having very similar RSEI scores. The
annual RSEI scores are not normally distributed; most blocks are clustered around the mean
and there are large standard deviations. It is possible that this clustering of RSEI values may
obscure differential price appreciation rates in some of the most toxic areas. Unless the
environmental is extremely hazardous, it may be difficult or impossible for homeowners to
recognize their potential environmental risks and exposure.
To investigate this, properties were marked with a binary identifier according to their
RSEI value in comparison to a set of determined thresholds. Properties were marked at
fourteen different thresholds and the frequencies of transactions above these thresholds are
reported in the figure below.
18
Figure 4 – Property Transactions Above RSEI Thresholds by Year
The spatial distribution of census blocks with RSEI values one or more standard deviations
from the mean were also mapped. Neighborhoods with the highest scores, furthest from the
means are in South and East Los Angeles including South Gate, Compton, Torrance and
others. This spatial distribution aligns with previous studies on environmental distributions in
Los Angeles. A map of average RSEI values by block appears below.
83,541
79,502
70,628
58,031
48,489
41,542
21,007
11,902 11,823
2,372
1,200
454 23 17
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
6,000
Mean
7,000
8,000
9,000
10,000
15,000
20,000
Mean + 1
S.D.
Mean + 2
S.D.
50,000
100,000
500,000
1,000,000
NUMBER OF TRANSACTIONS
RSEI SCORE
Number of transactions with RSEI toxicity value greater than
threshold
19
Figure 5 – Map of Census Blocks by RSEI Score
A second set of tests were performed to examine whether house price appreciation is different
across properties, above the various RSEI thresholds. This second set of models includes
dummies, Ijt, that are equal to 1 if the RSEI value at property j and time t is above the
determined RSEI threshold. This second set of regressions are estimated using:
ln#
𝑃′
!"
𝑃
!"
&=𝛽
#
+*𝐷
!$
%
$&'
+ *𝐼
!"
%
$&'
+𝑒
!"")
Tests of differential appreciation rates are performed at each of the RSEI toxicity thresholds of:
6,000; above the mean; 7,000; 8,000; 9,000; 10,000; 15,000; 20,000; 50,000; 100,000; 500,000;
and 1,000,000. Thresholds at the mean, mean + 1 standard deviation, and mean + 2 standard
deviations were also used. Thresholds were determined based on distribution of RSEI scores.
20
In this second set of tests, RSEI values above a threshold of 10,000 have a statistically
significant impact on home price appreciation rates of about one percent (0.92%) per year
(Model 8). This constitutes about 17% of the sample. This effect on price appreciation continues
to hold at the 15,000 RSEI score threshold where the estimated impact is 1.85% per year
(Model 9) and at the 20,000 RSEI score threshold [Model 10]. For properties with RSEI
scores of 20,000 or higher, there is an estimated downward impact on home price appreciation
of 1.59% per year. The effect of RSEI score on price appreciation is no longer detectable at
RSEI scores above 20,000 due to a very limited sample size of 1,200 transactions or fewer
(Models 11-14).
I performed addition tests at RSEI thresholds in reference to the mean. To do this, I
marked property transactions according to whether or not the RSEI value at the property was
above the mean during the sale (Model 15), one standard deviation above the mean (Model 16),
two standard deviations above the mean (Model 17), or three standard deviations above the
mean (Model 18). Properties where the RSEI value was one standard deviation or more above
the mean exhibited house price appreciation rates that were about 4% per year slower than
properties below this threshold. This effect held at each of the three standard deviation
thresholds.
1.5 Conclusions & Future Work
House price appreciation rates are affected by high levels of environmental
contamination. Using the log of house sales prices, appreciation rates of homes in census blocks
with RSEI scores above 10,000 are on average, one percent per year slower than homes in areas
with near-average environmental hazard scores. Based on the results presented here on Los
Angeles County homes that sold between 1999 and 2010, we can expect about 17% of the
transacted houses had lower appreciation rates in association with environmental
contamination. The houses with high RSEI values and stalled appreciation rates are in places of
historic environmental justice disadvantage where residents bear disproportionate
21
environmental burden. There is now evidence suggesting that in addition to the justice
implications posed by high incidences of environmental harm, there are real, measurable
economic implications too.
These results also suggest that homebuyers and sellers may only begin to internalize
the costs of environmental harm when harm reaches a particularly high level. Home buyers and
sellers may be unaware of or potentially unbothered by a home’s proximity to environmental
hazards when exposures are close to the average for the area. Where homes are in
neighborhoods of relatively low RSEI scores, homeowners and homebuyers may discount
environmental contamination issues vis-à-vis the hundreds of other factors that go into the
home buying decision such as school availability, access to jobs, and proximity to family and
friends, among many others.
The full cost of siting and operation of environmentally hazardous land use may also
not be fully realized and the long-term impacts of hazardous land use on the residential real
estate market should be considered. The observed differences in appreciation rates are steady
over time in areas where environmental contamination is high suggesting long-term impacts to
the market.
These findings improve our understanding of the potential for longstanding impacts
from environmental injustice on real estate market dynamics and opportunities for wealth
building within communities of environmental justice concern. They also provide evidence to
support programs which do not just push homeownership as the path to upward mobility, but
force us to think critically about how home-owning may present differential opportunity
contingent upon the surrounding environmental quality. It is crucial for environmental justice
scholars, urban planning practitioners, community organizers and homebuyers to understand
the potential effects of disproportionate burdens of environmental harm not just on
environmental and human health, but also on overall land use, real estate, and economic
patterns.
22
Table 2 - Repeat Sales Models
Table 2. Repeat Sales Models
VARIABLES Model 1 Model 2 Model 3
Dummy 1999 0.182*** 0.183*** 0.183***
(0.010) (0.010) (0.010)
Dummy 2000 0.139*** 0.139*** 0.139***
(0.006) (0.006) (0.006)
Dummy 2001 0.0803*** 0.0832*** 0.0813***
(0.004) (0.004) (0.004)
Dummy 2002 0.132*** 0.136*** 0.134***
(0.004) (0.004) (0.004)
Dummy 2003 0.185*** 0.185*** 0.184***
(0.003) (0.003) (0.003)
Dummy 2004 0.343*** 0.352*** 0.350***
(0.005) (0.005) (0.005)
Dummy 2005 0.491*** 0.497*** 0.494***
(0.006) (0.006) (0.006)
Dummy 2006 0.503*** 0.509*** 0.506***
(0.007) (0.007) (0.007)
Dummy 2007 0.456*** 0.459*** 0.456***
(0.006) (0.006) (0.006)
Dummy 2008 0.155*** 0.164*** 0.161***
(0.004) (0.004) (0.004)
Dummy 2009 -0.0430*** -0.0312*** -0.0335***
(0.004) (0.004) (0.004)
Dummy 2010 -0.0197*** -0.0140*** -0.0161***
(0.004) (0.004) (0.004)
RSEI Toxicity 1999 1.70e-06** 0.00000124 1.33e-06*
(0.000) (0.000) (0.000)
RSEI Toxicity 2000 1.04e-06* 0.000000618 0.000000667
(0.000) (0.000) (0.000)
RSEI Toxicity 2001 4.77E-08 -0.000000447 -0.00000032
(0.000) (0.000) (0.000)
RSEI Toxicity 2002 -2.50e-06*** -3.33e-06*** -3.12e-06***
(0.000) (0.000) (0.000)
RSEI Toxicity 2003 -2.24E-08 -3.08E-08 -3.33E-08
(0.000) (0.000) (0.000)
RSEI Toxicity 2004 -1.68e-06*** -2.76e-06*** -2.62e-06***
(0.000) (0.000) (0.000)
RSEI Toxicity 2005 -0.000000169 -6.38e-07* -0.000000584
(0.000) (0.000) (0.000)
RSEI Toxicity 2006 1.20e-06** 0.000000453 0.000000531
(0.000) (0.000) (0.000)
RSEI Toxicity 2007 1.07e-06*** 0.000000652 7.18e-07*
(0.000) (0.000) (0.000)
RSEI Toxicity 2008 1.48e-06*** 8.97e-07*** 9.11e-07***
(0.000) (0.000) (0.000)
RSEI Toxicity 2009 3.78e-06*** 1.78e-06*** 1.87e-06***
(0.000) (0.000) (0.000)
RSEI Toxicity 2010 2.88e-06*** 1.85e-06*** 1.90e-06***
(0.000) (0.000) (0.000)
LAcity
-0.0110*** -0.00241
(0.003) (0.003)
Dist_Ocean
-0.00341*** -0.00348***
(0.000) (0.000)
rent_burden_hh_pct
-0.000796***
(0.000)
homeowner_pct
-0.000215***
(0.000)
no_vehicle_pct
-0.00187***
23
(0.000)
Constant 0.180*** 0.246*** 0.316***
Observations 132,389 132,389 132,257
VARIABLES Model 4 Model 5 Model 6 Model 7 Model 8 Model 9
Dummy
1999
0.195*** 0.195*** 0.195*** 0.195*** 0.195*** 0.195***
(0.008) (0.008) (0.008) (0.008) (0.008) (0.008)
Dummy
2000
0.144*** 0.144*** 0.144*** 0.144*** 0.144*** 0.144***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
Dummy
2001
0.0806*** 0.0806*** 0.0807*** 0.0807*** 0.0807*** 0.0808***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Dummy
2002
0.115*** 0.115*** 0.115*** 0.115*** 0.115*** 0.115***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Dummy
2003
0.185*** 0.185*** 0.185*** 0.185*** 0.185*** 0.185***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Dummy
2004
0.333*** 0.333*** 0.333*** 0.333*** 0.333*** 0.332***
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
Dummy
2005
0.490*** 0.490*** 0.489*** 0.489*** 0.489*** 0.489***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
Dummy
2006
0.511*** 0.511*** 0.511*** 0.511*** 0.511*** 0.511***
(0.006) (0.006) (0.006) (0.006) (0.006) (0.006)
Dummy
2007
0.462*** 0.462*** 0.462*** 0.462*** 0.461*** 0.461***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
Dummy
2008
0.163*** 0.162*** 0.161*** 0.161*** 0.161*** 0.161***
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
Dummy
2009
-
0.0279***
-
0.0286***
-
0.0294***
-
0.0296***
-0.0303*** -0.0306***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Dummy
2010
-
0.00748**
-
0.00797**
-
0.00871**
-
0.00882**
-
0.00928***
-
0.00984***
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
RSEI Toxicity
> 6k
0.00395*
(0.002)
RSEI Toxicity
> 7k
0.00142
(0.002)
RSEI Toxicity
> 8k
-0.00296
(0.002)
RSEI Toxicity
> 9k
-0.00447*
(0.003)
RSEI Toxicity
> 10k
-
0.00916***
(0.003)
RSEI Toxicity
> 15k
-0.0185***
(0.004)
Constant 0.178*** 0.179*** 0.181*** 0.181*** 0.182*** 0.182***
Observations 132,389 132,389 132,389 132,389 132,389 132,389
24
VARIABLES Model 10 Model 11 Model 12 Model 13 Model 14 Model 15
Dummy
1999
0.195*** 0.195*** 0.195*** 0.195*** 0.195*** 0.195***
(0.008) (0.008) (0.008) (0.008) (0.008) (0.008)
Dummy
2000
0.144*** 0.144*** 0.144*** 0.144*** 0.144*** 0.144***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
Dummy
2001
0.0807*** 0.0806*** 0.0806*** 0.0806*** 0.0806*** 0.0806***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Dummy
2002
0.115*** 0.115*** 0.115*** 0.115*** 0.115*** 0.115***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Dummy
2003
0.185*** 0.185*** 0.185*** 0.185*** 0.185*** 0.185***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Dummy
2004
0.333*** 0.333*** 0.333*** 0.333*** 0.333*** 0.333***
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
Dummy
2005
0.489*** 0.489*** 0.489*** 0.489*** 0.489*** 0.490***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
