.1.

. t&V\\.
3.1.}... .m
i N:
{I
,3

“s

.‘ fl .
huh
,9

 

 

}
.2

 

2- .
$413!...
avaiflil.‘
ii lint.

-‘-. m.
’~*9"-.~ m
)Ilhy‘t ll

 

\Ya .5: 1217 i
. 52.541.» ..

 

 

 

 

 

 

 

sWWMHqumnmw

 

, rum“... ..

kwéié , a,

fiflmuun:
.- i . t

 

 
 

, agggégwg i

.I

 

x

 

           
 

 

 

 

 

037%) Ill!lllllzlflllllllllWill/llllll/l/lll/lll

1405 7487

This is to certify that the

thesis entitled

SPATIAL PATTERNS AND DISTRIBUTIONS OF THE LOCATIONS
OF RACIAL GROUPS AND EMPLOYMENT OPPORTUNITIES
IN MICHIGAN, 1960-1990

presented by

Bradley J. Davis

has been accepted towards fulfillment
of the requirements for

MA degree in Geography-Urban Studies

 

 

    

‘ Major professor

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY

Michigan State
Unlverslty

 

 

 

PLACE ll RETURN BOX to roman this checkout from your record.
To AVOID FINES Mum on or before data duo.

DATE DUE DATE DUE DATE DUE

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

WWI

 

 

 

Spatial Patterns and Distributions of the Locations
of Racial Groups and Employment Opportunities
in Michigan, 1960-1990.

By:

Bradley James J. Davis

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

Master of Arts

Department of Geography and Urban Affairs Programs

1995

ABSTRACT
SPATIAL PATTERNS AND DISTRIBUTIONS OF THE LOCATIONS
OF RACIAL GROUPS AND EMPLOYMENT OPPORTUNITIES
IN MICHIGAN, 1960-1990
By

Bradley James J. Davis

Various population groups in Michigan during the last four decades have persistently
experienced high levels of residential and employment segregation. The research problem
of this study is to determine the spatial phenomena Of residential and employment locations
for the white, black, and other racial groups. Data are Obtained on the county level for the
Census years 1960 to 1990 from the United States Bureau Of Census, the Michigan Statistical
Abstract, and the Michigan Employment Security Commission. The method employed to
determine location, distances and direction are centrographic measures. Measures Of
centrality for the population and economic data were computed for blacks, whites, and other
racial groups along with total population and total employment. The results suggest that
blacks experienced the highest level of both residential and employment segregation for each

decade.

Copyright by
BRADLEY JAMES J. DAVIS
1995

ACKNOWLEDGMENTS

It has been a long and interesting journey to reach this stage of my educational goal. The
members Of my thesis committee were outstanding in their attitude and guidance to
complete this research. Sometimes it felt that I bit Off more than I could chew, but they
made me believe that this subject matter was well worth completing. I would also like to
thank my mother whose unwavering support became a comfort to rely on. And finally, a
big thanks to my father who was present during the two year period I attended this
university. He knows that my goal is yet completed and that I will carry on with

perseverance and the utmost credibility.

iv

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................. vii

LIST OF FIGURES ............................................................................................. viii

CHAPTER 1: INTRODUCTION ........................................................... 1
1.1 : Data and Method ........................................................................ 2
1.2: Research Hypotheses ................................................................... 3
1.3: Definition of Terms ..................................................................... 3
1.4: Significance Of the Research ....................................................... 4

1.5: Limitations of the Data and Research

Approach ..................................................................................... 5
1.6: Conceptual and Theoretical Framework .................................... 6
1.7: Centroids as Methods Of Measurement ....................................... 7

1.8: Background Related to Race, Residence
and Employment Opportunities ................................................... 9

1.9: Factors Influencing the Residential Location of Population

Groups and their Employment Opportunities ............................. 12

1.10: Appraisal of General Hypotheses ................................................. 14
CHAPTER 2: DATA SOURCES AND RESEARCH

METHODOLOGY .......................................................... 17

2.1: Data and Study Regions .............................................................. 18

2.2: The Issue of Scale in Measuring Segregation ............................. 21

2.3: The Centrographic Method at the State Level ............................ 22
2.3.1: Measuring Centrality ....................................................... 27
2.4: The Patterns Of Residential Location of Racial Groups and
Employment Opportunities: Previous Research ......................... 27
CHAPTER 3: Analysis of Residential and Employment
Locations in Michigan ..................................................... 31

3.1 : Demographic and Employment Trends,
1960-1990 ................................................................................... 32

3.2: Findings on the Spatial Distribution Patterns Between
Total Employment (TE) And Population Groups ....................... 39

3.3: Findings on the Spatial Phenomenon Between
Total Population Center’s Of Gravity and

Population Groups ....................................................................... 44

3.4: Mapping of Directions and Relative Locations
of All Centroids ........................................................................... 44
CHAPTER 4: SUMMARY AND CONCLUSION ................................. 51
4.1 : Conclusion .......................................................................................... 54
APPENDIX A: UTM Coordinates .............................................................. 56
APPENDIX B: Distance Formula ............................................................... 57
REFERENCES ...................................................................................................... 58

vi

2.1:

3.1:

3.2:

3.2(a):

3.3:

3.4:

3.5:

3.6:

Appendix A: UTM Coordinates .......................................................................

LIST OF TABLES

Percentage Distributions Of Labor Employment in Michigan,

1960-1990 ...........................................................................................

Percentage Population Profile of Michigan by Race,

1960-1990 ............................................................................................

Population Totals for Counties within the Detroit SMSA,

1960-1990 ............................................................................................

Percentage Of Population Change for Counties within the

Detroit SMSA, 1960-1990 ..................................................................

Michigan Employment Status by Major Racial Groups,

1990 Annual Averages ........................................................................

Distances in Miles and Directions in Degrees for Population
Groups from Total Employment (TE) "Centers of Gravity,"

1960-1990 ............................................................................................

Distances in Miles and Directions in Degrees for Population
Groups from Total Population (TP) "Centers Of Gravity,"

1 960-1990 ............................................................................................

Distances in Miles and Directions in Degrees between
Total Population (TP), Total Employment (TE) and Total
Employment\TOtal Population "Center's Of Gravity,"

1 960-1 990 ............................................................................................

vii

21

32

34

34

38

41

42

43

56

3.1:

3.2:

3.3:

3.4:

3.5:

3.6:

LIST OF FIGURES

Michigan Population Centroids for Total Population (TP),
Total Employment (TE), White Population (WP),
Black Population (BP), and Other Population (OP), 1960 .......................

Michigan Population Centroids for Total Population (TP),
Total Employment (TE), White Population (WP),
Black Population (BP), and Other Population (OP), 1970 .......................

Michigan Population Centroids for Total Population (TP),

Total Employment (TE), White Population (WP),

Black Population (BP), and Other Population (OP), 1980 .......................
Michigan Population Centroids for Total Population (TP),

Total Employment (TE), White Population (WP),

Black Population (BP), and Other Population (OP), 1990 .......................
Michigan Black Population Centroids, 1960 - 1990 ................................

Michigan - County Locator Map ..............................................................

Appendix B: Distance Formula .........................................................................

viii

47

47

48

48

49

50

57

CHAPTER 1

INTRODUCTION

The uneven spatial distribution of blacks and other minorities from whites is a
topic that has been a major focus of research in the United States, especially since the
19605. The processes Of and the mechanisms involved in residential locations of racial
groups have generally focused on areas other than, and usually smaller the state level.
More importantly, past research has not linked the uneven spatial distribution Of racial
groups directly to economic opportunities in society.

This research focuses on the spatial differentiation of racial groups at the state
level using centrographic techniques to analyze the distribution Of residential
concentrations and employment opportunities. Residential concentrations of various
population groups have been a continuing concern to urban geographers and thus far
studied primarily as an urban phenomenon. This urban bias has been portrayed as urban
geographic and urban social problems in the context of residential segregation. However,
the underlying dimensions Of residential location are social and geographic problems that

have particular spatial manifestations beyond the urban level.

2

The problem of this thesis is twofold. First, to determine the extent to which the
pattern of residential concentration Of racial groups is related to the pattern of
concentration Of employment opportunities. Second, to determine how such patterns have
changed over the decades, from 1960 to 1990. The study area of this investigation is the

state of Michigan.

LlLDAtaaniMethad

Data were collected at the county level to determine the geographic distribution of
population groups and employment Opportunities for the State. Three major sources Of
data were Obtained for this study which were the Bureau Of Census Reports on General
Population and Housing Characteristics, the Michigan Statistical Abstracts and reports
from the Michigan Employment Security Commission.

Centrographic measures are employed to determine the geographic centers of
gravity for employment and for the residential concentration of whites, blacks, and other
minority groups. These measures are used because they best display the spatial disparity
between various racial population groups and employment Opportunities at a state level of

analysis. The variables will be discussed in detail in chapter two.

W
The hypotheses to be tested in this thesis are:

1. The patterns of residential concentration of racial groups is related to the pattern
of location of employment Opportunities.

2. Some racial groups are located closer to employment Opportunities than other
racial groups.
3. The patterns of residential concentration Of racial groups and the patterns of

concentration Of employment Opportunities are dynamic; that is, changing from
decade to decade.

11‘ D E . . [I
The term segregation is used in this study to describe the uneven spatial
distribution between the total population as well as the majority population group and a

minority population group within a defined geographic area. The term employment
segregation is defined similarly as the uneven spatial distribution of job Opportunities
within a defined geographic area. Residential concentrations are determined by

aggregating county population data for state level patterns.

Prior studies of residential location have been couched within segregation
models. The most common Of these measures is the Index of Dissimilarity or Segregation
Index. Researchers of segregation have developed additional indices, (e. g., P" exposure
index) many of which have been shown to be highly correlated (Duncan and Duncan,
1955; Massey and Denton, 1989). The majority of this research, however, has been
concerned with the patterns Of segregation focusing on small-area data (e. g., census tracts,
census blocks) within cities.

