Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Integration And Conflict: Initial Perspectives On Urban Race Riots
(USC Thesis Other)
Integration And Conflict: Initial Perspectives On Urban Race Riots
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
This dissertation has been
microfilmed exactly as received
70-8528
JIOBU, Robert Masao, 1939-
INTEGRATION AND CONFLICT: INITIAL PER
SPECTIVES ON URBAN RACE RIOTS.
University of Southern California, Ph.D., 1969
Sociology, race question
University Microfilms, Inc., Ann Arbor, Michigan
INTEGRATION AND CONFLICT: INITIAL
PERSPECTIVES ON URBAN RACE RIOTS
by
Robert Masao Jiobu
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Sociology)
January 1970
UNIVERSITY O F S O U T H E R N C A LIFO R N IA
T H E G RA D U A TE S C H O O L
U N IV E R SIT Y PA RK
LO S A N G E L E S, C A L IF O R N IA 9 0 0 0 7
This dissertation, written by
Robert M. Jiobu
under the direction of h.13... Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by The Gradu
ate School, in partial fulfillment of require
ments for the degree of
D O C T O R O F P H I L O S O P H Y
D ean
August 1969
D ate...-...... °............
DISSERTATION COMMITTEE
ACKNOWLEDGEMENTS
I would like to thank my Chairman, Dr. Herman Turk,
whose encouragement and perceptive advice was invaluable
in writing this dissertation. I would also like to thank
my wife, Karen, for her understanding during the course
of completing this undertaking.
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ii
LIST OF TABLES V
LIST OF EXHIBIT vi
Chapter
I. THE NATURE AND APPROACH TO THE PROBLEM . . 1
Integration
The General Concept of Integration
Integration in the Urban Arena
Conflict
The General Concept of Conflict
Conflict in the Urban Arena
Conflict and Integration
Conflict and Integration as Joint
Processes
Conflict and Urban Community
Integration
Conclusions
Urban Race Riots
Accounts of Urban Race Riots
Interpretation and Conclusion
The Comparative Approach
Nature of the Problem
Plan of the Dissertation
II. INTEGRATION AND CONFLICT 6
iii
Chapter Page
III. A CONCEPTUALIZATION OF URBAN INTEGRATION
AND RACE RIOTS............................. 32
Urban Interdependence Model for
Race Riots
Contrasting Model
Conclusion
IV. METHOD AND PROCEDURES...................... 50
Universe of Inquiry
Measures of Conflict
Measures of Integration
Control Variables
V. RESULTS AND DISCUSSION.......................80
Riot Proneness and Severity
Extralocal Integration
Local Integration
Assessment of Spuriousness
VI. SUMMARY AND CONCLUSIONS.................. 101
Summary
Further Theoretical Comparisons
Further Implications
Conclusions
APPENDICES: PRELIMINARY RESULTS ................ 112
I. Prediction of Riot Activity: Low
Extralocal Integration ................ 113
II. Prediction of Riot Activity: Medium
Extralocal Integration ................ 114
III. Prediction of Riot Activity: High
Extralocal Integration ................ 115
BIBLIOGRAPHY ....................................... 116
iv
LIST OF TABLES
Table
1. Intercorrelations Among City Riot Prone
ness and Measures of Riot Severity .
2. Intercorrelation of All Variables .
3. The Relation between Extralocal Integration
and Riot Activity Holding Constant the
Effects of Racial Polarization
4. The Relation between Negro-White Polarity
and Riot Activity Holding Constant the
Effects of Extralocal Integration and
Number of Police .......................
Page
61
82
85
87
v
Exhibit
LIST OF EXHIBIT
Page
1. Description and Names of Variables . . . 52
CHAPTER I
THE NATURE AND APPROACH TO THE PROBLEM
The focus of this dissertation is on two questions.
First, do urban communities possess structural character
istics which facilitate the outbreak of race riots?
Second, if so, under what conditions are these character
istics operative? In attempting to answer these questions,
a model of integration and conflict is developed utilizing
the comparative approach.
The Comparative Approach
The comparative approach attempts to realize the
advantages of a contro-lied experiment but in a natural
setting (Wilier, 1967). As applied to urban studies, it
implies a comparison between many cities utilizing sys
tematic controls (Gibbs, 1961). While the comparative
approach has been treated at some length and applied to a
variety of topics,1 the study of urban communities has not
■*"For a more extensive discussion and examples of
its application, see Sirjamaki (1964) and Udy (1965).
1
been accorded such treatment on a large scale (Strauss,
1968) . While case studies often serve as rich repositories
of ideas and as sources for hypotheses, they are, of
course, limited in that no generalizations are possible
and no systematic controls may be applied (Stouffer, 1950;
Campbell, 1957). It is for these reasons that a compara
tive approach has been adopted. Moreover, the emphasis
on structural characteristics is more amenable to compara
tive analysis on the pragmatic, but very important, grounds
that to obtain data on the internal dynamics of a large
sample of urban communities would be prohibitively costly
and time consuming.
Implicit in the adoption of the comparative
approach is the emphasis on a macro level of theory. The
focus is on deriving and testing theoretical propositions
that may be applied to a large number of urban communi
ties. Inasmuch as the urban community in toto will serve
as the unit of analysis, the level of abstraction will
2
The term macro refers to the use of relatively
large scale units of analysis (e.g., cities, states or
total societies) (Martindale, 1960:464).
3
necessarily be at a macro level, especially when coupled
with the comparative approach.
Nature of the Problem
If there have been relatively few studies of urban
phenomena utilizing the comparative approach, it would be
expected that studies of urban race riots in particular
would be even fewer in number. The next chapter will show
that such is the case. As a result of this lack, there
is relatively little theory and data bearing on the
specific problem at hand.
This lack of knowledge has been undoubtedly
influenced by the lack of comparative data. Fortunately,
recent research has produced data in sufficient quantity
to make the present undertaking practicable, at least in
a preliminary way. The United States Census and various
other reports from previous research may be collated and
serve as the data source for this study.
As will be apparent in subsequent chapters, the
primary guidelines for this research is the utilization
of traditional sociological variables as these are applied
to urban social structure. As a result, integration and
conflict, two variables which have been used in sociology
for many decades, are discussed in some length. The
reasons for this emphasis are twofold. First, the lack
of previously applied theory in the case of the present
problem forces one to use that which is available. Second,
and perhaps more important, there is no evidence yet that
urban events differ from other phenomena in such a funda
mental way such that the notions of integration and con-
3
flict are inappropriate. With certain modifications, it
is felt that such concepts are applicable but have not
been explored to their fullest extent.
Another guideline utilized in this dissertation is
the idea that riots are not simply random, unpatterned,
bizarre forms of mass behavior. Evidence supporting this
contention has been presented by several researchers (cf.
Turner and Killian, 1957; Lieberson and Silverman, 1965;
Wanderer, 1969). Furthermore, were the assumption
accepted that riots are a random form of behavior, it
would make little sense to attempt to predict and explain
them in scientific terms.
3
This is not meant to imply that the specific
nature of the urban setting can be disregarded.
Plan of the Dissertation
The pages to follow present a review of the
relevant literature pertaining to integration, conflict,
and the relationship of urban social structure to race
riots. Next, a model based on suggestions stemming from
that review will be offered, and contrasted to a Marxian
based model. Testable hypotheses will then be derived.
Finally, an empirical test of the hypotheses will be
undertaken.
CHAPTER II
INTEGRATION AND CONFLICT
This chapter will present a portion of the
literature which is directly relevant to the purpose of
the current undertaking. In one sense, a review of the
literature concerning integration would encompass most of
what has been written in the field, since integration is
one of the basic independent and dependent variables of
theoretical sociology. Therefore, selectivity must be
exercised. The criterion of selection used is applicabil
ity towards the model to be proposed in the next chapter.
Integration
The general concept of integration will be examined
first. Next, applications of the concept to the urban
community will be reviewed.
The General Concept of Integration
Historically, one major approach to the study of
integration may be termed "typological" (Loomis, 1963).
In this approach, societies, communities or groups are
characterized by distinctive types of integration and
placed on opposite ends of a continuum or juxtaposed
against each other for the sake of contrast.
Durkheim (1933) discussed the following two types
of solidarity: mechanical solidarity, based upon shared
orientations towards common goals and norms; and organic
solidarity, based upon interdependence— a function of the
division of labor that arises, in turn, as a response to
the moral density which ensues from population density.
Along similar lines, Toennies (1957) utilized the
notions of community and society. The former was charac
terized by the solidarity arising from natural will and
the latter by solidarity arising from rational will. In
turn, natural will was that which was a way of life and
not a means to an end. Rational will, in contrast, was
will which viewed relationships as means to further ends.
Redfield (1947) argued for the folk-urban dimen
sion. The folk society was one which was small enough
that each person could know every other person; relatively
isolated, nonliterate and homogenous. By implication, the
urban society was one which was characterized by the oppo
site pole of these dimensions.
8
Finally, at the level of small groups, Cooley
(1909) developed the concepts of primary and secondary.
A primary group was one which was characterized by face
to face relations, intimacy, "we feeling," small numbers,
and unspecialized relations. In contrast, a secondary
group was one which did not have these characteristics.
Interpretation. It is interesting to note that
these theorists all utilize a typology which implies that
one rough polar type, the mechanical-community-folk-
primary mode of integration, is perhaps most characteristic
of older and/or more primitive societies. Since this
dissertation is concerned with modern urban communities,
that pole probably does not exist in pure form for present
purposes (if indeed, it ever did). It is certainly not
reasonable to assume that every person in an urban commu
nity knows every other person, as the label folk society
would require. Nor is it reasonable to view urban resi
dents as interacting with most people on the basis of
natural will. In other words, the defining characteristics
of one polar type of integration probably do not exist
for the modern urban community.
However, rather than view the polar types as ideal
types which do not vary, this dissertation views their
characteristics as variables, thereby allowing one to
speak of them as being more or less approximated in a
given community.
Integration in the Urban Arena
Concerning the integration of urban communities,
it may be seen that variations of the preceding ideas have
been applied by a group of theorists utilizing the indi
vidual as the basic unit of analysis. In this view, urban
integration is seen as being composed of the integration
of each individual with other individuals.
Another group of theorists have generalized the
concept in a manner that permits the use of organizations
as the basic unit. Integration is viewed as being derived
from the relationships between organizations as well as
from the relationships between organizations and indi
viduals .
The individual as the unit of analysis. Landecker
(1951) has claimed that integration is not a single
variable; rather, he argues that it is more meaningful to
speak of four types of integration. The first, cultural
integration, is defined by the degree of consistency
among the standards existing within a given culture. The
second, normative integration, refers to the degree to
which these standards constitute effective norms for the
population. The third, communicative integration, is
defined by the degree to which there exists an exchange
of meaning amongst the populace. The fourth, functional
integration, refers to the compatibility of the exchange
of goods and services, or mutual exchanges, amongst the
populace. In all four cases, it may be seen that Landecker
is referring to the integration of the populace rather
than that of organizations.
Angell (1951) has spoken of the "moral integra
tion" of American cities. Using a ratio of crime rates to
welfare efforts as a measure of integration, he found that
this index was negatively associated with in-migration
and ethnic heterogeneity. The general interpretation of
this work is that stability amongst the population affects
the moral integration of the city.
Hillery (1968) has constructed a model encompassing
both the folk village and the modern city. Utilizing case
study materials on a comparative basis, he claims that
city integration is best viewed as a variation of the same
11
components that comprise the integration of more primitive
villages. These components are localized space, mutual
interaction and activity, and common sentiment. These he
subsumes under the term localized system. When applied
to the modern city, Hillery argues that these components
can be translated into contracts as the links between
people and families as the basic integrating unit.
Interpretation. Both Landecker and Angell speak
of integration in terms of the populace of the city.
Basically, the integration of the city is thus seen as
the sum total of the integration of the populace residing
within it. In other words, integration is achieved using
the individual as the basic unit of analysis.
Hillery moves a step away from this approach as
he speaks of contracts and families as basic to the inte
grative process. Contracts exist between people and in
this sense the basic unit is still the individual. On
the other hand, families are a type of group, and to the
extent that families are the basic unit of integration,
to that extent he moves towards a more macro level of
analysis.
12
Organizations as the unit of analysis. Warren
(1956a) has spoken of community integration in terms of
the organizations that reside within it. In this view,
integration at the community level is achieved via the
relation of the individual to the organization, and the
subsequent integration of the organization with other
organizations.
Emery and Trist (1965) have explained a related
aspect in their argument that the environment of a commu
nity may be characterized by the richness of the organiza
tions within it. Furthermore, the movement of the
environment (termed a turbulent field) causes organiza
tions to continuously respond to change and to interact
with other organizations.
Turk (1969a, 1969b) has woven together the ideas
of Warrei}, and Emery and Trist, pointing out that the
greater the number of organizations within a community,
the greater the likelihood that an individual will belong
to one. Therefore, the greater the interorganizational
integration, the greater the likelihood that one can speak
of the total community as being integrated.
Speaking of a subpart of a total community,
Stinchcombe (1965) has theorized that the integration of
ethnic communities is a function of that community1s
penetration by organizations. In this view Stinchcombe
is not alone. Kitano (1969) has remarked on the widespread
number of ethnically oriented organizations among the
Japanese-American community on the West Coast. He has
attributed the solidarity of this group, in part, to the
presence of these organizations. Lopata (1964), in study
ing the Polish-American community in Chicago, concluded
that the vast number of ethnically oriented voluntary
associations within that community functioned to both
integrate the community and, at the same time, help link
it to broader American society.
Viewing community integration in organizational
terms, Warren (1963) has further used the concept of
horizontal integration to refer to the integration of the
community qua community, and vertical integration to refer
to the integration of the community into the broader
society. In a study of "Dairyville" (1956b), he proposed
that the local autonomy of a community is disrupted by a
number of organizations which reside within the community
but which do not respond to community needs. As an exam
ple of vertical integration, all the churches in Dairyville
were affiliated with headquarters outside the community.
14
In contrast, many of the smaller industries were locally
owned and responded to community needs, thus contributing
to the horizontal integration of Dairyville. Turk and
Jiobu (1969), and Turk (1969b) have used similar con
cepts. "Extralocal integration" was viewed as the number
of linkages a community had to the broader society, while
"local integration" described the integration of the
community independent of its extralocal linkages.
Interpretation. Considering the present state of
knowledge, it might seem that it is a matter of choice
which unit of analysis, the individual or organization,
one chooses to accept.
However, the choice is really not an arbitrary
one. Integration may exist at different levels of
abstraction. It would appear that integration in terms
of organizations is more, appropriate for a systems or
macro approach because organizations can then be viewed
as parts of a system. This is certainly more parsimonious
than taking each individual as a separate part of the
system.
Furthermore, it would probably be a mistake to
uncritically accept integration based on likeness of
15
individuals, common orientation, common goals and values,
and contracts to be the same phenomenon as integration
based on the mutual compatibility and smooth articulation
of organizations with each other. The former bespeaks
the integration of individuals, the latter bespeaks the
integration of a system.
