A communications model of democratic political development in the United States

communicationsmo00marqrich ( .pdf )

A Cor.cunications Model of Democratic Political Develocrment
in the United States: 1870 1960













.By

JESSE F. MiARQUETTE












A DISSERTATION PRESENTED TO 'liE GrADUATei 2 COUNCIL OF
Ti1E UNIVERSITY OF FP LRI IN PARTIAL
FULFILLMENT OF THE REQUIRE S FOR THE -DEGREl-E OF
DOCTOR O1 PIl LOSOPtY


UNIVERSITY OF 'LORIDA
1971













ACKNOWLEDGMENTS


Two people provided the support and encouragement which

enabled me to complete this work: Dr. Alfred Clubok, the Chairman

of n.y Dissertation Committee, and Penny Marquette, my best friend.

T wold also like to thank the members of my Con iitteo,

Professors Frank Mlunger, Manning Dauer, John Spanier and Joseph

Vandiver for giving me their suggestions and their time.











TABLE OF CONTENTS



ACK"NOiL EDGMENTS ............... ................. ................ ii

LIST OF TABLES...... ........ ................ ... ............. iv

LIST OF FIGURES........................................ vi

ABSTRACT.. . ............... ........ .. ... .. ....... ....... x

INTRODUCTION. ..................................... ..... ........ .

CHAPTER I: A Philosophical Framework and a Definition of
Political Development................................4

CHAPTER II: Social Change and the Insatiable Personality..........29

CHAPTER III: Indices and Methodology ..................... .....55

Data Sources ....................... .............. 57

Variables .......................... .............. 59

Methodology.. ............ ......... ..... ......94

CHIAPTER IV: Models of Political Development .......................99

Basic Hypotheses.................. ............... 100

An Inter-Temporal Test of the Model of Development...133

CHAPTER V: Discussion and Conclusions ......................... 140

Conclusions...................................... .. 149

APPENDIX.......... .... .. ... ................... ...... ....... 161

BIBLIOGRAPHY....................................... ....... ... .... 164

BIOGRAPHICAL SKETCH ..................... .............. .......... 170










iii











LIST OF TABLES


I Decennial Correlations of Inter-Party
Competition and Education ......................55

II Number of Observations for Each
Decennial Census Period.......................57

III Original Variables Selected as
Possible Indicators ............................60

IV Means and Standard Deviations of Indicators
of Urbanization by Census Years................67

V iMans and Standard Deviations of the Index
of Government Activity by Census Years......... 71

VI Variables Included in Each Final Index
with Years of Inclusion............. ..........93

VII Causal Inference Prediction Equations
for Each Hypothesis ............................ 101

VIII Aggregate Correlation Matrix of the Six
Indices for the Period 1870 to 1960.............102

IX Aggregate Correlation Matrix of Six Indices
for the Period, 1870 to 1910...................108

X Aggregate Correlation Matrix of Six Indices
for the Period, 1910 to 1960...................110

XI Prediction Equations for the First and Second
Models of Development ......................... 113

XII Inter-Censal Correlations of Governmcut
Activity at T1 with Political Participation
at T2 for the Period, 1870 to 1960.............. 119

XIII Multiple Correlation Coefficients of Political
Participation with Indices of Education, Media
Consumption, Government Activity, Urbanization
and Industrialization by Census Years..........120












XIV Multiple Correlation Coefficients of
Government Activity with Media Consumption,
Urbanization and Industrialization by
Census Years ...................... ........... 321

XV Cross Year Correlation Matrix for the
Period 1870 to 1960............................122

XVI Correlation Matrix of Original Variables
with Logarithmic Transformed Counterparts......126

XVII Correlation Matrix of Six Variables
Using All Observations.......................129

XVIII Multiple Correlation Coefficients for the
Three Models of Development....................132

XIX Aggregate Correlation Matrix of the Inter-
Temporal Correlations: 1870 to 1910.............134

XX Aggregate Correlation Matrix of the Inter-
Temporal Correlations: 1920 to 1960........... 136

XXI Correlation Matrix of Media Consumption,
Government Activity, and Political
Participation, 1960.......................... 146

XXII Relationships of Mobilization Levels and
System Types as Crisis Outcomes in Six Nations.153













LIST OF FIGURES


1 Total Pig Iron Shipments in the United
States: 1820 1960 ...........................45

2 Causal Model of Three Concepts with
Simultaneous Interactions...................... 46

3 Causal Model of Three Concepts with One-Way
Causation...................................... 46

4 Causal Model of Three Concepts with Inter-
Temporal Effects ............................. 47

5 Inter-Temporal Effects of Three Operational
Indicators.................................... 50

6 Inter-Temporal Effects of Four Socio-
Economic Indicators ........................... 51

7 Causal Relations of Four Variables
and Government Activity.........................52

8 Final Representation of a Six-Variable
Model of Political Development.................54

9 Mean Per Capita Productivity from
Manufacturing.................................... 64

10 Mean Percent of Population in Cities over
2,500 and over 25,000 by Census Years..........66

11 Mean Values of the Index of Urbanization
by Census Years ............................... 70

12 Mean Daily Newspaper Circulation per 1,000
Population by Census Years....................73

13 Mean Per Capita Postal Receipts by Census
Years .............. ... ............ ...........75

14 Mean Number of Telephones per 1,000
Population by Census Years ..................... 77








15 :loan Values of the Index of Media
Constu option by Census Years ................... 79

16 %lean Per Capita Government Expenditures
by Census Years ...............................81

17 'an Percent of Population Who Are Literate
and Mean Percent of Population Attending
Iligh School by Census Years ................... .81

18 Mean Values of the Indcx of Education
by Census Years................................ 86

19 Mean Values ot Turnout in Presidential
Elections by Census Years ...................... 90

20 Mean Values of Congressional Election
Turnout by Census Years ........................ 92

21 Empirically Identified Causal lodel of
Development in the United States for the
Period 1870 to 1960............................104

22 Plot of Correlation Coefficients of Media
Consumption, Government Activity, Education,
and Political Participation Versus Time:
1870 1960....................................107

23 Empirically Identified Model of the
Development Process in the United States:
Phase I 1870 to 1910......................... 110

24 Summary Presentation of Both Phases of the
Causal Mode 1 ................................ 114

25 Mean Values of the Index of Education:
1370 to 1960 .................................. 124

26 Logarithmic Transformations of the Mean Values
of Media Consumption, Industrialization and
Government Activity: 1870 to 1960............... 128

27 Empirically Identified Model of the Development
Process in the United States Using All
Observations as Data Base. ... ..................131


vii












28 Empirically Identified Model of the
Development Process in Phase I, 1870 to
1910, Using Inter-Temporal Correlations ........135

29 Empirically Identified Model of the
Development Process in Phase II, 1920 to
1960, Using Inter-Temporal Correlations ........138

30 First Section Causal Model Resulting
from Total Observations Test.................. 141

31 Summary Presentation of Phase I Model,
1870 to 1910 .................................143

32 Summary Presentation of Phase II Model,
1920 to 1960................ .................144

33 Summary Presentation of Model Generated
from Total Observations Matrix ............... 144

34 Model of Expected Third Phase Relations ........146


viii
















Abstract of Dissertation Presented to the
Graduate Council of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy


A COMMUNICATIONS MODEL OF DEMOCRATIC POLITICAL DEVELOPMENT
IN THE UNITED STATES: 1870 1960

By

Jesse F. Marquette

August, 1971

Chairman: Dr. Alfred B. Clubok
Major Department: Political Science

Political development theory has consistently suffered from

a lack of reliable empirical referents against which to check proposed

concepts. One major reason has been that comparative research has

generally been cross-sectional. A second major reason has been that

thes research has been performed rising nations as data points. From

this cross-societal sample, attempts have been made to infer intra-

societal developmental sequences.

This research has attempted to overcome these failings. Only

one society, the United States, is used as a referent. Data points are

the states of the United States. The time period of the research

extends from 1870 to 1960. Obviously, generalizing from one society's

past experiences with change to the experiences of other societies is







somewhat questionable, but certainly less so than the opposite

process. The time period used is adequately long to justify expecta-

tions that thi generalizations will be reliable. Explanations of the

relationships examined are as parsimonious as possible, and generally

refer only to the physical necessity of certain processes. A constant

attempt is made to avoid any explanations which are ethnocentric

despite the obviously culture-bound basis of the data.

The primary objective has been to develop a causal model of the

relationship amongst four social variables: industrialization, urban-

ization, media consumption, and education; and two political variables,

government expenditures and political participation. In essence, ten

tests of the proposed model were made, one test for each decennial

census period from 1870 to 1960.

Two now, consistent, sequential and simple models of develop-

ment are apparent. The first phase model extends from 1870 to 1910,

and the second extends from 1920 to 1960.

A reformulation of developmental theory is undertaken in the

light of these findings.












INTRODUCTION


Theories of political development have consistently suffered

from a lack of reliable empirical referents against which to check

proposed concepts. One methodological failing has been that while

development implies change, most comparative research has been cross-

sectional. A second major failing relates to the use of nations as

data points, attempting to infer intra-societal sequences of development

from cross-societal samples. When both these failings occur in the

same research the results are, at best, confusing.

The research reported here attempts to overcome these failings.

Only one society, the United States, is used as a referent. The time

period of the research extends from 1870 to 1960. with the individual

states as data points.

Although generalizing from one society's experiences with change

to the experiences of other societies is somewhat questionable, it is

certainly less questionable than the opposite process. This research

has attempted to deal with these problems of cross-societal comparisons

in several ways. The time period used is adequately long to give the

generalizations high reliability. Further, the explanations of the

relationships found are as parsimonious as possible, and generally

refer only to the existential necessity of certain processes. In

addition, a conscious attempt has been made to avoid explanations which










are ethnocentric despite the obviously culture bound basis of the

data.

The objective of this research is the development of a causal

model of the relationship among four social variables (industrialization,

urbanization, media consumption, and education) and two political

variables (government activity and political participation).

The study is founded on a proposed model derived from work by

Cnudde and McCronel and an earlier work by this author,2 while the

theoretical basis is derived from a conceptualization advanced by

Lerner.3 In essence, ten tests of the proposed model were made, one

test for each decennial census period from 1870 to 1960.

Empirical research quickly proved both the model and the theory

to be unfounded. However, two new, consistent, sequential and simple

alternative models became apparent. Using these two models an attempt

is made to reformulate developmental theory in the light of the findings.

Chapters one and two present part of this reformulated theory as a

basis for understanding the models which follow. After presenting

the reformulated theory and the static models, a series of hypotheses

will be presented regarding the relations of rates of change. These

hypotheses will then be tested over the full hundred year time span.



Donald J. McCrone and Charles F. Cnudde, "Toward a Communications Theory
of Democratic Political Development: A Causal Model," American Political
Science Review (March, 1967), 72-79.

Jesse F. Marquette, "Social Mobilization and the Philippine Political
System," Comparative Political Studies (July, 1971), in press.

Daniel Lerner, "Communications Systems and Social Systems. A Statistical
Exploration in History and Policy," Behavioral Science (March, 1957),
266-275.





3





In this manner, an attempt is made to maintain a constant and reinforcing

interaction between theory and research.











CHAPTER I

A PHILOSOPHICAL FRAMEWORK AND A DEFINITION
OF POLITICAL DEVELOPMENT



An effective definition of political development requires an

effective definition of a developed polity. Development implies change

fr1o one state of political reality to another and we should be able

to define both states. Definitions of political development have

proliferated due to an inability to settle on one definition of a

developed polity.1 This results from the generally accepted opinion

that no fully developed polity currently exists. If, in fact, there is

no fully developed polity, then our definitions and descriptions must

be prescriptive; not what is, but what should be. The referents of

"what sho-ld be" ere highly conditioned by the intellectual heritage from

which we proceed. Thus, an economist like Rostow2 sees political change

primarily as an epiphenomenon of social change. A political scientist

like Organski3 sees political development as a necessary concomitant of



Lucian W. Pye, Aspects of Political Development (Boston: Little, Brown
and Compmny, 1966), 31-48.
2
Walter I. Rostow, The Stages of Economic Growth: A Non-Communist
n;lfe sto (Cambridge, England: Cambridge University Press, 1966).

A.F.K. Organski, The Stages of Political Development (New York: Alfred
A. Knopf, 1965).











economic development. A psychologist like McClelland4 views

development as a result of altered personality patterns, while a

political sociologist like Apter5 sees development as changing

patterns of group conflict.

There are many other ways to categorize existing definitions

of development, and rather than examining them as above, they can be

categorized according to the type of explanation used. Holt and Turner6

argue that the works of Pye, Hagen and McClelland are reductive

explanations;7 attempts to explain macro-social phenomena in terms of

generalizations founded on micro-social phenomena. Holt and Turner

argue that this form of explanation is ineffective because it lacks

composition laws. They suggest that emergent explanations are the best

currently available in social research.8

Depending upon the purpose, many more systems of categorization

can be developed. In developing a causal model of the socio-economic

bases for political change, there is one system of categorization which

clarifies much of the previous work. This system of categorization is

4

David MicClelland, The Achieving Society (Princeton: D. Van Nostrand
Company, Inc., 1961).
5
David Apter, The Politics of Modernization (Chicago: Chicago University
Press, 1965).
6
Robert T. Holt and John E. Turner, The Political Basis of Economic
Development (Princeton: D. Van Nostrand Company, Inc., 1966).
7
Ibid., 26-27.
8I
Ibid., 28.









based on the alternative perceptions of the social locus of political

change held by various authors. These divergent perceptions lead to two

different models of the causal relations between social change and

the political order.9 One alternative emphasizes the significance

of minority innovation and support of change. The minority emphasis

sees either an elite or a special minority group as a continuous element

in the process of modernization. This minority constantly provides the

drive for modernization over long periods of time. Mass behavior is

affected by minority desires, but the mass is not expected to become a

causal agent in further change.

A second alternative sees each newly mobilized individual as

a causal agent in the further process of mobilizing even more individuals.

This opposite perception emphasizes the significance of mass attitudinal

and behavioral change as fundamental to continuous political change.

Generally, the practice has been to emphasize one view or the other;

however, two authors, Nettl and Apter, have attempted to synthesize

these divergent perceptions.

J.P. Nettl outlines two alternative forms of political mobiliza-

tion, one referred to as stalactite and the other as stalagmite.10 In

stalactite mobilization an elite is responsible for deliberately moving


9
The minority emphasis is represented by such authors as Daniel Lerner,
E.E. Hagen and Lucian Pye, while the mass emphasis is represented by
Karl Deutsch, David Mc.Clelland and Gabriel Almond.
10
J.P. Nettl, Political Mobilization: A Sociological Analysis of Methods
and Concepts (New York: Basic Books, 1967), 288-316.










the mass of the populace toward political activity. This activity is

directed and focused according to the desires of the mobilizing elite.

Stalagmite mobilization, on the other hand, is mass-based and anti-

elitist; it is the awakening of the mass of the populace, shaped by

other forces. The institutional forms of political life resulting from

these alternative forms of mobilization are radically different. While

stalactite mobilization is seen as eventuating in undifferentiated,

elitist polities, stalagmite mobilization results in differentiated,

constitutionally based polities.11

David Apter attempts a similar synthesis with his concepts of

the reconciliation system12 and the mobilization system.13 In Apter's

rubric, the term reconciliation system is analogous to Nettl's

stalagmite mobilization. Here an attempt is being made to gradually

accommodate to the rising demands of the populace. Over a period of

time, demands are articulated by more and more groups in the system.

Since the general value orientation of this system is pragmatic and

meets already extant problems, Apter regards the value system as

instrumental.14 Apter's mobilization system is analogous to Nettl's


11
Ibid., 315.
12
Apter, op.cit., 391-421.
13
ibid., 357-390.
14
Apter defines an instrumental value system as one vhich emphasizes
the solutions to relatively mundane problems. Here, politics is
viewed as the process of solving short-term, existing problems. A
consumatory value system, on the other hand, is expected to place
great emphasis on realizing potentiality, and the ethical basis of
politics lies in the activity of achieving specified ultimate goals.









stalactite mobilization. Mobilization systems are driven by elite

desires to achieve certain goals, notably the political mobilization of

the mass. Insofar as the system is not responding to, but rather

initiating, demands, the structures which evolve tend to be undifferen-

tiated. The value system is seen by Apter to be consummatory ---

ultimate-goal oriented.

Both Apter and Nettl are unusual in that they include opposite

perceptions of the social locus of political change in their schema.

While other authors have, to some degree, discussed mass and minority

loci, they have generally emphasized one at the expense of the other.

The work of Daniel Lorner,15 E.E. Hagen,16 and Lucian Pye17

emphasize the significance of minority or-elite activity in the process

of modernization. Lerner and Pye are expressly concerned with the

activities of elites in initiating and sustaining modern behavior.

Lerner's emphasis on elite activity seems to stem from the focus of

his investigation: the significance of mass communication in the

developmental process.18 Since'the production and dissemination of


15
Daniel Lerner, The Passing of Traditional Society: Modernizing the
Middle East (New York: The Free Press, 1958).
16
E.E. Hagen, On the Theory of Social Change (Homewood, Ill.: 'The
Dorsey Press, 1962).
17
Lucian Pye, Politics, Personality and Nation-Building: Burma's Search
for Identity (New Haven: Yale University Press, 1962).
1i
Lorner, The Passing of Traditional Society; Lerner's research centers
on the Miiddle Eastern countries, all with relatively new mass
communications nets.










iass media is an elite activity, Lerner has tended to emphasize elite

behavior in explaining the process of development. Pye has also
19
suggested elite significance in the developmental process,9 equating

the identity crisis of the elite with the identity crisis of the nation.

Hagen,20 on the other hand, does not limit minority interpretation to

a discussion of elites. The minority groups which Hagen sees as

instrumental in the process of change are characterized by radically

different attitudes, but not by elevated position.