An exception is one that focused on residential locations of blacks and whites in
rural neighborhoods at a county level of analysis (Lang, 1979). Consequently, the
identification and observation Of residential and employment data at the state level have
been largely ignored. In general, segregation research has made little attempt to focus on
residential locations at other than an urban level of analysis (Kain, 1965, 1968, 1974;
Darden, 1987; Darden and Kamiar, 1988). This study expands our knowledge of
segregation or residential location of racial groups by focusing on a state level of analysis.
The significance stems from aggregating data to look at broader scales of analysis. This
allows fiJture research to determine if concentration of minority population groups is

correlated with concentration of employment opportunities at a state level.

5

Economic data availability for racial employment and unemployment figures is
provided for large metropolitan areas but not necessarily for all counties within the State.
Because Of the sensitivity Of the data, individual states mainly provide county level
figures for those counties that make-up metropolitan statistical areas (MSAS). Usually,
the decision to use county level data is based on various population or economic
thresholds, such as percentage of urbanization. For example, the state Of Michigan
provides such economic data for counties over 50,000 total population.

The inherent problem of spatial segregation analysis at larger scales is that data
for racial and occupational groups may only be provided for selected geographic parcels
within metropolitan areas. Therefore, not all socioeconomic variables present in small-
area data may be available at larger scales. In addition, common measures of spatial
segregation at the state level only provide relative proportions Of racial groups between
other counties. This study is directed towards aggregating selected county population
groups and selected occupational data in order to apply centrographic techniques for
analyses. Both residential concentration of racial populations as well as employment, or
where the majority of jobs are located in the state are calculated. The employment
centroid is represented by those industries that account for the highest percentage of
employment activity in the State of Michigan. County level employment data are gathered
for each of Michigan's eighty-three counties during each census period. The Total

Employment (TE) centroid explains over 86 percent on average of all employment

6

opportunities in the state of Michigan during the period of this study. The industries that
make up the TB centroid are discussed in detail in Chapter two

The methodological contribution of this research is in analyzing the patterns and
distributions Of residential and employment concentration among whites, blacks, and
other population groups of a state. Investigating the directions and patterns of residential
and employment concentrations within the State Of Michigan will allow researchers a
unique longitudinal view Of a particular state's residential and economic patterns of

growth and their respective spatial distribution.

LEW

Research concerning the geographical approach to residential locations has
focused on segregation theory and the adaptation of a particular group to certain spatial
constraints to both residential (Darden, et.al., 1980, 1988; Clark, 1984) and employment
(Kain, 1968, 1987; Farley, 1987; Hughes and Madden, 1991) Opportunities at various
urban geographic or "community" levels.

Segregation research has sought to describe the pattern Of location and changes of
some portion Of the population in the world about us. According to theories related to
residential segregation, spatial variation in segregation between population groups relates
directly to measurable differences associated with social and economic variables (Massey
and Eggers, 1990; White, 1983; Burgess, 1969; Parks, 1969, 1969a). For example, the
higher the occupational level, the greater the distance a group will live from the central

business district. It is generally supported by prior research that groups’ movement

7

outward from the central city is usually preceded by an increase in the level Of education,
occupation, and income (Burgess 1969a; Parks 1969a). If the distribution patterns and
employment opportunities available to various racial groups are uneven, this does not
necessarily imply discrimination, but may imply an historical attitude based on traditional
ignorance. In fact, to measure discrimination, one must be able to measure the productive
capacity of people and the Opportunity structure they face. Such measurement is beyond

the scope of this study.

W

The most important characteristics of spatial phenomena can be reduced to three
criteria: location, shape, direction, concentration. A spatial distribution may change
symmetrically over time and there will be no change in location. For example,
technically, a centroid possesses shape although it is not being measured here, the varying
symmetry crosses borders, such as a plane or line, or around a point or axis. Although the
symmetry of a centroid is structurally or functionally independent of direction, both
casual and more robust empirical Observation, as well as theory, suggests that
symmetrical change will be the unusual rather than the norm (Jones, Manson and
Coccossis 1991). Since the factors leading to asymmetrical change will be persistent and
enduring, the change of location is likely to be symmetric and unidirectional rather than
sporadic or random.

There is little in theory or the Observation Of real phenomena to suggest that

spatial variables will be shapeless--or without shape or circular. However, since there are

8

few tools for analyzing shape in any efficient way, there have been a few models as to the
nature of the shape to be anticipated. For example, the various geographic processes of
urbanization depart substantially from the original circular model of Christaller (1966).
The importance rises from the potentiality that certain models may give distorted or
misleading values to other summary measures which implicitly assume circularity. The
degree of direction or dispersion of a spatial phenomenon and the way in which it iS
changing over time is a critical element and has been a subject of speculation in urban
theory since its very origin. In addition, this has been the subject Of a large number of
attempts at calibration and measurement over time.

The prominent measurement for the spatial distribution of residential populations
during the last forty years has been the use of the Index Of Dissimilarity. Inherent
criticisms Of the index have evolved around a number of issues of interpretation (Hughes
1987), including the arbitrary choice of cities for analysis and/or comparison, and the fact
that small-units Of geographically defined areas of analysis do not necessarily constitute
cohesive entities (White, 1983). For example, such areas as census blocks and census
tracts may be homogeneous with respect to certain specified variables, such as income,
education, and household size (Greene, 1991; Holzer, 1991). Yet, they may not be
characterized by a high degree of a "community" or internal interaction such as religion,
political agenda and family structure (Wilson, 1978).

The Index of Dissimilarity is computed between two population groups or a sub-
group Of a population and the remainder of the population. It is a measure which has been

used particularly to assess the degree of residential segregation of a sub-group within the

9

total population. This study differs somewhat in that it employs centrographic techniques
where residential and employment concentrations are computed to determine the spatial
distribution and patterns. The significance of this research is in the scale and method of
measurement, which allows one to determine both the location and direction over time Of
a spatial entity. In addition, one can apply this measure to compute results for the total
population and total employment distribution, and therefore, allow one to compare and
contrast the results over time to other population sub-groups. The evolution Of
centrographic measures have clearly developed for measuring Spatial differentiation such
as the three characteristics Of location, shape and concentration (Tordella 1987, Jones,
Manson and Coccossis, 1991).

- :. ,9. H... :- ._ -. 01,. - :- 'q-. -._.I_ m . m-. an...“ ’-

During the 19605, the United States experienced an unprecedented Civil Rights
Movement. This movement was an outcry from racial minority groups within American
society against discrimination and segregation practices that isolated them from the social
and economic benefits experienced by the white majority population (Wilson 1987).
Several commissions were appointed to examine the causes behind the violence (riots)
stemming from this movement. The Kemer Commission reported that the major
grievances of residents in the worst affected areas were police brutality, unemployment,
and poor housing (National Advisory Commission on Civil disorders, 1968). In
describing employment growth within metropolitan areas, for example, the Kemer

Commission observed:

10

"Most new employment opportunities do not occur in central cities, near all-Negro
neighborhoods. They are being created in suburbs and outlying areas--and this
trend is likely to continue indefinitely. New Office buildings have risen in the
downtown of large cities, often near all-Negro areas. But the outflow of
manufacturing and retailing facilities normally offsets this addition significantly--
and in many cases has caused a net loss of jobs in central cities.

Providing employment for the swelling Negro ghetto population will require
society to link these potential workers more closely with job locations. This can
be done in three ways: by developing incentives to industry to create new
employment centers near Negro residential areas; by Opening suburban residential
areas to Negroes and encouraging them to move closer to industrial centers; or by

creating better transportation between ghetto neighborhoods and new job
locations." (p. 392)

In addition to the Kemer Commission, the McCone Commission was also responsible for
examining the causes Of the Watts (Los Angeles) riots in the summer of 1965. Not unlike
the Kemer Commission, the McCone Commission found that the most serious immediate
problem that faced the black population was unemployment (National Advisory
Commission on Civil Disorders, 1968). Most notable, is the fact that the majority of civil
unrest incidents in major US. cities took place after the passage of the Civil Rights Act in
1964 (National Advisory Commission on Civil Disorders 1968). Coupled with growing
civil unrest is a deconcentration trend of population and employment which is also having
a profound impact on central cities.

During the 19605 and into the 19905, segregation between various racial
minorities and whites continued (Deskins, 1972; Massey and Eggers, 1990; O’Hara
1991). In addition, population growth is continuing at a faster rate in exurbia than in

metro areas, a trend not witnessed for more than 160 years (Williams 1987). Other

11

common barriers that face racial minorities today are their social isolation from the better-
Off population, with higher incomes and higher educational achievement (Wilson, 1987)
along with the decline of jobs Offered by the labor market combined with the lack of
skills possessed by the job seekers (Kasarda, 1986). Many Observers feel that the two
presidential administrations (i.e., Ronald Reagan and George Bush) which were in power
during the 19805, perpetuated the social environment that harmed black socioeconomic
advancement, voiding civil rights gains, and promoting a general anti-minority climate
(Updegrave, 1989). In fact, these administrations eliminated a major portion of each
state's federal fimding for these programs along with their federal enforcement
responsibility (Bell 1 991).

The result Of these actions was greater emphasis on the influence of a state's fiscal
and policy decision making processes (Darden, 1980; Cason, 1980; Clark, 1984 and
Hollister 1989). Despite the plethora Of federal and state work and aid programs for the
poor and disadvantaged such as Work Incentive programs (WIN), Aid for Dependent
Children (AFDC), Community Employment Training Aid (CETA), Workfare and Food
Stamps, unemployment for blacks, and other minorities continued to rise at an alarming
rate (Wilson, 1987).

