\IHHHIHlllHlHIIIHIHllNlHll‘UHIHHHIHIHIIIWI

      

THS

This is to certify that the
thesis entitled

ENVIRONMENTAL INEQUALITY AND BROWNFIELD
REDEVELOPMENT IN METRO DETROIT

presented by
BETH DIELEMAN DYKSTRA

has been accepted towards fulfillment
of the requirements for the

MA. degree in Geography

W

' Major Professor’s Signature

3- II- 0%

Date

 

MSU is an Affinnative Action/Equal Opportunity Institution

 

 

I LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

am it???»

 

$13? 1 5 2033

 

ME 2 a 3191i?

 

Itov o 3 22;;

“it"fj lo

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/CIFiC/DateDue.p65-p.15

ENVIRONMENTAL INEQUALITY AND BROWNFIELD
REDEVELOPMENT IN METRO DETROIT

By

Beth Dieleman Dykstra

A THESIS
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of
MASTER OF ARTS
Departments of Geography and Urban Affairs

2004

ABSTRACT

ENVIRONMENTAL INEQUALITY AND BROWNFIELD
REDEVELOPMENT IN METRO DETROIT

By

Beth Dieleman Dykstra

This study examines the demographics of brownfield communities in Metro
Detroit to determine whether disadvantaged groups are more likely to live in close
proximity to these brownfields. The findings are compared to the demographics of
communities where state funded brownfield redevelopment is occurring to determine
whether redevelopment is equitably distributed. Data were obtained from the 1990 and
2000 Census of Population and Housing to assess demographic characteristics. Data on
brownfields were obtained from Michigan's Department of Environmental Quality.
Difference of means tests are used to measure significance in demographic differences
between communities with varying numbers of brownfields. The Index of Dissimilarity
is used to determine whether the distribution of various population groups is similar to
the distribution of brownfields. The results show poorer communities with lower housing
values, rents, owner or occupied housing are more likely to have brownfields. Analysis
of state redevelopment efforts show that for most communities, state brownfield
redevelopment efforts have occurred in communities in greatest need of economic

redevelopment and revitalization.

TABLE OF CONTENTS

LIST OF TABLES ............................................................................. v
LIST OF FIGURES ............................................................................ vi
1. Introduction ............................................................................ 1
H. Significance of the Research .......................................................... 3
HI. Background to the Research Questions .............................................. 5
IV. The Research Questions ............................................................... 6
V. Literature Review .................................................................... 7
A. Clarifying the terms .................................................................... 7
B. Environmental Justice Studies ................................................. 10
l. The Three Key Studies ........................................................ 10
2. Michigan Research .............................................................. 11
3. Research on Brownfields ...................................................... 12
4. Research on Scale .............................................................. 13
VI. Federal and State Brownfield Redevelopment Policy:
Economic Development, Environmental Justice, or Both? ........................ 15
VII. Summary ............................................................................ 18
VIII. Data and Analysis ................................................................... 20
A. Data ................................................................................. 20
B. Hypothesis .......................................................................... 23
C. Research Objectives ............................................................... 23
D. Methodology .................................................................... 24
E. Unit of Analysis and Study Area ................................................. 26
XI. Findings ..................................................................................... 29
A. Question 1 ............................................................................. 29
B. Question 2 ............................................................................. 44
X. Discussion ............................................................................. 50
A. Does environmental inequality exist in Metro Detroit? ....................... 50

B. Have the state’s efforts to redevelop brownfields been equitably
distributed? Do state redevelopment efforts directly address
inequalities that exist? .......................................................... 53

iii

XI. Conclusions and the Need for Further Research .............................. 56

BIBLIOGRAPHY ............................................................................ 58

iv

LIST OF TABLES

Table 1: Percentage of total population living in tracts with BEAs ..................... 30
Table 2: All Counties ........................................................................... 33
Table 3: Individual Counties .................................................................. 34
Table 4: Macomb County Municipalities ................................................ 35
Table 5: Oakland County Municipalities ................................................ 36
Table 6a. Wayne County Municipalities ................................................ 37
Table 6b. Wayne County Municipalities ................................................ 38
Table 7: D Values for the Index of Dissimilarity Text .............................. 41
Table 8: All Counties ........................................................................... 44
Table 9: Individual Counties .................................................................. 45
Table 10: Oakland County Municipalities ................................................ 46
Table 11: Wayne County Municipalities ................................................ 47

LIST OF FIGURES

Figure 1: Map of Study Area .................................................................. 28

vi

I. Introduction

Most central cities in the United States experienced a prolonged period of
disinvestment and out-migration of population and economic activity in the latter half of
the twentieth century. Since then, efforts to redevelop the central city have largely been
ineffective. One explanation for failed efforts to revitalize them is that industrial
restructuring has eroded the economic base of central cities, eliminated manufacturing
jobs, and left vacant industrial properties called brownfields (Leigh and Gradeck 1996).
Brownfields are “abandoned, idled, or underused industrial and commercial facilities
where expansion or redevelopment is complicated by real or perceived environmental
contamination” (EPA 2000). Earlier decades of industry and manufacturing have left
many brownfield properties environmentally degraded; contaminated with heavy metals,
organic and inorganic chemicals, and petroleum constituents; and littered with
dilapidated buildings and debris (EPA 2000). Environmental contamination is a negative
neighborhood extemality and a source of real or perceived health risks, lowers property
values, and acts as a force in widening inequalities between central cities and the rest of
the nation (Leigh and Gradeck 1996).

Two recent responses to the problem of brownfields are distinct, but not
necessarily opposed. The first is to view them as opportunities for economic
development. In this strategy, both the removal of site contamination and the potential
use of the cleaned site are considered (Hula 2001). The second response is concern for
environmental justice as many brownfield sites are located in poor and minority
communities (EPA 2000). Environmental justice advocates seek to “develop a paradigm

to achieve socially equitable, environmentally healthy, economically secure,

psychologically vital, spiritually whole, and ecologically sustainable communities”
(NEJ AC 1996). They argue that brownfield redevelopment must be linked with the
values cf environmental justice to address the needs and concerns of all, particularly
minorities and the poor, exposed to brownfield risks.

Many states, including Michigan, have created brownfield redevelopment
programs aimed at cleaning up contamination while spurring economic growth (Baumer
et a1 1999). However, according to Leigh (2000), “the current practice of many
brownfield redevelopment projects is to select only the most marketable sites for
remediation and redevelopment”. This practice “excludes disadvantaged neighborhoods
from programs aimed at redeveloping brownfields. Doing so creates the potential for
widening existing inequalities between better-off and worse-off neighborhoods.”

Though Leigh may be correct in her assertions, a focus on economic development
means that brownfield efforts are more likely to be integrated into state and community
planning where additional concerns, including environmental justice, can be considered.
The National Governors Association (2000) brief on brownfields showed states how they
can make brownfield redevlopment a key component in their state growth management
initiatives. They argue that brownfield urban redevelopment projects “help preserve
farmland, rural communities, and open spaces.” In addition, the focus on economic
development has encouraged support for brownfield redevelopment from diverse interest
groups. In Michigan, the state manufactuers association, state and local chambers of
commerce, local political leaders, community groups, and the rural community show
strong support for Michigan’s brownfield development plan, called the Michigan

Brownfield Initiative (Hula 2000).

II. Significance of the Research

The estimated 450,000 brownfields in the United States affect both individuals
and communities. If contaminated, these sites can pose health risks to those working or
living nearby. Before any site became a brownfield, there was a viable industry that
provided jobs for community members and contributed to the community’s identity. The
industry may have played an active role in the community through support of parks,
recreation, schools, and cultural programs. When the business closed and abandoned the
site, they became blights on the communities they once supported (Gist 1999).
Brownfields found in environmentally or economically distressed urban areas compound
problems the community is already experiencing, such as crime or a decline in business.
The negative effects of brownfields on individuals and communities raise serious
questions regarding whether any part of the p0pulation is more likely to live in close
proximity to brownfields.

Brownfield redevelopment provides benefits to individuals and communities.
Public health and safety concerns are addressed since the presence and extent of
contamination is identified. That contamination is removed or capped to prevent
exposure. The surrounding community benefits from increased property values and a
renewed tax base. Brownfield redevelopment removes blight, contributing to an
improved community image. Brownfield redevelopment efforts reduce the development
of greenfields, farmland, and open space, preventing the loss of diversity, habitat, and
wetlands. When businesses locate on brownfields they can use existing infrastructure

and public services. Such redevelopment can correct inequities between central cities

and the suburbs. Since brownfield redevelopment provides so many benefits, it is
important to know whether such efforts are equitably distributed.

This study determines the extent to which certain populations groups are exposed
to brownfields in Metro Detroit. The study also determines whether the State of
Michigan is addressing any inequalities in exposure through brownfield redevelopment
projects. Lastly, this research provides a model the MDEQ can use to continually
monitor the demographics of communities where brownfields in the state are located and
the demographics of communities where they fund brownfield redevelopment. Such
knowledge would provide an additional factor to consider when evaluating potential
redevelopment projects, putting the state at the forefront of linking the values of

economic development with environmental justice.

III. Background to the Research Questions

In 1999, the Anderson Economic Group examined the demographics of
Michigan’s urban brownfield communities to determine whether minorities or the poor
were more likely to live in communities with brownfields. They defined brownfields as
“contaminated, or potentially contaminated, properties" (1 999:2) as a result of historic
industrial activity. They identified all cities within Michigan that had brownfield
authorities or were included in larger brownfield authorities. They then aggregated and
analyzed the demographic characteristics of those cities and compared them to the
characteristics of the state. Their findings indicated that a slightly larger percentage of
the non-white population lives in brownfield authority areas compared to the state as a
whole. They also found that medium household income in brownfield cities is less than
that for the state (Anderson and Clemens 1999).