An opposite emphasis on social locus is represented by the

conceptual schema of Karl Deutsch,21 David McClelland22 and Gabriel

Almond.23 Their theoretical concern is with the significance of mass

attitudinal and behavioral changes in effecting the process of modern-

ization. Deutsch dwells on the exposure of unnobilized individuals

to aspects of modernity.24 lie is clearly concerned with the process

of mobilization, the spread of attitudes which are distinctly non-

traditional. Deutsch stresses the necessity of mass attitudinal change,

believing that while a minority .may provide the spark, significant change



19
Pye, Pc!litics, Personality and Nation Building; Pye's research group
consisted of a small sample of Burmese national government figures.
20
Hagen, op. cit.
21
Karl W. Deutsch, "Social Mobilization and Political Development,"
American Political Science Review (September, 1961), 493-514.
22
McClelland, ao. cit.
23
Gabriel A. Almond and Sidney Verba, The Civic Culture: Political
Attitudes and Democracy in Five Nations (Boston: Little, Brown and
Company, 1965).
24
Deutsch, op. cit.





10


is obtained only through mass mobilization. Gabriel Almond's presen-

tation25 is typological and little concerned with the process of

change. His primary effort involves the explication of possible

political system states, which he arrays in a sequence from less to

more developed. He attempts to describe development in terms of

mass attitudes conducive to democratic forms of political organization.

Both Netti26 and Holt and Turner27 have specifically criticized Almond

for this rather myopic view of development.

McClelland's theory of the basis of economic development is

clearly one of the most effective explanations yet offered.28 It

combines both an emphasis on the necessity of group change (cultural

values) and psychological patterns (socialization). McClelland has

attempted to combine both the forms of explanation described by Holt

and Turner; the analysis of cultural values is an attempt at emergent

explanation, while the analysis of socialization patterns is an attempt

at reductive explanation. McClelland expects different patterns of

socialization to lead to different levels of n-achievement. It is not

clear why some families adopt one pattern of socialization instead of



25
Almond and Verba, op. cit.; and also Gabriel A. Almond and G. Bingham
Powell, Comparative Politics: A Developmental Approach (Boston: Little,
Brown and Company, 1965), especially 16-41.
26
Nettl, oD. cit., 110.
27
Holt and Turner, op. cit., 13-16.
28
McCleiland, op. cit., 1-62.











another, unless there is a differential diffusion of cultural values.

At this point it is obvious that Holt and Turner's argument for the

necessity of composition laws is well taken.29 Had McClellard been

able to produce an effective explanation for the differential diffusion

of socialization patterns, his theory would have been immeasurably more

powerful. The degree of success which McClelland achieved, using a

combination of types of explanations, suggests that Holt and Turner's

use of only emergent explanations is unnecessarily restrictive.

As suggested earlier, alternative perceptions of the social

locus of change lead to different expected causal paths of change.

This contention may be illustrated using the formulations of Lerner30

and Deutsch.31 Cnudde and McCrone32 have attempted to test a causal

model derived from Lerner's work using data drawn from Cutright's

seventy-six nation study.33 Using Cutright's operational indicators of

urbanization, education, media consumption and political development,

they test a causal path represented by urbanization (U) education (E)-

media consumption (M) political development (D), conceptually derived



29
Holt and Turner, op. cit., 28.
30
Daniel Lerner, "Communications Systems and Social Systems: A Statistical
Exploration in History and Policy," Behavioral Science (March, 1957).
31
Deutsch, op. cit.
32
Donald J. McCrone and Charles F. Cnudde, "Toward a Communications Theory
of Democratic Political Development: A Causal Model," American Political
Science Review (March, 1967), 72-79.
33
Philips Cutright, "National Political Development: Measurement and
Analysis," in Nelson Polsby et al. (editors), Politics and Social Life
(Boston: Houghron Mifflin, 1963), 569-582.




12


from the following passage from Lerner.

...urbanization requires rising literacy for
individual participation. At a certain point, when
urbanization has done its work, literacy becomes the
independent variable in the process of growth and a new
phase of modernization begins. But the growth of literacy
itself, in this phase, soon becomes closely associated
with the growth of media. The media teach literacy and
growing literacy develops the market which consumes the
media product.34

The opposite path of U M -I E D, is implicit in the work of Karl

Deutsch.

...two distinct stages of the process: (1) the stage
of uprboting or breaking away from old settings, habits and
commitments; and (2) the induction of the mobilized persons
into some relatively stable new patterns of group membership,
organization and commitmentt.5

Here, the first stage is represented by exposure to aspects of modernity,

an essentially passive process. The second stage suggests a more

institutionalized resocialization of the individual. In the society

as a whole, the second stage suggests the process of education.

Although education as envisioned here is not necessarily formal education,

it must be relatively formalized (represented by stable and accepted

socialization processes), if the individuals are to be inducted into

stable new patterns of group commitments. Over time, the best measure

of this process would be the spread of formal education.


34
Lerner, The Passing of Traditional Society, 60.
35 c, ct.,
Deutsch, op. cit., 494.










These different perceptions of the path of change could have

arisen either from the authors' opposing views of the relation between

media consumption and education, or from the authors' different emphases

on social loci.

First, it is questionable that media consumption requires literacy

on any significant scale. Lerner himself describes an effective

melding of the modern media form of radio and the traditional oral

communication network when he describes the Chieftain of Balgat.36

As the nddel individual in an oral communication net, the Chief's

possession of a radio carried the entire net into a media system. There

is no reason why one literate individual could not substitute for the

Chieftain's radio in the communication network of Balgat, or any other

village.

Secondly, if the primary function of media consumption and

education is to increase individual empathy, there is no reason why

education must precede media consumption. An alternative formulation

might substitute "aspects of modernity" for "media consumption," as

does Deutsch. In Deutsch's formulation, media consumption is merely

one method of exposing the individual to modern behavior patterns.

Having become socially mobilized through exposure to aspects of modernity,

the individual would be expected to become politically mobilized after

exposure to educational opportunity. This substitution does no serious


36
Lerner, The Passing of Traditional Society, Chapter 1.




14


damage to Lerner's argument and, viewed in this way, the authors do not

seem to have fundamentally opposed views of the significance of media

consumption and education.

The preceding discussion was designed to emphasize the necessity

of delineating why a particular problem is approached in a specific

manner when evaluating the resultant conclusions. IWhen there are many

descriptions of an object, many definitions of a concept, there are

two choices in explaining the multiplicity. Either the object or

concept is extremely complex and invites many partial descriptions, or

there are many opposing viewpoints about the nature of the object.

In the case of political development theory, the existence of many view-

points has created a lengthy catalog of definitions.37

A definition is a brief description, adequate to the degree that

the thing described becomes generally recognizable. It is possible to

define political development in many ways. Certain aspects may be

chosen while others are ignored, depending upon our perceptual vantage

point. To advance a generally useful definition of political develop-

ment, the primary concern must be not only with the definition itself,

but also with making explicit the viewpoint from which the definition

emerges. In this sense the definition is a road map; if others are


37
Pye, Aspects of Political Development, 33-42.










to follow the definition and reach a specified goal, they must start

from the same point as the writer. Therefore, the following short

excursion into epistemology will attempt to clarify this author's

basic viewpoints.

The closer we are to an object or event, either in space or time,

the more important the object or event appears. When considering a

social fact, it is important to place it in proper temporal context

to avoid overly inflating its significance. As we include greater

reaches of the past in our research we are able to eliminate the evanescent

from consideration. Perhaps Davis gives the best illustration of the

significance of time.3 When we plot the mean values of socio-economic

attributes for many points in time we usually observe an extremely

ragged curve. It is possible to reduce the raggedness of the curve by

examining the trend line; the general tendency of the curve to increase

or decrease through time. The longer the time period examined, the

more reliable the trend line becomes. Conversely, the more foreshortened

the time frame, the more likely we are to read too great a significance

into transient events and the more inaccurate our generalizations will be.

The number of individuals included in the data base is at least

as important as the time frame used. The adequacy of our generalizations

can be as adversely affected by concentration on elite behavior as by



38
Harold T. Davis, The Analysis of Economic Time Series (Bloomington,
Indiana: The Principia Press, 1941), Chapter I.










concentration on too short a time span.

Political development theory must have the total society as its

basic referent; it must refer to the effects of political and social

change on the mass of men in society. Political change has occurred in

many different ways in many different societies. Many of these patterns

of change have benefited some members of the societies where they

occurred, but until political change benefits a significant proportion of

all the members of a society we are not justified in talking about

political development. Our ability to compare levels of development

will be severely hampered if the same measuring stick is not used to

make the comparison.

In this author's view, social theories are useful only to the

degree to which they aid us in explaining why some social act renders an

individual or group more or less content with its existence. As the

work of the psychologically oriented theorists demonstrates, whatever

the first cause, the ferment of social change lies within individuals.

Society is the primary environment of the individual and the texture

of his life is predominantly the texture of his social surroundings.

Yet his social environment is the totality of the characteristics of

socially relevant others. It seems reasonable that social science should

view the world primarily in terms of the benefit or harm individuals

may expect in a social situation. In this view, the individual is one

element in a social ratio. He.is part of the social whole, yet in

terms of enumerable attributes he is always a subset. Concentration

on either the individual or the society ignores the relation between




17


the two .-- the ratio which is the basis of social change.

In order to proceed from a useful vantage point it will be

necessary to specify the characteristics of our basic perceptual matrix.

First, it must force us to focus on the process of social change, and

not merely on static situations. Hence, it should allow us to follow

the transition of a society through time, and should be generalizable

to other societies. (Unfortunately, most of the comparative, empirical

work done to date focuses on different societies in different stages

of change.)39 Secondly, it must force us to focus our attention on

the relation between the individual and society rather than on one or

the other. It should allow us to discuss social change in terms of

individual satisfactions, forwhatever the first cause, social change

results from the frustrations or satisfactions of individuals. Finally,

the basic elements of the matrix should be abstracted, not culture

bound; a problem which most of the systems theorists have encountered.

A usable vantage point, partially specified and clarified, has

been advanced, first by Weber40 and later others,41 as the concept of



39
See, for example, Cnudde and McCrone, op. cit.; Cutright, op. cit.;
Arthur K. Smith, Jr., "Socioeconomic Development and Political
Democracy: A Causal Analysis," Midwest Journal of Political Science
(February, 1969), 95-125.
40
Max Weber, "Class, Status and Party," in R. Bendix and S.M. Lipset (eds),
Class, Status and Power (Glencoe, Ill.: The Free Press of Glencoe, 1953),
63-75.
41
Melvin M. Tumin, Social Stratification: The Forms and Functions of
Inequality (Englewood Cliffs, N.J.: Prentice-Hall, 1967); Gerhard
E. Lenski, Power and Privilege: A Theory of Social Stratification (New
York: McGraw-Hill Book Company, 1966); and ioan Davies, Social Mobility
and Political Change (New York: Praeger Publishers, 1970).





18


life chances. Weber describes life chances as "the typical chances for

a supply of goods, external living conditions and personal life exper-

iences."42 Ieber used the concept of shared life chances as a basis for

defining social class. Later theorists have used the distribution of

available life chances for defining the social stratification system.

When we describe the range of life chances accessed by an individual,

we are describing the texture of his life. We are assessing the general

possibility of his happiness and the range of gratifying activities

open to him. In determining the degree of social stability, two variables

must be considered: the distribution of available life chances and the

normative justification for the distribution.

An existing distribution of life chances is supported by values
43
and norms which tend to legitimize the distribution.43 Hence, the

life chances available to any individual are, in part, socially determined,

for no society conforms to complete achievement orientation. Essentially,

the concept of life chances rests on the relation between total available

possibilities for individual satisfaction and the possibilities actually

available to each individual. Since social systems are hierarchical, the

norm structure must legitimize inequality. Norm structures supportive of

a particular stratification system must provide the justification for



42
Weber, op. cit.
43
Tumin, op. cit., 14.










restricting life chances.

Stable social systems result from mutual reinforcement between the

actual distribution of life chances and the normative support for that

distribution. Norm and value support are the key determinants within a

wide range of actual distributions.44 While the society's norms and

values convey a sense of "oughtness" for the stratification system, the

distribution of life chances remains authoritative.45' 46, 47, 48 If

available opportunities are sufficient, and the distribution accepted,

the average individual will remain passive and the social system will

remain generally stable and slow to change. In truth, no social system

is stable in the sense of stasis; some change is always occurring, even

if that change is only biological replacement. There is no general reason

to expect a society to be stable as a besic tendency. The purpose of

developmental theory is to explain why some societies change and others

do not, and why some societies follow one path of change and others a

different path.



44
Ibid., 44.
45
David Easton, The Political System: An Inquiry into the State of
Political Science (New York: Alfred A. Knopf, 1953), 129-134.
46
Lenski, op. cit., 3 (footnote).
47
Harold D. Lasswell, Politics: Iho Gets What, When and How (New York:
McGraw-Hill Book Company, 1936).
48
Easton's definition of politics as "the authoritative allocation of
values" (op. cit.) is really a description of the stratification system.
Lenski (op. cit.) also argues that Lasswell ibidd.) has come to recognize
that his definition of politics as deciding "who gets what, when and how"
is also a description of the stratification system.




20


Morton's discussion of anomie49 as the incongruent relation

between cultural goals and institutional means for goal achievement

illustrates the necessity of balance.50 Lerner discusses the same

interaction under the intriguing title of the "Want/Get Ratio."51 The

result of an imbalance between the elements of Lerner's ratio is

individual frustration. Both authors argue that each individual's

reference groups and socialized values establish the basic goal expec-

tations appropriate to that individual's original position in the stratif-

ication system. The sanctions available to the roles occupied by the

individual provide the possibility for the satisfaction of the goal

expectations. Alterations in the elements of the ratio will lead to

individual frustration; widespread alterations will lead to social

stress.

Social stress can be defined as the result of interrupting the

mutual reinforcement between distribution and norm structure. Group

response to social stress is the simultaneous response of individuals

following existing patterns of orientation for or against change.



49
Robert K. Merton, Social Theory and Social Structure (Glencoe, Ill.:
The Frce Press of Glencoe, 1949), 125-149.
50
Balance here refers not only to the elements of Merton's paradigm, but
to the preceding discussion of the relationship between available
life chances and the normative justification for the current distribution.
51
Daniel Lerner, "Toward a Communications Theory of Modernization," in
Lucian W. Pye (ed), Communications and Political Development (Princeton:
Princeton University Press, 1965), 333-335.




21


Generally, the response to increased social stress tends to be

increased opportunity, due to a lack of any guidelines which could

dictate a new norm justification for the alternate distribution.

Essentially, the response to social stress is increased social mobility,

and social mobility is the redistribution of life chances. As the old

norm structure breaks down there is less support for traditional disabil-

ities and so the tendency is toward greater achievement orientation.

There is no expected shift in values toward a more favorable view of

mobility, rather there is simply no effective traditional inhibition on

the achievement motivation in individuals.

System accommodation to stress may be supported by elements of

the traditional norm structure. There are, of course, many forms of

mobility. Some forms of social mobility may be accepted, accommodative

modes. Either individual or group mobility may be valued.52 If mobility

is accepted, then stress may be relieved by a partial redistribution of

life chances under the aegis of tradition. Aspirant individuals may be

relieved of traditional disabilities and accommodated within traditionally

supported roles with concomitant sanctions.53 An alternative form of

mobility is the increase of sanctions for previously non-valued roles.54



52
Neil J. Smelser and Seymour M. Lipset, "Social Structure, Mobility and
Development," in Smelser and Lipset (eds), Social Structure and Mobility
in Economic Development (Chicago: Aldine Publishing Company, 1966), 1-50.
53
Ibid., 7.
54
Ibid., 7.





22


ilcre, the individual remains in the same .roles, but increases life

chances by virtue of increased sanctions. Or, lastly, neither the roles

nor the sanctions may be traditional and the individuals occupying

the new roles will attempt to insure high status for them.

The latter case provides the most interesting form of mobility

for political theorists. If neither the new role structure, nor the new

distribution of sanctions has general normative support, then purposive

system action is necessary to accommodate social stress. Purposive

action is required because neither the roles nor the sanctions have

normative support, and,therefore, there is no traditional justification

for the accommodative distribution. Whether or not this action occurs

will depend on the rate of change in the social stratification system.

Slow change may allow adaptation by the system even without traditional

support. The more rapid the rate of change, the less likely social

accommodation will become. Social acceptance of such a new distribution

of life chances may result from a generalized belief in its inherent

rectitude, or from the futility Of opposing the forces supporting it.

Social change and/or progress is rationally ordered only in

part. The bulk results from the interaction of many individuals

engaged in activities oriented toward personal rather than community

goals. This does not mean that change is not explainable or predictable --

it is. But, it must be kept clearly in mind that the paths discussed

are merely some of several alternatives, and this explanation attempts

only to make clear why one path eventuated rather than another. Social

change is deterministic only within the limits of the antecedent conditions

the theory will specify.










Although social systems are not purposive entities, obviously

a social group may exercise sufficient influence to affect the

behavior patterns of most individuals in a society. A group may force

general acceptance of a new distribution, thereby creating a semblance

of purposeful activity by a non-purposive entity.

In summary, social mobility may be sanctioned by traditional

norms and values or may occur so slowly that it does not cause significant

social stress; in either case it is not of particular interest to a

political scientist. Our interest focuses on social mobility which is

not sanctioned by traditional norms and values, and which occurs

rapidly enough to out-pace evolutionary accommodation.