Research has shown that among the educated, middle-class black population,
changes have resulted in greater residential mobility and employment Opportunities
(O'Hara, 1991). However, Updegrave's (1989) investigation of the Black educated
middle-class reveals it to be fragile and shrinking. Updegrave (1989) stated, that

economic inequalities are more a result of prejudicial behavior and a legacy of past

12

discrimination than back-to-back recessions or the federal government's hostility towards
affirmative action programs. Moreover, a large portion Of the black population, the urban
poor, appear stuck in a quagmire of unstable families, intermittent employment, welfare
dependence, and the temptations of crime. The popular, but erroneous assumption that if
America solves the problem of black poverty, economic parity between the races will be
achieved is further disputed by Updegrave (1989), showing that wide differences extend

beyond the 31% Of the black population that lives in poverty.

- -.‘H. I O .C
o UiOU‘Is‘I-..t..'.'.-_.U 01-.le

For the most part, we leave to private enterprise and entrepreneurship the task to
determining a city's organization Of its residential and industrial districts. It is true that a
vast majority Of larger cities have developed "master plans" that assist in urban growth
and change based on zoning laws and protective ordinances for its population. The
topography Of the land, natural advantages such as lakes, rivers, great bodies of water,
green areas, along with various methods Of transportation also influence the spatial
distribution of the population. New developments in transportation and communication
linkages along with new economic trade areas all contribute to the rapidly changing social
and industrial organization of the urban area. They have been the means of re-
concentrating traffic from the central business district (CBD) to the centers Of suburbs in
surrounding counties changing the whole character Of retail trade and industrial location.

They have also changed the overall pattern of residences and employment Opportunities

13

by multiplying suburbs away from the city centers. These forces have led to a change in
employment and population distribution accompanied by corresponding changes in the
habits, sentiments, and character of the urban population.

Both social and economic factors have been identified as measurable variables
that relate directly to segregation between population groups (Burgess, 1969). Factors
such as, employment trends in various sectors of the economy and concentration of
minority population groups in Older, central city neighborhoods are discemable patterns.
Parks (1969a) states, that spatial relations are highly correlated with social relations and
physical distance with social distance.

The Competitive Theory of Segregation suggests that white incomes probably rise
in the process of generating physical distance from blacks (Thurow, 1975). For example,
the greater the physical distance between whites and blacks, the higher the utility Of
whites, so whites are willing to pay a premium not to associate with blacks. And
sometimes whites are confronted with a choice between more physical distance or greater
incomes. By contrast, in a social-distance model (e.g. spatial mismatch hypothesis) whites
wish only to raise white incomes. Nevertheless, higher white incomes contribute to
greater social distance. It is also personal tastes, conveniences, economic interests, and
politics that tend to influence the environment of various population groups in our cities.
Physical, social, or economic pressures may enable whites to be segregated from blacks
or other minority populations.

Researches that measures variables concerned with poverty, unemployment and

income inequality in the United States have been conducted on a variety of fronts in

14

which the extent of income inequality and the degree of spatial segregation by income
have been related to levels and trends in residential segregation (Kain, 1965; 1968; 1974;
1986; 1987; Leonard. 1984; Darden and Karniar, 1988; Masssey and Denton, 1989;
Hughes and Madden, 1991).

Kain (1965) linked discrimination in the housing market of Detroit to the
distribution and level of non-white employment in urban areas. He found it was much
easier to get blacks and whites to work side by side than to live side by Side. Leonard
(1984) found that the distance of a workplace from the main ghetto is one Of the strongest
and most significant determinants of both changes over time and level Of racial
composition of the workforce. He extended Kain’s work by analyzing the impact of
distance on changes over time in employment share and impact of distance on the level Of
employment share in Metropolitan Statistical Areas (MSAS). Both found that the efficacy

of efforts to integrate the workplace are still limited by residential segregation.

WWW

The hypotheses in this study are concerned with both residential and employment
segregation. Their relationships are interrelated and interdependent in that methods Of
measurement, background and theoretical framework may be considered symbiotic. Such
hypotheses have their roots in the spatial mismatch hypothesis, white gains hypothesis
and the underclass hypothesis.

The spatial mismatch hypothesis argues that the extent to which serious

limitations on black residential choice, combined with the steady dispersal Of jobs from

15

the urban core, are responsible for the low rates of employment and low earnings Of the
black population. The spatial distribution of nonwhite employment and residential
choices of black households was first characterized by Kain (1965, 1968, 1974) in which
he proposed the spatial mismatch hypothesis. However, recent studies on spatial
mismatch have focused entirely on the black population deviating from Kain's original
characterizations of the interaction of residential segregation and employment
discrimination between various races and various income levels (Kain, 1986, 1987).

A second hypothesis, which may be called the "white gains" hypothesis, or split
labor market hypothesis argues that minority disadvantage occurs because the majority
group benefits from it (Farley, 1987). For example, a job not held by a black or other
minority group person is potentially available to the white majority member. This view
holds that whites have the most to gain where the minority population is the largest. This
differential thus predicts that the highest levels of segregation will occur in that situation.
Farley (1987) believes that very different geographic distributions Of jobs, and therefore
employment opportunities are being shifted into the suburbs. However, since findings of
this sort are usually restricted to data gathered for Metropolitan Statistical Areas (MSAS),
it ignores examination at a broader or finer, smaller scale, thus understating the true
magnitude of that shift.

A third hypothesis, stems largely from the work of Wilson who promoted the
term, "Underclass" (Wilson 1978, 1987). Wilson argues that minority economic
disadvantages today are not the product of overtly racial processes, but rather are due to

disadvantages associated with class. Wilson (1978) points out that the unemployment

l6

differential between blacks and whites is mainly a post-World War II phenomenon, and
links it to class based disadvantages among blacks with respect to job skills and
education. However, numerous research studies have shown that when socioeconomic
factors such as occupation, education, and income levels are controlled, racial segregation
still persists (Darden, 1987; O’Hara, 1991).

Based on this conceptual\theoretical background, this study examines the spatial
concentration of various racial groups and their relation to employment opportunities in

the state of Michigan.

CHAPTER 2

DATA SOURCES AND RESEARCH METHODOLOGY

As indicated in the previous chapter, the literature on the spatial concentrations
and patterns of various population groups has emphasized racial and economic
inequalities at the sub-state level; ignoring geographic processes taking place at broader
scales. This chapter is concerned with the structure of the data and that portion Of the

literature where scale is an issue.

17

18
WW
Three sources of data were Obtained and used for this study which were the
Bureau of Census Reports on Generalionulatinnandflnusingflhatameristicsm
Michigmfitanflicalflmm and reports from the Michigan Employment Security

Commission. The county level data were aggregated for a Set of five variables for the
State of Michigan: total population (TP); total employment (TE); white population (WP);
black population (BP) and other population (OP). The primary data source for population
figures was the United States Bureau of Census publications on W
W$ureau of Census, 1961; 1971; 1982; 1992) for the Census
years 1960, 1970, 1980, and 1990. The data consisted of county level figures for the
State of Michigan and the various racial populations as defined by the Bureau of the
Census. Two additional sources provided data for the State of Michigan on both
population and employment figures: the WW; and the Michigan
Employment Security Commission, Bureau Of Research and Statistics Labor Market
Analysis Section. These two sources combined, provide figures for total employment,
total employment rates, and total population at a county level for the period of this study.

Comparability between the various racial populations for each of the census
periods is based on a series of aggregated tabulations for the 1960, 1970, and 1980
censuses. The aggregated tabulations are patterned after the racial classifications
provided by the Bureau of the Census for the 1990 Census of Population (Bureau of
Census 1961a, 1993).

The classification of race as used by the Census Bureau reflects self-identification;

19
it does not denote any clearly defined scientific definition Of biological stock. Therefore,
the data for race represents self-classification by people according to the race with which
they most closely identify. Some of the definitions used in 1960 differ from those used in
subsequent census periods. For instance, the 1960 data on race were derived from that
which was commonly accepted by the general public. Divisions of the population were
categorized by color into two groups: white and nonwhite. The color group designated
as "nonwhite" included Negroes, American Indians, Japanese, Chinese, Filipinos,
Koreans, Hawaiians, Asian Indians, Malayans, Eskimos, Aleuts, etc. Persons of Mexican
birth or ancestry who were not definitely of Indian or other nonwhite race were classified
as white.

It was not until the 1990 Census that the specific term "Spanish/Hispanic" was
placed on the questionnaire. Persons who did not report a Specific race but identified
themselves as "Mexican," "Puerto Rican," or "Cuban" in the race question were assigned
to white in 1970, and for 1980 these persons remained in the "Other Races" category.
The 1990 data on Hispanic origin are generally comparable with those figures for the
1980 census. The 1990 and 1980 census data on the Hispanic population are not directly
comparable with the 1970 nor the 1960 nonwhite category data because Of a number of
factors: (1) over-all improvements in the 1980 and 1990 censuses, (2) better coverage Of
the population, (3) improved question designs, and (4) an effective public relations
campaign by the Census Bureau with the assistance Of national and local ethnic groups
(Bureau of Census, 1993).

Upon comparing the differences between the white and black populations for the

20
period of this study, it was necessary to identify those population groups that would fall
into the "other population" (OP) category. For the 1960 Census period the category of
Negro was identified as the "black population" (BP) variable. The data for the OP
variable was determined by a two step process; (1) subtracting the figures for Negro fiom
the total of all other racial categories thus deriving a figure for the BP variable; and then
(2) totaling the remaining racial categories in order to derive the OP variable. This
process was applied to the 1960, 1970, and 1980 censuses for both the population and
employment categories. Population groups included in this OP variable are all remaining
population groups other than the census definition of whites and blacks for each census
period.

