- ’ ... wv _,__'_

 

MIGRATION IN NONMETROPOLITAN
COUNTIES: AN ECOLOGICAL APPROACH

Dissertation for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY '
CAROLYN. TYIRIN KIRK ‘

1974

 

n—v-l‘

LIBRARY

IILIIIIIIIIIZIIIILIIIIIIIIIIIIIJIIIIIIIIIIIBIII M

University I 5-

 

This is to certify that the

thesis entitled

Migration in Nonmetropolitan Counties:
An Ecological Approach

presented by

Carolyn 'Iyirin Kirk

has been accepted towards fulfillment
of the requirements for

Ph.D. Sociology

Jegree in

2 75 Major more:

 

Date LI/16/7LI

 

0-7639

 

.. 800K BINUFW "IE I;
JEQARY v .‘JEI‘; :‘I

III ivnw.»-:~ ml

 

ABSTRACT

MIGRATION IN NONMETROPOLITAN COUNTIES:
AN ECOLOGICAL APPROACH

By
Carolyn Tyirin Kirk

Utilizing a modified version of Hawley's ecological
model of the process of territorial versus structural differ-
entiation, this study examined the relationship between both
organization and environment as well as changes in both and
the net-migration rate. Specifically, analyzing 227 nonmet-
ropolitan counties in the North Central Division during the
1960-70 decade through various techniques of correlational
analysis, the study tested the hypothesis that both posited
independent components of the ecological complex and changes
in each affect the net-migration rate directly with organiza-
tion having a stronger effect than environment.

Simple correlational analysis revealed indicators of
both organization and environment to be directly related to
migration in the posited directions based on the model with
the former having a greater impact than the latter indepen-
dent component. Moreover, diversity of structure, either of
or easily accessible to a pepulation, was the best pre-

dictor of the net-migration rate followed by variables

Carolyn Tyirin Kirk
measuring various aspects of manufacturing and institutional
populations. Dividing the sample into a rural and an urban
sub-sample showed few differences in the relative rank of
the factors most highly correlated with migration.

Stepwise multiple regression results showed that
organization explained over half and environment slightly
under a fourth of the variance in the dependent variable.
Combining both sets of independent variables and using
stepwise multiple regression and partial correlational
analysis revealed, however, that environment had a negli-
gible influence on the net-migration rate. 0n the other
hand, the partials for the most important organizational
variables showed almost no change between the analysis
utilizing organizational measures alone and the examination
employing both sets of independent variables. Such results,
coupled with a strong association between organizational
diversity and environmental nearness to an SMSA, indicate
a need to revise the model by positing organization to have
a direct impact on net-migration and environment to have an
indirect effect through its influence on organization.

Examination of measures of change also indicated
that organization has a direct influence on netdmigration,
although the two factors most highly correlated with
migration may measure components of the complex other than
those for which they were designated. Thus, this analysis

showed the continuing methodological problem of developing

Carolyn Tyirin Kirk
meaningful indices that clearly stand for only one compenent

of the ecological complex.

MIGRATION IN NONMETROPOLITAN COUNTIES:
AN ECOLOGICAL APPROACH

By

Carolyn Tyirin Kirk

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Sociology

1974

ACKNOWLEDGMENTS

Although it is not possible to acknowledge all who
have helped one throughout a dissertation and a graduate
career, there are several individuals whose help has been
invaluable. I especially thank Professor J. Allan Beegle
for his insights and suggestions on the dissertation and
for his guidance in formulating the study. From a more
practical standpoint, I also appreciate his efforts to
ease the strain of writing a dissertation away from
campus--efforts which included everything from providing
copying services to reading the first draft even more
quickly than I had hoped for. Thanks are also given
to Professors James J. Zuiches, James B. McKee, K. Dennis
Kelly and Harry Perlstadt for their critical insights
on both the original prOposal and the dissertation.

A very special appreciation is extended to Grafton
D. Trout who started it all when he unknowingly imparted
to me a love of sociology in an undergraduate urban
sociology course. I am also indebted to him for encouraging
me to pursue my interest in history along with my major
field in graduate school.

The greatest debt I owe and the largest thank you
I give is to Gordon W. Kirk, Jr., my husband, who served

ii

as research assistant and editor on this project even
though it meant countless hours of lost time on his own
research. I also especially thank him for his emotional
support when I was frustrated, his willingness to serve
as a sounding board, and his patience in listening to me
ramble seemingly without end about a subject in which he
has at best only a passing interest. To him I dedicate

this dissertation.

iii

TABLE OF CONTENTS

ACKNOWLEDGMENTS
LIST OF TABLES

Chapter
I. INTRODUCTION
PURPOSE OF THE STUDY
THEORY AND LITERATURE .
II. METHODOLOGY .
Unit of Analysis .
Description of the Sample .
Dependent Variable

Independent Variables
Measures of Organization

Measures of Organizational Change :

Measures of Environment

Measures of Environmental Change
Data .

Method of Analysis

III. ZERO-ORDER CORRELATIONAL ANALYSIS

Organization and Environment

Organizational and Environmental
Change

Summary .

IV. STEPWISE MULTIPLE REGRESSION AND
PARTIAL CORRELATIONAL ANALYSES

STEPWISE MULTIPLE REGRESSION ANALYSIS

Measures of Organization and
Environment . . .

Measures of Organizational and
Environmental Change

Summary .

iv

ii

vi

67
69

69

77
80

TABLE OF CONTENTS (Continued)

Chapter Page
PARTIAL CORRELATIONAL ANALYSIS . . . . . . . . . 81
Measures of Organization and
Environment . . . . . . . . . 82
Measures of Organizational and
Environmental Change . . . . . . . . . . . 85
Theoretical Implications . . . . . . . . . 89
Stepwise Multiple Regression of
Selected Variables . . . . . . . . . . . . 92
V. CONCLUSIONS . . . . . . . . . . . . . . . . . . 95
The Ecological Model . . . . . . . . . . 95
Methodological Considerations . . . . . . . . 96
Areas for Further Research . . . . . . . . . 98
Policy Implications . . . . . . . . . . . . . 108
Appendices
I. MEANS AND STANDARD DEVIATIONS . . . . . . . . . 111
II. CORRELATION MATRICES . . . . . . . . . . . . . . 117
III. LEVELS OF SIGNIFICANCE . . . . . . . . . . . . . 147

IV. STEPWISE MULTIPLE REGRESSION ANALYSIS
UTILIZING INDEPENDENT VARIABLES

GROUPED INTO SUBSTANTIVE CATEGORIES . . . . . 149
V. EXPLICATION OF VARIABLES . . . . . . . . . . . . 155
BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . 158

Table

10.

LIST OF TABLES

ZERO-ORDER CORRELATIONS BETWEEN
MEASURES OF ORGANIZATION AND
NET-MIGRATION . . . . .

ZERO-ORDER CORRELATIONS BETWEEN
MEASURES OF ENVIRONMENT AND
NET-MIGRATION . . . . .

ZERO- ORDER CORRELATIONS BETWEEN
MEASURES OF ORGANIZATIONAL CHANGE
AND NET- MIGRATION . .

RESULTS OF STEPWISE MULTIPLE REGRESSION
OF MEASURES OF ORGANIZATION WITH
NET-MIGRATION . . . . . . .

RESULTS OF STEPWISE MULTIPLE REGRESSION
OF MEASURES OF ENVIRONMENT WITH
NET- MIGRATION . . .

RESULTS OF STEPWISE MULTIPLE REGRESSION
OF MEASURES OF ORGANIZATION AND
ENVIRONMENT WITH NET-MIGRATION

RESULTS OF STEPWISE MULTIPLE REGRESSION
OF MEASURES OF ORGANIZATIONAL
CHANGE WITH NET-MIGRATION .

RESULTS OF STEPWISE MULTIPLE REGRESSION
OF MEASURES OF ORGANIZATIONAL AND
ENVIRONMENTAL CHANGE WITH NET-
MIGRATION . . . . . . . . .

PARTIAL CORRELATIONS BETWEEN MEASURES
OF ORGANIZATION AND NET-MIGRATION .

PARTIAL CORRELATIONS BETWEEN MEASURES
OF ORGANIZATION AND ENVIRONMENT
AND NET- MIGRATION .

vi

Page

53

57

6O

72

75

76

78

80

82

84

Table
11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

LIST OF TABLES (Continued)

PARTIAL CORRELATIONS BETWEEN MEASURES
OF ORGANIZATIONAL CHANGE AND
NET- MIGRATION . . . .

PARTIAL CORRELATIONS BETWEEN MEASURES
OF ORGANIZATIONAL AND ENVIRONMENTAL
CHANGE AND NET-MIGRATION .

RESULTS OF STEPWISE MULTIPLE REGRESSION
OF SELECTED MEASURES OF ORGANIZATION
AND ENVIRONMENT WITH NET-MIGRATION.

MEANS AND STANDARD DEVIATIONS FOR ALL
VARIABLES FOR THE ENTIRE SAMPLE
(N=227) . . .

MEANS AND STANDARD DEVIATIONS FOR ALL
VARIABLES FOR THE RURAL (AGRICUL-
TURALLY SPECIALIZED) COUNTIES
(N-167) . . .

MEANS AND STANDARD DEVIATIONS FOR.ALL
VARIABLES FOR THE URBAN (DIVERSIFIED)
COUNTIES (N860) . .

ZERO-ORDER CORRELATIONS FOR TOTAL
SAMPLE (N=227) . . .

ZERO~ORDER CORRELATIONS FOR RURAL
(AGRICULTURALLY SPECIALIZED)
SUB-SAMPLE (167 COUNTIES)

ZERO- ORDER CORRELATIONS FOR URBAN
(DIVERSIFIED) SUB- SAMPLE
(60 COUNTIES) . . . .

RESULTS OF STEPWISE MULTIPLE
REGRESSION . . .

vii

Page

86

87

93

104

106

109

111

121

131

147

CHAPTER I
INTRODUCTION

Human ecology has progressed from an emphasis on
community spatial and temporal relationships to a more
cogently defined perspective with a major emphasis on
explaining the causes and effects of organization in
relationship to population, environment, and technology.
The impetus for this redirection came largely from the

1950 publication of Hawley's Human Ecology: A Theory of

 

CommunityStructure.li Although still tied to some extent

 

to the earlier tendency of researchers to dwell primarily
on the spatial and temporal aspect of local community
structure, Hawley's approach marked a serious effort "to
restore a conceptual continuity with plant and animal
ecologies."2 It also resulted in emphasizing a broader
view of sustenance organization than had characterized
the earlier empirical studies of the Chicago School.
Following this direction, ecological theorists

since 1950 have developed a model consisting of four

 

1Amos H. Hawley, Human Ecology: A_Theory QEDCom-
munity Structure (New York, 1950).

 

2Amos H. Hawley, "Human Ecology," International
Encyclopedia pf the Social Sciences (New York, 1967), 319.

 

l

2

interacting components designated as the ecological complex;3
The variables include population, organization, environment
and technology. In this schema population refers to the
demographic characteristics of a set of inhabitants in a
given territory, e.g., age-sex structure and size. Further-
more, it is posited that a population's structure and size
is continually moving towards a state of equilibrium in
regard to the other three components while at the same time
inducing further change in the other three variables. In
‘moving toward this equilibrium, migration is the major means
by which a population changes in the short run.4 0n the
other hand, fertility and mortality (except under special
conditions where systematic policies of fertility and/or
‘mortality control are instituted) are long-run phenomena
effecting population change.

Agreement on the conceptualization of organization

is far less common than for population. Although all agree

 

3See in particular Otis Dudley Duncan, "Human

Ecology and Populations Studies," in The Stud of P0 ula-
tion, An Inventor and Appraisal, ed._Ey'PHIIIp_Mfl auser
and Otis DfidIey Duncan, (Chicago, 1959), 678-716 and Otis
Dudley Duncan and Leo F. Schnore, "Cultural, Behavioral,
and Ecological Perspectives in the Study of Social Organi-
zation," The American Journal g§_Sociology, 65, (September,
1959). 137:56.

 

4Donald J. Bogue, Components of Population Change
1940-50: Estimates gf Net-Migration afia NaturaI Increase
:23 Each Standard Metropolitan Area and State Economic
Area (Oxford, I957). Bogue finds, for exampIe, that in
the 1940-50 decade, percentage change in total population
over all nonmetropolitan state economic areas correlates
with net-migration at .917, p. 26.

 

3
it is a structural variable differing from what can be
called economic organization in the sense of pertaining
to factors amenable to cost analysis, the breadth of its
definition varies substantially among theorists. Gibbs
and Martin, for example, define organization narrowly in
terms of sustenance. More precisely, sustenance activities
are activities which provide a population with a livelihood.
In their illustration of this definition, they specifically
include types of economic concerns and occupations such as
a large department store, a municipal power company, an
independent taxi cab driven by its owner, and a house-
wife.5 Likewise, Duncan and Schnore view organization "as
a ramification of sustenance activities," but broadly
conceived. However, they do not specify the parameters
of the concept.6 Hawley, on the other hand, initially
defines the concept very broadly by stating that ”ecological
organization pertains to the total fabric of dependences
that exist within a population."7 Furthermore, both this
earlier definition (1950) and a later discussion (1967) of
the term imply that organization is similar to if not identi-
cal with social organization. Hawley does, however, note

that this ecological conceptualization emphasizes functional

 

5Jack P. Gibbs and Walter T. Martin, "Toward a
Theoretical System of Human Ecology," Pacific Sociological
Review, 2, (Spring, 1959), 30-3.

 

6Duncan and Schnore, 136.

7Hawley (1950). p. 179.

4
structures and excludes the normative aspects of social

8 Yet, like Duncan and Schnore, he does not

organization.
specify concretely what is included in ecological organiza-
tion.

Similar to the broader views of organization,
environment is easily defined in the abstract; the diffi-
culty comes in defining specifically what is included under
the rubric. Conceptually, environment refers to factors
both within and outside a unit under study which actually
influence or can potentially influence a population by
aiding or impeding the utilization of resources. Moreover,
it includes not only the physical environment but also
other populations or the social environment. Such factors
as climate and topography fall easily within the environ-
mental category. Others, particularly those pertaining
to social environment, are not easily classified. That is,
an apparently infinite number of historical situations with
regard to a given population's position vis-a-vis other
populations complicates the cataloging of specific social
environmental factors. For example, excluding the U.S.S.R.
should all Eastern EurOpean nations commonly referred to
as the Communist bloc be viewed as being in the same
ecological position vis-a-vis the U.S.S.R. or do the
various situations differ sufficiently to categorize the

countries into two, three or more groups in reference to

 

8Hawley (1967), 329-330, 337.

5

this particular social environmental variable, the Russian
population? Until more systematic research is done in this
area, it is perhaps only possible to conclude with Hawley
that environment "has no fixed content and must be defined
anew for each different object of investigation."9

Technology, the fourth component of the complex,
generally means technology in use. It includes both the
types and quantities of tools and techniques used in
exploiting resources and their effectiveness in providing
sustenance and in this manner places limits on both the
quantity and the quality of resource exploitation.10

The basic premise of ecology is that the four
components of the ecological complex are reciprocally
related to each other. That is, in order to analyze any
one element of the complex adequately, one has to consider
the other three components of the complex. Traditionally,
the major goal of human ecologists has been to explain
organization in terms of the other three variables. How-
ever, a second goal, alluded to above in the discussion of
population, "seeks to establish the consequences of the
presence or absence of particular characteristics of sus-
tenance organization within the ecological complex or

null

'ecosystem. The traditional primacy of the first goal,

 

9Hawley (1967), 330.

10Gibbs and Martin, 33; David F. Sly, "Migration and
the Ecological Complex," American Sociological Review, 37
(October, 1976), 617.

11Gibbs and Martin, 33.

 

6

as Hawley suggests, may be due largely to the convenience
in proceeding from the more operationally well-defined such
as population to the less well-defined such as organization.
However, as he points out in regard to population and
organization, for theoretical reasons "population is for
many purposes better regarded as the dependent variable,
delimited and regulated by organization."12

Despite the formulation of a cogent set of vari-
ables, human ecology remains both an heuristic device in
which the precise relationships among the components
remain unknown and basically an approach to urban systems
rather than to general social systems. Perhaps the major
hindrance to the development of a more formal theoretical
statement has been the recognition by ecologists that no
one variable can stand alone but must be considered in
relationship to the other three. However, to adhere to
this permise in conducting research entails the delineation
and measurement of four extremely inclusive variables.
For example, to include population in its totality neces-
sitates consideration of size, age-sex structure, in-,
out- and net-migration, fertility, and mortality. An
alternative strategy is to break down the complex as a
whole and each of the four variables into smaller component
parts, examine the relationships, and then put the complex
together again later in a more precise theoretical formula-

tion.

 

12Haw1ey (1967), 330.

7

In addition to the scarcity of research explicitly
examining the precise relationships among the components
of the ecological model, the focus of ecologists continues
to center on the city and its tributary area as the proto-
type of community or sustenance organization, even though
this focus of early ecologists was largely fortuitous.13
This continued emphasis may be due in part to the use of
the term community with its normative connotations and
equation with town or city rather than the more neutral
terms organization and/or social system. Also, the avail-
ability of data provided by the U.S. Census on cities,
SMA's, and later SMSA's has perhaps been instrumental in
sustaining such an emphasis. Because the delineation of
SMSA's includes a criterion based on sustenance dependence
of the population of surrounding counties on a particular
city of 50,000 or more, the problem of differentiating
between the ecological unit, defined by Hawley as "that
population which carries on its daily life through a given
system of relationships,"14 and a governmental unit for
which data are available is resolved to some extent.
Such a focus has resulted in a tendency to view social
systems as central place systems without recognizing the
ecological "situation" of nonmetropolitan populations

other than in relationship to the nearest city. The
ecological perspective itself, however, does not neces-

sitate such an emphasis.

 

13Hawley (1967), 331.
14Ibid., 33.

8

Although little research exists testing explicitly
the relationships among the four ecological components in
nonmetropolitan areas pg; s3, demographers have put forth
a number of generalizations concerning pOpulation and
sustenance organization in these areas.15 In general, loss
of population has been attributed to the increasing
mechanization of agriculture and the accompanying decline
of agricultural employment coupled with the inability of
nonmetropolitan areas to provide facilities attractive
to nonagricultural economic concerns. Such circumstances
have caused the young and educated to migrate out of rural
areas. This in turn has produced an age structure which
led to approximately 345 nonmetropolitan counties experi-
encing more deaths than births in 1967. On the other hand,
some nonmetropolitan areas have reversed this trend and
are both growing in population and attracting more migrants
than they are losing. These developments have been attrib-
uted to the ability of the population of such areas to
diversify by becoming commuter towns, retirement communities,
college or university towns, or by developing specialized
shopping facilities. In other cases, the existence of an

interstate highway seems to explain the divergence from the

 

15Two sources providing excellent summaries of these
generalizations are U.S. Department of Agriculture, Economic
Research Service, Economic Development Division, The Eco-
nomic 229 Social Condition 9: Rural America in thE—I97UTS,
(Washington, I971), Ch. 1; and the CommisSiOfiTon Population
Growth and the American Future, Population and the American
Future (New York, 1972), 30-33.

 

 

 

 

 

9
general pattern of population loss for both counties and
towns. Excluding the effects of age-structure, a popula-
tion variable, these generalizations suggest that both
organization and environment are important determinants
of population change. Specifically, the ability of a
population to reorganize its sustenance organization as
agricultural employment declines and the existence of
environmental factors whether natural such as climatic
factors amenable to retirement centers or manmade such
as interstates help to explain different rates of population
change and migration in these areas.

These demographic generalizations indicate that the
ecological complex may be able to provide a framework for
explaining more fully differential migration rates in non-
metropolitan areas. In addition, these findings as well
as the results of other specific studies discussed below
indicate that the study of nonmetropolitan areas can serve
to test explicitly the relationships posited within the
ecological complex in order to develop that useful heuristic

device into a more precise model.
PURPOSE OF THE STUDY

The purpose of this study is to test the relation-
ship between net-migration and both sustenance organization
and environment in nonmetropolitan counties in the North
Central Region during the 1960-70 decade. The analysis will

focus on several variables that have been measured in various

10

ways in studies utilizing nonmetropolitan counties, towns,
and State Economic Areas over several time periods. Other
factors which have been hypothesized as being of increasing
importance in explaining net-migration and can serve as
indices of organization, environment, and changes in each
will also be included. Furthermore, although employment
figures are the basis of several measures, the study also
incorporates several measures not based on employment in
determining manufacturing and agricultural specialization
and in dealing with recreation and governmental expenditures.

Testing these variables systematically over one
decade will help to clarify the importance of each variable
relative to other factors used as indicators of components
of the ecological complex. In turn, this will enhance our
understanding of why some nonmetropolitan counties are
attracting migrants while the vast majority are continuing
to lose population through net-migration. Moreover, the
results of such a study will contribute to defining more
precisely how two components of the ecological complex

affect one means through which population size changes.
THEORY AND LITERATURE

From a theoretical standpoint, the relationship
between migration and environmental resources is simply
the question of how large a population can the existing
environment sustain. If the population is too large for

available resources, then the population contracts through

ll
migration, decreased fertility, and increased mortality.
As pointed out earlier, however, out-migration is more
effective than either fertility or mortality in altering
population size in the short run. On the other hand, where
the environment can support a larger population than it is
sustaining, the population will tend to increase through
the acquisition of individuals migrating from areas unable
to support their p0pu1ations.

The relationship between environment and migration,
however, is more complex when other factors are considered.
Hawley has developed a model of this relationship that also
incorporates organization. According to his model, the
organizational process can be viewed in four stages begin-
ning with the competition of individuals or other units
with similar demands on a scarce resource supply so that
what one competitor gets necessarily decreases accordingly
the amount others can obtain. In the second stage both the
singularity of the supply and environmental factors impose
standard competitive conditions which lead to increasing
homogeneity among the competitors while in the third stage
congestion operates to eliminate the weakest competitors.
Finally, in the fourth stage either territorial or struc-
tural differentiation (or a combination of the two) appears
‘with migration providing the mechanism leading to the former

rather than the latter.16 That is, at the point at which

some members win and others lose in the competition over

 

16Hawley (1950), pp. 201-3.

 

Q
put

4 I
n. nu

q, v‘.
. V!
'H.‘

u.

.
‘no. u

m...‘

 

h“ . .
I“. ~

in”. l
a"): ‘

 

' Q
I I

h‘ ‘L
I -‘.
‘. .h

12
scarce (valuable) resources, the "deposed competitors"
have the option of either migrating into new territories
leading to territorial differentiation or remaining and
developing new skills in order to make oblique attacks
on the scarce resource supply leading to greater structural
differentiation. In addition, Hawley posits that the
characteristics of both the population and the environment
in which individuals compete have a direct although second-
ary influence of their own in determining which differ-
entiating process will predominate. On the other hand,
technology is considered only indirectly when he discusses
the development of new skills by individuals. Thus, in
this model the resolution of competition is the primary
Causal factor producing either structural or territorial
differentiation while environmental and population
Characteristics have secondary influences.”

As stated, the model offers little theoretical
insight into how organization and migration are directly
T:elated. This may be because it starts at a point at which
Either a new resource is discovered that totally dismantles
the existing sustenance structure or an undifferentiated

population enters a territory in order to exploit it for

k

17Both Jane Jacobs, The Death and Life of Great
American Cities, (New York, 1961), ch._I_3'; and Harriet
an erry King, "New Town, Mon Amour," The Chicagoan,
November, 1973, 78-83, provide illustraEions of this process
in Greenwich Village, New York, and in Chicago entertainment
areas respectively although both works emphasize the possi-
ble destructive elements of the process to a community
rather than the ecological processes involved.

 

 

13
the first time. In an ongoing system, however, competition
occurs within the context of an existing structure.

Hawley does suggest implicitly, however, that
organization has a direct effect on migration. That is,
the resolution of competition implies that at least a
crude form of organization exists at this point with the
population divided into two parts, successful and unsuc-
cessful sustenance-getters. Within the context of an
existing structure, moreover, it seems plausible to modify
the model by prOposing that if the new resource competi-
tion involves diversification of organization, more members
of the population can gain sufficient sustenance than in
a. situation where those receiving sufficient sustenance
increase their sustenance level while members receiving
little or no livelihood before remain in that position.
The research cited earlier on population change and migra-
tion in nonmetroplitan areas, moreover, supports the ,
Proposition that a more diversified sustenance organization
can support a larger population in a given territory than
can a highly specialized structure.

This relationship can perhaps be better stated
through an illustration involving specific territorial
units. If a state, for example, is divided into specialized
Units such as counties with each county specializing in a

different activity and consequentially each county being
structurally undifferentiated, then persons in any given

county not possessing skills for that county's specialty

14
will either have to develop such skills or migrate else-
where. The model further suggests that migration will be
greater in this situation than in the opposite polar case
where all activities are evenly distributed throughout all
the counties, i.e., structural differentiation is equal
among all counties. This is because in the territorially
differentiated state, persons will be more likely not to
reside initially in a county in which they can utilize
their skills whereas in an undifferentiated state where
all sustenance activities are equally distributed, i.e.,
each county population is maximally differentiated structur-
ally within the limits of the state structure, everybody
will reside initially at a point where he can potentially
utilize his skills in attacking the resouce supply. Further-
more, assuming an equal level of resources, this also
implies that in a situation in which different types of
t:erritorial units exist, some structurally differentiated
and some structurally undifferentiated, those that are
undifferentiated will lose peOple through migration and
those that are structurally differentiated will gain
p0pu1ation through migration since the latter have a wider
range of sustenance niches available and hence can both
attract and retain a wider range of skills.
This study focusses on defining more precisely the

causal relationships that Hawley's modified migration model
posits between organization and environment on one hand and

migration on the other. Specifically, it is hypothesized

15
that the greater the organizational diversification and the
higher the level of environmental resources, the more
positive will be the net-migration rate. Moreover, it is
posited that the influence of organization will be stronger
than that of environment.

The study also examines the relationship between
change in the two independent variables and migration.
Hawley neither specifies nor implies how changes in these
two components of the ecological complex relate to popula-
tion change through migration. Given the suggested
relationships between environment and organization at one
point in time followed by migration, however, it is plausi-
ble to assume that changes in these two variables will relate
tzo net-migration in similar ways. Thus, it is hypothesized
that the greater the changes in organization indicating
diversification and secondarily the greater the changes
in environment indicating increasing resource availability
to more members of a population, the more positive the net-
migration rate will be.

Both ecologists and demographers have taken some
Steps toward examining the relationships among the compo-
nents of the ecological complex in nonmetropolitan areas

although only a few have tested these relationships
explicitly. Among these studies are several utilizing
employment figures which support evidence presented
earlier that the major cause of differential migration

rates among rural areas is related directly to both the

16

decrease of opportunities in farming and the ability of
populations to diversify away from agriculture.18 Further-
‘more, they suggest that diversification involving the
acquisition of manufacturing concerns is an important factor
in explaining net-migration differentials. In general,
these studies taken together indicate that a nonmetropoli-
tan area that continues to rely heavily only on agriculture
and fails to augment this with manufacturing may be regarded
as a structurally undifferentiated area with a strong
negative net-migration rate. Conversely, an agricultural
population that includes some manufacturing and is increas-
ingly developing a more differentiated structure through
the acquisition of manufacturing concerns within its sus-
tenance structure loses less and/or gains p0pu1ation through
net-migration. On the other hand, only one of three studies
that also test standard of living indicates that this
factor plays a part in explaining net-migration.

Specifically, Bogue's study of correlates of net-
migration in nonmetropolitan economic regions from 1940 to
1950 indicates that net-migration correlates positively
with size of manufacturing labor force and negatively with

size of agricultural labor force in 1950. In addition, he

 

18Although a p0pu1ation in a given territory may
be undifferentiated structurally in regard to any sustenance
activity, nonmetropolitan p0pu1ations in the United States
tend to be involved in primary or extractive sustenance
activities, particularly agriculture. The degree to which
the sample utilized in this study conforms to this pattern
will be discussed in Chapter II. Also see Otis Dudley Duncan
and Albert J. Reiss, Jr., Social Characteristics of Urban
and Rural Communities, 1950 (New YorE} 1956), p. 2T5.

 

17

finds that the farm operator level-of—living index for 1945
correlates positively with net-migration. It should be
noted, however, that unless one assumes that the size of
both the agricultural and manufacturing labor forces remained
in relative proportion to each other, Bogue may be measuring
the organizational result of migration rather than its
cause; that is, the 1950 labor force size followed the 1940-
50 period in which migration is measured. Thus, care must
be taken in imputing causality. Furthermore, it should also
be noted that all three correlations vary from moderate to
strong within different regions although the direction of
all relationships are comparable between areas. 19

Levitan and Houghteling's study of the slower growth
rate of Missouri compared with the nation as a whole
suggests an explanation for Bogue's negative correlation
between net-migration and agricultural employment. They
find that the best explanation for this phenomenon rests
on the agricultural nature of the state. That is, slower
growth is due to the exodus of farmers primarily caused by
increased farm productivity, consolidation of farms, and a
corresponding higher birth rate in rural areas which has

forced migration of "excess farmers and/or farmers'
children to urban areas in search of employment both within

and outside the state.20

 

19Bogue, pp. 26-27.