A political act will be defined as -the attempted or actual use of

the authority of the state to redistribute the life chance matrix of a

society in an innovative manner. This definition has general significance

since it focuses on the redistribution of life chances, and therefore

will lead us to consider the effect of politics on all members of the

society. The more rapidly a system is changing, the more necessary

political activity becomes, since the re-institutionalization of norma-

tive support for a distribution becomes less likely as change occurs

more rapidly. There is absolutely no provision in this definition for an

expectation of either stability, equilibrium or success. It is only

suggested that political activity becomes necessary in an attempt to

accommodate to certain forms of social change. 'her. is no theoretical

basis for expecting this attempt to be successful in terms of contributing

to the reduction of stress.





24


In this definition, the state is understood to have geographically

defined jurisdiction. Any set of individuals who claim binding

jurisdiction over decision making for all individuals within a geograph-

ically delimited area, by virtue of their presence in that area, will be

considered a state. Only de facto jurisdiction is relevant; either the

claimants exercise effective decision making capacity or they do not

constitute a state.

Political activity will be sub-classified as either assumptive

or expansive. Assumptive political activity is an attempt to assume

distributive authority over a new area of life chances. For example,

the original struggle over whether the state would assume responsibility

for education constitutes assumptive political activity. An entirely

new range of life chances would be affected by the state's claim over

providing educational opportunity. Expansive political activity, on the

other hand, is the extension of previously accepted functions of

redistribution to larger subsets of the society. Continuing the example

of education, expansive political activity would be an attempt to

include more individuals within the state's education system. In most

societies, routine political activity will center on expansive politics.

It is, by its very nature, amenable to the Wildavskian treatment of

incrementalism.55



55
Otto A. Davis, M.A.H. Dempster and Aaron Wildavsky, "A Theory of the
Budgetary Process," American Political Science Review (September, 1966),
529-547.










Increasing the scope of either assumptive or expwasive politics

will increase the proportion of available life chances which the state

is attempting to redistribute. An increase in the scope of life chances

redistributed by political system activity will be termed political

permeance growth.5 Permeance growth is an obviously significant

element of political development. A political system is of consequence

to the degree that it affects the life chances of the average individual

in the society. One element of system consequence is the range of life

chances over which the state has assumed authority.

From the definition of politics it should be very clear that the

vast majority of activities undertaken by the state are not political.

Most state activities simply support the outcomes of previous political

acts. The level of earlier permeance growth may be measured by the

current level of government activity, for the range of activities

undertaken by the state as a result of previous political acts tends to

accrete like a palimpsest, layer upon layer.

There is no inherent expectation in this definition of politics

that political activity will tend toward greater equality of opportunity.

If the redistribution of life chances does result in a general tendency

toward greater equality of opportunity, then this tendency will be

termed congruent growth. This tendency toward equality is understood as



56
For an excellent essay on the concept of growth, see Karl De Schweinitz,
Jr., "Growth, Development and Political Modernization," World Politics
(July, 1970), 518-540.










growth simply because equality of opportunity is soon as socially

efficient. It is assumed that the most socially efficient utilization

of human ability is to be found in the achievement oriented society.

Congruent growth, therefore, employs a utopian referent; absolute

equality of opportunity. Thus, as traditional disabilities are removed,

the society more closely approximates the intelligent utilization of

individual talent.

Simultaneous growth on both the permeance and congruence dimensions

will be termed political development. Political development is the

activity of the state as it insures equality of opportunity over ever

larger ranges of life chances. This definition of political develop-

ment follows from the earlier definition of political activity.

Political activity is a response to patterns of social change which have

bred social stress. If the patterns of behavior undertaken in response

to that change do not relieve the original stress, this behavior cannot

be considered as conducive to further development. A good example of

inappropriate behavior patterns is the symbol manipulation of the

elites of many "developing" nations. Airlines, highways and modern

armies, all symbols of modern industrial societies, are resorted to in

response to the stresses found in the new nation.57 These manipulations

do not deal with the root causes of stress; rather like the individual



57
Herbert Feith, "Indonesia's Political Symbols and Their Wielders,"
World Politics (October, 1963), 79-97.










neurotic, they represent a retreat into compulsive symbolic activity

to relieve anxiety.

If political activity results from stress, then the more widespread

the stress, the more widespread will be the attempts to cope with it.

Hence, political permeance growth flows directly from social change.

On the other hand, congruence growth does not follow directly from

social change. Congruence growth is deemed necessary to development because

only by increasing equality of opportunity can there be effective relief

of stress. Since stress results from disruption of the ratio of

distribution to norm support, permeance growth alone is not sufficient

to reduce stress. Since re-institutionalization of normative supports

for a limited distribution is also unlikely during rapid change, the only

consistently effective stress reducing alternative is toward equality

of opportunity. In a modern industrial society, or in any rapidly

changing society, the individual will have no socialized limitiation on

goal expectations. Any socially supported disability of means will

perpetuate stress. Thus, permeance growth which does not tend toward

equality of opportunity will not only not relieve stress, it will

perpetuate stress.

It is by now obvious that political development cannot occur

without social development. By social development we mean an expansion

of the total range of available life chances in a society, both roles and

sanctions. Social development may, for a time, occur without political

development, but not vice versa.





28




We have now developed a linkage. Social change is social mobility,

widespread social mobility is social mobilization, and political

development is inextricably linked to social change. Our concern at this

point is to explain the process by which social change is generalized

and accelerated, and to explain why and how different rates of social

change lead to different forms of political change.












CHAPTER II

SOCIAL CHANGE AND THE INSATIABLE PERSONALITY



Deutschl and others2 inake the implicit assumption that social

mobilization has some positive end product. Deutsch defines social

mobilization as "The process in which major clusters of old social,

economic and psychological commitments are eroded or broken and people

become available for new patterns of socialization and behavior."3

Yet, the socialization of new values need have no particular signifi-

cance. In the Deutsch argument, these new values are supposedly

more "modern" and, therefore, more conducive to continued social

change.

However, the process of mobilization is significant, not

because it replaces traditional norms and values with new norms and

values, but rather because the essence of the process consists of

destroying certain specific, yet apparently universal elements of the




Karl W. Deutsch, "Social Mobilization and Political Development,"
.American Political Science Review (September, 1961), 493-514.
2
Bert F. Hoselitz, "Economic Growth and Development: Non-Economic
Factors in Economic Development," American Economic Review (May, 1957),
28- 41; and Manning Nash, Primitive and Peasant Economic Systems
(San Francisco: Chandler Publishing Company, 1966), 131-135.
3
Deutsch, op. cit. 494.


29




30


traditional norm structure and never replacing them.4

In discussing past periods of great economic expansion and

cultural prominence, IIagenS makes the simplistic, but significant obser-

vation that they have all come to an end. In explaining the current

development of societies, however, he is faced with a long standing

period of progress which has not ended, and perhaps may never end.

Hagen searches for a positive form of personality type, a "modern" man

who is in some way responsible for this sustained growth.6

An alternative explanation is that some particular set of char-

acteristics exists in traditional society placing limitations on social

change. Here, the elimination of certain strictures, rather than the

creation of a particular personality type, produces the original climate

for sustained growth. Although a new form of personality type is

produced, this personality type does not have any predisposition toward

positive new modes of behavior. Rather, the behaviors engaged in by this

individual will be largely dependent on his current social situation.

The individual's attempt to function in an alien social matrix may, in

fact, lead him to attempt to define his identity in socially dysfunctional



4
For a brief discussion of the social disorganization wrought by the in-
dustrialization of a traditional society, see George A. Theodorson,
"Acceptance of Industrialization and Its Attendant Consequences for
the Social Pattenms of Non-Western Societies," American Sociological
Review (October, 1953), 477-484.
5
E.E. Hagen, On the Theory of Social Change (Homewood, Ill.: The Dorsey
Press, 1962).
6
Hagen describes the developmentally conducive individual as the
"innovative personality."










modes of behavior.

Aside from occasional bursts of development, most societies

are marked by very low rates of social change. Through most of history,

this sub-culture of the peasantry has been the dominant culture of

society. Rogers7 has developed a list of attributes which apply to the

sub-culture of peasantry.8 In traditional peasant societies the

individual functions in a world frame of "perceived limited good."

In Rogers' argument, the individual assumes that there is an absolute

limit on the quantity of valued good. This perception is quite realistic

for the average peasant. Most traditional societies based on peasant

agriculture deal with fundamentally inelastic supplies of valued

goods. Both the quantity and variety of life chances and life styles

are highly restricted and the distribution of available life chances

is legitimized by tradition and religious sanctions. Although Rogers

sees "perceived limited good" as merely one element of a peasant sub-

culture,j0 it is this author's argument that given the other elements



7
Everett M. Rogers, Modernization Among Peasants: The Impact of
Communication (New York: Holt, Rinehart and Winston, Inc., 1969),24-39.
8
According to Rogers, "Central elements in this subculture of peasantry
are: (1) mutual distrust in interpersonal relations; (2) perceived
limited good; (3) dependence on and hostility toward government author-
ity; (4) family m; (5) lack of innovativeness; (6) fatalism; (7) limited
aspirations; (8) lack of deferred gratification; (9) limited view of
the world; (10) low empathy."
9
See for instance, Gerhard Lenski, Power and Frivilege: A 'Theory of Social
Stratification (New York: McGraw-Hill Book C0o.anlny, 1966); 'nd Xadcl De
Schweinitz, "Growth, Development and Political Modernizarion," Wlorld
Politics (July, 1970), 513-S40.
10
Rogers, op. cit., 25.











of the proposed sub-culture, "perceived limited good" is the central

factor defining peasant culture.

It should not be surprising that social mobility is very low in

a society where it is unrealistic to expect advancement and impious to

want it. Lenski has argued that "the fact is, unhappily, that in the

long run, in all of these societies, downward mobility was much more

frequent than upward."11 The other elements of the peasant sub-culture,

fatalism, amoral familism, distrust of others, all are closely related

to these two factors: the limited range of available good and the

rprpetually sanctioned inequality of distribution. In a peasant society

an individual's personal identity is largely defined by his position at

birth. The development of his self-concept consists of developing

subjectivee public identities" which are products of the roles he

occupies. With the stratification system rigidly restricting role assign-

ment, the individual's identity definition is also restricted. If

nothing else, the traditionally defined identity of the peasant leaves

him secure in his disadvantage.' The traditional individual knows who

he is and what is expected of him merely on the basis of his existence.

This form of identity definition will be called traditional identity.

Under the norm structures of perceived limited good, the vast majority

of any population experiences traditional identity.

If perceived limited good is the central element of a peasant



11
Gerhard E. Lenski, Power and Privilege: A Theory of Social Stratification
(New York: McGraw-Hill Book Company, 1966).





33


culture, we would expect to find a fatalistic, inward turned individual

with little capacity to visualize alternative life styles simply because

they are unrealistic. We would expect this individual to be primarily a

familist, if not an amoral familist, for in the struggle for survival,

few others are to be trusted. Finally, we would expect this individual

to be highly traditionalistic, not only because tradition is the

basis for his existence, but because sameness tends to reinforce itself.

It is reasonable, therefore, to accept that the peasant has, in fact,

managed to adapt realistically to his environment. Despite his

"ignorance," the culture of the traditional peasant is a realistic

reflection of the central fact of physical existence; good is, in

reality, limited, and the peasant culture, whatever its ethnic variations,

has the same central elements as a reflection of the restricted range

of life chances.

If a restricted range of life chances results in traditional

stasis, then change and growth could be based on increasing the range of

life chances. The norm structure of perceived limited good may be

disturbed in one of two ways: an alteration in the range of available

life chances, necessitating redistribution of the social matrix; or

some disturbance which destroys the old norm structure. Before the

industrial revolution, such redistributions were most frequently caused

by war and trade expansion, both of which acted against each element of

the social ratiu -- distribution and justification. After such a period

of expansion, there would necessarily be a re-traditionalization of the

norm structure of perceived limit d good simply because gcod, even between




34


societies, was limited. Most brilliant periods of human culture

closely accompany periods of conquest or trade expansion. General

social development, on the other hand, tended to be glacial until the

onslaught of the industrial revolution. Under the impact of industry

there has been an almost continuous expansion in the range of life

chances with a corresponding shift from traditional to situational

identity.

Modernization is the reinforcing interaction of social mobilization

and economic growth, destroying the norm structure of perceived limited

good for progressively more and more individuals. This process also

destroys the individual's social identity. For the peasant, social

identity is closely linked to the stratification system which is, in

turn, a product of perceived limited good. The difference between

situational and traditional identity lies in the rate of social change.

While the social situation is always changing, for the peasant the

change is so slow as to be almost imperceptible. For an individual in

a rapidly changing society, on the other hand, the elements of his

identity definition are changing at a very rapid rate; his identity is

no longer secure because his situation is no longer secure. The

socially mobilized individual has neither a secure identity, nor the

limitation on goal expectations which flow from the peasant's percep-

tion of limited good. It is this social situation which gives rise to the

"modern" man; an individual I call the Insatiable Personality. The

Insatiable Personality's primary characteristic is non-directed striving,

for only by striving can he develop a sense of selfless. He is, at once,





35


Hagen's innovative personality,12 McClelland's high n-achiever,13 Lerner's

high empathy type,14 and Pye's confused identity.15,16 He has high

empathy because the mobilization process has destroyed the blinders

of social strictures. He is innovative and achievement oriented

because he has no limitation on expectations. He is a confused identi-

ty because his social referents are in constant flux -- the more he

succeeds the more he has to strive. With no limitation on good

expectations, his reference groups and standards constantly rise as

each new goal is reached.



12
Hagen, op. cit.
13
David McClelland, The Achieving Society (Princeton: D. Van Nostrand
Company, Inc., 1961).
14
Daniel Lerner, The Passing of Traditional Society: Modernizing the
Middle East (New York: The Free Press, 1958).
15
Lucian W. Pye, Politics, Personality and Nation-Building:Burma's
Search for Identity (New Haven: Yale University Press, 1962).
16
Hagen uses status-deprived groups such as the Antioquenos of Latin
America and the Samurai of Japan as examples of highly innovative
groups. Yet, all that is unique about these groups is that they
were mobilized at an earlier date than other members of their
society. By virtue of their mobilization, they had no traditional
role to which they were still committed, and they were, therefore,
generally willing to take advantage of new situations as they arose.
The important point here is simply that these individuals did not
create the situations from which they profited. Their innovativeness
is passive not active. Lerner's concept of empathy also is generally
passive not active. This individual is able to consider new modes
of behavior, but he does not necessarily seek them out. McClelland's
high n-achiever, on the other hand, is aggressive in his striving
and he comes closer to the economically modernizing individual we
seek. Yet, McClelland's high n-achiever is willing to strive in
whatever social context he finds himself; he is success oriented,
but not necessarily innovative.










The Insatiable Personality exists in a social matrix which

places no normative or value limits on what he may expect from life.

His identity is insecure because it is defined by comparison; he sees

himself as he believes others see him. He strives constantly in an

attempt to obtain a more gratifying self-perception.17 Where

one peasant may be envious but not threatened by another peasant's

bigger harvest, one machinist will be both envious and threatened if

another machinist receives a raise. In both cases, identities are

defined relative to other people, but in the peasant case the relations

are fixed, while for the modern man they are variable and insecure.



17
There is an interesting parallel here between Lenski's discussion
of institutional evolution of stratification systems, the cultural
perception of limited good (op. cit.), and Veblen's discussion of
conspicuous consumption as a cultural artifact (Thorstein Veblen,
The Theory of the Leisure Class (New York: Viking Press, 1912),
60-70). At both ends of Lenski's spectrum, hunting and gathering
societies and mature industrial societies, we find relatively loose
stratification systems. Veblen's discussion of conspicuous consumption
suggests that in primitive societies and industrial societies,
conspicuous consumption is a right,whereas in traditional societies
(Lenski's agrarian societies) it is a privilege. In both forms
of society where conspicuous consumption is a cultural artifact of
the mass, we find loose stratification systems and no perception of
limited good. Given the absence of limited good and the flexibility
of the stratification systems, conspicuous consumption is an obvious
and probably necessary means of identity definition. Veblen argues
that conspicuous consumption is a means of maintaining reputation,
hence a primary component of individual identity. Reismaa's
discussion of the "other-directed" individual in modern society
(David Reisman, The Lonely Crowd: A Study of the Changing American
Character (New Haven: Yale University Press, 1964))parallels the
discussion of conspicuous consumption as a means of reputation building.
The other-directed individual is oriented to identity definition
by continuous comparison of outward cues.










In terms of individual economic behavior, modernization is, in

large degree, a potluck process. Not all Antiquenos, all Samurai, or

all Puritans were moderizers.18 However, if enough individuals attain

the characteristics of the Insatiable Personality, then it is probable

that some of them will engage in modernizing behavior. Including

a large number of Insatiable Personalities in a society is a necessary,

but not sufficient, condition for modernization; neither does the

mobilization of a group guarantee their future high economic activity.

As more and more individuals are socially mobilized, the

process becomes largely self-sustaining. Once this point is reached,

the key factor is no longer the destruction of the norm structure

of perceived limited good, but its general absence. An individual

socialized in modern industrial society automatically defines his

identity by comparison and his self-esteem by striving. In terms of

sustained social change, there is a continuous interaction between the

individual and society so that the individual may, at the same time,

be both cause and effect. By his very existence, the Insatiable Person-

ality creates an unstable social situation for other individuals since

he is, for them, part of society.

The significance of the Insatiable Personality as a conceptual

tool in explaining social and political change lies in its ability to

help us understand the generalized striving of mobilized individuals.


18
Hagen, op. cit.





38


Nash19 relates an interesting anecdote about New Guinea tribesmen

presented with metal knives. The resulting increase in efficiency was

expended, not in increased productivity, but in increased leisure. It

is hard to imagine an Insatiable Personality not putting a comparable

advantage to use to increase his productivity and thereby increase

his status. Productivity here is not confined to simple economic

increase, but to general efficiency in any line of endeavor. The

New Guinea natives were secure in their current situation, and even

presented with an aspect of modernity took a long time to use it.

The limitations of the psychologically oriented theories result

partially from their attempt to identify a first cause for modernization.