The method employed for calculating the variable total employment (TE)
consisted of an aggregation Of those industries combined that would employ over fifty
percent of the labor force in Michigan. As shown in Table 2.1, manufacturing has
consistently been the highest single category over trade and government. The category
"other" has consistently increased in proportion to decreasing labor rates in
manufacturing. The basic labor force employment categories consist of manufacturing,
government, and a combination of service industries headed under the "other"
employment category. With the exception of the 1960 census period that accounts for
nearly 57 percent, all other percentage figures for TB were well over the 1960 figure. The
variable total employment (TE) explains over 86 percent Of all employment activity for

the remaining census periods.

21

Table 2.1: Percentage Distributions of Labor Employment in Michigan, 1960-1990

 

Category 1960 1970 1980 1990
Manufacturing 38.8 30.2 23 .4 19.6
Other“ -- 39.4 42.3 58.3
Trade 1 8.0 -- -- --
Government -- 14.1 14.6 18.2
Total Percentages 56.8 83.7 80.3 96.1

 

*Note: The labor category "Other" as defined by the Bureau of Census Economic Data:
1970- construction, transportation, communication, and public utilities; 1980 and 1990-
wholesale/retail trade, finance/insurance, real estate services and mining.

Sources: MESC 1981, 1992; Michigan Statistical Abstracts 1961, 1971

 

22.111 [5].“ .5 .

In general, segregation research at the sub-state level of analysis has focused on
census or city blocks and census tracts within cities or across Standard Metropolitan
Statistical Areas (SMSA) which usually consist of two or more counties. Consequently,
the vast majority of segregation literature has generally focused on the largest urbanized
area within a State or has used a number Of SMSAs in states to compare and contrast their
characteristic differences or similarities. The literature proves insufficient when several
urbanized areas within a given state, or for that matter, the state itself is the scale of
research. F urtherrnore, these studies fail to compare social and economic characteristics

Of states that would allow future comparisons.

22

AWE/e1

Spatial segregation can be said to exist wherever the distribution of one group
differs from that Of another group. The methodology in this study applies centrographic
measures at the state level to population sub-groups and a total population and
employment centroid. The use of a centrographic technique results in the displaying Of
centroids allowing one to measure their spatial distributions and in their changes
continuous over time. These spatial distributions will be analyzed as stated previously, in
terms of their location, direction and distance. Each county is weighted according to
population and employment counts. The figures for each of Michigan's eighty-three
counties are aggregated to determine the center of the various population and employment
centroid which summarizes location. This is the center of “moments” similar to the center
of gravity which minimizes the sum of the squares of the distance of the Observations to a
point (Heyford 1902, Cooper 1963). Basically, it is the mathematical expectation of the
location of an Observation selected at random from a population. Changes in location and
direction of movement over time can be Observed by comparing the coordinates Of the
centers. Differences in the location of components Of the population such as racial or
employment groups at a single point in time can be compared in a Similar fashion. Similar
comparisons can be made for a variety of economic activities, physical facilities such as
residential housing, or any other phenomenon located in space. Centers are found in and
identified by the use of Universal Transverse Mercator coordinates as presented in

decimal fractions and degrees.

23

The formula for centrographic measurement is presented below:

Where: Pi = population or employment variable for county I
Xi = centroid for variables (e.g., total population) for county 1
I = 1,2,...83 since there are 83 counties in Michigan

 

Assuming planar or two dimension geometry, we can calculate the mean center by simply

calculating the mean of the X - coordinate (eastling), and the mean of the Y - coordinate

24
(northings). The two coordinates mark the location of the mean center, X.

Since the turn of the century, various scholars have embraced this method in
analyzing the centrality of an area or population (I-Iayford 1902; Bureau Of Census 1923;
Ross, 1930; Scates, 1933). During this earlier period, coordinate axis had to be estimated
since few maps were generated for micro-level analysis which were derived fi'om small-
scale maps. However, the basic formula for the mean center has not changed.

Centrographic measures have developed a substantial vocabulary over the last
century for measuring the three characteristics Of shape, dispersion and location. The
mean center can be described as the; center Of gravity (Lovelock and Weinberg 1985),
center Of minimum travel (Tordella 1987), a centroid (Jones, Manson, and Coccossis
1978), and more recently, the centre of minimum aggregate travel (Kumler and
GOOdchild 1990) or centroid ellipses (Greene 1991).

This study will calculate a series Of five state centroids for each of the census
periods, 1960 through 1990. The five centroids to be displayed are: 1) White population
(WP), 2) Black population (BP), 3) Other (OP) remaining minority population groups, 4)
total employment (TE), and 5) total population (TP). A precise geographic location or
point is determined for all centroids for each of the census periods. These geographical
centers or points permit a spatial comparison between the change of distribution of any

point(s) within any given census period and from one census period to another.

25
The Spatial patterns and distributions of the state’s racial population centroids for
each census period is determined from each county's total for that population group. The
total population and total employment centroid are figured Similarly. The formula (Clarke

1988) for determining the X/Y coordinates for each of the state centroids is presented.

 

erdinate:
1'1
2 x2
— .1. 1
D
Z.
.1.
i=1
Magic;

 

whereas:

i:
N
§=ZP
N
2:24
N

The variable Z is the mean/average Of all eighty-three counties in Michigan for
each of the five variables. This methodology generated a total of twenty centroids, four
sets for each census period and five variables for each set: TE, TP, WP, BP and OP. Each
UTM coordinate was input into a analytical and computer cartographic software package
called, Atlas Geographic Information System (AGIS). The term GIS stands for;
Geographic Information System (or Science). A combination of traditional and analytical
computer cartography were employed to generate the maps and determine directions. Two
other software packages, Corel Draw and PC Globe, were used to assist in generating

output and determining distances and directions.

27

2 3 l . I I . C l'

Centrographic measures have developed a substantial vocabulary over the last
century for measuring the characteristics of distributions such as the median center of
gravity or centroid for concentration of areas using spatial data. In addition, an ellipse of
density, may be derived from the root mean about axes through the center for shape; and
the average radius indicating the average distance and direction fi'om an Observation in a
spatial series from the center point.

A wide variety of measures of various characteristics can be generated to measure
a spatial distribution; however, the spatial phenomenon in a single point permits tracing
the change Of location Of that point from one period to another. The application of
centrographic techniques in this study provides observations for concentration and
direction of the various population groups and employment Opportunities over several
periods of time. This study will demonstrate the efficiency with which centrographic
measures display the distances and directions in the spatial distributions of various

phenomenon and convey large amounts of information simply and quickly.

I. 3 10' 0.0911.-.

'0
I.

o- o '
In snot Or. .11...

Studies at the micro-level have focused on Blacks, Hispanics, Native Americans,
and Asians by measuring the overall unevenness in the spatial distribution Of various
racial groups within cities or SMSAs. Darden (1987) has found that Asians experience

the least amount of residential segregation from Whites, have the highest level of

28
suburbanization, and are provided the greatest Opportunity for social, economic, and
spatial mobility. Blacks, on the other hand, experience the greatest degree Of residential
segregation, the lowest level of suburbanization, and the least opportunity for social,
economic and spatial mobility of all racial/ethnic minority groups.

Recently, research at the regional level in which states are aggregated into
economic blocks have focused their efforts on deindustrialization and the political
economy (Hill 1982; Hill and Negrey, 1987). These studies concentrated on employment
Opportunities and their relationship to economic regionalization or employment shifis.
Hill (1982) suggests the restructuring of the city caused by deindustrialization of the
Great Lakes region places city dwelling minorities at a great disadvantage both
economically and socially. Research that provided a state-wide (Michigan) focus in the
impact on black employment, concluded that new manufacturing jobs (i.e., related to the
automobile industry) where locating away from black population clusters (Cole and
Deskins 1988; Deskins 1988). Basically, jobs in the automobile and related industries
has been and continues to be a major provider of employment for black workers and
Deskins (1988) findings conclude that new venture sites are locating beyond normal
commuting distances set by the Equal Employment Opportunity Commission (EEOC).

The EEO laws (Title VII) since 1964 have mainly been examined by sociologists
and lawyers. Sociologists have examined the mobilization Of resources in order to
implement a legislative standard and questions Of the legitimacy of the administrative
process to the exclusion Of its effectiveness have occupied lawyers (Blumrosen 1993).

Both have examined the process by which legal policy is translated into “real world”

29
changes in sociO-economic behavior.

In sum, residential segregation and employment research has either focused on
groupings of multiple counties or Standard Metropolitan Statistical Areas (SMSAs), and
the aggregation of city census tracts or census blocks or regional economic groupings of
various states. As a result, this scale of research has ignored other scales Of analysis that
may be rooted in the influence of the state (Clark 1984, Hollister 1989) and to an extent,
federal-level policies on the sociO-economic and political-economic impact on the
population (Cason 1980). The present study is based on a scale of analysis which
addresses this concern.

A majority of the spatial segregation research for the State of Michigan has
generally focused on the urban areas Of Detroit, Pontiac, and Flint. Research at this
micro-level (Darden 1978, 1983, 1984; Darden and Kamiar, 1988) as well as the county
or MSA (Darden 1985, 1992) level has been on the major racial minority groups.
Inasmuch as not all of the population groups within Michigan have fared equally in their
experiences of economic prosperity and crisis, a major problem has lingered in that the
largest minority population cohort, blacks, have continued over the last forty years to
predominately reside in only a few counties. Consequently, the relatively high
unemployment rates among blacks and recently, Hispanics, is critical since Blacks reside
in a few concentrated areas. Research has Shown that the black population has
experienced a consistently lower rate of employment combined with consistently high
levels of segregation (Darden 1976). The primary reason why blacks have been

impacted more severely economically is segregation, both occupationally and

3O
residentially (Deskins 1988). Response to this lingering problem with existing models on
racial and employment inequalities ignore the realities of continued spatial segregation at
larger scales of analysis.
The analysis of residential and employment locations as well as employment

status by racial groups in Michigan will be discussed in the following chapter.