20Sar A. Levitan and Louis D. Houghteling, Factors
Affecting the Slower Growth g£_Missouri Population Compared
with the United—States, rev. ed., (Washington, 1961).

 

 

0‘

 
 
 
 
 

m

18

Two other studies add additional support to the
findings concerning employment but reveal that standard
of living may not be as strongly associated with net-migra-
tion as Bogue's results denote. In a study of migration
utilizing State Economic Areas in West Virginia, Rutman's
results indicate that population inflows are dependent on
economic opportunities available in general in the area of
destination in the 19508. However, he finds no statistically
significant relationship between migration and any of five
indicators of well-being based on percentage above or below
various income levels.21

Beegle, Marshall, and Rice have categorized non-
metropolitan counties in the North Central Region and
Kentucky on the basis of three variables, in- or out-
migration, high or low manufacturing, and high or low
standard of living based on the farm operator level-of-
living index over the 1940-50 and 1950-60 decades. On
this basis they find three prevailing patterns. The first
includes counties characterized by in-migration, high
standard of living and high proportions in manufacturing;
the other two patterns represent counties with out-migration,
low proportions in manufacturing, and either high or low
farm operator level of living. This suggests that a strong
and positive association exists between net-migration and

manufacturing employment, although as in the Bogue study

 

21Gilbert L. Rutman, "Migration and Economic
Opportunities in West Virginia: A Statistical Analysis,"
Rural Sociology, 35 (June, 1970), 206-17.

19

manufacturing percentages are based on end-of-decade data.
On the other hand, the results also suggest that the well-
being of an agricultural population has little influence
on migration. Because this study also includes the 1950s
while Bogue's encompasses only the 19403, the discrepancy
in results may be due to the use of several factors in
the farm level-of-living index which may no longer be
useful measures of well—being. That is, in 1959 the index
included percentage of farms with telephones, with freezers,
and with automobiles in addition to two items dealing with
average value of land and buildings and average value of
sales.22

Comparing the variables used in these four studies
‘with reference to the ecological model, Rutman deals
primarily with measures of sustenance level. On the
other hand, Bogue and Beegle, et.§l., have added indices
of organizational diversity or specialization based on
percentage employed in manufacturing and/or agriculture
while Levitan and Houghteling are only concerned with
diversity and specialization. Furthermore, although
various functional classifications of cities or central
places have been devised that consider the entire occupa-

tional structure, the vast majority of studies of

 

22Allan Beegle, Douglas Marshall, Rodger Rice,
"Selected Factors Related to County Migration Patterns in
the North Central States 1940-50 and 1950-60," Michigan
State University Agricultural Experiment Station Quarterly
Bulletin, 46 (November, 1963), 1-40.

 

20

nonmetropolitan migration, like the above studies, incor-
porate unidimensional indices of organization.23

There are two exceptions to this generalization.
Both Groth and Sly explicitly utilize measures of organiza-
tion embracing the entire occupational structure in migra-
tion studies based on county units of analysis. Groth has
developed a functional classification in which those counties
ranking in the top ten per cent in employment in any one
of six industry groups or having over ten per cent employed
in any one of three other ”low employment" categories are
designated as functionally specialized. He has further
dichotomized functional types into rural or urban counties
based on population size, nonagricultural labor force size,
or per cent commuting to urban centers and has compared the
resulting types with net-migration rates 1960-70 in counties
throughout the 48 contiguous States. In comparing rural
and urban subtypes where greater out-migration than in-
migration occurs, the loss is more severe for rural sub-
types; in other cases, controlling for functional type

reveals that rural subtypes show a net loss while their

urban counterparts show a net gain. Finally, his results

 

23For discussions and critiques of several of these
schema see Ralph Thomlinson, Urban Structure: The Social
and Spatial Character of Cities (New YOrk, I969YT 66-8;
AIbert J. Reiss, Jr., "Functional Specialization of Cities,"
Cities and Spgiety: The Revised Reader in Urban Sociolo ,
ed. by Paul K. Hatt and Albert J. ReissT_Jr. (New York,
1957), 555-75; Otis Dudley Duncan, et. al., Metropolis
and Re ion (Baltimore, 1960); Philip—G.—Groth: "Functional
CIass1 cations of Counties: Some Applications," Department
of Rural Sociology, University of Wisconsin, Madison,
May 26, 1972, 1-3.

 

 

 

 

 

21

show less variation in net-migration in rural subtypes than
in similar urban counties. He concludes from this that
"rurality pg£_§g exerts a stronger influence on net-migra-
tion than does functional specialization."24

Sly, in an explicit test of the relative importance
of the three nonpopulation components of the ecological
complex in relation to migration, finds that both occupa-
tional diversification and agricultural stability directly
affect the out-migration rate of the black male population
of Cotton Belt counties over the 1940-50 and 1950-60
decades. The first of the two measures, the index of
occupational dispersion, is based on the difference between
actual percentage of blacks in each occupational category
and the expected percentage assuming all workers have
equal access to all occupations and hence would be equally
distributed among all categories. The stability of the
agricultural structure over a decade is measured through
the development of a weighted index based on four agri-
cultural occupations in which the least stable occupation
is weighted most heavily and the most stable is weighted
least. On the basis of the ecological model, Sly hypothe-
sizes that the higher the former index indicating greater
diversity the lower the out-migration rate will be whereas
the higher the latter measure indicating greater occupa-
tional instability the greater the out-migration rate will

be.25

 

24Groth, 22.
25Sly, passim.

22

His results confirm the model with the index of
occupational dispersion correlating more strongly than the
agricultural index over both decades. Furthermore, path
analysis indicates that in general both organizational
factors have a direct relationship with migration while
technological and environmental factors affect migration
only indirectly through organization. There are two excep-
tions to this. The first is a technological variable, gas
consumption, which ranks between the two organizational
indices in the 19403 in direct influence, and the second
is white-nonwhite acreage ratio, an environmental factor,
which is the most important direct influence on migration
in the 19503.

Sly's results also reveal that the relationship
between both organizational indices and migration are
weaker in the 19503 than in the previous decade. He sug-
gests that the lessening influence of organization could
be due to the effectiveness of migration in the 19403 or
to greater discontent unrelated to the ecosystem among
Southern blacks in the 19503. Thirdly, given the pre-
dominance of the acreage ratio variable in the latter
decade, Sly suggests that possibly the 19503 witnessed a
reorganization of agriculture within these counties
accompanied by a lack of opportunity for blacks outside
of this sector.

It is difficult both to compare and to reconcile

the findings of the Groth and Sly studies unless one

23
concludes that organization is simply decreasing in influ-
ence on migration and hence population change. However,
the noncomparability of the populations under study (total
versus black male residents of county) as well as Sly's
third possible explanation for apparent declining organiza-
tional influence cautions against this. Furthermore, it
is not known how dispersed the Cotton Belt counties would
be throughout the Groth classification. If the Cotton
Belt counties fall into different functional categories
based on Groth's schema, it would suggest that the
organizational index used determines to some extent the
relationship found between organization and migration.
On the other hand, if the southern counties are all
specialized in one of Groth's categories, it would suggest
that Sly's two indices, and in particular his index of
occupational dispersion, which is meant to measure the
same dimension as Groth‘s index, may in fact measure a
different dimension of organization. Groth acknowledges
that it is only through the exploration of alternative
modes of classification that it can be discovered which
is the most fruitful measure of sustenance organization.26
The results of these two studies indicate that it is also
only through the exploration of alternative modes of
classification that the precise effects of sustenance

organization on migration can be determined.

 

26Groth, l9.

24

In addition to research based on occupational
indices, several ecological studies of nonmetropolitan
areas utilize other indicators of structural diversity.
A major focus of these studies is urban size as either
cause or effect of population change or migration. A3
Lampard points out, urbanization can be regarded as the
movement of people out of agricultural and into non-
agricultural occupational pursuits and generally larger
communities; moreover, this perspective "gives primary
recognition to the differential ordering of occupations
or industries within a given territorial space."27
Moreover, economic geographers have long noted a relation-
ship between size of urban place and function with larger
places providing more specialized services than smaller
places.28 From an ecological standpoint, it follows
that because larger urban places provide more services
they also provide a wider variety of occupational niches.
Thus, degree of urbanization and in particular size of

largest urban place can serve as a measure of occupational

diversity.

 

27Eric E. Lampard, "Historical Aspects of Urbaniza-
tion," The Stud of Urbanization, ed. by Philip M. Hauser
and Leo F. c nore (NewlYork,71965), 520.

 

28See in particular Brian J. L. Berry and Allen
Pred, Central Place Studies: A Biblio ra h of Theor and
Applications (Philadelphia, 196 ; an A en K. riEk,
"Principles of Areal Functional Organization in Regional
ggga§3geography," Economic Geography, 33 (October, 1957)

 

 

 

25

Two of the studies dealing with urban places also
incorporate other measures of structural diversity relating
to the existence of college, military, and/or other insti-
tutional facilities which can be viewed in two ways in
reference to migration. The existence of these institutions
draws migrants into an area while from an ecological
perspective they are also indices of structural diversity.
That is, they provide additional occupational niches for
the p0pu1ation both within the facilities themselves as
well as in supporting services that may arise due to
their location in a particular area, e.g., restaurants to
accommodate visitors to those within these three types of
facilities.

In an examination of nonmetropolitan counties
between 1960 and 1970 utilizing five organizational and
one environmental variable, Irwin finds that all measures
correlate positively with change in population size with
existence of a college being the most important factor
(r=.209). However, all correlations are low and the
multiple correlation coefficient for all six variables
is only .3087. The other four organizational variables
include military, large city, small city, and institution
with all variables set up as binary variables, i.e.,
existence or absence based upon various size criteria.
However, the study deals with population change rather than
migration. More importantly, as Irwin suggests, the use of

continuous variables might yield somewhat different results

26

than his binary variables and might clarify those relation-
ships he has found. For example, it seems likely that a
college of 1,000 students would generate fewer supporting
services both in terms of variety and quantity than would
a university of 40,000 students.29

In addition to the five organizational factors,
Irwin's results show that the existence of a freeway,
an environmental factor, is the second most important
variable explaining population change during the decade
(r=.l99). There are a number of possible explanations
for this phenomenon. Such access may induce manufacturing
concerns to locate near such interchanges, lead to the
creation of subsidiary road services for travelers, and/or
signify access to SMSA's or at least to a larger territory,
all of which would presumably promote greater structural
differentiation. Thus, this result in addition to Sly's
finding that the strongest direct influence on migration
in the 19503 was an environmental factor which may be
linked to organizational changes suggests that the rela-
tionship between environment and migration needs to be
explored more thoroughly relative to organization.

In a study of in-migration (rather than net-
migration) focussing on nonmetropolitan urban places,

Zuiches finds that both college and military activity

 

29Richard Irwin, "Nonmetropolitan Population Change:
1960-70," paper prepared for presentation at the annual
meeting of the Population Association of America, Washington,
D.C., April 23, 1971. ‘

27
are important explanatory factors, the former to intrastate
and the latter to interstate in-migration to urban places.
His analysis also indicates that those places farthest from
metropolitan complexes are experiencing higher levels of in-
migration than other urban places controlling for all other
variables. He concludes that these results suggest that the
remote urban places may act as central places in their own
right in dominating a rural hinterland.30
I Zuiches' explanation of the relationship between

remoteness and population has also been examined by
Burford, Lemon, and Fuguitt. Before discussing these
studies, however, it should be noted that the concept
of central place has two implications pertaining to
population. The first is that the more important the place
is as a center the larger it will be. Secondly, a central
place will be more diversified occupationally due to the
greater variety of services it performs for its hinterland
than will be a noncentral place of similar size at a given
point in time. Thus, if the ecological model is valid,
this diversity should induce greater growth through migra-
tion to the central place than to the comparable noncentral
place town.

Testing the relationship between county level net-

migration to urban centers and a remoteness index based

 

30James J. Zuiches, "In-Migration and Growth of
Nonmetropolitan Urban Places," Rural Sociology, 35
(September, 1970), 410-20.

28

in part on cities with a population as low as 10,000 for
the 1930-40 decade, Burford obtains a small but significant
parabolic correlation indicating that net-migration is
lowest in those areas both closest to and farthest from
large cities. In an analysis of the results, he suggests
that in those areas closest to cities, members can migrate
occupationally without necessarily migrating spatially.
Also, the tendency for places farthest from cities to have
lower net-migration rates than intermediate counties can
be attributed to a remoteness so great from large centers
that both in- and out-migration between the area and large
centers is discouraged. These results tend to confirm
partially Zuiches' suggested explanation of remoteness.
That is, Burford's findings suggest that isolation promotes
the fuller develOpment of urban places remote from.cities
as regional centers with greater differentiation than
urban places located in intermediate counties.31

Lemon's study of urbanization in southeastern
Pennsylvania in the eighteenth century indicates that
p0pu1ation size and central place functions are directly
related to each other and to remoteness from already exist-
ing central places. Through classifying towns by size and

function and comparing the actual distribution of towns

with that theoretically expected utilizing central place

theory, Lemon concludes that the "strong" fourth order

 

31Roger L. Burford, "An Index of Distance as
Related to Internal Migration," Southern Economic Journal,
29 (October, 1962), 77-81.

29

county seats in the backcountry which functioned as
commercial centers as well as political centers prevented
the expansion of other places. Similarly, the primacy of
Philadelphia prevented the growth of county seats near it
to fourth order central places as in the backcountry and
in general inhibited town growth in places within a 30-mile
radius of this fifth order metropolis.32

In a study exploring the relationship between
county seat status and growth Fuguitt suggests that although
federal and State governments have a greater influence on
local affairs today, it has been done generally through
the county,"so that county governmental functions have
been strengthened and augmented."33 In organizational
terms this strengthening and augmenting of functions also
suggests an increase and diversification of job opportuni-
ties and hence occupational niches within the structure.
In a test of the relationship between county seat status
and population growth or decline in the North Central
Region plus New Jersey, New York, and Pennsylvania, he
finds that such status is positively related to growth
of towns in counties more than fifty miles from SMSA

Central Cities.

 

32James T. Lemon, "Urbanization and the Development
of Eighteenth-Century Southeastern Pennsylvania and
Adjacent Delaware,” The William and Mary Quarterly, 3d.
Ser., XXIV (October,-I967), 562-337

33Glenn V. Fuguitt, "County Seat Status as a Factor
in Small Town Growth and Decline," Social Forces, 44
(December, 1965), 246.

 

 

 
 
 
 
 
 
 
 
 

30

Although the studies discussed in this survey vary
as to both the mechanism of population size change consid-
ered and the unit of analysis employed, several conclusions
can be drawn. In regard to environment, three of the
studies indicate that there is a relationship between
environment and population other than the effect of remote-
ness from metropolises leading to diversification.
Specifically, environmental expansion (interstates) is
associated with population expansion while environmental
restriction (white-nonwhite acreage ratio) correlates
with population decline through migration. In addition,
Fuguitt's results comparing towns with and without county
seat status suggest that another environmental factor,
the impact of differential outside influences from other
governmental units on counties, needs to be explored
further.

Utilizing different measures of structural diversity,
these studies also indicate that, despite Groth's conclusion
regarding the importance of rurality, counties more remote
from cities or SMSA's may possess towns that serve as
regional centers and consequently either retain more of
their populations and/or attract more migrants than less
remote towns. Furthermore, both empirical research and
theoretical work by economic geographers and ecologists
suggest that the size of urban places is related directly
and positively to diversity of function and hence organiza-

tional diversification. Finally, all of these studies

31
reveal that structural diversification away from agriculture
in nonmetropolitan areas is positively correlated with

both migration and population change.

CHAPTER II
METHODOLOGY

Employing ecological theory, this study will examine
the relationship between both organization and environment
and the net-migration rate. More specifically, various
methods of correlational analysis will be used in order to
measure the effects of specific variables (serving as indi-
cators of organization and environment) in explaining dif-
ferential migration rates in nonmetropolitan counties.
Moreover, the study will also focus on exploring the
relationships among various indices of organization and
environment that have been either explicitly or implicitly
suggested to be important factors contributing to the net-

migration rate.

Unit of Analysis

The study includes 227 counties, a one-quarter
sample of all nonmetropolitan counties in the North Central
Division as of 1970. Although a particular governmental
unit such as a county does not necessarily constitute an
ecological unit from an organizational standpoint, two
methodological considerations favor the use of governmental
units. As Gibbs and Martin point out, of major importance

32

nu.-

 
 

 

 
 

..:\

33
is the availability of data which is generally compiled by

governmental units.1

Furthermore, the use of such terri-
tories allows for easy comparability over time. That is,
the use of ecological units such as communities would entail
redefining population boundaries as neighboring populations
become integrated into one structure or part of a p0pu1ation
appears to break off into a new sustenance structure. The
use of governmental units with relatively stable boundaries
avoids this problem.

In addition, research indicates that the county
unit is not only a methodological convenience but also
serves as a basis for ecological organization. Fuguitt
has noted the strengthening of county level government as
a liason between federal and State agencies and the local
population. Brown, in an analysis of the political areal-
functional hierarchy in Minnesota also indicates two sub-
functions of counties themselves that affect sustenance
structure directly. The collection and disbursement of
tax money, he notes, can affect industry location, and
the county's power to create or dissolve school districts
can affect the sustenance structure in terms of both occupa-
tional niches and the training of prospective labor force

entrant S . 2

 

1Gibbs and Martin, 32.

2Fuguitt, p. 246; and Robert Harold Brown, Political
Areal Functional Organization: with special reference pg
Sp. Cloud, Minnesota (Chicago, I957), p. 110.

 

 

34

Lyford's study of Vandalia, Illinois, supports even
more strongly the idea that counties serve as ecological
units as well as political units. Specifically, he maintains
that there are strong ties between the town and farmers in
the rest of Fayette county. Regarding the farmer's decline,
he asserts that "Vandalia would suffer without its factories--
their 1033 would be a fearful blow to the town's hopes for
the future-~but it could not survive without its farmers.
As Dr. Josh Weiner puts it, 'the job of people in town is
to supply the farmer all the services he needs."'3 This
suggests that there are strong sustenance ties between rural
towns and the surrounding farm population beyond the

political ties of county government.

Description of the Sample

 

Demographically, the 227 counties include 179 with
negative net-migration rates in the 19603, 46 with positive
net-migration rates, and two that neither gained nor lost
population through net-migration during the decade. Compar-
ing counties over a two decade period indicates that 171
have followed the traditional pattern of losses through
net-migration in both decades; furthermore, for 55 of these
counties the percentage of negative net-migration increased
in the latter decade. On the other hand, 12 counties gained

through net-migration in both decades, 34 switched for losing

 

3Joseph P. Lyford, The Talk ip Vandalia: The Life
of 32 American Town (New York 1965), p. 12.

 

.

Isy~ A.

'v

ha.5.
-

 

-
"a .s

.
“«.‘

I.i.:

Q
C‘-

‘5']

 

 

 

(’3

 

s
§
.
u
.
‘J 5‘
. .
§
1,I
t
‘1 ’\
O‘ -
‘5-
s
l
5'}
.‘-‘
s I
. P
‘h

35
to gaining counties, and eight recorded negative net-migration
rates after gaining in the 19503. Finally, the two counties
that neither gained nor lost population in the 19603 through
migration had negative net-migration ratestfluaprevious decade.

Occupationally, the sample reflects the traditional
agricultural nature of the rural United States based on a
14 industry classification of occupations. Specifically,
agriculture ranked first in number employed in 165 counties
in 1960 while retailing ranked first in 26 counties, durable
manufacturing in 22, nondurable manufacturing in seven,
professional services in four, mining in two and contruction
in one. Transportation and communication, wholesale trade,
finance and insurance, business and repair services, personal
services, entertainment, and public administration did not
predominate in any counties. Agriculture dominated even
more in 1950 ranking first in employment in 202 counties;
durable manufacturing ranked first in ten counties, retail-
ing in eight, nondurable manufacturing in four, mining in
two and transportation and communication in one.

A comparison of the percentage of agricultural
employees in each county to that expected if all industries
of occupation were distributed evenly throughout the United
States further indicates that agriculture predominates.

Using this measure, 220 counties in both 1950 and 1960 had
more than the expected number of workers employed in agri-
culture; moreover, in both decades workers in two of the

seven remaining counties were principally engaged in mining.

F
In
said .

.‘I-A
A

In“.

'0‘}.
stub

I
«up. A
.1.
u

 

'o‘.‘
k..‘
v

I
0'...
«.g.

 

o'-

 

6
a
N d
‘

 

 

 

36
Thus, as measured by these indicators, the sample has been
primarily agricultural in nature although by 1960 27.3 per
cent of the counties had diversified to the extent that
agriculture did not rank first in primary industry of

employment.

Dependent Variable
The dependent variable in this study is the net-
migration rate computed using the residual method:
NM = PZ-P1+B-D
P1

 

Rates rather than absolute numbers are used in computing
net-migration as well as the majority of independent
variables because of the vast differences in size of the
base population among counties. For example, it would be
impossible for Arthur County, Nebraska, with 680 inhabitants
in 1960 to lose the 6,000 residents that Kankakee County,
Illinois, lost through net-migration during the decade.

If all variables were based on employment, these differences
would in effect be accounted for in the statistical analysis
since both net-migration and all independent variables

would reflect the limitations of various population sizes

in the counties. However, the inclusion of several non-
population based variables necessitates that rates rather
than absolute numbers be utilized in order to establish

a meaningful basis for comparison.

 

37

Independent Variables

 

The study utilizes eighteen measures of organiza-
tion, nine measures of changing organization, six environ-
mental variables and one variable measuring environmental
change. Each variable is stated below (the title in
parentheses is how it will appear in tables) along with
comments explaining either the measure itself and/or its
relevance to the study. Unless stated otherwise, it is
predicted that each variable will correlate positively
with net-migration. All predicted directions of associa-
tion are based on what should occur if the ecological model

is valid.

Measures of Organization

 

Because of the predominance of agriculture in 1960
as well as historically as indicated by 1950 data, all
structural measures pertaining to nonagricultural sustenance
are assumed to be indicators of diversity. For example,

a high percentage employed in manufacturing is assumed to
be an indication of greater structural diversity than a
low percentage similarly employed.

In addition, although it seems reasonable to assert
causation between independent variables measured in 1960
and the 1960-70 net-migration rate, caution must be exerted
in inferring causal relationships between mid-decade measures
and migration. However, because each variable is measured
for all counties at the same point in time, the data are

consistent among counties. Thus, relationships found

 

 

38

between these factors and net-migration can suggest further
areas of research in order to find more accurate measures
to clarify ecological relationships.

The specific variables which measure diversification,
size, or sustenance level and Opportunity are as follows:

1. Percentage employed in manufacturing, 1960
(Manufacturing Employment).

2. Degree of diversity, 1960 (Diversity).
This variable is measured by subtracting the Index of
Dissimilarity from one. The Index of Dissimilarity utilizes
the U.S. labor force structure categorized into the 14
industries employed earlier in this chapter to determine
primary industry of occupation. The index figure is the
percentage of workers who would have to transfer to other
industry groups for the county structure to duplicate the
national structure. It is determined by subtracting actual
rmmber of workers in a category from expected number based
on the U.S. structure, summing all positive differences,
and then dividing by the total number of labor force
participants in the county.4 The result measures specializa-
tion, while subtracting the index figure from one indicates
the extent of diversification of the structure within the
limits imposed by the national structure. For example,
if a country had an occupational profile such that 50 per

cent of the workers were engaged in agriculture and 50 per

 

aDuncan, 35. al., Metropolis and Region, 209-11.

39
cent in retail trade and a particular county's profile was
60-40 per cent respectively, 10 per cent of the county's
workers would have to switch occupations to duplicate the
national structure (Index of Dissimilarity) while 90 per
cent could remain in their present occupational niches
(degree of diversity).

3. Number of categories (21 possible types of
manufacturing production) in which at least one manufactur-
ing concern exists, 1967 (Manufacturing Categories).

This is a crude measure of diversity within the manufacturing
sector.

4. Number of manufacturing firms with at least 20
employees, 1967 (Manufacturing Firms).

This variable serves as an indicator of both size of the
manufacturing sector and diversity within it in terms of
number of facilities which can possibly offer employment.

5. Percentage of farm operator family income from
other employment, 1964 (Other Farm Income).

A negative association has been posited for this analysis,
although until it is tested the ecological model offers
arguments for correlation in either direction. The ability
to find work off the farm indicates structural differentia-
tion within a given territory. However, the low sustenance
level of farming in an area as indicated by the fact that
the farm operator's family needs other sources of income
and yet does not quit farming suggests that structural

differentiation is not great enough to induce an occupational

 

p
I
C

 

 

 

 

 

40
change although farming does not provide enough sustenance
for the family. Thus, a multiple income circumstance is
merely a step between full-time farm operator and migration.
6. Number of categories (eight possible types of
agricultural production) in which at least one farm exists,
1964 (Farm Categories).
This is a crude measure of diversity of farm land use
indicating differentiation within the agricultural sector.
7. Unemployment rate, 1960 (Unemployment).
This indicates the prOportion of the potential labor force
that cannot find employment and should correlate negatively
with net-migration.

8. Female participation rate in the labor force,
1960 (Female Participation).
A high female participation rate suggests the existence
of a more diversified structure in which a greater variety
of skills can be utilized.

9. Percentage of families with median income
under $3,000, 1960 (Income Under $3,000).

10. Percentage of public relief recipients, 1964.

Both this variable and the previous one should be negatively
correlated with net-migration. They are both indicators
of the sustenance structure's inability to accommodate
the population sufficiently.

11. Percentage employed in public administration
and education, 1960 (Public Administration).
Both this and the following variable test the effects of

the increasing role of governments as employers.

 

 

 

 

 

 

a
b
c

41

12. Percentage full-time equivalent employees in
local government, 1967 (Local Government Employment).

13. Number of hotels, tourist courts, motels,
trailer parks, camps, 1966 (Hotels).

14. Number of amusement and recreation services
excluding bowling alleys, billiard halls, movie theaters,
1966 (Amusement Places).

Both numbers 13 and 14 are indirect indices of the extent
to which a county serves as a resort or recreation area
and thus reflects employment Opportunities in tourist-
related businesses.

15. Percentage college poppulation, 1960 (College).

16. Percentage male military population, 1960
(Military).

17. Percentage institutionalized population, 1960
(Institutionalized).

Variables 15, 16 and 17 measure the impact of various insti—
tutionalized populations which have been posited as being
positively correlated with the net-migration rate.

18. Largest urban place, 1960 (Largest Town).
Size of the largest urban place is used as a measure of
diversification. As discussed in the previous chapter,
the positive correlation between size of place and
diversity of function implies a wider variety of
occupational niches leading to a more positive net-
migration rate for those counties with the more populous

largest urban places. Within this sample there are four

42
cases in which the largest urban place has a population
that resides in two counties; in such instances only that

part Of the population residing in the sample county is

included.

Measures of Organizational Change

 

With one exception, these variables designate changes
in various measures listed under organization. The nine
variables are as follows:

1. Change in percentage employed in manufacturing,
1960-70 (Change in Manufacturing Employment).

2. Change in land use, 1959-69 (Change in Land Use).
This is a binary variable indicating whether or not the major
type Of farm activity remains constant from the beginning‘
to the end Of the decade. Changing land use may mean a
change in skills needed which may in turn reflect itself
in the net-migration rate; that is, new farm Operators will
nfigrate into an area while former farm operators will be
more likely to look for new local jobs before migrating.5
The major problem in testing this relationship is that the
1959 data are for all farms and those for 1969 are only

fer farms with sales of $2,500 or more. This means that

 

5The results of studies by Isbell and by Bright
and‘Thomas suggest that individuals will migrate the least
distance possible in search of employment. See Eleanor
Collins Isbell, "Internal Migration in Sweden and Inter-
vening Opportunities," American Sociological Review, 9
(December, 1944), 627-39? and Margaret L. Bright and
Dorothy S. Thomas, "Interstate Migration and Intervening
Opportunities," American Sociological Review, 6 (December,
1941), 773-83.

1,
.4“-

" n
"-‘u--

r :E
LS

 

 

‘UO‘I‘
a

l‘u‘:
.

‘u‘

 

 

43
one must assume that smaller and part-time farms fall into
each category in similar proportion to larger farms.

3. Change in unemployment rate, 1960-70 (Change in
Unemployment).

Similar to the reasoning for the unemployment rate in 1960,
this variable should correlate negatively with net-migration.

4. Change in female participation rate, 1960-70
(Change in Female Participation).

The rationale for the posited positive correlation between
this variable and the net-migration rate as well as for the
remaining measures of organizational change and the dependent
variable corresponds to that given for each comparable static
measure of organization in the previous section.