Finally, it may well be that both types of inte
gration exist simultaneously. The type a researcher finds
may be a function of the level of abstraction with which
his theory deals. Theoretically, it is conceivable that
an urban community may be highly integrated vis-a-vis its
organizations but only slightly integrated vis-a-vis its
populace. Individuals may simply not belong to these
organizations in appreciable numbers. The converse of
this situation is also possible. In order to assess the
empirical relevance of these possibilities, a researcher
would have to examine both aspects.
Conflict
Turning now to the concept of conflict, it has
been pointed out (Coser, 1964:123-24) that conflict has
been viewed from two perspectives in the sociological
literature. One perspective views it as basically
disjunctive while the other views it as an integral part
of social behavior. The view of conflict as disjunctive
carries with it an implicit value judgment that conflict
is somehow "bad," "evil," or to be avoided.^ In order to
avoid this, as well as for other reasons to be explicated,
the view of conflict as a basic, integral component of all
social behavior is the one to be utilized here.
The General Concept of Conflict
In this section, the notion of conflict in general
will be examined. This is appropriate inasmuch as some of
these notions will be used later to build a model of urban
conflict and integration.
The ubiquity of conflict. The idea that conflict
is ubiquitous has been forcefully stated by Marx (1964:57)
The reason that this point is emphasized is that
the value implications of conflict are apparently quite
pervasive even in scientific works. For example, the
Journal of Conflict Resolution carries the word "resolu
tion" in its title implying that conflict should be re
solved. A standard social problems textbook (Merton and
Nisbet, 1966) has sections devoted to intergroup conflict,
community conflict and international conflict; all under
the guise of a "problem."
"The history of all hitherto existing society is the
history of class struggles." A similar point has been
made more recently by Dahrendorf (1958) who argues that
the absence of conflict is the abnormal state of society
rather than the reverse. The empirical documentation
of such assertions on the societal level would be an
enormous task. However, on a social psychological level
of analysis there has been much documentation of the
ubiquity of conflict. For example, in the field of
marriage and family, Blood and Wolfe (1960) view the
presence of some conflict as normal and inevitable in the
relation between man and wife; and their study empirically
demonstrates a wide variety of types. Similarly, in the
study of small groups, Bales (1949) has viewed and
measured the development of small groups as a response to
the various "strains" (implicit conflicts) generated by
the requirement that groups achieve objectives on the one
hand, and the need for tension release on the other. Also,
in an industrial setting, Blauner (1964) has shown that
the incompatibility of the technology and control of
automobile assembly lines, the need for self-fulfillment
on the part of workers, and the economic requirements of
18
production produce ever present conflict in the relation
between worker, union and management.
Conflict in the Urban Arena
Having thus examined the general concept, and
recalling that the purpose of this dissertation is to
examine the structural aspects of urban communities, the
focus of the next sections narrows to conflict and urban
social structure.
The ubiquity of urban conflict. If conflict is
ubiquitous on the societal and social psychological levels
of analysis, one would also expect it to be ubiquitous in
the urban arena. This would seem to be the case, at
least as far as the present state of knowledge allows a
judgment. Hillery (1968), in comparing five different
types of urban communities, concluded that conflict is
ever present. However, he went on to further argue that
the integrative components ameliorate and neutralize the
conflict components.
Coleman (1958), in a comparative study, demon
strated the widespread existence of conflict and set forth
a process model documenting it with evidence of school
conflict in a variety of communities. Even in what
19
Redfield (1941) described as a conflict free village,
Lewis (1951), upon reexamination of the same village
several years later, found a great deal of conflict which
was originally overlooked.
Interpretation. Considering both the general
concept and urban conflict, there seems to be little
doubt that it constitutes a major area of social behavior
which, until recently, has not been emphasized. By con
centrating on theories of integration, researchers have
thus overlooked a potential source of explanatory con
cepts .
Conflict and Integration
Thus far, conflict and integration have been
viewed as being mutually exclusive. The presence of one
precludes the presence of the other. It is possible,
however, to view the two concepts conjointly.
Conflict and Integration as Joint Processes
Over fifty years ago, Simmel (1955) approached
the question of conflict and integration arguing that
conflict is one portion of a social relationship. Speak
ing of such diverse groups as family and caste, he
20
maintained that conflict defined one part of the relation
ships while the other part was constituted by various
unifying forces such as love, common attitude and super
ordination and subordination. To overstate Simmel's point
for emphasis, one might argue that conflict is the rela
tionship and not something that destroys the relationship.
Coser (1956) has clarified and built upon
Simmel's (1955) themes. Coser points to many of the
integrative functions of social conflict. For example,
conflict may serve to clarify group boundaries and main
tain group identity by forcing recognition of one1s own
loyalties and by preventing group members from crossing
over to other groups. The necessity to cope with conflict
forces a pattern of social organization onto the group
(e.g., systems of authority and stratification). Further
more, Coser (1957) maintains that conflict is basically
progressive for the group. It prevents ossification of
existing group structure because the requirements of
coping with conflict are vital, thereby overriding any
inbred resistance to change. The process also generates
new norms, institutions and technology for the same
reason. In this context, it is interesting to note that
21
implicit in Marx's (1964:58, 60, 68) analysis is the
similar notion that conflicts integrates the proletariat
against the bourgeoisie in terms of class conflict.
North et al. (1960:359) point out that "some
measure of integration exists in any overt conflict situa
tion by virtue of the very fact that overt conflict implies
some contact between parties." Furthermore, they emphasize
that conflict may remain latent (i.e., neither party
attempts to change the existing conflict situation) or
manifest (i.e., either or both parties openly engage in
activities which will affect the conflict situation).
Dahrendorf (1958, 1959) has constructed what is
perhaps the most elegant model of conflict. Beginning
with Weber's definition of authority as the likelihood
that a command will be obeyed, he argues that subordina
tion and superordination are not necessarily related. A
gradual continuum between the two cannot always be con
structed. This in turn defines two quasi-groups with
incompatible interests. At this stage the interests are
still latent.
Given certain conditions, these quasi-groups are
then transformed into interest groups with manifest
interests. Such conditions may refer to a variety of
22
diverse elements. For example, the amount of political
freedom to organize would constitute a condition leading
to the development of interest groups, as would the
availability of the materials means of organization and
the level of technology (especially communication).
It is worth emphasizing that Dahrendorf's scheme
utilizes conflict as an integral part of the development
of groups. If subordination and superordination are true
dichotomies, then the relationship between the different
groups is inexorably locked in conflict.
Conflict and Urban Community Integration
In the urban arena, conflict has almost invariably
been seen as a disjunctive process. For example, Coleman's
(1958) model of community conflict generates the predic
tion that conflict prevents community action. Warren's
(1963) model is based on vertical and horizontal integra
tion with conflict left outside the system. Hillery's
(1968) model of cities is based upon the processes of
cooperation, contract and area, with conflict specifically
viewed as disruptive to community integration. In speaking
of Coleman (1958) he further goes on to say that "conflict
is a divisive force, and . . . only the presence of certain
23
other forces keeps the community from annihilation"
(Hillery, 1968:68).
Utilizing the disjunctive view of conflict, the
literature contains many examples of how it retarded or
prevented the initiation of community programs. Marris
and Rein (1967) have shown the negative impact of inter
agency conflict on the Ford Foundation "Grey Area
Projects" in New York. In the same city, Miller (1958)
has demonstrated the negative effects of interorganiza-
tional conflict on mobilizing juvenile delinquency
prevention programs. Crain et al. (1968) have studied
the fluoridation issue in hundreds of communities and
documented the conflict that often arises over that issue
and the negative results it has for the initiation of
such programs. Finally, Greer (1965) has shown the
negative results that conflict between state and commu
nity officials has for the implementation of urban renewal
programs in the South.
Conclusions
It is interesting to note that theorists (other
than Simmel) have either viewed conflict as disjunctive
or integrative in a special sense. The integrative
24
aspects have been approached from the viewpoint of
Sumner's (1906) distinction between the ingroup and out
group. Having made this distinction, they then proceed
to show that intergroup conflict can produce intragroup
integration.
Noteworthy by its absence is the notion that
intragroup integration can produce intragroup conflict.
In other words, there has been a neglect of the notion
that intragroup integration, in and of itself, can serve
as a source for intragroup conflict. Perhaps the reason
for the neglect of this topic is the hidden assumption
that conflict and integration are on opposite poles of
a continuum. From this assumption, it would follow that
conflict and integration are inversely related to each
other. Indeed, it could hardly be otherwise by the very
nature of how the continuum is constructed. It is one of
the theses of this paper that such an approach obscures
a potentially fruitful area of inquiry.
First, insofar as the inverse relation is a func
tion of apriori definitions, then the question pivots
around syntac and is not susceptible to any sort of
empirical argument. Acceptance of the definitions is
tantamount to acceptance of the inverse relation.
25
Second, it is not altogether clear, at least
conceptually, that conflict and integration are at sepa
rate ends of a continuum. Having previously defined
integration as the extent to which the parts of a system
are compatible with each other, it is apparent that the
opposite pole is not conflict, but anarchy or the absence
of compatibility. This is a far different state than the
definition of conflict postulated.
Taking the same problem from the side of conflict,
it is also apparent that the opposite condition is not1
integration but an absence of a particular kind of rivalry.
In either case, it is clear that the assumption of a
conflict-integration continuum requires more than just the
definition of terms. The approach of Simmel (1955) and
Dahrendorf (1958, 1959), both of whom have used conflict
as a defining element of a social relation, is congruent
with the argument made here. Furthermore, as Horowitz
(1962:180) has stated, "(to) see conflict as necessarily
destructive of the social organism, is to place a definite
premium on social equilibrium."
Third, it is felt that the seeming polarity
between conflict and integration should be approached
26
empirically. Examples of such empirical questions would
be as follows: Under what conditions does conflict and/or
integration characterize a social system? Can conflict
and integration co-reside in the same social system? If
so, at what levels of force are the two vis-a-vis each
other? Can integration be used to predict conflict and
vice versa? Such an approach would thus allow data to
determine whether or not there is indeed a polarity
between the two concepts rather than basing judgment on
apriori assumptions.
A point of view. From the preceding examination
of past theory, this dissertation argues that conflict
and integration have not been shown to be inversely
related. Furthermore, it is argued that conceiving of
the two concepts as separate dimensions without the
assumption of an inverse relationship, would allow a
fruitful expansion of current sociological theorizing.
In other words, conflict (or integration) can be approached
as an independent variable with the other as a dependent
variable. In the next chapter, a model will be proposed
which attempts to utilize this viewpoint.
Urban Race Riots
While the study of riots in general, and urban
race riots in particular, has a long history, the attempts
to systematically explore the topic from a macro approach
have been sparse. Furthermore, attempts to link theories
of urban community integration and conflict with the
occurrence of race riots have been almost nonexistent.
The reason for this may lie in the fact that
riots have been traditionally studied under the heading
of collective behavior. As such, explanatory factors
have been usually looked for from a social psychological
perspective. Such an approach, as mentioned previously,
is in itself unsuitable for use in macro level, compara
tive theory without modification.
Accounts of Urban Race Riots
Grimshaw (1960) attempted to delineate the eco
logical patterns related to race riots by examining the
2
For references to riot behavior m general, see
Turner and Killian (1957); Smelser (1962); and Grimshaw
(1960). Even these accounts are largely descriptive,
unsystematic and ad hoc in nature.
28
various types of neighborhoods that seem to be focal
points of racial violence (e.g., white, higher class;
contested areas; Negro slums; and stable, mixed neighbor
hoods) . Accordingly, as these types of neighborhoods are
contested by Negroes, racial violence is likely to occur.
While Grimshaw's account is instructive, it suffers from
the lack of empirical evidence to support the conclusions.
Perhaps the first attempt to systematically take
urban social structure into account on a empirical, com
parative basis is that of Lieberson and Silverman (1965).
Studying race riots from 1913 to 1963, they found that
the low proportion of Negroes in occupations traditionally
assigned to Negroes, low proportion of Negro store owner
ship, low Negro unemployment, low white income, high
proportion of Negro policemen, and high population per
councilman, all tended to distinguish matched control
cities from experimental cities (i.e., cities that had
experienced a riot).
Following the lead of Lieberson and Silverman
(1965), Bloombaum (1968) applied multidimensional scalo-
gram analysis to the same set of data. It was found that
simultaneously taking into account the several variables
mentioned above serves to distinguish control from
29
experimental cities much better than any one item alone.
This result was not surprising inasmuch as the original
findings were generally weak but in the predicted direc
tion. Perhaps more interesting is the implication that
these variables form a single dimension; although pre
cisely what that dimension would be is difficult to say.
There are two studies from the social psychologi
cal viewpoint which have implications for urban social
structure. In a survey of two matched neighborhoods in
Detroit, Street and Leggett (1961) ascertained the pre
disposition of Negro residents towards the use of violence.
One neighborhood was "considerably more active; it had a
neighborhood club (which) . . . was said to have attracted
an audience of 300 to 400" at general program meetings
(Street and Leggett, 1961:54). This neighborhood was
found to be more favorable towards the use of violence
than the control neighborhood. The integrating function
of the club and the relation of this integration to
violence is one possible interpretation of this study.
Where integration was high, the predisposition towards
30
3
violence was also high.
Ransford (1968), in a survey of Watts, Los
Angeles, related the subjective feelings of powerlessness
and dissatisfaction on the part of Negroes to their iso
lation from white society. Isolation was viewed as a
structural property. He found that Negroes who felt
powerless and dissatisfied were highly prone to favor
violence as a means to an end. This result was even more
pronounced under the condition of isolation.
Interpretation and Conclusion
While it is difficult to come to any firm conclu
sions on the basis of such scanty theory and research, it
does seem that a systematic theoretical structure relating
urban social structure to race riots is lacking. Lieberson
and Silverman (1965), and Bloombaum (1968) to a lesser
extent, have simply provided a series of unrelated ad hoc
hypotheses.
3
It should be pointed out that the authors inter
preted their findings in terms of economic deprivation
rather than in the terms presented here.
31
For present purposes, the interesting point of
the Ransford (1968) study is the attempt to link social
psychological variables (powerlessness and dissatisfac
tion) to the structural characteristics of the community
(isolation). The study suggests that structural charac
teristics are important variables and worthy of study in
their own right. Since integration is here viewed as a
structural characteristic, the Street and Leggett (1961)
study gives support to the use of organizations as a basic
unit of analysis and to notion that integration and con
flict are positively related.
CHAPTER III
A CONCEPTUALIZATION OP URBAN INTEGRATION
AND RACE RIOTS
The review of the literature showed that in any
conceptualization of urban social structure and race
riots, several factors must be taken into account. These
are (1) the dimensions of extralocal and local integra
tion, (2) the relationship between the various modes of
integration, and (3) the specification of the basic units
of analysis. In this chapter a series of concepts,
rationales, assumptions and hypotheses are proposed.