The first cause of modernization in any society is really unimportant.

From the point of view of the mass of men, the process of social change

is always one of diffusion. In examining a modern society there is

little consequence in discovering what came first, the chicken or the

egg -- our concern is with understanding how we got a whole flock.

While the process of social mobilization results in Insatiable

Personalities, social mobilization cannot long continue without continued

economic advancement. As previously stated, through most of history

the only way to increase wealth was by horizontal economic expansion.

In terms of any given area of population there was no significant

increase in the total quantity of valued goods. Horizontal expansion


19
Nash, oa. cit., 42-57.









led to a redistribution of wealth which could produce an increase in

the total available life chances for members of one social system.

However, since the total supply of goods was inelastic, there was

necessary deprivation of members of other systems. This inelasticity in

agrarian economic orders was the primary basis for the re-traditionaliza-

tion of the norm structure of perceived limited good, since all members

of competing societies were faced with what amounted to a zero-sum

game. (It should be noted that the situation in the United States

did not result in a zero-sum game due to the availability of unoccupied

land.)

Increasing wealth after the industrial revolution created a

non-zero-sum game. Economic expansion became a process of largely

vertical growth, increasing not only the quantity of life chances in

terms of real goods (growth), but the quantity of life chances in

terms of available roles (differentiation).20 The special increase in

wealth provided the necessary reinforcement for a continuous process

of social mobilization. Since a continuous increase in the range of

available life chances existed, the Insatiable Personality was not

faced with the reassertion of the social strictures resulting from the

former inelasticity of valued goods. There was the possibility for the

reinforcing interaction through time between the Insatiable Personality



20
Wilbert E. Moore, "Changes in Occupational Structures," in Neil J.
Smelser and Seymour M. Lipset (eds), Social Structure and Mobility in
Economic Development (Chicago: Aldine Publishing Company, 1966), 194-
248.




110


and increasing life chances which we have called modernization..

Social modernization is not a necessary result of either the

increase in the number of Insatiable Personalities nor the inception of

economic growth. Although aspiration is enough to ensure striving for

a limited period of time, ignorance can provide as effective a deterrent

to continued striving as did the formerly inelastic supply of valued

goods. We have argued that the Insatiable Personality is willing to

engage in new modes of behavior and to strive to achieve in order to

secure his identity. Enabling him to pursue achievement is a primary

linkage between politics and society.

We have defined politics as the attempted or actual use of the

authority of the state to redistribute the life chance matrix of a

society in an innovative manner. Social pressure for an increased

ability to satisfy current aspirations will lead to pressure for

greater educational opportunities, and in terms of increased

permeance, government activity will proceed in a very special direction

-- toward increased education. -There is no expectation that either

the pressure on the government, or its activity will be successful.

Success (in the sense of increased education) will depend upon the rate

of change in the society and the rate ef demand with which the political

situation can cope.21 If the pressure for greater education results

in higher levels of educational attainment, then the probability of



21.
Later discussion will outline the effect on political development
of very rapid rates of mobilization.





41


continued modernization is very high. If the pressure for greater

education does not result in higher educational attainment, then the

probability of continued modernization is almost non-existent.

The basic elements for an explanation of increasing political

development have new been outlined. Modernization is an interaction

between increasing wealth and increasing social mobilization; each

factor is dependent on the other and neither can long continue without

the other. Yet, the aspirations raised by social mobilization can only

be satisfied, through time, by increased individual capability arising

from increased educational opportunity. If the Insatiable Personality

is not to experience system de-stabilizing frustration which will

ultimately end the process of modernization, widespread education must

become normal. Since the state is the only agency which can supply

educational opportunities on a mass basis, extreme permeance growth

begins. The state thus becomes intimately involved in the redistribution

of life chances for the society, and by being involved, becomes an

integral part of the process of'modernization.

Political development is one possible result of modernization if

the rates of change allow balanced growth. Imbalances in economic

growth or social mobilization can result in hybrid forms of political

organization which may or may not become developed in time. Our concern

in the following pages will be to present a specific model of the

ideal-typical form of development.

To this point, our theoretical concern has centered upon the

general case of transition from peasant to industrial society. Peasant





42


societies were used as the base line to insure applicability to many

societies. In the specific case of the United States, we are examining

the transition from a society based upon agrarian enterprise to one

based on industrial production. Prior to the onslaught of massive

industrialization, the process of frontier expansion, indeed the very

process of colonization, had increased the range of life chances avail-

able to the average individual. American society was already experiencing

the process of social development resulting from increased wealth produc-

tion. The American was a farmer, not a peasant.

Although increased life chances were available to the average

American as a result of the abundance and quality of available land,

as in a peasant society, the means of wealth production were still

based on agriculture. Despite the increased availability of wealth,

the average American was still faced with a restricted set of altern-

ative life styles. So long as the society was based on agricultural

production, it was limited in its potential for social, and hence politi-

cal, development. Since political and social development are processes,

not events, there is no compelling reason to begin at the beginning

in order to conduct valid tests of models of these processes.

Accepting the preliminary changes in the social structure which

established America as a free agrarian, rather than peasant society, it

is still expected that the transition from a rural-agrarian to an

urban-industrial social order would be deeply traumatic. Although the

theoretical argument for the experiences which result in the Insatiable

Personality remain the same, it must be acknowledged that the average










American's situation in 1870 was not as restricted as the situation of

the average peasant.

To cover as long a time period as possible, it would be best to

commence the research at as early a date as possible. At the same time,

extension of the time period should not result in any loss of quality

in the variables used, nor too severe a loss in the number of available

data points. The specific time period of 1870 to 1960, was chosen for

the research as a result of several considerations.

Prior to 1850, data on newspaper consumption and literacy are not

available on a state by state basis. Therefore, two of the major variables

are unavailable prior to 1850. Comunencing the research with 1850

would have included the disturbing effects of the Civil War. Further,

since many of the states were not, in fact, states during the period

of secession, the political participation index would be truncated.

Consequently, it was decided that the earliest practical year to

include in the research was 1870.

Figure 1 presents the gross total pig-iron shipments in the United

States between 1820 and 1960, by census years. Pig-iron is used as

an index of basic industrial capacity since it provides the raw

material for all advanced industrial processing of steel and steel

related products. In addition, the basic metal used in the advancement

of the rail network in the United States was forged iron. From Figure 1

it can be seen that through 1870, little change had occurred in the level

of industrial activity. By beginning with 1870, we will be examining the

process of transition from .agrarian to industrial society from its

earliest stages.

















Legend for Figure 1



Production noted in thousands of short tons.

Raw data from the United States Bureau of the
Census, Historical Statistics of the United States,
Colonial Times to 1857 (Washington, D.C.: United
States Government Printing Office, 1960), and
Historical Statistics of the United States,
Continuation to 1962 and Revisions (thashington,
D.C.: United States Government Printing Office,
1965).










73,500


70,000


66,500


63,000


59 500



56,000


52 500


49,000


45 500


42,000


38,500


35,000


31, 50 0


28,000


24, 500


21, 50 0


17, 500


14, 000


1 C 0 5 0 0


7, 0 0


3, 50 0


1820 30 40 50 60 70 80 90 1900 10 20 30 40 50


Figure 1


Total Pig Iron Shipments in the United States: 1820 1960





46


Our previous discussion suggests the following linkage for the

theoretical elements: increasing wealth leads to increasing social

mobilization, which in turn aids in increasing wealth. If increasing

social mobilization leads to increased individual adaptability, then

social mobilization and wealth will likely continue increasing.22

Diagrammatically, we may represent the simultaneous occurrence of these

relationships by interconnected circularities.


Wealth -> Social ---- General
Production Mobilization Adaptability

Figure 2

Causal Model of Three Concepts with Simultaneous Interactions


Considering the above within the framework of passing time

makes the entire system clearer. For any single period, T, the

linkages are shown in Figure 3.


IWealth __-> Social _General
Production Mobilization Adaptability

Figure 3

Causal Model of Three Concepts with One-Way Causation



22 - -
For a similar, but confusing model, see Norman H. Nie, G. Bingham
Poweil, Jr., and Kenneth Prewitt, "Social Structure and Political
Participation: Developmental Relationships, I F, II," American
Political Science Review (June and September, 1969), 361-378 and 808-
318.




47


If we consider these same relationships for a period, T to T+1, we

would have the linkages shown in Figure 4.

Wealth Social General
Production Mobilization Adaptability
P T T

Wealth Social General
ProductionT+1 obilizationT+1 AdaptabilityT+1


Figure 4

Causal Model of Three Concepts with Inter-Temporal Effects



Through time, each variable's continued growth becomes a necessary

precondition for the continued growth of the other variables in the system.

This argument assumes that in past periods of rapid social change, the

process was interrupted by the inelasticity of available wealth. The

ensuing re-institutionalization of the norm structure of perceived

limited good led to the demise of further social mobilization and the

reassertion of the traditional agrarian social order.23 In considering

current modernization, wealth increases have become the result of in-

dustrialization, and therefore have been relatively continuous. In

current modernization, all the variables in the system have become

more closely interdependent since wealth production is now a highly com-

plex social process. The social organization of industrial society require

widespread social mobilization in order to continue increased industry.


23
Lenski, op. cit., 275.











Social mobilization can be broken down into two major, inter-

related processes, urbanization and mass communication, each more

specific and more easily operationalized. In the following

discussion, urbanization will be considered only in terms of intense

demographic concentration.24 Industrialization as it occurred under

past levels of technology required extreme demographic concentrations.

Large scale industrial production requires masses of workers concen-

trated in close proximity to the place of work. Throughout the history

of industrialism, the growth of cities has been the primary result of

stripping the hinterland of agrarian population. Urbanization did not

result merely from natural population increase in pre-existing cities.

Indeed, as Deutsch25 has noted, in all industrializing societies, new

cities sprang up as new industries were formed.26 Again, there is a

cycle; industrialization makes large scale urban populations possible,

and increasing urban populations are necessary for further industrial



24
Rather than treat urbanization as a syndrome, it was considered
wiser to examine the most basic characteristic of urban life and
then test for associated relationships.
25
Karl W. Deutsch, Nationalism and Social Communication. An Inquiry
into the Foundations of Nationality (New York: Wiley, 1953).
26
One must, of course, also consider the effect of urbanization on
peasants who are intermittently exposed to city life. The extraordi-
nary urban slum sprawl of many cities of the third world is an apparent
reflection of the effect of urban opulence on the unprepared rural
population. S-3 also, Philip M. Hauser, "The Social, Economic and
Technological Problems of Rapid Urbanization," in B.F. Hoselitz and
I.E. Moore (eds), Industrialization and Society (UNESCO: Houton,
1i963), 199-217.




49


expansion. The by-product of this process is an ever increasing

quantity of Insatiable Personalities who, by their activities,

will tend to reinforce both industrialization and urbanization.

Urbanization necessitates the second process of social mobiliza-

tion, mass communication. Mass communication is essentially the process

of large scale community integration. If cities were merely organized

in small enclaves, industry by industry, then integration would be less

important. However, even in a relatively small city, the extreme

complexity of the logistical system makes widespread communication

necessary. To create an urban society, mass communication is an absolute

essential. In earlier epochs, communication required literacy since

the media were print. This has led some theorists to presume that

literacy is necessary for mass communication. There is, in reality,

no reason for mass literacy to precede mass communication. Prior to

the development of electronic media, as long as the nodal personality

in a sociometric network was literate, the other persons in the net

could effectively be considered-part of the mass communications system.

A society cannot support urbanization without the community

integration carried on by the processes of mass communication.

Demographic concentration can continue for some time, but the process

of modernization will break down under the pressure and confusion of a

non-integrated environment.

A cycle is again apparent. Industrialization leads to urbanization

which in turn is necessary for further industrialization, Yet, continued

urbanization is not possible without the integrating functions of mass

communication. Diagramatically, in terms of two time periods, the relation





50


are shown in Figure 5.


Industrialization --- Urbanization ------ Commnunication
T T T

Industrialization ------- Urbanization ------ Communication
T+l T+1 T+o

Figure 5

Inter-Temporal Effects of Three Operational Indicators



Mass communication has a significant impact on the development

of the Insatiable Personality. Where urbanization acts as the first

stage of social mobilization, mass communication acts as the second

stage by increasing the individual's aspirations, making him aware of

currently available alternative life styles. At this juncture, social

stress becomes reflected in the political sphere. Aspirant individuals

are faced with the unpleasant realization that their abilities are

not commensurate with their aspirations.

It has been argued that social mobilization destroys the

individual's links with his previous social position and identity. In

order to exist in his new situation, the individual must find some

new form of identity in which to secure himself. He must, in short,

take on new roles. Yet, his ability to fill new roles is contingent

upon his level of general adaptability. Education may be learning through

experience, or formal classroom schooling. If the individual is educa-

ted through experience, it will be a process of trial and error.










Repeated exposure to modern life styles may simply inure the individ-

ual to the harassments of modernity. If the rate of mobilization is

slow, then learning from experience will be sufficient to perpetuate

modernization for a time. The more rapid the original rate of mobil-

ization, the more necessary an efficient form of education will be.

Insecurity and maladjustment are common products of repeated failures

to cope in the face of an incomprehensible social order. As the

number of maladjusted individuals increases, the ability of the remain-

ing mobilized individuals to sustain the process of modernization

decreases. Hence, the process of social mobilization creates a situation

in which education becomes increasingly necessary. In terms of an

entire society, mass formal schooling becomes necessary if the process

of modernization is to continue. Diagrammatically, the relations

explicated so far are shown in Figure 6, first for one time period, T,

and then for a time span, T to T+l.


For Time T

Industrialization--- Urbani nation --s Communication -. Education

For Time T+l

Industrialization--> Urbanization--> Communication--> Education
T T T T


Industrialization ---> Urbani zation--- Communi cation-- Education
T+1 T+1 Ttl T+1

Figure 6

Inter-Temporal Effects of Four Socio-Economic Indicators





52


While increasing communication produces the need for increased

education at a given time, T, continued increase in the level

of communication at time, T+1, is contingent upon prior increases in

education

Formal schooling provides the only method of increasing individ-

ual adaptability in the mass of the mobilized populace. The necessity

for increasing the level of individual adaptability generates the first

extreme permeance growth of the state. Government activity is obviously

necessary to the process of social modernization, yet it is not

a necessary link in the chain of causality. Mass education is a social

service which has no immediate return. Government activity is thus a

necessary intervening variable between the pressures resulting from

mass mobilization and the increased need for education. If it were

possible to increase the level of mass education without government

intervention, then the process of modernization would continue without

the state. It is precisely the process of modernization which renders

non-political solutions to the problem impossible. As the rate of

social change increases, political activity becomes more and more

necessary. These relations can be diagrammatically represented by

Figure 7, for any time, T.


Industrialization Media Consumption

___- Education
Urbanization Government Activity

Figure 7

Causal Relations of Four Variables and Gcvernment Activity




53


It is possible to extrapolate this argument one step further.

Having acquired some measure of education, the average individual will

be more willing to engage in political activity.27 In a rapidly

changing society many issues will have become politicized. The

average individual's interest in political activity will vary with the

number of issues which he perceives as affecting him. Previous expo-

sure to the activities of mass media and government will arouse his

awareness of the importance of political participation. In general,

however, the tendency will be against attempting to participate until

a certain measure of self-confidence is attained. For the average

citizen, that self-confidence will be the result of the process of

education. A generalized disposition toward attempting to affect

one's fate will be the result of education. The actual opportunity for

participation will be determined by the rate of social change which the

extant political institutions can accommodate. Presumably, if the

current political institutions have been able to provide a satisfactory

level of education, then the probability is high that they will be

able to accommodate mass political participation. In other words,

it is expected that if the developmental process continues through the

level of creating widespread education, then increased political

participation will follow as a matter of course. The expected final

model will follow the path indicated in Figure 8.



27
For the general evidence already relating education to participation
see, Lcster IV. Milbrath, -Political Participat.on. IIow and lWhy Do
People Get Involved in Politics? (Chicago: Rand, McNally and Company,
1965), 48-89.




54




Industrialization Media Caiisumption Education



Urbanization Governii;'nt Activity Political Participation

Figure 8

Final Representation of a Six-Variable I!odel
of Political Development


The ensuing discussion will describe the methodology and

operational indicators used to examine the adequacy of the above model.












CHAPTER III

INDICES AND METHODOLOGY



In the preceding chapters, some objections were raised to

the methods used to test models of development. One of those objec-

tions referred to the cross-sectional nature of the tests. Develop-

ment is a dynamic process, and it cannot be tested by an examination

confined to one point in time. In fact, it is all too easy to come to

false conclusions about a concept simply from examining its behavior at

only one point in time. As an example, let us examine the relationship

between the concept of inter-party competition and education. Inter-

party competition is used as an example because it was not included

in the model of development. Table I presents a list of correlations

between an indicator of inter-party competition and the level of

education in the United States, by state, from 1870 to 1960.

Table I

Decennial Correlations of Inter-Party Competition and Educationa,b


Year 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960

IPC
with -.20 .32 .33 .44 .76 .06 .64 .71 .25 -.05
education

The indicator of inter-party competition used was the losing
party's percentage of the winning party's vote.
Raw data from Sven Petersen, A Statistical History of the
American Presidenti.al Elections (New York: Ungar Publishers,
1963), ard U.S. Bureau of the Census, U.S. Census of Population
(Washington, D.C.: U.S. Government Printing Office, 1872, 1883,
1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962).





56


It is obvious that observing this correlation at only one point in

time could lead to erroneous conclusions. Observing the relation in

only 1940 would lead us to conclude that a very high relationship

exists between education and inter-party competition. If, however, we

confined our observation to 1920, we would conclude the exact opposite.

In order for a generalization to be valid, it must hold for more than

one point in time.