5. Change in percentage employed in public adminis-
tration and education, 1960-70 (Change in Public
Administration).

6. Change in percentage college population, 1960-70
(Change in College).

7. Change in percentage male military population,
1960-70 (Change in Military).

8. Change in percentage institutionalized population,
1960-70 (Change in Institutionalized).

9. Change in percentage of p0pu1ation in largest
urban place, 1960-70 (Change in Largest Town).

In five instances the largest place in 1960 did not remain
the largest place in 1970. Because the focus of this study

is the relationship between size of urban place as an index

44
of diversification and net-migration rather than on urban

place per pg, the percentage change is computed using one

place in 1960 and the other in 1970.

Measures g Environment

 

Five of the six environmental variables are indicators
Of resources related to access either to employment or to
aid which.may affect the sustenance level of a population.
The sixth variable, interstates, as noted in the discussion
of Irwin's study, can be viewed as an environmental factor
which has implications for both local structural change
and access to cities or larger organizational complexes.
Finally, none of the variables is based on data for 1960
which mitigates against inferring causality. However, in
several cases it seems reasonable to posit net-migration
as the dependent variables despite time of measurement.
The reasons for this are discussed under the specific
variables. The six environmental variables are as follows:
1. Amount of federal funds spent per capita in
fiscal 1970 (Federal Outlays).
This variable as well as the two that follow pertain spe-
cifically to the outside influence of federal and State
governments. Although counties are a part of these larger
units, the three variables are included under environment
because state and federal policies affecting a county are
not totally detenmined by the local population but in compe-

tition with other populations seeking both funds and jobs.

um...

UH ..A .

t
:i
V

   

 

 

45

Also, in regard to this one measure, because relative federal
expenditures have remained fairly consistent over time among
units, causality can be reasonably inferred.

2. Per capita revenue from state government in a
county, 1966-67 (State Revenues).

3. Percentage of federal government employees,
December 1965 (Federal Employment).

4. Existence of an interstate highway, 1970
(Interstate).
This binary variable includes only highways completed by
1970; thus, they existed at least during part Of the previous
decade. Also, although the measure is designated inter-
states, it includes other four-lane (or larger) highways
if they lead to places outside the county. On the other
hand, multi-lane roads that either encircle part of a
town or run between two nearby towns within the same county
are not coded as interstates. The major noninterstate
four-lane highway that is coded as an interstate is route
66 in Illinois and Missouri which be being superceded by
Interstates 55 and 44.

5. Nearness to closest SMSA, 1970 (Nearness to
SMSA).
Both numbers 5 and 6 are calculated by measuring the
distance from each nonmetropolitan county seat to the
central city of the nearest SMSA or city respectively.
It is hypothesized that nearness to an SMSA will be

positively correlated with net-migration. The use of

46

the 1970 SMSA's affects one sample county and the new SMSA
in 1970 already had a population of 130,020 in 1960 while
the two cities comprising the Central City had a larger
total population in 1960 than in 1970.

6. Nearness to closest city of 25,000 or more,
1970 (Nearness to City).
The relationship between this variable and net-migration
is posited to be similar to the association between SMSA's
and migration. Because cities of 25,000 or more exist in
nine counties in the sample, this variable overlaps with
largest urban place and thus blurs the distinction between
organization and environment in those cases. The use of
1970 data, moreover, entails the use of six cities that
‘moved into the 25,000 or more category during the decade.
On the other hand, no usable 1960 cities of 25,000 or more

fell below that population level by 1970.

Measure pf Environmental Change

 

1. Average change in acreage per farm, 1959-69
(Change in Farm Size).
It is posited that smaller increases in average farm size
(average size decreased in only two counties) will be
positively correlated with net-migration since a smaller
increase suggests that fewer farms are being consolidated
and fewer farm operators are being displaced through sales

of farms.

47

Egg

Various p0pulation data are available in the
published U.S. Census of Population for 1960 and 1970.
Local government employment and state funds data are found
in the 1967 U.S. Census of Governments while the 1967 Census
of Manufactures provides data on manufacturing concerns.
Information on recreation-related facilities is available
in the 1967 Census of County Business. Data pertaining to
farms and farm operators are provided by the Census of
Agriculture for 1959, 1964 and 1969. Since some of the
data have been compiled, percentages computed, and published
in the 1967 County and City Data Book, this source is
utilized where applicable. Data for federal expenditures
for 1970 are published by the National Technical Information
Service, and measures of distance and existence of an inter-
state have been determined with the use of the Rand-McNally

Road Atlas.

Method pf Ana lys is

The data will be analyzed through simple and partial
correlation and stepwise multiple regression analysis. The
only difficulty in using such techniques arises with two
binary variables, change in land use and existence of an
interstate highway. All other factors are interval level
variables and all relationships are assumed to be linear.
Furthermore, because in several cases confirmation of the
ecological model would be found in obtaining negative

correlations, those variables have been transposed by

48
multiplying by -1 so that all positive correlations appear-
ing in the tables support the model and negative correla-
tions do not. Furthermore, because of the large number of
variables used in the study, the transposed variables will

be marked by a (t) throughout the analysis to aid the

reader.

‘9‘» a

I‘
I...
v

.
“_

 

 

n,
0
.t
l
'0
.0 l
K I
I

 

CHAPTER III
ZERO-ORDER CORRELATIONAL ANALYSIS

In order to examine the relationships of various
organizational and environmental indices to migration,
Pearson's r will be calculated between all independent
variables and the net-migration rate. In addition to these
34 independent variables, correlations will also be
computed between migration and population change in both
the urban and rural non-farm and the rural farm sectors.
The analysis will also encompass the examination of two
sub-samples. The first consists of the 167 counties where
the 1960 labor force was principally engaged in agriculture
(165) or mining (2), and the second includes the 60 more
diversified counties where other types of industrial
employment predominated in l960--primarily manufacturing
(29) and retail trade (26). Underlying this division is
the assumption that the second sub-sample represents
diversification away from agriculture. This assumption
rests on the fact that noncity populations historically
in the United States have been primarily engaged in
agriculture. That 55 of the 60 counties in the more
diversified sub-sample, even though having fewer workers

49

50
in agriculture compared to other industries, still main-
tained more farm workers than would be expected based on
the U.S. occupational structure in 1960 and that 37 of these
counties would have been placed in the agricultural group
in 1950 indicate that this second sub-sample represents
diversification away from agriculture.

Since urbanization on one level can be viewed as
the movement of a p0pu1ation from an agricultural to a
nonagricultural sustenance base, such a schema can serve
to clarify and test Groth's suggestion that rurality may
be more important than functional diversity in determining
‘migration patterns. That is, the 167 agricultural or mining
counties can be designated as rural while the 60 more
diversified counties can be viewed as more urbanized non-
metropolitan areas.1 Thus, an analysis utilizing these
two sub-samples as well as the entire sample can Offer
further insights into not only the relationship between
rurality and functional specialization but also their
relative impact on migration.

The first step in this examination is to determine
how the farm and non-farm sectors of nonmetropolitan county
p0pu1ations relate to net-migration. Correlating migration
wdth.changes in the size of both the urban and rural non-

farm and the rural farm sectors not only measures the

 

1Although the criterion for determining a county's
rurality in this study differs from Groth's, one of his
three criteria is also based on occupation.

51
contribution of migration to population change but also
gives an indication of how changes in these sectors
influence migration. That is, from this second perspective
population size in a particular sector of the sustenance
structure serves as a surrogate measure for employment
opportunities. Examination of these variables demonstrates
that for the sample, rural and urban sub-samples respec-
tively migration is more highly associated with changes in
non-farm.population (.670, .573, .929) than with changes
in the farm population (.138, .270, .059). This indicates,
similar to previous studies, that migration is directly
and positively related to the ability of a pOpulation to
absorb displaced farmers and their children into a non-
agricultural structure when agricultural employment declines.
The relatively higher correlation between rural farm
population change and migration in the rural sub-sample
with its lower degree of urbanization than the entire sample,
however, also indicates that an expanding agricultural
sector is positively associated with migration. On the
other hand, the extremely high correlation between the non-
farm population and net-migration in urban counties suggests
that the lower fertility rates among nonagricultural popula-
tions may also affect the relationship between population
growth in that sector and migration. That is, because
population size is a function of fertility and mortality
as well as migration, a low fertility rate which makes a

comparatively small contribution will increase the relative

52
influence of migration in contributing to population growth.
However, the large differences in the correlations yielded
by the two independent variables in all three sets of
counties also indicate that, in addition to the possible
influence of differential fertility rates, migration is
positively related to the ability of nonmetropolitan popu-

lations to diversify away from agriculture.

Organization and Environment

Although moderate to low, correlations between
organizational measures for the entire sample and the two
sub-samples generally support the ecological model (see
Table 1).2 The index of diversity which encompasses a
county's entire employed population correlates most strongly
with the net-migration rate in all three groups (.517,
.503, .328 for the sample, rural and urban counties
respectively). That the association between this variable
and the net-migration rate is lower for the urban than for
the rural sub-sample further suggests that this particular
index may be less sensitive to structural differences in
diversified areas than in specialized areas.

Three of the four remaining variables correlating
‘with migration at .30 or above for the entire sample are
Ineasures of manufacturing. Percentage employed in manufactur-

ing is associated with migration at a level almost identical

 

2Significance levels for the results of this analysis
as well as for the results of the stepwise multiple regres-
sion and partial correlation analyses can be found in
Appendix III.

«A.
J .

 

53

TABLE 1

ZERO-ORDER CORRELATIONS BETWEEN

MEASURES OF ORGANIZATION

AND NET-MIGRATION

 

 

Rural

 

. Urban
Sample (Agrlculturally . . .
N 227 Specialized) (Diviiiifled)
N-167
Diversity .517 .503 .328
Manufacturing
Employment .516 .532 .329
Manufacturing
Categories .466 .445 .271
Other Farm Income (t) .427 -.420 -.270
Manufacturing Firms .301 .276 .132
College .298 .262 .300
Local Government
Employment .295 -.319 -.260
Hotels .250 .180 .149
Farm Categories .240 .210 .125
Female Participation .236 .163 .131
Largest Town .158 .173 -.122
Income Under $3000 (t) . 156 . 086 - . 154
Institutionalized .148 .190 .045
Public Administration
and Education .113 .029 .309
Unemployed (t) .104 -.087 .085
Amusement Places .059 .052 .142
‘Military .050 .106 -.068
Public Relief (t) .021 -.123 .150

 

to that between diversity and migration (.517 and .516) for

the entire sample. In addition, number of manufacturing

categories and number of manufacturing firms with at least

20 employees yield correlations of .466 and .301 respec-

tively for the 227 counties.

However, all three variables

have weaker relationships with migration in the urban sub-

sample. Since manufacturing activity has been accounted

 

In...
.a

'l

5...,
ii

'0).

...:

‘I
..~'

"n
5
'cl 3.

5"..

5..
l.-

u.-

‘7.
n
7%..

4' I.

i].

I.
5"

(In

54
for, in part, through the inclusion criterion for this
sub-sample, such results are to be expected; i.e., almost
half the urban counties are designated as high in manu-
facturing so that these variables are partially measuring
strength of manufacturing activity in comparatively strong
manufacturing areas. That a similar pattern holds between
diversity and migration comparing the urban sub-sample to
the other two groups also suggests not only a high degree
of relationship between diversity and manufacturing but
also the importance of manufacturing activity in the
diversification of nonmetropolitan counties.

Other farm income (transposed), the fourth variable,
contradicts the ecological model correlating at -.427 with
net-migration for the total sample. However, as pointed
out in the previous chapter, the model suggests a relation-
ship in either direction. Accordingly, the seemingly more
plausible negative correlation between farm income and
'migration indicating a multiple-income circumstance to be
a step between full-time farmer and migration was chosen
to be tested and the variable transposed accordingly.
Results now indicate, however, that it is more probable
that the availability of off—the-farm employment to members
of a farm family is either an inducement to other farmers
to migrate into a county and/or allows existing farm
operators to remain while other factors influence migration.

Such "flip-flop" explanations in the model indicate, though,

55
that the model has not been highly develOped and further
empirical explorations are needed.

Similar to the manufacturing variables, other farm
income yields a lower correlation with migration among urban
counties than overall. That it also correlates (transposed)
with manufacturing employment at -.528, -.524 and -.280
for the sample, rural and urban sub-samples suggests a
fairly strong association with manufacturing.3 Thus, the
lower correlation with migration in the urban sub-sample
‘may reflect other farm income's relationship with manu-
facturing. On the other hand, the lower correlation may
also be due to the lesser influence of farming in these
60 counties.

Both percentage college students (.298) and per-
centage employed in local government (-.295) correlate
with net-migration near .30 for the sample; the respective
correlations for the sub-samples are similar. That the
latter variable is negatively associated with migration
suggests that a minimal level of government services and
hence employees are maintained whether or not a county is
losing population through migration. It also suggests,
Inoreover, that there may be a lag in cutting back positions
in such counties while there may be a concomitant lag in
expanding local government in areas that are growing through

migration .

 

3Correlation matrices for the entire sample, rural
and.urban sub-samples can be found in Appendix II.

56

With one exception, the remaining eleven variables
correlate with net-migration at or below .25 for all three
groups. Percentage in public administration and education,
the only exception, correlates with net-migration at .309
in the urban sub-sample compared to .113 for the sample
and a negligible .029 for rural counties indicating that
public services may be a dimension of diversification that
develops where either diversification and/or urbanization
has reached some critical point. That is, a progression
in diversification from agriculture to industry or trade
to public services may exist.

Finally, three variables are associated with net-
migration in the urban sub-sample contrary to their relation-
ships in the other two groups. Size of largest town and
percentage with median income under three thousand correlate
negatively while percentage on public relief correlates
positively with net-migration among these 60 counties and
vice versa for the sample and rural counties. The low
correlations coupled with the small urban sample size,
however, mitigate against making assertions about these
differences or any similar differences concerning variables
under the other three rubrics.

An analysis of environmental variables shows that
they also tend to confirm the ecological model although
the most important of these factors have lower correlations
with migration than do the most important organizational

measures (see Table 2). Nearness to SMSA has the strongest

57
relationship with net-migration (.410, .400 and .303 for
the sample, rural and urban sub-samples respectively) among
environmental variables. Coupled with the strong associa-
tion between diversity of county structure and migration,
this suggests that diversity of structure, either of a
county's population or easily accessible to it, is the most

important factor in explaining the net-migration rate.
TABLE 2

ZERO-ORDER CORRELATIONS BETWEEN MEASURES
OF ENVIRONMENT AND NET-MIGRATION

 

 

Rural

 

. Urban
Sample (Agr1culturally . .
N=227 Specialized) (Diviifififled)
N-l67
Nearness to SMSA .410 .400 .303
Federal Outlays -.379 -.376 -.194
Nearness to City .304 .301 .110
Federal Employment -.l92 -.201 .060
Interstate .181 .125 .071
State Revenues .043 .003 .096

 

Federal outlays, the second most highly correlated
variable, is negatively associated with net-migration
(-.379, -.376, and -.l94 for all,rural and urban counties).
The reasons for this may be identical to those suggested
regarding local government employment, i.e., certain
minimal levels are maintained and/or expansion or contrac-
tion does not take place at the same rate as pOpulation

size change through migration. However, this variable

58
also includes two other dimensions that may account for
its negative correlation. Many of the costs of federal
projects such as highways tend to be similar wherever they.
are built; these stable costs will be reflected in a higher
amount spent per capita in areas with smaller populations
(the more rural areas) which demographers have indicated
to be the areas with greatest population loss through migra-
tion. The reduction in this correlation for urban counties
compared to the entire sample tends to support such an
explanation. In addition, rural congressmen may be better
able to solicit funds and jobs for their constituents than
urban legislators. Illustrative of this was the existence
of post offices and personnel in rural areas that served
extremely small numbers of residents in the 19603 compared
‘with the larger populations served per employee in large
urban centers. Such explanations would also explain the
negative correlations between federal employment and migra-

tion in the sample and rural sub-sample. Finally, the nega-
tive correlation between federal outlays and net-migration
may reflect high governmental subsidies to farmers in the
form of cash payments for crops and loans administered
through the Department of Agriculture.

One additional variable is associated with net-
migration above .30 for the entire sample; nearness to
cities of 25,000 or more correlates at .304. Comparing this
with the higher correlation for the presumably more diversi-
fied SMBATs in which the central city must have a 50,000
minimum.population supports the proposition that the greater

c.

'I

.‘J .

59

the organizational diversity, either of a population or
easily accessible to it, the more positive will be the net-
'migration rate. Furthermore, for both these variables the
correlations are lower in the urban sub-sample than in the
other two groups suggesting that where diversification

exists within a population, access to an even more diversi-
fied structure has less influence on net-migration than where
the population is more heavily specialized in agricultural

sustenance activities.

Opganizational and Environmental Change

 

Switching the focus to variables indicating rates
of organizational and environmental change provides an
indication of how these two processes relate to migration.
Although ecologists propose that on-going changes in popu-
lation, organization, and environment (as well as technology)
are reciprocally related, for the purpose of this study,
migration is viewed as the dependent variable. This some-
what arbitrary decision is based on the premise that it
seems more logical that a change in either of the two
posited independent variables will affect migration more
directly than migration will affect either organizational
or environmental change in nonmetropolitan areas. That is,
it is more likely that the decision to locate a factory or
establish an interstate will have a greater direct effect on
migration in a particular nonmetropolitan area than vice
versa.

Among measures of organizational change, change in
size of largest urban place correlates most strongly with

net-migration for all three groups--.528, .501 and .679.

on:
bii

(II

H:
w.

'OI

...

I
"I

n.‘
l

'1

//

 

 

60
for the sample, rural and urban sub-samples respectively
(see Table 3). The higher association for urban counties,
contrary to the pattern found among measures of organiza-
tion, suggests that this variable is either measuring a
different dimension of organization than the static vari-
ables or reflecting a dimension of population change rather
than organizational change. The low correlations between
largest urban place in 1960 and net-migration as well as
the negligible relationship between this independent
variable and change in size of largest urban p1ace--.03,
.07 and .03 respectively for the total sample, rural and
urban groups respectively--also indicate that these two
apparently related town size variables may be tapping

different dimensions in regard to net-migration.

TABLE 3

ZERO-ORDER CORRELATIONS BETWEEN MEASURES OF
ORGANIZATIONAL CHANGE AND NET-MIGRATION

 

 

Rural
Urban
Sample (Agriculturally - . .
N=227 Specialized) (Dlviiiifled)
N=167
Change in:
Largest Town .528 .501 .679
College .305 .301 .230
Military .238 .079 .450
Institutionalized -.179 .230 -.011
Female Participation .082 .010 -.129
Manufacturing
Employment .065 .142 .092
Unemployed (t) -.056 .002 -.226
Public Administration -.018 .019 -.175

Land Use .004 -.039 .103

 

61

Change in percentage college students, the only
other variable in this set to be related to net-migration
above .30 for the sample, ranks second in importance in
its association with migration both for all 227 counties
(.305) and for rural counties (.301) and third for urban
counties (.230). Furthermore, the correlation for the
sample is similar to that for percentage college students
in 1960 while the change variable is slightly higher for
rural and lower for urban counties than the static variable.

Although none of the seven other variables correlates
with net-migration above .30 for the entire sample, sub-
sample correlations reveal that various factors differ in
their relationships with migration in rural and urban
areas. Among the 167 rural counties percentage of insti-
tutionalized population is positively related to migration
(.230) although the comparable correlation is negative
overall (-.179) and negligible for urban counties (-.011).
This indicates that the establishment of such institutional
facilities is much more important in inducing in-migration
and/or in retarding out-migration of residents in rural
areas than in urban areas where diversity of structure has
developed along other lines.

The remaining discrepancies between sub-samples and
the entire sample follow the same pattern for measures of
organizational change as for static organizational variables
in that most of the differences occur in the urban sub-

sample. Most importantly, change in percentage of military

62
personnel has the second highest correlation with net-
migration among these 60 counties (.450). A comparison
‘with the negligible association in the rural sub-sample
(.079) and the lack of association between military person-
nel in 1960 and net-migration reflects the existence of
military bases in some of the urban counties by 1970 and
the build-up of armed forces personnel in the 19603.
Furthermore, this also suggests that urban or diversified
nonmetropolitan counties offering more services to military
personnel than rural counties while still possessing rela-
tively large amounts of unpOpulated land are attractive
sites for military bases.

Finally, two additional factors correlate more
strongly with net-migration in the urban sub-sample than
in the other two groups. In contrast to the negligible
relationships found in both the entire sample and rural
sub-sample, both change in the unemployment rate (-.226)
and in percentage employed in public administration and
education (-.l75) are more negatively associated with net-
nflgration in the urban group. The latter variable follows
the same pattern as percentage employed in public adminis-
tration and education in 1960 in its stronger correlation
for the urban group. However, the static variable is related
positively with net-migration while the change variable,

similar to other government-related variables, yields a

63

negative correlation.4 This indicates that although public
services may be a more important dimension of diversifica-
tion in urban counties compared to rural counties as dis-
cussed earlier, the negative relationship between change
in such employment and migration reflects either minimal
necessary levels and/or particularly an inability to expand
or contract public services quickly in response to popula-
tion changes due to migration in urban as well as rural
counties. A more precise analysis of these tentative
explanations regarding government-related variables, however,
must await future investigations encompassing population
change per se. To an even greater extent any explanation
regarding change in the unemployment rate must remain
tenuous. It is possible, however, that this measure may
reflect a time lag between individual migration and employ-
'ment among nonagricultural migrants.

Change in average size of farm (transposed), the
only measure of environmental change utilized in the
study, moderately supports the ecological model correlating
with net-migration at .361, .330 and .587 for the entire
sample, rural and urban sub-samples respectively. This
demonstrates that more positive net-migration rates are
associated with smaller increases in average farm size.

Such results also imply that a larger increase in farm

 

4Ten government-related correlations are negative
‘while five are positive; moreover, three of the five
positive correlations with net-migration concern public
administration and education employment in 1960.

64
size is related to a more negative net-migration rate.
If both this implication and the ecological model are
valid, the stronger association in the urban sub-sample
may reflect a strong stream of farm migrants to nonagricul-
tural niches in nearby towns which in turn would decrease
the availability of jobs for inter-county migrants and thus
stem.in-migration. The lower association between farm
size change and migration in rural counties, on the other
hand, would reflect the fewer nonagricultural employment
opportunities available to either displaced farmers or

inter-county migrants.

Theoretical Implications

This examination supports the two hypotheses drawn
from the ecological model although correlations generally
are low to moderate. Specifically, correlations support
the hypothesis that both organization and environment have
a direct effect on the net-migration rate, and the positive
associations between the vast majority of variables and
migration support the hypothesized direction of specific
relationships.5 The higher associations obtained for the
most important organizational variables compared with environ-

nmntal variables among both static and change factors also

5Among the twelve indicators of organization and
environment and changes in each that are most strongly
associated with migration, only other farm income (trans-

;msed) and federal outlays do not correlate in the hypothe-
sized direction.

1"

3:61

I?!

3"

‘O

’l

.\

\J

65
indicate that the direct influence of the former is greater
than the latter on net-migration.

Findings further reveal that within each rubric the
variable which measures total diversity of structure,
either of or easily accessible to a population (diversity,
SMSA, change in largest town), correlates most strongly
with net-migration although the organizational change in
largest town may be tapping population change rather than
functional diversification. Regarding other variables,
results indicate that manufacturing--measured in terms of
employment, diversity of enterprises, and/or size--as well
as the existence and growth of a college population are
important factors in yielding more positive net-migration
rates.

A comparison of sub-samples reveals that diversity
of structure has a greater impact on the net-migration rate
in rural than in urban areas. However, these differences
are better explained by the dimension of diversity away
from agriculture in the inclusion criterion for urban than
by urbanization itself. Such results point out the problems
inherent in attempting to discern the influence of various
ecological factors when utilizing a measure of rurality
versus urbanity based on occupation. Two solutions to this
dilemma are possible, however. One is to determine if
"urban" has a meaning other than one based on occupation,
and if so, to develOp indices based on these other dimensions.

Parenthetically it should be noted, though, that even Wirth's

66

classic criteria of size, density, and heterogeneity in
determining degree of urbanization rests to some extent on
an occupational assumption since the relatively larger
amount of unpOpulated land needed for farming mitigates
against as large and dense settlements as those comprising
populations not engaged in agriculture.6

The other alternative is suggested by the results
concerning percentage employed in public administration and
education. That is, there may be a pattern such that at one
structural point diversification tends in one direction,
e.g., agriculture to manufacturing or trade, and at other
points tends in other directions, e.g., manufacturing to
public services. Thus, by determining the critical structural
points at which diversification entails different dimensions
it may be possible to categorize counties on the basis of
occupational and occupation-related variables as to degree
of urbanization and then within different categories explore
various ecological relationships. Moreover, the establish-
ment of such critical points may help to clarify more fully
the process of a population's occupational movement from
primary to secondary to tertiary industries.

Finally, although the two sub-samples yield differ-
ent correlations between the various independent variables
and net-migration, the most important variables in one group
with few exceptions correspond to the most important in the
other. Because of this consistency, all multiple and partial

correlational analyses will encompass only the entire sample.

 

6Louis Wirth, "Urbanism.as a Way of Life," The
American Journal of Sociology, 44 (July, 1938), 1-24.

 

'-

 

 

 

CHAPTER IV

STEPWISE MULTIPLE REGRESSION AND PARTIAL
CORRELATIONAL ANALYSES

Extending the analysis of net-migration from
simple correlations to stepwise multiple regression will
not only further explicate the degree of variance in the
dependent variable explained by the three sets of eco-
logical variables and the one measure of environmental
change but will also indicate how much each independent
variable adds to the explanation provided by all previously
entered variables. Basically the computation of multiple
stepwise regression beyond the first step (where the
single best predictor of the dependent variable is entered)
entails adding each variable on the strength of the product
of its normalized beta squared if added at that step
multiplied by its tolerance or the degree to which the
measure taps a different linear dimension than those
variables already entered; this product equals the partial
correlation coefficient squared. In this examination, all
ecological variables will be included in the appropriate
equations regardless of the strength or direction of
association with net-migration shown by simple correlations
since it is possible that in controlling for factors

67

 

greviou
pendent
wrong?

uted I

tion c

(a? CI

beta c

also I

Vat-1

thoS

Ofv

the

68
previously entered into the equation, a particular inde-
pendent variable may be related to net-migration more
strongly or in the opposite linear direction than indi-
cated by Pearson's r.

In carrying out the analysis, the multiple correla-
tion coefficient (R), the variance explained (R2) and the
change in variance explained by the addition of a variable
(R2 Change) will be included in tables. The normalized
beta controlling for all other independent variables will
also be listed although the generally low to moderate
simple correlations caution against strong predictive
assertions based on this statistic. Moreover, the rela-
tively high degree of association among several of the
independent variables suggested by the analysis of zero-
order correlations also indicates that the beta weights
may not accurately reflect the actual relationship between
some independent variables and the net-migration rate.
Examination of migration using partial correlational
analysis in the second section of this chapter will help
to clarify the extent of such multicollinearity among those
independent variables explaining the greatest amount of
variance in the net-migration rate.

In addition to its use in examining more closely
those variables contributing most fully to the explanation
of variance, partial correlational analysis will be utilized
in the first section of the chapter to define more precisely

the relationship between the first variable entered into

69
each stepwise regression equation and other independent
variables highly correlated with the first variable regard-
less of their associations with net-migration. This is
because, as noted above, the variable entered on the first
step is simply that factor which is the best predictor,
based on the zero-order correlation coefficient, of the
dependent variable. However, it is possible that the
measure may, in fact, be a surrogate for or a composite
of other factors used in the study. Thus, prior to each
examination utilizing stepwise regression analysis, partial
correlations will be computed using net-migration as the
dependent variable in order to clarify the relationship
between the first variable entered and others closely
associated with it. Specifically, the other variables
include all independent measures under the same ecological
rubric correlating above .50 with the first independent

variable in the equation.

STEPWISE MULTIPLE REGRESSION ANALYSIS

Measures of Opganization and Environment

Among the eighteen organizational variables, the
index of diversity, correlating most highly with net-
migration at .517, is the first variable entered into the
regression equation. That the association between per-
centage employed in manufacturing and net-migration (.516)
was found to be almost as high and the two independent

variables correlate with each other at .668 necessitates

 

 

A n

v.

.1‘
-.

 

 

 

 

70

further investigation of these relationships. Although
partial correlational analysis of the two variables reveals
that each correlates at a much lower level with net-migration
suggesting that to some extent they are measuring the same
thing (or tapping the same linear dimension), they retain
their relative order of importance. Specifically, the
partial correlations for diversity and manufacturing equal
.271 and .268 respectively.