Taken collectively, these are referred to as the "Urban
Interdependence Model for Race Riots."
Urban Interdependence Model for Race Riots
The term "interdependence" is used as a generic
one to subsume both conflict and integration under a
single heading (Litwak and Hylton, 1962; Aiken and Hage,
1968). By "model" is meant a group of nominally defined
concepts which provide relationships and terms (Wilier,
32
33
1967) from which hypotheses are derived. By "rationales"
is meant the basic, underlying point of view or orienta
tion of the model (Wilier, 1967).
Concepts. The conceptualization presented here
uses five concepts as sources of relationships and terms.
They are as follows:
1. Community. There seems to be a fair amount
of consensus that the term denotes a geographic area
inhabited by persons in social interaction with at least
one common bond (cf., Parsons, 1951; Hillery, 1955;
Kaufman, 1959; Sjoberg, 1964). This concept of community
opens the way for the application of traditional socio
logical variables. Implicit in the definition is the
notion that communities are groups in the sense that
groups are defined by the interaction of their members
(Homans, 1950:82-86). This does not mean that all group
concepts can be indiscriminately applied to the analysis
of communities. Certain modifications are in order in
regards to such factors as formality of structure, impli
cations of sheer size, primacy of goal orientation
(Shepherd, 1964), and of course, the influence of the
34
areal restriction. Nonetheless, with suitable modifica
tion, group notions might be profitably applied to the
study of communities while at the same time, such appli
cation does not preclude the use of concepts which have
already been applied to community and society.
2. Structure. The research questions specify
structural characteristics. By structure is meant the
relatively unchanging aspects of social behavior (Homans,
1950:280-82), or, to phrase it in a slightly different
manner, it refers to the constants of social behavior.
A rigid application of such a definition would lead to
only descriptive studies; however, it is generally agreed
that to study structure implies something further— the
examination of the "interrelations of or arrangements of
'parts' in some total entity or 'whole'" (Nadel, 1964:4).
The research questions, then, deal with the interrelation
ships of constants and the implications of their relation-
for race riots.
3. Conflict. For purposes of this dissertation,
conflict may be defined as a "struggle over values and
claims to scarce status, power and resources in which the
aim of the opponents is to neutralize, injure or eliminate
35
their rivals" (Coser, 1956:8).1 As distinguished from
competition which has as its aim the attainment of a
goal, conflict has as its aim the neutralization, injury
or elimination of rivals (Williams, 1947:43). Obviously,
it is quite easy to move from competition to conflict,
especially on a psychological level, by a simple shift of
goals. The two concepts are independent, however, and
should be treated as such.
4. Integration. Integration may be defined as
the extent to which the parts of a system are compatible
with each other (Parsons, 1951:36n). Another term,
consensus, is sometimes used in this area. It generally
refers to the extent to which there is common agreement
or common opinion amongst the actors of a social system
(Horowitz, 1962). It may be that integration and con
sensus, are empirically interwoven.
This definition of conflict should be kept sepa
rate from conflict viewed in Lewinian terms. In Lewin's
(1951) view, conflict results from the placement of an
actor in relation to goals which cannot be simultaneously
maximized.
In addition, this definition of conflict is differ
ent from the game theory approach which views it as a
result of two or more parties seeking goals which both
parties cannot simultaneously attain (Schelling, 1963).
36
5. Riot. A riot may be defined as an "assault
upon persons and property because they are part of a
given subgroup of the community" (Lieberson and Silverman,
1365:887) . A race riot would be an assault in which the
subgroup is demarcated on the basis of racial character
istics .
Rationales. From consideration of the preceding
concepts, two basic rationales emerge. They are as
follows:
1. Race riots are a form of conflict. A race
riot is an assault (implying that the intent of the actors
is to do harm) upon persons or property delineated by
racial characteristics or upon their symbolic surrogates.
In the normal sense of the concept, riots refer to a
group of persons and not to an individual. It would make
little sense to speak of a riot composed of one person.
2. Conflict implies the existence of integration.
This is the rationale derived from the basic concepts and
as implied by such authors as Simmel (1955) and Coser
(1956, 1957).
As applied to urban community integration, it is
37
argued that if a community were totally fragmented— i.e.,
an aggregate of unrelated units— one could not speak of
the community as being either integrated or in conflict.
However, to the extent that integration is present, to
that extent the possibility of conflict is also present.
Each integrative relation is thereby potentially capable
of being transformed into one of conflict by its very
existence. In other words, the existence of integrative
relations defines, in and of itself, a potential for
conflict. Without integration there can be no interaction.
Without interaction there can be no relationships which
can become conflictual in nature.
Furthermore, as the number of integrative rela
tions increases (i.e., as the system becomes more complex),
the number of relations tends to increase geometrically.
Therefore, in complex integrated systems, there is an
increased probability that some of the relationships will
become conflictual. This is simply the application of
probability to a set of relations.
Assumptions. Before proceeding to the specific
assumptions, urban integration in toto may be defined as
the degree to which communitywide organizations within an
38
urban setting articulate, or are interlocked with each
other in an active way. This is explicitly a systems
approach inasmuch as each organization is conceived of
as a part within some totality of parts (cf. Buckley,
1967). In other words, each part is contributing to the
overall articulation of the total system.
Turning now to the specific assumptions of the
urban interdependence model, it may be seen that they
represent the specification of integration and conflict
for urban phenomenon. They are as follows:
1. Urban integration is composed of two compo
nents: extralocal and local integration (Warren, 1963;
Turk and Jiobu, 1969). First, extralocal integration
refers to the degree to which the community articulates
with, or is interwoven with the broader society. An
extreme case would be the existence of actors under com
plete dominance of the broader society and without any
relations with the local community.
The second component, the local, refers to the
intra-community articulation and compatibility of system
parts— i.e., of the elements within urban community—
independently of any extralocal integration. Such
39
integration is defined as high if local actors have many
relationships with actors within the community, whether
or not these same actors also have many relationships
with actors outside the community. Organizations are
viewed as the primary determinants of both kinds of
integration. Integration of the populace is seen as
2
being mediated by these organizations. As an example of
extralocal integration via organizations, consider that a
military base located in a community is controlled by the
military establishment, an organization of the broader
society. The base is almost totally oriented towards
integration into the broader military network. In con
trast, a chain of locally owned supermarkets may have
very little orientation towards the goals and norms of
the larger society but must be very responsive to those
of the local community.
There is a third possibility, that of the mixed
category of integration deriving from the simultaneous
integration along both the extralocal and local dimensions.
2
See Chapter II for further discussion and docu
mentation.
40
this undoubtedly occurs empirically. However, it is
possible to speak of communities as being more or less
integrated along either dimension. The notion that com
munities must be all one or all the other is not realis
tic. ^ For example, a military base, whose ultimate
orientation and integration is extralocal, may attempt
to interact with local organizations in order to purchase
certain supply items, thus effecting a measure of local
integration. By the same token, the supermarket chain
may have to purchase items from organizations outside of
the local community, thus effecting a measure of extra
local integration.
2. Concerning the relationship between extralocal
and local integration, it has been argued that the two
are inversely related to each other (Turk and Jiobu, 1969).
Gains alone one dimension are at the expense of the other.
This represents an application of ingroup-outgroup theory
to the special case of community. As mentioned in the
discussion of assumption 1, there is no definitional
3
Suggested by Hall's (1963) discussion of Weber's
ideal type bureaucracy.
41
reason why an extralocally integrated actor has to con
tribute to the local integration of the community, nor
for the converse. More important to the argument is the
idea that the inverse relation exists because integration
in the one dimension disrupts integration in the other.
As extralocally oriented actors penetrate a community,
they draw off resources (either monetary or social) from
the community, thereby disrupting local integration. The
very act of penetration is also disruptive because it may
tend to break up extant local relations as well as intro
duce elements with which the community is not prepared
to cope.
Hypotheses. The preceding considerations allow
the derivation of two fundamental hypotheses:
1. Communities characterized by high local inte
gration will experience (a) a greater number of race riots,
and (b) more severe ones than communities characterized by
low local integration.
2. Communities characterized by high extralocal
integration will experience (a) a greater number of race
riots, and (b) more severe ones than communities charac-
terized by low extralocal integration.
These hypotheses are derived from the two ration
ales and the various assumptions and concepts. Considering
the rationale that race riots are a form of conflict in
conjunction with the rationale that conflict implies
integration, it follows that race riots imply integration.
This is the fundamental assertion underlying both hypoth
eses. However, the general concept of integration must be
specified further in order to accommodate the urban con
text in which race riots occur. In other words, urban
integration is the relevant type vis-a-vis race riots.
Urban integration, it was argued, is composed of two
components, extralocal and local (assumption 1). There
fore, both types of integration have been taken into
account in formulating the above hypotheses.
Discussion of the hypotheses. There are several
points concerning these hypotheses which require addi
tional comment. First, it should be noted that the hypoth
eses do not distinguish between the relative strength of
local and extralocal integration. Since there are no
firm guidelines to specify whether extralocal or local
integration is the more potent causative factor in regards
43
to race riots, the strategy adopted here is to consider
them equally potent and allow empirical outcomes to
determine which is the more potent.
Second, one might argue that local integration is
the more relevant type because race riots occur within the
context of a community and directly affect it. However,
it should be kept in mind that the current race riots
are thought to be heavily influenced by the tenor and
tempo of the current civil rights movement. Recent riots
probably have an altogether different quality than earlier
ones. For example, the Chicago race riot of 1919, and
the Detroit race riots of 1919, and 1943 have been attrib
uted to a variety of causes, notably economic competition
between Negroes and whites. However, the issue of civil
rights and racial integration have not been offered as an
underlying causative factor (cf. Turner and Killian, 1957:
104-106; 110-112). In contrast, heavy emphasis is given to
these issues insofar as more recent (1967) riots are con
cerned (cf. National Advisory Commission on Civil Dis
orders, 196 8).
Furthermore, civil rights organizations such as
the National Association for the Advancement of Colored
44
Peoples, the Urban League, and the Congress of Racial
Equality, are nationwide in scope. The organizational
structure of these groups, while having local outlets,
transcends any one community. The civil rights movement,
then, encompasses both types of integration implying that
race riots are also affected by both types.
In terms of local integration, examination of
preliminary empirical results suggested that the relevant
type is that between Negroes and whites, rather than that
4
of the overall community qua community. Although initial
focus was placed on overall integration, for the case of
race riots which are, by definition, closely tied to
racial situations, the appropriate type was seen to be
racial while the overall integration appeared inappro
priate.
In this light, racial integration reflects the
articulation of two system subparts (Negro and white).
The obverse of integration is racial polarity— i.e., the
extent to which the two subparts are unrelated and isolated
Overall integration in the sense of viewing the
community as totality in a summary fashion rather than
examining the specific subdimensions which comprise the
totality.
45
from each other. There is no necessary reason why inte
gration along racial lines should be related to integra
tion along other dimensions. For example, one may conceive
of a society polarized along class, age, or religious
dimensions in addition to race. If it were to turn out
that all members of a given race were also of the same
religion as all members of another race, then one could
expect a relation between religious and racial integration.
However, the relationships may be due to historical factors
and could be just as well that both races are of a differ
ent religion. In other words, there is nothing inherent
in the concept of integration along diverse dimensions
that requires them to be related.
Finally, it should be pointed out integration as
conceived above is different from integration deriving
from a caste system. For example, the social organization
of plantations during slavery may be thought of as highly
integrated in that the plantation constituted a single
social unit but with a clear, unpenetrable boundary
between master and slave. Nonetheless, in its totality,
the plantation might be conceived of as highly integrated.
Third, it should also be noted that in any empir
ical situation, assessment of the effects of local
46
integration requires that the effects of extralocal
integration be held constant (or vice versa). This
necessity follows from the inverse relation between the
two types of integration (assumption 2). Were they not
held constant, the effects of one could negate the effects
of the other.
While the hypotheses specify that integration must
precede the outburst of race riots, it is not possible to
specify the exact time interval required, again due to a
lack of theory and empirical evidence. Nonetheless, one
would expect that a certain (unspecifiable) time interval
is required between a given level of integration and the
outburst of race riots. This expectation rests upon the
sheer amount of time required for actors to develop issues
of conflict out of a base of integration.
Contrasting Model
As this study progressed, it was realized that the
basic statement of the urban interdependence model (i.e.,
the greater the integration, the greater the conflict)
could be compared to three basic statements of the Marxian
model— (a) the greater the interstratum polarity, the
greater the interstratum conflict; (b) the greater the
47
intrastratum solidarity, the greater the interstratum
conflict; and (c) the greater the exploitation of one
stratum by another, the greater the likelihood of inter-
5
stratum conflict.
Apart from the economic aspects of Marxian theory,
these three major ideas have been explicit or implicit in
the writings of such conflict theorists as Dahrendorf
(1958, 1959), Simmel (1955), and Coser (1956), all reviewed
in Chapter II. Taking the form of the argument rather
than its specific content and substituting the idea of
racial polarization for interstratum polarity, the idea
of racial solidarity for class solidarity, and the idea
of Negro deprivation for interstratum exploitation,
generates the basic idea of many race relations theorists
who find the cause of racial violence in Negro deprivation
(exploitation) and ghettoization (interstratum distance
and solidarity) (cf. Danlke, 1969; Rainwater, 1969; Buggs,
5
Mills (1962:85-87), in inventorying Marx's ideas
has phrased these as "the class structure becomes more and
more polarized, thus increasing the chance for revolution";
the wage worker— class-in-itself— will be transformed into
the proletariat, a class-for-itself"; and "exploitation is
built into capitalism as an economic system, thus increas
ing the chances for revolution."
48
1969). In other words, the quasi-Marxian statements
become as follows: (a) the greater the intraracial
solidarity, the greater the interracial conflict; (b) the
greater the segregation of the Negro, the greater the
interracial conflict; and (c) the greater the deprivation
of the Negro, the greater the interracial conflict.
When we consider two segments of a single community,
Negro and white, the urban interdependence model would
generate predictions opposite those of the quasi-Marxian
model. Under the former approach, where the polarity
between Negroes and whites is high, it would be expected
that the incidence of racial conflict would be low and
vice versa. The logic of this expectation is that Negro-
white polarity indicates the relative absence of racial
integration, and such absence would mean little conflict.
This is simply a restatement of the argument that integra
tion is positively associated with conflict, but phrased
in terms of low integration and low conflict rather than
high integration and high conflict.
Thus, the exciting opportunity offered itself to
compare a quasi-Marxian model predicting that the exclusion
of Negroes would lead to racial conflict with the present
model which predicts exactly the opposite.
Conclusion
It is felt that the urban interdependence model
for race riots provides a means of examining traditional
sociological variables in a new light. Past theory and
research have not examined integration and conflict in
precisely this way. Furthermore, the opportunity to
compare this model with other quasi-Marxian approaches
manifested itself, adding new importance to the under
taking.
In modern society, the urban setting obviously
constitutes an interesting and important locus for the
testing of such models, and race riots provide a socio
logically important dependent variable.