To avoid falling prey to this same criticism, the current

research uses data drawn from a period from 1870 to 1960. Observations

were made at ten year intervals beginning in 1870. The ten year

interval was used primarily for the sake of convenience. Data for the

socio-demographic variables were readily available at the ten year

intervals of the census. It would have been possible to generate

the necessary data for shorter intervals by interpolation. However,

since interpolated data would have been based on the decennial census

figures, it was felt that little would be gained by increasing the

number of observation points.

A second objection raised against previous work was the use

of cross-societal samples as a data base for creating intra-societal

generalizations. This research used the United States, by state, as a

data base. Table II presents the number of data points used for each




United States Bureau of the Census, U.S. Census of Population
(Washington, D.C.: U.S. Government Printing Office, 1972, 1883, 1895,
1901, 1913, 1922, 1933, 1943, 1952, 1962).





57


census period.


Table II

Number of Observations for Each Decennial Census Period


Year 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960


N 37 38 44 45 48 48 48 48 48 48



A state became a data point if, at the time of the census, it was

legally a state. During the 1870's, several of the southern states

had still not been readmitted to the Union. Although the socio-

demographic data were available for these states, no election figures

were available, and,therefore, the last stages of the model could

not be tested. The same situation prevailed in regard to the various

territories which had not yet been admitted to the Union.



Data Sources



All socio--demographic data used in this analysis were taken

froma official government publications. The primary sources of infor-

mation were the various volumes of the Census of the United States.2

At times, certain series of statistics were more readily available in


2
Ibid.





58


the United States Statistical Abstract.3 While the figures in. the

abstract are generally based on the data reported in the Census, the

abstract figures are often presented in a more usable form, especially

in regard to state summary totals. In certain cases, notably early

school attendance figures, it was necessary to use special reports.

Data on school attendance were retrieved from the Report of the United

States Commissioner of Education.4 Figures for postal receipts, by

state, were obtained from the Report of the Postmaster General.5

Beginning in 1850, the Bureau of the Census was directed to collect

figures on social statistics, and there are data on literacy, news-

paper circulation, church attendance, etc. starting with 1850.

Oddly enough, the figures on Presidential and Congressional

elections were not available on a state by state basis from official

publications before 1920. Data for the Presidential elections were

drawn from Petersen, A Statistical History of the American Presidential

Elections, and were checked against the figures presented by W.D.



3
U.S. Bureau of the Census, Statistical Abstract of the United States
(Washington, D.C.: U.S. Government Printing Office, 1883, 1893. 1903,
1913, 1923, 1933, 1943, 1953, 1963).
4
U.S. Commissioner of Education, Annual Report of the Commissioner of
Education (Washington, D.C.: U.S. Government Printing Office, 1871,
1881).
5
U.S. Postmaster General, Annual Report of the Postmaster General
(Washington, D.C.: U.S. Government Printing Office, 1871, 1881).
6
Sven Petersen, A Statistical History of the American Presidential
Elections (New York: Ungar Publishers, 1963).











Burnham, Presidential Ballots: 1836-1892.7 Both authors were forced to

rely on newspaper accounts of the various elections in those states

where no special totals were kept by a state agency. Data for the

Congressional elections were taken from an archival record supplied by

the Inter-University Consortium for Political Research.8 The tape

presented only congressional district totals, but for the purposes

of this research the district totals were aggregated to the state level.



Variables



A series of fourteen socio-demographic variables was collected

for the entire period. This series included data on government

employment and government expenditutes which were available from census

reports. Table III presents a list of the variables originally

selected for examination as possible operational indicators of the six

major variables: industrialization, urbanization, education, media

consumption, government activity and political participation. First

it was necessary to select a series of variables, appropriate as

indicators, both in terms of variable behavior through time, and in

terns of the relationships among the several variables. Second, it was



7
Walter D. Burnham, Presidential Ballots: 1836-1892 (Baltimore: Johns
Hopkins University Press, 1955).
8
Inter-University Consortium for Political Research. Historical
Data Archives, Statistics of the American Congressional Elections,
1324-1968 (Computer tape).





60


necessary to choose either a single operational indicator for each

concept, or to create an index of the concept.


Table III

Original Variables Selected as Possible Indicators



1. Per capital value added from manufacturing

2. Percent of the total population resident in cities greater
than 2,500

3. Percent of the total population resident in cities greater
than 25,000

4. Average daily newspaper circulation per 1,000 population

5. Telephones per 1,000 population

6. Radios per 1,000 population

7. Postal receipts per capital

8. Total horsepower in use in manufacturing per capital

9. Percent of the working age population in manufacturing

10. Percent of the population who are literate

11. Percent of the school age population in elementary school

12. Percent of the school age population in high school

13. Percent of the age group 19-24 enrolled in college

14. Total state and local government expenditures per capital

15. State and local government employees as a percent of working
age population

16. Turnout in Presidential elections as a percent of population

17. Turnout in Congressional (off-year) elections as a percent
of population

aAvailable from. 1880 on
b
Available from 1930 on
CAvailable from 1910 on





61


Of the seventeen variables presented in Table III, four were

found unsatisfactory and were not included in the analysis. Two

variables, government employees and elementary school attendance, were

discarded due to the extremely erratic nature of their mean values

through time. Both of these variables were anticipated to demonstrate

a relatively consistent growth pattern through time. However, this

was not the case. The difficulty probably lies in the sources of the

statistics. The elementary school attendance figures were often

incomplete during the early period covered in the research. The em-

ployment figures were probably correct, but erratic due to constantly

shifting definitions of the census category during the early period.

Perhaps reflecting the stable nature of American industrialism,

or more probably an artifact of the census categories, the

percent of working age population in manufacturing has remained remark-

ably steady over the past hundred years. If a variable does not change,

it cannot be examined. Since the mean value of this variable remained

steady at around twelve percent'of the working age population, it could

contribute no additional information to the analysis. The fourth

variable, horsepower in use in manufacturing, presented a different

problem, also related to its variance. There was no effective means

of combining this variable with the other usable indicator of indus-

trialization, value added per capital. If the total horsepower were

measured against a per capital base, then the value added variable, being

measured in dollars, would dominate the index. If horsepower ;'ere

measured against some other base, then it would most probably dominate









9
the index. Since these two variables were highly correlated (0.85),

it was decided to use just one and not try to combine them. Since the

primary theoretical concern is with industrialization as a source of

wealth production, it was decided to use value added per capital as the

major indicator of industrialization. Figure 9 presents the mean

values of the index of industrialization from 1870 to 1960.

In this research, urbanization is viewed simply as a measure of

demographic concentration, and does not represent any complex of

expected associated traits. Two indicators, percent of population in

cities over 2,500 and percent of population in cities over 25,000

were combined into an additive index of urbanization. Figure 10

presents the mean values for the two indicators from 1870 to 1960.

It is necessary to refer to the concept of increasing scale to

explain the manner in which several of the indices, including

urbanization,were constructed. At any one time, T, a variable will

have some variance, S. Through time, the value of S may alter. While

we refer to the same conceptual,variable at many times, T, T+1,...T+N,

the variance of that variable at different times may be different. If

the variance of the variable changes in a steady and predictable manner,

due to increasing scale, it may be necessary to build into the index

a means of compensating for the alteration in the variance of one of the

sub-indicators. Specifically, if the component variables of the index

have a naturally occurring saturation point, then it will be necessary



9
For my objections to the use of standard scoring techniques of index
construction, see the. Appendix.

















Legend for Figure 9



Units of measure in dollars

Raw data from the U.S. Bureau of the Census,
U.S. Census of Population (Washington, D.C.:
U.S. Government Printing Office, 1872, 1883,
1895, 1901, 1913, 1922, 1933, 1943, 1952, and
1962), and the U.S. Bureau of the Census, Sta-
tistical Abstract of the United States (Washington,
D.C.: U.S. Government Printing Office, 1883, 1893,
1903, 1913, 1923, 1933, 1943, 1953, and 1963).

Means computed using Norman H. Nie et al., Statis-
tical Package for the Social- Sciences (New York:
McGraw-Hill Book Company, 1970).





















































































1870 80 90 1900 10 20 30 40 50 60


Figure 9


Mean Per Capita Productivity from Manufacturing


5 5 0


so ol



450'



4 0 0



350



30 0



250

















Legend for Figure 10



Percent of population in places over 2,500
population, 1870-1960.

Percent of population in places over 25,000
population, 1870-1960.

Raw data from the U.S. Bureau of the Census, Statistical
Abstract of the United States, 1883, 1893, 1903, 1913,
1923, 1933, 1943, 1953, 1963; and U.S. Bureau of the
Census, U.S. Census of Population, 1872, 1883, 1895, 1901,
1913, 1922, 1933, 1943, 1952, 1962.

Means computed using the Statistical Package for the Social
Sciences.





66


Figure 10

Mean Percent of Population in Cities over 2,500
and over 25,000 by Census Years


100





67


to continue adding co.locnent variables to maintain the variance of the

index. For example, during the early phases of urbanization in a so-

ciety, residence in a place of more than 2,500 people may be an adequate

index of urbanization. However, as the percentage of the population

resident in places of more than 2,500 approaches 100, then the index

is no longer an adequate basis for differentiating what is urban. In

short, the variable, through time, will cease to vary. Examine the

data presented in Table IV, presenting the means and standard deviations

of the two component variables of the index of urbanization.


Table IV
Means and Standard Deviations of Indicators of
Urbanization by Census Yearsa,b


Cities over 2,500
Mean Standard Deviation


21.4

24.3

32.2

33.8

37.9

42.2

46.0

47.3

55.6

61.9


16.8

19.0

22.2

21.7

21.4

21.4

19.9

18.3

16.0

14.8


Cities over 25,000
Mean Standard Deviation


10.2

13.0

15.9

18.4

21.9

25.4

28.7

29.2

31.3

35.5


12.6

13.6

15.8

17.7

19.6

20.8

20.3

19.4

15.5

14.7


aPaw data from the Bureau of the Census, U.S. Census of
Population, 1872, 1883, 1895, 1901, 1913, 1933, 1943,
1952, and 1962.
bComputations performed using the Statistical Package for the
Social Sciences.


1870

1880

1890

1900

1910

1920

1930

1940

1950

1960


I





68


Had we depended on the single variable, percent in cities over 2,500,

as an index of urbanization, our index would have begun to saturate.

We would have been led to infer that the importance of the conceptual

variable, urbanization, was decreasing, when in fact what was occurring

was the saturation of a society where approximately the same proportion

of the populace was resident in places over 2,500. In short, inter-

temporal research requires that operational indicators take account both

of increasing scale and, thus, decreasing variance in component variables.

The index of urbanization was created by summing the percent of

the population resident in cities greater than 2,500 and the percent of

the population in cities greater than 25,000. The majority of the

variance of this index from 1870 to 1930 'was contributed by the percent

in cities greater than 2,500. Since 1930, the percent in cities greater

than 25,000 has contributed the greater portion of the variance. It

should be clear that to continue the research for a time period beyond

1960, a further variable would have to be added to the index. Figure 11

presents the mean values of the index of urbanization from 1870 to 1960.

An index of media constumption was created following the same

procedure as used for urbanization. For 1870, the index of media

consumption was created by summing the per capital postal receiptsl0 and



10
Postal receipts per capital were used as an index of media consumption
following the argument that postal correspondence is a measure of
the level of community integration. Since our primary concern was
with media consumption as a means of community integration as well as
a means of modern behavior pattern transmission, the level of
postal activity was included.

















Legend for Figure 11



Raw data from the U.S. Bureau of the Census,
U.S. Census of Population, 1872, 1883, 1895,
1901, 1913, 1922, 1933, 1943, 1952, 1962; and
Bureau of the Census, Statistical Abstract of
the United States, 1883, 1893, 1903, 1913, 1923,
1933, 1943, 1953, 1963.

Computations performed using the Statistical
Package for the Social Sciences.













100


95


90'


85


80


75


70



G 5'


60


55


50





4 0


35


30'


25


20


15


10


1870 80 90 1900 10 20 30 40 50 60

Figure 11


Mean Values of the Index of Urbanization by
Census Years










the average daily newspaper circulation per 1,000 population. Figures

12 and 13 present the mean values of the newspaper and postal receipts

for 1870 to 1960.

In 1880, the number of telephones per 1,000 population were

added to the newspaper and postal receipts to create the index. This

same configuration was used for 1890, 1910 and 1920. Figure 14

presents the mean values of the telephone variable from 1880 to 1960.

Beginning in 1930, the number of radios per 1,000 population

was added to the index of media consumption. The composition of the

index from 1930 to 1960 was the sum of the newspaper, postal receipt,

telephone and radio variables. Figure 15 presents the mean values

for the total index of media consumption from 1870 to 1960.

Government activity was indexed by per capital state and local

government expenditures. It was unnecessary to use more than one

variable for this index since it had no saturation point during the

period under study. Table V presents the means and standard deviations

of the index of government activity for 1870 to 1960. Figure 16

presents the mean values of the index for the same period.


Table V
Means and Standard Deviations (S.D.) of the Index of Government Activity
by Census Years


Year 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960

:.ean 6.5 5.9 9.1 14.5 15.8 33.9 74.0 92.9 69.4 330.7

S.D. 3.6 3.5 5.2 7.5 11.4 13.5 25.8 29.7 19.3 70.2
















Legend for Figure 12



Quantities in real numbers.

Raw data from the U.S. Bureau of the Census,
U.S. Census of Population, 1872, 1883, 1895,
1901, 1913, 1922, 1933, 1943, 1952, 1962; and
Bureau of the Census, Statistical Abstract of
the United States, 1883, 1893, 1903, 1913, 1923.
1933, 1943, 1953, 1963.

Computations performed using the Statistical
Package for the Social Sciences.
























400-











300'











200











10 0-


0870 ~ 90 1900 0 0 'o

Figure 12

M an Daily Newspaper Circulation per 1,000
Population by Census Years

















Legend for Figure 13



Quantities in dollars.

Raw data from the U.S. Bureau of the Census,
Statistical Abstract of the United States,
1883, 1893, 1903, 1913, 1923, 1933, 1943, 1953,
1963; and U.S. Postmaster General, Annual Report
of the Postmaster General (Washington: U.S.
Government Printing Office, 1871, 1881).

Computations performed using the Statistical
Package for the Social Sciences.





75

















8




71




6





















2









----i --p---j----T------- -1-oTo---T----------T--'-----'----| 6%
1870 80 90 1900 10 20 30 40 50 60

Figure 13

Mean Per Capita Postal Receipts by Census Years

















Legend for Figure 14



Quantities in real numbers.

Raw data from the U.S. Bureau of the Census,
U.S. Census of Population, 1872, 1883, 1895,
1901, 1913, 1922, 1933, 1943, 1952, 1962; and
Bureau of the Census, Statistical Abstract of
the United States, 1883, 1893, 1903, 1913, 1923,
1933, 1913, 1953, 1963.

Computations performed using the Statistical
Package for the Social Sciences.






77






500











40





35











250





200





150





1 00




50






1870 30 90 1900 10 20 30 40 50 60


Figure 14

Mean Number of Telephones per 1,000 Population
by Census Years
















Legend for Figure 15



Quantities in index numbers.

Raw data from the U.S. Bureau of the Census,
U.S. Census of Population, 1872, 1883, 1895,
1901, 1913, 1922, 1933, 1943, 1952, 1962; and
Bureau of the Census, Statistical Abstract of
the United States, 1883, 1893, 1903, 1913, 1923,
1933, 1943, 1953, 1963.

Computations performed using the Statistical
Package for the Social Sciences.












1,500


1 ,000


95o


900


850


800


75 0


700


650-



600


550-


500'

450-






350


300


250-


200'


150'


100o


1970 80 90 1900 10 20 30 40 50 60


Figure 15


Mean Values of the Index of Media Consumption by Census Years

















Legend for Figure 16



Quantities in dollars.

Raw data from the U.S. Bureau of the Census,
U.S. Census of Population, 1872, 1883, 1895,
1901, 1913, 1922, 1933, 1943, 1952, 1962; and
Bureau of the Census, Statistical Abstract of the
United States, 1883, 1893, 1903, 1913, 1923, 1933,
1943, 1953, 1963.

Computations performed using the Statistical
Package for the Social Sciences.






81


325



300



275



250



225



200



175



150



125



100



75



50



25 -




1870 80 90 1900 10 20 30 4 0 5 0 6 0


Figure 16


Mean Per Capita Government Expenditures
by Census Years










For the years 1870 to 1910, the index of education was

created by summing the percent of the population who were literate

and the percent of the school age population attending high school.

Beginning in 1910, the percent of persons, age 19 to 24, attending

college was added to the index. Figure 17 presents the mean values

for the entire index of education for the years 1870 to 1960. Figure

18 presents the mean values for the index for the same period of time.

Electoral participation is the final conceptual variable in

the model. It was indexed by two different series of electoral

statistics. The first series was turnout in Presidential elections

as a percent of the total populace. Turnout in Presidential elections

is an obvious measure of participation; however, the base used to

compute a percentage turnout may occasion some comment. The decision

to ignore the obvious differences in electoral laws regulating the

suffrage was deliberately made. The most compelling argument for using

the total population, rather than the legally eligible electorate, as

a data base lies in the definition of political development on which

this work is based. Political participation is understood as one

element in the process of development. Participation is seen as a

political response to a change in the social environment. It is,

of course, possible to disenfranchise certain groups in the society.

The basis for disenfranchiscment may be color, language, religion or

sex, but, whatever the reason, exclusion of a social group from political

participation does not exclude them from the society. There are

reasonable bases for exclusion, such as age or mental incapacity, but
















Legend for Figure 17



Percent literate.

Percent attending High School.

Definition of literacy changed in 1950, to
functional literacy.