The Anderson Economic Group study is the only one of its kind for Michigan.
Since the study only looks at environmental inequality at the scale of the city, there is
need for further analysis at a finer scale to obtain more in depth information. At the city
scale, the study was not able to show who is living in closest proximity to the brownfields
within the city. Such knowledge would give a clearer idea of the extent to which
minorities and the poor may or may not be disproportionately exposed to the negative
effects of brownfields. No studies have attempted to analyze Michigan’s brownfield
redevelopment efforts. Such a study would provide insight into whether Michigan’s goal
of economic development through brownfield redevelopment also addresses

environmental justice concerns.

IV.

The Research Questions

Given the need for further study, my research will answer the following questions:

1)

2)

Is there a relationship between the spatial distribution of brownfields and the
demographic and socioeconomic characteristics of neighborhoods? Specifically,
is there a difference between the racial, economic, and housing characteristics of
communities with and without brownfields. Also, is the percentage of minorities
in a neighborhood positively related to the percentage of brownfields in a
neighborhood? Is the percentage of low-income residents in a neighborhood
positively related to the percentage of brownfields in the neighborhood? Is the
average home value and monthly rent, percentage of occupied housing, and
percentage of owner occupied housing related to the percentage of brownfields in
the neighborhood?

What are the demographics of communities where the state has invested in
redevelopment? Are investments directed to neighborhoods that are already
“better off”, essentially ignoring environmental justice concerns, or, are the
projects specifically targeted to disadvantaged neighborhoods (i.e., those occupied
by a higher percentage of racial minorities and the poor)? Does the distribution

fall somewhere in between?

V. Literature Review:
A. Clarifidng the terms

Most research designed to determine whether some socioeconomic or racial
groups are more likely to be exposed to the negative effects of pollutants does not
distinguish between the terms “environmental racism” and “environmental justice”.
Even less research uses or defines the terms “environmental equity” and “environmental
inequality”. It is important to define these terms to ensure a consistent understanding of
them.

The United States Environmental Protection Agency (2003) defines
environmental justice as “the fair distribution of environmental risks across
socioeconomic and racial groups” and “the fair treatment of people of all races, cultures,
incomes, and educational levels with respect to the development and enforcement of
environmental laws, regulations, and policies. Fair treatment implies that no population
should be forced to shoulder a disproportionate share of exposure to the negative effects
of pollution due to lack of political or economic strengt ”.

Bryant expands the definition to include “cultural norms and values, rules,
regulations, behaviors, policies, and decisions to support sustainable communities, where
people can interact with confidence that their environment is safe, nurturing, and
productive” (Bryant 199526). According to environmental justice values, communities
and populations that experience unequal exposure to any environmental hazard are
suffering from an environmental injustice that should be removed. Specific examples of
these environmental injustices include environmental racism and environmental

inequality.

Environmental racism—unequal and deliberate exposure to toxic and hazardous
waste on people of color—emerged in the 19805 as a fi'amework to understand how
environmental risks, known or potential, are distributed differently across demographic
groups. Since 1983, many studies designed to detect environmental racism have
analyzed the demographics of communities surrounding solid waste sites (dumps and
landfills) and hazardous waste sites (facilities that transport, store, and dispose of
hazardous industrial wastes) to determine whether such sites are more likely to be placed
in poor and minority neighborhoods. A common assumption of this research is these
communities were deliberately targeted for least desirable land uses.

A number of academics and organizations have begun challenging both the idea
and existence of environmental racism (Anderton et a1. 1994; Anderson et a1. 1999).
Their challenge is based on various studies that have found income, not race, to be the
variable most highly correlated to the location of waste facilities. Since intentional
placement is an important component of environmental racism, other studies have
discredited the existence of environmental racism by focusing on what came first, the
non-white population or the hazard (Been 1993; Hurley 1997; Cutter 1995). Bullard has
consistently stressed that this debate is irrelevant (1994) since the end result is a
discriminatory pattern (Pulido 1996). As a result, researchers concerned with
discriminatory patterns have begun to change the focus from environmental racism to
environmental inequality.

“Environmental inequality” expands on environmental racism to include any form
of environmental hazard that burdens one social group more than another (Fellow 2000).

While environmental racism focuses on the intentional placement of hazardous sites in

poor and minority communities, environmental inequality focuses on the broader

dimensions of the intersection between environmental quality and social hierarchies. It

addresses structural questions that focus on social inequality (the unequal distribution of

power and resources in society) and environmental burdens. Environmental inequality

research is interested in both the descriptive (whether injustices exists) and process (how

those injustices came to be).

argues:

One complex process that results in inequalities is white privilege. Pulido (2000)

“...a focus on white privilege enables us to develop a more structural, less
conscious, and more deeply historicized understanding of racism. It differs from
a hostile, individual, discriminatory act, in that it refers to the privileges and
benefits that accrue to white people by virtue of their whiteness... Hence, instead
of asking if an incinerator was placed in a Latino community because the owner
was prejudiced, I ask, why is it that whites are not comparably burdened with
pollution? ...Industrialization, decentralization, and residential segregation are
keys to this puzzle. Because industrial land use is highly correlated with pollution
concentrations and people of color, the crucial question becomes, how did whites
distance themselves from both industrial pollution and nonwhites? ...White
privilege allows us to see how environmental racism has been produced—not
only by consciously targeting people of color (as in the incinerator cases)—but by
the larger processes of urban development, including white flight, in which whites

have sought to fully exploit the benefits of their whiteness.” (Pulido 2000).

Throughout this study, I use the term environmental justice as a broad term to
describe the values that motivate research that measures unequal exposure to any
environmental hazards. I use the term environmental inequality in my analysis of the
demographics of brownfield communities in Metro Detroit. Lastly, this study measures
“environmental equity”, equal protection of communities through environmental laws
(Bryant 1995), by looking at the distribution of government funded brownfield
redevelopment.

B. Environmental Justice Studies

1. The Three Key Studies

The first major study of community demographics near facilities for treatment,
storage, and disposal of hazardous wastes was conducted by the US. Government
Accounting Office (GAO 1983). The objective of this study “was to determine the
correlation between the location of hazardous waste landfills and the racial and economic
status of surrounding communities” (GAO 1983z2). Researchers compiled population
data for census areas surrounding four hazardous waste facilities in EPA region IV (the
Southeastern states). The GAO concluded that the majority of the population
surrounding three of the four facilities was black (GAO 1983).

In 1987, the United Church of Christ Commission for Racial Justice published a
report documenting racial and socio-economic characteristics of communities throughout
the United States with commercial hazardous waste sites and uncontrolled toxic waste
sites. They studied the demographic characteristics of zip code areas containing one or
more of these sites. They found that race proved to be the factor most significantly

correlated to the location of commercial hazardous waste facilities. Socio-economic

10

status was less correlated to the location of waste facilities, though incomes and home
values were substantially lower in communities with commercial facilities than those
without. They also found that the majority of Black and Hispanic Americans live in
communities with uncontrolled toxic waste sites (1987).

A third classic study (Bullard 1990) analyzed the demographics of the census
tracts that made up the 25 neighborhoods where various types of solid waste facilities
were found. While neighborhoods with majority Black residents compose just over one-
fourth of the city’s population, 21 out of 25 of those facilities were found in these
neighborhoods. Bullard’s work has been monumental in that he was the first to really
examine why such inequalities were occurring. His research has consistently explored
the thesis that “black communities, because of their economic and political vulnerability,
have been routinely targeted for the siting of noxious facilities, locally unwanted land
uses, and environmental hazards” (Bullard 1990:14).

2. Michigan Research

In Michigan, Mohai and Bryant (1992) assessed racial biases in distribution of
commercial hazardous waste facilities within three counties surrounding the city of
Detroit. They used a mail survey to obtain racial and economic data for 793 individuals.
From this survey, the authors compiled the population composition within various
distances of the hazardous waste facilities. Descriptive analysis found that only 3% of all
whites in the three-county area lived within one mile of a facility compared with 11% of
the minority population. They used multiple regression analysis with the distance of the
resident to the facility as the dependent variable and race and income of the resident as

the independent variables. Their analysis showed “the relationship between race and

11

location of commercial hazardous waste facilities in the Detroit area is independent of
income” and that “race is more importantly related to the distribution of these hazards
than income” (Mohai and Bryant 1992: 1 74).

3. Research on Brownfields

A few studies have looked at the demographics of brownfield communities. Leigh
and Gradeck (1996) compared the means of demographic characteristics of census tracts
with varying numbers of environmentally suspect and delinquent sites in Milwaukee.
They found that blacks were more likely to live in tracts containing three or more such
sites than whites. As the number of sites per tract rose, so too did the tract’s percentage
of black population. They also found the economic status of the population in the areas
most plagued by such property was lower than that of the population of tracts with few
environmentally suspect and tax-delinquent sites. As the number of sites per tract
increased, median household income decreased. Education levels were lower in tracts
with more potentially contaminated properties. These tracts also had higher
unemployment rates and lower labor force participation rates (Leigh and Gradeck 1996).

Andrew Hurley’s (1997) historical study of the abandoned Wagner Electric
Company site Wellston, Missouri, a neighborhood of St. Louis, supports the view that the
role of race in skewing environmental experience is far more complicated than is usually
acknowledged. His case study shows that “historic causes have more to do with deep-
seated structural inequalities than with overt racial discrimination in either the siting or
remediation process.” He found that real estate dynamics rather than discriminatory
siting decisions were largely responsible for bringing minority populations into this

brownfield community. The result was that 96.7% of the population in Wellston was

12

African American by 1990. Political disempowerment further sustained environmental
inequality.