CHAPTER IV
METHOD AND PROCEDURES
In this chapter, the hypotheses derived in the
preceding section are given their operational definitions.
The discussion will concern the measures selected and the
universe of inquiry.
Universe of Inquiry
The 114 United States cities with 1960 populations
of 100,000 or more inhabitants, and with a Negro population
of 2 per cent or more,-*- comprise the group of cities inves
tigated. A city is defined as a subdivision of a State
within an area over which a municipal corporation has been
established to provide government for a specific population
Because this study deals with race riots, there
must obviously be some minimum number of Negroes present
for a riot to occur. The criterion of 2 per cent was
selected because no city in which a riot occurred in the
first nine months of 1967 had a population less than 2.4
per cent Negro. The 2 per cent criterion thus included
all cities experiencing a riot.
50
51
concentration (United States Bureau of Census, 1967).
The term city thus refers to the political boundaries of
a municipal government.
Adopting the city as the areal unit is reasonable
for two reasons. First, historically, urban race riots
have largely taken place in Negro ghettos and ghettos are
"almost everywhere (in) the oldest part of the central
city" (Banefield and Wilson, 1966:174). This suggests
that city parameters are the most directly relevant.
Use of other units may obscure these more direct charac
teristics .
Second, because riots take place within cities,
responses on any governmental level involve the municipal
government of the city. Although state and federal forces
may be involved, the city government is the one which is
always called upon to make some sort of response.
Although their selection is arbitrary to some
extent, it is felt that cities of 100,000 or more inhabit
ants essentially encompass all of the "large" cities of
the country. When size decreases below a certain point,
the city may begin to differ qualitatively in regards to
such factors as cosmopolitanism and economic orientation
(cf. Keyes, 1958). Furthermore, it was felt that cities
52
of this size range would generally be more self-sufficient
in terms of having such things as their own school system,
local television station, and own water and power supply.
The specific measures to be discussed are summa
rized in Exhibit 1 along with the details of their con
struction and their sources.
EXHIBIT 1
DESCRIPTION AND NAMES OF VARIABLES
Short Name Description of Variable
Measures of Conflict
Riot Commission Index Four point, ordinal scale
based on descriptive cate
gories of riot behavior.a
Journalistic Severity
Index
Four point, ordinal scale of
riot severity from journal
istic accounts; based on
the reported number of
arrests and casualties.*3
Injury Index Ten point, ordinal scale of
riot severity based on the
number of riot related
injuries.a
Wanderer Severity
Index
Eight point, Guttman scale
of riot severity based on a
survey of city mayors;
reported by Wanderer.0
53
EXHIBIT 1 (Continued)
Short Name Description of Variable
Riot Proneness Number of riots per city,
January 1, 1969 to September
1, 1969.a
Measures of Integration
and Negro-White Polarity
Extralocal Integration
Ghettoization
Occupational Segregation
Relative Unemployment
Relative Fertility
Ten point, ordinal scale
based on the number of head
quarters of national volun
tary associations located
with the city.d
Ten point, ordinal scale
based on the Taeuber and
Taeuber index values of
residential segregation.®
Per cent of Negroes in the
U.S. Census categories of
private household workers,
laborers (excluding farm and
mine), and farm laborers
(excluding unpaid family
help and foremen).^
Per cent of Negroes unemployed
minus the per cent of whites
unemployed.9
Number of Negro children
under 5 years divided by the
number of Negro women ages
15 through 49; minus the
number of white children
under 5 years divided by the
number of white women ages
15 through 49.^
54
EXHIBIT 1 (Continued)
Short Name Description of Variable
Nonwhite Poverty
City Density
Per cent of nonwhite families
earning less than $3,000 in
1959.h
Persons per square mile
h
Control Measures
Nonwhite Population
Change
Nonwhite population in 1950
minus the nonwhite popula
tion in I960.*1
Total Population
Change
Total population in 1950
minus the total population
in 1960.h
Migration Per cent of persons 5 years
and over from a different
county, 1955-1960. *1
Police Per capita number of police
officers,_uniformed and
civilian.1
Northeast City location in Connecticut,
Massachusetts, New Jersey,
New York, Pennsylvania, or
Rhode Island; dichotomously
coded.
East North Central City location in Illinois,
Indiana, Michigan, Ohio, or
Wisconsin; dichotomously
coded.
55
EXHIBIT 1 (Continued)
Short Name Description of Variable
Southeast City location in Alabama,
District of Columbia,
Florida, Georgia, Kentucky,
Mary1and, Mi s s i s sippi, North
Carolina, Tennessee, or
Virginia; dichotomously
coded.
Southwest City location in Arkansas,
Lousiana, Oklahoma, or
Texas; dichotomously coded.
West City location in Arizona,
California, Colorado, Hawaii,
Iowa, Kansas, Minnesota,
Missouri, Nebraska, New
Mexico, Oregon, Utah, or
Washington; dichotomously
coded.
Sources:
aNational Advisory Commission on Civil Disorders
(1968).
Christian Science Monitor, New York Times, U.S.
News and World Report; all from January 1, 1967
to August 1, 1967.
cWanderer (1968).
^Encyclopedia of Associations (1961).
eTaeuber and Taeuber (1965:32-34).
•^U.S. Bureau of Census (1963) .
56
EXHIBIT 1 (Continued)
Sources (continued):
^Derived from data in the U.S. Bureau of Census
(1962) and the U.S. Bureau of Census (1963).
^U.S. Bureau of Census (1962).
1U.S. Department of Justice (1961).
Measures of Conflict
Primary consideration was given to those race
riots occurring in the first nine months of 1967. This
time period was chosen for two reasons. First, data per
taining to that period were available from two independent
sources, thereby making it possible to assess reliability.
Second, as previously discussed, the nationwide civil
rights movement contributes to the extralocal integration
of race riots. The Los Angeles Watts riot (1965) is con
sidered to be one of the first major riots given a civil
rights interpretation (Buggs, 1969). Since 1967 riots
occur some two years later, the time order allows for the
57
operation of civil rights influences, an allowance which
o
was probably lacking for riots occurring prior to 1964.
The hypotheses call for the measurement of both
severity and number of race riots. The following four
3
measures of severity were available:
1. The Report of the National Advisory Commission
on Civil Disorders (1968:158-1^9) provided severity cate
gories based on such factors as number of fires, amount
of looting, and number and size of crowds. With these
categories, it was possible to construct a four point,
4
ordinal scale ranging from "no riot" to "severe riot."
Mean score was 1.6 for 65 cities experiencing at least
2
Previously discussed m Chapters II and III.
O . . . .
With all severity measures, a city experiencing
multiple riots was given the score of the most severe riot.
^If no riot occurred, the city was scored "0."
A minor riot, scored "1," was described as few
fires, few broken windows, generally lasting less than one
day, involving small numbers of people, use of local
police.
A moderate riot, scored "2j" was described as
isolated looting, some rock throwing and fires, lasting
one to two days, one sizeable crowd or many small ones,
use of State police but usually not the National Guard.
A severe riot, scored "3," was described as many
fires, intensive looting, reports of sniping, lasting more
than two days, sizeable crowds, use of National Guard and/
or other Federal forces,.
58
one riot. This measure is called the "riot commission
index."
2. A second severity measure was derived from
descriptions of riots found in the news media. A four
point, ordinal scale was constructed using the number of
arrests and casualties as criteria.Mean score was 1.3
based on 32 cities for which data were available.
3. Additionally, it was possible to construct a
severity index based on riot related injuries as reported
by the National Advisory Commission on Civil Disorders
(1968:162-163). A ten point, ordinal scale was devised
ranging from "no riot" through "riot with no injuries,"
g
to "riot with many injuries." Mean score was 3.0 based
5
If no riot occurred, the city was scored "0."
A minor riot, scored "1," was described as having
less than 50 casualties and less than 100 arrests.
A moderate riot, scored "2," was described as
having 50 or more casualties and 100 or more arrests.
Only Detroit and Newark were considered severe and
scored "3," based on the overall assessment of the news
media reports.
£
A ten point scale was used rather than raw figures
because the distribution was heavily skewed. Detroit and
Newark accounted for 60 per cent of the 1,736 injuries
relevant to this study. Mean number of injuries was 26.7
per riot. There was no way of ascertaining whether in
juries were accidental or the result of deliberate, riot
59
on 65 cities experiencing at least one riot.
4. Finally, Wanderer (1969) constructed a Guttman
scale of riot severity using survey data from city mayors.
Of the 75 cities for which he reported scores, 44 were in
the current study. The scale ranged from "1," not severe,
to "8," very severe. Mean score was 4.1 for 44 cities.
It should be noted that the Guttman criterion was
originally applied to a set of 75 cities. There is no
necessary reason for the scale items to satisfy the Guttman
criterion in the subset of 44 cities used in this study.
Moreover, the scores may not be entirely independent of
the data used by the National Advisory Commission.
Wanderer's data were gathered for a Senate subcommittee
studying urban racial violence (Wanderer, 1969; 500n) and
bear a strong resemblance to the data used by the Advisory
related actions.
7
Items, in order of severity were as follows: no
scale item, vandalism; interference with firemen; looting,
sniping, called State police; called National Guard, law
officer or civilian killed. Scores were reversed to make
a high number indicate a severe riot.
60
Commission. In regards to both points, the present
assessment is an empirical one.
Turning now to the second major dimension of
conflict, that of riot proneness, this aspect of conflict
was measured by the number of separate riots experienced
by a city in the first nine months of 1967. There were
47 cities which had one riot. Of the 18 cities experienc
ing multiple riots, the number ranged from two to five.
Mean number for the 114 cities under consideration was .82.
In order to assess reliability, the correlations
between the four severity measures, along with the riot
proneness measure, are presented in Table 1. Considering
patterns of correlations, Table 1 shows that the injury
based index has higher correlations with the other severity
measures than does any other severity measure. For exam
ple, the injury based index correlates with the other
severity measures .59, .53, and .48; while the Wanderer
index correlates with the other severity measures .39, .32,
and .53. Between these two indexes, it can be seen that
the injury based index has the greater number of higher
correlations. If this pattern of analysis is continued
for all four severity measures shown on Table 1, it can be
observed that after the injury based index, the commission
61
TABLE 1
INTERCORRELATIONS AMONG CITY RIOT PRONENESS
AND MEASURES OF RIOT SEVERITY3
1 2 3 4 5
1. Commission
Severity Index
.56
(32)b
.59
(65) c
.39
(44) d
.09
(65)
2. Journalistic
Index of Riot
Severity
.48
(32)
.32
(26) e
-.21
(32)
3. Injury Index of
Riot Severity
.53
(44)
.21
(32)
4. Wanderer Index of
Riot Severity
.04
(44)
5. Number of Riots
per City
aValues in parentheses indicate the number of
cases upon which the correlation is based. Correlations
are based on those cities for which scores on both pairs
of variables were available.
T_
With an N of 32, correlations of .34 and .44 are
significant at the .05 and .01 levels respectively.
cWith an N of 65, correlations of .24 and .31 are
significant at the .05 and .01 levels respectively.
c ^With an N of 44, correlations of .29 and .38 are
significant at the .05 and .01 levels respectively.
eWith an N of 26, correlations of .38 and .48 are
significant at the .05 and .01 levels respectively.
62
severity index is the second most correlated measure,
higher than either the Wanderer index or the journalistic
index.
The fact that one measure of severity is based on
descriptive categories of riot behavior while the other
is based on injuries, lead to adopting both. The former
measure refers to descriptions of the riot itself while
the latter refers to riot consequences. One would expect
the two measures to be substantially related, as they are
(r = .59). However, it is possible to have a severe riot
resulting in no injuries or many injuries resulting from
a moderate riot. For this reason, it seemed prudent to
use both measures of severity.
The journalistic index, while showing lower
correlations than those measures based on Advisory Com
mission data, was still substantially correlated with
other measures of severity. This suggests that journal
istic accounts, although perhaps exaggerated and inaccu
rate, can provide the basis for deriving suitable measures
when alternative sources are not available.
Notable by its relatively low correlation with
other severity measures is the Wanderer index. This may
be the result of using Guttman scores originally derived
63
for a set of 75 cities on a subset of 44 cities. For the
44 cities, the data may not have satisfied the Guttman
criteria. On the other hand, the source of Wanderer's
data was a mail questionnaire asking city mayors for their
description of riot activities. One would expect that
mayors, having a vested interest in their cities, would
tend to describe these types of events in as favorable a
light as possible, thereby biasing the index.
The severity index and injury indexes, both based
on Advisory Commission data, were adopted. While the
journalistic measure was found to be potentially useful,
in the present case the Advisory Commission data had the
advantages of greater reliability, as well as being more
systematic and extensive in the number of cities con
sidered. Moreover, the Advisory Commission data are
presumably the more accurate since it is based on commit
tee testimony and repeated investigation (National Advisory
Commission on Civil Disorders, 1968:574-581).
Turning to the number of riots per city— i.e.,
riot proness— Table 1 shows that there is virtually no
relation between proneness and severity (regardless of
how severity is measured). How many riots occur in a city
64
o
neither increases nor decreases with riot severity.
It is difficult to interpret this result. Prone
ness may be a separate dimension of riot conflict. On
the other hand, it is also possible that cities experienc
ing multiple riots are those which, through time, have
become most capable of controlling riot severity. Sever
ity may thus be a function of a city's response to rioting
while proneness may be a function of a city's social
organization. In later pages, this relation will be re
examined taking all 114 cities into account.
Measures of Integration
Extralocal integration. Turk and Jiobu (1969)
employed the measure of extralocal integration used here.
In that study, the level of a city's extralocal integra
tion was found to be predictive of the magnitude and
volume of Federal funding of local poverty organizations,
while the level of local integration was predictive of
g
The negative correlation between number of riots
and the journalistic index is attributable to seven cases
out of the thirty-two upon which the correlation is based.
If those seven cases are removed, the correlation falls to
zero.
i
65
the complexity of the interorganizational network through
which these funds flowed.
Several reasons exist for using the Turk measure.
First, since it successfully predicted a different depend
ent variable in a manner consonant with its nominal
definition, evidence for its validity is already at hand.
If the measure can predict conflict, then further evidence
becomes available. Second, although Turk and Jiobu (1969)
were not interested in examining conflict in great detail,
their treatment of integration is sufficiently similar to
the present conceptualization that differences in theory
do not pose a problem. Third, these authors investigated
the 130 largest United States cities. Of those cities,
114 are used in the current study. Thus, the samples of
both studies are highly comparable.
For these reasons, headquarters of national asso
ciations was adopted as the measure of extralocal integra
tion.^ A high score (i.e., many headquarters) indicated
Examples which are familiar to the reader would be
the American Sociological Association, the American Statis
tical Association, and the National Training Laboratories
Institute for Applied Behavioral Science.