CHAPTER 3

ANALYSIS OF RESIDENTIAL AND EMPLOYMENT LOCATIONS IN MICHIGAN

The results Of determining the Spatial patterns and distributions Of employment
and residential locations in Michigan are discussed. The first section is concerned with
demographic and economic trends within the State between 1960 and 1990. The
following sections discuss findings relating to centrographic measures in distances and
directions Of population groups from total employment, distances and directions of
population groups from total population, and the mapping Of centroids. A county locator
map for the state of Michigan may be found at the end of this chapter. In addition, raw
data for the centroid coordinates at the state level, and the algorithm employed to

determine distance is located in the appendices.

31

32

The racial proportion of the population within Michigan has increased
proportionately to overall population growth between 1960 to 1990, and Michigan
experienced a total population growth Of nearly 1.5 million during this period. The
percentage increase of population during this period was 18.8 percent and the percentage
of minority population was 16.6 percent. Since 1960, the racial demographics Of
Michigan's population as seen in Table 3.1 have lightly decreased in its proportion Of the
white majority by nearly 6.5 percentage points while slightly increasing its proportion of
minority populations by over 4.5 percentage point for blacks and nearly 3 percentage

points for other minority groups.

Table 3.1: Percentage Population Profile of Michigan by Race, 1960-1990

 

 

Census Year 1960 1970 1980 1990
M299: 7,823,194 8,881,826 9,262,078 9,295,297
White 90.6 88.3 85.0 83.4
Black 9.2 11.2 12.9 13.9
Other 0.2 0.5 2.1 2.7

 

Note: "Other" represents all the racial categories not represented in the above Table and
as categorized by the Bureau of Census for each of the census years '

 

 

Source: US. Bureau Of the Census, W
W.

33

In fact, it is projected that by the year 2010, (MESC, 1992) Michigan's racial
population distribution will experience even a smaller proportion of whites, 79.4 percent,
compared tO a higher proportion of nonwhite population of 20.6 percent. Geographically,
Michigan's total population is very unevenly distributed. Historically, the majority Of the
people have resided in major cities located in the southern third Of the state. During the
period Of 1960 to 1990, an average between 40 to 45 percent Of the residential population
lived in the three counties of Wayne, Oakland, and Macomb as shown in Table 3.2. The
greatest population densities are located in the southeastern parts of the state; the lowest,
in the northern Lower Peninsula and the Upper Peninsula. However, residential
population growth has been increasing in the southwestern portion of Michigan since the
mid-19805.

The highest average population density is in Wayne County and also the location
of Michigan's largest city, Detroit. In fact, between 1960 and 1990 nearly three- fourths of
the black population in Michigan continually resided in Wayne county, but only one-third
of the county's population is black. In addition to blacks being heavily concentrated in
southeastern Michigan, other minorities were also concentrated in both urban and rural
areas in the southeastern portion Of the state. The majority Of other minority populations
also resided in Wayne county. However, other minority populations on average have
made up less than two percent Of the county's total population during the period of this
study.

Major regional changes in population between 1960 and 1970 were beginning in

areas in the northern Lower Peninsula, the southwestern portion of the state, and the

34

TABLE 3.2: Population Totals for Counties within the Detroit SMSA, 1960-1990

 

 

 

 

SMSA

Counties 1960 1970 1980 1990

Lapeer 41,926 52,361 70,038 74,700
Livingston 38,233 58,967 100,289 114,050
Macomb 405,804 625,309 694,600 704,100
Monroe 101,120 1 18,479 134,659 135,660
Oakland 690,259 907,871 1,011,793 1,529,400
St. Clair 107,201 120,175 138,802 145,300
Wayne 2,666,297 2,669,604 2,337,240 2,164,400
Totals 4,050,840 4,552,766 4,487,421 4.867,610

 

TABLE 3.2 (a): Percentage Of Population Change for Counties within
the Detroit SMSA, 1960-1990

 

 

 

SMSA %Change %Change %Change
Counties 1960-70 1970-80 1980-90
Lapeer 20.0 25.3 6.3
Livingston 26.2 41.3 12.1
Macomb 25.2 10.0 1.7
Monroe 14.7 12.1 0.8
Oakland 23.0 10.3 32.9
St. Clair 10.8 13.5 4.5
Wayne 0.5 -12.5 -7.4
Totals 11.1 -1.5 7.9

 

 

SOURCE: US. Bureau of Census, 1960-1990 Census of Emulation and W,
Washington D.C.; 1970-1990 Population Estimates and Projections, Michigan Employment
Security Commission, Detroit, Michigan

35
surrounding urban-suburban areas Of Detroit. For example, in Table 3.2, four of the seven
counties within the Detroit SMSA increased their population base by over twenty percent
during that period. Between 1960 to 1990, Livingston county nearly trebled its population
base, Oakland county doubled its 1960 population of 690,259 to over 1.5 million while
only Wayne county showed signs Of a declining population base.

The character of population change was in the form of suburbanization into areas
extending beyond Wayne county. Counties that contained relatively small urban
settlements or were somewhat rural in nature such as; Livingston, Washtenaw, Genesee,
and St. Clair in the southeastern section as well as Ottawa and Muskegon in the
southwest, continued to grow during the 19705 and 19805. Movement from the larger
cities was a major characteristic during this period. Consequently, industry as well as
other economic activities and residential housing shifted from the central cities, most
notably, Detroit in Wayne county.

Michigan's economy during the period of this study has provided jobs in primarily
three sectors; manufacturing, agriculture (including agribusiness), along with recreation
and tourism. The largest sector, manufacturing, has been overdependent on the
automotive industry which has experienced problems in recession years. It has also been
impacted by fierce global competition. The distribution of manufacturing jobs within the
automotive industry is predominately concentrated in southeastern Michigan.
Consequently, the average income per capita by counties was highest in the southern half
of the Lower Peninsula and much lower in the remaining counties Of the state. In the

19605 and 19705 nearly 50 percent of Michigan's jobs in manufacturing were in the three

36
counties of Wayne, Oakland, and Macomb, which correlates closely with the fact that on
average, about 40 percent of the state's total population lived in these three counties.
Other clusters of manufacturing activity are centered within the Grand Rapids
metropolitan statistical area (MSA), along with Lansing, Jackson, Flint and Saginaw.

The emergence of a world economy has challenged the ability of Michigan tO
adapt and compete in manufacturing, which requires new technologies and new capital
and in turn a more highly skilled labor force. Historically, it is important to understand
that Michigan's automobile industry has been and continues to be a major provider of
employment for black workers (Deskins, 1988). Michigan's heavy dependence upon
manufacturing, especially automobiles and trucks, rapidly changed during the early 19805
along with a rapid change in residential population location. Comparably, national and
regional (Great Lakes) figures for manufacturing jobs shows a nearly nine percent decline
for black employees compared to all other employees (Deskins, 1988). Combined with a
diminishing population and tax base, the economy, schools, and government services in
Michigan, especially, of Older cities such as Detroit, Grand Rapids, Flint and Pontiac have
experienced a major exodus Of job opportunities combined with a declining population
base.

One reason is that the racial make-up in Michigan's urbanized counties include a
higher proportion of blacks and other racial minorities contrasted with a lowering
proportion of the white majority population when compared to the state wide percentages
(Public Sector Reports, 1987). For example, in 1990 the city Of Detroit's (Wayne county)

racial distribution was 63 percent black, 29 percent white, and 8 percent other minority

37

groups. Furthermore, Table 3.3 reveals that the 1990 annual unemployment average for
black male adults was 17.9 percent, nearly tripling the white male adult rate Of 6.8
percent. The corresponding figures for women were 14.2 percent for blacks and 5.7
percent for whites. The unemployment rates among Hispanics fell somewhere in between
the black and white unemployment rates. The Hispanic rates for both sexes (including
teenagers 16-19 years) were approximately 11.9 percent. Historically, Hispanic‘s
minority population has ranked second to Blacks in proportion in Michigan. However,
for the period between 1970 to 1990, the highest rate of growth among all minority
population groups in Michigan has been for Asian and/or Pacific Islander's (MESC 1991).

The employment status for Michigan's labor force clearly indicates the effects Of
this economic and spatial transformation. For example, in Table 3.3, the unemployment
rates for black men and women far exceed the rates in all categories for the white and
total population in general. F urtherrnore, black teenagers experience the highest
unemployment rate of 35.9 percent compared to the white teenage unemployment rate of
14.6 percent. The Hispanic unemployment rate reveals a significant disparity between

women and men, however, their rates average between blacks and the white majority.

38

TABLE 3.3: Michigan Employment Status by Major Racial Groups,
1990 Annual Averages.

 

 

 

 

 

EmnlnxmenL‘ZuiLLQD- EnemnloxmenLRate
Total 60.5 7.5
Men 69.2 8. 1
Women 52.3 6.8
Teenagers 48.6 17.3
White 62.9 6.3
Men 71 .9 6.8
Women 54.2 5.7
Teenagers 54.7 14.6
Black 46.3 16.1
Men 5 1 .4 1 7.9
Women 42.2 14.2
Teenagers 24.5 35.9
Hispanic 59.3 1 1.9
Men 66.0 13.9
Women 53. 1 9.6
Teenagers N/A N/A

 

Source: Michigan Employment Security Commission 1992. Note: The Teenagers
category is for both sexes, 16 to 19 years of age. No Teenage data was provided
for Hispanics

 

...°., H J. --. .' u, -” .4." :.h.-' .1 m . n... II 1”
Regularignfirsmps

Distances in miles and direction in degrees fi'om the location of the TE centroids
are calculated for each census year. Distances between centroids are determined
geometrically, basically, by subtracting the X and Y UTM coordinates from each
centroid. Direction of movement for each centroid is determined by its origin of location
in a given census period and its direction of movement to its new location in the
following census period. Direction is equivalent to 360 degrees in which 0 degree equals
due north, 90 degrees equals due east, 180 degrees equals due south and 270 degrees
equals due west. The distance and direction are depicted from each of the centroids for
total employment (TE), white population (WP), black population (BP) and the population
of all other minorities (OP) for the census periods 1960 to 1990. Theoretically, if an even
distribution of population groups and employment Opportunities existed, distances
between centroids would be minimal and their directions of movement between census
years would be relatively the same. Whereas, the center of gravity for TP and TE should
be relatively the same. An uneven distribution of these centroids would reveal the
Opposite effect with separate locations between population groups and employment mean
centers. All centroids during each census year demonstrates their mean locations at the
county level and are aggregated to determine their mean center of gravity for the state.
One can roughly equate a change in distance of one mile in any direction to a change in

the mean population center by 960 persons (Jones, Manson, Coccossis, 1978).