4. Research on Scale

One important methodological consideration increasingly being addressed in
research on environmental justice is the appropriate scale or geographic unit of analysis
to be studied. Several studies have found that different size units produce different
results for the same area. Anderton and others (1994) compared analyses of data
aggregated to the zip code level with results of aggregating to the census tract level.
Disproportionate exposure depended on how the areas were defined. In 1995, Bowen
and others analyzed toxic release amounts with both counties and census tracts as the unit
of analysis for Ohio and Cuyahoga County (Cleveland) respectively. When the entire
state was studied, the correlation between minority concentration and toxic release
amounts was high, largely because industry, minority populations, and toxic releases are
concentrated in urban areas. When they studied Cleveland alone at the census tract level,
they found that toxic industrial facilities were more likely to be located in poorer and less
affluent areas than in areas with minority concentrations.

In the Cleveland study Bowen and others argue “smaller units of spatial
aggregation are more satisfying because they require more modest assumptions about
causal and statistical variations in local phenomena. Also these units tend to reduce
information loss regarding locational differences.” Larger units, such as counties dwarf
the size of what is being studied. In Bowen’s 2002 analysis of environmental justice
research he argues that studies should use geographical units of analysis that capture the

underlying spatial process involved in exposure. He found that studies that relied on

13

aggregate data to zip codes or counties did not capture the processes of interest. Studies
that focused on the census tract level were more appropriate. Monmonier (1994) also
argues that census tract areas are more appropriate than zip code analysis for analysis of

commercial transfer, storage, and disposal facilities.

14

VI. Federal and State Brownfield Redevelopment Policy: Economic

Development, Environmental Justice, or Both?

The Environmental Protection Agency (EPA) is the federal agency that is
primarily concerned with brownfield redevelopment. In 1994, the agency developed an
Environmental Justice Strategy in response to concern that minority populations and/or
low-income populations bear a disproportionate amount of adverse health and
environmental effects from brownfields. In this strategy they have conducted
environmental risk studies in communities where there are environmental justice
concerns and use GIS to identify geographic areas where sources of pollution appear to
have a disproportionate effect on minority, low-income, and educationally disadvantaged
populations. They co-sponsored a series of public dialogues on issues of urban
revitalization and strategies to create healthy and sustainable communities. They are
~ examining economic redevelopment opportunities to ensure that they complement
environmental justice. They are developing methods to expand public involvement in
siting and permitting (EPA 1996). Though the EPA is committed to environmental
justice in theory, the decision to invest in brownfield cleanup is driven by the economic
potential of the site, not the level of on-site contamination or the demographics of the
surrounding community (Hula 1999).

The EPA does not have authority to finance redevelopment of non-Superfimd
brownfields and has allowed state authorities to take a leading role in redevelopment
efforts. The state of Michigan has implemented a comprehensive state-level program to
encourage brownfield redevelopment run by the Michigan Department of Environmental

Quality (MDEQ). The primary focus is economic development. The MDEQ encourages

15

redevelopment by protecting buyers from liability for the cleanup and remediation costs
of previous contamination. They have created flexible cleanup standards related to the
proposed use of the land. Commercial and industrial standards are less demanding than
those for residential development. They provide public funding for brownfield
redevelopment projects through a bond of $335 million targeted directly to brownfield
redevelopment. Lastly, they allow Michigan municipalities to create brownfield
redevelopment authorities (BRAs) (Hula 1999).

BRAs create a specialized institutional structure to promote local planning and
implementation of brownfield redevelopment. Brownfield authorities can pay or
reimburse private or public parties for cleanup activities; lease, purchase or convey
property; accept grants and donations of property or labor fiom public or private sources;
invest the authority’s money; borrow money; engage in lending and mortgage activities
associated with property it acquires; and create revolving loan funds to finance projects.
Brownfield authorities develop a plan for redeveloping eligible properties within its
jurisdiction. Ifthe MDEQ approves the plan, they can collect funds through increases in
state and local taxes. Though the BRA has the freedom to develop any site, most develop
sites in which businesses and industry express interest. No BRA in Michigan targets
brownfields in communities that have a high number of brownfields for redevelopment
(Hula 1999).

Though the federal government, state government and localities are committed to
brownfield redevelopment, studies have found that brownfield redevelopment efforts are
not always compatible with the goals of environmental justice. The nature of brownfield

remediation has been found to be handled differently when it affects minorities and the

16

poor than when it affects other groups. Lavelle and Coyle (1992) analyZed every US.
environmental lawsuit concluded from 1985-1992 and found penalties against pollution-
law violators in minority areas were lower than those imposed for violations in largely
white areas. In their analysis of every residential toxic—waste site from 1980 to 1992,
they found the government takes longer to address hazards in minority communities than
white communities. In addition, they were more likely to accept less stringent solutions,
such as capping or other techniques that contain contamination at sites located near
minority neighborhoods or in minority neighborhoods. In white neighborhoods, they
were more likely to require removal. Hurley (1997) found evidence of such “white-
privilege” in his study of Wellston, Missouri. The contamination at the Wagner Electric
site in Wellston, a primarily African American community, was capped while the EPA
removed contamination and bought the homes of property owners in the neighboring

white community of Times Beach, Missouri.

17

VII. Summary

Past research shows that a higher percentage of minorities than whites live in
close proximity to hazardous waste facilities throughout the United States (GAO 1983,
UCC 1987, Bullard 1990) including the Metro Detroit area (Mohai and Bryant 1992). In
Milwaukee, minorities and the poor are more likely than others to live in neighborhoods
with a high number of brownfields (Leigh and Gradek 1996). In Michigan, communities
with Brownfield Redevelopment Authorities (BRAs) have a higher percentage of
minorities and poor than those without BRAs (Anderson and Clemens 1999). Hurley’s
case study indicates the reasons for these inequalities are often complex (1997).

Though the federal government and Michigan is concerned about environmental
justice, they admit that economic development is the primary concern. Leigh (2000)
claims this practice “creates the potential for widening existing inequalities between
better-off and worse-off neighborhoods.” In addition, Lavelle and Coyle’s 1992] study
showed the penalties against polluters were found to be lower and less stringent clean-up
of pollution was required in minority neighborhoods than in white neighborhoods.

Since few studies have addressed environmental inequality and brownfields, there
are many gaps in the research. Leigh and Gradeck’s 1996 study does not indicate
whether such inequalities exist outside of Milwaukee. Though Anderson and Clemens
(1999) studied brownfields in Michigan, the course scale of their analysis does not
provide insight into whether individual communities and neighborhoods are more likely
to experience the negative effects of brownfields. Leigh’s claims are based on
speculation, not evidence. Lavelle and Coyle’s (1992) analysis of brownfield

remediation is dated, only examine federal policy, and does not indicate whether such

18

practices have changed. No one has analyzed how well Michigan’s brownfield
redevelopment program addresses environmental justice concerns.

This study will expand on the current research to address some of these gaps. It
will show the distribution of brownfields within Metro Detroit at the census tract level to
determine whether some communities and groups of people are more likely to live in
close proximity to them. The research will also provide insight into the brownfield
remediation policies and practices of the state of Michigan by determining whether their

goal of economic development is also compatible with environmental justice concerns.

19

VIII. Data and Analysis
A. Data

Currently, Michigan does not have a complete list of brownfields within the state.
A number of Brownfield Redevelopment Authorities are in the process of creating
databases of brownfields within their community but have not completed the task. The
Michigan Department of Environmental Quality (MDEQ) has developed a database of
locations where Baseline Environmental Assessments (BEAs) have been performed and
then submitted to the MDEQ fiom 1995 to October, 2001. A BEA is an evaluation of the
environmental conditions that exist at a facility at the time of purchase, occupancy, or
foreclosure. It provides liability protection for new purchases, occupants, or lenders who
are foreclosing. Though this is not a comprehensive list of all brownfields (and may even
include some sites that do not fit the state’s definition of a brownfield), it is the largest list
available, and is what I use in the analysis to answer the first research questions. There
are a total of 3,747 unique sites used in this study for the entire state that were submitted
to the database from 1995 to October, 2001. 513 unique sites are in Wayne County, 311
in Oakland County, and 170 in Macomb County. Though each site is unique in size and
level of contamination, each site is given equal value in the analysis.

To answer the second set of questions I used a database of brownfield sites where
the state has allocated money for redevelopment or assessment fi'om 1995-2000. This
database, compiled by the Michigan Department of Environmental Quality, lists the
addresses for a number of proj ect funding types, including brownfield redevelopment,
site reclamation, revitalization revolving loan fund, site assessment, and waterfront.

There are 254 projects in the state at 248 unique sites. Thirty-seven of those projects are

20

in 35 unique sites in Wayne County, 12 in Oakland County, and l is in Macomb County.
Though the state may have invested different amounts of money and time into each site,
each site is given equal value in the analysis.

To make the data useable, duplicate sites from the database of Baseline
Environmental Assessments database and the Site Reclamation Project database were
first removed. Both databases list the address and county of each brownfield, but do not
list the census tract the brownfield is located in. The addresses of the unique sites were
then mapped using Tiger files in ArcView. These points were then matched to the 1990
and 2000 census tract they are located in. The accuracy of the placements was cross-
checked. In the first check, some sites where the specific community the site was located
in did not match the census tract assigned to it by ArcView. Those sites were located,
remapped and assigned the proper census tract where neccesary. In the second check,
sites that had different 1990 and 2000 census tracts were located. If an error had
occurred, the actual tract was reassigned to the site. For some sites, the address
information was ambiguous (i.e., a site located at the intersection of two roads where the
exact corner was not specified). When this happened, a portion of the site was assigned
to the tracts the site could be located in, under the assumption that the site is probably
affecting both tracts. When the site could have been in either of 2 tracts, each tract was
assigned a value of .5. When it could have been in one of 4 tracts, each tract was
assigned a value of .25. The end results were that 994 unique BEA sites were placed in
census tracts. Forty four out of 50 site reclamation projects were mapped and placed in
census tracts. The six that were not assigned census tracts had incomplete address

information.