66
a city within which broad societal concerns came to a
focus and were dealt with. In this sense, one would
expect Washington, D.C., as the Nation's Capital, to be
highly integrated into the broader society. Additionally,
in a slightly different sense, one would expect cities
such as New York, Chicago, and Houston to have a large
number of external ties to the broader society. All of
these cities ranked highest in the number of national
headquarters located within them. In contrast, Wichita
Falls, Texas; Columbus, Georgia; and Hammond, Indiana
ranked lowest on this measure. Furthermore, a national
voluntary association of the type that was counted repre
sents the common element of a variety of similar organiza
tions. It is therefore likely to reflect one or more
societal institutions. Taken together, such associations
are likely to encompass the full range of societal insti
tutions (Turk, 1969b).
Turk and Jiobu (1969) found this measure to be
associated with a series of variables which provided
additional evidence of its validity. These correlations
were recomputed for the 114 cities in the current study.
As expected, only trivial differences were found between
the two sets of correlations.
67
One would expect that cities characterized by
high centrality in the society would be also characterized
by high extralocal integration. Ratings of the cities'
national centrality by two sociologists— one using loca
tion in the transportation network and the other using
prestige as the criterion— were averaged. These mean
ratings were correlated at the level of .57^ with the
national headquarters index.
One would also expect the larger and older of
these cities to be the most highly integrated extralocally.
These expectations were confirmed as both city size and
age were highly correlated with the national headquarters
index (r's = .73, and .71 respectively). Banking activity
may be taken as another index of a city1s external linkages
and regional influence. Per capita bank deposits proved
to be correlated at the level of .55 with the headquarters
measure, thus confirming the expectation.
In order to further assess validity, cities were
ranked according to the number of export establishments—
■^All of the values reported in this section are
based on the product moment formula. A value of r = .19
is significant at the .05 level or beyond.
68
i.e., establishments serving nonlocal clientele and
organizations^— located within them. These ranks, in
themselves measuring a city's extralocal ties, were used
to compute an index of diversity.^ This index proved to
be highly correlated with the national headquarters
measure (r = .53). By way of further control, a similar
index of diversity was computed using the ranks of cities
according to the number of m a in t e n a n c e -^— i.e., nonexport—
establishments located within them. As expected, this
index was unrelated to the headquarters measure (r = -.15).
Finally, industrialization, a factor which could confound
these measures, was also found to be unrelated to the
headquarters index (r = .11).
11
Manufacturing; lumber, building, and hardware;
automotive dealers; apparel and accessory stores; eating
places, merchant wholesalers; wholesale trade; hotels and
motels; tourist camps; motion picture theaters; other
amusement and recreational establishments; hospitals.
Source: U.S. Bureau of Census (1967).
The formula for computing the diversification
indexes is as follows: I.D. = 1 - (£x2/(£x)2). This is
the method used by Gibbs and Martin (1962) to compute
differentiation with the labor force.
13
General merchandise stores; food stores; gasoline
service stations; drug stores; personal services; auto
repairs and services, garages. Source: U.S. Bureau of
Census (1967).
69
The use of this index in the present case is
theoretically justified for two interrelated reasons.
First, the model predicts that as extralocal integration
increases, conflict will increase. One such form of
conflict is race riots. This line of reasoning makes no
distinction between types of conflict, and rioting is
justified simply as a measurable dependent variable that
is consistent with the theory.
Second, above and beyond that, it is possible to
specify a more direct link to rioting. At least since
1954, the year of the Supreme Court decision to integrate
public schools, it may be argued that the goals of the
broader society, as proclaimed by the Federal Government,
have been towards racial integration. Subsequent Federal
legislation since 1954 has made various other forms of
discrimination illegal in such spheres of life as public
accommodations, voting rights, and restrictive housing
agreements.
Since it has been argued that the national head
quarters index ties broad societal institutions and
concerns together, and presumably national oriented
organizations, by and large, reflect the dictums of the
larger society, communities high on the national
70
headquarters index are thus simultaneously oriented
towards these dictums of racial integration. This line of
reasoning, along with the preceding argument, constitute
additional face validity for the national headquarters
measure.
In summary, the following four sources of validity
are used to justify the national headquarters index:
(1) the fact that the index successfully predicted in the
past; (2) the face validity concerning the index as linking
a community to the broader society; (3) the empirical
correlates of the index; and (4) the face validity specify
ing that national headquarters reflect the dictums of the
larger society in terms of favoring racial integration.
Local integration. The polarization between
Negroes and whites took on increased importance for the
local integration of a community as the study progressed.
The examination of preliminary results shown in the
Appendix, highlighted the polarity measures to be described
below. Their importance lies in permitting the comparison
of the present model to a quasi-Marxian model. Under the
quasi-Marxian model, discussed in Chapter III, the greater
the polarity— i.e., the less the integration— , the
71
greater the likelihood of conflict. Under the present
model, it would be expected that the less the polarity—
i.e., the greater the integration— , the greater the
likelihood of conflict. In other words, two opposite
predictions derived from two competing theoretical struc
tures may be compared.
The literature on race relations suggested the
following two measures as important to the polarity— i.e.,
nonintegration— of Negroes and whites.
1. As a measure of ghettoization, the Taeuber
and Taeuber (1965) index of residential segregation was
adopted. The index value represents the percentage of
persons that would have to move in order to obtain a
homogenous pattern of residential integration. It was
felt that where segregation was high, ghettos would pre
dominate. The deprivation associated with ghetto life as
14
While it is possible to have a low measure of
segregation resulting from a perfect checkerboard pattern
of black and white blocks, the empirical likelihood of such
a situation is small. As one would expect though, an occa
sional Southern city exhibits a low segregation index value
because Negroes live in close proximity to whites as a
result of residential patterns inherited from slavery
(Taeuber and Taeuber, 1965:49-53).
72
well as the opportunity for the development of intra-
Negro organizations and socialization (i.e., development
of solidarity amongst the exploited), has been cited as
contributing to riots by several authors (cf. Street and
Leggett, 1961; Dahlke, 1969; Hundley, 1969), and is con
gruent with the quasi-Marxian viewpoint.
2. The prevalence of occupational restrictions
imposed on Negroes— i.e., "occupational segregation"—
was quantified by the percentage of Negroes employed in
the United States Census (1963) categories of private
household workers, laborers, and farm laborers. Lieberson
and Silverman (1965) reported that as the percentage of
Negroes in these occupations decreases, the likelihood of
riot increases because whites perceive this diffusion as
an economic or other social threat.
The latter two measures represent the polarity
between Negroes and whites on the local level. The next
two refer to their cultural assimilation with respect to
life styles. Along this dimension, it is argued that
where the assimilation of the Negro is relatively high,
the polarity between Negroes and whites is relatively low,
and vice versa. Disparate life styles would indicate
73
high polarization along that dimension while similar life
styles would indicate low polarization. The following
two measures were thus included:
3. The relative deprivation which exists between
Negroes and whites has often been cited as a causative
factor in regards to race riots (Grindstaff, 1968; Hundley,
1969; Rainwater, 1969). To the extent that the relative
differences are high, it is felt that a high degree of
Negro-white polarity exists. The converse is also true.
Relative deprivation is a social psychological
concept and it is impossible to make statements about the
individual from the present data. Nonetheless, the con
cept did suggest the construction of a measure to repre
sent the relative differential that exists between white
and Negro unemployment rates. Because unemployment has
impact on many aspects of life style, it was felt that it
should represent an important aspect of polarity. In nine
cases this difference was positive showing that the white
rate was lower than the Negro rate. However, in only
three cases did the positive difference exceed one per
centage point. Thus, for all practical purposes, this
index measures the degree to which the Negro unemployment
74
rate is higher than the white rate.
4. Another measure indexing Negro-white polarity
was the difference between the Negro fertility ratio and
the white fertility ratio. It was felt that the greater
this difference, the greater would be the polarity.
Discussions of assimilation by Gordon (1964),
and Timms (1969) suggested the importance of assimilation
in terms of life styles while Bogue (1969:656-751) has
pointed out that changes in fertility imply that life
styles are changing because it is known that different
socioeconomic groups have different fertility ratios. In
all but three cases the difference between the Negro and
white ratio was positive, indicating that Negro fertility
is the higher of the two.
It will be recalled that the quasi-Marxian model
places a great deal of emphasis on violence as a conse
quence of exploitation. It has also been suggested as
a causative factor specifically related to the American
Negro (Buggs, 1969; National Advisory Commission on Civil
Disorders, 1968). For these reasons, the following meas
ure was included:
5. It was felt that Negro poverty reflects one
75
aspect of Negro exploitation. Negro poverty was measured
by the per cent of nonwhite families earning less than
$3,000 per year in 1959. This criterion was selected
because $3,000 per year is often considered the dividing
point between the poor and not poor.
Turk and Jiobu (1969), and Turk (1969b) used the
prevalence of communitywide civic associations as an over
all measure of local, urban integration. This measure
was originally included but then excluded when it became
apparent that there was no way of knowing whether civic
associations integrated the community in terms of including
or excluding Negroes and, therefore, whether these asso
ciations might not sometimes have signified white solid
arity and Negro fragmentation.
Control Variables
As safeguards against spuriousness, several vari
ables were considered. Although no specific hypotheses
were formulated for these control variables, their omnibus
nature made it prudent to take them into account. Another
group of variables for which it is possible to make tenta
tive theoretical statements were also included as further
safeguards against spuriousness at the very least and for
76
the possible theoretical clarification they might add.
Three variables fell into the latter category.
They are as follows:
1. The amount of total migration, both nonwhite
and white, from a different county in the previous (to
1960) five-year period was included because it may be
argued that as communities experience an influx of migrants,
the stability and integration of those communities is
disrupted.
2. The total increase of city population (i.e.,
nonwhite and white combined) between 1950.and 1960 was
included for a similar reason. As the population of the
community changes, it may be that it disrupts the total,
overall integration of the community.
3. A further specification of total population
increase is that of the nonwhite population between 1950
and 1960. This variable was included because it may be
that the specific subdimension of Negro population increase
is the more relevant to racial violence than that of the
total population.
77
These three variables concerning population
changes may be a reflection of overall system nonintegra
tion. On the other hand, they may simply be a reflection
of such factors as region, employment or other measures
of polarity.
A further variable with an unclear theoretical
interpretation, but potentially important, is that of the
police. The role of the police has been considered crucial
to riot precipitation (Lieberson and Silverman, 1965;
Dahlke, 1969; Hundley, 1969). This suggests the following
measure:
4. The prevalence of police officers was measured
by the per capita number of police, both uniformed and
civilian in 1960. The present approach precludes state
ments linking individual police officers or acts to other
variables; however, it is possible that the number of
police create either a meliorative or inflamatory atmos
phere vis-a-vis race riots.
An alternative interpretation of the police vari
able is that it is a measure of imperative or forced
78
integration.^ As the official representatives of
legitimated authority and permitted the use of legal
violence, the police may act as agents which force the
articulation of system subparts. Since the police variable
has at least two interpretations and some empirical evi
dence taken as congruent with the former (Lieberson and
Silverman, 1965), it was included.
A final variable was that of location within the
geographic regions of Northeast, East North Central, South
east, Southwest, and West. These were included by the
dichotomous coding of a city1s location within one of
the five regions. Different regions may be characterized
by different, broad-based subcultural and historical
patterns and were included because of their omnibus
nature.
While one might be tempted to control city age
and population size, it should be recalled that these
Originally included as a measure of imperative
or forced integration was the per capita amount of munici
pal revenues. During the course of this research, analysis
by Turk (1969b) showed this measure too omnibus in nature
to be useful. It was therefore excluded.
79
variables were theoretically and empirically related to
the national headquarters index as evidence for the
validity of that index. Statistically, since the cor
relations were of a large magnitude (greater than .40),
the effect of such control would be to diminish the effects
of the headquarters index and might be tantamount to a
"partialling fallacy" (Gordon, 1968). At the same time,
it would unnecessarily introduce the confounding effect
of correlation between predictor variables (i.e., multi-
collinearity).
CHAPTER V
RESULTS AND DISCUSSION
Since this study uses three indexes of riot
activity, it is possible to assess both the absolute values
of product moment correlation coefficients and the pattern
of the coefficients. Variables take on added meaning if
they are all related to the dependent variable in a con
sistent manner regardless of absolute size.
For purposes of exposition only, a correlation
greater than .39 is termed "high," or "large," or "strong."
One from .29 through .38 is termed "moderate"; and one
from .19 through .28 is termed "slight," or "low."
Although arbitrary, these criteria are probably within
the range of values normally found in sociological
research, and are based on the per cent of variance
explained interpretation of product moment correlations
(McNemar, 1962). In terms of probability, values from
+ .53 through + .22 fall within the 95 per cent confi
dence interval, and values from + .51 through + .16 fall
within the 99 per cent confidence interval.
80
81
Riot Pron.en.ess and Severity
The intercorrelation matrix for all of the
variables considered— both hypothesized and those included
as safeguards against spuriousness— are presented on
Table 2. The table is based upon the 114 cities studied.
It may be noted that the number of riots per city is
highly correlated with both the riot commission index of
severity and riot injury indexes. This is in contrast
to the low correlations shown on Table I."*" Since Table 2
takes into account all of the cities while Table 1 does
not, the high relationships may be attributed to the
inclusion of cities which have not experienced any riot
at all. These cities have a score of zero for riot prone
ness, and therefore, also have a score of zero for the
riot severity and number of riot related injuries.
Indeed, 56 per cent of the cities fell into this category.
The high correlation coefficients between these measures
are thus a function of the scoring process and should not
be interpreted as reflecting a relation between two
J, *
On Table 1 the correlations were based only on
those cities for which scores on each pair of variables
were avai1able.
82
TABLE 2
INTERCORRELATION OF ALL VARIABLES3
1 2 3 4 5 6 7 8 9 10 11
1. Riot Proneness 77 56 48 -21 -38 -33 -12 12 36 -22
2. Commission Index 78 47 -30 -37 -27 -17 23 37 -17
3. Injury Index 34 -25 -26 -14 -14 18 30 -08
4. National Headquarters -22 -47 -28 -09 10 43 -39
5. Ghettoization 47 48 14 -13 -41 31
6. Occupational Segregation 67 29 -24 -54 55
7. Nonwhite Poverty 25 -28 -46 49
8. Relative Unemployment -28 -26 23
9. Relative Fertility 27 -33
10. City Density -42
11. Nonwhite Population Change
12. Total Population Change
13. Migration
14. Police
15. Northeast
16. East North
17. Southeast
18. Southwest
19. West
Zero order, product moment coefficients with decimal po
based on 114 cities.