Two other variables, number of manufacturing
categories (.751) and number of manufacturing firms with
at least 20 employees (.512), are also highly associated
with diversity. Including all four variables, the partial
correlation for each with migration equals .201, .247,
.079 and -.110 respectively for diversity, manufacturing
employment, manufacturing categories and manufacturing
firms of 20 or more. In other words, the correlations
between the last two variables and net-migration fall to
extremely low levels compared with both their simple
correlations (.466 and .301 respectively) and the partial
correlations for the first two measures; however, the
association between diversity and migration also is weakened
more than that between manufacturing employment and the
dependent variable yielding a reversal in their relative
positions. These results suggest that, although the
structural index contains other dimensions, to a great
extent it is measuring the influence of various aspects

of manufacturing on net-migration; such results are not

'r' "

n

.O

'I

.4
-

 

71
surprising since manufacturing is one of the two major
avenues diversification takes in these nonmetropolitan
counties. However, that diversity is still the best
single predictor of net-migration, encompasses more aspects
of manufacturing than percentage in manufacturing, and
maintains that position vis-a-vis partial correlational
analysis between only diversity and percentage in manufactur-
ing indicates it is a better overall measure than the latter
to-enter first into the equation in seeking the optimal
solution to explaining the greatest amount of variance
with the fewest independent variables.

Results of stepwise multiple regression show that
all eighteen organizational variables collectively explain
almost 54 per cent of the variance in net-migration (see
Table 4). However, seven of these measures account for
more than 50 per cent with each adding at least one per
cent to the explained variance. In contrast, the eleven
remaining variables only explain an additional 3.5 per
cent and none increases the amount of variance explained
by at least one per cent.

Comparing the seven highest ranking variables
with their relative position utilizing Pearson's r demon-
strates that five of them--diversity, percentage local
government employment, percentage college population, per-
centage employed in manufacturing, number of manufacturing
categories--also correlate individually close to or above

.30 with net-migration. Two of these five, however,

 
 
 

72
TABLE 4

RESULTS OF STEPWISE MULTIPLE REGRESSION

OF MEASURES OF ORGANIZATION WITH

NET-MIGRATION

 

 

 

. 2 R2 Beta
Multiple R R Change Weight

Diversity .518 .268 .268 .271

Local Government

Employment .575 .331 .063 -.203
College .617 .381 .050 .239
Manufacturing Employment .658 .433 .052 .261
Largest Town .689 .474 .041 -.375
Manufacturing Categories .701 .491 .017 .273
Public Relief (t) .710 .504 .013 -.077
Unemployment (t) .716 .513 .009 .161
Other Farm Income (t) .723 .522 .009 -.082
Hotels .728 .530 .008 .150
Public Administration .730 .533 .003 .068
Military .732 .535 .002 .055
Amusement Places .733 .537 .002 .070
Income Under $3000 (t) .733 .538 .001 .047
Institutionalized .734 .539 .001 .026
Female Participation .734 .539 .000 -.035
Manufacturing Firms .734 .539 .000 .040
Farm Categories .734 .539 .000 .006

 

contribute less to the explanation than their simple cor-

relations with net-migration would indicate.

Both percent-

age in manufacturing and number of manufacturing categories,

the second and third most important variables in the simple

correlational analysis, are entered at steps four and six

respectively. In regard to the former, the above partial

correlational analysis indicates that the high relationship

between diversity and percentage in manufacturing coupled

with diversity being entered first would decrease the

 

U1

u,

 

 

1':

73

manufacturing variable's contribution since the structural
index, to some extent, measures the same linear dimension.
Similarly, the extremely low partial correlation for manu-
facturing categories and net-migration controlling for
diversity and manufacturing employment as well as number
of manufacturing firms of 20 or more indicates that manu-
facturing categories is tapping a dimension similar to
these other variables.

On the other hand, two variables that ranked in
the lower half in the simple correlational analysis collec-
tively contribute more than five per cent to the explained
variance. Specifically, size of largest urban place
(r = .158) is entered on the fifth step, and percentage
on public relief (r = -.021) is entered on the seventh.
This suggests that these two variables tap linear dimensions
both different from other variables more highly correlated
with net-migration and more strongly than other low-
correlating measures.

By contrast, two variables correlating at more than
.30 in the earlier analysis add less than one per cent
apiece to the explained variance. The negligible contribu-
tion of number of manufacturing firms with at least 20
employees can be attributed to its high association with
diversity and the other two manufacturing variables as
partial correlational analysis illustrates. Similarly,

the small contribution of other farm income may result

74
from its fairly strong relationship with percentage employed
inmmanufacturing.

Turning attention to the six environmental variables,
the first factor to be entered into the stepwise multiple
regression equation is nearness to an SMSA. Among other .
measures under this rubric, only nearness to a city of
25,000 or more, correlating at .766, is associated with
this independent variable at a level above .50. Partial
correlational analysis yields relationships of .289 between
SMSA and net-migration and -.016 between city of 25,000
or more and migration compared with simple correlations
of .410 and .304 respectively. Thus, despite the reduction
in association between SMSA and net-migration, it remains
stronger than the relationship between city of 25,000 or
more and net-migration when the influence of the other
independent variable is controlled.

A perusal of the results of stepwise multiple
regression shows that the six environmental variables
collectively explain only 23.3 per cent of the variance
in net-migration (see Table 5). Moreover, nearness to an
SMSA and federal outlays account for 21.6 per cent of this
and are the only variables individually contributing at
least one per cent to the explanation. These findings
also reveal that the rank order of independent variables
is identical to those based on Pearson's r with one excep-
tion, the decline in rank from third to fourth of nearest

city of 25,000 or more. This is most likely due to the

75
close relationship between city of 25,000 or more and
SMSA as illustrated both by the high correlations between
them and the negligible association between the former and

net-migration controlling for the latter.
TABLE 5

RESULTS OF STEPWISE MULTIPLE REGRESSION
OF MEASURES OF ENVIRONMENT
WITH NET-MIGRATION

 

 

 

2
. 2 R Beta
Multiple R R Change Weight
Nearness to SMSA .410 .168 .168 .357
Federal Outlays .465 .216 .048 -.256
Federal Employment .471 .222 .006 -.075
Nearness to City .476 .227 .005 -.130
Interstate .482 .233 .006 .078
State Revenues .482 .233 .000 .010

 

Comparing the two sets of static variables reveals
that organization explains more than twice as much of the
variance in net-migration as does environment thus support-
ing the hypothesis that the former has a stronger direct
effect on migration than the latter. Including all vari-
ables under both rubics in one stepwise multiple regression
equation illustrates even more clearly the difference in
their relative explanatory power. SMSA, the highest ranking
environmental variable, is entered at step eleven and adds
less than one per cent to the variance explained by the ten

organizational variables preceding it while other

76

environmental variables contribute even less (see Table 6).

Furthermore, including the environmental variables increases

the amount of variance explained by organizational measures

TABLE 6

RESULTS OF STEPWISE MULTIPLE REGRESSION OF

MEASURES OF ORGANIZATION AND ENVIRONMENT

WITH NET-MIGRATIONa

 

 

 

, 2 R2 Beta
Mu1t1ple R R Change Weight
Diversity .518 .268 .268 .215
Local Government
Employment 575 .331 .063 -.210
College .617 .381 .050 .211
Manufacturing Employment .658 .433 .052 .203
Largest Town .659 .474 .041 -.368
Manufacturing Categories .701 .491 .017 .252
Public Relief (t) .710 .504 .013 -.096
Unemployment (t) .716 .513 .009 .165
Other Farm Income (t) .723 .522 .009 -.097
Hotels .728 .530 .008 .161
Nearness to SMSA .732 .536 .006 .163
Public Administration .736 .541 .005 .084
Nearness to City .737 .544 .003 -.099
Federal Outlays .738 .545 .001 -.072
Military .740 .547 .002 .048
Amusement Places .740 .548 .001 -.064
Female Participation .741 .549 .001 -.046
Interstate .741 .550 .001 .031
Manufacturing Firms .742 .550 .000 .054
Institutionalized .742 .550 .000 .014
Farm Categories .742 .550 .000 -.013
Income Under $3000 (t) .742 .550 .000 .011

 

aState Revenues and Federal Employment, two environmental

variables, were not entered in the equation because of

extremely low F-Levels (.001) after the last step listed;
the beta weight of each if entered at this step would be
.001 and -.002 respectively.

77
only 1.1 per cent. In addition to demonstrating the over-
whelming influence of organization, these findings also
indicate that the two highest ranking variables in the
analysis of environment alone may be highly associated
with high-ranking organizational factors. The correlations
between both SMSA and federal outlays and diversity (.507
and -.519 respectively) tend to confirm this suggestion.
Because of these correlations and in spite of the results
of the combined regression equation showing the negligible
influence of environment a partial correlational analysis
will be included in the second section of this chapter
encompassing both the seven organizational and the two
environmental factors contributing over one per cent to
the explained variance in separate analyses in order to
clarify further relationships between organization and
environment.

iMeasures of Organizational and
Environmental Change

 

 

Shifting the focus of analysis to measures of
ecological change indicates that change in largest urban
place is the best single predictor of net-migration among
organizational change variables; moreover, no other inde-
pendent variable correlates with it above .50.1 Overall,

the nine variables explain only 37.2 per cent of the

 

1As pointed out in Chapter III, the relationship
between change in size of largest urban place and migration
ummeore accurately reflect population change than
organizational change.

78

variation in net-migration, and five of these--change in

size of largest town, military, institutionalized, college

population, and percentage in public administration and

education--explain almost 36 per cent of the variance

with each contributing at least one per cent (see Table 7).

TABLE 7

RESULTS OF STEPWISE MULTIPLE REGRESSION OF

MEASURES OF ORGANIZATIONAL CHANGE
WITH NET-MIGRATION

 

 

 

- 2 R2 Beta
Multiple R R Change Weight
Change in:
Largest Town .528 .299 .279 .445
Military .564 .318 .039 .196
Institutionalized .573 .328 .010 -.129
College .582 .338 .010 .195
Public Administration .600 .360 .022 -.l6l
Land Use .604 .365 .005 -.078
Manufacturing
Employment .608 .370 .005 .076
Unemployment (t) .609 .371 .001 .043
Female Participation .610 .372 .001 .037

 

A comparison among the five variables with their

relative strength utilizing Pearson's r reveals that four

of them were also the most important factors in the earlier

examination although the order varies.

Specifically, change

in percentage college students ranked second but is entered

on the fourth step in the regression equation after change

in percentage military and in percentage institutionalized.

In addition to these four variables, percentage employed in

79
public administration and education, one of the remaining
variables each of which yields a Pearson's r under .10 with
net-migration, is entered on the fifth step and is the
only other variable adding at least one per cent to the
explained variance. This indicates that it measures not
only a linear dimension different from the four variables
previously entered but also one that none of the other
extremely low correlated factors either taps or measures
as strongly.

Adding change in average farm size, the only measure
of environmental change, to the organizational change
variables raises the total variance explained from 37.2
to 46.1 per cent (see Table 8). Furthermore, this variable
ranks second in importance to change in largest urban place
while two other factors-~change in percentage military and
in the female participation rate--rank third and fourth
respectively and also contribute over one per cent to the
explained variance. Comparing this regression equation to
that generated by organizational change variables alone
reveals that in addition to the largest town variable
only change in percentage military retains a similar rank
in the second equation. On the other hand, the other three
major variables in the first equation contribute less to
the total explanation and are entered at least two steps
later in this equation while change in the female partici-
pation rate, in addition to being entered earlier, increases

its contribution from almost zero to 2.4 per cent. These

‘ilrl i. A

‘1

~\

 

 

80
TABLE 8

RESULTS OF STEPWISE MULTIPLE REGRESSION OF
MEASURES OF ORGANIZATIONAL AND
ENVIRONMENTAL CHANGE WITH
NET-MIGRATION

 

 

2
. 2 R Beta
Mult1ple R R Change Weight
Change in:
Largest Town .528 .279 .279 .424
Farm Size (t) .603 .364 .085 .328
Military .636 .405 .041 .205
Female Participation .655 .429 .024 .150
Land Use .661 .438 .009 -.095
Institutionalized .665 .442 .004 -.103
Public Administration .668 .447 .005 -.116
College .675 .456 .009 .124
Unemployment (t) .678 .460 .004 .072
Manufacturing
Employment .679 .461 .001 .030

 

changes in relative ranking and contribution, moreover,
suggest that the differences in general between the two
regressions encompassing ecological changes may be due to
average farm size being differentially associated with

various measures of organizational change.

Theoretical Implications

The findings using stepwise multiple regression
analysis support the hypothesis that organization has a
stronger direct effect than environment on migration.
Among static variables, the seven most important factors

explain half the variance in the net-migration rate while

81

all six environmental variables explain slightly under one-
fourth. That the relationship between environment and migra-
tion becomes negligible in the combined regression equation,
however, suggests that a high degree of collinearity exists
between indicators of organization and of environment; such
possible associations will be explored further in the follow-
ing section.

Contrary to the regression analysis of static
variables, this examination shows that the one environ-
mental change variable, change in farm size, maintains its
relative importance, as indicated by zero-order correla-
tions, and adds to the total variance explained when it is
included in the ecological change equation. Its inclusion
also influences the impact of various organizational change
variables on the dependent variable. This suggests, similar
to the static analysis, a high degree of collinearity
between variables under both independent rubrics and/or
the possibility that changes in land patterns has both a
direct impact on migration and an indirect effect through
organization.

Among the sixteen variables contributing over one
per cent to the explanation of variance in the net-migration
rate in the five equations, beta weights indicate that six
factors do not support the hypothesized direction of rela-
tionship posited on the basis of the ecological model.

Three of the six measure governmental inputs. A fourth

factor, percentage on public relief, may reflect governmental

82
activity as well as level of economic well-being. Thus, the
explanations for the negative relationship between govern-
'ment and migration given in the previous chapter may also
apply to this variable. The negative beta weight for change
in institutionalized population reflects the negative zero-
order correlation in the sample and urban sub-sample.
Coupled with the positive simple coefficient for the rural
sub-sample, this suggests that although the location of
special institutions in rural areas may help to stem popu-
lation losses through migration in those counties losing
population most rapidly, they are of little importance
in bringing about a more positive net-migration rate in
nonmetropolitan counties in general. Therefore, that this
variable contributes one per cent to the explained variance
yet is negatively associated with net-migration reflects
both its positive impact in the more rural counties and
the fact that such counties are experiencing higher popu-
lation losses through migration than more diversified
nonmetropolitan counties. Finally, the negative beta
‘weight for size of largest urban place may reflect chang-

ing patterns of urbanization.2

PARTIAL CORRELATIONAL ANALYSIS
In order to clarify the relationships between various

independent variables and their effects on the variance in

 

2A full discussion of this possible explanation based
on results of simple, regression and partial correlational
analysis will be included in the last chapter.

83

net-migration, partial correlations of all variables in each
regression equation contributing at least one per cent to the
explained variance of the dependent variable will be examined
in this section. Specifically;each independent variable will
be correlated with net-migration while controlling for all
other major factors in the same regression equation. In addi-
tion, each variable's zero-order correlation for the entire
sample will also be included in tables, and all measures will
be listed in the order in which they were entered into the
respective regression equations to provide for easier com-

parability with previous analyses.

Measures pf Organization and Environment

 

Examination of the seven organizational variables
contributing over one per cent to the explained variance of
net-migration reveals several differences with the results
of the two previous analyses (see Table 9). Percentage
college students yields the highest partial correlation
followed by size of largest urban place; furthermore, both

register higher partial correlations than their respective
TABLE 9

PARTIAL CORRELATIONS BETWEEN MEASURES OF
ORGANIZATION AND NET-MIGRATION

 

 

 

Partial r Simple r
Diversity .204 .517
Local Government Employment . -.292 -.295
College .409 .298
Manufacturing Employment .218 .516
Largest Town -.316 .158
Manufacturing Categories .199 .466

Public Relief (t) -.158 -.021

 

84
simple correlations and are the only two measures correlating
above .30 with net-migration in this analysis. lkiaddition,it
should be noted that contrary to its low positive simple cor-
relation, size of largest urban place has a negative associa-
tion with net-migration here and thus runs counter to the
ecological model. It may be, as other studies suggest, that
this relationship reflects the existence of regional centers
in counties more distant from SMSAH; that is,counties gen-
erally characterized by less positive or more negative net-
migration rates compared with those located nearer metro-
polises. However, a Pearson's r of .232 between SMSA and
size of largest urban place indicates that, if anything,
urban places closer to SMSA's tend to be larger.

Among the remaining five variables, percentage on
public relief also explains a higher percentage of the
unexplained variance than its simple correlation with
migration indicates while percentage in local government
yields almost identical associations with the dependent
variable. By contrast, reduced partial correlations at
about .20 for diversity, percentage in manufacturing and
number of manufacturing categories reflect the high inter-
relationships earlier found between these variables which
in turn lessen the independent influence of each on
migration.

In contrast to the inconsistencies among measures
of organization, the two most important environmental factors
yield similarly weakened associations with net-migration.

Specifically, the partial correlations for SMSA and federal

85
outlays are .292 and -.240 compared with simple correlations
of .410 and -.379 respectively. That these two variables
‘maintain their relative ranks supports the previous examina-
tions showing SMSA to be the most important environmental
factor affecting net-migration.

Combining the two sets of variables indicates that
environment has little direct influence on the relationship
between organization and migration; all partial correlations
remain within .03 of what they yield in the organizational
analysis alone (see Table 10). Furthermore, that the partial
correlation for largest urban place is reduced only slightly
does not give further insight into the negative association
found above between that variable and migration. On the

other hand, results show that the two environmental variables
TABLE 10

PARTIAL CORRELATIONS BETWEEN MEASURES OF
ORGANIZATION AND ENVIRONMENT
AND NET-MIGRATION

 

 

Partial r Simple r
Diversity .179 .517
Local Government Employment -.288 -.295
College .401 .298
Manufacturing Employment .189 .516
Largest Town -.307 .158
Manufacturing Categories .192 .466
Public Relief (t) -.159 -.021
Nearness to SMSA .081 .410

Federal Outlays -.019 -.379

 

86
are negligibly associated with net-migration when control-
ling for organization. Such findings suggest that federal
expenditures, as might be expected, are relatively consistent
among counties with similar sustenance structures. This
analysis also indicates even more than the sub-sample
comparisons of simple correlations that the internal sus-
tenance structure has a stronger direct influence on migra-
tion than access to an SMSA. That is, regardless of the
difference in distance from a metropolis of two county
populations, if they have similar sustenance structures,
population change through migration will be affected
similarly. However, that nearness to SMSA and diversity
correlate at .507 suggests that the former variable may
have an indirect influence on migration in that popula-
tions near SMSA's may be organized into more diversified
structures than populations farther away.

Measures ofO anizational and
nv ronmenta rhange

 

Examining indices of organizational change reveals
only slight differences between these results and those of
the two previous analyses (see Table 11). Although its
association with migration declines slightly, change in
size of largest urban place remains the most important
variable and the only one with a partial above .30.

Change in percentage military ranks second, and the other

three measures yield partial correlations between .15

and .20.

87
TABLE 11

PARTIAL CORRELATIONS BETWEEN MEASURES OF
ORGANIZATIONAL CHANGE AND
NET-MIGRATION

 

 

 

Partial r Simple r
Change in:
Largest Town .449 .528
Military .227 .238
Institutionalized -.157 -.l79
College .188 .305
Public Administration -.l79 -.018

 

The most noticeable differences between these
findings and earlier examinations are the increase in
explanatory power of change in percentage employed in
public administration and education (although still low)
and the decrease of change in percentage college students.
Similar to the stepwise regression results, this indicates
the former measures a linear dimension different from vari-
ables entered before it and the latter taps to some extent
the same linearity as other variables of organizational
change. Finally, while never strong influences, change in
percentage institutionalized and in employees in public
administration and education yield negative simple and
partial correlations with net-migration contrary to what
was hypothesized. Regarding the former, it may reflect
the possibly greater predominance of old-age facilities

in some counties characterized by high age-structures

88

which are in turn induced by high out-migration of the young;
thus, this variable may represent more directly changes in
population rather than in organization. The latter inde-
pendent variable, on the other hand, reflects the generally
negative relationships between public service or governmental
variables and migration discussed in the previous chapter.

Partial correlations of the highest ranking measures
of change incorporating the environmental change in average
farm size also yield results similar to those found in the
earlier analyses (see Table 12). The only exception to
this is the much higher association between change in the
female participation rate and net-migration in this examina-
tion. Although still fairly low, the partial correlation
suggests that when controlling for other changes in organiza-
tion and environment, net-migration is positively related
to the availability to women of occupational niches in the

sustenance structure.
TABLE 12

PARTIAL CORRELATIONS BETWEEN MEASURES OF
ORGANIZATIONAL AND ENVIRONMENTAL
CHANGE AND NET-MIGRATION

 

 

 

Partial r Simple r
Change in:
Largest Town .505 .528
Farm Size (t) .397 .361
Military .266 .238

Female Participation .203 .082

 

89

In addition, the slight increase in the correlation
between farm size change and migration as well as its effect
on other measures or organizational change run counter to
the results combining static organizational and environ-
mental variables in which the influence of the latter on
the former is negligible and the relationship between environ-
ment and net-migration is drastically reduced. This indi-
cates that the particular variables used in this study may
not represent environment consistently well. However,
since the three environmental factors used in the partial
correlational analyses represent very different aspects
of environment, it seems more likely that not all facets
of environment affect migration equally. Specifically,
SMSA represents access to more diversified organizational
structures; federal outlays, the impact of other govern-
mental units; and farm size change, internal land pattern
change. On the other hand, the different results may also
indicate that the interrelationships among the processes
of pOpulation, organizational and environmental change are
of a different order than that reflected by indicators of

organization and environment measured at one point in time.

Theoretical Implications

Partial correlational analysis, similar to the
previous two examinations, indicates that organization has
a direct effect on migration as hypothesized. The differ-

ences between simple and partial correlation coefficients,

90
though, reveals close relationships among several variables.
The lowered partials for diversity and the two manufactur-
ing variables demonstrate that all three to some extent
are measuring manufacturing activity. On the other hand,
higher partials for college, largest town and public
relief indicate that each has a stronger impact on net-
migration when controlling for other organizational factors
than each has independently. Finally, the consistency
between coefficients for local government employment
suggests that this variable has a relationship with net-
migration which remains unaffected by other independent
variables.

The examination of static variables also indicates
that environment has little impact on migration when con-
trolling for organization; both regression and partial
correlational analysis show a necessity for modifying
this part of the ecological model. That is, the regres-
sion equation and partials combining static organizational
and environmental variables indicate that, if a direct
relationship exists between environment and migration,
it is negligible, although the strong correlation between
diversity and SMSA suggests that environment does affect
migration indirectly through its impact on organization.

Among variables measuring ecological change, change
in size of largest urban place and change in farm size

consistently rank first and second in all three analyses

91
in their impact on migration. However, difficulties exist
in both cases regarding just what part of the ecological
model is being measured. As suggested before, change in
largest town may tap population rather than organizational
change.

Perhaps more importantly from a theoretical view-
point, change in farm size, the only measure of environ-
mental change, simultaneously reflects changes in organiza-
tion, i.e., a change in land patterns implies a change in
sustenance organization. Furthermore, if farm size change
more accurately measures organizational than environmental
change, this study indicates that environment only affects
migration indirectly through organization as shown by re-
sults of the analysis of static variables. Such an inter-
pretation of the variable supports Sly's suggestion that
the land pattern variable he used may more accurately
reflect organizational than strictly environmental change.3
If this is the case, than it is neither the different facets
of environment being tested nor the difference between
static and processual variables that have produced differing
results concerning the impact of environment on migration
but rather that change in farm size more appropriately
measures organizational change. This study then indicates
that organizational change has a direct effect on net-migra-
tion but the impact of environmental change remains to be

tested.

 

3s1y, 627.

92

Stepwise Multiple Regression of Selected
Variables

To clarify further the relative impact of organiza-
tion and environment beyond the simple and multiple corre-
lational analyses employing all variables, those static
factors which previous examinations have shown to be most
closely associated with migration have been placed into
three groups within a stepwise multiple regression equation
with each set being forced into the equation on the basis
of the importance of each in previous analyses. In order
of inclusion the three sets are organization (six vari-
ables), environmental locational factors (two) and environ-
mental inputs (one). The specific variables correspond
to those utilized in the partial correlational analysis
with two exceptions. Percentage on public relief is not
included because both its simple and partial correlation
coefficients are low. On the other hand, nearness to a
city of 25,000 is added because of its relatively high
simple correlation with net-migration although its
extremely close association with nearness to an SMSA (.765)
decreases its influence in the environmental equation.
Furthermore, it is included because this examination is
concerned with the impact of location near central places,
as measured by population size.

Results reveal, similar to previous examinations,
that organization has a direct influence on net-migration

while the impact of environment, whether locational or

93

relating to federal inputs, is negligible when organization
explains as much as it can (see Table 13). Specifically,
organization explains almost half the variance in the net-
migration rate while each set of environmental variables
adds less than one per cent to the explanation. Thus, this
analysis indicates again the need to modify the ecological
model by positing a negligible direct relationship between

environment and migration.

TABLE 13

RESULTS OF STEPWISE MULTIPLE REGRESSION OF
SELECTED MEASURES OF ORGANIZATION AND
ENVIRONMENT WITH NET-MIGRATION

 

 

 

 

 

Multiple R2 R2 Beta
R Change Weight
Organizational Factors
Diversity .518 .268 .268 .208
Local Government Employment .575 .331 .063 -.204
College .617 .381 .050 .321
Manufacturing Employment .658 .433 .052 .236
Largest Town .689 .474 .042 -.304
Manufacturing Categories .701 .491 .017 .254
Environmental Locational
Factors
Nearness to SMSA .703 .494 .003 .138
Nearness to City .705 .498 .003 -.115
Environmental External
npu ut actors
ederal Outlays .707 .500 .002 -.058

 

The direction of association, as indicated by beta

weights, also remains unchanged. It should be noted, however,

94
that nearness to a city of 25,000, not discussed before,
yields a negative beta weight contrasted to a positive
simple correlation with migration. That the distance to
a city is identical to the distance to an SMSA in 135
cases, i.e., the closest large central place is an SMSA,
indicates that small cities remote from SMSA's have little
impact on reversing high negative net-migration rates in
nearby counties. Furthermore, this suggests, as does the
negative relationship between largest urban place and net-
‘migration, that large towns and/or small cities do not
serve as central places in the fullest sense of not only
providing a wide variety of services but also attracting
migrants to surrounding areas because of such services.
One explanation for such findings and their implications
for population policies will be discussed in the conclud-

ing'chapter.

CHAPTER V
CONCLUSIONS

The Ecological Model

Hawley's ecological model posits with respect to
migration that although the competitive process is the
primary factor leading to either territorial or structural
differentiation, environment, population, and secondarily
individual technologies have a direct yet weaker influence
on the differentiating process. Because competition occurs
within an organizational structure which sets the rules of
competition, the model has been modified and organization
has been posited to have both a direct and a stronger
impact upon differentiation than does environment.

Utilizing the modified model, this study focusses
on that part concerned with both the direct and relative
influence of organization and environment on migration or
the process of territorial differentiation. Specifically,
it is posited that the higher the level of organizational
diversity and the higher the level of environmental
resources, the more positive will be the net-migration rate.

Results indicate that when considered alone, both
organization and environment have a direct impact on
net-migration with the direction of association between the

95

96
majority of indicators of each and migration supporting the
‘model. Lastly, among both sets of factors, that variable
which measures diversity of structure, either of or easily
accessible to a population, correlates most strongly with
the dependent variable.

When the two sets of variables are combined utiliz-
ing stepwise multiple regression and partial correlational
analyses, the relationship between organization and net-
migration is maintained while the influence of environment
becomes almost negligible. On the other hand, the high
association between the environmental variable SMSA and
organizational diversity indicates that environment may
affect migration indirectly through its impact on organiza-
tion. That is, where a population is located may be either
more or less conducive to the development of a sustenance
structure that will in turn have a direct impact on net-
migration. This study then indicates a need to modify the
ecological model by positing that organization has a rela-
tively strong direct effect on migration while environment
affects territorial differentiation indirectly through its

effect in bringing about organizational diversification.

Methodological Considerations

Methodologically, the correlations between several
variables and net-migration in this study point out the
difficulty in constructing variables meant to measure one
particular aspect of only one component of the ecological

complex. For example, among variables of change results

97
suggest that change in size of largest urban place and in
percentage institutionalized may also be measuring aspects
of population in addition to organizational change while
change in farm size may more accurately represent organiza-
tional rather than environmental change. Somewhat dif-
ferently, the negative partial between percentage on public
relief (transposed) and net-migration corresponds to the
negative associations between most government-related
variables and the dependent variable. This suggests that,
although percentage on public relief still measures a facet
of organization, it may more accurately represent the level
of local governmental welfare services in a county than
sustenance level.