21

Demographic and socioeconomic variables pertaining to the census tracts were
obtained from the 1990 census and 2000 census, as available. The census variables
chosen and analyzed were grouped into five distinct chategories. The 1990 variables end
in 90, while 2000 variables end in 00.

Group 1: Population

 

POPULATION90 Count of total population

 

POPULATIONOO Count of total population

 

 

 

 

Group 2: Class

 

 

 

 

 

 

 

 

 

 

MEDHHI90 Median household income

POVERTY90 Count of population in Poverty
Group 3: Race

WHITEOO Count of population that is White

BLACKOO Count of population that is Black

ASIAN 00 Count of population that is Asian/Pacific Islander
HISPANICOO Count of population that is Hispanic

 

 

 

 

Group 4: Race and Class

 

WHITEPOVERTY90 Count of white population in Poverty

 

BLACKPOVERTY90 Count of Black population in Poverty

 

 

 

22

 

Group 5: Housing

 

 

 

 

 

lhfiiDIANVALUE9O IMedian housing values

MEDIANRENT90 'Median rent

OCCUPIEDOO Count of occupied housing units

OWNEROO Count of occupied housing units that are owner occupied

 

 

B. Hypothesis

I hypothesize that:

(1) Minorities, specifically African Americans, and the poor are more likely than

whites and those with higher incomes to live in communities with

brownfields. A higher percentage of African Americans and the poor live in

communities with a percentage of brownfields. Those communities with

brownfields will also have lower average housing values, lower rents, and less

occupied and owner-occupied housing.

(2) Brownfield redevelopment efforts are positively related to corrrrnunities that

have a lower percentage of minorities and are less economically distressed (as

measured by income, housing values, and percentage of occupied and owner-

occupied housing).

C. Research Objectives

There are two primary research objectives in this study. The first is to determine

whether exposure or proximity in regard to the location of brownfields, exists in Metro

Detroit for certain population groups. Specifically, the study will determine whether

minorities and the poor are more likely to live in communities with brownfields, and

23

 

whether their spatial distribution matches the spatial distribution of brownfields. It will
also determine whether communities with brownfields have different median housing
values and rent, and percentage of occupied and owner occupied housing than those
without. The second objective is to analyze whether brownfield redevelopment projects
and money is equitably distributed across the area, and whether these efforts serve to
address inequalities that may be present.
D. Methodology

To determine the first objective, I used exploratory and statistical analysis. I used
exploratory analysis on the first four groups, population, class, race, and race and class to
determine the percentage of total population living in tracts with brownfields for all of
my study areas. Two statistical tests, the Difference of Means test, and Index of
Dissimilarity were also used to measure whether some groups experience unequal
exposure to these brownfields.

The two-sample Difference of Means test compares two populations with one
another to determine whether the two groups are statistically different from each other.

The two-sample t test for unpaired data with unequal variances is defined as:

 

r=Z—Z/(Js,2/N, +s,’/N,)

with degrees of freedom:

v = (sf /N. +s.’ /N.)2 «cf /N.)2 /N. —1)+<s.’ /N.>/N. —1>>
where N1 and N2 are the sample sizes, Y1 and Y2 are the sample means, and $21 and 822

are the sample variances. The test statistic T was used to determine the p-value using the

two-tailed table. Significant differences are those with a p value of less than or equal to

24

.05. Before calculating the means for Y1 and Y2, I transformed many of my variables.
The variables POVERTY, WHITEPOVERTY, BLACKPOVERTY, WHITE, BLACK,
ASIAN, and HISPANIC were divided by the total population in each tract to determine
the percentage of total population each of those variables were in each tract. The variable
OCCUPIED in each tract was divided by the count of total housing units in the tract to
determine the percentage of occupied housing units. The variable OWNER was divided
by OCCUPIED for each tract to determine the percentage of occupied housing units that
are owner occupied. The variables POPULATION90, POPULATIONOO,
TRACTAREA, MEDIANVALUE, and MEDIANRENT were not changed. The mean of
all the variables were compared for tracts with no brownfields to those with one or more
brownfields.

The index of dissimilarity D, is used to measure the degree to which population
groups by race and class and brownfields are distributed evenly within a metropolitan
area by looking at the location of brownfields. The test is defined as:

D= 0.5 23 I xi-yil where, for this test:

xi = the percentage of total study area’s population group by race and class in a given
census tract;

y,- = the percentage of the study area’s brownfields in the same census tract; and.

D = the index of dissimilarity (Darden & Tabachneck, 1980; Duncan & Duncan, 1955).
The value for the index of dissimilarity D can range from 0 indicating maximum
brownfield exposure to 100 indicating minimum brownfield exposure. The lower the
index of dissimilarity, the higher the level of brownfield exposure for a particular

population group. I analyzed all tracts with brownfields for each of the study areas

25

because my purpose was to determine the unevenness in the spatial distribution of
brownfields compared to the spatial distribution of certain population groups.

To determine the second objective regarding the equitable distribution of
brownfield redevelopment projects in Michigan, I determined which tracts have received
state money for brownfield redevelopment and which have not from the database
provided by the MDEQ. I first compared the differences in means for all variables of
tracts where state money has been spent to all other tracts with brownfields where no
state money has been spent. The same tests, transformation of data, and parameters as
used for the first objective were used.

E. Unit of Analysis and Study Area

Analysis was performed on the 1,088 census tracts in 3 of the Counties, Macomb
County, Oakland County and Wayne County that traditionally have composed metro
Detroit. Census tracts were chosen for a number of reasons. First, an environmental
justice study in the Cleveland area measured environmental equity at both the county and
census tract level. From their analysis, Bowen et a1 (1995) found that environmental
equity research is not amenable to a county-level or city level analysis due to the
courseness of spatial association at such a large scale. Analysis using smaller spatial
units, such as census tracts, was more appropriate. Second, census tracts were chosen
since demographic information from the US. Decentennial census is readily available at
the level. Third, analysis in Michigan at the city level by Anderson eta12000 study has
already been done.

Lastly, to determine which unit of analysis is most appropriate for brownfields, I

looked at the impact brownfields have at a variety of scales. The state or federal

26

government is often responsible for allocating money for clean up. In Michigan, the

County or Municipality organizes Brownfield Redevelopment Authorities. However, it is

the individual neighborhoods and communities that surround a brownfield that are

impacted the most on a daily basis, and is the level this study is primarily concerned.

Since census tracts are defined in part by homogeneity of housing and population within

communities, they are an appropriate unit of analysis when doing research on the

demographics and neighborhood characteristics of brownfield communities.

Analysis was done first for combined counties of Macomb, Oakland, and Wayne.

To get an idea of what was happening at a smaller geographic level, analysis was also

done for the sub-areas of three counties individually and for 10 individual municipalities.

The municipalities that met the following criteria were chosen: 1)The percent of people

living in poverty or percent of non-white population was above the county average; 2)

The municipalities had 7 or more census tracts; and 3) There was sufficient variability in

the number of BEAs per census tract (i.e., at least one tract had no BEAs and at least one

tract had 3 or more). The municipalities that met all three criteria are in grey on the map

and include:

 

Macomb County’s

Oakland Coung’s

Wayne Coung’s

 

 

 

 

 

 

Clinton Township Pontiac Detroit
Warren/Centerline F emdale Dearbom
Southfield Romulus
Westland
Wyandotte

 

 

 

27

 

Figure 1: Map of Study Area

 

Oakland County

Pontiac

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Rom UI U S ’ ”gage;
Wyandoj . tte

Wayne County

 

 

28

Macomb County

    
 

Clinton

Westland Dearbom . '/

 

IX. Findings
A. Question I :

The findings for the tests used to answer question 1 are displayed in seven tables.
Table 1 shows the results of exploratory analysis. Tables 2-6 shows the results of the

difference of means tests. Tables 7 shows the results from the Index of Dissimilarity test.

29

.532» no can: 2 2880a E ounBobE 05 ".7...

$32» no can mm 2.58% E ooaobte 05 u...

$285505 5:» Bomb E wE>=
cos—£38 =Bo>o 55 wEomEpoE :23 Bomb 5 wE>= Bacon x we owfiaouaoa .63on a BEBE momeoBoa BEE—FEE fixes.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

wcm 1th ad» *odm ode .4me Qwv Wow now 3.25:
ed affine Wmv cdw www.mn Q3 Inn» m? 03 283%.:
to.8_ is. E «NW Em 28$ can :35. Ba we. 52.5:
v.0 was Qua o.m _ mi m. 8 oém EON 3cm 25:83 83:36:...2
osm 5.; *wév mdm 56m rmév. hum w.wm Ywm :86: 35:5 «55$
3.2. 3.3 {now $.59 mm w v. S Ede m. E wdo 235.....— 85:35.52
Nam mam n? a? 2:. Se mam mow 2.9. 22:38 . Beau
*véh _ .vo Law ads _ db «.2. «.3 WE ode 2:25.— 2::an
, 85.“ 6:52
...NHw wamdv 93.. ...wdm 16.9. m.mm $.19. imm odm 3:..8595ta3 .3550
1.23 ENS msm «*mgm .3..th N.wm 1a.? wdv 21v aim—Eek :SED ..Eean
ER *9: .5.? ham w.wm N. _ v now Yes Exam 5580 a???»
".3..de ndm Swim». ..iadn Sade cfim *mdv 93m wém .3550 2.2an 8::on
*3de thin adv mdv 19$ 9% *amdm ode v.3 3550 :5..an 333:...—
mozuaoo a: «3 1::
mdm *Ndn and”. m. 3 o. 3 nmv «.3 ES v.3 6:2an £5832 8:580 =<
co co 0
>Hmm>0m >Hmm>0m Q>me>0m ooZO ooZOE.
MU<qm ”+523 ooU_Z<mm_: ooZ<_m< Gov—Olefin camp—:3 4<HOP EKJDQOL «JDAOQ bEsEEoU
8.59 859 BuSENmK
33 week ..V 33.5 35‘ ..m 53.5 ..N macaw ..~ 53.5

 

 

34H: 53 39s.: 5 war:— nous—.52. .88 we oufisoaom "u 039—.