TABLE 2
INTERCORRELATION OF ALL VARIABLES3
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
77 56 48 -21 -38 -33 -12 12 36 -22 -23 -24 29 21 11 -21 -28 14
78 47 -30 -37 -27 -17 23 37 -17 -26 -25 32 20 13 -17 -35 15
34 -25 -26 -14 -14 18 30 -08 -24 -22 31 20 11 -12 -27 04
rs -22 -47 -28 -09 10 43 -39 -43 -28 47 08 01 -13 -18 20
47 48 14 -13 -41 31 44 25 -28 -46 12 46 18 -26
ation 67 29 -24 -54 55 53 43 -41 -38 -18 48 39 -27
25 -28 -46 49 28 22 -29 -25 -23 56 30 -34
nt -28 -26 23 28 51 -04 -12 -41 29 30 -03
27 -33
-42
-30
-73
-30
-57
32
59
27
53
04
15
-17
-20
-29
-45
10
-08
Change 51 49 -44 -25 -22 48 06 -07
ange 74 -67
-52
-52
-53
51
-09
-32
-07
-27
20
25
-08
-28
40
34
-35
-23
07
31
-07
-28
-26 -20 -25
-22 -27
-22
der, product moment coefficients with decimal points omitted. Correlations
ities.
independently defined variables.
Extralocal Integration
It was initially hypothesized that extralocal
integration would be positively associated with the fre
quency and severity of race riots. Further examination
of Table 2 shows that this is the case. It may be seen
that the index of national headquarters (i.e., extralocal
integration) is highly correlated with both the number of
riots per city (r = .48), and with the commission index
(r = .47). Also, the headquarters index is moderately
related to the index of riot injuries (r = .35).
The headquarters index was also negatively related
to the measures of Negro-white polarity. The correlation
between the headquarters index and the Negro poverty
measure attained the level of -.28. It was also cor
related at the level of -.47 with the occupational segre
gation measure and at the level of -.22 with the index of
ghettoization. Because these same variables were also
negatively related to the three measures of riot activity,
there existed the possibility that the headquarters index
had its effect through them.
However, when the effects of these relationships
84
are controlled using partial correlation coefficients,
Table 3 shows that the original correlations between the
national headquarters index and the riot activity indi
cators decreased, as expected, but still remained moderate
ly strong. It may be observed on the table that only one
of the partial correlations decreased below .27 while
the other eight remained between .30 and .44, thus showing
that the headquarters index continues to make an independ
ent contribution.
Table 3 also shows that two variables included as
safeguards against spuriousness, the change in the non
white population and the per capita number of police,
failed to confound the results. These findings, then,
clearly support the original hypothesis that extralocal
integration is positively associated with both riot prone
ness and riot severity. Even when the relationship is
assessed independently of the effects of Negro-white
polarity, extralocal integration continues to be an
important predictor variable.
Local Integration
Concerning the polarity (i.e., nonintegration)
between Negroes and whites on the local level, it was
85
TABLE 3
THE RELATION BETWEEN EXTRALOCAL INTEGRATION AND RIOT
ACTIVITY HOLDING CONSTANT THE EFFECTS OF
RACIAL POLARIZATION
(N=114)
Independent and
Control Variables
Correlation between National
Headquarters and:
Number of Commission
Riots per Index
City
Injury
Index
Zero Order Correlation . 48a .47 .34
Partial Correlation
Holding Constant:
Occupational Segregation .37b .36 .26
Nonwhite Poverty .43 .43 .32
Ghettoization .35 .43 .31
Change in Nonwhite
Population .46 .45 .35
Police .41 .38 .24
aZero order, product moment correlation coeffi
cients .
r_
Partial correlation coefficients.
86
previously noted that the measures permitted the comparison
of two contrasting models of conflict. Under the urban
interdependence model, a high level of polarity should
be associated with a low level of riot activity. In
contrast, under a quasi-Marxian model, a high level of
polarity should be associated with a high level of riot
activity. Finally, under the quasi-Marxian model, depri
vation of Negroes should be positively associated with
riot activity.
Table 4 shows the various measures of Negro-white
polarity holding constant the effects of the headquarters
index. This control was required because of the inverse
relationship between extralocal integration and racial
polarization. Assessment of the overall results shown on
the table reveals that all nine partial correlation co
efficients continue to make a negative contribution to
riot activity, although several of the partial coefficients
fall below .19. Nonetheless, the fact that three measures
of racial polarization produce nine negative coefficients
even when the strong effect of the headquarters measure
is controlled clearly suggests that the quasi-Marxian
approach is not operative in the case of riot activity.
With that approach, one would expect the signs of the
87
TABLE 4
THE RELATION BETWEEN NEGRO-WHITE POLARITY AND RIOT
ACTIVITY HOLDING CONSTANT THE EFFECTS OF
EXTRALOCAL INTEGRATION AND NUMBER OF
POLICE
(N=114)
Independent Variables Dependent Variable
Zero Order Correlation
of:
Number of
Riots per
City
Commission
Index
Injury
Index
Occupational Segrega
1
CO
CO
-.37 -.26
tion
Nonwhite Poverty -.33 -.27 -.14
Ghettoization -.21 -.30 -.25
Police .29 .32 .31
Partial Correlation Holding
Constant National Headquarters
Occupational Segrega -. 20b -.20 -.12
tion
Nonwhite Poverty -.23 -.16 -.04
Ghettoization -.12 -.23 -.19
Partial Correlation Holding
Constant Number of Police
Occupational Segregation -.30 -.28 -.15
Nonwhite Poverty -.27 -.19 -.05
Ghettoization -.14 -.22 -.07
aZero order, product moment correlation coeffi
cients .
T_
Partial correlation coefficients.
88
coefficients to be positive rather than negative.
However, according to the present model, these signs are
in the expected direction. Thus, on an overall basis, it
may be concluded that the urban interdependence model is
the one supported by the pattern of results shown on
Table 4.
It is possible, however, to specify the relation
ships even further. In terms of the number of riots per
city (i.e., riot proneness), it may be noted on the table
that ghettoization has a partial correlation of -.12
while holding constant the effects of extralocal integra
tion. This implies that the ghettoization of Negroes is
a function of the community's extralocal integration. The
fact that those cities high on the extralocal integration
dimension tend to be less ghettoized (r = -.22 between
the two measures) indicates that these cities reflect the
national dictums of integration more than those low on
the dimension. This is a further specification of the
result found on Table 4 and is congruent with the present
model as high levels of integration should be positively
associated with high levels of riot proneness. In con
trast, under the quasi-Marxian model, one would expect
89
just the opposite because ghettoization implies intra-
racial solidarity.
Table 4 also shows that the index of occupational
segregation continues to make an independent contribution
to riots per city when controlling the effects of the
headquarters measure (partial r = -.20). This constitutes
further evidence against the quasi-Marxian approach.
Furthermore, the measure is even closer to Marx's main
ideas than the ghettoization index. As discussed in
Chapters II and III, Marx formulated his classic distinc-
tion between the worker and management. The categories
which comprise the occupational segregation index are
those of the worker, and if the quasi-Marxian approach is
correct, it should be here more than with ghettoization,
that polarity is positively related to conflict. Because
occupational segregation is negatively related to riot
proneness, support for the present model— which argues
t
that the relationship should be positive— is strongly
shown.
2
Private household workers, laborers, and farm
laborers.
90
Additional evidence against the quasi-Marxian
model is shown by the negative partial correlation of
-.23 between riots per city and the Negro poverty measure.
A fundamental statement of the Marxian approach is that
conflict is a means of alleviating deprivation and ex
ploitation. This statement would lead one to expect a
positive relation between Negro poverty and riot proneness
while just the opposite is shown on Table 4.
This negative relation is interpretable, however,
in terms of the present model. It was noted that the
measure of nonwhite poverty is not a pure one of economic
status per se. It is highly correlated with occupational
segregation (r = .67), and with the ghettoization index
(r = .48). These correlations indicate that nonwhite
poverty reflects a large degree of Negro exclusion from
white society on both residential and occupational dimen
sions. This is hardly surprising in itself, but it does
suggest two possible, although interrelated, interpreta
tions .
On the one hand, assuming that nonwhite poverty
reflects the exclusion of the Negro, the urban inter
dependence model would indicate that poverty is negatively
associated with riot activity. This is the same argument
91
made for the occupational segregation and ghettoization
indexes. In other words, the poverty measure may simply
be interpreted as another measure of Negro-white polariza
tion and the negative partial correlation on Table 4
constitutes evidence supporting the present model.
On the other hand, assume that nonwhite poverty
does reflect poverty (along with a component of segrega
tion) . This partially defines what is meant by a ghetto.
The concept of a ghetto refers to the isolation from the
larger community and usually implies that residents tend
to be impoverished (although this does not have to apply
universally). Furthermore, the Negro ghetto is thought to
be characterized by social integration (D. Warren, 1969;
Blauner, 1969). To view the Negro ghetto as a fragmented,
disorganized mass would be contrary to the evidence pre
sented in the literature.
If this is the case, it may be argued that the
intragroup social integration of the impoverished, ghetto
Negro leads to the weakening of intergroup integration
between blacks and whites, thereby causing a reduction
in the incidence of conflict. This argument, of course,
is based on the urban interdependence model and is counter
to that of a quasi-Marxian model.
92
Neither of the preceding two arguments can be
made in their pure form from the present data; and, in
actual fact, it is likely that both are interwoven. For
example, Negro poverty results in restricted employment
opportunities and restricted employment opportunities
result in poverty. Both aspects are likely to be opera
tive socially. Nonetheless, whether one accepts the first
argument, or the second, (or both conjointly), the evi
dence on Table 4 supports the urban interdependence model
over the quasi-Marxian.
Turning to the effects of the polarity measures on
the commission index of riot severity, Table 4 shows that
these measures are all negatively associated with severity.
Occupational segregation has a partial correlation of -.20
with the commission index of severity; nonwhite poverty
shows a similar partial correlation of -.16; and the
ghettoization index exhibits a partial correlation of -.23.
As discussed previously, this constitutes evidence against
the quasi-Marxian approach and supports the present model.
It may also be noted that while ghettoization has
negative effects on riot proneness through extralocal
integration, it is still negatively related to the com
mission index even when assessed independently of the
93
headquarters measure. This correlation allows further
specification of the ghettoization variable. Given that
a riot takes place, which is a function of extralocal
integration as far as ghettoization is concerned, high
levels of racial polarization are negatively related to
severity. In other words, ghettoization contributes
negatively to the severity of a riot but not to the riot
proneness of a city. This finding is congruent with the
urban interdependence model in terms of severity but not
with the quasi-Marxian framework.
Finally, Table 4 shows the effects of the polarity
measures controlling for the per capita number of police.
It may be seen that this control has no effect on the
measures of occupational segregation and nonwhite poverty.
Given that the police measure is highly related to the
index of national headquarters (r = .47), this is not
surprising; nor is it surprising that it reduces the value
of the ghettoization index from a zero order correlation
of -.21 to a partial correlation of -.14.
An interesting pattern of results is shown by the
reduction of all three zero order correlations concerning
the polarity measures and the injury index. When the
police variable is controlled, the independent effects of
94
the three polarity measures are greatly reduced (partial
r's = -.15, -.05, and -.07 for occupational segregation,
nonwhite poverty, and ghettoization, respectively).
It would appear that insofar as riot related
injuries are concerned, the police variable is the meaning
ful one. However, it may be noted that the riot commission
index is independent of the police measure (partial r's =
-.28, -.19, and -.22, for occupational segregation, non
white poverty, and ghettoization, respectively). Taking
both of these findings together, one is tempted to con
clude that the police are responsible for injuries related
to riots. While this may be considered, it should be
noted that the present data do not allow the linking of
police acts to riot consequences.
A more defensible interpretation is suggested by
the fact that the original bivariate correlations of .29,
.32, and .31, between the police measure and the number of
riots per city, commission index, and injury index, are
reduced to partial correlations of .08, .13, and .19 when
the effects of the headquarters measure are taken into
account. This implies that as the police measure is either
a spurious attribute of extralocal integration or that it
is a subdimension of extralocal integration. The latter
95
interpretation is congruent with the present model because,
as previously discussed, the police variable may be inter
preted as a measure of forced or imperative integration.
It is interesting to theorize that communities which are
extralocally integrated require a degree of imperative
integration on the local level; especially since extra
local integration tends to be inversely related to overall,
local integration (Turk, 1969b).
The obverse of segregation is integration. The
previous measures of segregation dealt with residential
and economic restrictions. The two measures of the
relative differences between Negroes and whites prove to
bear very little relation to the measures of riot activity.
Table 2 shows that the difference between white and Negro
unemployment rates (relative unemployment) was correlated
at the levels of -.12, -.17, and -.14, with the number of
riots per city, the commission index of riot severity,
The relation of the police to riot activity was
examined independently of the racial polarization measures
as a safeguard. The partial correlations, although slight
ly decreased, did not alter the basic zero order relation
ships. Furthermore, examining the effect of municipal
revenue taken as a crude index of governmental elaborate
ness (Turk, 1969b) on the police measure did not substan
tially alter the findings.
and riot injury index, respectively. The direction of
these correlations support the present model although
4
their absolute values are not large. However, the
differences between the Negro and white fertility ratios
(relative fertility) showed similarly low correlations but
in a positive direction. The correlations attained the
level of .12, .23, and .18, with the same measures of riot
activity, respectively. The direction here supports the
quasi-Marxian model. Although the .23 correlation between
the index of relative fertility and the commission index
constitutes a slight relation in the direction predicted
under the Marxian model, it would be unwarranted, given
the size of direction of the remaining five correlations,
to place too much credence on that single value. These
results, then, neither support nor fail to support either
the Marxian or present approaches.
Indeed, using partial correlation coefficients to
assess the measure of relatively unemployment independently
Under the present model, an increase in integra
tion should lead to a decrease in the relative differences
between Negroes and white, and therefore an increase in
the amount of riot activity. The opposite is true for
the Marxian based model.
97
of its correlation with migration decreases the original
bivariate correlations to .00, .04, and .05 for riots per
city, commission index and riot injuries. Since it is
known that migrants are often nonwhite and of relatively
low income (Bogue, 1969; 759-771), the relative unemploy
ment measure may reflect these migration patterns to a
degree that makes it unwise to place a great deal of faith
in it.
The same argument may be made concerning the
relation between the relative fertility measure and
migration. High Negro fertility has tended to be asso
ciated with high Negro migration (Bogue, 1969:652-683).
The largest bivariate correlation of .23 between the
relative fertility measure and the commission index of
riot severity is reduced to .15 when the effect of migra
tion is held constant. Unfortunately, data pertaining
directly to Negro migration were not available, and it
must be assumed that the migration of the total population
reflects Negro migration.
Assessment of Spuriousness
Included as safeguards against spuriousness was
city location in the five geographic regions of Northeast,
98
East North Central, Southeast, Southwest, and West.
Location, in itself, is hardly interesting, but it does
take on meaning as the social and cultural aspects related
to location can be identified and shown to be relevant.
Table 2 shows that location in the East North
Central and Western regions have little relationship with
riot activity. None of the correlations exceeded a value
of .15. However, for location in the Northeastern region,
it may be noted that a slight, positive relation exists
as the correlations are .21, .20, and .20 for riots per
city, the commission index of severity, and the riot
injury index, respectively. At the same time, Southeastern
location is correlated at the level of -.21, -.17, and
-.12 for the same respective variables, while Southwestern
location is correlated at the levels of -.28, -.35, and
-.27, again, for the same respective indexes of riot
activity. Although all of these correlations are of a
rather low magnitude, their pattern is interesting.