Dummy
2006
0.511*** 0.511*** 0.511*** 0.511*** 0.511*** 0.511***
(0.006) (0.006) (0.006) (0.006) (0.006) (0.006)
Dummy
2007
0.461*** 0.462*** 0.462*** 0.462*** 0.462*** 0.462***
(0.005) (0.005) (0.005) (0.005) (0.005) (0.005)
Dummy
2008
0.161*** 0.162*** 0.162*** 0.162*** 0.162*** 0.162***
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
Dummy
2009
-0.0297*** -0.0288*** -0.0289*** -0.0289*** -0.0289*** -0.0288***
(0.003) (0.003) (0.003) (0.003) (0.003) (0.003)
Dummy
2010
-0.00904** -0.00820** -0.00823** -0.00823** -0.00824** -0.00813**
(0.004) (0.004) (0.004) (0.004) (0.004) (0.004)
RSEI Toxicity
> 20k
-0.0159***
(0.005)
RSEI Toxicity
> 50k
0.0066
(0.014)
RSEI Toxicity
> 100k
0.00222
(0.022)
RSEI Toxicity
> 500k
-0.0204
(0.100)
RSEI Toxicity
> 1m
-0.0529
(0.117)
RSEI Toxicity
Above Mean
0.00287
(0.002)
Constant 0.181*** 0.180*** 0.180*** 0.180*** 0.180*** 0.179***
Observations 132,389 132,389 132,389 132,389 132,389 132,389
R-squared 0.201 0.201 0.201 0.201 0.201 0.201
25
VARIABLES Model 16 Model 17 Model 18
Dummy
1999
0.195*** 0.195*** 0.195***
(0.008) (0.008) (0.008)
Dummy
2000
0.144*** 0.145*** 0.145***
(0.005) (0.005) (0.005)
Dummy
2001
0.0810*** 0.0807*** 0.0807***
(0.003) (0.003) (0.003)
Dummy
2002
0.116*** 0.115*** 0.115***
(0.003) (0.003) (0.003)
Dummy
2003
0.183*** 0.185*** 0.185***
(0.003) (0.003) (0.003)
Dummy
2004
0.332*** 0.333*** 0.333***
(0.004) (0.004) (0.004)
Dummy
2005
0.487*** 0.489*** 0.489***
(0.005) (0.005) (0.005)
Dummy
2006
0.509*** 0.511*** 0.511***
(0.006) (0.006) (0.006)
Dummy
2007
0.459*** 0.462*** 0.462***
(0.005) (0.005) (0.005)
Dummy
2008
0.159*** 0.162*** 0.162***
(0.004) (0.004) (0.004)
Dummy
2009
-0.0317*** -0.0292*** -0.0292***
(0.003) (0.003) (0.003)
Dummy
2010
-0.0111*** -0.00861** -0.00861**
(0.004) (0.004) (0.004)
RSEI Toxicity
> 1 StdDev
-0.0486***
(0.005)
RSEI Toxicity
> 2 StdDev
-0.0424***
(0.011)
RSEI Toxicity
> 3 StdDev
-0.0424***
(0.011)
Constant 0.184*** 0.181*** 0.181***
Observations 132,389 132,389 132,389
1
2 Chapter 2: Park Access for the Transit-Dependent. Do
communities really have access?
2.1 Abstract
This paper introduces a new measure of park access that improves upon existing
measures to analyze and understand neighborhoods of environmental justice concern and their
access to recreational opportunities. Previous park equity research has relied on measures of
linear distance between populations and parks, park service areas, or regional approaches.
These past measures lack precision and do not consider the fact that most low-income people of
color are less likely to travel by car and more likely to be transit-dependent in urban settings.
This new model of park access relies on travel times via public transportation to deepen our
understanding of park access, carries our conversations of environmental justice from defining
the problem to identifying solutions, and opens up urban planning conversations on how we
might expand park equity planning to include new measures of success.
2.2 Introduction
The distribution of environmental hazards has long been the focus of environmental
justice research and activism. Over forty years of research reveals environmental burdens such
as toxic waste generation facilities, landfills, and Superfund sites. In addition, toxic air and
water are patterned across space in a way that disproportionately burdens low-income
communities of color. These spatial distributions can be attributed to institutionalized racism,
commoditization of land and other natural resources, power inequality, and inadequate
legislation. Generally, minority and low-income status are both strong predictors of exposure
to environmental harm, with race being a slightly stronger predictor. However, it is worth
noting that evidence of minority move-in to hazardous areas is not found to be a significant
2
contributor to these unequal land use patterns. With these patterns well-established, there is
tremendous work to be done to improve environmental quality and increase equity.
In this paper, I argue that research which goes beyond the classic environmental justice
studies of environmental burdens and includes investigation of environmental goods helps to
provide a more holistic understanding of environmental justice. To do this, I examine the
spatial distribution of parks and open space in Los Angeles and use a new method to measure
park accessibility for the transit-dependent. I show that parks are well-distributed throughout
Los Angeles County and raise additional questions about the distribution of environmental
goods in the region.
This paper contains three parts. Part one is a brief literature review of the existing
environmental justice literature in Los Angeles, investigations of park availability, and tools
that have been used to measure park access. Part two presents my use of the automated Remix
access tool (aRat) to measure access to parks using public transit. The application of aRat is an
innovative, computer-coding and open data approach to more precisely measure access. Part
three is a presentation and discussion of the results.
2.3 Literature
Scholars working in the environmental justice field have investigated exposures to
environmental hazards including toxic waste facilities, landfills, water and air contamination,
and Superfund sites. New, emerging literature investigates access to environmental goods and
desirable land uses including quality, affordable housing; transportation; jobs; healthy food; and
recreational opportunities, providing us with a more well-rounded perspective of
environmental justice which goes beyond equal protection from harm and moves towards
promotion of real equity. In other words, expanding the environmental justice literature to
3
consider environmental amenities in addition to environmental harm pushes us from an
equality perspective to an equity perspective. This includes an understanding that there may be
communities of people who need parks more than others — these communities might include
low-income neighborhoods, people under age 18 or over the age of 65, and people without
access to a car (Rigolon 2016; Boone et al., 2009; Talen, 2003; Wolch et al., 2005).
Additionally, people living in communities of environmental justice concern are
generally in denser, more urban environments and have limited access to private recreational
sites like backyards. Focus on park equity includes an understanding that low-income minority
people may have transportation limitations and limited access to private recreational
opportunities like backyards (Loukaitou-Sideris & Stieglitz, 2002). Because of their unique
setting in the urban environment, individuals living in these neighborhoods spend more time in
public spaces and are more dependent on public recreation facilities such as parks for physical
activity, social ties, and leisure (Rigolon 2016, Karsten, 2005; Loukaitou-Sideris & Stieglitz,
2002, Loukaitou-Sideris, 2009).
Regular physical activity has been shown to increase health outcomes – reducing the
likelihood of heart disease, diabetes, high blood pressure, colon cancer, anxiety and depression
and obesity (Bedimo-Rung et Al., 2005). The benefits of physical activity are increasingly
important as people become more likely to work sedentary jobs. As of 2010, just over one
quarter of the U.S. adult population reports having had no physical activity in the last month.
In California, reports are slightly better with only 22 percent of adults reporting no physical
activity in the last month (Centers for Disease Control and Prevention, 2020). For
disadvantaged populations, this rate could be higher (Brownson et Al., 2001). However,
evidence suggests park facilities offer opportunities for physical exercise and convenient access
to parks is positively associated with reported exercise frequency (Brownson et Al., 2001).
4
In addition to the physical health benefits associated with parks, additional benefits to
psychological health, social interactions, economic activity, and environmental quality should
be considered (Bedimo-Rung, et Al., 2005; Boone et al., 2009). And even where parks have been
unmaintained to the point they are considered blight, the existence of parks still presents
opportunities to create new community-service spaces (Boone et al., 2009). These multifaceted
benefits suggest parks’ importance in understanding and building equity our in communities.
2.3.1 Measurement tools
Previous studies of park proximity have relied on a variety of GIS methods to measure access
but many provide us with proxies of access based on linear distance or park service areas.
Accessibility can be defined in a number of ways but generally, accessibility is understood to be
the total number of destinations accessible within a given travel time (Levinson, 2015). This
measure and understanding of accessibility is simple, comprehensible, and comparable between
cities but has not been applied to investigations of access to parks. This paper provides a first of
its kind investigation of access to park using transit travel times, refining previous park access
work.
2.3.2 Distance-Based Methods
A quarter mile or ten minute walk has historically been conceived as the distance people are
willing to travel by foot to reach a park or recreational facility (Boone et al., 2009; Forsyth
2000; Nicholls 2001; Lindsey, Maraj, and Kuan 2001; The Trust for Public Land 2004; Wolch,
Wilson, and Fehrenbach 2005). This has become the standard by which cities and other public
agencies determine communities who do and do not have “adequate” access to parks and has
become an advocacy goal for nonprofits concerned with park access including National
Recreation and Parks Association and Trust for Public Land (Boone et al., 2009). Some
evidence however suggests that walkability might not be the ultimate indicator of physical
5
activity and access to parks (Adams, et Al., 2015). Instead, patterns of physical activity can be
predicted by a suite of built environment features including public transportation (Adams, et
Al., 2015). We can enhance our existing understanding of access to parks by foot with further
measures of access to parks by public transit to develop a more complete understanding of the
urban fabric and its relationship to opportunities for recreation.
2.3.3 Park Service Areas
Park service areas (PSAs), generated in GIS, have been used to map park locations and the local
residents they serve (Sister et al., 2009) which assigns every resident in a region to his/her
closest park. The number of people served by each PSA is then summed and used to estimate
“park pressure” or congestion; PSAs with higher congestion are then deemed to be
disadvantaged. Arguments against the PSA approach argue that residents might not always
visit their nearest park, However, there is good evidence the park proximity is a good
determinant of park usage (Giles-Cort and Donovan 2002).
PSAs assume that the nearest park is automatically accessible for nearby residents. For
people who rely on public transit or lack access to a car for other reasons, this might not
actually be the case. Children, often the target audience for many park facilities are too young.
This research offers a more accurate representation of park access based on actual transport
methods.
Travel time is more relevant to understanding accessibility than physical distance.