40

Table 3.4 reveals the distances and directions for the WP, BP and OP centroids
from the TE centroid for each census year is determined. Direction in degrees is simply
determined in that a 360 circle, zero degrees is due north, ninety degrees is due east and
so forth. The OP centroid for 1960 is located the furthest from TE, 48 miles from the
center of gravity. However, during the course of the next three census periods, the OP
centroid not only decreases in distance but continually moves in the same direction as the
TE centroid, whereas, by 1990, the OP and TE centroids are approximately in the same
location. On the other hand, the WP centroid fluctuates between 33 to 37 miles from the
TE centroid either moving away or closer in a ESE direction. The WP centroid is always
in very close proximity of the TE, while continuing to move in a NNW direction.
Interestingly, the WP centroid moves in the opposite direction of BP centroid for each of

the census years.

41

TABLE 3.4: DISTANCES IN MILES AND DIRECTIONS IN DEGREES FOR
POPULATION GROUPS FROM TOTAL EMPLOYMENT (TE) "CENTERS OF
GRAVITY," 1960-1990

 

 

 

 

 

 

 

 

 

YEAR WHITE Dist\Dir BLACK Dist\Dir OTHER Dist\Dir
Miles Degrees Miles Degrees Miles Degrees

1960 7 [ 330° 35 l 133° 48 1 183°

1970 4 J 278° 33 | 134° 10 | 335°

1980 3 ] 324° 37 1 136° 3 L 36°

1990 11 | 320° =1 33 £13? 0 | 0°

 

 

 

42

TABLE 3.5: DISTANCES IN MILES AND DIRECTIONS IN DEGREES FOR
POPULATION GROUPS FROM TOTAL POPULATION (TP) "CENTERS OF
GRAVITY," 1960-1990

 

 

 

 

 

 

 

 

 

 

 

YEAR WHITE Dist\Dir BLACK Dist\Dir OTHER Dist\Dir '
Miles Degrees Miles Degrees Miles Degrees

1960 4 1 323° 374 135° 63 1 138°

1970 5 1 224° 36 1 133° 7 1 330°

1980 6 1 318° 33 1 133° 2 | 90°

1990 12 1 327° 36 1 134° 4 | 55° .

 

 

43

TABLE 3.6: DISTANCES IN MILES AND DIRECTION IN DEGREES BETWEEN
TOTAL POPULATION (TP), TOTAL EMPLOYMENT (TE) AND TOTAL
EMPLOYMENTTTOTAL POPULATION "CENTERS OF GRAVITY," 1960-1990

 

 

 

 

 

 

 

 

 

 

TP\TP Dist\Dir TE\TE Dist\Dir TE\TP Dist\Dir
YEAR Miles Degrees Miles Degrees Miles Degrees
1960 Origin Origin 2 1 340°
1970 1 | 90° 1 1 144° 4 1 14°
1980 4 L324° 6 1 324° 2 | 0°
1990 3 [216° 3 1 323° 2 1 180° 1

 

 

44

"'uw H 1-1.9,! ...“"..H :11“. .1. ”_ 1 H ..- . 1.311
40mm

Four indexes are used to determine distance in miles and direction in degrees from
the locations of the TP centroid. Distances and Directions are determined by the same
methodology for the TE centroids (see Section 3.2).

In Table 3.5, the WP centroid distances and directions from the TP center's of
gravity for each of the census periods is nearly identical to that Of the TB center's of
gravity. This situation is equally true for the BP centroids, again for each of the census
periods. The OP centroid's also display similar patterns Of distance and direction indexes
for each of the census periods. Unlike the 1990 index for the TB center of gravity in
Table 3.4, the OP centroid slightly increases in distance from the 1980 census period and

it is not identical in locale with the TP center of gravity.

WAN...” [- .H it“:_{._ .3” . 6 “I“
The main purpose behind the mapping of the centroids is to geographically
display the locations allowing an easier comparison Of centroids between each of the
census periods. In addition to the tables in the previous section, they act as a supplement
so one can quickly determine the relative location of each centroid for comparison Of
trends between each census period. Three sets of maps are used to display the relative
locations and directions for all centroids. The first set of maps, Figures 3.1 and 3.2,
depict all centroid locations and directions for the 1960 and 1970 census years. The

second set of maps, Figures 3.3 and 3.4, display centroid locations and directions for the

45
1980 and 1990 census years. The last set, Figure 3.5, highlights the BP centroids for each
Of the four census periods, 1960 to 1990. All centroids for each census period
demonstrates their average weighted locations and one can roughly equate a change in
distance of one mile in any direction to a movement in the mean population center by 960
residents.

In Figure 3.1, a tight clustering of centroids exist for TE and WP centroids. This
indicates that the center of gravity of the white population as well as the center of gravity
for employment locations are evenly distributed for these centroids. The general location
Of these centroids is best described as to where the borders Of Livingston, Ingham, and
Shiawassee counties converge. Figure 3.2 for 1970 reveals the continued clustering Of
the WP with TP and TE centroids.

The next set of maps, Figures 3.3 and 3.4, are for the census periods 1980 and
1990. In Figure 3.3, the WP centroid has reversed its direction slightly increasing its
distance from the TP and TE centroids. This trend continued for 1990 (Figure 3.4)
resulting in the greatest distance of 11 miles from TE centroid and 12 miles from the TP
centroid.

The locations Of the BP centroids in the eastern portion of Washtenaw for 1960 to
1990 are quite different from that of all other centroids. Figures 3.1 through 3.4 reveal
that the BP centroid is consistently located near the borders of Washtenaw and Wayne
counties. This is more clearly displayed in Figure 3.5, revealing that the BP centroid had
moved an average Of less than 2.1 miles during all four census periods indicating that the

black population in Michigan has experienced little or no residential change in their

46
center Of gravity or concentration since 1960. In addition, 1990 data in Table 3.4
demonstrates that the uneven distribution of employment Opportunities for the BP
centroid is over ten times the level for that of other minorities and nearly three times that
of the white majority population.

The OP centroid between 1960 to 1990 experienced the greatest change in
distribution patterns between both the TP and TE centroid. The OP centroid located in
Shiawassee county for 1970 compared to the 1960 location in Jackson county,
dramatically changed to a more even distribution with the TP and TE centroids. It seems
that a large out migration from southern counties such as; Jackson, Hillsdale, Lenawee,
and Washtenaw, took place towards where the jobs were. The largest other minority
group consist of Hispanics for all census periods. Two speculative theories as to the
major migration Of this centroid are Offered; 1) a transformation of predominately
agricultural labor force to a manufacturing laborforce, and 2) a major growth movement
of Michigan's fruit industry (located within those counties bordering Lake Michigan) to
the southeastern counties Of Michigan. The 1980 centroid location was also in Shiawasee
county. Thus, by the 1990 census year, the OP centroid also represented the most evenly
distributed group in terms of both residential and job locations.

A summary of the data and methodology is discussed in Chapter four along with

concluding remarks on the social and economic consequences of the findings.

coma .Amov coflunasaom nonuo can .Ammv
coaunasuom Ronam .Amzv eoflunasaom oufinz.amev
ucosxoaaem Hmuoe .Amev :OfiumHsmom annoy you
mowouucou coaumazmom :mmfizowz um.m ousmflm

‘
t“

 

_
a;

b‘

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

47

 

 

 

 

 

coma .Amov coaunasaom cacao can .Amm.
coflonasdom Roman .Amzv coacnasaom moanz.xmav
ucoemoamem Hmuoe .Amev :Oflumasmom Hmuoe now
moflouucoo coauoasmom cmmflnoflz "H.m ousmflm

h. L. 5.. 5.. .8

 

J
.'

 

     

 

i]

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«l
_ g8
hf

 

 

 

48

r.“

ommfl .Amov conunasaom nonuo can
coaumasmom xomam

.Ammv

.Anzv conunasaon ouanz.xmev
ucoeonmem Hmuoa .Amev cofiumasmom Hmuoe you
moflouucmu coflumHsmom cmmfl20fiz "v.m ousmflm

 

«
a.