Why location in Southern regions should be nega
tively associated with riot activity while location in the
East is positively associated, defies certain interpreta
tion with the current data. There are, undoubtedly,
several historical factors such as slavery and legalized
99
segregation involved. These factors are outside the scope
of the present data. However, it is interesting to note
that the present data do show that Southern location is
negatively correlated with city density. This correlation
has a value of -.20 for location in the Southeast, and a
value of -.45 for location in the Southwest. In contrast,
Eastern location is positively correlated with density
(r = .53).
Population density immediately suggests the
Durkheimian concept of organic solidarity. It may be
that Southern regions are characterized by a low level
of organic solidarity while the Northeastern region is
not. Given this as a premise, it would then follow from
the present approach that location in the Northeast would
indicate a higher level of organic solidarity, and would
therefore be positively correlated with riot activity.
In contrast, location in Southern regions, indicating a
lower level of organic solidarity, would therefore be
negatively correlated with riot activity.
Indeed, when the relationship between regional
location and riot activity is assessed independently of
the density factor, the correlations between these three
locations and riot activity decrease. The partial
correlations vary from .00 to .15 in contrast to the
original bivariate correlations which varied from -.12 to
-.35. This behavior strongly suggests density and its
analog of integration as a more meaningful interpretation
than the omnibus nature of region qua region.
CHAPTER VI
SUMMARY AND CONCLUSIONS
In this final chapter, the results will be briefly
summarized and broad based conclusions derived. This
study has major implications for further theory and
research. These will also be discussed.
Summary
It was hypothesized that the level of a commu
nity's extralocal integration is positively associated
with conflict. Extralocal integration was measured by
the number of national headquarters located within the
community, while conflict was measured by the number of
riots per city, a severity index based on descriptive
categories, and an index based on the number of riot
related injuries. It was found that the headquarters
measure was positively related to the riot measures,
thus supporting the hypothesis.
The polarity (i.e., nonintegration) between
Negroes and whites was assessed by an index of ghettoiza-
101
102
tion, an index of occupational segregation, and a measure
of Negro family poverty. Racial polarity and extralocal
integration were found to be inversely related. Nonethe
less, even when the effect of this inverse relation was
controlled, extralocal integration continued to be an
important predictor of conflict.
Further results showed that the community1s level
of local integration (i.e., nonpolarity) was positively
associated with the incidence of conflict. These unex
pected results clearly pointed to the inadequacy of a
quasi-Marxian model of racial polarization and racial
solidarity leading to interracial conflict and, at the
same time, lent strong support to the present model.
In particular, it was found that occupational
segregation was negatively related to riot activity.
Since the quasi-Marxian model predicts a positive rela
tionship between these variables while the present model
predicts a negative relationship, these results supported
the latter model. Further evidence along these lines
was found by consideration of the correlation between
Negro poverty and riot activity. It was argued that Negro
povdrty should be positively related to riot activity
103
under the quasi-Marxian model but negatively related under
the present model. The negative relation found supports
the current model. It was also found that ghettoization
was only slightly, but negatively, related to riot activity
when extralocal integration was controlled, thus implying
that for residential segregation, extralocal integration
was the explanatory variable.
The per capita number of police, interpreted
finally as a measure of imperative or forced integration
was found to be positively related to the number of riot
related injuries, while density, interpreted in terms of
Durkheim's concept of organic solidarity, was found to
help explain the effects of location in the Southern
region of the country.
Further Theoretical Comparisons
The model proposed in this study is a very broad
one establishing a theoretical framework that may be
applied to a variety of social relationships. It repre
sents a first attempt to theoretically define and empiri
cally examine the proposition that conflict is positively
associated with integration. As such, the mechanisms by
which integration leads to conflict in a given situation
104
are unspecified.
A major theoretical position other than the quasi-
Marxian one, is that of relative deprivation. Tocqueville
(1964 translation), in a classic study, argued that it is
not the completely oppressed who engage in revolution,
but those oppressed who, for some reason, have the burden
of their oppression lifted slightly. In other words, as
the lot of the oppressed begins to improve, the prob
ability of revolution increases.
The form of this argument has been applied to the
case of race riots. It has been maintained that it is
the Negro whose socioeconomic status is improving who is
predisposed towards racial violence. Studies of riot
participants using survey techniques and based on this
approach have come to slightly mixed conclusions. On the
one hand, the studies of Newark and Detroit conducted for
the National Advisory Commission on Civil Disorders (1968)
revealed that Negro riot participants tended to be of
higher socioeconomic status than other ghetto residents.
Ransford (1968) demonstrated a positive relation between
subjective feelings of alienation and powerlessness, and
the advocacy of violence on the part Negroes in Watts,
105
Los Angeles. On the other hand, A. Campbell and Schuman
(1968) surveyed fifteen cities, again for the National
Advisory Commission on Civil Disorders. They point out
that economic and educational variables are not related
to riot participation in a simple, linear fashion but were
unable, in that preliminary report, to suggest precisely
what the form of the relationships are.
These studies would indicate that some variant of
relative deprivation theory is an important one; although
it is not possible to specify precisely the form of the
relationships. Also, relative deprivation implies a
reference group and it is not possible to specify who
constitutes the relevant reference group for deprived
ghetto residents. The reference group may be whites, or
it may be middle-class Negroes, or it may be relatively
well-off ghetto residents. However, assuming that the
bulk of the evidence is correct in the direction of their
relationship— i.e., as Negroes begin to attain socio
economic statuses equivalent to whites, the likelihood of
participating in, or advocating, racial violence
increases— one might ask how this affects the present
model.
106
It is felt that relative deprivation theories
constitute intervening variables between integration and
conflict. All the race riots studied here have occurred
in ghettos, although frequently directed against symbolic
surrogates of the white community, notably white merchants
and white police officers (National Advisory Commission
on Civil Disorders, 1968). Naive interpretation of this
fact would claim a paradox here with the present finding
that ghettoization is inversely associated with riots.
But the normative structure of the black ghetto is one of
interracial integration as discussed in Chapters III and
IV, and this is assumed to be especially true in cities
where most Negroes live outside the ghetto. In such
situations, two factors would be would be operative in
explaining the paradoxical findings that the Negro who is
more advanced socioeconomically in the ghetto, advocates
racial violence. First, where norms and examples of
interracial solidarity are most salient, any deviations
from these norms and examples are also likely to be salient
because they stand out as being counter to the norms of
integration, and because greater deprivation is experienced
relative to other Negroes. Being more visible, these
deviations are therefore more likely to be detected.
107
Furthermore, since behavior which is counter to the norms
of interracial solidarity are in violation of cherished
norms and interests, deviations are also likely to be
heavily sanctioned. The effect of these two processes
would be to increase the level of conflict. It should be
emphasized that the higher the integration, the greater
would be the relative impact of any remaining deprivation.
In this sense, integration is positively related to con
flict, even though it is the ghetto resident who riots.
Second, advocates and participants of race riots
are those who are relatively well off by ghetto standards
but not by absolute standards of the larger society. This
specifies the present theory by suggesting that those
whose oppression is still high but not absolute are the
most likely to react to violations of norms of integra
tion, especially where the extra-ghetto environment is
most likely to show these norms to be upheld, thereby
elevating aspirations.
Both of these factors are viewed as a further
specification of the variables which intervene between
integration and conflict. They serve to explain how it
is that integration leads to conflict while the present
108
model serves to focus on the existence of the relationship
and emphasize the possibility that such a relationship is
a valuable way of approaching phenomena previously concep
tualized as being on opposite ends of a continuum.
Further Implications
There is no reason why the fundamental statement
of the present model (i.e., the greater the integration,
the greater the conflict) must be restricted to urban
matters. The model implies that the relationship between
integration and conflict should be operative at other
levels of abstraction as well. Since the present study
was restricted to racial violence approached from a com
parative perspective, it was impossible to make any
statements about the individual from the data used. It
would be interesting in further research to assess the
isomorphism of the model by applying it to such diverse
social groupings as husbands and wives, friendship cliques
and work organizations. At this level of abstraction,
interdependence has been implied but not explicated (to
the author's knowledge). For example, it has long been
noted that sensitivity (or T) groups are never character
ized by conflict until the group has formed some kind of
109
cohesive bonds between members (Tannenbaum et al., 1961:
123-221). In the relation between client and psycho
therapist, a similar phenomenon has been noted. Until
the client comes to trust his therapist, conflict cannot
enter the dyadic relationship (Rogers, 1961:59-69).
This study implies that recognition of an inter
dependent relation between integration and conflict would
constitute a new and important way of viewing a wide
variety of phenomena.
Conclusions
The major conclusion to be derived from this study
is that integration and conflict are interdependent social
processes which are predictable one from the other.
Empirical support was found for the argument that the
presence of integration defines a potential for conflict,
and where conflict exists, integration must also exist.
Previous research and theory have failed to take
this interdependence into account in a large scale, sys
tematic fashion as, except in the reverse direction of
causation— i.e., the integrative function of social con
flict (cf. Simmel, 1955; Coser, 1956). Particularly in
the case of the empirical events studied here, racial
110
conflict has been seen as a response to deprivation and
segregation in quasi-Marxian terms. This research would
suggest the inadequacy of this approach. What the quasi-
Marxian approach fails to take into account is that
ghettoization and deprivation results in the isolation of
the Negro from the rest of the community.
Further generalization by examining other forms
of conflict such as management-worker strife, labor
strikes, and revolution would expand the present approach,
especially if they could be examined from both the view
point of the individual as well as the group.
The results of this study also showed the neces
sity for further theoretical delimitation of the various
types of conflict. For example, what role does moral
integration (Angell, 1951) play? What implications does
Durkheim's classic ideas of mechanical and organic solid
arity have? What is the relation between the types of
integration and the realistic and nonrealistic conflict
of which Coser (1956) speaks? It was not possible, in
this first exploration, to examine, let alone answer, any
of these questions in more than a briefly suggestive
fashion.
Finally, while the original model did require
further specification as the research progressed, it was
still true that a rather abstract, macro model did provide
the foundations for concrete, empirical research.
APPENDICES
PRELIMINARY RESULTS
112
113
APPENDIX I
PREDICTION OF RIOT ACTIVITY: LOW EXTRALOCAL INTEGRATIO
Number of Riots
per City
(R = .67) a
Commission
Index
(R = .48)
Independent
Variable
rb
Beta
Weight0
Independent
Variable
r
Beta
Weight
Indepe:
Varia]
Relative
fertility .38 .31
Relative
fertility .34 .31
Relati'
ferti
Relative
unemployment -.51 -.54
Relative
unemployment -.35 -.38
Relati'
unempli
Per capita
welfare ex
penditures -.10 -.20
Population
increase -.01 .29
Mentio:
civic
Population
increase -.07 .39
Mention of
civic groups -.04 .00d Migrat;
Mention of
civic groups .05 . 14d
Ghettoization -.16 .28
Multiple correlation coefficient.
Zero order correlation coefficient.
Q
Standardized regression coefficient.
P>.05? all other variables in the equation achieve the .05 level <
or beyond.
I
113
I
APPENDIX I
PREDICTION OF RIOT ACTIVITY: LOW EXTRALOCAL INTEGRATION
(N=38)
of Riots
2ity
. 67) a
Commission
Index
(R = .48)
Index of Riot
Injuries
(R = .47)
Beta
ffeightc
Independent
Variable r
Beta
Weight
Independent
Variable
r
Beta
Weight
.31
Relative
fertility .34 .31
Relative
fertility .26 .25
-.54
Relative
unemployment -.35 -.38
Relative
unemployment -.27 -.50
-.20
Population
increase -.01 .29
Mention of
civic groups .15 .21
.39
Mention of
civic groups -.04 .00d Migration -.04 .45
.14d
.28
orrelation coefficient,
correlation coefficient,
ad regression coefficient.
1 other variables in the equation achieve the .05 level of significance
114
APPENDIX II
PREDICTION OF RIOT ACTIVITY: MEDIUM EXTRALOCAL INTEGRAT
Number of Riots
per City
(R = .67)a
Commission
Index
(R = .54)
Independent
Variable rb
Beta
weight0
Independent
Variable r
Beta
Weight
Independent
Variable
Relative
fertility .34 .51
Relative
fertility .28 .64
Relative
fertility
Relative
unemployment .17 .25
Index of reform
government .15 .45
Index of re
government
Index of reform
government .32 .60
Mention of
civic groups .00 -.47
Mention of
groups
Per capita edu
cation expend. .20 .28
Nonwhite popu
lation change .13 .42
Nonwhite pc
lation chs
Mention of
civic groups .12 -.25
Multiple correlation coefficient.
T_
Zero order correlation coefficient.
Standardized regression coefficient.
j
P>.05; all other variables in the equation achieve the .05 level
or beyond.
APPENDIX II
PREDICTION OF RIOT ACTIVITY: MEDIUM EXTRALOCAL INTEGRATION
(N=35)
of Riots
City
= . 67)a
Commission
Index
(R = .54)
Index of Riot
Injuries
(R = .59)
Beta
weight0
Independent
Variable r
Beta
Weight
Independent
Variable r
Beta
Weight
.51
Relative
fertility .28 .64
Relative
fertility .36 .61
.25
Index of reform
government .15 .45
Index of reform
government .23 .36
.60
Mention of
civic groups .00 -.47
Mention of civic
groups .28 -.14d
.28
Nonwhite popu
lation change .13 .42
Nonwhite popu
lation change .21 .43
-.25
orrelation coefficient.
correlation coefficient,
ed regression coefficient.
.1 other variables in the equation achieve the .05 level of significance
115
■ 3
APPENDIX III
PREDICTION OF RIOT ACTIVITY: HIGH EXTRALOCAL INTEGRATI
Number
per
(R =
of Riots
City
. 52) a
Commission
Index
(R = .48)
Independent
Variable rb
Beta
Weight0
Independent
Variable
Beta
r Weight
Independi
Variabli
Nonwhite
poverty -.38 -.39
Occupational
segregation -.40 -.25
Per capi-
hospital
penditu]
Relative
fertility -.26 -.24
Mention of
civic groups -.09 .94
Mention <
civic g]
Mention of
civic groups -.03 -.04d Ghettoization -.41 -.29 Migratior
Migration -.19 -.23 Per capit
governme
expend.
Ghettoiza
aMultiple correlation coefficient.
Zero order correlation coefficient.
Standardized regression coefficient.
dP>.05; all other variables in the equation achieve the .05 level o
or beyond.
115
APPENDIX III
PREDICTION OP RIOT ACTIVITY: HIGH EXTRALOCAL INTEGRATION
(N=41)
Riots
Y
)a
Commission
Index
(R = .48)
Index of Riot
Injuries
(R = .67)
Beta
eight0
Independent
Variable
Beta
r Weight
Independent
Variable
r
Beta
Weight
.39
Occupational
segregation .40 -.25
Per capita
hospital ex
penditures .39 .31
.24
Mention of
civic groups .09 .94
Mention of
civic groups .05 -.14
. 04d Ghettoization .41 -.29 Migration -.40 -.26
.23 Per capita
government
expend.