This study has also experimented with several
nonemployment based variables, some of which have been
used before and others that have not. Two measures of
manufacturing not based on employment are incorporated
in addition to the more commonly used percentage of manu-
facturing employees. Number of manufacturing categories
correlates relatively highly while number of firms.with 20
or more employees yields a moderate association with net-
migration. However, partial correlational analysis of all
three manufacturing variables and diversity greatly reduces
the influence of the nonemployment based measures. On
the other hand, the greater combined influence of the
two on the relationship between diversity and migration
compared with percentage employed in manufacturing and the

dependent variable indicates that the two nonemployment

98

based variables measure to some extent linear dimensions of
manufacturing not totally accounted for in the employment
variable. Among other nonemployment based factors, per-
centage on public relief, other farm income and federal
outlays vary in importance among the three analyses, and
change in average farm size consistently yields moderate
relationships with net-migration while none of the other

seven such variables is closely associated with migration.

Areas for Further Research

The results of this study also indicate the need
for further research in several areas. Despite its rela-
tively high correlation with net-migration and the inclusion
of several aspects of manufacturing, the particular index
of diversity used in this study apparently is not sensitive
to all structural nuances. This is most clearly illustrated
by both the increased partial correlation of college pOpula-
tion and the lesser impact of diversity in the urban sub-
sample. Such results indicate the need for further explora-
tion in two directions. Specifically, since no set of
variables comparable to those measuring manufacturing
represents retail trade although ahmost half the more

diversified counties in the sample could be characterized
as being engaged in such activity, further investigation
of the index should be undertaken encompassing various
facets of both manufacturing and trade in nonmetropolitan
counties. Moreover, the results indicate, similar to
Groth's findings, the continuing need to test different

types of indices until one is constructued that both

99

incorporates other aspects of diversification, such as
colleges, not tapped by this index and is equally useful
at all levels of urbanization as measured by nonagricultural
employment.1

Although it is expected that different independent
variables will correlate with the dependent variable at
different levels in any study of this type, comparison
among static and processual variables of organization and
environment suggest two specific areas of inquiry. The
differential influences of static organizational variables
and their counterparts measuring change emphasize the need
for more rigorous examination of the interrelationships
between the two. Specifically, it may be, for example,
that starting at a base of 15 per‘cent, a 10 per cent
increase in manufacturing employment may have a different
impact on the migration rate than starting at a base of
35 per cent; thus, research should focus on determining
if and where critical points of changing influence exist.
The dissimilar impact of environment in the two types of
combined regression and partial correlational examinatiOns
also indicates the need for a clearer understanding of why
environment measured at one point in time has a negligible
influence on migration while change in environment has a
relatively strong impact, although as noted earlier this
difference may be due to the dissimilarity of measures

used in the study.

 

1Groth, l9.

100

In addition, among static organizational variables
a general pattern emerges where migration is associated
‘most strongly with that variable measuring structural
diversity most completely followed by those measuring
‘manufacturing while other factors representing other
specific areas of diversification correlate at lower levels.
Such results could reasonably be expected. On the other
hand, no comparable rationale exists for explaining the
dissimilar correlations between various environmental
variables and net-migration underscoring Hawley's comment
that environment is so diffuse it needs redefining for
every investigation. Moreover, until the various facets
of environment represented in this study and others are
explored more thoroughly and delineated precisely, the
theoretical utility of this component of the ecological
complex will remain weak.2

The generally negative associations between
government-related variables and net-migration indicate
that the causal relationship between public or governmental
services and population change through migration also
requires clarification. In particular, the consistently
negative relationship between percentage in local govern-
ment, measured in mid-decade, and migration suggests that
the direction of association may be related to the relative

inflexibility of government to expand or contract in response

to migration rather than increases or decreases in

 

2Hawley, 1967, 330.

lOl
governmental employment opportunities affecting migration.
Yet, the small positive simple correlation for percentage
in public administration and education in 1960 suggests
that government activity does affect the dependent variable.

This apparent discrepancy can be resolved in at
least two ways. First, the process of migration may yield
differences in the local government employment to population
ratio that are only "corrected" after the decadal census
counts. Such an eXplanation could also account for the
negative correlations yielded by three other variables--
federal employment, change in percentage employed in public
administration and education, and federal outlays for fiscal
1970. However, this explanation does not account for the
positive correlation between migration and 1960 employment
in public administration and education which is measured
by precisely that population data which would be used to
make adjustments.

On the other hand, since public administration and
education employment is actually a composite of employment
stemming from all branches of government, it is also
plausible that the whole may be related positively with
migration while at least the local and federal parts are
not. That no variable measures county governmental employ-
‘ment and the only measure of state inputs refers to revenues
which also includes monetary inputs other than payrolls
preclude further analysis here. However, this explanation

does not account for the negative correlation between

102
change in percentage employed in public administration and
education and net-migration which should be positive if it
were totally valid. Thus, both these tentative explanations
of government-related factors and the variables themselves
require further examination.

Diversification into the establishment of resort
or recreational activities is represented in this study
by number of both hotels and amusement places although
neither correlates above .25 with net-migration. Moreover,
excluding their strong intercorrelation, both correlate
above .50 only with size of largest urban place and number
of manufacturing firms with 20 or more employees among all
independent variables. This suggests that both these
recreational variables may actually tap a dimension of
diversification in towns particularly associated with
medium-sized or larger manufacturing concerns. Thus, it
appears that neither variable represents recreational/
resort activity so much as urban-industrial activity indi-
cating the need for developing other indices to determine
the effect of recreational/resort facilities on population
change through net-migration.

Apparent inconsistencies in the correlations among
largest urban place, nearness to SMSA, and net-migration
suggest another area requiring further exploration. In
this study the simple correlation for the urban (more
diversified) sub-sample as well as the partial correlation

for the entire sample show a negative relationship between

103

size of largest urban place and net-migration. Yet size of
largest urban place and net-migration are both positively
related to nearness to an SMSA for this sub-sample and the
entire sample. In otheruwords,although larger urban places
(as measured by the largest town in each county) tend to

be located in those counties nearer to an SMSA which also
correlates more positively with net-migration, size of
largest place, controlling for diversification, is related
to more negative net-migration rates.

These inconsistencies appear to be the result of
changing patterns of urbanization or small town growth
stemming from changing patterns and modes of transportation.
Specifically, Lemon notes that in the eighteenth century
fourth-order central places were located on either major
roads or navigable streams and served primarily as commercial
centers while the preemption of transport and commercial
functions by Philadelphia prevented the growth to fourth-
order centers of towns near that city. In the twentieth
century Irwin finds a small positive correlation between
interstate location and county population growth comparable
to that found in this study between interstates and migra-
tion while the U.S. Department of Agriculture's report on
rural America in the 19703 indicates that such roads are

particularly important in explaining growth of Southern towns3

 

3Lemon, 502-3, 510; Irwin, 9; U.S. Department of
Agriculture, p. 22. The correlation between interstates and
net-migration in this study has not been discussed so far
because it is below .30 and does not add at least one per
cent to the explanation of variance.

104

In addition, Warner argues that transportational
changes after 1920 have affected the physical form of the
city and by inference the pattern of urbanization with
regard to smaller centers. That is, the increasing availa-
bility of the automobile after that date has made living
in nearby towns and commuting to jobs in large cities more
feasible inducing the growth of towns near metropolises.
MOreover, the introduction of the motor truck in the first
thirty years of the century and the development of the U.S.
Route System in the 19203, primarily through improvements
on existing roads, provided an alternative to railroad
transport of freight. The railroad system, however,
remained intact until after 1948 due to the delay in con-
struction of a fuller highway network prompted by the
Depression and World War II. Thus, after 1920 and
particularly after 1948 the transport function of regional
railroad centers declined. Finally, Warner suggests that
perhaps the most important ramification of this change in
transportation was the lengthening of distance and the
lessening of costs involved in short-haul freight transport
as trucks superceded handcarts and horse-drawn wagons in

cities. This in turn allowed manufacturing firms to enjoy
greater spatial freedom in regard to location beginning

in the 19203 and again especially after 1948 with the
improvement of roads and highways within and around

‘metropolises.4

 

4Sam Bass Warner, Jr., The Urban Wilderness, A
History of the American Cipy (New York: 1972), pp. 113-9.

 

 

105

Such changing patterns and modes of transportation
can resolve the apparent discrepancy concerning the negative
partial correlation between size of largest urban place and
net-migration as well as the negative simple correlation for
the urban sub-sample contrasted to the positive relation-
ships between nearness to an SMSA and both migration and
largest urban place. Specifically, it may be that although
towns nearer SMSA's tend slightly to be larger than those
farther from such centers. some more remote urban places
may still be larger than others closer to SMSA's due to
their earlier importance as railroad and commercial centers.
That is, although the largest urban place in counties at
the fringe of metropolitan areas may, on the average, be
larger than those in more remote areas, the largest non-
‘metropolitan urban places may lie in remote areas; moreover,
such towns which served as more diversified central places
during the railroad era may be losing inhabitants through
migration with the decline of the railroad and thus their
loss of attractiveness to other industries. Both the find-
ing of this study that counties with smaller urban places
are growing more rapidly or losing less of their populations
through net-migration and Zuiches' results indicating that
smaller towns tend to attract proportionately more in-
‘migrants than larger places support such an explanation

that large remote centers and their hinterlands, i.e.,

106

the county, in general have declined with changing trends
in transportation.5

The strong partial correlation between percentage
in colleges in this examination as well as the positive
relationship found by Zuiches between colleges and intra-
state in-migration, however, also indicate that some smaller
remote towns may attract migrants by serving as training
grounds for the development of general "urban" and particu-
larly college-related skills which will be transferred to
jobs in metropolises at a later stage of individual migra-
tion. Thus, the tendency for more remote towns to gain more
in-migrants than less remote towns between 1955 and 1960
while such counties in the North Central Region in the
19603 lost rather than gained population through net-
migration may result from those places continually attract-
ing students to college or university facilities located
there yet not being able to retain them upon graduation.
MOreover, the seemingly contradictory correlations may also
be due in part to the fact that the measurement of in-
migration includes only those entering a county while the
component of out-migration in the net-migration rate

includes both former in-migrants and county residents

that leave.6

 

5The decline of remote towns as central places due
to transportational changes would also explain the negative
beta weight for nearness to city; that is, small cities
remote from SMSA's do not provide the central place services
to attract and retain populations in surrounding counties.

6Zuiches, 410-20.

107

On the other hand, the positive association between
size of largest urban place and SMSA may reflect the greater
feasibility of commuting longer distances to work in cities
or metropolitan areas. The low level of this correlation,
moreover, probably mirrors the fact that once commuting
reaches a high level, the county is incorporated into an
SMSA as well as the continued existence of some large rural
centers of an earlier era. Similarly, the negative correla-
tion between largest urban place andmigration may reflect
in addition to the decline of large rural centers and the
incorporation of high-commuting counties into SMSA's, the
increasing spatial flexibility of commuters in continguous
nonmetropolitan counties due to the automobile and an intri-
cate system of roads which no longer makes it necessary to
live close to commuter lines or shopping facilities.
Furthermore, the strong positive correlation found between
diversity and SMSA similarly demonstrates the movement of
manufacturing firms to outlying counties as more multi-
lane highway networks link these counties to the metropolitan
area. Finally, although the various results suggest such
an explanation they also indicate that the relationships
may be extremely complex. Thus, a more precise understand-
ing must await the undertaking of longitudinal studies
addressed specifically to how transportational changes,
urbanization, and population redistribution relate with one

another on a state or regional basis.

108
Policy Implications

 

The results of this study are similar to the findings
of the Commission on Population Growth and the American
Future and indicate that its policy recommendations concern-
ing depressed rural areas can be implemented without great
difficulty. On the other hand, findings also show that
current federal monetary policies, if anything, promote a
population distribution pattern contrary to the Commission's
recommendations. Specifically, counties more remote from
metropolitan areas tend to have more negative net-migration
rates than those closer to SMSA's. However, as farm con-
solidation continues, these findings demonstrate that heavy
population losses from rural areas can be stemmed when
employment opportunities, particularly in manufacturing
and secondarily in public services, are available. That
this study indicates that particular counties have already
diversified away from agriculture and have also stemmed
the tide of heavy rural out-migration suggests that it
is quite feasible to pinpoint those areas that have dis-
played the potential for growth and could be developed
further in line with the Commission's recommendation of
employing a growth center strategy for such places in
depressed areas. The promotion of selected centers as
governmental policy would not only enhance employment
opportunities but also would concentrate public services
in such towns/cities making them central places in the

fullest sense of providing greater services for surrounding

cent
and

RICE

ian
poli
cont

larg

CORC

cone
EOVe:
that
10am:
if CI
of 11‘
an as
of c
at we
Badge
tion :
promoz

0t IQr

H
O

109
rural residents. Such a concentration may also make such
centers more attractive to residents and potential migrants
and thus alleviate pOpulation pressures on metropolitan
areas.

The findings of this study concerning federal
inputs, however, demonstrate the lack of any such growth
policy by the federal government at this time. That is,
controlling for the level of manufacturing and size of
largest place, federal inputs do not vary in relationship
to migration trends. Moreover, that such monies tend to be
concentrated in agricultural areas yielding the highest
losses through net-migration suggests that aiding farmers
has a negative effect on making the most rural areas attrac-
tive residential locations and underscores the lack of
concern for population distribution problems by the federal
government in its monetary policies. This is not to argue
that the Midwestern farmer should not be aided through
loans and cash payments for crOps but to point out that
if this country is committed to producing a higher quality
of life for its citizens through population redistribution,
an assumption the Commission maintains, the current pattern
of fund allocations is inadequate at best and detrimental
at worst to promoting such a policy. Thus, either current
budgeting policies must be redirected or additional taxa-
tion imposed to be earmarked specifically for growth
promotion purposes in specific areas through the building

or repair of existing transportation networks, incentives

110
to businesses to relocate, and the provision of social

services to residents of nonmetropolitan areas.

APPENDICES

APPENDIX I

MEANS AND STANDARD DEVIATIONS

APPENDIX I
TABLE 14

MEANS AND STANDARD DEVIATIONS FOR ALL VARIABLES
FOR THE ENTIRE SAMPLE (N=227)a

 

 

 

 

Mean Standard
Deviation
Variable
Net-Migrationb b .922 .129
Non-Farm Population C nge 1.080 .195
Farm Population Change .841 .483
Or anization
-_%iversity .699 .116
Manufacturing Employment .142 .115
Manufacturing Firms 7.784 11.662
Manufacturing Categories 7.366 4.284
Other Farm Income (t)c .775 .128
College .010 .023
Institutionalized .008 .016
Military .002 .011
Local Government Employment .031 .020
Public Administration .084 .032
Farm Categories 5.247 1.097
Hotels 3.595 7.707
Amusement Places 1.370 4.630
Female Participation .276 .048
Largest Town 5943.9 6900.7
Income Under $3000 (t)C .656 .113
Unemployment (t)c .957 .025
Public Relief (t)c .964 .031
Environment
Nearness to SMSAd 902.180 70.093
Nearness to Cityd 928.079 55.346
Federal Outlays 961.06 654.53
Federal Employment .006 .005

111

112
APPENDIX I

TABLE 14 (Continued)

 

 

 

 

 

Standard
Mean Deviation
State Revertluese 89.25 42.23
Interstate .308 .463
anizational Change
orhange in:
Largest Townb 1.062 .173
Collegeb 1.014 .024
Militaryb 1.000 .013
Institutionalizedb 1.005 .011
Female Participationb b 1.075 .044
Manufacturing Employmfint 1.031 .040
Public Administration 1.021 .029
Land Usef .163 .370
Unemployment (t)c 1.002 .017
Environmental Change
Change in Farm Size (t)d 9909.8 173.7

 

aAllmeans and standard deviations for variables
computed as percentages are given in decimal form.

bDue to negative figures 1.000 has been added to
yield all positive figures for computational purposes.

cBecause of the hypothesized negative correlation
with netdmigration and negative figures, each figure has
been.mu1tip1ied by -1 and 1.000 has been added to yield
all positive figures.

dBecause of hypothesized relationship with net-
‘migration and negative figures, each figure has been
multiplied by -1 and 1000. has been added to yield all
positive figures.

eFigures are in terms of dollars and cents.
fThis is a binary variable.

gAcres have been computed through one decimal place.
multiplied by -1 due to hypothesized relationship with net-

migration, and 10,000 has been added to yield all positive
figures

APPENDIX I
TABLE 1 5

MEANS AND STANDARD DEVIATIONS FOR ALL VARIABLES
FOR THE RURAL (AGRICULTURALLY SPECIALIZED)
COUNTIES (N=167)a

 

 

 

 

 

 

Mean Standard
Dev1ation
Variable
Net4Migration .900 .119
Non-Farm Population Change 1.067 .197
Farm Population Change .786 .139
Organization
Diversity .663 .109
Manufacturing Employment .099 .081
Manufacturing Firms 3.940 5.890
Manufacturing Categories 5.755 3.142
Other Farm Income (t) .802 .109
College .008 .020
Institutionalized .007 .015
Military .002 .008
Local Government Employment .033 .023
Public Administration .085 .028
Farm Categories 5.084 1.020
Hotels 1.707 4.620
Amusement Places .359 2.450
Female Participation .266 .047
Largest Town 3884.6 3806.1
Income Under $3000 (t) .623 .102
Unemployment (t) .962 .018
Public Relief (t) .963 .031
Environment
Nearness to SMSA 892.742 72.969
Nearness to City 919.611 56.203
Federal Outlays 1065.09 695.03

113

114
APPENDIX I

TABLE 15 (Continued)

 

 

 

 

 

Mean Standard
DeVIation
Federal Employment .007 .006
State Revenues 87.71 43.76
Interstate .228 .421
Or anizational Change
ange in:
Largest Town 1.063 .171
College 1.012 .022
Military 1.000 .009
Institutionalized 1.006 .010
Female Participation 1.072 .048
Manufacturing Employment 1.031 .040
Public Administration 1.020 .030
Land Use .162 .369
Unemployment (t) 1.002 .017
Environmental Change
Change in Farm Size (t) 9896.8 185.9

 

aSee footnotes to Appendix I, Table 13.

APPENDIX I
TABLE16

MEANS AND STANDARD DEVIATIONS FOR ALL VARIABLES
FOR THE URBAN (DIVERSIFIED)
COUNTIES (N=60)a

 

 

Mean Standard
Dev1ation
Variable
Net-Migration .983 .137
Non-Farm Population Change 1.119 .185
Farm Population Change .997 .897
Organization
Diversity .798 .062
Manufacturing Employment .259 .118
Manufacturing Firms 18.483 16.282
Manufacturing Categories 11.850 3.835
Other Farm Income (t) .701 .146
College .015 .030
Institutionalized .009 .020
Military .004 .016
Local Government Employment .028 .006
Public Administration .081 .040
Farm Categories 5.700 1.183
Hotels 8.850 11.372
Amusement Places 4.183 7.368
Female Participation .304 .041
Largest Town 11675.6 9810.3
Income Under $3000 (t) .747 .089
Unemployment (t) .943 .035
Public Relief (t) .968 .032
Environment
Nearness to SMSA 928.450 53.709
Nearness to City 951.650 45.598
Federal Outlays 671.56 407.45

115

116
APPENDIX I

TABLE 16 (Continued)

 

 

 

 

 

Mean Standard
Deviation
Federal Employment .004 .003
State Revenues 93.54 37.66
Interstate .533 .503
Or anizational Change
ange in:
Largest Town 1.060 .177
College 1.020 .028
Military 1.001 .019
Institutionalized 1.003 .014
Female Participation 1.083 .029
Manufacturing Employment 1.030 .041
Public Administration 1.022 .024
Land Use .167 .376
Unemployment (t) 1.003 .018
Environmental Chan e
Change In Farm Size (t) 9946.7 84.4

 

aSee footnotes to Appendix I, Table 13.

APPENDIX II

CORRELATION MATRICES

117

Aomscouaooo

 

mom. mom. NoH. ooo. moo. Rho. ooH.- HNo.- AHNV ADV mmflamm ooflnso
Hao. omH.- Noo.- omN.- Nmm.- oao.- omo.- ooH.- Aooo ADV DamsaoHosmao
Boo. oao. Boo. mom. son. ooa. ooo. omH. Aoao ADV ooo.mo Amos: maouaH
Hoo.- woo. an“. mom. one. was. Boo. mos. Aoao pace ummmumq
ooo. mom. Hmo. ooo. woo. omo. Hoo. omo. ARHV aooumoooouumm BHBBBE
ooo. moo. Hao. How. oHo. «no. ooo.- omo. AoHo mmomam uaoammaa<
ooo.- omo. oom. ooH. mom. mom. Non. omm. Away oflmuom
ooH.- moo. ooo. owe. moo. ooo. ooH. oo~. ASHV mmouowmuoo sumo
ooo.- oH~.- ooH.- oNN.- Hoo.- omo. oHH. oHH. Amav aosumuomflcoae< ofiansm
ooa. Hoo.- Noo.- ooH.- ooo.- moo.- mo~.- mo~.- ANHV Damaaofloam Bamacu0>oo Hmuoq
mmo.- moo. moo. oNo.- ooo. ooo. Hao. omo. AHHV AABDHHHz
Nmo.- How. oHN. NHH. omo. ooo. moo. oofl. Aoov omnsamaoflusufioocH
moo. ooo. moo. oHo.- ooo. «Ho. moo. moo. on mmmaaou
ooo.H soN.- moH.- oom.- omo.- ooo.- om~.- am¢.- on ADV maoocH sumo Bogue
ooN.- ooo.H oom. mow. Hoe. HNH. oHN. ooo. ARV mmouowmumo mcwusuommsamz
oeH.- ooa. ooo.H “so. Nam. ooo. ooH. Hom. Aoo mango waousuommacmz
omm.- mos. Rao. ooo.H moo. oRH. Now. on. on Damaaofloem waounuomwaamz
omo.- Ham. Nam. moo. ooo.H oNH. omH. Ram. Aoo AOHBHB>Ao
COHumecmeO
oNo.- HNH. oma. oRH. oNH. ooo.H ooo.- ooH. on swamno coflumaaoom spam
omN.- oHN. ooH. Now. ooH. omo.- ooo.H oao. ANV mmcmno cooomaaoom anom-aoz
“No.- ooo. Hon. oon. Ram. ooH. oao. ooo.H AHV cooomuwoz-umz

 

Amy AAV on Amy Aqv Amv ANV .AHV

 

 

ANNNHZV m4m2<m A<HOH mom mZOHH<Ammmou MMQMOIOmMN

ma mam<H

HH XHszmm<

118

momocmsaooo

 

 

 

NoH.- «mm. ooH. mom. moo. mNo.- moo. mom. Anmv ADV mumm aumm,cm «woman

0 cmzo HmucmEcoum>cm
mmo.- mao.- oNo.- Hoo.- omo. .moo. omm.- mmo.- Ammo ADV Damamomoamco
mmH.- moo. Hmo.- mao. moo. muo.- mmo. ooo. Ammo moo oamg
omo. mmo. NNo. Noo.- o~o.- moo.- Nao. omo.- memo comumuummamao< ammosa
mom.- omo. omo. oNH. omH. Hoo.- mNo. mmo. Ammo Damamomoam mcmunuommsamz
moo. moo. AHA. “mm. ooo. moo. mmo.- «mo. Ammo comumomomuumm wmoamm
5mm. om~.- oam.- amo.- oNN.- ouo.- mmH. omH.- Ammo omnmflmcomusuoumcm
moH.- mao. mRo. oom. moo. moo. omH. moo. Aomv mumummmz
moo. mom. woo. woo. mom. omo. omm. mom. Ammo mwmmmoo
oH~.- mom. who. “mm. mom. Hmo.- omo. mom. AmNV mace Dmmwumq
"aw mwamao

mmwcmnu Hmoomumumcmwmm
moo.- omm. Nam. mom. mom. omo. ooo. mom. ARNV mumumumuam
mm~.- mmo. omo.- Hmo. ooH. Nmo. mHo. moo. Ammo mmscm>mm momum
ooo. omm.- mom.- oRN.- omN.- moo.- ooo.- NoH.- Ammo OsmamoooEM Hmumomm
«mo. mmo.- oom.- omo.- omm.- nmo.- mom.- oum.- memo mmmmuoo Hmumomm
moo.- omo. ooo. omo. mmo. moo. mom. oom. Ammo momo on mmmaummz
ooN.- moo. mom. mom. Rom. moo. mmH. omo. Ammo <m2m ou mmmcummz

ucoacoum>cm

 

va Amy on Amv Aqv Amv ANV Aav

 

 

Anmmuzv mqmz<m A<HOH mom mZOHB<AmMMOU MMQMOIOMmN

396358 2 Sea.

HH xHszmm<

”q x~.fihzu.~n~1~<

119

mooocoocooo

 

 

omo. mom. moo. omo.- omo.- moo. ooo. mom. Ammo moo ommmom ommoso
ooo.- BAH.- Nom.- ooo.- omo. AHH.- ooo. Hmo. Aouo Aoo Dcosmomoaoc:
omo. mom. mom. Roo.- HHH.- mom. omm. mom. AoHo moo ooo.mm Amoco oaooam
oom. mmo. ooo. omo. ooo.- mmN. omo. moo. Ammo czom ooomuoa
ooo. mom. mom. ooo.- ooo.- ooo. «ma. ooo. Ammo comuoomumoumo oaoaom
ooo.H on. ooo. omo.- ooo.- moo. mom. ooo. AoHo oooomo unmamo9a<
omm. ooo.H mom. Now. ooo.- mam. mom. omo. Ammo Amuse:
omo. mom. ooo.m ooo.- mmH.- omo.- mom. ooo.- memo oomuowmuoo atom
omo.- Now. ooo.- ooo.m ooo.- ohm. mmo.- mom. Ammo aofiumuuoocmao< oomoao
ooo.- ooo.- RoH.- ooo.- ooo.H Hmo.- moH.- ooo.- Ammo ocosmomoam ucoecuo>oo Hones
moo. mam. ooo.- ofim. omo.- ooo.H ooo.- mom. Ammo muooommz
mom. mom. mom. amo.- moH.- ooo.- ooo.m moo. Aomo omummocomosooumam
ooo. omo. omo.- mom. ooo.- mom. Hoo. ooo.H moo owmmmoo
omo. omo.- ooH.- ooo.- ooo. mho.- omo.- moo. Aoo ADV maooam spam umouo
ooo. omo. mmo. om~.- Hoo.- mom. omo. ooo. ARV omouommomo mamasoomoscoz
mno. oom. mom. ooo.- ooo.- ooo. omo. ooo. moo AEDHE wamusuomoncmz
moo. ooo. omo. ooo.- ooo.- omo.- Nam. omo.- Amo Damamomoam wcmuaooooscmz
omm. moo. mom. ooo.- ooo.- ooo. omN. ooo. Aoo mumoum>mo
comumnmcmwmm
omo. omN. ooo. ooo. moo.- ooo. ooo. omo. Amo owamoo cooommsooo 80mm
ooo.- omo. ooo. «Hm. moN.- omo. mmo. moo. ANV owaooo nomoomsoom auom-coz
omo. omN. ooo. mom. mo~.- omo. ooo. ooo. moo comoouwmz-umz
momo Ammo memo Ammo Amoo Ammo AoHo moo

 

 

Anmmuzv mqmz<m A<HOH mom mZOHH<Ammmou anmOIOmwN

826388 2 mama.