30

Table 1 shows the percentage of total x population (in columns) for the study area
(in rows) living in tracts with BEAs. The percentages that are greater than the percentage
of total population living in that community for the respective year are highlighted,
indicating overrepresentation for that specific group. One asteric indicates a difference/
over representation of 5% or more; two indicates a difference of 10% or more. Based on
the hypothesis that minorities and the poor would be more likely to experience higher
exposure to brownfields, it was expected that there would be over representation for
people living in poverty, White and Black people living in poverty, and Blacks, Asians,
and Hispanics; and under representation of Whites. Over representation of minorities and
the poor and under representation of Whites would indicate this expected inequality.

The results for All Counties do not show the expected inequality. Only Hispanics
and Whites living in Poverty are over represented by more than 5% in tracts with
brownfields than the population as a whole.

The results for the Counties individually indicate County differences. Macomb
and Oakland County have largely similar results that demonstrate the expected inequality
in most areas, except for the case of Asians and Hispanics in Macomb, and Whites in
Oakland. The results show inequality in Wayne County in regard to class and race and
class, but not for race alone.

The results for the municipalities also indicate city differences. Macomb’s
Clinton Township and Warren-Centerline, and Wayne County’s Dearbom demonstrate
inequality in all areas. In Pontiac, Femdale, and Wyandotte, there is inequality in regard
to class, but not race. In Detroit, Southfield, Westland, and Romulus, the results indicate

very little inequality.

31

There are contradictory results based on the size of the community studied. The
most inequality is found at the individual County level, where all three Counties
demonstrate inequality in class, and the two Counties that have a small Black population
(Macomb and Oakland) show inequality in regard to race. Though inequality is present
in the counties separately, when they are studied together as a region, none of the
inequality is visible. This loss of information indicates that inequality is more difficult to
discover at a regional level. Analysis at the individual community level provides greater
clarification than at the county level by providing insight into which communities do
have inequality (contributing to the overall County assessment) and those that do not.

There are an equal number of communities (six) with Whites over represented as
communities where Blacks are over represented. While the count is the same, the degree
to which Blacks are over represented is much greater than Whites. Whites are only
represented by more than 5% in the case of Detroit and Romulus. Blacks are over
represented by over 10% in all six communities. In Oakland County, where both Blacks
and Whites are over represented, Whites are over represented by 3 % while Blacks are by
more than 30%.

Another important consideration is the fact that there are many ethnic minorities
in today’s society with non-European ancestry that are classified as White in the United
States census. One specific group in this study is Arab-Americans living in Dearbom,
Michigan. In Dearbom, the two tracts with the most BEA sites (7 and 4 respectively)
have the highest percentage of population with Arab ancestry (68% and 48%), lowest

median incomes ($15,000 and $20,000) and highest percentage of people living in

32

poverty (45% and 27%). Thus, caution is advised when interpreting the results of the

analysis for Dearbom.

Tables 2-6: Difference of mean test comparing tracts without BEA sites to tracts

with BEA sites

 

Key for Tables 2-6

 

Meanl: Mean value for tracts without BEAs

 

Mean2: Mean value for tracts with BEAs

 

 

p-value: p-value for Difference of Means test
between Meanl and Mean2 (highlighted values

have a p-value of .05 or less)

 

 

Table 2: All Counties

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Meanl Mean2 p-value
Group 1: Population POPULATION90 3538.2 3673.6 0.165
POPULATIONOO 3428.8 353 1.5
Group 2:Class TOTALPOVERTY9O 14.1 15.7 .
MEDHHI90 41003.5 37069.8 ,;..,,;;:
Group 3: Race WHITEOO 66.7 65.2 0.507
BLACKOO 27.8 28.4 0.817
ASIANOO 2.3 2.3
HISPANICOO 2.5 3.4
Group 4: Race and WHITEPOVERTY90 4.3 6.3
Class BLACKPOVERTY90 9.4 8.8
Group 5: Housing MEDRENT90 502.8 474.9
MEDVALUE9O 71433.2 62094.4
OCCUPIEDOO 94.5 93.4
OWNEROO 73.8 65.9

 

 

 

 

33

A
Lommv

fimvfimm

 

82.55 33.52: a 2.3

34

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

was wow oboo moo mi. oommZBO
moo v.3 ammo Ono woo oonEDUUO
NEW? @er $8 $88 383 osmB<>om2 wees:
mots mdmm vono $me msmm oohzmmnmz ”m 980
no no 9: .o M: .o.o oo>kmm>OmMU<Am mam—U 23
Rod 3 2. mafia 3 an oofimm>oamtm3 83. a. 9.20
Boo 04 No God o4 mg 8092me
Sod 9m Nd :ld md m4 ooZ<~m<
Soda Na 2 «.35 3 3 855m
mood flow so ”Nod v.3 nmo oombg comm mm 955
«So anmom meow: mmoo odmmom osmnov ooimamz
mmoqo ms was hood to o.m oo>kmm>OmA<HOH mam—Una 380
Bed onmm osvom mono ogmmv mince ooZOE<ADmOm screamed
3nd msme v.39” mead «so: momma ooZOF<ADmOm ; 9.80
nos—go N582 2822 nigh N582 2:32

 

 

05—33095me—

QEmEsoH c955

 

 

 

moan—ammomuzz .3580 .15an 3 ~35.

35

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Sued mdo Q: emwd v.3 Nmm mowd mém WE comm—Z590
mnmd New wdm and 03 Ema bmmd mda fig con—WEDUUO
mnvd adoomm odofw mwod Seaman odmoww Svd Wmenm Ohm—Q. oom54<>am2 mausom
wood N63 Womv mmbd 3‘3 odao 8md ONE» Doom ooHZmMn—mz ”w 9.80
$20
$06 md Nd wmmd wd v; _S .o 0.2 ms oarhmmN/Omvaaim can 8am
9:3 @6— Ns 3 md Wm o6 wood 2: Z: om>bmm>0mm§$>> ”v @590
momd 9N v; 526 v; : 366 M22 fin ooUHZ<mmE
29o mg m; Sued Wm wN mmwd m.N ON ocz<~m<
ommd 1m Em @506 new 56m $56 NS v.9 ooMU<4m 8am
mend ado odo mend 9.5m Qmm mmmd Yam 0.? COMP—I3 “m @580
ovho NAmwNm o.mw_mm mad ”Ev? Wham—m Sod mega fivvowm ooEEQNE wan—UN
936 mg _ as 3; .o 5“ 0.0 med wfim No— oowhmmsonfiflrpw 9.80
03.6 3:8 mdwmv awed oémmm 05mm ~36 Dom; OBNVN ooZOEkflEOm cosflsmom
20o v.3”; mvov ahwd 563m wdim 685 5.2.3 ZEN ooZOF<ADmOm ; macho
o=_m>-m N532 332 023$ , Name—2 ~ 532— o=_a>-m Name: 2822
036.com Eocfizofi 23:8

 

 

 

 

 

8::«3352 5.55 @5an 6 «any

36

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

08.0 mac o. H w coed vow v. H 5 head ofiv com commZBO
owmd mdo ”mo momd Kano ado NH fl d ddw vow ooQMEDUUO
vomd odoooo dddmmo owmd odmnmo mowdem mmmd N. ~ me~ _.voomN oomDA<>QmE mango:
ovmd o; E odom vad fidvv #th Rod monm N.mwm ooEZmMsz “m ode0
$20
oEmozoo< 82 End md m4 osd odm o.\lm do>Hmm>OAMU<Am can 08%
vmdd 92 N4» hmmd vd 9m vomd v6 m6 do>Hmm>Ommt§ ”v HEB0
El d Ym wN owed Wm Nd dead 06 o.m ddU~Z<mmHE
chd o; o; emvd o. _ Wm Rod _._ od OdZ<Hm<
oomd ON NA mood HS dd hood flow o.mw ddM0<Am 8mm
3 fid m.mw _o ode mfiw Wow mood m4: odd domrEEB nm A390
00— d wfimwhm Ndwbwv 05d fiwwdom Nflwm ocod Ndommm comma doHIEQmE mam—UHN
vmdd 0.2 m6 mwvd No m8 Ewd mgmm «.mm do>Pmm>qu<HOE o380
odbd m.womv oomom 55d cdfl m OSwom Nmmd momom mdwdm ooZOEkJDmOm coca—smog
and ~6de wdvnm vvmd odmom v.3; wad ode?” \Lcmm doZOF<ADmOL ; 9.80
o:_m>-m N532 :32 “vs—go, N532 2:82 u=_m>-m Name—2 2:32
50%on 95:83 :8qu

 

 

 

 

 

mam—36352 5580 a??? use flash.

 

37

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$3 25 $5 2 S «.8 haw comm—230
end :5 5.8 $3 5.3 955 858580
593 38$ 3%: 85.5 8m? 28? 553.5555 9:95:
:3 25m 3.? 23 58m 38 89255: ”n 955
and 3. 3 558993 52 535555933 3.90 95
83 n: 55 5.: 3 5599555953 82 5 955
2:5 2 3 MG 5.». 822955
:55 no to «$5 to no 8252
83 5.2 55 ~85 mo 3 853m
:3 5.2 ~59 ~86 5.8 25 SEE? 85m ”m 955
£2 58% 322 25.5 395 38: 85:52
553 «.2 5.2 M55 2 3 55955051909 5555 955
and 35m imam «$5 R: 5.88 82092555 595595
52.5 022 5.8% 52.5 new? 25% 552221555 9 955
0335 «:32 2:82 0395 N582 2822

3383: 885:5»?