 

h-

 

54

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

omma

.Amov coannasaom cacao can .Ammv
coflpnasaom Roman .Amzv conunasmom mannz.Auev
DCOE>OHQEM Hmuoe .Amav :Oflumasmom Hmuoa mom

coaumasmom ComanOAE "m.m ousmflm

moflouuaoo

 

49

 

1—“0

87'

 

 

Lt——_—— .
“2, sow-s

 

    

8i
a

 

 

 

 

 

 

)=>z

 

 

 

 

 

 

 

 

 

 

 

 

 

L. ..... -

"Term '

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O ”Wu-tor:
P‘ 9' P’ T’ P' E‘ P‘
Figure 3.5: Michigan Black Population Centroids, 1960 - 1990

5O

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

fl’ 0' l’ 31' 8' 0" or I?
P
I l T
0' l I T I
‘r" “an" m A“.
? lune-uh
le“ 0
2“”...""' ""‘.......°"""
ma Inn—o- :I‘uh—
Gal-- Oh-
! 0* anon-O-
no...— Iona—-
. OI..-
“WV 1”
- run-- In.
on... 0"—
aa—uv-uuu '0'"-
uua-o-u "fl."
47‘1— 00—0—- ru-I-o-o 497'
nun—-
I”
m m
coon-e M
-
‘C‘h— Wm -‘“.
anon-n
‘5'i— _"o
LEGEND
— Canon-roommate!!!”
”Woman-tun
-"'" www.m(m)
“. @ manna-mm J“.
o maumuruomw
o mama-mm nu-
o WWWIHMMM V"
on
”mm
“WM-IOUML' w: .
now-amnion w \ “w
nun-u- rum
”.— aolu n-o .m— «4:-
333 o “5' ""
o 20 co .0 fl tutti-noun .m- ~— .-
1 1 r n m l O m M a...
o 20 co m D twill”
l - j 1 1 L J I-
t“ t
a... an non-i 1 on
0o
um ; ~
on- o
”.- _ 0'- ulorron m a
a... uml anna- m DHRO‘T' 4r
‘2"- limos '—' Au"
moon a. III- “I...“ ” ARBOR
V I j
‘ n \J
"o 01.8.de 1 1 1 1 J wvmw-lu'
fl’ 0" U' 01' I' I‘ 00‘ 9‘

 

 

Figure.3.6: Michigan - County Locator Map

CHAPTER 4

SUMMARY AND CONCLUSION

The Objectives of this study were to determine: (1) the extent Of residential and
employment segregation at the county level in the State Of Michigan; (2) the patterns and
distributions Of both population and employment centroids; and (3) if Significant
differences exist in the distances and the directions for each centroid during each census
period on these factors. Three different sources Of data were used to determine population
characteristics and employment Opportunities for the period of this study. Centrographic
methods were employed to determine location and distances of the White, Black,
Hispanic and total populations for the State of Michigan. In addition, a total employment

centroid was determined.

51

52

The following hypotheses were tested:

1. The patterns of residential location of racial groups is related to the pattern Of
location of employment Opportunities.

2. Some racial groups are located closer to employment opportunities than other
racial groups.

3. The patterns of residential concentration of racial groups and the patterns of
location of employment oppurtunities are dynamic, that is changing fiom decade
to decade.

Contrary to the first stated hypothesis, it was found that the BP centroid was the
only minority population group segregated both residentially from all other population
groups and economically from employment Opportunities. During the 1960 census
period, the OP centroid exhibited the greatest distance from both TP and TE. However,
by the 1990 census period, the OP centroid exhibited the least differentiation both in
direction and distance from the TP as well as the TB centroid.

The test of the second hypothesis revealed that both the WP and the OP centroids
exhibited the least distances from the TE centroid. In fact, by the 1990 census period, the
location of the OP centroid was nearly identical to that the of TE and TP centroids.

The third hypothesis held true for both the WP and OP centroids as well as the TE
centroid. These centroids exhibited a significant change in location during each of the
census periods. On the other hand, the BP centroid location changed an average distance
of one mile during each census period. Basically, the BP centroid’s change in location

during each census period was negligible indicating that their pattern of concentration has

53
remained relatively constant.

During the forty year period Of this study two important findings were revealed.
First, the black population within the state of Michigan has been, and continues to be
segregated from the center Of gravity of the total population and the center Of gravity for
employment Opportunities as well as from all other minority population groups. Since
1960, black residential location has centered in Wayne county as evidence in Figure 3.5.
Secondly, the change of location and direction of the WP centroid precedes the change Of
direction and location of the TE and TP centroids. In other words, where the center Of
gravity of the white population resides, the center of gravity of the TP and TE centroids
follow. On the other hand, where the white population resides, employment opportunities
will follow. The Opposite scenario has been true for blacks.

. The concern of this research was measuring the economic Opportunity structure
within the state of Michigan for various racial groups. Although previous research
methods were adequate for determining small-area patterns and distributions of economic
Opportunities between various population groups, it was necessary to use centrographic
methods to better assess broad patterns of economic Opportunities.

Understanding the dominance of Detroit and the southeast Michigan region is
crucial since not only does this area contain a large share Of the population, industry, and
income Of the state, it is also the largest city1 in the United States in which the majority Of

the total population is black. A major source Of new jobs in Michigan is and will

 

The District of Columbia contains the highest proportion of Blacks amongst a given
population but is not considered a city but rather a "District".

54
continue to be small businesses (those employing fewer than 100 workers, MESC 1992).
The location Of these small businesses has depended heavily on those areas that Offer not
only tax credits and tax abatements but a clean and safe environment as well as good

schools and nearby recreational opportunities.

4.1:.C9nszlusinn

The social and economic consequences for blacks is crucial as the findings
indicate. The general spatial trend of the TE and the TP centroids seems to indicate an
increase in spatial segregation from employment Opportunities and fi'om the remaining
population groups Of the state since the 1970 census year. If this trend continues into the
next century and beyond, the black population would continue to experience higher
unemployment rates than the white majority, and other minority population groups.

Socially, blacks, the largest minority population group, has been
disproportionately isolated in one county in the state. As other racial groups migrated
westward to other counties within the state, blacks seem to have maintained their same
location. Economically, the same case may be presented for employment opportunities.
Since, the mid 19705, the state has Slowly begun a transformation of their economic
structure from the intensive automobile manufacturing environment Of the past into
various other technologies and service industries. Spatially, the location Of these new
economic ventures have move outward fiom their original locations, further isolating the

black labor force from employment opportunities.

55

Results from this study clearly indicate that blacks are located where jobs and the
overall distribution of the state's population are not. Moreover, it is a trend that seems to
be continuing. The location of all other racial population groups and jobs are becoming
more and more spatially segregated from the black population. The findings here reveal
that at this broader scale of analysis, the extent of residential and employment segregation
between blacks and whites and between blacks and other racial population groups is a
state wide phenomenon. Consequently, one can determine that progress for blacks since
the Civil Rights Act Of 1964 has been more symbolic than real as indicated during the
forty year period of this study. Furthermore, residential and employment segregation
remains virtually as pervasive today as over a quarter-century ago.

Further research might examine the state of Michigan's political processes by
which legal policy is translated into "real world" changes in social-economic behavior.
Also, similar centrographic measures might be employed to determine the extent in which

blacks experience similar spatial phenomenon in other states.

APPENDICES

56

APPENDIX A

UTM Decimal/Cartesian Coordinates of Centroids for Total Population (TP), White
Population (WP), Black Population (BP), Other Population (OP), and Total Employment

 

TE , 1960-1990

YEAR / POP x COORD Y COORD x COORD Y COORD f
_____ ______ _________________ __ ______________~_-__o-_-.. 1

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1960 TP 84.1 1423 42. 80426 84°06'51" 42°48'15"
WP 84.16544 42.84304 84°09'55" 42°50'34"
BP 83.60158 42.40958 83 °36'05" 42°24'34"
OP 84.39354 38.75419 84°23'36" 40°45'15"
TE 84.09844 42.76194 84°05'54” 42°45'42" I
1970 TP 84.09810 42.80523 84°05'53" 42°48'18"
WP 84.15896 42.19611 84°09'31" 42°45'15" l
BP 83.61360 42.41857 83°36'48" 42°25'06"
OP 84.15866 42.87854 84 ° 09'31" 42 ° 52'42" I
TE 84.07694 42.75723 84°04'36" 42°45'26" I
1980 TP 84.15066 42.84632 84°09'02" 42°50'46" I
WP 84.22824 42.90976 84°13'41" 42° 54'35" "
BP 83.64477 42.42929 83 ° 38'41" 42 °25'45" 1
OP 84.1 1294 42.84784 84°06'46" 42°50'52"
TE 84.14464 42.81111 84°08'40" 42°48'39"
1990 TP 84.18149 42.82328 84° 10'53" 42°49'23" I
WP 84.31020 42.96581 84°18'36" 42°57'56" J
BP 83.66751 42.44931 83 °40'03 " 42°26'57" I
OP 84.12417 42.84522 84°07'27" 42° 50'42" I
TE 84.18548 42.84260 84° 1 1'07" 42° 50'33 "

5 7
Appendix B

The following is the formula to determine distance between any two given UTM
coordinate of centroids or points:

(X-X)‘°+ (Y-Yf'

D
II

= X + Y

= r2“

LIST OF REFERENCES

LIST OF REFERENCES

Blumrosen, AW. 1993. Medern Law. University of Wisconsin Press

Bowyer, A. and Woodwark, J. 1983. A Eregrgmrnerjs Geometry. Butler and Tanner Ltd.:
Fromme and London

Burgess, E.W. 1969. "Behavior and Conduct," in Innedjrcjjen 19 the Seience ef
$93191ng (Ed.) Morris Jankowitz. Chicago: The University of Chicago Press

. 1969a. "Segregation as a Process," in Ingednerien In the Seienee Qf Snejnlngy
(Ed.) Morris Jankowitz. Chicago: The University Of Chicago Press

Cason, D., Jr. "Michigan's State Urban Strategies," prepared for: The National Academy
Of Public Administration. The States and Urban Strategies: Case Studies and
Comparative Analysis. Germantown, Maryland: HUD USER. September 1980

Christaller, W. 1966. Central Blnees in Sthern Germany. Translated by CW. Baskin.
Englewood Cliffs, N.J.: Prentice-Hall.