Transit-dependent populations do not have access to personal automobiles and instead rely on
buses, trains, light rail and walking to move around the city. The transit-dependent may
include people too young to drive (youth), people too old to drive, and people without the
6
financial means necessary for a car. Having nearby, transit accessible parks provides
recreational opportunities for these populations.
The goal of this analysis is to develop a web-based open data method for measuring park access
based on the actual transit network. This involves developing a process that serves three
primary research objectives:
1. Develop access isochrones based on the transit network.
2. Assess neighborhood populations and estimate transit-dependent population.
3. Estimate the quality of parks available to the transit-dependent within a 30-minute
transit access shed.
Performing this work using open data and repeatable computing processes will allow for
continued analysis of park access for the transit-dependent in a variety of locations even with
alterations and deletions to the transit network.
The second aim of this study is to test the quality of parks in relation to traditional
environmental justice hypotheses. This part of the study aims to understand if neighborhoods
of environmental justice concern suffer from restricted access to quality parks in their
neighborhoods. Park acreage and other measures of park quality including availability of
recreational equipment have been included in the data to estimate park quality. A 2007 study
by Boone et al. found that a higher proportion of African Americans had access to parks within
walking distance than whites, but whites had access to more park acreage. This study expands
upon this work by not only considering parks within walking distance, but parks within transit
sheds too. Further, through consideration of other park characteristics beyond acreage, this
study deepens our understanding of who has access to parks and to what degree those parks
provide an amenity.
7
2.4 Methods
2.4.1 Study area
Los Angeles County, the most populous county in the United States and one of the most
economically active areas, is used here for analysis. The County was selected not only for its
population size and economy, but also for its importance in the environmental justice literature.
A number of important environmental justice studies have been done in this region establishing
firm evidence of environmental discrimination against the County’s most vulnerable
populations (Pastor, Sadd & Hipp, 2001; Sadd, Pastor, Morello-Frosh, Scoggins & Jesdale,
2011; Morello-Frosh, Pastor, Porras & Sadd, 2002). The existing evidence of unequal
distribution of environmental harms warrants closer inquiry into distributions of
environmental goods in the same region.
Los Angeles is also one of the most ethnically diverse areas in the country with over
half of the population being black or Hispanic. A table showing selected demographic
characteristics of Los Angeles County compared to the United States a whole is presented
below. On average, LA County is more ethnically diverse and has a larger share of people living
under the poverty line.
8
Table 3 - Demographic Characteristics of Los Angeles County and the United States
Table 3.
Demographic Characteristics of Los Angeles County and the United States
Los Angeles
County
United
States
Black population - The percentage of the population who
identifies as Non-Hispanic Black
7.91% 12.29%
Hispanic population - The percentage of the population (of any
race) who identifies as Hispanic
48.42% 17.60%
Percent of population under 18 22.50% 22.90%
Percent of population over 65 12.50% 14.90%
Median household income $61,015 $57,652
Percentage of households under the poverty line 17.00% 14.60%
Percentage of people under age 18 and under the poverty line 24.00% 20.30%
Percentage of people over age 65 and under the poverty line 13.40% 9.30%
Source: United states Census Bureau American Community Survey, 2017 5-Year Estimates
In 2019 the City of Los Angeles was ranked 55 out of 100 as part by ParkScore, a park
ranking system developed by The Trust for Public Land which includes data on park access,
park acreage, investment, and park amenities. LA’s score has been steadily climbing over the
past few years; in 2017 LA was ranked 74
th
and in 2018, 66
th
. The City has increased its
ranking as new park facilities have opened up and more people are now living within a ten
minute walk to a local park. While this is overall good news for the City, Los Angeles still lags
behind neighboring Southern Californian cities on the list including Irvine (5), San Diego (16),
and Long Beach (18). Additionally, the ParkScore for Los Angeles only includes parks within
Los Angeles City – it does not consider the County as a whole.
Countywide, there have been some efforts to identify places that need additional parks
and to better connect people to the parks that do exist. In 2016, the Los Angeles County
Department of Parks and Recreation completed a countywide Parks Need Assessment which
included a new series of metrics to determine park need. These new metrics were designed to
9
guide future, needs-based allocation of funding for parks and recreation facilities. The Parks
Needs Assessment found that 80% of the existing parks should be considered “high pressure
parks” with less than 3.3 acres of park area available per one thousand residents (Metro, 2019).
The national average is 6.9 acres per one thousand residents. Additionally, 49% of the
countywide population was determined to live more than one half mile from a park. The
neighborhoods identified as high need areas were clustered mainly in Central, South, and
Southeast Los Angeles; these are the same places that have previously been identified as places
of environmental justice concern.
Dovetailing on the Department of Recreation and Parks analysis, Los Angeles Metro
(Metro), the county’s public transit agency, has identified transit access to parks as a strategic
goal for the agency. In May 2019, Metro released a strategic plan document where, “Metro
envision a Los Angeles in which people can use a network of varied and connected
transportation options to get to parks and open spaces” (Metro, 2019). Within this document,
Metro recognized the need for an equity-based approach to park access where vast disparities
between communities are recognized and solutions are developed accordingly. For the facilities
that do exist, lack of public transportation access to parks was found to be a top barrier to
accessing the public parks and open spaces (Metro, 2019). The Metro strategic plan used a case
study approach to identify and highlight transit-to-parks access solutions with an emphasis on
some of Los Angeles’ largest parks. My hope is that this paper supports the ongoing regional
efforts to enhance public transit access to parks by providing comprehensive, countywide data
on the existing park accessibility landscape for communities with and without environmental
justice concern and provide data-driven momentum towards equitable park access.
10
2.4.2 Data
The Los Angeles County Department of Recreation and Parks maintains a database of
the parks and open spaces located throughout the County. This database is the same one that
was used in the 2016 Park Needs Assessment. The parks database was downloaded from the
Los Angeles County GIS data portal and geocoded using geospatial software. The data includes
2,827 unique facilities which range in size from 0.004 to 642,078 acres and each of the parks are
classified into four access types with 55.8% classified as local parks; 22.3% regional open spaces,
21.2% natural areas; and 0.7% regional recreation parks.
Park data were merged to 2010 United States Census Bureau census tract geographies
for integration with sociodemographic information. The American Community Survey (ACS)
2017 5-year estimates data were then used for sociodemographic characteristics of each tract
including racial composition; household income and poverty estimates; and estimates of the
transit-dependent population in each tract — the number of people below age 18 or above age
65.
Transit system data were taken from General Transit Feed Specification (GTFS) data
from each transit provider in the Los Angeles region. GTFS is a data structure developed in
2005 to make it easier for public transit providers to maintain and distribute data on their
systems. The uniform structure of GTFS data makes it easy to aggregate transit data across
different agencies, making large-scale analysis of public transit provision easier and less error
prone. The mapped transit system for Los Angeles County included August 2018 GTFS data
from each of the agencies listed below. The total number of lines provided by each agency are
shown in parentheses. A map showing all park locations overlain with the Los Angeles County
transit system is presented in Figure 1.
11
Table 4 - Los Angeles County Transit Providers
Table 4. Los Angeles County Transit Providers
Antelope Valley Transit Authority
(20)
LA Metro Bus (139) Orange County Transit Authority
(82)
Beeline (13) LA Metro Rail (6) Pasadena Transit (10)
Carson Circuit (8) LADOT (45) Santa Clarita Transit (42)
City of Commerce Municipal Bus
Lines (11)
Long Beach Transit (35) Santa Monica Big Blue Bus (20)
Culver City Bus (8) Metrolink (7) Torrance Transit (11)
Foothill Transit (37) Norwalk Transit System (6) UCLA (8)
GTrans (5)
Figure 6 – Parks and Public Transit in Los Angeles County
12
2.4.3 Measuring access to parks
I used the automated Remix access tool (aRat) developed by Swayne and Kundaliya
(2020), to measure transit access to parks for each census tract in the County. aRat is a Python
application which interfaces with Remix to automate rapid downloads of transit access
isochrones for thousands of origin points.
Remix is a private, online transit planning tool developed to make transit scenario
planning easier and more accessible for transit planners. Remix’s online platform aggregates
and plots GTFS data on the existing transit system. Remix users can then make alterations to
the existing transit system through point-and-click operations. Possible transit system
alterations including adding a new line, deleting existing lines, changing rights of way, and
making schedule changes. In addition to these point-and-click operations, Remix can also
generate transit access isochrones – calculating the total area a passenger can travel using
transit in 15, 30, 45, or 60-minutes. While these isochrone calculations in Remix are quick,
users can only visualize and download one isochrone at a time.
aRat makes it possible understand access via transit from thousands of census tract
origin points. Through Python code, aRat automatically moves the isochrone origin point in
Remix to the centroid of every census tract, eliminating the need for onscreen operations. As
aRat moves the isochrone origin point across the map, it downloads the resultant isochrone
shapefile to the user’s PC for further analysis. It is important to download isochrones for every
census tract as the underlying public transit availability can dramatically influence the
isochrone shapes and sizes. Figure X below shows access isochrones for two points in Los
Angeles County. The isochrone in the north was generated for a census tract with very public
transit and the total accessible area is limited. The isochrone to the south with an origin point
in Downtown Los Angeles is much bigger since public transit is denser in that area.
13
Figure 7 – Representative Isochrones
Remix makes a series of assumptions during isochrone generation. Remix routes its
passengers over the fastest combination of bus, rail and walking. The rider will board the first
bus or train according to the published schedule and if the rider needs to make a transfer, they
are assumed to wait half of the headway time before boarding their next bus or train. If the
rider disembarks from a bus or train and continues walking until their commute time runs out,
they will walk at a pace of three miles per hour. Since each of the isochrone calculations in
Remix rely on published transit schedules, the Remix user must define the time of day for
isochrone generation. All access calculations in this paper were modeled using a peak weekday
travel time of 5:00 pm, representing a best-case transit scenario.
14
After isochrone download is complete, aRat clips the total isochrone area to
underlying census tract geographies and other destination data. While aRat was originally
developed to calculate access to jobs, the tool allows for users to input shapefiles with other
destination data of interest— examples include parks, libraries, financial institutions and
grocery stores. In this case, the Los Angeles County parks database was used. After clipping
isochrones to destination data, aRat outputs a tabular data file with the number and names of
parks accessible from each census tract in 15, 30, 45, or 60 minutes using transit.
In contrast to the park service area approach to calculating park access, aRat allows for
consideration of multiple accessible parks for each census tract. The PSA approach assigns each
resident to his/her nearest park while aRat tallies up all accessible parks within a given travel
time. While there may be evidence that people living in close proximity to a park have better
recreational habits and outcomes, aRat does not impose park destination decisions on residents,
instead letting the public transit system determine which park facilities are accessible.
2.5 Results
Table 5 - Count of tracts with total number of parks accessible
Table 5.