 

 

mvzm—aamumnaz 5550 2:55 :5 035,—.

38

Before doing the difference of mean tests (Tables 2—6), the data was transformed.
All three variables that looked at poverty were divided by the 1990 population to give the
percentage living in poverty per census tract. The 2000 race variables were divided by
the 2000 population to give the respective percentage of Whites, Blacks, Asian/Pacific
Islanders, and Hispanics living per census tract. In the housing variables, the count of
occupied housing was divided by the total count of housing per census tract, while the
count of owner occupied housing was divided by the count of occupied housing units per
census tract.

There are significant differences for the entire study area, All Counties (Table 2)
in Class, Race and Class, and Housing. Specifically, Median household income is lower,
there are more Whites living in poverty, and lower median housing values in tracts with
BEAs than in tracts without.

Within the Counties studied individually (Table 3), there are many similarities. In
all, there are significant differences in Class (tracts with BEA sites are poorer), Race and
Class (tracts with BEA sites have more Whites living in poverty), and Housing (in tracts
with BEA sites, rents and values are lower, and there are smaller percentages of owner-
occupied housing). In addition, Macomb County has a significantly smaller percentage
of Whites and greater percentage of Blacks and Blacks living in poverty in communities
with BEAs. Oakland County has a significantly higher percentage of Hispanics. Both
Oakland and Wayne County have significantly less occupied housing in tracts with BEAs
than in those without.

The Macomb County communities (Table 4) of Clinton Township and Warren-

Centerline have significant differences in Class (tracts with BEAs are poorer), Race

39

(there are fewer Whites and more Blacks in tracts with BEAs), Race and Class (there are
more Whites living in Poverty in tracts with BEAs), and in Housing (the median value of
housing is lower in tracts with BEAs). Oakland County municipalities (Table 5)
and Wayne County municipalities (Tables 6a and 6b) have few significant differences. In
Pontiac, there are significant differences in population (the population is greater in tracts
with BEAs), and in Race (there are more Hispanics in tracts with BEAs). In Wyandotte,
there are significant differences in Class (tracts with BEAs are poorer), Race (there are
fewer Whites in tracts with BEAs), and in Race and Class (a larger percentages of Whites
live in Poverty in tracts with BEAs).

Though the remaining results do not show significant differences, there are certain
trends that can be seen in Tables 4-6. In most municipalities, tracts with more people
living in poverty with a lower median household income have BEAs. Many communities
have fewer Whites and more Blacks. Exceptions are Southfield, Detroit, Westland, and
Wyandotte. All communities but Southfield have a greater percentage of Whites living in .
Poverty in tracts with BEAs than in those without. More communities, such as All
Counties, Southfield, Detroit, Westland, and Romulus have a smaller percentage of
Blacks living in tracts with poverty.

As expected, all municipalities but Femdale have lower rent, and all communities
but Westland and Romulus have lower values in communities with brownfields than
those without. There is also less occupied housing in all communities, and less owner-
occupied housing in all communities but Pontiac in communities with brownfields than in

those without.

40

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

QS Nwm 5.3 _.mm Yam 2.1.83—
v.? 9mm 50¢ 9mm w.mm Sue—Ea»?
w.mm m.wm 1mm tom 5.2.. 52:39
OSN NdN 9mm 5.5 hNN can—«v.95 meg—533.52
Mfl mm 5.9, 9.3, m? .553 3.5.5 2:55
adm vfim a. a N.m~ 5.8 23:59 82—53252
5mm adm m. fl m # .om mdm EELS—Em 5.550
54% 54% 5.8 vdm 5.0.5. 25.59 .52an

2:13:90 mag—2.53:2
mdv oém cam adv 0mm v.3 Tmm QNm .5.—53> 3550
mm cm QR Ném mNN fiwm Nd— QVN 35:33. 5.5.0 aEcoaS
MAM Harm wéw WON E 5.9“ hum v.5“ 5580 2:83
mdm 5.? odm marmw 0.5m mdv 06m wém fiasco 2.2an 8:550
9mm Won 0.3 mam v. fl m mém gem flow. 55.50 :5..an 3:231:—

lt 8:550 “Emma
«Ea 6.53.90
an. 5.3. me new film 39 :9 98 Rm 5585: 5:955 =<
>Hmm>Om >Hmm>Om DE >Hmm>Om 829 320
v8.35 E {imam ZSm< Mufim BE A<HOH 535909 5215909
2.59 3.59
ES» 35‘ ..V 5.36 REE ..m. 58.5 .N 3.36 =§=§m ..~ 53.5

 

 

25... 9:25:55 .5 555 2: .8 825 a K 5359

41

Table 7 gives the D values from the Index of Dissimilarity Test. The test
indicates whether the distribution of the variables x (in the column) is similar to the
distribution of BEA sites. D can range from O to 100. The lower the value, the more
similar the distribution, indicating exposure to brownfields for the specific population
group. Values greater than 50 indicate a high dissimilar distribution, i.e., a lack of
exposure to brownfields and are highlighted in the table. Since the values are dependent
on the size of the spatial units used, the values can only be compared for variables
analyzed using spatial units of comparable size.

In All Counties, the distribution of BEAs is similar to the distribution of the 1990
and 2000 population. The distribution is similar for Group 2, Class, as well as the
distribution of Total poverty is somewhat similar to that of BEAs. In regard to race, only
the White population has a D value less than 50. When race and class intersect, only
Whites living in poverty have a similar distribution with a D value of 44.7.

The results for the Counties of Macomb, Oakland, and Wayne County are similar
to the results for All Counties with a BEA distribution similar to the 1990 and 2000
population, and individuals living in poverty. The results for race and the intersection of
race and class show County differences. The White and Hispanic population, and Whites
living in Poverty in Oakland and Macomb County are similar to that of BEAs. The D-
value for Asians in Oakland County is also below 50. In Wayne County, none of the D-
values for race were below 50. In addition, for all three Counties, the D-value for Blacks
living in poverty is above 50.

The results for the individual municipalities also indicate that the distribution of

BEAs and the 1990 and 2000 population are similar. As expected, the distribution of

42

BEA sites and persons living below poverty is also similar (D values range from 14.7 in
Romulus to 46.4 in Warren-Centerline), indicating that the poor are likely to live in tracts
with brownfields.

When race is examined alone, the results for Whites were unexpected. In all
communities, except for Detroit which does not have a large percentage White
population, the distribution of Whites and brownfields is similar. Only Blacks living in
Pontiac, Southfield, Femdale, Detroit, Westland, and Wyandotte are more likely to live in
communities with more brownfields than Whites are. In addition, the distribution of
Blacks in Clinton Township, Warren-Centerline, and Dearbom is similar to the
distribution of brownfields, though not as much so for the Whites. The distribution of
other non-White groups studied, Asian Islanders and Hispanics, does match the
distribution of BEA sites in many communities, as hypothesized.

When poverty and race intersect, the distribution of Whites living in poverty is
similar to the distribution of brownfields in all communities Detroit. In all of the
communities that have a larger percentage of Blacks distributed throughout the
community (Pontiac, Southfield, and Detroit), the distribution of Blacks living in poverty

and brownfields is similar.

43

B. Question 2

The findings for the Difference of Means tests used to answer question 2 are

displayed in Tables 8-11.

Tables 8-11: Difference of mean test comparing tracts with BEA sites

with projects to tracts with BEA sites without projects

 

Key for Tables 8-11

 

Meanl: Mean value for tracts with BEAs without redevelopment projects

 

Mean2: Mean value for tracts with BEAs with redevelopment projects

 

 

(highlighted values have a p-value of .05 or less)

p-value: p-value for Difference of Means test between Meanl and Mean2

 

Table 8: All Counties

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Variable Meanl Mean2 p—value

Group 1: POPULATION90 3712.7 3259.0 0.075

Population POPULATIONOO 3563.8 3240.2

Group 2:Class TOTALPOVERTY90 15.0 21.9
MEDHHI90 37839.8 28939.2

Group 3: Race WI-IITEOO 65.5 65.8
BLACKOO 28.2 26.4 .
ASIANOO 2.3 1.2 :
HISPANICOO 3.2 6.0 0.167

Group 4: Race WHITEPOVERTY90 6.1 9.6

and Class BLACKPOVERTY90 8.5 10.8

Group 5: MEDRENT 90 479.1 417.2

Housing MEDVALUE90 64282.2 37339.4
OCCUPIEDOO 93.5 92.3
OWNEROO 66.6 58.9

Group 6: BEA BEA90 2.1 2.4

Count BEAOO 2.0 2.4

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.0 0.0 5.0 0.0 00.000 2500
0.0 0.0 5.0 v.0 00<mm <00 ”0 0:80
0.50 0.00 0.00 v.05 00000230
0.00 v.00 v.00 v.00 000300000
3.8% :33. .H. 380m 388 85.3582 @380
0.5; 0.0? 0.500v 5.550 000200005 H0 0:80
5.00 V00 #00 .0 0.3 0.0 00>00m>00v00§m $30 05
v.00 0.0 000.0 5.5 v0 00>Hmm>00mH§ 0802a 0:80
0.5 0.0 050.0 0.0 0.0 0000720000
0.0 0.0 0.0 5.0 002504..
. 0.00 0.50» 000.0 0.00 v.00 0300.040
...m. v.00 0.0V 5000 0.00 v.00 00.00.0003 08.0 H0 0:80
000 .0 0.00000 0.00000 % 0.00500 0 .00N0v 000.000va
050.0 0.00 0.0.0. 000.0 5.00 0.0 00>00m0>00A<HOH mam—00 0:80
000.0 0.300 0.300 005.0 0.0000 0.0050 002000015000 5023000
0000 0.0000 0.0000 000.0 0.0000. 0.0N0v 002000230000 “0 0:80
02570 0:32 0532 03?-0 0532 0532
2:33 0:200

 

 

855.5 33.5....— "a 2.3.