HH Nanmmm<

I . R-Zvnlnllllu‘

120

Aoosomocooo

 

 

 

 

omo. moo. mam. mam.u Hmo.u omo. moo. omo. AmmV ADV oumm shah a“ owcmno
ownmno Hmucoacoum>cm
omo. «no.1 moo. Hoo. omH.n HHo.n Hoo.n ooo.: AomV AuV uaoaxoaoamcs
omo.- ooo.: NHo.- ooo.- NmH. omo. omo.- moa.n Ammo om: mama
noa. mum. ooo.: omo. moo. oom. «ma. onm. Aqmv comumuuwmcmafi< omandm
mma.u oH~.u Hoo. oNN.u who. oHH.u mHH.u mmN.u Ammo ucoaooamam wcmuouommsamz
ooo. ooo. omo.- ooo. mmo. ooo.: ooo. Hmo. Ammo comummmomuumm mamamm
NoH.u Noa.u wHH.u qoa. moH. Noo.: ooo.: omo.- AHmV omnmamcowusumumcH
HoH.u now. omo. oqm. mHo.n ~m~.u ooo. mom. AomV zumumawz
moo. ooo. Hmo. oaq. omo.: NNH. woo. woo. AoNv owoaaoo
ooo.- mmH. moo. Nam. owa.u Hoo. moo. omo. Awmv c308 umowumq
”amowcmmmll
mowamsu Hmcomumumcm Ho
HHN. How. Hwa. omo. mmo. moH. «ma. mmH. Anmv ouwumuoucH
ooo.- moo. moH. ooo.- moa. mmo. ooo.- oma.u Ammo mosco>om mumum
NHH.1 Hmo.- Noo.- mom. wmm. mmH. mma. ooo.- Ammv ucoshoaoam Hmuooom
Hoa.n oma.- Nam.u moo. qu. qu. oNH.u ooo.- Aqmv mmmauso amuooom
oqm. NmH. Hum. oma.u mna.: omo. omH. BNH. Ammo oumu ou mmoaHmOZ
oma. omo. mom. oa~.- moH.n Nmo. NNH. moo. ANNV «mzm ou mmocumoz
uaoauoum>cm
AoHo Ammo Aomv AmHV Ammo AHHV AoHo Aoo

 

 

AmNNuzv mum2<m A<HOH mom monH<Ammmoo MMQMOIOMMN

Aomsamucoooao mqmoe

HH anzmmm<

121

AooscAocooo

 

mmo. ooA. mmo. ooo.A moN. oom. NoA. on. AANV ADV omAAom 0AAosm
ooN. mNo.- ooo.- moN. ooo.A «No.- Amo.- Noo. AoNo ADV ocmamvoamao
on.- NoA. Aoo. oom. oNo.- ooo.A ooo. Nmo. Avo Ago ooo.mo Amoco oaoocA
ooN.- moo. NmN. NoA. Amo.- ooo. ooo.A moo. Avo c309 “momumA
Noo.- oNN. ANN. oNo. Noo. Nmo. moo. ooo.A ANAV coAuovoAuumm onaom
AoA.- ooN. oNA. ooA. mNA.- mmo. oom. ooN. Avo omooAm osmamosao
oNA.- NoA. omo. mmA. NNA.- Noo. mmo. mom. Aon vouom
NAo.- Aom. Nom. Noo. NoA.- AoA. mNN. NoA. Avo omAAomouoo sumo
Noo. omA.- oAN.- omo.- Noo.- Noo.- omo. Noo.- Aon BOAomAAoAsAao< UAAAsA
moA. NAA.- AoA.- ANN.- vo. AAA. Noo.- Aoo.- ANAV AcoENOAoam uamaauo>oo AoooA
Aoo. omo. Nmo. Aoo. NAA.- moA. mmN. ooo. AAAV AquAAAz
oNA.- omA. NNA. ooo. Noo. omA. omN. omA. AoAv ooAAAocoAusoAuocA
ooo.- ANA. moo. mmA. Amo. AoA. moN. ooN. Aoo mwmAAoo
NoN. mmo.- ooN.- mmo. ANA. Noo. Amo.- omo. Amo ADV maooaA sumo Amouo
mmo.- ooo. Noo. ANA. moA.- oNo. ooo. mom. ANV omAAowmooo chusoommsaoz
mmN.- ooo. omo. NoA. Noo.- Noo. NoN. ANA. Aoo mauAm chusAomosamz
omo.- ooo. Aom. ooo. omN.- mom. mom. ooN. Amo ocoaavoam wcAusuuoosaoz
oAm.- mmo. Nom. omo. Nmm.- ooo. omo. ooo. Aoo AvoAm>Ao
comumumcmwmm
Noo.- moo. on. NNo. oNo.- omA. mAA. omo. Amo mwamoo cvooAsoom sumo
ooo.- ooA. omA. ooA.- omo.- ooo. Noo. Aoo. ANV omcooo cOAooAsoom auom-coz
on.- ooo. vo. ANo.- ooA.- omA. omA. omN. AAV EOAoouwAz-omz

 

AoNV Ava ANNV AANV AoNo Avo Avo ANAV

 

 

 

 

ANNNuzo mAmz<m A<Aoe mom ozoAmoAmmmoo mmnmo-ommN

€263.88 NA ”39%

AA onzmmm<

122

Aooovocooo

 

 

 

 

qoq.1 own. moo. omo. Hoo.1 woo.1 omH. mom. Ammo Auv oNHm Show CH owcmao
«wwmso HmucoscouH>cm
omo.1 Hoo.1 mmo. NNH. mm¢.1 mHo. mmo.1 mmo. Homo Auv ucmB%OHoBocD
mmo. 1 oHo. NHo. HoH.1 «no.1 NoH.1mmo.1moo.1 Ammo mm: oon
«oo. 1 ooo.1 Noo.1 omo. ooH.1 moH. mHH. mmH. Aqu coHuoHumHSHat< OHHnnm
me.1 qu. owH. NHH. 1 mmo. on. 1 ooH. 1 ooo. Ammo ooo850Hoam onunuommscmz
omH. woo.1 woo. qu. moo. mmN. ooo. mom. 1 AmmV coHumoHoHuumm OHmBmm
ooH. mmH.qNH.1 ooo.1 omo. moH.1 moH.1 omo. 1 AHmv wouHHmaomuouHumcH
ooo. mmo. 1 Hoo. omo. 00H.1 omo. «co.1 ooo. AomV oumuHHHz
omA.- mAA. omo. omA. vo. mNN. ooo. moN. AoNo ommAAoo
NoH.1 mmo. qu. NNH. 1 ooo. ooo.1 omo. HoH. Ammv c309 umowumg
.pm owcmWWIl
mowcmso HmaoHumNHam Ho
HmH. 1 mom. «mm. mmo. mmo.1 NoN. mHm. mom. Homo oumumuouaH
«Ho. 1 noH. 1 omo. Nmo.1 oom.1 moo. wHH.1 mmo. Aomv monoo>om oumum
mom. HwH.1 mmN. omH.1 qu. NNH.1 on.1 HoH.1 Ammo ucoa%OHaEm Hmuooom
ooo.H oo¢.1 oqq.1 mmo. oom. mHo.1 wom.1 mwm.1 Aqmv momHuso Hmuooom
oo¢.1 ooo.H mom. ooH. mmo.1 NoH. mom. onm. Ammo KnuHo ou mmocumoz
o¢¢.1 mom. ooo.H moo. ooo.1 Hoo. Now. HNN. Ammo <m2m 0» mmocumoz
ucoacouH>am
AoNv AmNo ANNV AANV NoNo AoAV AmAv ANAV

 

 

ANNNHZV MHmz<m H¢HOH mom mZOHH<Hmmmoo mmnMO1omMN

935988 NA 5mg

HH anzmmm<

H

-

VfihA~ZH.—1~uh<

123

AomocAucooo

 

 

Aoo. ooo.- Noo. omA. NNA.- omo. Nmo.- omA.- AANV ADV ooAAom UAAABN
moo. omo. ooA.- vo. omo. Noo.- ooo.- vo. AoNo Auo Aaoaovoamco
NmN. NoA.- omo. mNN. ooo.- NoN. moo. NNA.- Avo AAV ooo.om Amoco oaoocA
ooo. moA.- ooo.- oom. omo. NAm. mAA.- ooA.- Avo coco Aoowqu
ooo.- omo.- ooo. moN. AoA. ooN. Nmo. AoA.- ANAV coAuovoAAAoo voaom
ooo. NoA.- AoA.- moo. ooo.- AAN. ooo.- NAA.- Nvo omovo Damamo:a<
ooo. NoA.- NoN. ooo. mNA. AoN. moo. Amo.1 Aon vooom
Nmo.- mAA.- Nmo. Aoo. moo. AoA. moA. Noo.- Avo ooAAoooomo sumo
ooo. ooA. ooN. vo. NAN. oNo. Noo.- NoN. Aon EOAAoAAoABAao< 0AAoso
moo. NoA. vo.- Noo.- ooA.- mmo. moA. oNN. ANAo oaoaNvoam ocmaa0o>oo AoooA
ooo.- Noo.- NoN.- NNA. Aoo. moA. mNo. moA. AAAo NAouAAAz
ooo. ooo.- ooo. Noo. mNo. ooA. ooo.- NmA. Avo ooNAAmcoAusoAuocA
Aoo. mmo.- moN. NoN. omo. mmA. oNA.- ooo.- Aoo mooAAoo
ooo. NmA. moA.- Noo. oAN.- on.- ooN.- ooo. Aoo Nov oaoosA spam Aoooo
ooo. omN.- mNo. NoA. NNA. ooo. mNo. ooN.- ANo ooAAommooo maAusoomosaoz
NAA. oNA.- oNo. NoA. NNo. NAm. vo.- ooA.- Aoo oaAAm wcAusoomoacmz
NmA. NoN.- ooA. Noo. NoA. moN. Aoo. ANN.- Amo DaoaNvoam waAAauoooacoz
oNo. oNN.- ooo. moA. mNA. NoN. ooA. omN.- Aoo NAAoAm>Ao
aoHuMNHcmwuo
ooo. oNo.- Noo. vo. Aoo.- vo. Noo. vo.- Aoo omaooo cvoonooo sumo
mmo.- omA. ooA. omN. ooo. ooo. on. ooo.- ANV mwaooo BOAAoAsooo anom1aoz
Noo. oNA.- omN. mom. oNo. AoA. moo. NoA.- AAo BOAAoAmA21omz
ANmo AAmo Aomo AoNo AoNv ANNV AoNo AmNo

 

 

AnNNuzv MHmZ<m H<HOH mom monH<Hmmmoo MMQMO1OMMN

AomsaAuaooo NA Soon.

HH xHazmmm<

Viv ~ Anzu.~1~a.~f\

124

AooscAocoov

 

NNN.- NNo.- vo. ooA. ooA. oAA. omA. mmo.- ANmo Aoo omwm1mmom cm owaooo

m fimno H mquECO RH >Cm
ooA.- omA. oNo.- AAo.- NAA.- Nmo.- Noo. oNo.- Aooo on oaoaNvoaoao
NNo. ooo.- ooo. vo.1 NmA. omo.- Aoo. ooo. Ammo moo oamA
moo. ooA.- moo. NAA. ooo. NNo. oNo. ooN. Aomo avooAqucAao< 0AAono
Nmo. mmo.- Aoo.- ooN.- omo. vo.1 ooA. Noo.- Ammo AcoaNvoam chusuomoscmz

ooo.A ooA.- ooo.- mAA. omo. ooo. Noo.- omo. ANoo BOAAmvoAuumo oAmamm
ooA.- ooo A Noo.- Amo.- mNA.- ANo.- omo. ooo. AAmo omuAAocoAunuAuoaA
ooo.- Noo.- ooo.A omo. Aoo. Aoo.- Nmo. ooo. Aomv ououAAAz
mAA. Amo.- ooo. ooo.A Nom. AoA. ooA.- ooo. AoNo owoAAoo
omo. mNA.- Aoo. Nom. ooo.A Noo. NNo.- Noo. NoNo pace “mowqu

“CH owcmmw
mowcmno HmcoHumuHcm no

 

Ammo oumumuoucH
Homo mosco>om oumum

omo. Hmo.1 Hoo.1 HmH. moo. ooo.H omo. ooo.
Noo.1 mmo. mmo. qu.1 550.1 omo. ooo.H qu.

 

omo.- ooo.- moo. ooo.- Noo.- ooo.- Aoo.- ooo.A AmNo DamaNvoam Aouooom
omA. ooA. ooo. omA.- NoA.- AoA.- vo.- ooN. AoNv oNvoso Aoumoom
ooo.- ooA.- Noo.- mAA. oNo. omA. NoA.- AoA.- AoNo NAAo on oomcuooz
Noo. oNA.- Aoo. omo. Noo. oNN. oNo. oNN.- ANNV «mzm 0A ommcuooz
ucoacouH>cm

 

33 33 Homo 33 $3 93 $3 3.3

 

 

ANNNuzo mAmz<m A<Hoo mom szAmoAmmmoo mmom01ommN

AomscAucoov oAmAm<e

HH NHszmm<

> ,h Ii .1. 1: en Ah 4‘

125

Aooovocooo

Nmo. NNA. AoA.- oNo. NAA.- AANV ADV omAAmm UAAoso
Aoo.- NmA.- ooo.- ooA.- ooo. AoNo ADV ocmaovoEmao
ooo.- on. NoA.- moA. ooA.- Avo ADV ooo.om Amoco maooam

 

 

oNA. omo.- mmo.- mAA. ooA.- Avo agoa ooowqu
NNN. omo. Noo.- mmA. ooo. ANAo BOAUmvoAAAmo voaom
NNo. oNo. mmo.- moA. omA.- Nvo ooovo “cosmosa<
moo. Noo.- ooo.- oNN. oAN.- Aon vooom
oAm. Noo. NAo.- ooo.- Aoo. Avo omAuommuoo spam
oAN.- Aoo. ooo.- oNo. oNN.- Aon cOAAoAqucAaoo 0AAoao
ANo.- ooA.- NmA. ooo. mNo. ANAo ocoaNvoam ocmaauo>oo AoooA
omo. AAo.- mNo. ooN. oAA.- AAAV ououAAAz
ooo. Aoo.- omo.- omA. AAA.- Avo omNAAmcoAuooAumaA
mmo. Noo.- moA.- oNN. ooN.- Aoo owmAAoo
NoA.- mmo.- mmA.- ooo. moA.- Nov ADV m585 anon Aoooo
oom. mNo.- ANo. omo. ooo. ANV ooAuoooooo waAAsAommsaoz
ooA. oNo.- Aoo.- NNo. vo. Aoo oaDAm wcAusuooosamz
mNo. Aoo.- oNo. Noo.- oNA. Amo oaoaNOAoam waAAsooooscmz
moo. ooo. ooo. vo.- ooA. Aoo NAAouoooo
coHumuHmemm
mNo.- ooo. mNo.- vo.- Aoo.- Nov mocooo BOAooAsooo sumo
NoN. oNA.- omo. NNo. oNo. ANo mocooo cvomAsooo apom1coz
Aoo. omo.- ooo. vo.- moo. AAV aOAAoAmA21Amz
ANmV Aomo Ammo Aomo Ammo

 

 

Ammwuzv mHmz<m H<HOH mom monH<Hmmmoo meMO1OMMN

AoosaAAaoovNA mAm<a

HH NHszmm<

l

l

Al’l'I-INI) l X

126

 

 

 

 

 

ooo.H ANo.- ooo. mmo.1 NNA. ANmV ADV mNAm Show cw mwamso
w CNSU H mquEOHHNFHm
ANo.- ooo.A vo.- mNo. AoA.- Aoov ADV ocoaNvoamao
ooo. vo.1 ooo A ooo.- oNo. Ammo moo oaoA
mmo.- mNo. ooo.- ooo.A oNN.- Aomo :oAAmAquBAeo< 0AAo=o
NNA. AoA.- oNo. oNN.- ooo.A Ammo unmsNvoam wcAusuoooncoz
NNN.- ooA.- NNo. moo. Nmo. ANov aoAuooAOAAAmo onamm
NNo.- omA. ooo.- ooA.- mmo.- AAmo ooNAAocOAusoAuocA
vo. oNo.- ooo. moo. Aoo.- Aooo NAooAAAz
ooA. AAo.- vo.- NAo. ooN.- AoNo ommAAoo
ooA. NAA.- NoA. ooo. omo. AoNo ESom oomwumA
"am owama
mowfimfio HmGOHumuHcm no
oAA. Nmo.- ooo.- NNo. vo.- ANNV ooououoocA
omA. Noo. Aoo. oNo. ooA. AoNo monso>om oomoo
mom.- oNo.- ooo. ooN. Noo.- AmNo AcoaNvoam Aouoooo
ooo.- omo.- Noo.- ooo.- ooA.- AoNo oNvoso Amuoomm
oom. Aoo.- vo. ooo.- NoA. AoNo NAAo 0“ ooocuooz
Noo. moo. NAo. Noo.- ooA. NNNo <mzm oo oooapooz
ufimEflohH>cm
ANmo Aomo Ammo Aooo Ammo

 

Anmmnzv MHmz<m H<HOH mom mZOHB<Hmmmoo MMQMOIOMMN

325388 NA 5mg

HH XHQmem<

127

AomacAocooo

 

mNo. moo. omo. oNA.- NNo. moN. on.- mNA.- AANV Aoo ooAAmm oAAoso
oom. omN.- moA.- oom.- mmm.- NNA.- Noo.- Noo.- AoNo Aoo ucoaNOAoaoco
AoN. omN. moN. NoA. oNN. oom. ooo.- ooo. Avo Aoo ooomm Amos: maooaA
Amo. Nom. NoN. Nom. ooo. oNA. Aoo. mNA. Aon mace AoowpoA
ooo. omm. NoN. ANA. Noo. mNo. mNo. moA. ANAV BOAAmvoAoAmo onsmm
mNo. omm. ooo. oAN. AoA. Aoo. moo. Nmo. Avo omuon Dcmamosao
oNo. Nom. mNm. NmA. mmN. moN. moA. omA. Aon vouom
moA.- oNN. omN. ooN. NoN. moo.- mmA. oAN. Avo ooAAomoumu anom
Aoo.- AmN.1 NoA.- ooN.- mmo.- NAA. Noo. oNo. AmAV coAuouooaAeo< 0AAoso
NoN. Amo. Amo. NmA.- ooo. Nom.moN.1 oAm.- ANAV DamaNvosm unmaauo>oo AmooA
oAA.- moA. ooo. Nmo. om9 NoA. moo. ooA. AAAV NAouAAAz
NoA.- NoN. oNA. ooA. ooN. mmA. ooo. ooA. Avo ooNAAABOAoauAAoaA
Aoo. oNo. Noo. on.- oNo. NoA. moN. NoN. Aoo mooAAoo
ooo.A NoN.- oAA.- on.- ooo.- moo. omN.- oNo.- Amo ADV oaooaA sumo Aoouo
NoN.- ooo.A ooo. omo. oAN. NoA. oNA. moo. ANV ooAAomoooo maAunooooscmz
oAA.- ooo. ooo. A NAm. ooo. mmA. moA. oNN. Aoo mauAm chAsuoomsaoz
on.- omo. NAm. ooo.A oAN. moA. oNN. Nmm. Amo ucoamvoam wcAusuummscoz
ooo.- oAN. ooo. oAN. ooo.A ooo. ooo. mom. Aoo mAAoAm>Ao
coHumuHcmmll
moo. NoA. mmA moA. ooo. oo9 A ooo. oNN. Aoo omcooo aoAAoAsooo sumo
omN.- oNA. moA. oNN. ooo. ooo. ooo.A mNm. ANo omcmoo BOAumAsooo anom1coz
oNo.- moo. oNN. Nmm. mom. oNN. mNm. ooo.A AAo avooAmA21ooz

 

Amo Amv on Amv Aqv Amv ANV AAV

 

 

Amonzoou NoAV mAmz<m1mom AomNAAvomom
AAA<movooAmo<o A<mom mom mzoAa<Ammmoo mmom01ommN

mH MHm<H

HH xHazmmm<

128

AooscAocooo

 

 

 

 

mNA.- ooo. AAN. omm. omo. oAN.- oNN. omm. ANmo Aoo oNAm sumo aA omcmoo
oMCmno Hmuaoacoum>cm
mm9 oNo.- moo.- on.1 ooo. moo. mNo.- Noo. Aomo Aoo oamaNvoaoco
Nm9 1 ooo. ooo. NAo. moo. moA.- Nmo. omo.- Ammo woo oaoA
mmo. ooA. mNo. moo.1 ooo.- oNo. oNA. vo. Aomo aOAuonoAcAao< 0AAA=A
ooA. 1 oNA. mNo. oAN. moA. mmA.- mNo. NoA. Ammo Damamvoam waAunuoooaamz
oNo. mNo. mmo. AoA. ooo. mAN. mmo.- ooA. ANmo avomoAUAuuoo «Anson
ooN. Nmm. NoN.- oom.- mmN.- ooA.- omA.- omN.- AAmV omNAAocvosquoaA
ooA.- NmA. oNo. mAA. moo. oAA. ooo. oNo. Aomo NABAAAAz
omo. AmA. AAA. vo. NmA. oAA. omN. Aom. AoNo owmAAoo
mmN.- oAA. mmo. mAN. mmA. NNo. oom. Aom. AoNV caom oomwumA
.GH owcm
«mango HmGoHumuHammuo
Nmo. mNN. mAA. Amo. moA. NNo.1oNo. mNA. ANNo muoomuouaA
ooA. 1 AoA. oNA. omo. omA. omo. Noo. moo. AoNo masco>om oooum
ooo. AoA. mNA.- NAN.1 ooA. 1 mNo. - oo9 AoN.- AmNo oaoamvoam Aouooom
moN. NNo. mmN.- Aoo.- oAm. NAo. oNo.- on.1 AoNo omonso Amuooom
ooo.- oom. oAm. omo. omm. Nmo. mmA. Aom. AmNo mAAo on moocuooz
mNA.- Aoo. AoN. oom. AAm. ooo. ooA. ooo. ANNV «mzm ou oomauomz
quEfiOHriGm
Aoo. ANV Aoo Amo Aoo Amo ANV AAV

 

 

AmMAozooo NoAv mAm2¢m1mom AnmNAAvommm
oAA<momA=oAmu<o Aomom mom mzoAm<Ammmoo mmnm01ommN

AomacAucoov mA mAm<a

HH mezmmm<

129

AooscAocooo

 

oNo. mmA. ooo. Aoo. moN.- mmo. moo. oAA. AANV Aoo AoAAoA UAAoso
moo. 1 oNN.1omN.-mmo.1 Nmo. ooo.- vo.1 on. AoNo ADV AcoamoAQEAAo
NNA. mmN. Nmo. oNo. mmo.- NNo. ooA. NNA. Avo ADV ooomm Amoco meoocA
ooo. mmo. AAA. Noo.1 Nmo. ooA. AoN. omN. Aon asoo AomonA
moo. ooN. AmA.ooo.1 Amo.1 moo. AoA. ooN. ANAV coAAAoAOAAAmA oAmsmm
ooo.A mmo. Noo. ooo.- mmo.- Noo.1 vo.1 mNo.1 AoAv moquo Acoamosa<
mmo. ooo.A mmA. mmo. ooo.- ooo. ooA. NNN. Aon oAonm
Noo. mmA. ooo.A NNA.- omA.- NNo.- ooA. omo. 1 Avo omAAomoumo sumo
ooo.- mmo. NNA.- ooo A Noo.1 AmN. omo.- Noo. Aon aoAAmAAoAaAao< OAAAsA
mmo.- ooo.- omA.- Noo.1 ooo.A ooo.- ooo.- ooo.- ANAV unmamvoam AcmauAm>oo AmooA
Noo.1 mNo.1 NNo.- AmN. vo.- ooo A mNo.1 oNN. AAAV mAmAAAA:
vo.1 ooA. oAA. omo.- ooo.- mNo.1 ooo.A ooo. Avo ooAAAmcoAABAAuoaA
moo. NNN.omo.1 Noo. ooo.- oNN. ooo. ooo. A Nov owmAAoo
mNo. oN9 moA.- Aoo.- NoN. oAA.- NoA.- Aoo. Amo Auo me83 Sumo Aoouo
omm. Nom. oNN. AmN.- Amo. moA. NoN. oNo. ANV ooAAommuoo chusuooAscmz
ooo. mNm omN. NoA.- Amo.1 ooo. oNA. Noo. Aoo oaAAA waAusoomAsaoz
oAN. NmA ooN. ooN.- NmA.- Nmo. ooA. vo.1 Amv Aaoamvoam mAAAauomoscoz
AoA. mmN. NoN. mmo.- ooo.- omo. ooN. oNo. on NAAoAm>Ao
aomumuHcmwll
Aoo. moN. moo.1 NAA. Nom.- NoA. mmA. NoA. Amo mmcooo coAAmAsooo sumo
moo. moA. mmA. Noo. moN.- moo. ooo. moN. ANV mmcmoo coAAmAsooo summ1aOz
Nmo. omA. oAN. oNo. oAm.- ooA. ooA. NoN. AAV coAAoAmAz1Aoz

 

SC SC SC 83 SC SC 83 A3

 

 

AmmAazooo NoAv AAmz<m1mom AomNAAAAumom
AAAomoAAouAmo<o A<mom moo ozvooAmmmoo Amamo1ommN

AomacAucoov «A 33¢.

HH Nanmmm¢

130

AooscAAcooo

 

 

 

 

omo. omo. mom. omm.- ooo.- AAo. moo. on. ANmo Aoo omwm1mmom AHA owaono
m CmSU HmquBfiOHHNrCm
mNo.1 omo.- NAo. mNo. mmA.- omo.- ooA.- mNo.1 Aomo ADV Acoemvosmao
ooo. Nmo.1 omo.- vo.1 oNA. Amo.1 NAo.- ooo.- Ammo moo ocoA
ooo.- ooN. ooo. AAo.1 omo. Nmm. Noo. oom. AAmv coAAmAqucAaoA 0AAoso
Nmo.1 NoA.- moA. ooA.- moo. ooA.- AmA.- oAN.- Ammo AaoaNvoam waAAsuomosaoz
ooo. on. ooA.- mmA. NAo. vo.1 NAA. NNo. ANmo aoAumoAOAAAoA voamm
NNo.- mmA.- mAA.- AAA. oNo. on.1 Noo.1 AAo.1 AAmo omAAAAAOAosAAAocA
vo. moo. oNA. omo. oNo. ANA.1 ANo. omo.- Aomo AAAAAAAz
vo.1 AAm. mmo.- Nom. oNo.- ooN. NAo. NmN. AoNo omoAAoo
NNo.- ooo. ANo.- omA. NoN.- mAA. mmo. Amm. AmNo mace Aomwqu
"CH owamcu
owcmnu Hm:OHumuHcmwmm
ANo.- ooA. ooo. mAA. ooA. AmA. oAN. oNA. ANNV mAonAmuaA
Aoo. ANA. mmN. NNA.- mmA. NAo. vo.1 moA.- AoNv Amoco>mm «Doom
oNo.- NAo.- on.1 NAN. NNN. omA. mNA. Nmo.1 AmNo Asmamvoam Aoumoom
Nmo.1 NAA.- moN.- mNo. ANA. Amo. NNA.- Nmo.1 AANo mNo.38 AmAmomA
AAA. NoA. NmN. AoA.- NmA.- Noo. oNA. NoA. AmNo NDAG ou oomaumoz
NAo. AAo. Aom. NmN.- omo.- omo. AoA. oNo. ANNV Amzm oA momenomz
uSmBGOHHNer
AoAv Aon AAAV Aon ANAV AAAV Avo Aoo

I
ill

AmmAAzooo NoAv mAAz<m1mom AQANAAAAomAm
AAAAmovooAmuAV AAmom mom mzoAAAAmmmoo Amam01ommN

AmmaaHucoov wH MHM<H

HH xHszmm<

131

AooAAAvooo

 

 

omA. Nmo. ooo. ooo.A mNA. Nom. NNo. Nmm. AANV Auo AmAAom oAAooA
AoN. oAA.- omA.- oNA. ooo.A Amo. oNN.- Noo.1 AoNo AAV AaoaNvoamco
ooA. moo.1 moA.- Nom. Amo. ooo A NAm. AoN. Avo Ago ooomm Aooco oaoocA
NoN.- mAN. ooN. NNo. oNN.- NAm. ooo A omm. Aon c309 oomonA
mmm.1 ooA. NoA. Nmm. Noo.1 AoN. omm. ooo.A ANAV coAumoAUAAAoA onamA
Nmo.1 AAA. NAo. oNo. moo.1 NNA. ooA. moo. Avo omovo oaoamo9a<
NAA.- NoA. AAo. mmA. oNN.- mmN. mmo. oAN. Aon oAonm
moN.- NmN. Aom. ooo. omN.- Nmo. AAA. AmA. AAAV omAAomouoo spam
mNo. AoA.- NmN.- Aoo. mmo.- oNo. Noo.1 ooo.- Aon aOAAonoAcABAA oAAoso
ANA. NmA.- omo.- mAN.- Nmo. omo.- Nmo. Amo.1 ANAo AcosNvoam Acoacuo>oo Aoqu
Amo. Noo. NoA. mmo. Aoo.- NNo. AoA. moo. AAAV AAAAAAAz
NNA.- oNA. AoA. on. vo.1 AoA. AmN. AAA. AoAv ooNAAmcoAAsAAAocA
Nmo.1 omo. oNo. oAA. on. NNA. omN. ooN. Aoo mmoAAoo
mAN. vo.1 mNA.- mNA. oAm. AoN. Amo. 39 A8 A“: 988A 531A Aofio
NNA.- oom. AmA. moo. omN.- omN. Nom. Aom. ANV omAAomono maAusuooAaaoz
mmN.- mAm. AoN. omo. moA.- mAN. NmN. NmN. Aoo oaAAA maAAAAomAasoz
AAA.1 omA. oAm. ANA.1 Aom.- NoA. Nom. ANA. Amo Acmemvoam wcAAsuomAscmz
oAm.- omo. AAm. NNo. mmm.1 oNN. ooA. NoA. AAo NAAvo>mo
coHumuchwmm
NAo. Nmo. vo. moN. NNA.- oom. oNA. mNo. Amo omcooo coAquaooA sumo
oNA.- omA. ooA. on.1 Noo.1 ooo.- Aoo. mNo. ANV mocmoo coAAoAsooo apom1aoz
on.1 Aom. ooA. mNA.- Nmo.1 omo. mNA. moA. AAV cvoonAz1uoz
AANV AmNo ANNo AANV AoNo Avo Aon ANAV

 

 

 

 

 

1"
I

AmmAazooo Nvo mAAz<m1mom AomNAAAAomAm

ngPHHDUHMws Hébm mom monHSMMMOU mmGMOIOMMN

GoficHuaoov mH 333.