Count of tracts with total number of parks accessible
Number of parks
accessible within 15
minutes
Number of parks accessible
within 30 minutes
Freq. Percent Freq. Percent
0 237 10.23 0 - -
1 462 19.95 1 25 1.08
2 471 20.34 2 43 1.86
3 350 15.11 3 45 1.94
4 229 9.89 4 62 2.68
5 171 7.38 5 80 3.45
>5 396 17.08 >5 2061 88.95
Total 2,316 100 Total 2,316 100
15
Figures 8 and 9 displays counts of tracts according to the number of parks accessible within 15
or 30 minutes. Analysis of tract demographic data according to park access validates previous
studies which have found that lower-income and minority populations tend to have slightly
better access to parks. These findings from previous papers which relied on linear distance and
park service areas hold when travel time via transit is instead used as the access measure. The
map below shows each census tract colored according to the number of parks accessible from
that tract in 15 minutes via transit. Generally, tracts with access to more parks are located in
central Los Angeles with some pockets of good access in the communities of Santa Monica,
Long Beach, and the coastal neighborhoods in between. Park access in the northeast portion of
the county in neighborhoods near Santa Clarita and Palmdale is limited. These communities
have access to zero or one park – Angeles National Forest.
16
Figure 8 – Access to Parks Using Transit – 15 minutes
17
Table 6 - Census tracts descriptives t-test for equality of means
Table 6.
Census tract descriptives t-test for equality of means
0 parks
accessible -
15 minutes
1+ parks
accessible -
15 minutes
Variable Mean Mean Difference
Median household income $ 72,699 $ 65,159 $ 7,540
***
Percent of population under 18 21.0 22.0 -1.1 **
Percent of population over 65 13.1 12.7 0.4
% white 52.9 51.5 1.4
% Hispanic 42.2 48.4 -6.2 ***
% black 6.8 8.1 -1.3
N tracts 237 2,077
*** p<0.01, ** p<0.05, * p<0.1
T-tests comparing tracts with access to zero parks in 15 minutes to tracts with access to
one or more parks reveal significant differences in the tracts by median household income, the
percent of the population under 18, and the percent of the population who are Hispanic. Tracts
with access to zero parks have higher median incomes, fewer children, and fewer Hispanic
people.
Looking at park accessibility for 30 minutes of transit travel time, demographic
differences generally hold but differences in median household income lose significance. Since
every tract in Los Angeles County had access to at least one park after 30 minutes of travel
time via transit, tracts with access to one park are compared to tracts with access to two or
more parks. Tracts with access to two or more parks had larger populations under 18, fewer
white people, and more Hispanic people. These differences may be explained by the spatial
location of predominantly white neighborhoods (greater than 50% white) further away from the
central city in suburban environments (Figure 8).
18
Table 7 - Census tract descriptives t-test for equality of means
Table 7.
Census tract descriptives t-test for equality of Means
1 park
accessible -
30 minutes
2+ parks
accessible -
30 minutes
Variable Mean Mean Difference
Median household income $ 76,245 $ 65,795 $ 10,450
Percent of population under 18 19.6 22.0 -2.4 *
Percent of population over 65 13.4 12.8 0.6
% white 67.9 51.5 16.4 ***
% Hispanic 30.1 47.9 -17.9 ***
% black 7.0 8.0 -1.0
N tracts 25 2288
*** p<0.01, ** p<0.05, * p<0.1
19
Figure 9 - Access to Parks Using Transit – 30 minutes
Regression results display the relationships between the number of parks accessible using
transit and census tract characteristics. In Model 1, 15 minutes is used as the travel time and in
Model 2, 30 minutes is used. For park catchment areas of 15 minutes, the proportion of
population under 18 and over 65 is inversely related to the number of parks accessible but the
magnitude of this relationship is quite small. Statistically significant relationships are also
found for the percent of white residents, black residents, percentage of people living below the
poverty line, and high school completion rates but the coefficients are less than .1 park per
percentage point. At 30 minutes, the associations are magnified.
20
Table 8 - Linear Regression Results of Park Access
(Model 1) (Model 2)
VARIABLES Number of Parks
Accessible –
15 minutes
Number of Parks
Accessible –
30 minutes
% Under 18 -0.0736*** -0.658***
(0.0147) (0.0708)
% Over 65 -0.00906 -0.334***
(0.0157) (0.0758)
% White 0.0101** -0.0588***
(0.00416) (0.0200)
% Hispanic -0.00163 -0.0258
(0.00418) (0.0202)
% Black 0.0147** 0.102***
(0.00686) (0.0331)
Median household income -6.55e-06 -0.000148***
(4.23e-06) (2.04e-05)
% Poverty 0.0662*** 0.437***
(0.0101) (0.0486)
% High school graduate -0.0688*** -0.612***
(0.00746) (0.0360)
Constant 7.129*** 76.65***
(0.697) (3.363)
Observations 2,283 2,283
R-squared 0.119 0.344
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
These results suggest that neighborhoods of environmental justice concern are located in
places where residents can more easily access parks via transit. With this evidence that low-
income, communities of color may have better access to parks via transit compared to other
neighborhoods, it is important to remember that not all parks are created equal.
To proxy for park quality, I relied on the Parks Needs Assessment data on existing park
amenities. The Needs Assessment included counts of sixteen important park facilities. Park
facilities include tennis courts, basketball courts, baseball fields, soccer fields, multipurpose
21
fields, fitness equipment, skate parks, picnic areas, playgrounds, pools, splash pads, dog parks,
gymnasiums, community/recreation centers, senior centers, and restrooms (County Recreation
& Parks, 2016). Needs Assessment data include qualitative ratings of the facilities at each park.
Qualitative rankings include counts of facilities in good, fair, or poor condition.
I compared one hundred parks in the parks needs assessment to online, aerial
photographs of parks. In many cases, specific facilities were over or under counted. For
example, the data file may have indicated six tennis courts but when I examined the park via
aerial photographs, only four courts were visible. In no cases did I find that an amenity was
counted that simply did not exist. Due to these data reliability issues, I collapsed the Needs
Assessment data on park amenities into a simpler amenity index, much like the QUINPY
developed by Rigolon and Németh (2016). The park index used in this paper collapses the
Needs Assessment data into categorical groups in order to reduce the uncertainty associated
with the original data.
Table 9 - Park Quality Index
Table 9.
Park Quality Index
Original Variables New Variable Scoring Description
Playground - GOOD
Playground
0: no playground
Playground - FAIR 1: 1 playground
Playground - POOR 2: 2 or more playgrounds
Pools - GOOD
Pools
0: no pool
Pools - FAIR 1: 1 or more pools
Pools - POOR
Soccer - GOOD
Sports Field
0: no sports field
Soccer - FAIR 1: 1 sports field
Soccer - POOR 2: 2 or more sports fields
Baseball - GOOD
Baseball - FAIR
Baseball - POOR
Multipurpose Field - GOOD
Multipurpose Field - FAIR
22
Multipurpose Field - POOR
Basketball - GOOD
Sport Courts
0: no sport courts
Basketball - FAIR 1: 1 sport court
Basketball - POOR 2: 2 or more sport courts
Tennis - GOOD
Tennis - FAIR
Tennis - POOR
Skate Park - GOOD
Skate Park - FAIR
Skate Park - POOR
Gymnasium - GOOD
Gymnasium - FAIR
Gymnasium - POOR
Restroom - GOOD
Restroom
0: no restroom
Restroom - FAIR 1: 1 or more restrooms
Restroom - POOR
Dog Park - GOOD
Dog Park
0: no dog parks
Dog Park - FAIR 1: 1 or more dog parks
Dog Park - POOR
Senior Center - GOOD
Community
Senior Centers
0: no community or senior centers
Senior Center - FAIR 1: 1 or more community or senior centers
Senior Center - POOR
Community Center - GOOD
Community Center - FAIR
Community Center - POOR
Park Size Park Size 1: park in the lowest quartile by size
2: park in the second quartile by size
3: park in the third quartile by size
4: park in the highest quartile by size
This park index system provides us with a proxy for park quality where each park can score up
to 14 points. Griffith Park is the only park in the dataset to earn a perfect score and the mean
score for all parks in the set is four. The distribution of parks by quality score is shown below.
Table 10 - Parks by Quality Score
Table 10 .
Parks by quality score
Score No. of Parks Percent
0 12 0.42
23
1 424 15
2 450 15.92
3 390 13.8
4 675 23.88
5 132 4.67
6 144 5.09
7 126 4.46
8 114 4.03
9 111 3.93
10 103 3.64
11 91 3.22
12 40 1.41
13 14 0.5
14 1 0.04
Total 2,827 100
Park quality scores for all parks accessible from each tract were summed and used in regression
analysis. Results are limited but suggest for tracts with higher proportions of people under age
18 or over age 65, overall quality of parks accessible within 30 minutes is lower. Tracts with
more white and black residents have access to more quality parks and tracts with higher
proportions of people living in poverty are subject to lesser park quality.
Table 11 - Linear Regression Results of Park Quality
(Model 3) (Model 4)
VARIABLES Total Park Quality
– 15 minutes
Total Park Quality
– 30 minutes
% Under 18 -0.0220 -0.0358***
(0.0178) (0.00898)
% Over 65 -0.00252 -0.0418***
(0.0191) (0.00962)
% White -0.00121 0.00571**
(0.00505) (0.00254)
% Hispanic 0.0145*** -0.00535**
(0.00508) (0.00256)
% Black 0.0167** 0.0145***
(0.00834) (0.00420)
Median household income -2.04e-05*** -3.50e-05***
24
(5.14e-06) (2.59e-06)
% Poverty -0.00359 -0.0275***
(0.0122) (0.00616)
% High school graduate -0.0610*** -0.0492***
(0.00907) (0.00457)
Constant 11.51*** 17.35***
(0.847) (0.426)
Observations 2,283 2,283
R-squared 0.054 0.175
After analyzing for park quality, limited environmental justice claims can be made. Tracts with high
incidence of poverty and minority racial status have access to lower quality parks. These park quality
findings track well with other studies which find that white neighborhoods are co-located with more
facilities (Rigolon, 2017; Sister et al., 2010).
Conclusion
This paper contributes to our understanding of park access by providing a new measure of park
accessibility for transit-dependent populations. I build on existing methods for measuring
access to more closely approximate how people experience physical proximity to parks, with
broader applications to environmental hazards and benefits.
By applying this tool to park access, as well as measuring park quality in a more nuanced way, I
show that park access in Los Angeles is a more complex story than the traditional
environmental justice narrative, in which low-income communities of color disproportionately
live in neighborhoods with concentrated environmental harms and lower-quality
environmental amenities. Rather, I show that neighborhoods of environmental justice concern
experience denser opportunities to access parks via public transit. However, these parks are
generally lower-quality than those that exist in wealthier neighborhoods. Therefore, while
communities that disproportionately experience environmental justice issues—those with a
25
greater concentration of environmental harms—the distribution and nature of environmental
benefits differs. Therefore, these neighborhoods may have superior access to a park—which is
good, because the land is already dedicated to that purpose—a park in a low-income
neighborhood qualitatively differs from its counterpart in a wealthier neighborhood.
In Los Angeles County, these results suggest more work needs to be done to enhance
the quality of existing parks in low-income communities. This is a low-hanging policy solution.