Raw data from the U.S. Bureau of the Census,
U.S. Census of Population, 1872, 1883, 1895,
1901, 1913, 1922, 1933, 1943, 1952, 1962; and
Bureau of the Census, Statistical Abstract of
the United States, 1883, 1893, 1903, 1913, 1923,
1933, 1943, 1953, 1963; and U.S. Commissioner
of Education, Annual Report of the Commissioner
of Education (Washington, D.C.: U.S. Government
Printing Office, 1871, 1881).

Computations performed using the Statistical
Package for the Social Sciences.





84


,- .-..--


.
J


/
... ~ ,-- .--.


1870 80 90 1900 10 20 30 40 50 60

Figure 17

Mean Percent of Population Who Are Literate and
Mean Percent of Population Attending High School
bv Census Years


100





10




80





70





60





50




40


30





20"





10

















Legend for Figure 18



Quantities in index numbers.

Raw data from the U.S. Bureau of the Census,
U.S. Census of Population, 1872, 1883, 1895,
1901, 1913, 1922, 1933, 1943, 1952, 1962; and
Bureau of the Census, Statistical Abstract of
the United States, 1883, 1893, 1903, 1913, 1923,
1933, 1943, 1953, 1963; and the Report of the
Commissioner of Education, 1871, 1881.

Computations performed using the Statistical
Package for the Social Sciences.





















1,300



1,200



1,100-



1, 0 0 0-



90 0



800




700



600



500



400



300-



200



100


1870 80 90 1900 10 20 30 40 50 60

Figure 18


Mean Values of the Index of Edu.cation
by Census Years





87


even here, the criteria may yield to a changing social milieu.

We may be forced to argue that the eighteen-year-old vote is more

developed than the twenty-one-year criterion. If it becomes clear

that the desire for participation of the eighteen-year-old population

is high, then their exclusion from political decision making indicates

a lack of development. The argument can be raised that in a caste

system, it is possible for one caste to advance while a lesser caste

remains backward -- perhaps. But basically, such an argument seems

to claim that the upper caste constitutes the society. For this

argument to be valid, we must be willing to equate political non-

existence with social non-existence.

It is precisely the increasing self-awareness of social existence

upon which we have predicated our argument of increasing pressure

for political growth on both the permeance and participation dimensions.

Hence it was deemed permissible to use the total population as a data

base. This base was also thought most reliable, since an enumeration

of the population would tend to.be more valid than an estimate of the

eligible electorate.11



11
Walter Burnham argues (Walter D. Burnham, "The Changing Shape of the
American Political Universe," American Political Science Review
(March, 1965), 7-28) that the American political universe altered
from a predominantly high participatory state to a less participatory
state around the turn of the century. This change can partly be
attributed to a decline in participation rates by eligible white
males.





88


Operating on the assumption that alternative electoral arenas

might reflect different developmental patterns, it was deemed

advisable to check the model by using another series of electoral

statistics. The second series used was turnout in off year Congression-

al elections. Congressional elections were chosen since the constituency

represented is at the sub-state level and the pressures in this type

of election would seem to be the most unlike the Presidential race.

Yet, the office is significant enough to merit general attention. Off

year elections were used to minimize the pulling effect that a

Presidential contest is believed to exercise on the electorate. Thus,

the pressures to participate were not expected to come from the effects

of a possibly exciting Presidential contest.

To further minimize the deviations inherent in single election

results, the electoral data were aggregated around the appropriate

decennial census periods. For example, the elections for the Presidency

for 1888 and 1892 were averaged to give one figure per state, per

decade. In this way, minor variations in turnout were smoothed over,

and the effects of particular campaigns on the level of participation

were partially controlled. Figures 19 and 20 give the mean values

of the participation indices for the period 1870 to 1960.

Before discussing the methodology used, a summary statement of

the variables and their operational indicators, as used in the

research, is presented in Table VI.

















Legend for Figure 19



Raw data from Sven Petersen, A Statistical
History of the American Presidential Elections
(New York: Ungar Publishers, 1963).

Computations performed using the Statistical
Package for the Social Sciences.





90


35




30 "










20




15

















1870 80 90 1900 10 20 30 40 5i G0

Figure 19

Mean Values of Turnout in Presidential
Elections by Census Years

PAGE 1

Cor.!;iuini cat ions Model of Democratic Political De in the United States: 1870 1960 •By 5E F. I'Li-RQUnTie DISSERTATION PRliSBNTED TO THE G'^ADijATl: COUNCIL OF TiiE UNIVnilSITY OF FLORIDA IN PARTIAL FULFILLMENT OF TlIC RLOUIRU:.ix:NTS FOR TAii n^iGaUH OF DOCiOR OF PlIIi.OSOPHY UNiyizRsn'Y OF fi.orida

PAGE 2

ACKNOWLEDGMENTS Two people provided the support and encouragement v.-hich enabled me to complete this work: Dr. Alfred Clubok, the Chainnan of Vi.y Dissertation Comiittee, and Penny Mai-quette, my best friend. T wo\'ld also like to tliank the members of my Coiruiu tteo. Professors Frank Munger, Manning Dauer, John Spanier and Joseph Vandiver for giving me their suggestions and tiieir time.

PAGE 3

TABIH OP CO>ri']:NTS ACK>;OiVLEDG^ft:NTS ii LIST OF TABLF.S iv LIST OF FIGURr:S vi ABSTRACT ix INTRODUCTION 1 CHAPFFR I: A Philosophical Framework paid a Definition of Political Development. . , 4 CHAJ'TCR 11: Social Change and the Insatiable Personality 29 CHAPTHR III: Indices and Methodology 55 Data Sources 57 Variables 59 Metliodo logy 94 CIlAPTliR IV: Models of Political Development 99 Basic H)potheses 100 An Inter-Temporal Test of the Model of Development. .. 133 CHAPTHi; V: Discussion and Conclusions 140 Conclusions , 149 APPENDIX 1^BISLIOGRAPHY -^'^ BIOCRAPinCAL SKETCH 1^0

PAGE 4

LIST OF TABLES I Doc'^nnial Correlations of Inter-Party CoiTipetition and Education 55 II Number of Observations for Each Decennial Census Period 57 III Original Variables Selected as Possible Indicators 60 IV Means and Standard Deviations of Indicators of Urbanization by Census Years 67 V M<^ans aiid Standard Deviations of the Index of Government Activity by Census Years 71 VI Variables Included in Each Final Index with Years of Inclusion 93 VII Causal Inference Prediction Equations for Each Hypothesis 101 VIII Aggregate Correlation Matrix of the Six Indices for the Period 1870 to 1960 102 IX Aggregate Correlation Matrix of Six Indices for the Period, 1870 to 1910 108 X Aggregate Correlation Matrix of Six Indices for the Period, 1910 to 1960 110 XI Prediction Equations for the First and Second Models of Developiuent 113 XII Inter-Censal Correlations of Government Activity at Ti^ with Political Participation at T2 for the Period, 1870 to 1960 119 XIII Multiple Correlation Coefficients of Political Participation with Indices of Education, Media Consum.ption, Government Activity, Urbanization and Industrialization by Census Years 120

PAGE 5

Xiy Multiple Correlation Coefficients of Government Activity with I'ledia Consv.mpt Jon, Urbanization and Industrialization by Census Years '21 XV Cross Year Correlation Matrix for the Period 1870 to 1960 122 XVI Correlation Matrix of Original Variables with Logarithmic Transformed Counterparts 126 XVII Correlation Matrix of Six Variables Using All Observations 129 XVIII Multiple Correlation Coefficients for the Three Models of Development , 132 XIX Aggregate Correlation i^!atrix of the InterTemporal Correlations: 1870 to 1910 134 XX Aggregate Correlation Matrix of the InterTemporal Correlations: 1920 to I960., 136 XXI Correlation Matrix of Media Consumption, Government Activity, and Political Participation, 1960 146 XXII Relationships of Mobilization Levels and System Types as Crisis Outcomes in Six Nations. 153

PAGE 6

LIST OF FIGURES 1 Total Pig Iron Shipments in the United States: 1820 1960 45 2 Causal ficdel of Three Concepts with SiiPultaneous Interactions 46 3 Causal Model of Three Concepts with One-Way Causation 46 4 Causal Model of Three Concepts witli InterTemporal Effects 47 5 Inter-Temporal Effects of Three Operational Indicators 50 6 Inter-Temporal Effects of Four SocioEconomic Indicators 51 7 Causal Relations of Four Variables and Government Activity 52 8 Final Representation of a Six-Variable Model of Political Development 54 9 Mean Per Capita Productivity from Manufacturing 64 10 Mean Percent of Population in Cities over 2,500 and over 25 ,000 by Census Years 66 11 Mean Values of the Index of Urbanization by Census Years 70 12 Mean Daily Nevvspaper Circulation per 1,000 Populat ion by Census Years 73 13 Mean Per Capita Postal Receipts by Census Years 75 14 Mean Number of Telephones per 1,000 Population by Census Years 77

PAGE 7

15 Mean Values of tlic Index of Media Consiunpt ion by Census Years 79 16 Mean Per Capita Government Expenditures by Census Years 81 17 ^lean Percent of Population Ulio Arc Literate and Mean Percent of Population Attending High School by Census Years 34 18 Mean Values of the I.idcx of Education by Census Years 86 19 Mean Values of Turnout in Presidential Elections by Census Years 90 20 Mean Values of Congressional Election Turnout by Census Years 92 21 Empirically Identified Causal Model of Development in the United States for the Period 1870 to 1960 104 22 Plot of Correlation Coefficients of J!cdia Consumption, Government Activity, Education, and Political Participatiun Versus Tin,e: 1870 1960 107 25 Empirically Identified Model of the Development Process in the United States: Phase I 1870 to 1910 110 24 Summary Presciitation of Both Pliases of the Causal Model .' 114 25 Mean Values of the Index of Education: 1370 to 1960 124 26 Logarithmic Transformations of the f^ean Values of Media Consumption, Industrialization and Government Activity: 1870 to 1960 128 27 Empirically Identified Model of the Development Process in the United States Using All Observations as Data Base 131

PAGE 8

28 Fiiipirically Identified Model of the Development Process in Phase I, 1870 to 1910, Using Inter-Teinpoi'al Correlations 135 29 Empirically Identified Model of the Development Process in Phase II, 1920 to 1960, Using Inter-Temporal Correlations 138 30 First Section Causal Model Resulting from Total Observations Test 141 31 SuiiiiTiary Presentation of Phase I Model, 1870 to 1910 143 32 Summary Presentation of Phase II Model, 1920 to 1960 ..144 33 Suinmary Presentation of Model Generated from Total Observations Matrix 144 34 Model of Expected Third Phase Relations 146

PAGE 9

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfil liaent of the Requirements for the Degree of Doctor of Philosophy A COMMUNICATIONS MODFL OF DEMOCRATIC POLITICAL DEVELOPMENT IN THE UNITED STATES: 1870 1960 By Jesse F. Marquette August, 1971 Chairman: Dr. Alfred B. Clubok Major Department: Political Science Political development theory has consistently suffered from a lack of reliable e;npirical referents against v\]-iich to check proposed concepts. One major rea-jon has been that comparative lesearch lias generall}' been cross-sectional. A second major reason has been that the research has been performed 'using nations as data points. From this crosssocietal sample, attempts have been made to infer intrasocietal developmental sequences. Tills research iias atte.mpted to overcome these failings. Only one society, the United States, is used as a referent. Data points are the states of th.e United States. The time period of the research extends from. 1S70 to i960. Obviously, generalizing from one society's past experiences with chaiige to the experiences of other societies is

PAGE 10

somewhat questi.onable, but certainly less so than the opposite p^•c;ccss. Tl\e time period used is adequately long to justify expectations that the generalizatioiis \.'ill be reliable. I-xplaiiations of tlie relationships examined are as parsimonious as possible, and generally refer only to the physical necessity of certain processes. A constant atterw.pt is ii'ide to avoid any explanations which are ethnocentric despite the obviour^ly culture-bound basis of the data. The primary objective has been to develop a causal model of the relationship amongst four social variables : industrialization, urbanisation, media consuuiption, and education; and two political variables j govenimcnt expenditures and political participation. Tn essence, ten tests of the proposed model were made, one test for each decennial census period from 1870 to I960. T'..-o nev;, consistent, sequential c.md simple models of development arc apparent. Tlie first phase model extends from 1370 to 1910, and the second extends from 1920 to 1960, A reforiiinlaticn of de\ elopmental theory is undertaken in the light of these findings.

PAGE 11

INTRODUCTION T'ncories of political development liave consistently suffered from a lack of reliable erapirical referents against v;hich to check px'oposed concepts. One methodological failing has been tliat while development implies cliange, most comparative research has been cross sectional. A second major failing relates to the use of nations as data points, attempting to infer intra-societal sequences of dc/elopment from cross-societal samples. Mien both these failings occur in the same research the results are, at best, confusing. The research reported here attempts to overcome these failings. Only one society, the United States, is used as a referent. The time period of the research extends from 1870 to I960, vdth the individual states as data points. Although generalizing from one society's experiences with change to the experiences of other societies is somewhat questionable, it is certainly less questionable than the opposite process. This research has attempted to deal with these problems of cross -societal comparisons iu several ways. The time period used is adequately long to give the generalizations high reliability. Further, the explanations of the relationships found are as parsimonious as possible, and generally refer only to the existential necessity of certain processes. In addition, a conscious attem.pt has been made to avoid explanations which

PAGE 12

are ethnocentric despite tlie obviously culture bound basL-^ of the data. The objective of this research is tlie development of a causal model of the relationship among four social variables ( indust rial i zation , urbaniza tion, media consumption , and education) and two political variables [go vernment activity and political participation) . Tli.e study is founded on a proposed model derived from work by Cnudde ajid McCrone^ and an earlier vvoi-k by this author, ^ while tlie theoretical basis is derived from a conceptualization advanced by Lerner. In essence, ten tests of tlie proposed model were made, one test for each decennial census period from 1370 to 1960. Empirical research quickly proved both the model and the theory to be unfounded. Ilo'.v'evcr, two new, consistent, sequential and simple alternative models became apparent. Using these two models an attempt is made to reformulate developmental theory in tlie light of the findings. Chapters one and two present part of this reformulated tlicory as a basis for understanding the models '.;hich follow. After presenting the reformulated theory and the static models, a series of h.ypotheses will be presented regarding the relations of rates of cliange. These ]\ypotheses will then be tested over the full hvmdred year time span. 1 Donald J. McCrone aiid Charles F. Cnudde, "Toward a Coimriunications Theory of Democratic Political Develcpm.ent : A Causal Model," Ame rican Political Science Review (March, 1967) , 72-79. 2 Jesse F. Marquette, "Social Mobilization and tlie Philippine Political System," Comparativ e Political Studies puly, 1971}, in pi-ess. Daniel Lerner, "Communications Systems and Social Systems. A Statistical Exploration in History aaid Policy," Behavioral Sci ence (March, 1957), 266-275.

PAGE 13

In this iiianner, an attempt is made to iriaintain a constant and reinf'">rcing interaction between theory and research.

PAGE 14

CHAPTER I A PHILOSOPHICAL FMMEWORK AND A DEFINITION OF POLITICAL DEVELOPMENT Aa effective definition of political development requires an effective definition of a developed polity. Development implies change fron one state of political realiity to another and we should be able to define both states. Definitions of political develcp?iicnt have proliferated due to an inability to settle on one definition of a developed polity. lliis results from the generally accepted opinion that no tally developed polity currently exists. If, in fact, there is no fully developed polity, then our definitions and descriptions must be prescriptive; not v;hat is, but what should be. The referents of "whet should be" -"re highly conditioned by the intellectual heritage from which we proceed. Thus, an economist like Rostow sees political change primarily as an epiphenomcnon of social cliange. A political scientist like Organski sees political development as a necessary concomitant of 1 Lucian iv'. Pye, Aspects of Poli tical Deve lopment (Boston: Little, Brown and Company, 1966) ,~3"l -48. 2 Vv'alter Vv'. Rostow, The Stages of Econom ic Gro wt h: A Non-Communist ^•|anifesto (Cambridge7 England: Cambridge University Press, 1966). 5 A.F.K. Ciganski, The Stages of P olit ical Development (New York: Alfred A. Knopf, 1965). " ' "'

PAGE 15

4 economic development. A psychologist like McClelland views development as a result of altered personality patterns, while a political sociologist like Aptcr sees development as changing patterns of group conflict. There are many other ways to categorize existing definitions of development, and rather tlian examining them as above, they can be categorized according to the type_ of explanation used. Holt and Turner argue that the v.-orks of Pye, Hagen and McClelland are reductive explanations; attempts to explain macro-social phenomena in terms of generalizations foimded on mici-o-social phenomena. Flolt and Turner argue tliat this form of explanation is ineffective because it lacks composition laws. They suggest that emergent explanations are the best currently available in social research. Depending upon the purpose, many more systems of categorization can be developed. In developing a causal model of the socio-economic bases for political chaTige, there is one system of categorization v/hich clarifies much of t!\c previous work. This system of categorization is Tiavid McClelland, The Achieving Society (Princeton: D. Van Nostrand Company, Inc., 1961). 5 David Apter, The Politics of Modernization (Chicago: Chicago University Press, 1965). 6 Robert T. Holt and John E. Turner, The Politica l Basis of Econ omic D evel opm ent (Princeton: D. Van Nostrand Company, Inc., 1965). 7 Ihid. , 26-27. Ibid., 28.