Ghettoization
.44
-.39
.21
-.21
rrelation coefficient,
correlation coefficient.
1 regression coefficient.
other variables in the equation achieve the .05 level of significance
BIBLIOGRAPHY
BIBLIOGRAPHY
Aiken, Michael, and Jerald Hage
1968 "Organizational interdependence and intra-
organizational structure." American Sociological
Review 33 (December):912-931.
Angell, Robert C.
1951 "The moral integration of American cities."
American Journal of Sociology 57 (July):part 2.
Bales, Robert F.
1949 Interaction Process Analysis. Cambridge, Mass.:
Addison-Wesley.
Banefield, Edward C., and James Q. Wilson
1966 "Ethnic membership and urban voting." Pp. 168-177
in Bernard E. Segal (ed.), Racial and Ethnic
Relations. New York: Crowell.
Blauner, Robert
1964 Alienation and Freedom. Chicago: University of
Chicago Press.
1969 "Internal colonialism and ghetto revolt." Social
Problems 16 (Spring):464-483.
Blood, Robert 0., and Donald M. Wolfe
1960 Husbands and Wives. New York: The Free Press.
Bloombaum, Milton
1968 "The conditions underlying race riots as portrayed
by multidimensional scalogram analysis." American
Sociological Review 33 (February):76-92.
Bogue, Donald J.
1969 Principles of Demography. New York: Wiley.
117
118
Buckley, Walter
1967 Sociology and Modern Systems Theory. New York:
Prentice-Hall.
Buggs, John A.
1969 "Report from Los Angeles." Pp. 493-506 in Robert
R. Evans (ed.), Collective Behavior. Chicago:
Rand McNally.
Campbell, Angus, and Howard Schuman
1968 Racial Attitudes in Fifteen American Cities.
Ann Arbor: The University of Michigan Institute
for Social Research.
Campbell, Donald T.
1957 "Factors relevant to the validity of experiments
in social settings." Psychological Bulletin 54
(July):298-312.
Clark, Terry N.
1968 "Community structure, decision making, budget
expenditures, and urban renewal in 51 American
communities." American Sociological Review 33
(August):576-593.
Coleman, James S.
1958 Community Conflict. New York: The Free Press.
Cooley, Charles H.
1909 Social Organization. New York: Scribners.
Coser, Lewis
1956 The Function of Social Conflict. New York:
The Free Press.
1957 "Social conflict and the theory of social change."
British Journal of Sociology 8 (September):197-207.
1964 "Conflict." Pp. 123-124 in Julius Gould and
William L. Kolb (eds.), A Dictionary of the Social
Sciences. New York: The Free Press.
119
Crain, Robert L., Elihu Katz, and Donald B. Rosenthal
1968 Community Structure and Innovation. Indianapolis:
Bobbs-Merrill.
Dahlke, Otto H.
1969 "Race and minority riots." Pp. 394-404 in Robert
R. Evans (ed.), Collective Behavior. Chicago:
Rand McNally.
Dahrendorf, Ralf
1958 "Out of utopia." American Journal of Sociology
64 (September):116-127.
1959 Class and Class Conflict in Industrial Society.
Stanford: Stanford University Press.
Durkheim, Emile
1933 The Division of Labor in Society. George Simpson
(trans.). New York: Macmillan.
Emery, F. E., and E. C. Trist
1965 "Causal texture of organizational environment."
Human Relations 18 (July):21-32.
Encyclopedia of Associations
1961 Geographic and Executive Index, Vol. II.
Detroit: Gale.
Gibbs, Jack P.
1961 "Introduction." Pp. 1-11 in Jack P. Gibbs (ed.),
Urban Research Methods. New York: Van Nostrand.
Gibbs, Jack P., and Walter T. Martin
1962 "Urbanization, technology, and the division of
labor." American Sociological Review 27 (October):
667-677.
Gordon, Milton
1964 Assimilation in American Life. New York: Oxford
University Press.
Gordon, Robert A.
1968 "Issues in multiple regression." American Journal
of Sociology 73 (March):592-616.
120
Greer, Scott
1965 Urban Renewal in American Cities. Indianapolis:
Bobbs-Merrill.
Grimshaw, Allen D.
I960 "Urban racial violence in the United States."
American Journal of Sociology 66 (September):
109-119.
Grindstaff, Carl F.
1968 "The Negro, urbanization, and relative depriva
tion in the Deep South." Social Problems 15
(Winter):342-352.
Hall, Richard H.
1963 "The concept of bureaucracy." American Sociologi
cal Review 69 (July):32-40.
Hillery, George A., Jr.
1955 "Definitions of community." Rural Sociology 20
(June):11.
1968 Community Organization. Chicago: University of
Chicago Press.
Homans, George C.
1950 The Human Group. New York: Harcourt, Brace and
World.
Horowitz, Irving L.
1962 "Consensus, conflict and cooperation." Social
Forces 41 (December):177-188.
Hundley, James R., Jr.
1969 "The dynamics of recent ghetto riot." Pp. 480-492
in Robert R. Evans (ed.), Collective Behavior.
Chicago: Rand McNally.
Kaufman, Harold F.
1959 "Toward an interactional conception of community."
Social Forces 38 (October):9-17.
Kitano, Harry H. L.
1969 Japanese Americans. Englewood Cliffs: Prentice-
Hall.
121
Keyes, Fenton
1958 "The correlation of social phenomena with commu
nity size." Social Forces 36 (May):111-115.
Landecker, Werner S.
1951 "Types of integration and their measurement."
American Journal of Sociology 56 (January) :332-
340.
Lewin, Kurt
1951 Field Theory in Social Science. Dorwin Cartwright
(ed.). New York: Harper and Row.
Lewis, Oscar
1951 Life in a Mexican Village. Urbana: University
of Illinois Press.
Lieberson, Stanley, and Arnold R. Silverman
1965 "The precipitant and underlying conditions of
race riots." American Sociological Review 30
(December):887-899.
Litwak, Eugene, and Lydia F. Hylton
1962 "Interorganizational analysis." Administrative
Science Quarterly 6 (March) :395-426 .
Loomis, Charles P.
1963 "Introduction." Pp. 1-29 in Ferdinand Toennies,
Community and Society. Charles P. Loomis (trans.)
New York: Harper Torchbooks.
Lopata, Helena Znaniecki
1964 "The function of voluntary associations in an
ethnic community." Pp. 203-223 in Ernest W.
Burgess, and Donald J. Bogue (eds.), Contributions
to Urban Sociology. Chicago: University of
Chicago Press.
Maclver, Robert M.
1928 Community. New York: Macmillan.
Marris, Peter, and Martin Rein
1967 Dilemmas of Social Reform. New York: Atherton.
122
Martindale, Don
1960 The Nature and Types of Sociological Theory.
Boston: Houghton Mifflin.
Marx, Karl, and Friedrich Engles
1964 The Communist Manifest. Francis Bu Randall
(trans.). Joseph Katz (ed.). New York:
Washington Square Press.
McNemar, Quinn
1962 Psychological Statistics. 3d ed. New York:
Wiley.
Merton, Robert K., and Robert A. Nisbet
1966 Contemporary Social Problems. 2d ed. New York:
Harcourt, Brace and World.
Miller, Walter B.
1958 "Inter-institutional conflict as a major impedi
ment to delinquency prevention." Human Organiza
tion 17 (Fall):20-23.
Mills, C. Wright
1962 The Marxists. New York: Dell.
Nadel, S. F.
1964 The Theory of Social Structure. New York: The
Free Press.
National Advisory Commission on Civil Disorders
1968 Report to the President. Washington, D.C.:
U.S. Government Printing Office.
North, Robert C.; Koch, Howard E., Jr.; and Dinna A.
Zinnes
I960 "The integrative functions of conflict." Journal
of Conflict Resolution 4 (September):355-374.
Parsons, Talcott
1951 The Social Systems. New York: The Free Press.
Parsons, Talcott, and Neil J. Smelser
1956 Economy and Society. New York: The Free Press.
123
Rainwater, Lee
1969 "Open Letter on white justice and the riots."
Pp. 507-520 in Robert R. Evans (ed.), Collective
Behavior. Chicago: Rand McNally.
Ransford, Edward H.
1968 "Isolation, powerlessness, and violence."
American Journal of Sociology 73 (March):581-591.
Redfield, Robert
1941 The Folk Culture of Yucatan. Chicago: University
of Chicago Press.
1947 "The folk society." American Journal of Sociology
12 (January):293-308.
Rogers, Carl R.
1961 On Becoming a Person. Boston: Houghton Mifflin.
Schelling, Thomas C.
1963 The Strategy of Conflict. New York: Oxford
University Press.
Shepherd, Clovis R.
1964 Small Groups. San Francisco: Chandler
Simmel, Georg,
1955 Conflict. Kurt H. Wolff (trans.). New York:
The Free Press.
Sirjamaki, John
1964 "The institutional approach." Pp. 33-50 in
Harold T. Christensen (ed.), Handbook of Marriage
and the Family. Chicago: Rand McNally.
Sjoberg, Gideon
1964 "Community." Pp. 114-115 in Julius Gould and
William L. Kolb (eds.), A Dictionary of the
Social Sciences. New York: The Free Press.
Smelser, Neil J.
1962 Theory of Collective Behavior. New York: The
Free Press.
124
Stinchcombe, Arthur L.
1965 "Social structure and organizations." Pp. 142-193
in James G. March (ed.), Handbook of Organizations.
Chicago: Ranc McNally.
Stouffer, Samuel A.
1950 "Some observations on study design." American
Journal of Sociology 40 (January):356-361.
Strauss, Anselm L.
1968 "Strategies for discovering urban theory."
Pp. 78-98 in Leo F. Schnore (ed.), Social Science
and the City. New York: Praeger.
Street, David, and John C. Leggett
1961 "Economic deprivation and extremism." American
Journal of Sociology 67 (July):53-57.
Sumner, William G.
1906 Folkways. New York: Ginn.
Taeuber, Karl E., and Alma F. Taeuber
1965 Negroes in Cities. Chicago: Aldine
Tannebaum, Robert; Weschler, Irving R.; and Fred Massarik
1961 Leadership and Organization. New York: McGraw-
Hill.
Timms, Duncan W. G.
1969 "The dissimilarity between overseas-born and
Australian-born in Queensland." Sociology and
Social Research 53 (April):363-374.
Toennies, Ferdinand
1957 Community and Society. Charles P. Loomis (trans.).
East Lansing: Michigan State University Press.
Tocqueville, Alexis de
1964 Democracy in America. Francis Brown (ed.).
Andrew Hacker (trans.). New York: Washington
Square Press.
125
Turk, Herman
1969a "Comparative urban studies in interorganizational
relations." Sociological Inquiry 39 (Winter):
101-110.
1969b "Interorganizational networks in urban society."
American Sociological Review, forthcoming.
Turk, Herman, and Robert Jiobu
1969 "Interorganizational systems in urban society."
Paper presented at the Annual Meeting of the
American Sociological Society, San Francisco,
California.
Turner, Ralph H., and Lewis M. Killian
1957 Collective Behavior. Englewood Cliffs: Prentice-
Hall.
Udy, Stanley H., Jr.
1965 "The comparative analysis of organizations."
Pp. 678-709 in James G. March (ed.), Handbook of
Organizations. Chicago: Rand McNally.
U.S. Bureau of the Census
1962 County and City Data Book, 1962. Washington, D.C.:
U.S. Government Printing Office.
1963 U.S. Census of Population, 1960. Number of
Inhabitants. Washington, D.C.: U.S. Government
Printing Office.
1967 County and City Data Book, 1967. Washington, D.C.:
U.S. Government Printing Office.
U.S. Department of Justice
1961 Crime in the United States. Washington, D.C.:
Federal Bureau of Investigation.
Wanderer, Jules J.
1969 "Index of riot severity." American Journal of
Sociology 74 (March):500-505.
Warren, Donald I.
1969 "Detroit." Social Problems 16 (Spring):464-483.
126
Warren,
1956a
1956b
1963
Wilier,
1967
Roland L.
"Toward a reformulation of community theory."
Human Organization 15 (Summer):8-11.
"Toward a typology of extra-community controls
limiting local community autonomy." Social Forces
34 (May):338-341.
The American Community. Chicago: Rand McNally.
David
Scientific Sociology. Englewood Cliffs: Prentice-
Hall.
Williams, Robin
1947 Reduction of Intergroup Tensions. New York:
Social Science Research Council.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Some Behavioral Consequences Of Career Success: A Synthesis Of Reward Andbalance Approaches
PDF
Attitudes Of Ministers And Lay Leaders Of The American Baptist Conventionof The State Of Washington On Selected Social Issues
PDF
The Effect Of Differential Treatment On Attitudes, Personality Traits, And Behavior Of Adult Parolees
PDF
Referential Dissociation And Response To Stress
PDF
Perception Of The Power Structure By Social Class In A California Community
PDF
Effects Of Integrated School Experience On Interaction In Small Bi-Racialgroups
PDF
Ecology Of Negro Communities In Los Angeles County: 1940-1959
PDF
Delinquency And Rationalization: A Study Of The Delinquent Act
PDF
Normative values of selected law enforcement officers and adult male offenders
PDF
Group Factors And Individual Internalization Of A Value
PDF
Passage Through The Expressive Student Subculture: The Initiation And Cessation Of Marijuana Use
PDF
Economic Differentiation And Social Organization Of Standard Metropolitanareas In The United States: 1950
PDF
Social Components Of Housing Cost In The Western Metropolis
PDF
Role Conflict Resolution In Complex Social Systems: A Polyarchical Systems Model For Organizational And Social Change
PDF
S.T.P.: A Simulation Of Treatment Processes
PDF
A Study Of Factors Related To Police Diversion Of Juveniles: Departmentalpolicy And Structure, Community Attachment, And Professionalization Of Police
PDF
Status Consistency Among The Clergy
PDF
The Construction And Empirical Test Of A Theory Based On Selected Variables In Small-Group Interaction
PDF
European Economic Integration And African States
PDF
Role Expectations Of American Undergraduate College Women In A Western Coeducational Institution
Asset Metadata
Creator
Jiobu, Robert Masao
(author)
Core Title
Integration And Conflict: Initial Perspectives On Urban Race Riots
Degree
Doctor of Philosophy
Degree Program
Sociology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
OAI-PMH Harvest,sociology, ethnic and racial studies
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Turk, Herman (
committee chair
), Sabagh, Georges (
committee member
), Storm, William Bruce (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c18-390439
Unique identifier
UC11362218
Identifier
7008528.pdf (filename),usctheses-c18-390439 (legacy record id)
Legacy Identifier
7008528.pdf
Dmrecord
390439
Document Type
Dissertation
Rights
Jiobu, Robert Masao
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
sociology, ethnic and racial studies