Positively, local parks with higher quality amenities may be possible with some investment in
existing public spaces. Transit may be one option Los Angeles has to make up for current gaps
in park provision. It takes time and money to develop new parks – especially if land acquisition
is required. In the meantime, transit may be able to help provide access to existing public lands.
Future work could leverage online, crowd-sourced data on park quality from platforms such as
Yelp and Google reviews. These online review platforms could give important information on
perceived park quality with opportunities for enhancements.
26
3 Chapter 3: Access to Environmental Law: Incidence and
distribution of California Environmental Quality Act (CEQA)
Environmental Impact Reviews and CEQA litigation
3.1 Introduction:
The California Environmental Quality Act (CEQA) was signed into law fifty years ago
in response to mounting fears about water quality, smog, and other environmental quality
concerns in California. The law requires documentation of the potential environmental impacts
associated with development projects, opening development up to public opinion and litigation.
Since the law was passed, environmental quality has improved in California – this is a win.
Simultaneously, California has become synonymous with soaring rents, slow residential
construction, and a growing housing affordability crisis. CEQA is repeatedly blamed for these
issues. With strong environmental wins and potential affordability losses, there remains no
clear path forward for the future of CEQA. Anecdotal evidence of CEQA’s impacts on
development abounds but systematic data on CEQA review and CEQA litigation is lacking.
The tension between environmental protection and development remains unresolved reflecting
a broader planning dilemma between environmental regulation and development.
In this paper, I build a primary dataset of CEQA environmental impact reports and
associated litigation. With this dataset, I can uniquely trace how CEQA is interacting with the
development process. This systematic evaluation of CEQA, California’s most important
environmental law, had yet to be done. I fill this gap and broaden our understanding of the
actual extent of CEQA litigation. This work is essential to forming data-driven discourse on
CEQA and its broader effects on urban development.
In this paper, I build and analyze a dataset comprised of five years (2010-2015) of CEQA
environmental impact reviews, CEQA litigation, and development projects in California. I
27
include data on all EIRs and CEQA petitions filed in the state, matching petitions to the
specific EIRs challenged, to deepen our understanding of CEQA litigation’s role on various
project types and project locations as well as project timing. With this dataset, I address the
following questions about CEQA litigation and its role in development and communities across
the state.
1. What is the nature and extent of CEQA litigation?
2. How does CEQA litigation delay time to development?
3. What is the spatial distribution of CEQA development and litigation?
I find no overwhelming evidence of the CEQA EIR process or CEQA litigation delaying
residential development. I instead find evidence that CEQA litigation is more often used on
mixed use and public infrastructure projects – projects that may require additional
discretionary review beyond what is provided by a city’s General Plan in California. I do find
that litigation can substantially alter development timelines and suggest future pathways for
research investigating ways to reduce CEQA litigation uncertainty.
3.2 Background:
3.2.1 What is CEQA
The California Environmental Quality Act (CEQA) is a state statute enacted in September
1970. CEQA came just months after President Nixon passed the National Environmental
Protection Act (NEPA). CEQA was designed to be a more rigid version of the national law.
Both CEQA and NEPA require the incorporation of environmental values into government
decision making on development and require disclosure of the potential environmental impacts
of these decisions (OPR). In the case of CEQA, the potential environmental impacts of a project
should be mitigated to the extent feasible. However, the statute recognizes that mitigation is
not always possible and it is therefore not always a prerequisite for project approval. However,
28
informing the public of the potential environmental risks associated with a project is a
prerequisite for project approval.
The fundamental purposes of CEQA as outlined in Title 14 of California’s Code of Regulations,
section 15000 et seq. are:
1. “Inform governmental decision-makers and the public about the potential, significant
environmental effects of proposed activities.”
2. “Identify the ways that environmental damage can be avoided or significantly reduced.”
3. “Prevent significant, avoidable damage to the environment by requiring changes in
projects through the use of alternatives or mitigation measures when the governmental
agency finds the changes to be feasible.”
4. Disclose to the public the reasons why a governmental agency approved the project in
the manner the agency chose if significant environmental effects are involved.”
CEQA’s jurisdiction over development in the state is limited. CEQA only regulates
development requiring governmental discretionary approval on whether of how to carry out a
project (Guidelines § 15337).
Under CEQA guidelines, discretionary projects must be reviewed for environmental
impacts. This environmental review can be achieved in one of three ways, depending on the
expected severity of impacts. An Initial Study is prepared for every discretionary project. The
initial study is a preliminary analysis of the environmental effects, and pending results of it, a
negative declaration, mitigated negative declaration, or environmental impact report (EIR) may
be recommended. Negative declarations are the least time-intensive CEQA review. These
documents apply to projects where there is no substantial evidence of significant effects on the
environment (Guidelines § 21080). An EIR must be prepared when initial study reveals
evidence of significant, un-mitigatable effects (Guidelines § 21080). In other words, EIRs
are only prepared for CEQA projects expected to have significant impacts on the environment.
Many projects do not require an EIR, but we do not have systematic data on the percentage of
29
projects which skip the EIR process. A 2018 paper (O’Neill et al.) estimated CEQA’s interaction
with residential development projects of over five units over three years in five Bay Area,
California cities. They found that all housing projects of over five units were required to
undergo discretionary review, but city planning departments mandated very few projects to
complete an Environmental Impact Review. The housing projects that were subject to EIR
tended to be larger than exempt projects, and compliance routes other than EIRs were used
more often than not to approve residential developments and infill (O’Neill et al., 2018).
The primary purpose of EIR documents is to, “inform decision makers and the public
about a project’s significant environmental effects and ways to reduce them, to demonstrate to
the public that the environment is being protected, and to ensure political accountability by
disclosing to citizens the environmental values held by their elected and appointed officials”
(Bass et al., 2012). EIRs are tedious and include multiple public notices and noticing periods
alongside numerous drafts of the EIR documents. The complexity of the EIR process requires
technical expertise from environmental planners, consulting firms, scientific specialists, and
lawyers, making the undertaking of an EIR cost-prohibitive for smaller development projects.
Developers are very motivated to avoid EIRs, and are often amenable to modifying their plans
to prevent the EIR process by instead agreeing to environmental mitigation measures by way
of a mitigated negative declaration (Olshansky, 1996).
The CEQA statute creates an investigative reporting process for environmental harm; it
does not specify acceptable thresholds of environmental harm. Localities are left to make
decisions regarding what levels and types of harm are “acceptable” in light of environmental,
economic, and social considerations. At its core, CEQA is a right-to-know provision ensuring
that public officials and citizens have access to information on development and its impacts.
30
Faced with the facts of environmental impacts associated with CEQA projects, the public has
opportunities to express their opinions on a project. The public has access to information about
all CEQA projects. It has the opportunity to contest a project’s approval, no matter which
CEQA document is prepared (negative declaration, mitigated negative declaration, or
environmental impact review). Projects with an EIR, however, have the added requirement of
responding to public comment. The Final Environmental Impact Review (FEIR) must also
incorporate these comments. While comments must be responded to, changes to the project in
response to every comment are not mandatory. A lead agency can approve a project even when
developers do not modify their plan in response to all EIR comments. Projects with an EIR
include the most opportunity for interactions with the public (Barbour & Teitz, 2005). The
general CEQA is below.
Figure 10 - Overview of CEQA Process
The public can also challenge a CEQA project through litigation. While litigation does
not automatically stop a project from going forward, project developers often hesitate to move
forward during the litigation process due to the legal and financial risks involved. Litigation is
often the most powerful way for communities to press their demands on project proponents.
31
3.2.2 CEQA Litigation
While environmentalists often tout CEQA as a powerful tool for environmental protection in
California, which has contributed to the natural beauty that the state has been able to maintain,
business and developers view it differently due to the risks litigation poses to project success.
Environmental groups rely on CEQA to “check” business, open the development process to
public review, and to comment on all development. Environmentalists maintain that litigation
is reserved for the most environmentally concerning projects. However, CEQA opponents,
including project developers, claim that they are slowed by CEQA and made vulnerable to
costly and unnecessary lawsuits. Indeed, CEQA has since been blamed for rising home prices,
increased development uncertainty, and slower development times in California. CEQA
opponents argue the policy deters much-needed development in the state by allowing project
opponents broad grounds under which to dispute a project.
Litigation under CEQA is subject to unique rules which set it apart from most other
policies. Even though CEQA is often thought of as an environmental protection law, CEQA
statute does not require petitioners to file for purely environmental reasons. There have been
documented cases of labor unions and other business groups filing CEQA petitions to protect
personal business interests (CITE). Anonymous filing to protect litigant identity is also
allowed. This anonymity can serve two purposes – to protect vulnerable groups, or to provide
cover for nefarious groups – but this is incredibly hard to track in the data. Sometimes CEQA
petitions are brought by environmental groups founded for the sole purpose of bringing a
petition forward. Again, this can bring collective bargaining power to local communities, or it
can provide a guise. CEQA’s unique rules surrounding litigation give communities a rare lever
to influence development, and protections for petitioners are essential for this reason.
32
The extent to which CEQA litigation occurs has been a subject of tense debate.
Olshansky (1996) suggests that CEQA litigation may be less frequent than opponents contend.
His study found a mean of 1.01 CEQA-lawsuits being filed per jurisdictional area from 1985 to
1990 through a survey of California planners. In his survey, litigation occurred on one of every
354 CEQA documents. Of all California planner survey respondents, 80.1% responded that they
had one or zero CEQA-lawsuits in their jurisdiction during the study period. However, CEQA
critics still express significant concern for CEQA litigation abuse, and other studies hold that
litigation under CEQA is both more frequent and problematic than Olshansky implies (Holland
& Knight, 2015; Barbour & Teitz, 2005). California attorneys have investigated CEQA
litigation incidence continue to disagree on its extent. Jennifer Hernandez, an attorney and
advocate for CEQA reform claims this environmental law is “being hijacked to advance
economic interests” (Bradford, 2018). Berkeley attorney and researcher Eric Biber believes
CEQA “gets more attention than it deserves” and is “not a cause of the underlying challenges
we face in producing more housing in urban areas in California” (Bradford, 2018).
Barbour and Teitz (2005) discuss how CEQA litigation offers a “double-edged sword.”
The law was designed to encourage the public to take a more democratic role in the
development process, and it may be better than more dogmatic, technical rules to limit
development. However, since it is a self-enforcing statute, enforcement often occurs through
court challenges, which are expensive and time-consuming. Developers, to protect themselves
from being sued, have taken to creating extremely dense, sometimes incomprehensible EIRs.
Extreme documentation of potential impacts and scientific studies are thought to “bullet-proof”
the document against litigation but ironically makes it more difficult for citizens to provide
meaningful input to the project (Barbour and Teitz 2005). Even more ironically, Holland &
Knight find that EIRs are wrapped up in litigation more often than other CEQA planning
33
documents and only 43% of EIRs survive in court (2015). This suggests that EIRs may not be
the way for developers to ensure the success of their project
Furthermore, some evidence illustrates how project opponents have abused the law to
advance their individual interests, sometimes producing environmentally inferior outcomes.