PAGE 16

based on the alternative perceptions of the social locus of political change lield by various authors. Tnese divergent perceptions lead to two different riodels of tlie causal relations between social change and the political order. One alternative empl:asizes tl^e significance of minority innovation and support of change. T\\e minority emphasis sees either an elite or a special minority group as a continuous element in tlie process of modernization. Tliis minority constantly provides the drive for m.odei-nization over long periods of time. Mass behavior is affected by minority desires, but tlie mass is net expected to become a causal agent in further change. A second alternative sees each newly mobilized individual as a cousal agent in the furtlier process of mobilizing even more individuals, This opposite perception emphasizes the significance of mass attitudinal and behavioral change as fundamental to continuous political change. Generally, the practice has been to emphasize one view or tlie other; 'however, two authors, Nettl and Apter, have attempted to synthesize these divergent perceptions. J. P. h?ettl outlines two alternative forms of political mobiliza10 tion, one referred to as stalactite and the other as stalagmite. In stalactite mobilization an elite is responsible for deliberately moving 9 The ;iunority emphasis is represented by such autliors as Daniel Lerner, E.H. Hagen and Lucian Pye, while the mass emphasis is represented by Karl Deutsch, David McClelland and Gabriel Almond. 10 J . P . Nettl, Po l_it ical Mobil izat i o n: A Sociological A nalysis of Me t hod s and Concepts' [New York: Basic Books, 1967), 288-316.

PAGE 17

the mass of the populace tov.'ard political activity. This activity is directed and focused according to the desires of the niobilizing elite. Stalagmite mobilization, on the other hand, is mass-based and antielitist; it is the av;akoning of the mass of the populace, shaped by other forces. The institutional forms of political life resulting from tliese alternative forms of mobilization are radically different. I'/liile stalactite mobilization is seen as eventuating in undifferentiated, elitist polities, stalagmite mobilization results in differentiated, constitutionally based polities. David Apter attempts a similar synthesis with his concepts of the reconciliation systera^^ and the mobilization system. 13 in Apter's rubric, the term reconciliation system is xinalogous to Nettl's stalagmite mobilization. Here an attempt is being inade to gradually accom.modate to the rising demands of the populace. Over a period of time, demands are articulated by more and more groups in the system. Since the general value orientation of this system is pragmatic and meets already extant problems, Apter regards the value system as instriunental. Apter's mobilization system is analogous to Nettl's U rbid_. , 515. 12 Apter, op.ci_t_. , 391-421. 15 "" Ibid.. 357-590. 14 /\.pter defines an instrujnental value system as one vhich emphasizes the solutions to relatively mundane problems. Here, politics is vie-wed as the process of solving short-tern, existing problems. A consumatovy value system, on the other hand, is expected to place great emphasis on realizing potentiality, and the ethical basis of politics lies in the activity of achieving specified ultir.iate goals

PAGE 18

stalactite iiiobilization. Mobilization systems are di'iven by elite desires to adiieve certain goals, notably the political mobilization of the mass. Insofar as the system is not responding to, but rather initiating, demands, the structures wliich evolve tend to be undifferentiated. Tlie value system is seen by Apter to be consummatory -ultimatego? 1 oriented. Both Apter and Nettl are unusual in that they include opposite perceptions of the social locus of political change in their scliema. Uliile other authors have, to some degree, discussed mass and minority loci, they have generally empliasized one at the expense of the other. The woi'k of Daniel bemier, E.E. Hagen, and Lucian Pyo emphasize tlie significance of minority or -elite activity in the process of modernization. Lerner and Pye are expressly concerned with the activities of elites in initiating and sustaining modern behavior. Lerner's empliasis on elite activity seems to stem from the focus of his investigation: the significance of mass communication in the dcvelcpm.ental process. '^ Since' the production and dissemination of 15 Daniel Lerner, T he Passing of Traditi onal Society : Moderni zing the Middle Gast (NevNf York: The Free Press, 1958). 16 E.E. Hagen, On the Theo ry of Social Cha nge (I^omewood, 111.: Tlie Uorsey Press, 1962) . " ' 17 Lucian Pye, Politics, Personality and Nation-Building: Burma's Sea rch for I dentity (New Haven: Yale University Press, 1962). 18 Lerner, Hie Pas sing of Traditional Soci ety; Lerner's research centers un tlie Middle Eastern countries, all with relatively new mass cornmun i c a t ion s nets.

PAGE 19

I'nass media is an elite activity, Lcrner has tended to emphasize elite behavior in explaining the process of development. Pye has also 19 suggested elite significance in the developmental process, equating the identity crisis of the elite with the identity crisis of tlie nation. 20 Hagen, on the other hand, does not limit minority interpretation to a discussion of elites. Tlie minority groups which Hagen sees as instrumental in the process of change are characterized by radically different attitudes, but not by elevaced position. An opposite emphasis on social locus is represented by the conceptual scliema of Karl Deutsch,-^ David McClellaiid'^ and Gabriel Almond.--^ Their theoretical concern is with the significc^nce of mass attitudinal and beh.avioral clianges in effecting the process of modernization. Deutsch dwells on the exposure of unnobilized individuals to aspects of modernity. He is clearly concerned with the pr ocess of lAobilization, the spread of attitudes which are distinctly nonLraditional. Deutsch stresses the necessity of mass attitudinal change, believing that while a minority .may provide the spark, significant change 19 Pye, Pol itics, Personality and Nation Building; Pye's research group consisted of a small sample of Burmese national government figures. 20 Hagen, oo. cit. 21 '^ Karl W. Deutsch, "Social Mobilization and Political Development," Aj.erican Po litical Science Review (September, 1961), 493-514. 22 " McClelland, 0£. c_it. 23 Gabriel A. Almond and SiuJiey Verba, The Civic C ul ture: Political Attitudes a nd Do inocra cy in Five Nations (Boston: Little, Bro\;n and Company, 1965). 24 ' ' Deutsch, op. cit.

PAGE 20

10 is obtained only through mass mobilization. Gabriel Almond's presents tation" is typological and little concerned with the process of change. His primary effort involves the explication of possible political system states, which he arrays in a sec)itence from less to more developed. He attempts to describe development in terms of nass attitudes conducive to democratic forms of political organization. Both Netti ° and Holt and Turner '^ have specifically criticized Almond for tl;i s rather myopic view of development. McClelland' s theory of the basis of economic development is clearly one of the most effective explanations yet offered. ^^ It combines botli an emphasis on the necessity of group cliango (cultural values) and psychological patterns (socialization). McClelland has attempted to combine both tlie forms of explanation described by Holt and Turner; the analysis of cultural values is an attempt at emergent explanation, while the analysis of socialization patterns is an attempt at reductive explanation. McClelland expects different patteiris of socialization to lead to different levels of n-achievement. It is not clear why some families adopt one j^attern of socialization instead of 25 Almond tuid Verba, 02_. cit . ; and also Gabriel A. Almond and G. Bingliam Powell, Comparati ve Politics: A Develop menta l Approach (Boston: Little, Brown and Company, 1965), especially 16-41. 26 Nettl, ou. cit., 110. 27 ""* " Holt and Turner, op_. cit . , 13-16. 28 McClelland, op. cit., 1-62.

PAGE 21

another, unless there is a differential diffusion of cultural values. At this point it is obvious tiiat Molt and Turner's argument for the 9Q necessity of composition laws is well taken." Had McClelland been able to produce an effective explanation for tlie differential diffusion of socialization patterns, his theory v.-ould have been immeasurably more po'.vcrful. The degree of success which McClelland achieved, using a combination of types of explanations, suggests that Holt and Turner's use of only emergent explanations is unnecessarily restrictive. As suggested earlier, alternative perceptions of the social • locus of change lead to different expected causal paths of change. Tliis contention may be illustrated using the formulations of Lerner-^^ and Deutsch."^ Cnudde and McCrone have attempted to test a causal ri:odel derived from Lcmer's work using data dravm from Cutright's seventy-six nation study. "^ Using Cutright's operational indicators of g rb an i z at i on , education , media consu mp tion and politic al development , they test a causal path represented by urbanization (U) -> educ ation (E)->m edia consum.ption (M) -> politica l developm ent (D) , conceptually derived 29 Holt ;ind Turner, og_. cit. , 28. 30 Daniel Lerner, "Communications Systems and Social Systems: A Statistical Exploration in History and Policy," B ehavioral Science (March, 1957). 31 Deutsch, op. cit . 32 " Donald J. McCrone and Charles F. Cnudde, "Toward a Communications Theory of Democratic Political Development: A Causal Model," American Political Science Review (March, 1967), 72-79. 33" Philips Cutright, "National Political Development: Measurement and Analysis," in Nelson Polsby et al_. (editors). Politics a nd Social Life (Boston: Houghron Mifflin, l'963) , 569-582.

PAGE 22

12 from the following passage fi-om Leiner. . ...urbanization requires rising literacy for individual participation. At a certain point, when urbanization has done its woi-k, literacy becomes the independent variable in the process of growth and a new phase of modenii zation begins. But the growth of literacy itself, in this phase, soon becomes closely associated with the growth of media. Tlie media teach literacy and growing literacy develops the market which consiunes the media product. ^^ 71ie opposite path of U -> M -> E -> D, is implicit in the work of Karl Deutsch. ...two distinct stages of the process: (1) the stage of uprooting or breaking away from old settings, Iiabits and committments; and (2) the induction of the mol)llized persons into some relatively stable new patterns of group membership, organization and committment. 35 Here, tlie first stage is represented by exposure to aspects of modernity, an essentially passive process. Tlie second stage suggests a more institutionalized resocialization of the individual. In the society as a wliole, the second stage suggests the process of education. Although education as envisioned here is not necessarily formal educationj it must be relatively formalized (represented by stable and accepted socialization processes), if the individuals are to be inducted into stable new patterns of group coimnittments . Over time, the best measure of tliis process would be the spread of formal education. 34 Lsrner, The Passi ng of Traditional S ociety , 60. 35 ' "" * Deutsch, op. cit., 494.

PAGE 23

13 Tliese differenj: perceptions of the path of change could have ai'isen either froin the authors' opposing views of the relation between media consumption and education, or from the autliors' different emphases on social loci. First, it is questionable that media consumption requires literacy on any significant scale. Lerner himself describes an effective melding of the modern media form of radio and the traditional oral coimTiunication network when he describes the Chieftain of Balgat. As the mdel individual in an oral conmunication net, the Chief's possession of a radio carried the entire net into a media system. There is no reason why one literate individual could not substitute for the Chieftain's radio in the coiamimication network of Balgat, or any other village. Secondly, if the primary function of media consiunption and education is to increase individual empathy, there is no reason why education must precede media consimiption. An alternative formulation might substitute "aspects of modernity" for "media consumption," as does Deutsch. In Deutsch's formulation, media consumption is merely one method of exposing the individual to modern behavior patterns. Having become socially mobilized through exposure to aspects of modernity, the individual would be expected to becom.e politically mobili;^ed after e.-posure to educational opportuiiity. 'fliis substitution does no serious Lerner, The Pass i ng of T raditiona l Society, Chapter 1

PAGE 24

damage to Lcrner's argument and, viewed in tliis way, the authors do not seem to have fundamentally opposed views of the significance of media consumption and education. The preceding discussion was designed to emphasize the necessity of delineating why a particular problem is approached in a specific manner v;hen evaluating the resultant conclusions. \'"\en there are many descriptions of an object, many definitions of a concept, tliere are two choices in explaining the multiplicity. Either tlie object or concept is extremely complex and invites many partial descriptions, or there are many opposing viewpoints about the nature of the object. In the case of political development theory, the existence of many viewpoints has created a lengtliy catalog of definitions. A definition is a brief description, adequate to the degree that the thing described becomes generally recognizable. It is possible to define political development in many ways. Certain aspects may be chosen v;hile others are ignored, depending upon our perceptual vantage point. To advance a generally useful definiti on of political development, the primar y concern must be not only with the definition itself , b ut also w it h making explici t the viewpoint from which th e d efinition emerges . In this sense the definition is a road map; if others are 37 Pye, Aspects of Political Development, 33-42.

PAGE 25

15 to follow the definition and reach a specified goal, they must start from the same point as the v.'riter. Therefore, the following short ex-cursion into cpistcmology will attempt to clarify this author's basic viev.-points. The closer we are to an object or event, either in space or time, the more importajit the object or event appears, hlien considering a social fact, it is important to place it in proper temporal context to avoid overly inflating its significance. As we include greater reaches of the past in our research we are able to eliminate tlie evanescent from consideration. Perhaps Davis gives the best illustration of the significance of time.^ Ulien we plot the mean values of sccio-economic attributes for many points in tiip-e we usually observe an extremely ragged cur've. It is possible to reduce the raggedness of the curve by examining the trend line; the general tendency of the curve to increase or deci-ease through time. Vae longer the time period examined, tlie more reliable the trend line becomes. Conversely, the more foresliortened the tJme frame, the more likely we are to read too great a significance into transient events and the more inaccurate our generalizations will be. The number of individuals included in tlie data base is at least as important as the time frame used. The adequacy of our generalizations can be as adversely affected by concentration on elite behavior as by 38 Jiarold T. Davis, The .\na l ysis of Economic Time Series (Bloomingtonj Indiana: Tne Principia Press, 1941), Chapter I.

PAGE 26

concentration on too short a time span. Political development theory must have the total society as its basic referent; it must refer to the effects of political and social cliango on the niass of men in society. Political change has occurred in maFiy different ways in many different societies. Many of these patterns of change have benefited some members of the societies where they occurred, but luitil political change benefits a significant proportion of all the members of a society we are not justified in talking about political development. Our ability to compare levels of development will be severely hampered if the same measuring stick is not used to m.ake the comparison. In this author's view, social theories are useful only to the degree to which they aid us in explaining why some social act renders an individual or group more or less content with its existence. As the work of t'ne psycliologically oriented theorists demonstrates, wliatcver the first cause, the ferment of social cluinge lies within individuals. Society is the primary environment of the individual and the texture of his life is predominantly the texture of his social surroundings. Yet his social environment is the totality of the characteristics of socially relevant others. It seems reasonable that social science should view the world primarily in term.s of the benefit or harm individuals may expect in a social situation. In this view, the individual is one element in a social ratio. Me. is part of the soc.'.al whole, yet in term.s of enumerable attributes he is always a subset. Concentration on eithec the individual or the society ii^nores the relation between

PAGE 27

17 tlie two -the ratio, which is the basis of social change. In o?;der to proceed from a ui^eful vantage point it will be ]iecessary to specify the characteristics of our basic perceptual matrix. First, it must force us to focus on tlie process of social change, and not merely on static situations. Hence, it should allow us to follow the transition of a society through time, and should be generalizable to other societies. (Unfortunately, most of the comparative, empirical vjork done to date focuses on different societies in different stages of change.)^ Secondly, it must force us to focus our attention on tiie relation between the individual and society rather than on one or tlie other. It should allow us to discuss social change in terms of individual satisfactions, for whatever the first cause, social change r esults from the fr ustrations or satisfactions of indiv iduals. Finally, the basic elements of the matrix should be abstracted, not culture bound; a problem which most of the systems theorists have encountered. A usable vantage point, partially specified and clarified, has been advanced, first by IVeber^'and later others/^ as the concept of 59 See, for example, Cnudde and McCrone, 0£, cit . ; Outright, 0£. cit.; Arthur K. Smith, Jr., "Socioeconomic Development and Political Democracy: A Causal Analysis," M idwest Journal of Polit ical Science (February, 1969), 95-125'. ~ 40 Max Weber, "Class, Status and Party," in R. Bendix and S.M. Lipset (eds) , Class, Status an d Povvrer (Glencoe, 111.: The Free Press of Glencoe, 1953), 63-75. 41 Melvin M. Tumin, So cial Stratification: The Fo rms and Fu nctions o f Inequality (Englewood Cliffs, N.J.: Prentice-Hall, 1967); Gerhard E. Lenski, Power and P rivilege: A TTieor y of Social Stratification (New York: McGraw-Hiir^ook Company, 1966); and loan Davies, Social Mobility^ and Political Change (New York: Fraeger Publishers, 1970).

PAGE 28

life chances. IVebcr describes life chances as "the typical chances for a supply of goods, external living conditions and personal life experiences." V.'eber used the concept of shared life chances as a basis for defining social class. Later theorists have used the distribution of available life chances for defining the social stratification system. 'I'll-ien ue describe tlie rajige of life chances accessed by im individual, we are describing the texture of his life. We are assessing tlie general possibility of his happiness and the range of gratifying activities open to him. In determining the degree of social stability, two variables must be considered: the distribution of available life chances and the normative justification for the distribution. An existing distribution of life cliances is supported by values 43 and norms which tend to legitimize the distribution. Hence, the life chances available to any individual are, in part, socially determined, for no society conforms to complete acliievement orientation. Essentially, tlie concept of life ch.ances rests on the relation between total available possibilities for individual satisfaction and the possibilities actually available to each individual. Since social systems are hierarchical, the norm structure must legitimize inequality. Norm structures supportive of a particular stratification system must provide the justification for 42 V.'eber, op. c it . 43 Tu;nin, oo. cit., 14,

PAGE 29

19 restricting life chances. Stable social systems result from mutual reinforcement between the actual distribution of life chances and the normative support for that distribution. Norm and value support are the key determinants within a wide range of actual distributions.'^ Miile the society's norms and values convey a sense of "oughtiiess" for the stratification system, the distribution of life chances remains authoritative. ' ' ' If available opportunities are sufficient, and the distribution accepted, the average individual will remain passive and the social system will remain generally stable and slow to change. In truth, no social system is stable in the sense of stasis; some change is always occurring, even if that change is only biological replacement. There is no general reason to expect a society to be stable as a basic tendency. The pui-pose of develorjmental theory is to explain why some societies change and others do not, and w-hy some societies follow one patli of change and others a different path. 44 Ibid., 44. 45 David Easton, The Political System: An Inquiry into the State o f Po litical Science (New York: Alfred A. Knopf, 19S3) , 129-134. 46 Lenski, o£_. £21_. , 3 (footnote). 47 Harold D. Lassweil, Politics: Mio Gets Vfnat, UTien a nd How (New York: McGraw-Hill Book Company, 1936). 48 Easton 's definition of politics as "the autl^oritative allocation of values" (o£. cit .) is really a description of the stratification system. Lenski (ojo. ci t_. ) also argues that Lassweil (ibid.) lias come to recognize that his definition of politics as deciding "who gets what, when and how" is also a dtscrintion of the stratification s/stem.