Clarke, KC. 1988. Analnjeal and Cempnter Cartography. Prentice Hall: Englewood
Cliffs, New Jersey. Chapter 9

Clark, W.A.V. 1984. "Population Geography at Micro-Scale: Residential Mobility and

Public Policy," in Geography and Penulatmn (Ed.) John 1. Clarke. Oxford,
England: Pergamon Press Ltd. pp. 77-84

Cole, RE. and Deskins, D.R., Jr. 1988. “Racial Factors in Site Location and Employment
Patterns of Japanese Auto Firms in America.” California Management Review.
31(1): 9-22

Cooper, L. 1963. "Location and Allocation Problems" Qperatiensfieseareh. 11:331-43

Court, A. 1964. "The Elusive Point of Minimum Travel" Annals ef the Assneinrjen ef
Amman Shimmer:- 54(3):400-03

58

59

Darden, J. T. 1980. ”Lending Practices and Policies Afl'ecting the American
Metropolitan System,” In 8D. Brunn and J .0. Wheeler (Eds), The Ameriean
Metropolitan System: Present and Eurure. New York: Halstead Press. Chapter 7

. 1987. "Choosing neighbors and neighborhoods: the role of race in housing
preference." In Disaster! Neighborhoods: The Charming Patterns ofRacial
Segregarien, G. Tobin (Ed) Urban Annual Review 32. Newbury Park: Sage pp.
15-42

, and Kamiar, MS. 1988. "SociO-Economic Development of Black Ghetto
Tracts in Michigan: The Microcosm Model Reconsidered," Ihe East Lakes

Cmegmnher. 23:71-84

and Tabachneck, A. 1980. "Algorithm: Graphic and Mathematical
Descriptions of Inequality, Dissimilarity, Segregation, or
Concentration. "Enyirenmem and Planning. 22227-234

 

Deskins, D. R, Jr. 1972. "Race, Residence, and Workplace in Detroit, 1880 to 1965."
Emnomicfieozrnohy 48179-94

. “Michigan’s Restructured Automotive Industry: It’s Impact on Black
Employment,” in The State of Black Michigan: 1233. F. S. Thomas (ed)-
Publisher: Urban Afl’airs Programs, Michigan State University. Chapter Three.

Duncan OD. and Duncan B. 1955. “A Methodological Analysis of Segregation Indices.”
American Sociological Renew. 202210-217

Farley, J. E. 1987 . "Disproportionate Black and Hispanic Unemployment in US.
Metr0P01itan Areas," The American Joumal of Economics and Sociology.
46(2): 129-50

Greene, R. 1991. "Poverty Concentration Measures and the Urban Underclass,"
Economic Geomnbx- (”(3)1240-52

Harrison, B. 1974. Urhan Eeonernie Dexeleprnent. Washington, DC: Urban Institute
Press

Hawley, A. 1950. Human Ecolorn. New York, NY: Ronald Press.

Heyford, J.F. 1902. "What is the Center of an Area, or the Center of a Population,"
Journal of the Amnican Statistical Association New Series, 58:47-58

60

Hill, RC. 1982. ”The Crisis of Industrial Cities,” In Qn'rienl Issues in the 1289}:
Problems and Resmnses, Thomas, Francis S. (Ed) A Discussion Series of the
Urban Afi‘airs Programs, Michigan State University.

Hill, RC. and Negrey, C. 1987. ”Deindustrialization in the Great Lakes" Urban Afi‘airs
ommly. 22: 580-597

Hollister, D. C. 1989. "Fiscal Update #10," State W Ruhliearions. Michigan
House of Representatives. Lansing: Michigan

Holzer, H. J. 1991. "The Spatial Mismatch hypothesis:What Has the Evidence Shown?"
Urban Studies 28(1)1105-7-2

Hughes, MA. 1987. "Moving Up and Moving Out: Confilsing Ends and Means About
Ghetto Dispersal," Urban Studies. 24:503-17

. 1989. "Missng Truth To Power: A Geographical Perspective On The
‘Underclass' Fallacy," Eeonomie Qeograrmy 65(3):l87-207

and Madden, J. F. 1991. "Residential Segregation and the Economic Status of
Black Workers: New Evidence for an Old Debate," Journal offlrhnn Eeonomies.
29:28-49

 

Jackson, P. 1984. Exploring Social W. Boston: G. Allen & Urwin Publishers.
Chapter 8

Jones, B.G., Manson, D.M. and Coccossis, RN. 1978. “Applications of Centrographic
Techniques To The Study Of Urban Development: The Expansion Of

Metropolitan Atlanta, l940-197S.”_H11D Research Grant W. Task 111
Report. Published: Program in Urban and Regional Studies: Cornell University.

Kain, J. F. 1965. "The Efl‘ect of the Ghetto on the Distribution and Level of Nonwhite
Employment in Urban Areas. " Harvard University Consultant: The RAND
Corporation.

. 1968. Housing Segregation, Negro Employment, and MetrOpOlitan
Decentralization. Quarterly Journal ofEeonomies: 82(2): 175-97

. 1974. Housing Segregation, Negro Employment, and Metropolitan
Development: A Retrospective View. In Eanems official Discrimination, Mo]. 11:

Employment and Income, (Ed) George M. Von Furstenberg. Lexington: D.C.
Heath

61

. 1986. The Influence of Race and Income on Racial Segregation and Policy. In
Housing Winn, Race, and Eederal Policies (Ed-1101111 M- Goering-
Chapel Hill: University of North Carolina Press

. 1987. The Interaction of Residential Segregation and Employment
Discrimination. Jonmal ofllrban Economies 21 :323-46

Kasarda, J. 1989. "Urban Industrial Transition and the Underclass," The Annals of the
AmencanAcademyoonliticalandSocialScience 501:26-47

Kumler, MP. and Goodchild, MF. (1990) “The Elusive Population Centre of Canada.”
The University of California at Santa Barbara: An Unpublished Paper

Lang, M. 1979. "Historic Settlement and Residential Segregation in Rural Neighborhoods
of Jasper County, Mississippi." PhD Diss., Michigan State University

Leonard, Jonathan S. 1984. "The Interaction of Residential Segregation and Employment 1
Discrimination" BEER NO. 1274. ’

Lovelock, CH. and Weinberg, CB. 1985. Marketing Challenges: Cases and Exercises.
McGraw Hill Co. pp. 273-295

Massey, D. S., and Denton, N. A. 1989. "Hypersegregation in US. Metropolitan Areas:
Black and Hispanic Segregation Along Five Dimensions," Demography.
26(3):373-9l

., and Eggers, M. L. 1990. "The Ecology of Inequality: Minorities and the

Concentration Of Poverty, 1970-1980,” Ameriean Journal of Soeiology. 95: l 153-
87

Michigan Employment Security Commission (MESC). 1992. Area m Srarisries,
Bureau of Research and Statistics, Labor Market Analysis Section. Publisher;
MESC: Detroit, Michigan.

(MESC). 1981. Area Labor Srarisries, Bureau of Research and Statistics, Labor
Market Analysis Section. Publisher; MESC: Detroit, Michigan.

Michigan Statistical Abstract. 1961. Michigan State University. Graduate School of
Business, Division of Research. Published by Bureau of Business Research,
College of Business and Public Service: East Lansing.

62

. 1971. Michigan State University. Graduate School of Business, Division of
Research. Published by Bureau of Business Research, College of Business and
Public Service: East Lansing.

. 1981. Michigan State University. Graduate School of Business, Division of
Research. Published by Bureau of Business Research, College of Business and
Public Service: East Lansing.

National Advisory Commission on Civil Disorders. 1968. Began; of rhe Nan'onal
Adxisont Commission on Ciril Disordm. New York: Bantam Books

O’Hara, W. P., (et.al.) 1991. “African Americans in the 19905.” Population Bullenrin.
(July) 46:1

Palumbo, G.; Sacks, S.; and Wasylenko, M. 1990. "Population Decentralization within
Metropolitan Areas: 1970-1980," lourml offlrban Eeonornies. 27: 151-167

Park, R. E. 1969. "Temperament, Tradition, and Nationality," in Inrroduerion to me
Science of Sociology (Ed) Morris Jankowitz. Chicago: Univ. of Chicago Press

. 1969a. "Behavior and Conduct," in Introduerion to the Seienee of Sociology
(Ed) Morris Jankowitz. Chicago: The University of Chicago Press

. 1969b. ”Segregation as a Process," in Introduerion to the Science omeiology
(Ed) Morris Jankowitz. Chicago: The University of Chicago Press

. 1969c. "Racial Conflicts," in Introduction to the Science of Sociolomt (Ed-J
Morris J ankowitz. Chicago: The University of Chicago Press

Soja, E. W. 1971. "The Political Organization of Space,"Commission on College
Geography. Resource Paper No. 8. Association of American Geographers.
Washington, DC.

Thurow, L. 1975. Generating Inequality. New York: Basic Books

Tordella, S. J. 1987. "How To Relate To Centroids," Demographics. 9(5):46-50

US. Bureau of the Census. 1961. ”1960 Census of Population and Housing,” General
Population and Characteristics, Michigan. Washington, DC: US. Government
Printing Ofice.

US. Bureau of the Census. 1961a. "United States Summary," Volume lzChapters B & D.

63

US. Bureau of the Census. 1971. ”1970 Census of Population and Housing," General
Population and Characteristics, Michigan. Washington, DC: US. Government
Printing Ofice.

US. Bureau of the Census. 1982. "1980 Census of Population and Housing," General
Population and Characteristics, Michigan. Washington, DC: US. Government
Printing Ofice.

US. Bureau of the Census. 1992. "1990 Census of Population and Housing," General
Population and Characteristics, Michigan. Washington, DC: US. Government
Printing Ofice.

US. Bureau Of the Census. 1993. "Definitions of Subject Characteristics," Tables P-9 and
P-lO. Washington, DC: US. Government Printing Ofice.

Updegrave, W. L. 1989. "Race and Money," Money. Chicago: Time Inc. (12):22-3O

White, M. J. 1983. "The Measurement of Spatial Segregation," Amman lournal of
Sociology. 88(5):]008-19

Williams, B. B. 1987. Blank Worker in an Indusrrial Suburb. Rutgers University Press:

New Brunswick.

Wilson. WJ. 1978. The Declaring Significance ofBacc: Blacks and Changing American
Instinrtions. Chicago: University of Chicago Press.

.1987. TheInrlxDisaoxantagedHTthnnerCitx tthJnrlerclais. andhlhlic
Poliey. Chicago: University of Chicago Press

 

"IurnTrill/.11“