Lawsuits can be filed anonymously, which further enables project opposition unrelated to
environmental concerns to proceed under CEQA. While this makes plaintiff data hard to track,
Holland and Knight (2015) suggest that between 2010-2015, 45 percent of CEQA litigation (or
over 250 cases) was instigated by “some organization or association” (a separate classification
from “established environmental advocacy organization”) including business competitors;
regulated party practitioners including regulated industries; construction trade unions; non-
construction trade unions; NIMBYs; and “greenmail” and “bounty hunter lawyers” (Holland
and Knight, 2015). Since EIRs are the most likely form of CEQA documentation to get
wrapped up in litigation, avoiding them can protect a project against costly court-related
delays. However, this is antithetical to CEQA’s core principles of well-reasoned, participatory
environmental planning. Thus, while CEQA was designed to balance principles of
environmental, social, and economic values in development projects, existing litigation data
suggests that this original intent is getting lost in practice (Hollland & Knight, 2015; Barbour
and Teitz 2005).
Access to CEQA
This paper presents a first of its kind database on all CEQA EIRs and subsequent legal
petitions filed in California from 2010 to 2015. The CEQA litigation data used in this paper
provides us with the most complete understanding of CEQA project locations, where litigation
is taking place, the types of projects most often sued, and the incremental time delay to projects
34
subject to dispute. These data allow us to deepen our understanding of which communities have
access to this crucial piece of environmental legislation and open up new possibilities for data
driven CEQA reform.
3.3 Data:
Table 12- EIRs by Year
I obtained Environmental Impact Report data from CEQAnet, an
online searchable database of all CEQA documents submitted to the
State Clearinghouse for review. CEQAnet includes summaries of all
documents including EIR project titles, project locations, lead agency
names, contact information, and project descriptions. For this project,
I downloaded all EIR documents filed between 1 January 2010 and 31
December 2015 (N = 2,349). To focus on CEQA litigation and its
interaction with specific development projects, I removed EIRs filed in reference to General
Plan or Specific Plan for a final database of 1,964 EIR documents.
Using EIR document texts and web articles online, additional information was collected
on each EIR. EIRs were hand-coded by project type into one of eleven categories. In cases
where a project did not neatly fit into a single project category, it was coded as mixed use
(mixed use projects with house were kept separate). The distribution of EIRs by project type
(Figure 1) and the project type coding schema (Table 2) are below.
Table 12.
EIRs by Year including
General/Specific Plans
Filing Year Total EIRs Percent
2010 464 19.75
2011 395 16.82
2012 373 15.88
2013 348 14.81
2014 406 17.28
2015 363 15.45
Total 2,349 100
35
Figure 11 - EIRs by Project Type
Table 13 - EIR Cording Schema
Table 2.
EIR Coding Schema
1. Mixed Use with Residential Mixed use projects may include housing
2. Residential Strictly residential projects with no commercial
components. Senior housing included.
3. Commercial Commercial development includes shopping, office,
church/religious uses
4. Mining/Extraction/Oil Energy Mining and extraction of energy sources. Includes
renewable energy developments.
5. Auto Transportation Auto transportation infrastructure includes bridge
projects, freeways, and roads.
6. Public Services and Infrastructure Schools, and other educational facilities, hospitals,
public buildings, convention centers, city/community
centers, fire/police stations. This category also
includes public utility upgrades.
7. Park/Recreation & Environment Parks, recreation facilities, cemeteries, and
environmental remediation projects.
8. Public Transportation Public transportation facilities for rail, bike, and bus.
Airport facilities.
20%
12%
12%
10%
5%
20%
7%
4%
2%
6%
2%
EIRs by Project Type
Mixed Use with Residential
Residential
Commercial
Mining/Extraction/Oil Energy
Auto Transportation
Public Services and Infrastructure
Park/Recreation & Environment
Public Transportation
Agriculture
Noxious Uses
Mixed Use
36
9. Agriculture Agricultural projects including vineyards and ranching
activities.
10. Noxious Uses Development of landfills, recycling centers,
composting centers, logging, industrial manufacturing
plants, and correctional facilities.
11.
Mixed Use Mixed use projects which do not include housing
San Francisco-based law firm, Holland & Knight provided data on CEQA litigation. The
law firm filed two Freedom of Information Act (FOIA) requests to the California Attorney
General’s office. The first FOIA request included all CEQA petitions filed from 2010 to 2012.
The second request included CEQA petitions for the years 2013 through 2015. I combined
these two petition databases and removed CEQA petitions filed against CEQA documents
other than EIRs. The remaining petition database included the case name, party names, county
court with jurisdiction, petition filing date, and a few other pieces of identifying information for
each of the cases filed against EIRs (N = 565).
Table 14 - CEQA Petition Database Summary
It is possible not all CEQA petitions are in this dataset. The petition records were
collected on behalf of Holland & Knight by the California Attorney General. Since the petitions
are not all stored digitally, some may have been left out of the FOIA request due to
inconsistencies in legal document storage practices.
Matching EIRs and Petitions
Table 14.
CEQA Petition Database Summary
CEQA petitions filed from 2010-2012 542
Petitions filed against an EIR 256
CEQA petitions filed from 2013-2015 613
Petitions filed against an EIR 309
37
Unfortunately, CEQA petitioners are not required to identify individuals involved by
name, nor is it required to name the Environmental Impact Report (EIR) that is the subject of
this petition. CEQA allows for the anonymous filing of suits against projects. While this
provision was designed to protect vulnerable individuals and communities, it also allows
individuals to band together and form temporary “organizations” to file a CEQA suit. This
makes it challenging to track who is bringing lawsuits against projects. We had to hand match
every CEQA petition with its associated EIR through a process of web media searches, court
document review, and other online sources.
3.4 Findings
3.4.1 Nature and Extent of CEQA Litigation
One way to evaluate CEQA litigation is to look at the incidence of CEQA petitions, but without
coincident consideration of the CEQA document under scrutiny, there is no way to capture the
full impacts of CEQA litigation on development in California. The EIR-petition matching
process undertaken for this paper is critical to understanding the locations and specific types of
projects being tried under CEQA.
38
Figure 12 - Distribution of EIR Project Types
The EIR-petition matching process reveals some surprising findings on which types of
CEQA projects are most often litigated. CEQA litigation by project type is shown in table 4
and the results are surprising. Twenty-nine percent of commercial projects, auto transportation
projects, and mixed use projects with EIRs were litigated in the 2010-2015 study period. Mixed
use projects with a residential component were also sued at a high rate of 28-percent. This is
ten percentage points higher than the average litigation rate by project type of 18 percent.
Some of the lowest litigation rates were observed for public service and infrastructure projects;
park/recreation projects; and agriculture projects. On top of this public service projects and
park/recreation projects made up almost thirty percent of all EIRs. Noxious use EIRs also have
low litigation rates but EIRs done on these types of projects represent less than ten percent of
all EIRs. Curiously, litigation rates for residential projects were slightly above average, but
0
50
100
150
200
250
300
350
400
450
Mixed Use with Residential
Residential
Commercial
Mining/Extraction/Oil Energy
Auto Transportation
Public Services and Infrastructure
Park/Recreation & Environment
Public Transportation
Agriculture
Noxious Uses
Mixed Use
Distribution of Project Types
EIRs without Petitions EIRs with Petitions
39
were not overwhelmingly large in light of public criticism blaming CEQA litigation for the
high cost of housing in the state. Housing developments subject to public criticism under
CEQA may avoid the CEQA litigation process through negotiated means such as community
benefits agreements. Large portions of California’s residential development may skip the CEQA
process altogether, through by-right development in accordance with a City’s General Plan
document or via other, city-specific programs (O’Neill et al., 2018).
Table 15 - Litigation by Project Type
3.4.2 CEQA Litigation & Time to Development
During the petition coding process, we collected information on the total time each petition
spent in court in order to track the incremental time delay associated with a CEQA lawsuit on a
development project. Tracking how long lawsuits spend in courts requires looking up each
individual case in court records and looking at the suit register. The petition data acquired via
Table 15.
Litigation by Project Type
Project Classification
EIRs by
Type
% of
EIRs by
Type
EIRs
without
Petitions
EIRs with
Petitions Litigation Rate
Mixed Use with Residential 400 0.20 312 88 0.28
Residential 235 0.12 195 40 0.21
Commercial 235 0.12 182 53 0.29
Mining/Extraction/Oil
Energy 187 0.10 152 35 0.23
Auto Transportation 103 0.05 80 23 0.29
Public Services and
Infrastructure 388 0.20 340 48 0.14
Park/Recreation &
Environment 142 0.07 125 17 0.14
Public Transportation 79 0.04 62 17 0.27
Agriculture 35 0.02 31 4 0.13
Noxious Uses 124 0.06 104 20 0.19
Mixed Use 36 0.02 28 8 0.29
N = 1964 N = 1611 N = 353 Avg. = 0.18
40
the FOIA request included many, but not all, court case numbers. Using court case numbers
and other identifying information associated with the petitions, I queried each County court to
find the suit’s entry and exit date from the court. In cases where the petition database did not
include the case number, I first used court database to find case numbers by party names. If
court searches were not successful, I then relied on web and media searches.
Across all petitions, there was a measured average time delay of 598 days (1.6 years)
associated with litigation. The delays range from 25 days to 2,212 days (6.06 yeas). The
uncertainty inherent to this large range is reason for developers to be concerned. This is a
significant time delay for developers trying to make their projects financially feasible, with
important implications for project feasibility and construction throughout the state. The
incremental costs to projects wrapped up in litigation will vary widely depending on the
financing arrangement on a project and the location.
3.4.3 Access to Environmental Law
What about the distribution of EIRs and CEQA litigation? Who is using this law? During the
coding process, EIRs and petitions were geocoded according to their reported project locations
and mapped using geospatial software. EIR locations are plotted on the map below.
41
Figure 13 - EIR and CEQA Petition Locations
Not surprisingly, large clusters of CEQA activity popped up in the most economically active
locations in California – Los Angeles, San Diego, San Francisco, and Sacramento. Development
and EIR activity is intimately tied to the availability of land for development and the economic
pressure to develop but EIR activity might also correlate to the status of a City’s General Plan
(Olshansky, 1996). As General Plans become outdated, CEQA bears an increasing burden in
guiding development decisions. California guidelines require cities to update their General
Plans every five years, but they are instead updated about every 12 years. This gap leaves many
years in between where city planning agencies make ad hoc decisions on development, instead of
42
relying on a general plan (Olshansky, 1996). So while EIR activity in this database may reflect
urban economic pressures, the role of General Plans should also be considered.
Examining the top five counties by number of EIRs, EIRs per square mile, and EIRs per
person reveals three different lists making qualitative discussions about CEQA activity difficult
and subjective. Los Angeles ranked number one in terms of raw EIR counts while San
Francisco County had the largest number of EIRs per square mile.
Table 16 - EIR Counts
Top 5 Counties by Number of EIRs
Rank County
EIR
Count
1. Los Angeles County 365
2. San Diego County 179
3. Riverside County 165
4. Santa Clara County 120
5. Orange County 119
Top 5 Counties by EIRs/Square Mile
EIR
Count
Area (Sq.