45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm; m: _o.~ «8.: m 3 8<mm :58
mm; m: .3 one: 0 ca oa<mm <mm a 955
$3 4.: So . 0.8 30 Emmi/o
on; 0.8 08 and 4.3 08 8:80:80
83 :68? 3.28 . . ._ @829 338 combi>nm2 mason
$3 35. 5%. «.8: :02 was. 895:8: ”m 955
30.: no No 23 3:. ma om>emm>o§o<qm v.36 Ea
$3 «.0 2: :3 M3 0: 892850953 833. 985
23 E 3. $3 E 4.2 SBZEmE
an}. N: m: 8...... 3 3. 8252
«who 00 mm . ,. 3.x ed... 855m
33 3a 0.8 v ,o E: «.3 SEE? 833 955
Sod c.0800 0.:va 0mg 38: $38 SEQ:

. 2: q: N8: 000 0.8 899mm>05<909 $200 955
269. 0E: and 03$ 2%.... 8200950000 83:93“:
new? 5%: 30¢ 3.30 :05. 829950000 : 955
02?-0 0:32 0:32 0.5—«>0 0:32 0:32

DEF—om 0002.50

 

 

80599.52 0.55 :5an a: use...

46

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mwfid md m.m mmnd 5N 0N madd m4» fin dd<mm 2500 <mm
m3d md Wm Sad EN 9N hmdd Ev dd da<mm no 9.80
maod mdc v.2. dmvd Nab ode mm M d v.Nm :3 ddmmzao
wmad Qma Ema End ma Eda wed mdw fiww ddeEDUUO
an ddmwwv dddmmv _o_ d ndcdmm dddw: bland ”mmmdfi WNW Sum damDA<>DmE 9:303
ded dddv Wm; wawd 0.5m mddm mmod mdmv vdbm daHZmMDmnz nm @280
538:5? 52 258.53 52 fig 55 55.5. 325
dawhm§0§0<qm can 08%
Rod 0: v.3 wood 0.3 5.: ddmd a.w Wm da>kmm>OmmE “v @380
End a.~ EN wmmd d.m Wm vomd d.a~ dd dd0~Z<mmE
haad vd vd cbdd ad Nd amod v; o; ddZSm<
mad vd ad m9 d d.“ MN dd _d v.3 mdw ddMU<Am comm
Svd a.ma ~Vma vwnd dew flow 2 _d wdm We do; Hm a=80
mvmd 9:1me mdwmam vavd fimammm fidmvam wand mdwmdm ddthN 8%ng $30
dmod fig a.E Smd WC NS bond ddv ddm da>Hmm>On§<HOH ”N 9:80
wmvd m.$©m Ewmwm nwmd 2mm w.wmwm 85 5mm: :uddm ddZOF<ADmOm confismom
havd déddv danmm Namd 5am: dganm 93d mail dag m daZOHH/EDQOA ; A“:80
u:_m>-m N582 2:52 0335 N532 2802 _o:_a>-m $822 :32
0:8:ng =53:on— :8qu

 

 

 

mug—«38:32 5550 235$ "2 ”Sub

47

Tables 8-11 gives the results for difference of mean tests where the mean of the
variables for tracts with BEA sites where state money has been invested in
redevelopment, assessment, etc. was compared to the mean in tracts with BEA sites
where no state money has been invested. Significant differences would indicate that the
state has invested in communities that are significantly more likely to have a higher
percentage of population with certain characteristics. Some communities studied in
previous tests do not have any redevelopment and were not analyzed in this test.

The results indicate that in the entire region (Table 8), Class and Housing plays a
role in where the state has provided redevelopment, as poorer areas with lower housing
values were more likely to receive state money. The race of the community does not
seem to matter, except for Whites living in poverty.

The results for the Counties individually (Table 9), demonstrate County
differences. In Oakland County, Class and Housing differences are significant as poorer
areas with lower housing values were more likely to have state projects. In Wayne
County, Race plays a role, while class does not. Tracts with a greater percentage of
Whites and a lower percentage of Blacks are more likely to have state investment in
redevelopment. Significantly more Whites living in poverty are in tracts with
redevelopment than those without.

Only the individual municipalities (Tables 10 and 11) of Pontiac, Femdale, and
Detroit show significant differences. In Pontiac, Class, Race, and Housing must be
considered. Poorer tracts with smaller amounts of Whites and greater amounts of Blacks,
with lower housing values and fewer owner occupied units are more likely to experience

redevelopment. In Femdale, the differences are only in regard to population, as more

48

people live in tracts where redevelopment has occurred. In Detroit, there are fewer
people in those tracts, and significantly lower housing values.

The remaining results do not have significant differences. However, trends can be
identified. Most communities have fewer people living in tracts where the state has
invested in redevelopment. In most communities, except for Femdale and Wyandotte,
the tracts have a poorer population. Like Wayne County, All Counties, Detroit, and
Wyandotte have more whites and fewer Blacks in tracts where the state has invested in
redevelopment. Like Pontiac, Oakland County has fewer Whites and more Blacks in
tracts with redevelopment. All of the communities but Wyandotte have a higher mean
of White Poverty and/or Black Poverty in tracts with BEA sites than in tracts without.
Most of the communities have lower median rents and/or lower median values in
communities with redevelopment sites than in those without. Femdale and Wyandotte
demonstrate the opposite. All communities except for Femdale, Wayne County, and
Dearbom have less occupied and owner-occupied housing in tracts with BEA sites than
those without. In all communities but Femdale and Wyandotte, there are more BEA sites

in the tracts where the state has invested money than in tracts where the state has not.

49

 

IX. Discussion
A. Does environmental inequality exist in Metro Detroit?

The first objective of this study was to determine whether exposure to the location
of brownfields exists for certain population groups in Metro Detroit. The results indicate
that in most of the communities, exposure is greater for some population groups and less
for others.

The following table shows these results. For the first three factors (Race, Class, and
Race and Class) YES means that exposure is greater due to the factors specified
according to two of three tests. Specifically, the percent .of the specific population living
in tracts with brownfields is at least 5 percent greater than the percent of population
living in tracts with out brownfields; and/or the p value is less than .1 for the difference of
means test; and/or, the D value is less than 50. For the fourth category, Housing, YES
means that at least one of the Housing factors had a value p value of less than .1 in the

difference of mean test.

50

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Community Race Class Race and Housing
Class
All Counties Macomb, YES YES YES
Oakland, and (White)
Wayne Counties
Individual Macomb County YES (Black, YES YES YES
Counties Hispanic) (White,
Black)
Oakland County YES YES YES YES
(Hispanic) (White,
Black)
Wayne County YES YES YES
(White)
Macomb Clinton Township YES (Black, YES YES YES
County Asian) (Black)
Municipalities Warren- YES (Black) YES YES YES
Centerline (White)
Oakland Pontiac YES YES YES '
County (Hispanic) (Black)
Municipalities Femdale YES YES
(White,
Black)
Southfield
Wayne Detroit YES
County Dearbom YES (Black) YES YES YES
Municipalities (White,
Black)
Romulus YES (White)
Wyandotte YES (Black) YES YES
(White)
Westland YES

 

All communities but Southfield demonstrate some level of population exposure in .

regard to the location of brownfields. In Macomb County (both the entire county and

individual communities), the unequal exposure to brownfields is severe. Minorities and

the poor are both more likely to experience the negative effects of brownfields than

Whites or those with higher incomes. In Oakland County communities, the poor and

51

 

 

Hispanics experience environmental inequality. In Wayne County communities, the
results are mixed.

Though inequalities do exist, the only clear pattern that exists is in regard to
class, regardless of race, and housing, as poorer communities with lower housing values,
rents, owner or occupied housing are more likely to have brownfields. A clear pattern of
Black disadvantage is not present outside of Macomb County.

The absence of a pattern may be in part a reflection of the data available for study.
BEA sites are properties that are assessed when being purchased, changing occupancy, or
being foreclosed. These are potentially contaminated properties with the highest
economic potential, and would most likely not be found in the most depressed
neighborhoods. In a 1991 assessment of a sample of BEA sites filed from 1995-99, Hula
et al found that of the 300 observed sites, 203, or 69 percent, showed some level of
economic activity. Only 14 percent were reported as vacant or abandoned. These
findings indicate that there may be many brownfields, particularly those in communities
like Detroit that are overrun with abandoned properties that have little commercial
interest to developers, that remain undiscovered and have not yet been assessed for

contamination.

52

 

B. Have the state ’s efi‘orts to redevelop brownfields been equitably distributed? Do state
redevelopment efiorts directly address inequalities that exist?

The second objective of this study is to determine whether Michigan’s brownfield
redevelopment policies have been able to link the values of environmental justice and
economic development. The analysis of environmental inequality in the area does
indicate that some groups are more likely to experience the negative effects of
brownfields. Brownfield redevelopment efforts have the potential to correct or
perpetuate those inequities.

The following chart describes the demographics of the communities where

 

redevelopment has occurred. “NO” indicates that for the specific variable, the mean is
significantly lower in communities with brownfields where redevelopment has occurred
than in those where they have not. The implication of “NO” is that redevelopment efforts
have not been equitably distributed for the specified group. “YES” indicates that there is

a higher mean percentage in communities with redevelopment than those without.

53

 

 

 

 

 

 

 

 

 

 

 

 

 

Community Race Class Race and Housing
Class
All Macomb, NO YES (poorer) YES (White) YES
Counties Oakland, and (Asian) (lower
Wayne values)
Counties
Individual Oakland NO YES (poorer) p YES
Counties County (Asian) (lower
values)
Wayne NO YES (White)
CountL (Black)
Oakland Pontiac NO YES (poorer) YES
County (White, (lower
Municipal Asian) values)
ities YES
(Black)
Femdale
Wayne Detroit YES
County Dearbom N 0
Municipal (Asian)
ities Wyandotte

 

 

 

 

 

The results indicate that for most communities, state brownfield redevelopment
efforts have occurred in communities in greatest need of economic redevelopment and
revitalization. These are communities that are poorer, have lower median incomes, lower
median rents and housing values, and less occupied and owner occupied housing.