HH Nanmmg

132

Aooovocooo

 

mmm.1 omm. Hom. moo.1 NHo. moo.1 AHN. omm. Ammo Auv mNHm Bumm CH owcmnu
wwdmno HousmaaoAH>cm
omo.1 oHo. ooo. Hmm. HN¢.1 Noo. 0Ho.1 «NH. Aomv Auv ucoehoHQEoco
moo.1 omo. ooo. HNN.1 omo. omH.1 Noo.1 Hmo.1 Anmv om: vcmH
NHo. moo.1 moo.1 NHo. ~0N.1 moo. NHN. oMH. Aqmv coHumuuwHaHafid omHnam
mHN.1 mmH. Now. me.1 mHo. NoH.1 oAH.1 HNo. AmmV ucoakonfim wcHAnuommnamz
mNN. N0H.1 mHo.1 mNo.1 omo. NmH. oNo. mom.1 Ammv aOHummHoHunm onBom
mmH. omH.1 omH.1 ooo. HAH. AOH.1 mmm.1 omo. AHmV vONHHmGOHunuHumGH
moo.1 Amo.1 on.1 «Ho.1 wMH.1 Hwo. ooo. NHH. Aomv humuHHHz
«MH.1 omo. omo. NNH. moo.1 owH. mom. me. AoNV oonHoo
omH.1 moo. oHH. mm~.1 moo.1 oNH.1 oNo. Noo. Aowv mace umownoq

 

AaH owcosm
owcmno HmcomuoNHcm no

 

moA.- mmA. mmA. mmo. NAo.- moA. ooA. AmN. ANNA momquoABA
AAo.1 oNo.- Nmo. mmo. ooA.- AAA. on.1 omA. AoNV monum>mm muoom
NmA. NmA.- moN.- ooA.- mmo.- AAo 1 NAo.- Nmo.1 AmNo uaoamvoam Amumomo
ooo.A NNm.1 NmA.- omA. AoN. ooA. NoN.- mmm.1 AANV omvoso Amuooom
NNm.1 ooo.A AAN. Nmo. oAA.- moo.1 mAN. ooA. AmNo NAAo oA ooocuooz
NmA.- AAN. ooo A moo. omA.- moA.- ooN. NoA. ANNV Amzm oo ooocuooz
uaoacoum>am

 

33 33 Ammo 33 8.3 83 33 CC

 

 

AmmAazooo Nvo AAAz<m1mom AomNAAAAomAm
AAAAmoAAooAAAAV AAmom moo ozoAAAAmmmoo mmnm01ommN

AvoacHucoov mH mHm<H

HH Nanmmm<

x.~flfi’4~!.~b§!~<

133

AvoscHucouv

 

 

oNo.- ooo. AAo.1 NNA. mmN.- mmo. mmo. ooA.- AANo AAV AoAAom OAAAAA
omo. AAA. mmA.- ooo.- ooo.- NAo.- ooA.- mmo.- AoNo ADV Acmamvoamco
NNA. AoA.- Aoo. omA. oNA.- moA. AAA. AAo.1 Avo ADV ooomm Amoco mamucA
oNo. NmN.- vo. mom. oNo. ooA. on.1 NAo.- Aon :soa AomvoA
Nom.1 omo. NAA. mmA. Noo. AmN. omA. Nmo.1 ANAV coAAmoAOAAAoA onamm
ooo. NNo.- mmA.- vo.1 NNo.- ANo.- Aoo. oNo.- Avo omomAA AsmaBA=a<
on. vo. moo. AAm. ooo. ooA. ANA. NAo.- Aon oAonm
ooA.- mAA.- oNA. mmo.- ANo.- ooo. mmN. on.1 AAAV ooAAowmAAo sumo
mmA. AAA. omo. Nom. omA. mAA. NNA.- NAN. Aon coADmAAoAaAaoA oAAAAA
NAo. oNo. oNo. oNo.- NoN.- ooA. mmA. NNN. ANAV Acoamvoam Acmacuo>oo AoooA
vo.1 on.1 ANA.1 ooN. mAA. AmA. NAo. omA. AAAV mAAAAAAz
NAA. Noo.1 ANo. NAo. mmo. oAN. vo.1 mNA. Avo omAAAmcoAusuAuch
NNo. AAo.1 omo.- NmN. Amm. oNA. moA.- Nmo.1 Aoo omoAAoo
oNo. moN. oAA.- omo. mmN.- Nmo. ooA.- ooo. Amy AAV maooaA sumo Aonuo
mNo. Nmm.- NmA. AmA. oAA. mNN. AmA. AoA.- ANo ooAAommuoo chAAAooAscoz
mmo. NoN.- oNo. AAA. mmo. mAA. oNA. mNA.- Aoo oaAAA chAaAooAaamz
AoA. oAm.- mAA. vo. mAN. Amo. mmo. NAN.1 Amo AAAENOAoam maAAsAomAAcoz
ooo. mmN.- moo. NmA. mmA. moA. omA. ooA.- AAV NAAoAm>Ao
coHumNHcmmmm
mAN. ooA.- mAA. oAA. NNo. NNo. omo.- mNo.1 Amo omcmoo aoADoAaooo sumo
mmo.- omA.- AAo. omN. omm. oNo. Noo.1 ooo. ANV omcooo coAAmAnooo aAmA1coz
ooA. omN.- oNo. Aom. Aom. mNA. moo. AoN.- AAA soAAmAmAz1Amz
88 38 Aomo 89 $8 ANNA $9 89

 

 

AmmAAzooo NoAV AAAz<m1mom AamNAAAAomAm
AAA<ADAAooAmu<V AAmom mom mzoAAAAmmmoo mmom01ommm

AmmaaHucoova MHm<H

HH Nanmmm<

134

AomscAucooo

 

 

 

 

on.1 NOH.1 moH.1 omo. ooH. Aoo. omH. omm.1 Ammo Auv oNHm anon GH owamsm
owcmno HmucoacouH>am
NNN.1 moN. moo.1 ooo. mNo.1mmo.1 moo.1ooo.1 Ammo Huv uaoahonaocD
Noo. NNo. moo.1 Hmo. mNH. AAo. oHo. moo.1 Ammo om: pcmH
mqo. AmH.1 omo. qu. me. mmo. NNH. oNN. Aqmv noHumuumHaHafl< oHHnam
omo. mNo.1 HHo. NNN.1 mmH. moo. omH. moH.1 AmmV ucoE%OHQEm waHAnuoomacoz
ooo.H AoH.1 Nmo.1 NHH. omo. omo. mHo. mHo.1 ava aoHuoaHoHuHom OHmaom
AoH.1 ooo.H oHH.1 ooo. 1 moH.1 NHo.1 mHo. moo.1 AHmV UONHHwaomuauHumoH
Nmo.1 oHH.1 ooo.H ooo. 1 mNo.1 Nmo. mmo. moo. Homo meuHHHZ
NHH. moo.1 moo.1 ooo.H ooo. HHH. moH.1 moo. AoNv oonHoU
omo. moH.1 mNo.1 ooA. ooo.H vo. Amo.1 moH. 1 AmNV asos umowuoH
.GH omaono
wwaono Ho50AumuHcowII
omo. NHo.1 Nmo. HHH. vo. ooo.H HNo. NNo. ANNV oumumumucH
mHo. NHo. mmo. omH. 1 Amo.1 Hmo. ooo.H mmo.1 AmNV monso>om oumum
mHo. moo.1 moo.moo.1moH.1 NNo. mmo.1 ooo.H Ava ucoamoHoam Honooom
oNN. an. ooo. 1 AMH. moH.1 moH.1 mHo.1 NmH. AANV m.mmHAAAAo Hmuovom
NoH.1 omH.1 Aoo. 1 omo. moo. me. moo.1 NmH.1 AmNV quo ou mmoauooz
mHo.1 moH.1 moo.1 oNo. oHH. me. Nmo. mON.1 ANNV <mzm ou mmoaummz
uaoacouH>cm
Ava AAmv Homo AoNv Ava ANNV AoNo Ava

I‘ I] 9
I I I

AmmAAzooo Nvo AAAz<m1mom AAANAAAAomAm

 

NHH<MDHHDOHmw<V H<MDM mom monH<Hmmmoo MMQMO1OMMN

HomscAuaooo mA mAA<A

HH Nanmmm<

135

AomscAocooo

 

 

Noo.1AmN. ANN.1 NAo. mmA.1AANo ADV AmAAom 0AAAAA
NA9 ANA. oNo. NoN.1on. AoNo Auo oaoaNvoamco
oNo. 1 NAo.omA.1 No9 NoA. 1 Avo ADV ooomm Amoco oaooaA
AAN. vo.1 Noo.1 NAN. oAA. 1 Aon ozom ooomAAA
AmN. ANA. Amo.1 omA. ANo. ANAV aoAuooAOAuAmo «AoAAA
omo. mNo.1ooo. ooo.- Nmo.1 Avo ooooAA Acmaoona<
omo. mmo. Nmo.1 ooN. NoA.- Aon oAonm
mom. NAo.omo.1 ooo.1moA. AAAV ooAAomoAmo sumo
omm. 1 mNo. on.1 AA9 ooA.- AmAv coAAmAAoAaAaoA OAAAAA
ooo.- mmA. 1 oNA. om9 mmo. ANAo Acmsmvoam Acoacum>oo AmooA
AAo. mmo.- Amo.1 Nmm.ooA.1 AAAV NAAAAAAz
moo. moA.- NAo.- Nmo. AmA.- Avo omAAAmcoAusoAuocA
on.1mNo.1 moo.1 oom. oAN.- Aoo omoAAoo
mNA.mmo.1 Nmo.1 mmo. AoA.- Amy AAA oaooaA sumo Aoooo
oo9 ANo.- ooo. AoA. oNA. ANV ooAAommAmo maAAaAoomscmz
AAN. on.1 ooo.- mNo. mNo. Aoo AEAAA chusuomAscAz
omm. on.1 NAo. on.1 oAN. Amo Acmamvoam chAAAquscm:
omA. Amo. on. ooo.- moA. AAV mAAoAm>Ao
coHumuHGMMII
mAN.- moo. moA.- mNo. mmA.- Amo mmcooo BOAAvoooo sumo
oNN. mNo.1 Nmo. oNA. mNo. ANV omcooo aoAAmAsooo aAAA1coz
omm. Noo. omo.- vo. NAA. AAV avomAmAz1uoz
ANmo Aomo Ammo AAmv Ammo

 

 

AmmHHZDOU noHv mHmz<m1mDm AQMNHH<Hommm
NHH<MDHHDUHMU¢V H<MDM Mom monH<Hmmmoo mmnm01OMMN

AvoncHuaoov mH mHm<H

HH NHszmm<

1!

V8 InI~Zw.-n.-n.<

136

 

 

 

 

 

ooo.A mmA.- mNo. mmo.- oAA. ANmo AAA ouwm aumm aA omcmoo
m SwSU HMUGGECOHHN/fim
mmA.- ooo A ANo.- Noo.1 AmN.- Aomo Nov AcosNvoaoco
mNo. ANo.- ooo.A Noo.1 mmo. Ammo moo oaoA
mmo.- Noo.1 Noo.1 ooo.A oNN.- AAmo :oAAAAAoAaABAA oAAosA
oAA. AmN.- mmo. oNN.- ooo.A Ammo Anaemvoam maAusuumAscoz
oAm.- NNN.- Noo. on. Amo. ANmo aoAAooAOAAAmA oAmsmA
moA.- moN. NNo. AmA.- mNo.1 AAmo omuAAocvosoAAoaA
moA.- mNo.1 ooo.- omo. AAo. Aomo NAAAAAAz
moo. ooo. Aoo. AoA. NNN.- AoNo ommAAoo
ooA. NNo.- mNA. mmA. mmA. AmNo cane oooonA
A GLAIowAAm
uwdmno HmGOHumuHGmeo
ANo. mNo.1 AAo.1 mmo. moo. ANNV onAoAmAaA
omA. vo.1 vo. NNA. mmA. AoNo monum>om «Doom
on.1 ooo.- mmo.- oNN. ooA.- AmNo Acmamvoam AAAmomm
Nmm.- omo.- on.1 NAo. mAN.- AANV omonno Amuooom
on. vo. omo. ooo.- NmA. AmNo mAAo ou oooaumoz
Aom. vo. ooo. Noo.1 NoN. ANNV «mzm oA ooocumoz
ufiwEGOHHNrCm
Ava Aomo Ammo AAmo Ammo

 

 

 

 

AmmAAzooo NoAv AAAz<m1mom AnmmAA<AomAm

AmoacHHcoov mH MHde .

HH anzmmm<

mAAAAoAAooAAo<V AAADA mom mzoAmAAmmmoo mmnmo1ommm

H

x “WNZn-hO-a.i<

137

AooscAoaooo

 

 

Amm. mNm. oAm. AAN. Aoo. 1 moo. mAN. omA. AANV AAV AoAAmA 0AAono
mmA. omA. oAm. oNN. NAo. 1 mNo. oNA. mmo. AoNo ADV Aaoamvoaoao
mom. ooN. moA. AmA. ANN. 1 omo. mNo.1 AmA.- Avo AAA ooomm Amoco oaooaA
NoN. NNA. mmm. on. ooA.Aoo.1 AoA.- NNA.- Aon pace oooonA
mmm. Nm9 NAA. on.1 NNo. AAo.1 omA. AmA. ANAV coAumoAUAAAmA oAmamA
oom. AoA. moo. on.1 NmN. Amo.1 mmA.- NAA Avo ooovo Acoaoosao
oNA. mmA. oom. oNN.- mNA. ooN. moA. oAA. Aon vouom
moo. Nmm. mmA. AoA. oNN. mNo.1 oAA. mNA. AAAV ooAAomoAAo sumo
omN.- omN. 1 NAA.- mAm. 1 vo.1 NAo. NmN. oom. Aon coAAoAquAABAA oAAoso
moo. NoN.- mmN.1AmA.ANo.1 on.1 ooN 1 ooN.- ANAo oaoamvosm Acoacuo>oo AoooA
NAo. mNo. NAo.omN.1 Nmo. Aoo. mNo. moo.1 AAAV NAAAAAA:
oNo. moN. Amm. omo. oAm. ANo. moo. on. Avo ooAAAAAOAAsAAuocA
NAo. Noo.1 No9 mmN.- ANo.- on. 1 Nom. oom. . on omoAAou
ooo.A mmA. ooN.omN.1 AoA.- Nmo.ooA.1 oNN.- Amo AAA oaooaA aumm Amnuo
moA. ooo. A mmN. oNA. oNA. mmo. 1 moA. ANN. ANV moAAomvoo chABAoAAsamz
ooN. mmN. ooo.A oom. on. mNo. NNo. NmA. Aoo oaAAA chAAAooAacmz
ooN 1 oNA. oom. ooo.A ooo. Aoo. mAm. on. Amo AcmaNvoam maAAsuomAacmz
AoA.- oNA. on. ooo. ooo.A Aoo. moA. mNm AAV NAAoAm>Ao

GOAumNHcmwuo
Nmo. mmo. mNo. Aoo. Aoo. ooo.A mNA.- omo Amo mmcmoo EOAAoAsooo Sumo
ooA 1 moA. NNo. mAm. moA. mNA.- ooo.A oNo ANo omcooo avooAsooo sAoA1aoz
oNN.- ANN. NmA. on. mNm. omo. oNo. ooo A AAV BOAAmAmA21omz

Amo ANA Aoo Amo AAA Amo ANA AAo

 

 

AmmAAzooo ooo mAmz<m1mom AomAAAmmm>on z<mmo mom mzoAAAAmmmoo mmnmo1ommm

oH MHm<H

HH Nanmmm<

138

AooscAocooo

 

 

NAN.1 NmA. ANN. oAA. mmm. Noo.- mNA. Nmm. ANmo AAA onwm aAoA AA mmcooo

m GNSO H qumEGOHHN/Gm
ANA.1 mNN.1 mmo.1NmA.1 Amo.1 Nmo. mmN.- oNN. Aomo Auo Acoamvoaoao
omm.- mmo. mmo. AAN. Aoo.1oN9 1 oNA. moA. Ammo moo ocoA
NmN. mmA.- Nm9 1 mmm.1 ANN. ooo. 1 NNA.- mNA.- AAmo aoAAAAAoAcAEAA UAAABA
mmo. moo. ooo.- Aoo. AmA. oNo. AAA.1 Noo.1 Ammo ocoaNvoam chAsoooAsaoz
oNA. moo. NAA. moA. Nom. 1 ooo.- NAo.- oNA.- ANmo coAAooAOAAAoA onamA
Noo.1 omo.- omo.- Aoo. 1 mNo.1 mNo.1 vo.1 AAo.1 AAmo omNAAocoAAoAAAAAA
mmN. 1 moo.1 moo. mmo. AAo.1 NNo. ooA. omA. Aomv NAAAAAAz
ANN. mNo. Aoo. omA.- ooo.- omo.- omN. omN. AoNo 0mmAAoo
NNA. 1 moN. omA. oom. AmA. omo.- AAN. oNo. AmNo c309 AooonA
.GH owcmao

owamnu HmcoHuoNHcmwll
AoN. No9 NoN. omA. AAo. oAA. 1 NAA. ANo. ANNA onAvoAcA
oAA.- Nom. 1 ooN.- NAo.- NAA. ANo. omo. ooo. AoNo Amocopmm «Baum
mo9 1 mmo.1 AmA.- mmN.- NmA. Aoo. omo. ooo. AmNo AcmaNoAAam Amuooom
Amo.omN.1 NNA.- NAm.- NAo. moo.1 omA.- AoA.- AANV ommAuso Amuooom
mmA. AoA. AoA. oAA. omN. mNo. Noo. oAA. AmNo NAAo oA oochmmz
ooo.- Nom. Amm. mmm. NoA. Noo.1 moN. mom. ANNV «mzm oo oooaumoz

 

ucoacoHH>cm

 

Amy ANV on Amy AAV Any fiNv aHv

 

 

AmMHHZDoo oev MHm24m1mDm AanhHmmm>HQv z<mMD mom monH<Hmmmoo mmnM01omMN

AooncAoaooo oA AAAAA

HH anzmmm<

139

AomAAAoaooo

 

 

mmA. oNA. Noo. ooA.- oAA.- NAo. mAA. AAA. AANV Auo AoAAmm 0AAoso
mNo. mmo. mmA. NoA.- mom.1 ooo.- AmA. AmA. AoNo Aoo ocmsNvoemao
mAm. oAN. NAo.- NNA.- Amo. Noo. oAA. moA. Avo AAV ooomm Amoco oaoocA
mom. oom. mAA. ooA. Noo.1 NmN. NNN. mom. Aon mace AooonA
NmN. omA. mNo. NNA. NoA.- moo. Nmo. AAA. ANAV aoAAovoAAAoo voamm
ooo.A AmA. mmN. omA.- NAo. ooA. mmm. mmo.1 Avo omovo ocoamosa<
AmA. ooo A ooA.- on. mAA.- ooA. omA. on. Aon voAom
mmN. ooA.- ooo.A oAN.- ooo.- moN.- Aoo. NNA.- AAAV omAAommuoo anom
omA.- on. oAN.- ooo.A vo.1 AAA. Noo.1 ooN. Aon aoAuwuquaAaoA 0AAAAA
NAo. mAA.- ooo.- vo.1 ooo.A Amo. NmN.- ooN.1 ANAV Acmamvoam Acoaauo>oo AoooA
ooA. ooA. moN.- AAA. Amo. ooo.A NNo.- NAo.- AAAV NAAAAAAz
mmm. omA. Amo. Noo.1 NmN.- NNo.- ooo.A ooo. Avo ooNAAocoAAAAAAoaA
mmo.1 on. NNA.- ooN. ooN.1 NAo.- ooo. ooo.A Aoo omoAAoo
oom. oNA. moo. omN.- moo. NAo. oNo. NAo. Amo Aoo oaooaA sumo Aonuo
AoA. mmA. Nmm. omN.- NoN.- mNo. moN. Noo.1 ANV mmAAomono maAABAooAsamz
moo. oom. mmA. NAA.- mmN.- NAo.- Amm. Noo. on ABAAA maAAsAoAAsawz
on.1 oNN.- AoA. mAm.- AmA.- omN.- omo. mmN.- Amo Aaoamvosm maAAsuooAAaoz
NmN. mNA. oNN. vo.1 ANo.- Nmo. mAm. ANo.- AAV NAAvo>A
GOHumNHAh—m 1H0

Amo.1 ooN. mNo.1 NAo. on.1 Aoo. ANo. on.1 Amo omaooo cvomAsooA sumo
mmA.- moA. oAA. NmN. ooN.1 mNo. moo. Nom. ANV «mango aoAAmAsooA BAAA1coz
NAA.- oAA. mNA. oom. ooN.1 moo.1 on. oom. AAo :OAAoAmAz1omz
Avo Aon AAAV Aon ANAV AAAV Avo Aoo

I
il’

 

 

AmmHHZDoo ocv MHmz<m1mDm AanmHmmm>an z<mmD mom monH<Hmmmoo MMQMOIOmmN

AmmacHuaoov oH MHm<H

HH XHszmm<

140

AomsaAucooo

 

 

 

 

vo. AAo. mom. NmA. mmo.1 moo. mmo. Amo. ANmo ADV oNAm sumo AA mmcooo
m amno HmucmacoHH>am
Noo. mmo.1 mHH.1 NHo.1 HAo. AAo. oNo.1 ooN.1 Ammo Auv unmamoHoamao
moH.1 Amo.1 omo. moH.1 NNo. oAH. NAN.1 NNH.1 Ammo om: oomH
oNN. AoN. moo.1 omH. mmo.1 oNo. ooN. mHN. AAmV GOAumAuonAao¢ OAHnom
ooN.1 oHo.1 mNo.1 AAo.1 ooo. NNH.1 Nmo.1 ooN.1 Ammo ucoahoHoam wcAnnuommocmz
mMH.1 mmo.1 oNo.1 oAN.1 NAo. HoH.1 vo. oHo.1 Ava aOAumvoAuumm mHmamm
Noo.1 moo.1 vo.1 Noo. HHA. oHo.1 mmo.1 omo.1 AHmv ooNAHmCOAuouAumaH
NoH.1 NNN. moo.1 AmA. oNN.1 on.1 moo. NHA. Aomv mumuAHHz
moo. HoA. Aoo. mNm. moH.1 Nmo.1 on. mNN. AoNV oonHou
NHo. mmm. oNN. on. NAH.1 ooH.1 NmH. NmA. Ava osoe ummwan
AnH owcmm
owamno HmcoHumuHam no
mAN. ooH. omH. Amo.1 vo.1 NNo. Aoo. AAo. ANNV oumumuouoH
moH.1 mHH.1 on. omH. oAm. omo. ooN.1 moH.1 AoNV mosoo>mm oumum
omo.- omA. Amo.1 oAA. NNo. NoA. NNN. on. AmNo AcoaNvoam Amuooom
moH.1 Noo. oHo.1 HoH. ooH. oNH. moo.1 mHo.1 AANV mmmHuso Hmuooom
mmN. Hoo. Hmm. moo.1 moo.1 ooo. moH. ooo. Ava AAuAu ou mmooumoz
omo. moH.1 mAm. NmH.1 NNN.1 oNo.1 moH. mHo. aNNV «mzm ou mmocumoz
UGQBCOHHNAEM
AoAv AmAV AAAV AmAv ANAV AAAV HoAv on

 

 

AmmHHZDOU oov MHmz<m1mDm AanmHmmm>Hav z<mms Mom

AnosaHuaoov oH MHm<H

HH xHszmm<

monH<Hmmmoo MMQMOIQMMN

141

AomscAAcooo

 

 

Hom.1 omN. oom. ooo.H omA. NAN. mmN. Hmm. AHNV Auv moHHmm omHan
omH.1 mom. NmH. omA. ooo.H Nmm. Hmm. HAm. AONV Auv ucm8%0HQEmGD
mmo.1 mNH. 5N0. NAN. Nmm. ooo.H NBA. 0mm. AoHV Auv 000mm Hows: maoofiH
oHo.1 mAA. mNH. mmN. Hmm. NmA. ooo.H HwA. AoHv QBOH umomeH
m0N.1 mAN. mHH. Hmm. HAm. omm. HwA. ooo.H AnHv COHumaHoHuHmm mHmBmh
moH.1 moN. omo. moH. oNo. oAm. mom. NmN. AoHv moomHm ucoaomna<
n00. H00. mmH.1 0NH. mmo. oAN. 00m. omA. AmHv mHouom
oHo.1 Hmm. mAm. Noo. me. NHo.1 mHH. who. AAHV moHuowoumo sham
HoH. moo.1 NmH.1 moH.1 NoH.1 NNH.1 ooH. NNH. AmHV aoHumuumHnHamA OHHnom
ooH. moo.1 NAN.1 oAH.1 moo.1 Amo. Noo.1 NoH.1 ANHV ucoa%OHaam ucoecuo>ou HmooH
oNH. mmo. oNo.1 NAo. moo.1 mmo. NmN. moo. aHHv mumuAHHz
moo.1 me. 00H. mHH. HmH. oHH. NNN. mmo. AOHV mmNHHmcowuduHuwfiH
mHo.1 ooo. mHo. AAH. AoH. moH. oom. AHA. on omoHHoo
Amo. wMH. 000.1 Hmm. mmA. mom. moN. mmm. va Auv oaoocH Show Hmnuo
00N.1 HoA. Nmm. mNm. 0mA. 00m. NBA. NmA. Amv moHHowmumU mcHuauommaamz
NNH.1 AoA. Amm. oAm. 0Am. m0A. mmm. NHA. A00 mahHm waHufluommficmz
mHm.1 0HA. mmm. HAN. CNN. HwH. mHo. MA0.1 Amv ucmE%OHmBm wcwunuommscmz
mHo. 0mm. moH. Amo.1 NHo.1 HNN.1 00H. Noo. AAV hummum>wn
coHumNHnmwmm
moo.1 mNo. Noo.1 woo. mNo. omo. Aoo.1 HAo.1 AmV owfimnu coHuoHsoom Show
0mH.1 N00. moN. mAN. mNH. mNo.1 A0H.1 omH. ANV mmcmao COHumHamom Ehmmunoz
AoH.1 0HH. mom. 0mH. mmo. AmH.1 NNH.1 HMH. AHV nomumeHSIuoz
AANV Ava ANNV AHNV AoNv HoHv Ava ANHV

 

 

 

AmmHHZDOU 00v MHmZ<m1mDm AQmHmHmmm>HQV z<mmm mom monH<Hmmmoo MMQMOIOMNN

AvoncHucoov oH 53E.