Mi.) EIRs/Square Mile
1. San Francisco County 59 47 1.26
2. San Mateo County 59 449 0.13
3. Orange County 119 948 0.13
4. Alameda County 89 738 0.12
5. Contra Costa County 72 720 0.10
Top 5 Counties by EIRs/Person
EIR
Count Population EIRs/Thousand People
1. Alpine County 1 1120 0.89
2. Inyo County 12 18026 0.67
3. Mono County 6 14168 0.42
4. Sierra County 1 2999 0.33
5. Amador County 9 38626 0.23
A linear regression model was prepared to better understand characteristics of census tracts
with and without EIRs. The dependent variable is equal to the total number of EIRs in each
43
tract and range from 0 to 15. Census tract characteristics including the college graduation rate,
homeownership rate, racial status, and median household income were taken from the US
Census Bureau American Community Survey 2017 5-year estimates. The proportions of college
graduates and homeowners in a tract are associated with more EIRs being filed, however these
relationships are not statistically significant at the p<0.01 level. Tracts with larger black and
Latino populations are less likely to have EIRs and these relationships are statistically
significant. Surprisingly, areas with higher median household income have a statistically
significant and negative coefficient but the measured effect is nearly zero.
Table 17 - Linear Regression of EIR Projects
(1)
VARIABLES Model 1
% college graduate 0.00167*
(0.000900)
% homeowners 0.000326
(0.000511)
% black -0.00366***
(0.00104)
% latino -0.00219***
(0.000485)
Median household income -1.42e-06***
(4.92e-07)
Constant 0.407***
(0.0449)
Observations 8,057
R-squared 0.009
Standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1
3.5 Conclusion:
This paper describes CEQA, CEQA litigation, and its interaction with urban development
process across the state. The EIR database reflects a landscape of development where EIRs are
44
most often being used for mixed use residential (400 EIRs) and public service and
infrastructure (388 EIRs) projects. There is no overwhelming use of EIRs for pure residential
development (235 EIRs), standing in contrast to much of the media attention CEQA has
received. This pattern in the EIR database may reflect development trends towards certain
types of projects – developers are simply more interested in building mixed use and
infrastructure. Or, it might reflect gaps in other city planning documents such as General
Plans. In cities where the General Plan is outdated or lacks specificity when it comes to mixed
use and infrastructure projects, CEQA may be the only tool available for planners to approving
these project types. It is important that city planning agencies review their General Plans to
untangle which mechanism is patterning environmental impact reviews under their
jurisdiction. For the residential projects which do undergo the EIR process, about 20-percent
are litigated. It does not seem that residential project EIRs are specifically targeted by
litigators; the global EIR litigation rate is 21%. Previous inquiries into CEQA litigation’s role
in deterring residential development relied only on counts of CEQA petitions filed against all
CEQA document types, and did not link every petition to an EIR (Holland and Knight, 2015).
Having this linked data of petitions and EIRs gives us a much closer look at how CEQA
petitions are being used.
Media reports of CEQA litigation posing serious delays to development timelines are
validated by the data on CEQA lawsuit timing. For EIRs that are petitioned, developers may be
subject to delays of up to six years. Certainly, this cannot be ignored. For the projects that do
end up in court, there may be ways for developers and communities to reach an agreement
more quickly. Negotiated Community Benefits Agreements (CBAs) may be one tool
communities have to bargain for better development outcomes including enhanced
environmental quality or additional project amenities. In cases where labor interests are
45
bringing suits under CEQA, Project Labor Agreements (PLAs) could also reduce the
development uncertainty associated with litigation. Developers might have opportunities to
negotiate both CBAs and PLAs before litigation is threatened to avoid going to court
completely or they might negotiate a CBA/PLA during litigation to reduce the total time delay
and bring a resolution more quickly. Future research on the incidence of PLAs and CBAs to
neutralize EIR opponents is justified and could provide important insights for communities,
labor interests, and developers. Additional work should also assess whether these agreements
alter the environmental outcomes produced by the EIR process. In other words, by potentially
altering how litigation occurs, do environmental outcomes change?
CEQA litigation gets a lot of buzz in urban areas such as Los Angeles and San
Francisco. Geocoded EIR and petition data validates this but raises issue with simple counts of
petitions by jurisdictional areas. The high number of petitions in Los Angeles County may be
due only to the County’s huge geographic coverage. All but two California counties had one or
more EIRs during the study period with an average per county EIR count of 38. Census tracts
with more black and brown residents are associated with fewer EIRs. This may suggest
racialized city planning departmental practices which are requiring less environmental
oversight in minority communities, giving developers the option to file negative declarations or
mitigated negative declarations instead of EIRs. This could also be reflective of slower, less
intense development in minority neighborhoods. The research design presented here does not
offer an opportunity to untangle this dynamic but future work should consider the relative
rates of negative declarations, mitigated negative declarations, and environmental impact
reviews according to neighborhood characteristics. CEQA may be providing some communities
with a valuable tool and insight into local development while others are left without a window
into development processes.
46
This paper has attempted to more completely document the functioning of CEQA, the
EIR process, and CEQA litigation across the state. While CEQA is not perfect, it does not
appear that the law is targeting systematically challenging affordable development in the state.
CEQA is the only lever communities have to guide local development and repealing it would be
unsubstantiated based on what we know about litigation rates. City planners with large CEQA
caseloads should look closely at the types of projects under suit in light of their General Plan.
Piecemealed development fettered by CEQA litigation may be streamlined by updates to
General Plans which promote sustainable, affordable development with well-laid plans.
47
4 References:
Adams, M. A., Todd, M., Kurka, J., Conway, T. L., Cain, K. L., Frank, L. D., & Sallis, J. F.
(2015). Patterns of walkability, transit, and recreation environment for physical
activity. American journal of preventive medicine, 49(6), 878-887.
Allen, D. 2001. Social class, race, and toxic releases in American counties, 1995. Social Science
Journal 38(1):13-25.
Ash, M., Boyce, J. K., Chang, G., Pastor, M., Scoggins, J., & Tran, J. (2009). Justice in the air:
tracking toxic pollution from America’s industries and companies to our states, cities,
and neighborhoods. Political Economy Research Institute, Amherst, MA.
Barbour, E. and Teitz, M. (2005). CEQA reform: Issues and Opinions. Public Policy Institute of
California.
Bass, R. E., Bogdan, K and Terry Rivasplata. CEQA Deskbook, Third Edition, Solano Press,
2012.
Bedimo-Rung, A. L., Mowen, A. J., & Cohen, D. A. (2005). The significance of parks to physical
activity and public health: a conceptual model. American journal of preventive
medicine, 28(2), 159-168.
Been, V. (1994). Locally undesirable land uses in minority neighborhoods: disproportionate
siting or market dynamics?. Yale law journal, 1383-1422.
Been, V., & Gupta, F. (1997). Coming to the nuisance or going to the barrios? A longitudinal
analysis of environmental justice claims. Ecology law quarterly, 24(1), 1-56.
Boer, J. T., Pastor, M., Sadd, J. L., & Snyder, L. D. (1997). Is there environmental racism? The
demographics of hazardous waste in Los Angeles County. Social Science Quarterly, 78(4),
793-810.
Bostic, R., & Lee, K. O. (2009). Homeownership: America’s Dream?. Insufficient Funds: Savings,
Assets, Credit, and Banking among Low-Income Households, Rebecca M. Blank and Michael S.
Barr (218–56). New York: Russell Sage Foundation.
Boone, C. G., Buckley, G. L., Grove, J. M., & Sister, C. (2009). Parks and people: An
environmental justice inquiry in Baltimore, Maryland. Annals of the Association of
American Geographers, 99(4), 767-787.
Bowen, W. M., Salling, M. J., Haynes, K. E., & Cyran, E. J. (1995). Toward environmental
justice: Spatial equity in Ohio and Cleveland. Annals of the Association of American
Geographers, 85(4), 641-663.
Bradford, B. (2018). Is California’s CEQA environmental law protecting natural beauty — or
blocking affordable housing?. Capital Public Radio. Retrieved from
48
https://www.scpr.org/news/2018/07/11/84603/is-california-s-ceqa-environmental-
law-protecting/
Brownson, R., Baker, E., Housemann, R., Brennan, L., & Bacak, S. (2001). Environmental and
policy determinants of physical activity in the United States. (Research). The American
Journal of Public Health, 91(12), 1995–2003. https://doi.org/10.2105/AJPH.91.12.1995
Bullard, R. D. (1983). Solid waste sites and the black Houston community. Sociological
inquiry, 53(2‐3), 273-288.
California Environmental Quality Act, California Pub. Res. Code § 21065; Guidelines §
15378(a)
California Forward (2012). ‘Time for California to get on CEQA reform bus?’ [Video].
Cutter, S. L., Holm, D., & Clark, L. (1996). The role of geographic scale in monitoring
environmental justice. Risk Analysis, 16(4), 517-526.Boyle, M., & Kiel, K. (2001). A
survey of house price hedonic studies of the impact of environmental
externalities. Journal of real estate literature, 9(2), 117-144.
Chatman, D. G., Tulach, N. K., & Kim, K. (2012). Evaluating the economic impacts of light rail
by measuring home appreciation: A first look at New Jersey’s River Line. Urban studies,
49(3), 467-487.
Collins, M. B., Munoz, I., & JaJa, J. (2016). Linking ‘toxic outliers’ to environmental justice
communities. Environmental Research Letters, 11(1), 015004.
Curry, M. (2019). Is CEQA Preventing Housing from Being Built in California? Retrieved from
https://cal.streetsblog.org/2019/03/13/is-ceqa-preventing-housing-from-being-built-
in-california/
Di, Z. X. (2005). Does housing wealth contribute to or temper the widening wealth gap in
America?. Housing Policy Debate, 16(2), 281-296.
Division of Adult and Community Health, National Center for Chronic Disease Prevention and
Health Promotion, Centers for Disease Control and Prevention. (2020). Behavioral Risk
Factor Surveillance System online prevalence data, 1995–2010. Retrieved from
https://www.cdc.gov/brfss/brfssprevalence/
Elkind, E. (2014). Misleading Attacks on California’s New Transportation Analysis Under
CEQA. Retrieved from https://legal-planet.org/2014/08/28/misleading-attacks-on-
californias-new-transportation-analysis-under-ceqa/
Esri (2018). Spatial Autocorrelation (Moran’s I) Retrieved from http://pro.arcgis.com/en/pro-
app/tool-reference/spatial-statistics/spatial-autocorrelation.htm
Fischel, W. A. (2001). The homevoter hypothesis: How home values influence local
government taxation, school finance, and land use policies. Cambridge, MA: Harvard
University Press.
49
Forsyth, A. (2000). Analyzing public space at a metropolitan scale: Notes on the potential for
using GIS. Urban geography, 21(2), 121-147.
Janet E. Kohlhase, 1991. The impact of toxic waste sites on housing values. Journal of Urban
Economics, 30(1), 1-26.