In All Counties and in Detroit, redevelopment efforts are directly addressing
inequalities that exist. Other communities are directly addressing some inequalities. To
address all, they would need to target specific groups. Oakland County would need to
target redevelopment to Hispanic communities. Pontiac would need to target
communities with poor Blacks, while Femdale should target poorer communities.

Wayne County needs to target poor communities more extensively. Wyandotte and

Dearbom need to target minority and poor communities.

54

A “NO” response indicates that brownfield redevelopment efforts are not
equitably distributed for the specific group. Though the Asian community is relatively
small throughout the region, they are living in tracts with brownfields that are less likely
to have redevelopment activities. In addition, Whites living in Pontiac are living in
communities that have less redevelopment, and those communities should be targeted if
equity is a goal. In Wayne County, Black communities are less likely to have brownfield
redevelopment. Fortunately, all of the “NO” responses in the table are for groups that do
not already experience environmental inequality in regard to BEA sites within the area

studied.

55

 

XI. Conclusions and the Need for Further Research

This study has addressed the descriptive aspect of environmental inequality
research to show where inequality is present in metro Detroit’s brownfield communities.
The methods used in this study could be used in additional communities to determine the
extent of environmental inequality in other areas. The results indicate that for the most
part, state brownfield redevelopment efforts are equitably distributed and are targeting
communities in greatest need of redevelopment. Michigan has achieved success in
linking environmental justice with economic development through Brownfield
Redevelopment Grants funded by the Clean Michigan Initiative bond fund. However,
there is more communities could do to correct the inequalities that do exist. Most
troubling is Macomb County and its communities. The results show evidence of
extensive environmental inequality in those areas, but no state funded brownfield
redevelopment initiatives have occurred.

In Detroit, the next step for future research is to look at process by focusing on the
history of communities where environmental inequality is present. A case-study similar
to Hurley’s study of the Wagner Electric site in Wellston, Missouri would provide useful
information into some of the complex processes at work in creating and sustaining
environmental inequality in Metro Detroit. This history could also be compared to that in
communities where inequality has not been found.

Future research is needed to analyze why come communities have had significant
redevelopment that is compatible with environmental justice values while others have
not. A study comparing local policies, the role of Brownfield Redevelopment Authorities

in various communities, the extent of community participation and creative land

56

 

purchasing strategies as well as other factors would provide greater insight into this
question (Godschalk 1994, Kirshenberg et al 1997, Davies 1999, Leitmann 1999, Leigh
2000, Rogers 2000). The understanding gained from this analysis would provide
additional knowledge to tailor new and creative redevelopment strategies to where they

are most needed.

57

BIBLIOGRAPHY

Anderson, Patrick L. and Ian Clemens. 1999. A report on the demographics of
Michigan ’s urban brownfield communities. Lansing: Anderson Economic Group.

Anderton, Douglas L., Andy B. Anderson, John Michael Oakes, and Michael R. Fraser.
1994. Environmental equity: The demographics of dumping. Demography 31(2): 229-
248.

Baumer, Amy, Brian Cole, Jared Cypher, Yvonne Fleener, Summer Hallwood, Daniel W.
Linna Jr, Brian McGrain, Mandy M. Medina, Bruce Roberton. 1999. Michigan
Brownfield Redevelopment: A study of the impact of legislative change East Lansing:
Program in Public Policy and Administration, Michigan State University.

Been, Vicki. 1994. Locally undesirable land uses in minority neighborhoods. Yale Law
Review 103:1383.

Bowen, William. 2002. An analytical review of environmental justice research: What
do we really know? Environmental Management. 29(1):3-15.

Bowen, W.M., M.J. Salling, K.E. Haynes, and E.J. Cyran. 1995. Toward environmental
justice: Spatial equity in Ohio and Cleveland. Annals of the Association of American
Geographers. 85(4):641-663.

Bryant B. (Ed). 1995. Environmental justice: Issues, policies, and solutions.
Washington DC: Island Press.

Bullard, Robert. 1990. Dumping in Dixie: Race, Class, and Environmental Quality.
Boulder: Westview Press.

Bullard, Robert. 1994. A new ‘chicken-or-egg’ debate: Which came first—the
neighborhood, or the toxic dump? The Workbook 19: 60-62.

Cutter, Susan L. 1995. Race, class and environmental justice. Progress in Human
Geography 19(1): 111-122.

Darden, J. T., & Tabachneck, A. (1980). Algorithm 8: Graphic and mathematical
descriptions of inequality, dissimilarity, segregation or concentration. Environment and
Planning Annals. 12: 227-234.

Davies, Lincoln L. 1999. Working toward a common goal? Three case studies of
brownfields redevelopment in environmental justice communities. Stanford
Environmental Law Journal. 18(2).

58

Duncan, 0., & Duncan, B. (1955, April). Methodological analysis of segregation indexes.
American Sociological Review. 20: 210-217.

General Accounting Office (GAO). 1983. Siting of hazardous waste landfills and their
correlation with racial and economic status of surrounding communities; Washington,
DC: GAO

Gist, GL. 1999. Another aspect of sustainable development—Recycling land. Journal
of Environmental Health. 61(9).

Godschalk David R, et a1. 1994. Pulling together: A planning and development
consensus building manual. Washington, DC: Urban Land Institute and Program for
Community Problem Solving.

Hula, Richard C. 1999. An assessment of brownfield redevelopment policies: The
Michigan experience. The Pricewaterhouse Coopers Endowment for the Business of
Government. Arlington, VA.

Hula, Richard C., Phil Davis, and Bruce Robertson. 2001. Brownfield Initiatives
Proving Successful. The Victor Institute for Responsible Land Development and Use,
Michigan State University. East Lansing, MI.

Hurley, Andrew. 1997. Fiasco at Wagner Electric: Environmental justice and urban
geography in St. Louis. Environmental History 2(4): 460-481.

Lavelle, M., and M. Coyle. 1992. A special investigation; unequal protection; the racial
divide in environmental law. National Law Journal (21 September): 81-16.

Leigh, Nancy Green and Gradeck. 1996. Urban neighborhood demographics associated
with environmentally suspect, tax delinquent properties. Equity and redevelopment
implications. The Review of Black Political Economy. Summer: 61-81.

Leigh, Nancy Green. 2000. Promoting more equitable brownfield redevelopment:
Promising approaches for land banks and other community land development entities.
Working Papers, Lincoln Institute of Land Policy. Cambridge, MA.

Leitmann, Joseph. 1999. Sustaining cities: Environmental planning and management in
urban design. New York: McGraw Hill.

Kirshenberg, Seth, William Fischer, Charlie Bartsch, and Elizabeth Collaton. 1997.
Brownfields redevelopment: A guidebook for local governments and communities.
Washington, DC: International City/County Management Association.

Mohai, P. and B. Bryant. 1992. Environmental racism: Reviewing the evidence.

Pp. 163-76 in Race and the incidence of environmental hazards: A time for discourse.
Edited by B. Bryant and P. Mohai. Boulder: Westview.

59

Ill‘lll‘l‘l I111 ‘1 l! ‘I‘III“‘ 111‘ I

ll“l“..“..<ul‘{lll.t 1‘.-.‘ (4|

Monmonier, M. 1994. Zip codes, data compatibility, and environmental racism.
Multivariate data analysis. Macmillan Publishing: New York.

National Environmental Justice Advisory Council Waste and Facility Siting
Subcommittee (NEJAC). 1996. Environmental justice, urban revitalization and
brownfields: The search for authentic signs of hope. Washington, DC.

Pellow, David N. 2000. Environmental Inequality Formation: Toward a theory of
environmental injustice. American Behavioral Scientist 43(1): 581-601.

Pulido, Laura. 2000. Rethinking Environmental Racism. Annals of the Association of
American Geographers 90(1): 12-40.

Pulido, Laura. 1996. A critical review of the methodology of environmental racism
research. Antipode 28(2): 142-159.

Pulido, Laura. 1996. Introduction: Environmental Racism. Urban Geography 17(5):
377—379.

Rogers, Will. 2000. Economic benefits of parks and open space. San Francisco: Trust
for Public Land. http:www.tpl.org. Accessed December, 2000. .

Szasz, Andrew, and Michael Meuser. 1997. Environmental inequalities: Literature
review and proposals for new directions in research and theory. Current Sociology 45(3):
99-120.

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. New York: Commission for Racial Justice,
United Church of Christ. ‘

US. Bureau of the Census. 1990. Census of Population, Social and Economic
Characteristics from 1990 Census of Housing, General Characteristics.

US. Bureau of. the Census. 2000. Census of Population, Social and Economic
Charecteristics from 2000 Census of Housing, General Characteristics

US. Environmental Protection Agency. 1996. 1996 Waste Programs Environmental
Justice Accomplishments Report & Executive Summary. Washington, DC.

US. Environmental Protection Agency. 2000. A quick view of key brownfields issues.
http://www.epa.gov/R5Brownfields/quick.html. Accessed April, 2000.

US. Environmental Protection Agency Office of Enforcement and Compliance. 2003.

Assurance Guide to Environmental Issues: Glossary of Terms & Acronyms.
http://www.epa.gov. Accessed November, 2003.

60

..1z||. It! I‘ll. ll|lli all
('1 l1

Itl.’xl'. «til. 'Iy .1 "t III‘IO.‘
ill‘ll

 

IlllllllllllllIlllll

 

.-._‘--_‘.——_-_ l__._ A