HH anzmmm<

 

 

142

AomssAAaooo

 

 

 

 

AmN.- oAA. moA. Nmo. moo. NoN.- NNo. mmA. ANmo ADV ouwm aAmA1cA omaon
m fimfio HmquEGOHHNer
ooo. ooo.- mmo.1 moA.- omm.- ANA.1 oAA.- oom.- Aomo AAA AcmaNvoaAao
Aoo.- Nmo.1 NNo. ooo. ooN.1 oNo.1 omA.1 AAA.1 Ammo omo oaoA
NAo.- Noo.1 ANA.1 ooo. NmA.- oNA. vo. ooA. AAmo aoAAAAAoAnAEAA oAAAso
NoA.- omo. mNA. mmo. moo. AmA.- oAm.- moo. Ammo AcmaNvoam maAAsAooosamz
Nmo.1 Noo.1 moo.1 NAm. oAN. NmA. mmo.1 mmo. ANmo aoAAooAUAAAmo mAmamm
omA. NoA.- ANo.- AmA.- mAA.- moA.- mNo.1 moA.- AAmo ooAAAmaoAusuAuoaA
ooo. on.1 Noo. NoA. moo.1 oNo.1 omA.1 vo. Aomo NAAAAAAz
mNA.- omo. vo. moA. ANN. AmA. omN. AoA. AoNV omoAAoo
oAA.- omA. AmN. omA. AmN. Noo. mNo. NoN. AmNo csom AoomAmA
"cm owcono
owcmno HmGOHumNHamwmm
moo. AAA. ANm. Aoo.- mmo. Aoo. mAN. AAA. ANNA oAonAoAAA
NAA. mmm.1 AoA.- mAm.- ooo.- Nmm.- NmA.- oom.- AoNo Amoco>om mumum
moN. ooo.- NmA.- Noo. Amo.1 AAo.- mmo. vo.1 AmNo Acoamvoam Amumomm
ooo.A ooo.- mNA.- Aom.- omA.1 omo.- on.- moN.- AANV omonno AvoomA
ooo.- ooo A omN. omN. Nom. mNA. mAA. mAN. AmNo mqu 0A oomcumoz
mNA.- omN. ooo.A oom. NmA. NNo. mNA. mAA. ANNV Amzm 0» mmocummz
ufimEfiOHfiNfium
AANV AmNo ANNV AANV AoNV Avo Aon ANAV

 

 

AmmHHZDOU 000 MHmz<m1mDm AQmHmHmMm>HQV z<mms mom

Ammaawucoov oH mHm<H

HH XHszmm<

monH<Hmmmoo MMQMO1OMMN

143

AomsaAucooo

 

 

NAm. AmA.- NoA. moA. omA. Aoo.- mAm.- Noo. AANo ADV AoAAom 0AAo=A
oAN. mAA.- moo.1 ANN. AmN. mmo. ooo.- Amo.1 AoNo AAA Acoamvoamco
NmA. moA.- oNo.1 AmA. Noo. Aoo. Nmm.- AAo.1 Avo Ago ooomm Amoco m383
mmo.1 mNo.1 omA.1 omN. mNo. mAN. NmA.- mmo. Aon csom AommAmA
mmo. moA.- vo. AoA. NoN. AAA. oom.- vo.1 ANAV aoAAovoAAAoo «Aoamm
mmA.- Noo.1 NoA.- moo. NAo. omm. mmA.- omo.- Avo omoon Acoaoosa<
ooo.- moo.1 NNN. AoA. mAN. ooA. mAA.- omA. Aon oAmuom
oNo.1 on.1 moo.1 Aoo. mNN. omA. vo. Amo.1 AAAV moAAoumAmo aAoA
mAN.- Noo. AmA. mNm. on. ANo.- omA. oAA. Aon avooAAoAaAaoo 0AAAAA
NAo. AAA. mNN.- omA.- NAA.- vo.1 oAm. NNo. ANAV AcoaNvoam Acoaaum>oo Amqu
AmA.- vo.1 oAm.- Nmo.1. ooA.- NNo. omo. NoA. AAAV AAAAAAAz
vo. mmN.- moo. on. NoA. Aoo. NoN.- NNN. Avo omuAAmaoAAoAAuocA
vo.1 omo.- NAA. mNN. NmA. AAo. moA.- on. Aoo ommAAoo
oNA. Noo.1 mmN.- ANN. NNA.- AoN. oAA.- moo.1 Amo ADV oaooaA sumo Amouo
moo. omo.- moo.1 mNo. moN. NoN. Nom.1 mmo.- ANV ooAAomvoo chAAAooAscoz
NAA. omo.- moo. Aoo. omA. NmN. ooN.1 AmA.- Aoo mauAm chAsAooAaaoz
moA. Aoo.- mmo. omA.1 oom. omA. NAo.- mmN.- Amo osmamvoam chAaAqusam:
Nom.1 mNo.1 AAo.1 ooo.- AmA. AAo. NAA. NmA. AAV mvoAo>Ao
:oHumNHamwuo
ooo.- mNo.1 NNo. omo.- omo.- oAA.- ANo. Aoo. Amo omamoo aoAAmAsooo sumo
NAo.- vo.1 ooA. omN. AAN. NAA. omo. omo. ANo omaooo aoAqusooo 5Aom1aoz
oNA.- AAo.1 omA. omN. oNo. ANo. ooo. ooo. AAV cvooAmAz1Aoz
ANmo AAmo Aomo AoNo AmNo ANNV AoNo Ammo

 

 

 

 

 

 

AmMHHZDOU 00v MHmz<m1mDmAEMHmHmmm>Hav z<mm= mom mZOHH<Hmmmoo meMOIOMMN

AooaaAAaoov oA AAAAA

HH XHazmmm<

144

AooAAAAAooo

 

 

 

 

mmA.- Aoo.- on. mmo. NmA. omA. NNA. Nmo. ANmo AAV ouwm auom aA omamoo
o mono kucoacouH>¢m
vo.1 ooo. omo. oAN.- NNm.1 mmo.1 AoA. Amo. Aomo ADV Acoamvoaoao
Aoo. Amo.1 NoA. NAN.1 vo. omo.- moA. ooA. Ammo ooo oaoA
ooA. ooN.1 Nmo. oom. mNo.1 mAA. NAA.- AoA. AAmv aOAAoAquaAaAA oAAAsA
NmA. oAA.- mAN.- NoN.- NmA.- NAo.- vo.1 mNo.1 Ammo Acmamvoam maAAsAooosamz
ooo A moo.1 mmo.1 Aoo. Noo. mmA. NAA.- NmA.- ANmo noAAooAUAAAAA onamm
moo.1 ooo.A NAo.- moo.1 on.1 ooo. ooA. mmN.- AAmo ooAAAoaoAusoAuoaA
mmo.1 NAo.- ooo.A NAN. omN. mmN.- mmo. Noo. Aomo NAAAAAAz
Aoo. moo.1 NAN. ooo A oom. mmo. ooN.1 mNo.1 AoNo mmoAAoo
Noo. on.1 omN. oom. ooo A mmN. Amo.1 on.1 AmNo azom AooonA
"cH owams
omnono HmcoHuoNHam Ho
mmA. ooo. mmN.- mmo. mmN. ooo.A NAA.- oNA.- ANNA ouoovoAaA
NAA.1 ooA. mmo. ooN.1 Amo.1 NAA.- ooo.A moA. AoNo monso>mm oAAAm
NmA.- mmN.- Noo. mNo.1 on.1 oNAe- moA. ooo.A AmNo Acmamvosm Amuoomm
Nmo.1 omA. ooo. mNA.- oAA.- moo. NAA. moN. AANV oNvoso AvoooA
Noo.1 NoA.- on.1 omo. omA. AAA. mmm.1 ooo.- AmNo NAAo ou oooauooz
moo.1 ANo.- Noo. vo. AmN. ANm. AoA.- NmA.- ANNo Amzm ou oochomz
HGGBAHOHHNer
ANmo AAmo Aomo AoNo AmNo ANNA AoNo AmNo

 

 

 

 

 

AmmAmzooo ooo mAAz<m1mom AamAmAmmm>Ano z<mmo mom mzoAAAAmmmoo mmnm01ommm

AomsaAuaooo oA mAAAA

HH NHQ2mmm<

145

AAAAAAAAooo

 

 

Nmo. moA.- ooo. ooo. mmo. AANV AAA AAAAAA 0AAAAA
moo. omm.- ooN.1 NmA.- moo. AoNo ADV Asoamvoaoco
NoN.- ANA.1 oNo.- oNA. AmA.- Avo AAV ooomm Amoco AscocA
NNo. oAA.- omA.1 vo. oAm.- Non pace AomonA
mmA. oom.- AAA.1 ooA. moo. ANAo aoAumoAUAAAAA voamA
vo. Noo. moA.- oNN. oAN.- Avo ooovo Acmaoo=a<
AAo. moo.1 Amo.1 AoN. oAm.- Aon voAom
mom. mAA.- mmo. moo.1 mNo.1 AAAV omAAomono auom
NmA. NAo.- ooA.- omA. AAm.- Aon aoAquAoAcAEAA 0AAAAA
mmo.- AAo. NNo. mmo.1 ooo. ANAV 23325 “858?vo A33
moo. AAo. NAA. mNo. NNA.- AAAV m..ASAAAz
mmo. oNo.1 NAN.1 omN. Nmo.1 Avo omNAAmaoAAAAAAocA
Amo. ooN.1 NNA.- mAN. ooN.1 Aoo ommAAoo
NAN.1 ANA.1 omm.- NmN. mmo. Amo ADV oaouaA sumo Aonuo
NmA. mNN.- mmo. mmA.- moo. ANV mmAAomoAmo maAusuoomsaoz
ANN. mmo.1 mmo.1 Nmo.1 ooo.- Aoo oeuAA maAAaAumoaamz
oAA. NmA.- AAN. mmm.1 Aoo. Amo AaoaNOAoam chAsAooAAamz
mmm. Amo.1 Aoo. ANN.1 AmA. AAV NAAvo>A
AHOHUNNHAHW HO

Noo.1 Nmo. oNo.- ooo.- oNo. Amo omavo coAqunooo sumo
mNA. mmN.- oNA. NNA.- AAA.1 ANV omavo aoAqusooo summ1aoz
Nmm. oNN.- moA. mNA.- Noo.1 AAo AOAooAmAz1Amz
ANmo Aomo Ammo AAmo Ammo

 

 

 

 

AmmAazooo ooo mAAz<m1mom AamAmAmmm>Ano zAAmo mom ozoAAAAmmmoo mmamo1ommm

AomsaAucooo oA AAAAA

HH NHszmm<

 

146

ooo.A mNA.- Aoo. NAA.- ooo. ANmo AAV mwwmomwmwowwAmmwmmwam

mNA.- ooo.A ANo. oAA. Nmo. Aomo AAV AamaNvoamao
Amo. ANo. ooo A omo.- mNA. Ammo ooo oaoA
NAA.- oAA. omo.- ooo.A mAN.- AAmo aoAumAAoAcAaoA AAAAAA
ooo. Nmo. mNA. mAN.- ooo.A Ammo unmamvoam waAAsAomAAaoz
mmA.- vo.1 Aoo. ooA. NmA. ANmo AOAumoAuAAAmA voamm
Aoo.- ooo. Amo.1 ooN.- oAA.- AAmo omNAAoaoAusoAuoaA
on. omo. NoA. Nmo. mAN.- Aomo NAAAAAAz
mmo. oAN.- NAN.1 oom. NoN.- AoNo ommAAoo
NmA. NNm.1 vo. mNo.1 NmA.- AmNV asom AommAmA
AGH owed

 

owawzo HmcoHumNAcm A0

 

 

omH. mmo.1 omo.1 mHH. vo.1 ANNV oumumnoucH
NNA. AoA. moA. NAA.- vo.1 AoNo Amoco>om onAm
mmo. Amo. ooH. HoH. mNo.1 Ava ucoahoHaam Houooom
AmN.- ooo. Aoo.- NAo.- NoA.- AANV mNmAuso Amumoom
oAA. ooo.- Nmo.1 Noo.1 omo. AmNo mAAo oA mooauooz
moA. mmo.1 NNo. ANA.1 mNA. ANNV <mzm ou ooocummz
unmadouH>cm
ANmo Aomv Ammo AAmv Ammo

 

iihlI
I‘HI‘

AmmAAzooo ooo mAmz<m1mom AnmAmAmmm>on zoom: mom mzoAAAAmmmoo mmnmo1ommm

 

 

AomacAuaooo oA mAm<A

HH xHazmmm<

APPENDIX III

LEVELS OF SIGNIFICANCE

APPENDIX III
LEVELS OF SIGNIFICANCE

Although this study has been concerned with explain-
ing the variance in net-migration among sample counties,
levels of significance have been computed to test for
generalizability for all zero-order correlation coefficients,
change in R2 in the regression analysis for all variables
adding at least one per cent to the explanation of variance
in the dependent variable and all partial correlation
coefficients. Results indicate that the power of the tests,
in this case through the sample size, has a strong effect
on the levels of Significance found.

Specifically, for the zero-order correlations, all
associations between independent variables and net-migration
for the entire sample (N=227), rural sub-sample (N=167) and
urban sub-sample (N=60) at or above .113, .142 and .230
respectively are significant at the .05 level. Moreover,
correlations of .236, .230 and .450 or higher for the sample,
rural and urban sub-samples respectively are also signifi-
cant at the .001 level.

All variables explaining at least one per cent of
the variance in their respective regression equations are
significant in terms of change in percentage explained at

147

148

the .001 level with two exceptions; among variables of
organizational change, both change in institutionalized

and change in college population are significant at the

.025 level. All partial correlation coefficients similarly
are significant at the .001 level with two exceptions, near-
ness to SMSA (F=l.450) and federal outlays (F=.O82). These
two exceptions, however, would be expected given both
variables' negligible partials and the points at which they
were entered into the multiple regression equation combining

measures of organization and environment.

APPENDIX IV

STEPWISE MULTIPLE REGRESSION ANALYSIS
UTILIZING INDEPENDENT VARIABLES
GROUPED INTO SUBSTANTIVE
CATEGORIES

APPENDIX IV

STEPWISE MULTIPLE REGRESSION ANALYSIS
UTILIZING INDEPENDENT VARIABLES
GROUPED INTO SUBSTANTIVE

CATEGORIES

Although the variables used in this study corres-
pond to particular components of the ecological complex,
analysis shows a strong degree of collinearity among some
variables such as those measuring manufacturing. Somewhat
differently, other variables both similar to each other
yet included under different ecological rubrics show similar
relationships with migration such as several of the indices
of governmental inputs. In order to understand more fully
how such related variables affect the net-migration rate,
all variables have been grouped into five categories
(employment and economic differentiation, institutional,
governmental, urban access, recreational) according to
their substantive characteristics ignoring their placement
in the ecological schema. Using stepwise multiple regres-
sion to analyze these five sets, each set has been forced
into the equation at a particular time due to theoretical
considerations and empirical results from this study and

others; however, within the respective sets each variable

149

150

has been entered on the basis of its partial correlation
coefficient squared as in the ecological analysis. The

categories and rationale for entering each set at a par-
ticular step are listed as follows in the order in which

they are entered into the equation:

Employment and Economic Differentiation Variables

 

This set has been entered first because of the
explanatory importance of particularly diversity and
manufacturing in this and other studies. Moreover,
diversification, manufacturing activity and a high
standard of living have long been posited to be of major

importance in attracting migrants.

Institutional Variables

The relative importance of variables pertaining
to institutions in this study and others suggests that
their existence contributes to stemming the tide of out-
migration from nonmetropolitan counties, although the
influence of institutions appears not to be as strong

as the previous set of variables.

Governmental Variables

 

Entered on the third step, this set is meant to
tap the observed increasing influence of government in
terms of both employment and expenditures. However, the
ecological analysis indicates that governmental inputs
are not as important in explaining the variance in the

net—migration rate as are the two sets entered before it.

151

Urban Access Variables

 

Although change in size of largest town consistently
yields the highest correlation with net-migration through-
out the ecological analysis, the negligible effect of the
other urban access variables when combined with measures
included within the three preceding categories suggests
that these factors have a high degree of collinearity with
other variables that explain the variance in migration

more fully.

Recreational Variables

 

Despite observations that some rural areas are
beginning to grow and prosper due to recreational activi-
ties, either the two particular variables used here do not
tap this prosperity or the development of recreational
facilities and increasing prosperity have little effect on
population change through migration.

Results of stepwise multiple regression show that
all 34 independent variables together explain 64.5 per cent
of the variance in the net-migration rate (see Table 19).
MOreover, measures of employment and economic differentia-
tion explain 40.9 per cent of the variance with diversity
and manufacturing employment contributing the greatest
amount while the addition of institutional variables,
particularly those pertaining to military and college
population change, raises the percentage explained to
50.4 per cent. That the contribution of governmental

variables is smaller than those measuring urban access

152
suggests that perhaps the latter set Should have been entered
before the former. However, among urban access variables
change in size of largest town contributes by far the
greatest amount to the explanation of variance, and, as
noted in the ecological analysis, it appears that this
variable may be more of an indicator of population change
than of diversity or urbanity. Finally, the two measures
of recreational activity add almost nothing to the amount
of variance explained by the four sets of variables entered

before them.

APPENDIX IV

TABLE 20

RESULTS OF STEPWISE MULTIPLE REGRESSION

 

 

 

 

 

 

Multiple R2 R2 Beta
R Change Weight
Employment and Economic
erentIationa
DiVersity .518 .268 .268 .166
Manufacturing Employment .566 .321 .053 .097
Other Farm Income (t) .586 .344 .023 '.O82
Unemployment (t) .609 .371 .027 .060
Change in Farm Size (t) .619 .383 .012 .111
Change in Female
Participation .624 .390 .007 .054
Female Participation .628 .394 .004 -.O47
Change in Land Use ' .631 .398 .004 -.082
Change in Manufacturing
Employment .633 .400 .002 -.020
Income Under $3000 (t) .635 .403 .003 .068
Manufacturing Firms .636 .404 .001 .038
Manufacturing Categories .639 .408 .004 .232
Change in Unemployment (t) .639 .409 .001 .016
Public Relief (t) -.O82
Farm Categoriesb -.023
Institutional8
Change in College .687 .471 .062 .102
Change in Military .703 .495 .024 .140
College .707 .500 .005 .060
Military .709 .502 .002 .093
Institutional .710 .503 .001 .029
Change in Institutional .710 .504 .001 .036
Governmenta
Local Government
Employment .736 .541 .037 -.l61
Public Administration .738 .545 .004 .004
Change in Public
Administration .740 .547 .002 -.O9l
State Revenues .740 .548 .001 .007
Federal Employment .741 .549 .001 .030
Federal Outlays .741 .549 .000 -.051

153

154
APPENDIX IV

TABLE 20 (Continued)

RESULTS OF STEPWISE MULTIPLE REGRESSION

 

 

 

 

 

Multiple R2 R2 Beta
R Change Weight
Urban Accessa
’CHange in Largest Town .788 .622 .073 .317
Largest Town .800 .640 .018 -.274
Nearness to SMSA .801 .642 .002 .091
Interstate .801 .642 .000 .017
Nearness to City .801 .642 .000 -.020
Recreation
HOtéIs .802 .644 .002 .070
Amusement Places .803 .645 .001 -.052

 

aR2 Change computed as the difference between R2 of
last variable entered in this set and the previous set is
significant at the .001 level.

bAlthough the computer program employed should have
forced all variables in this set into the equation before
entering variables in the next set, these two variables,
due to a combination of low tolerances and small F-ratios,
were added with institutional variables. See Norman Nie,
Dale H. Bent, C. Hadlai Hull, Statistical Packa e for the
Social Sciences (New York, 1970), p. 183. SpecifiEEIly,
public reIIef was entered after change in college and farm
categorfies was added after college; in both cases Multiple
R and R were identical to those of the respective variables
preceding each and neither added anything to the explained
variance. In determining significance levels of R2 Change,
however, these two variables were included with institu-
tional variables; such inclusion would have the effect of
decreasing the significance of R2 Change contributed by
this second set.

 

APPENDIX V

EXPLICATION OF VARIABLES

APPENDIX V
EXPLICATION OF VARIABLES

The definitions of variables used in this study can
be found in the various sources and for the most part are
selfevident. However, several factors may need further
explication for those not familar with these sources.

Netdmigration refers to the differential between
those entering and those leaving a county. In other words,
if there were two counties such that one experienced no
‘migration while the other received 100 inrmigrants but lost
100 people through outamigration, net-migration in both cases
would equal zero. The net-migration rate in this study is
measured as net numbers per 1000 inhabitants.

Number of manufacturing categories includes 21
different industrial groups. They are as follows: ordnance
and accessories; food and kindred products; tobacco manu-
facturers; textile mill products; apparel and other textile
products; lumber and.wood products; furniture and fixtures;
paper and allied products; printing and publishing;
chemicals and allied products; petroleum and coal products;
rubber and plastics products, n.e.c.; leather and leather
products; stone, clay, and glass products; primary metal

industries; fabricated metal products; machinery, except

155

156

exectrical; electrical equipment and supplies; transporta-
tion equipment; instruments and related products; miscel-
laneous manufacturing industries.

Change in farm category is based on an eight-fold
classification of farm type. The eight categories are as
follows: cash-grain, tobacco, other field crop, vegetable,
fruit and nut, poultry, dairy, and livestock other than
poultry and dairy.

The unemployment rate refers to the percentage of
the work force unemployed in a county. Specifically, the
Census Bureau counts as unemployed all civilians not "at
work" but looking for work during the four weeks preceding
the census count and available to accept jobs in the
civilian labor force. The minimal age for being considered
in the work force was 14 in 1960 and 16 in 1970.

The female participation rate is the percentage of
women who are in either the armed forces or civilian labor
force. It includes women who are both employed and un-
employed (see previous paragraph).

Percentage public relief recipients is the percent-
age of county residents receiving aid from either joint
federal, state and local undertakings or from local programs
without federal participation. Such aid includes old age
assistance, medical assistance for the aged, aid to
dependent children, the blind, and the permanently and

totally disabled.

157

Three variables deal with institutionalized popula-
tions. Percentage military includes all residents who are
in the armed services. Percentage college includes all
those to age 34 enrolled in a college as a percentage of
the entire population. Percentage institutionalized
refers to all institutionalized individuals with the excep-
tion of those in either the armed forces or college dormi-
tories. This variable encompasses such institutions as
homes, schools, hospitals or wards for juveniles, the handi-
capped, mental and chronic diesease patients, and tubercu-
losis patients; homes for unwed mothers and for the aged
and dependent; and correctional institutions.

Two variables refer to outlays from nonlocal govern-
mekts. State revenues includes general revenue received
from the state government, usually fiscal aid in the form
of grants-in-aid and shared tax proceeds. Secondarily it
also includes amounts received for services performed for
one government by another on a reimbursement or cost sharing
basis and payments received by the county in lieu of taxes.
Federal outlays encompasses all federal funds received by
local governments and individuals in a county except for
those monies connected with "federal influence" activites.
Such activities indlude the current market value of
donated commodities, the original acquisition cost of
donated surplus real and personal property, and the face

value or contingent liability of guaranteed/insured loans.

BIBLIOGRAPHY

BIBLIOGRAPHY

Beegle, J. A.; Marshall, Douglas; and Rice, Rodger.
"Selected Factors Related to County Migration
Patterns in the North Central States 1940-50 and
1950-60." Michigan State University Agricultural
Experiment Station Quarterly Bulletin, 46 (November,
1963), 206-41.

Berry, Brian J.L.; and Pred, Allen. Central Place Studies:
A Biblio ra h of Theor and Apfilications.
PhIIadeIpEIa: Regiona cience Institute, 1965.

Blalock, Hubert M., Jr. Social Statistics. New York:
McGraw-Hill, 1960.

 

 

 

Bogue, Donald J. Components of Po ulation Change, 1940-50:
Estimates nget Migration an Natura Increase for
Each’Standard Metropolitan Area and State EConomic
Area. OxfOrd: Scripps Foundation, 1

 

 

Bright, Margaret L., and Thomas, Dorothy S. "Interstate
Migration and Intervening Opportunities." American
Sociological Review, 6 (December, 1941), 773-83.

 

Brown, Robert Harold. Political Areal Functional Organiza-
tion: with special reference to St. Cloud,
Minnesota. Chicago: University Of CEIcago Press,
I957.

Burford, Roger L. "An Index of Distance as Related to
Internal Migration." Southern Economic Journal,
29 (October, 1962), 77-81.

Commission on Population Growth and the American Future.
Po ulation and the American Future. New York:
TEe New Afierican Library, 1972.

Duncan, Otis Dudley. "Human Ecology and Population Studies."
The Stud of POpulation, an Inventory and Appraisal.
‘EHIte y Philip M. Hauser and OEIS DuaIEy Duncan.
Chicago: University of Chicago Press, 1959.

158

 

 

159

Duncan, Otis Dudley, and Schnore, Leo F. "Cultural,
Behavioral, and Ecological Perspectives in the
Study of Social Organization." The American
Journal 9f Sociology, 65 (SeptemBET, I959),

 

Duncan, Otis Dudlye; Scott, W. Richard; Lieberson, Stanley;
Duncan, Beverly D.; and Winsborough, Hal H.
Metropolis and Region. Baltimore: Johns Hopkins,
1960.

Duncan, Otis Dudley, and Reiss, Albert J., Jr. Social

Characteristics of Urban and Rural Communities,
1950. New York: John Wiley &*SOns, I956.

 

 

 

 

Federal Outla s, a report of the Federal Government's
Im

act by State, Cafinty andIIar e City, Fiscal
Year I9 . washington: U.S. ice of Economic

Opportunity, n.d.

 

Fuguitt, Glenn V. "County Seat Status as a Factor in
Small Town Growth and Decline." Social Forces,
44 (December, 1965), 245-51.

Gibbs, Jack P., and Martin, Walter T. "Toward a
Theoretical System of Human Ecology." Pacific
Sociological Review, 2 (Spring, 1959), 29-36

Gordon, Robert A. "Issues in Multiple Regression." The
American Journal 2; Sociology, 73 (March, 1968),

 

Groth, Philip G. "Functional Classifications of Counties:
Some Applications." Unpublished paper, Department

of Rural Sociology, University of Wisconsin, Madison,
May 26, 1972.

Hawley, Amos H. "Human Ecology." International Encyclopedia
of the Social Sciences. VoI. 4, New York:
MECMIIlan, I967, 328-337.

 

Hawley, Amos H. Human Ecology: A Theor of Community

Structure. New or : RonEl ress, 1950.

 

Irwin, Richard. "Nonmetropolitan Population Change:
1960-70." Paper prepared for presentation at the
annual meeting of the Population Association of
America, Washington, D.C., April 23, 1971.

Isbell, Eleanor Collins. "Internal Migration in Sweden
and Intervening Opportunities." American
Sociological Review, 9 (December,Il944), 627-39.

 

160

Fuguitt,Glenn V. "County Seat Status as a Factor in Small
Town Growth and Decline." Social Forces, 44
(December, 1965), 245-51.

 

Jacobs, Jane. The Death and Life pf Great American Cities.
New York: Vintage, l96I.

 

 

King, Harriet, and King, Terry. "New Town, Mon Amour."
The Chicagoan, November, 1973, pp. 78-83.

Lampard, Eric E. "Historical Aspects of Urbanization."
The Study pf Urbanization. Edited by Philip M.
Hauser and Leo F. Schnore. New York: John Wileyé:
Sons, 1965.

 

 

Lemon, James T. "Urbanization and the Development of
Eighteenth-Century Southeastern Pennsylvania and
Adjacent Delaware." The William and Mary Quarterly,
3d ser., XXIV (OctobefT—19675, 507:33.

 

Levitan, Sar A., and Houghteling, Louise D. Factors
Affecting the Slower Growth pf Missouri Population
ComparevaIth Ehe United States. Rev. ed.
Washington: Library of7Congress Legislative
Reference Service, 1961.

 

 

 

Lyford, Joseph P. The Talk in Vandalia: The Life p£_§p
American Town. New YErE: Harper, I965.

 

Philbrick, Allen K. "Principles of Areal Functional
Organization in Regional Human Geogra hy."
Economic Geography, 33 (October, 1956 , 299-336.

Rand-McNally Road Atlas. Chicago: Rand-McNally, 1971.

Reiss, Albert J., Jr. "Functional Specialization of
Cities." Cities and Society: The Revised Reader
in Urban Sociology. Edited by Paul K. Hatt and
AIbert J. Reiss, Jr. New York: The Free Press
of Glencoe, 1957.

 

 

Rutman, Gilbert L. "Migration and Economic Opportunities
in West Virginia: A Statistical Analysis," Rural
Sociology, 35 (June, 1970), 206-17.

 

Sly, David F. "Migration and the Ecological Complex."
American Sociological Review, (October, 1972),

 

Thomlinson, Ralph. Urban Structure: The Social and
Spatial Character pf Cities. New YOrk: Random
HOuse, 1969.

 

 

 

161

U.S. Bureau of the Census. County and City Data Book,

 

1967. (A Statistical Abstract Supplement).

Washington: Government Printing Office, 1967.

U.S. Bureau of the Census. U.S. Census of A riculture,
I OffIce

 

1959. Washington: Government Pfint ng , 1961.

U.S. Bureau of the Census. U.S. Census of Agriculture,

 

 

1964. Washington: Government PrIfiting Office, 1967.

 

U.S. Bureau of the Census. U.S. Census of Agriculture,

1969. Washington: Government Prifit1ng 1ce, 1972.

U.S. Bureau of the Census. U.S. Census of County Business,

1967. Washington: GovernmEfit Prififing ice,

U.S. Bureau of the Census. U.S. Census of Governments,

 

 

1967. Washington: Government Prififing Office, 1968.

U.S. Bureau of the Census. U.S. Census of Manufactures,

 

 

1967. Washington: Government Prififing Office, 1971.

U.S. Bureau of the Census. U. 8. Census of Population:

 

 

1950. Washington: Government Printing Office, 1952.

U.S. Bureau of the Census. U. S. Census of Population:

 

 

1960. Washington: Government’Printing Office, 1964.

U.S. Bureau of the Census. U. S. Census of Population:

 

1970. Washington: GovernmentIPfinting Office, 1971.

U.S. Department of Agriculture, Economic Research Service,

Warner,

Economic Development Division. The Economic and
Social Condition of Rural America In tEe I970's
Prepared fOr’Ehe U. —S. Senate Committee on Government

Operations. Washington: Government Printing Office,
1971.

 

Sam Bass, Jr. The Urban Wilderness. New York:
Harper and Row, I972.

 

Wirth, Louis. "Urbanism as a Way of Life." The American

Zuiches,

 

Journal pf Sociology, 44 (July, 1938), 1-24.

 

James J. "In-Migration and Growth of Nonmetro-
politan Urban Places." Rural Sociology, 35
(September, 1970),410-20

 

   

HICHIGRN STRTE UNIV. LIBRQRIES
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIWIIIIIIIIIHIIIIHIIIII
31293101270688