Gatzlaff, D., Green, R., & Ling, D. (1998). Cross-Tenure Differences in Home Maintenance and
Appreciation. Land Economics, 74(3), 328-342.
Gayer, T., Hamilton, J. T., & Viscusi, W. K. (2000). Private values of risk tradeoffs at superfund
sites: housing market evidence on learning about risk. Review of Economics and
Statistics, 82(3), 439-451.
General Accounting Office, U. S. (1983). Siting of hazardous waste landfills and their
correlation with racial and economic status of surrounding communities. United States
General Accounting Office.
Gertler, M., & Gilchrist, S. (2018). What happened: Financial factors in the great
recession. Journal of Economic Perspectives, 32(3), 3-30.
Goodman, C. J., & Mance, S. M. (2011). Employment loss and the 2007–09 recession: an
overview. Monthly Labor Review, 134(4), 3-12.
Gyourko, J., Linneman, P., & Wachter, S. (1999). Analyzing the relationships among race,
wealth, and home ownership in America. Journal of Housing Economics, 8(2), 63-89.
Hamilton, J. T. (1995). Testing for environmental racism: prejudice, profits, political
power?. Journal of Policy Analysis and Management, 14(1), 107-132.
Hecht, S. (2015). Anti-CEQA Lobbyists Turn to Empirical Analysis, But Are Their
Conclusions Sound? Retrieved from https://legal-planet.org/2015/09/28/anti-ceqa-
lobbyists-turn-to-empirical-analysis-but-are-their-conclusions-sound/
Hird, J. A. (1993). Environmental policy and equity: The case of Superfund. Journal of Policy
Analysis and Management, 12(2), 323-343.
Hite, D., Chern, W., Hitzhusen, F., & Randall, A. (2001). Property-value impacts of an
environmental disamenity: the case of landfills. The Journal of Real Estate Finance and
Economics, 22(2-3), 185-202.
Holland & Knight (2015). In the name of the environment: How litigation abuse under the
California Environmental Quality Act undermines California’s environmental, social
equity and economic priorities.
Johnson, H. B. (1995). Making the Case for Transit: Emphasizing the" Public" in Public
Transportation. The Urban Lawyer, 1009-1014.
Kiel, K. A., & McClain, K. T. (1995). House Prices during Siting Decision Stages: The Case of
an lncinerator from Rumor through Operation.
Kim, Sunwoong. Race and Home Price Appreciation in Urban Neighborhoods: Evidence from
Milwaukee, Wisconsin.
50
Lindsey, G., Maraj, M., & Kuan, S. (2001). Access, equity, and urban greenways: An
exploratory investigation. The Professional Geographer, 53(3), 332-346.
Loukaitou-Sideris, A., 1995. Urban form and social context: Cultural differentiation in the uses
of urban parks. Journal of Planning Education and Research, Vol. 14, 89-102.
Melissa Boyle and Katherine Kiel, 2001. A Survey of House Price Hedonic Studies of the
Impact of Environmental Externalities. Journal of Real Estate Literature: 2001, 9(2), 117-
144.
Mendelsohn, R., Hellerstein, D., Huguenin, M., Unsworth, R., & Brazee, R. (1992). Measuring
hazardous waste damages with panel models. Journal of Environmental Economics and
Management, 22(3), 259-271.
Mennis, J. 2002. Using geographic information systems to create and analyze statistical
surfaces of population and risk for environmental justice analysis. Social Science
Quarterly, 83(1): 281-297.
Mennis, J. L., & Jordan, L. (2005). The distribution of environmental equity: Exploring spatial
nonstationarity in multivariate models of air toxic releases. Annals of the Association of
American Geographers, 95(2), 249-268.
Metro. (2019). Next stop: more access to open spaces. Retrieved from
http://media.metro.net/projects_studies/toc/images/nextStop_transitToParks_05-
2019.pdf
Millner, F. (2011). Access to environmental justice. Deakin L. Rev., 16, 189.
Mohai, P., & Bryant, B. (1992). Environmental injustice: weighing race and class as factors in
the distribution of environmental hazards. U. Colo. L. Rev., 63, 921.
Morello-Frosch, R., Pastor, M., & Sadd, J. (2001). Environmental justice and Southern
California’s “riskscape” the distribution of air toxics exposures and health risks among
diverse communities. Urban Affairs Review, 36(4), 551-578.
Morello-Frosch, R., Pastor, M., Porras, C., & Sadd, J. (2002). Environmental justice and
regional inequality in southern California: implications for future research.
Environmental Health Perspectives, 110(Supplement 2), 149–154.
Nicholls, S. (2001). Measuring the accessibility and equity of public parks: A case study using
GIS. Managing leisure, 6(4), 201-219.
Olshansky, R. (1996). The California Environmental Quality Act and local planning. Journal of
the American Planning Association, 62(3), 313-330.
Palmquist, R. B., Roka, F. M., & Vukina, T. (1997). Hog operations, environmental effects, and
residential property values. Land economics, 114-124.
Parcel, T. L. (1982). Wealth accumulation of black and white men: The case of housing
equity. Social Problems, 30(2), 199-211.
51
Pastor, M., Morello‐Frosch, R., & Sadd, J. L. (2005). The air is always cleaner on the other side:
Race, space, and ambient air toxics exposures in California. Journal of Urban Affairs,
27(2), 127-148.
Pastor, M., Sadd, J., & Hipp, J. (2001). Which came first? Toxic facilities, minority move‐in, and
environmental justice. Journal of urban affairs, 23(1), 1-21.
Political Economy Research Institute (2019). How accurate are the RSEI data on toxic air
pollution? Retrieved from https://www.peri.umass.edu/how-accurate-are-the-rsei-data-
on-toxic-air-pollution
Pulido, L. (1996). A critical review of the methodology of environmental racism research.
Antipode, 28(2), 142-159.
Rigolon, A. (2016). A complex landscape of inequity in access to urban parks: A literature
review. Landscape and Urban Planning, 153, 160-169.
Rigolon, A., & Németh, J. (2018). A QUality INdex of Parks for Youth (QUINPY): Evaluating
urban parks through geographic information systems. Environment and Planning B:
Urban Analytics and City Science, 45(2), 275-294.
Sadd, J. L., Pastor Jr, M., Boer, J. T., & Snyder, L. D. (1999). “Every breath you take...”: the
demographics of toxic air releases in Southern California. Economic Development
Quarterly, 13(2), 107-123.
Sadd, J. L., Pastor, M., Morello-Frosch, R., Scoggins, J., & Jesdale, B. (2011). Playing It Safe:
Assessing Cumulative Impact and Social Vulnerability through an Environmental
Justice Screening Method in the South Coast Air Basin, California. International Journal
of Environmental Research and Public Health, 8(5), 1441–1459.
Schweitzer, Lisa, and Max Stephenson Jr. "Right answers, wrong questions: Environmental
justice as urban research." Urban Studies 44.2 (2007): 319-337.
Shapiro, T., Meschede, T., & Osoro, S. (2013). The roots of the widening racial wealth gap:
Explaining the black-white economic divide. Research and policy brief.
Sheppard, E., Leitner, H., McMaster, R. B., & Tian, H. (1999). GIS-based measures of
environmental equity: exploring their sensitivity and significance. Journal of Exposure
Science & Environmental Epidemiology, 9(1), 18-28.
Theberge, JP. (2019). Environmental Laws Don’t Impede Development. Retrieved from
https://www.voiceofsandiego.org/topics/opinion/environmental-laws-dont-impede-
development/
United Church of Christ. Commission for Racial Justice. (1987). Toxic wastes and race in the
United States: A national report on the racial and socio-economic characteristics of communities
with hazardous waste sites. Public Data Access.
52
US General Accounting Office. (1983). Siting of hazardous waste landfills and their correlation
with racial and economic status of surrounding communities. GAOfRCED-83-168.
Washington, DC: US. General Accounting Office.
United States Environmental Protection Agency (2018). Basics of TRI Reporting. Retrieved
from https://www.epa.gov/toxics-release-inventory-tri-program/basics-tri-reporting
United States Environmental Protection Agency (2018b). What is RSEI? Retrieved from
https://www.epa.gov/rsei/learn-about-rsei
Wolch, J., Wilson, J. P., & Fehrenbach, J. (2005). Parks and park funding in Los Angeles: An
equity-mapping analysis. Urban geography, 26(1), 4-35.
Zabel, J. E., & Kiel, K. A. (2000). Estimating the demand for air quality in four US cities. Land
Economics, 174-194.
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Healthy mobility: untangling the relationships between the built environment, travel behavior, and environmental health
PDF
Location of warehouses and environmental justice: Three essays
PDF
Urban air pollution and environmental justice: three essays
PDF
Spatial and temporal expenditure-pricing equity of rail transit fare policies
PDF
The demand for reliable travel: evidence from Los Angeles, and implications for public transit policy
PDF
The role of social media for community-based organizations focused on environmental justice in southeast Los Angeles
PDF
The built environment, tour complexity, and active travel
PDF
Household mobility and neighborhood impacts
PDF
Productive frictions and urbanism in transition: planning lessons from traffic flows and urban street life in Ho Chi Minh City, Vietnam
PDF
Diverging pathways to citizenship and immigrant integration in the U.S.
PDF
The long-term impact of COVID-19 on commute, employment, housing, and environment in the post-pandemic era
PDF
The impact of demographic shifts on automobile travel in the United States: three empirical essays
PDF
Insights into residential mobility and pricing of rental housing: the role of gentrification, home-ownership barriers, and market concentrations in low-income household welfare
PDF
Three essays in United States real estate markets
PDF
Essays on congestion, agglomeration, and urban spatial structure
PDF
Cities on the environmental justice frontline: the intractability of hazards and the governability of amenities
PDF
Community development agreements: addressing inequality through urban development projects
PDF
The flexible workplace: regional tendencies and daily travel implications
PDF
Who learns where: understanding the equity implications of charter school reform in the District of Columbia
PDF
Unraveling decentralization of warehousing and distribution centers: three essays
Asset Metadata
Creator
Swayne, Madison R.E.
(author)
Core Title
Environmental justice in real estate, public services, and policy
School
School of Policy, Planning and Development
Degree
Doctor of Philosophy
Degree Program
Urban Planning and Development
Publication Date
04/21/2020
Defense Date
04/21/2020
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
accessibility,environmental justice,environmental policy,Housing,OAI-PMH Harvest,parks,public services,Public transportation,Real estate,repeat-sales,Transportation
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Schweitzer, Lisa (
committee chair
), Boarnet, Marlon (
committee member
), Green, Richard (
committee member
), Pastor, Manuel (
committee member
)
Creator Email
madi.swayne@gmail.com,swayne@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-286313
Unique identifier
UC11673087
Identifier
etd-SwayneMadi-8294.pdf (filename),usctheses-c89-286313 (legacy record id)
Legacy Identifier
etd-SwayneMadi-8294.pdf
Dmrecord
286313
Document Type
Dissertation
Rights
Swayne, Madison R.E.
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
accessibility
environmental justice
environmental policy
public services
repeat-sales