PAGE 30

20 Merton's discussion of anomie'^^ as the incongruent relation between cultural goals and institutional means for goal achievcircnt illustrates the necessity of balance.^ Lerner discusses the same interaction under the intriguing title of the "Wrint/Get Ratio. "^^ The result of an imbalance between the elements of Lerner' s ratio is individual frustration. Both authors argue that each individual's reference groups and socialized values establish the basic goal expectations appropriate to that individual's original position in tlie stratification system. The sanctions available to the roles occupied by the individual provide the possibility for the satisfaction of the goal expectations. Alterations in the elements of the ratio will lead to individual frustration; widespread alterations will lead to social stress. Social stress can be defined as the result of interrupting the mutual reinforcem.ent between distribution and norni structure. Group response to social stress is the simultaneous response of individuals follo-.ving existing patteiTis of orientation for or against chajige. 49 Robert K. Merton, So cial Theory and S o cial Structure (Glencoe, 111.: The Free Press of Glencoe, 1949), 125-149. 50 Balance here refers not only to tl.'e elements of i'>!erton's paradigm, but to the preceding discussion of the relationship between available life chances and the normative justification for the current distribution. 51 Daniel Lemer, "Toward a Communications Theory cf Modernization," in Lucian W. Pye (ed) , Commun i cations and Pol itical Dev elopment (Princeton: Princeton University Press, 1965), 333-3:^5.

PAGE 31

21 Generally, the response to increased social stress tends to be increased opportunity, due to a lack of any guidelines which could dictate a ne;; norm justification for the alternate distribution, nssentially, the response to social stress is increased social inoh.ility, and social mobility is the redistribution of life chances. As the old norm structure breaks down there is less support for traditional disabilities and so the tendency is toward greater achievement orientation. There is no expected shift in values toward a more favorable view of mobility, rather there is simply no effective traditional inhibition on the acliievement motivation in individuals. System accommodation to stress may be supported by elements of the traditional norm structure. There are, of course, many forms of mobility. Some forms of social mobility may be accepted, accommodative 52 modes. Either individual or group mobility may be valued. If mobility is accepted, then stress may be relieved by a partial redistribution of life chances imder tlie aegis of tradition. Aspirant individuals may be relieved of traditional disabilities and acconmodated within traditionally supported roles with concomitant sanctions. ^^ An alternative form of 54 mobility is the increase of sanctions for previously non-valued roles. 52 Neil J. Sm.elser and Se)TTiour M. Lipset, "Social Structure, Mobility and Development," in Smelser and Lipset (eds) , Social Structur e and Mobility in Economic Dev elopment (Chicago: Aldine Publishing Company, 1966), 1-50. 53 Ibid ., 7. 54 Ibid. , 7.

PAGE 32

22 More, the individual remains in the same roles, but increases life clicinces by virtue oJi increased sanctions. Or, lastly, neith.er the roles nor the sanctions may be traditional and tlie individuals occupying the new roles will attempt to insure high status for tlicm. Tlie latter case provides tlie most interesting form of mobility for political theorists. If neither the new role structure, nor the new distribution of sanctions has general normative support, then purposive system action is necessary to accommodate social stress. Purposive action is required because neither the roles nor the sanctions have normative support, and , the re fore, there is no traditional justification for the accommodative distribution. Miether or not this action occurs will depend on the rate of change in tlie social stratification system. Slow change may allow adaptation by the system even without traditional support. The more rapid the rate of change, the less likely social accommodation will become. Social acceptance of such a new distribution of life chances may result from a generalized belief in its inlierent rectitude, or from tlie futility Of opposing the forces supporting it. Social change and/or progress is rationally ordered only in part. The bulk results from the interaction of many individuals engaged in activities oriented toward personal rather than community goals. This does not mean that change is not explainable or predictable -it is. But, it must be kept clearly in mind that the paths discussed are merely some of several alternatives, and tb.is explanation attempts only to make clear why one path eventuated rather than another. Social change is deteriiiinis tic only within the limits of the antecedent conditions tJie theory will specify.

PAGE 33

23 Although social systems are not pui-posive entities, obviously a social group inay exercise sufficient influence to affect the behavior patterns of most individuals in a society. A group may force general acceptance of a new distribution, thereby creating a seiablance of purposeful activity by a non-pui-posive entity. In summary, social mobility may be sanctioned by traditional norms and values or may occur so slowly that it does not cause significant social stress; in either case it is not of particular interest to a political scientist. Our intei'est focuses on social mobility which is not sanctioned by traditional noniis and values, and v,'liich occurs rapidly enough to out-pace evolutionary accommodation. A political act will be defined as -the attempted or actual use of the autliority of the stat e to redistribute the li fe chan ce matrix of a society in an innovative manner . This definition has general significance since it focuses on the redistribution of life chances, and therefore will lead us to consider the effect of politics on all members of the society, ll^e more rapidly a system, is changing, the more necessaiy political activity becomes, since the re-institutionalization of normative suppom; for a distribution becomes less likely as change occurs more rapidly. There is absolutely no provision in this definition for an ' expectation of either stability, equilibrium or su ccess . It is only suggested tiiat political activity becomes necessaiy in an attempt to accommodate to certain forms of social change, 'ihere is no theoretical basis for expecting this attempt to be successful in terms of contributing to the reduction of stress.

PAGE 34

24 In this definition, the state is undei'stood to have geographically defined jurisdiction. Any set of individuals \vho claim binding jurisdiction over decision making for all individuals witliin a geographically delimited area, by virtue of their presence in that area, v.-ill be considered a state. Only de facto jurisdiction is relevant; either the claimants exercise effective decision making capacity or they do not constitute a state. Political activity will be sub-classified as either assumptive or expansive. Assumptive political activity is an attempt to assume distributive authority over a new area of life chances. For example, the original struggle over whether the state would assujne responsibility for education constitutes assumptive political activity. An entirely new range of life chances would be affected by the state's claim over providing educational opportunity. Expansive political activit)', on the other liand, is the extension of previously accepted functions of redistribution to larger subsets of the society. (Continuing the example of education, expansive political activity would be an attempt to include m.ore individuals within the state's education system. In most societies, routine political activity will center on expansive politics. It is, by its very nature, amenable to the Wildavskia^n treatment of incrementalisra. " 15 Otto A. Davis, M.A.H. Dempster and Aaron Wildavsky, "A Theory of the Budgetary Process," Ameri c an Political Sc ience Review (September, 1966), 529-547.

PAGE 35

25 Increasing the scope of either assuraptive or expajisive politics i;iil increase the proportion of available life chances which the state is attenrpting to redistribute. An increase in the scope of life chances redistributed by political system activity will be termed political permeance growtli. Permeance growth is an obviously significant element of political development. A political system is of conserpaence to the degree that it affects the life chances of the average individual in the society. One element of system consequence is the range of life chances over wliich the state has assumed authority. From, tlie definition of politics it should be very clear that tlie vast majority of activities undertaken by the state are not political. Most state activities simply support the o<.itcom.es of previous political acts. The level of earlier permeance growth may be measured by the current level of government activity, for the range of activities uiidertakcn by tlie state as a result of previous political acts tends to accrete like a palimpsest, layer upon layer. Tlicre is no inherent expectation in this definition of politics that political activity -./ill tend toward greater equality of opportunity. If the redistribution of life chances does result in a general tendency toward gi eater equality of opportunity, then this ten;iency will be termed congruent growth. This tendency toward equality is uiiderstood as 56 For an excellent essay on the concept of gi'ovjth, see Karl De Schweinitz, Jr., "Growth, Development ajid Political Modernization," '\'orld Pol it ics (July, 1970), 513-S40.

PAGE 36

26 growth simply because equality of oppovtuuit)' is seen as socially efficient. It is assumed t'nat tlie most socially efficient utilization of liuman ability is to be found in the achievement oriented society. Congruent growth, therefore, employs a Utopian referent; absolute equality of opportunity. Thus, as traditional disabilities are removed, the society more closely approximates the intelligent utilization of individual talent. Simultaneous growth on both the pex-neance and congruence dimensions will be termed political development. Po litical development is the activity of the st ate as it insures equ ality o f opportunity over ever larger ranges o f life chances . This definition of political development follows from the earlier definition of political activity. Political activity is a rcbponse to patterns of social change which have bred social stress. If the patterns of behavior underti''ls'on in response to that cliange do not relieve tlie original stress, this behavior crnnot be considered as conducive to further development. A good example of inappropriate behavior patterns is the symbol manipulation of the elites of many "developing" nations. Airlines, highways and modern armies, all symbols of modern industrial societies, are resorted to in S7 response to the stresses found in the new nation. These m.anipulations do not deal with the root causes of stress; rather like tlie individual 57 Herbert Feith, "Indonesia's Political Symbols and Their Wielders," '.far Id Politics (October, 1963), 79-97.

PAGE 37

27 neurotic, they represent a retreat into -compulsive symbolic activity to relieve anxiety. If political activity results from stress, then the more widespread tlie stress, the more widespread will be the atteripts to cope with it. Hence, political permeance growth flows directly from social change. On the other hand, congruence growth does not follow directly from social cliange. Congruence growth is deemed necessai-y to develop ment because only by increasing equality of opportunity can there be effective relief of stress. Since stress results fi'om disruption of the ratio of distribution to norm support, permeance gro\rth alone is not sufficient to reduce stress. Since re-institutionalization of normative supports for a limited distribution is also unlikely during rapid change, the only consistently effective stress reducing alternative is toward equality of opportunity. In a modern industrial society, or in any rapidly changing society, the individual will liave no socialized limitiation on goal expectations. Any socially supported disability of maans will perpetuate stress. Thus, permeance growth which does not tend toward equality of opportunity will not only not relieve stress, it will pei^petuatc stress. It is by now obvious that political developmient cannot occur without social development. By social development we mean an expansion of the total range of available life chances in a society, both roles and sanctions. Social development may, for a time, occur without political development, but not vice versa.

PAGE 38

28 V.'e have now developed a linkage. Social change is sccia.l mobility, widespi'ead social mobility is social mobilization, and political development is inextricably linked to social change. Our concern at this poiiit is to explain the process by \chich social change is generalized and accelerated, and to explain why and how different races of social change lead to different forms of political change.

PAGE 39

CHAPTER II SOCIAL CHANGE AND THE INSATIABLE PERSONALITY Deutsch-*and othei's^ make the implicit assumption that social mobilization has some positive end product. Deutsch defines social mobilization as ^'Vao process in which major clusters of old social, economic and psychological commitments are eroded or broken and people become available for new patterns of socialization and behavior."^ Yet, the socialization of new values need have no particular significance. In the Deutsch argmne.nt, these new values are supposedly more "modern" and, therefore, more conducive to continued social change. However, t'ne process of mobilization is significant, not because it replaces traditional norms and values with new norms -iiid values, but rather because the essence of the proce ss consists of destroyi ng certain specific, yet apparently uni v ersal elements of th e 1 Karl W. Deutsch, "Social Mobilization and Political Development," M.erican Political Science Review (September, 1961), 493-514. Bert F. Hoselitz, "Economic Growth and Development: Non-Economic Factors in Economic Development," ^toerican Ec onom ic Review (May, 1957) ^ 2841; and Manning Nash, Primiti ve and Pea san t Eco nomic Systems (San Francisco: ChandlePublishing Company, 1966), 131-135. 3 Deutsch, 0£. cit. , 494.

PAGE 40

30 traditional a orm st ruct:ure an d never replacing tliem.'^ In discussing past periods of great economic expansion and cultural prominence, ilagen makes the simplistic, but sigiiificant observation that they have all come to an end. In explaining the current development of societies, however, he is faced vitli a long standing period of progress which has not ended, and perliaps may never end, Hagen searclies for a positive form of personality type, a "modern" man who is in some way responsible for this sustained growth. .\n alternative explanation is that some particular set of characteristics exists in traditional society placing limitations on social cliange. Here, the elimination of certain strictures, rather than the creation of a particular personality t>'pe, produces the original climate for sustained growth. Altiiough a new form of personality type is produced, this personality type does r^ot have any predisposition toward positive new modes of behavior. Rather, the behaviors engaged in by this individual will be largely dependent on his current social situation, llie individual's attempt to funiition in an alien social matrix may, in fact, lead him to attempt to define his identity in socially dysfunctional 4 For a brief discussion of the social disorganization wrought by the industrialization of a traditional society, see George A. Theodorson, "Acceptance of Industrialization and Its Attendant Consequences for the Social Patterns of Non-Western Societies," T^jner ican Sociological Review (October, 1953), 477-484. 5 R.H. Hagen, On the The ory of Soc isl Chang e (Homewood, 111.: Tlie Dorsey Press, 1962). 6 Hagen describes the developmentally conducive individual as the "innovative personality."

PAGE 41

31 laodes of behavior. Aside fro;;i occasional bursts of development, most societies are marked by very low rates of social change. Tlirough most of history, this sub-culture of tlie peasantry has been the dominant culture of society. Rogers has developed a list of attributes which apply to the sub-culture of peasantry. ^ In traditional peasant societies the individual functions in a v.'orld frame of "perceived limited good." In Rogers' argument, the individual assumes that there is an absolute limit on the quantity of valued good. 'Fliis perception is quite realistic for tlie average peasant. Most traditional societies based on peasant agriculture deal with fundamentally inelastic supplies of valued goods. Botli tlie quantity and variety of life chances and life styles are hlglily restricted ajid the distribution of available life chances is legitimized by tradition and religious sanctions. Alth;ugh Rogers sees "perceived limited good" as merely one elem.ent of a peasant sub7 Everett M. Rogers, Moder n ization Among Peasants: The Impa ct of Communicatio n (New York: Holt, Rinehart and Winston, Inc., 196!J) ,24-S9. 8 According to Rogers, "Central elements in this subculture of peasantry are: (1) mutual distrust in interpersonal relations; (2) perceived limited good; (3) dependence on and hostility toivard government authority; [4) familism; (5) lack of inncvativeness; (6) fatalism; (7) limited aspirations; (8) lack of deferred gratification; (9) limited vic\; of the world; (10) low empathy." 9 See for instance, Gerliard Lenski, Power and Privi lege: A Tl-ifcor y of Socinl Stratification (L'ew York: McGrau--MiH Book Goiap any, 1966); ond Karl De Schweinitz, "Growth, DevelopjT:ont and Political Modernizarion," '£o2.J:d Politics (July, 1970), Si3-S40. 10" Rogers, o£. cit . , 25.

PAGE 42

32 of the proposed sub-culture, "per ceived limi t ed go od" is the cen tral factor defining peasant culture. It should not be surprising that social mobility is very low in a society where it is unrealistic to expect advancement and impious to want it. Lenski has argued tliat "the fact is, unhappily, that in the long run, in all of tliese societies, downv/ard m.obility was much more frequent than upward. "-^^ Tlie other elements of the peasant sub-culture, fatalism, amoral familism, disti-ust of others, all are closely related to these two factors: the limited range of available good and tlie perpetually sanctioned inequality of distribution. In a peasant society an individual's personal identity is largely defined by his position at birth. The development of his self-concept consists of developing "subjective public identities" which are products of the roles he occupies. With the stratification system rigidly restricting role assignment, the individual's identity definition is also restricted. If nothing else, the traditionally defined identity of the peasant leaves him secure in liis disadvantage. ' The traditional individual knows who he is and what is expected of him merely on the basis of his existence. This foi-m of identity definition will be called traditional identity. Under the norm structures of perceived limited good, the vast majority of any populatioii experiences traditional identity. If perceived limited good is the central ele;nent of a peasant 11 Gerhard E. Lenski, Pow er and Privilege: A Theory of Social StratJ fication (New York: McGraw-Hill Book Company, 1966).

PAGE 43

33 culture, u'e v.oulJ expect to find a fatalistic, iiward turned individual with little capacity to visualize alternative life styles simply because they are unrealistic. V/e v/ould expect this individual to be primarily a faiTiilist, if not an amoral famllist, for in the struggle for survival, fcv; others are to be trusted. Finally, vre would expect this individual to be highly traditionalistic, not only because tradition is the basis for his existence, but because sameness tends to reinforce itself. It is reasonable, therefore, to accept that the peasant has, in fact, managed to adapt realistically to his environmeiit . Despite his "ignorance," tlie culture of the traditional peasant is a realistic reflection of tlie central fact of pliysical existence; good is, in reality, limited, and the peasant culture, whatever its ethnic variations, has the same central elements as a reflection of the restricted range of life chances. If a restricted range of life chcinces results in traditional stasis, then change and growth could be based on increasing the range of life chances. The norm structure of perceived limited good may be disturbed in one of two ways: an alteration in the range of available life chances, necessitating redistribution of tiie social matrix; or some disturbance which destroys the old norm structure. Before the industrial revolution, such redistributions were most frequently caused by war and trade expansion, both of wliich acted against each element of the social ratio -distribution and justification. After such a period of expansion, there would necessarily be a re-traditionali/:at:ion of the norm structure cf perceived limit 3:d good simply because gccd, even between

PAGE 44

34 societies J was limited. Most brilliant periods of hiunan culture closely accompany periods of conquest or trade expansion. General social development, on the other hand, tended to be glacial until the onslaught of the industrial revolution. Under the impact of industry there has been an almost continuous expansion in the range of life chances with a corresponding shift from traditional to situational identity. Modernisation is the reinforcing interaction of social mobilization and economic growth, destroying the norm structure of perceived limited good for progressively m.ore and more individuals. This process also destroys the individual's social identity. For the peasant, social identity is closely linked to the stratification system which is, in turn, a product of perceived limited good. Tlie difference between situational and traditional identity lies in the rate of social change. Viliile the social situation is always changing, for the peasant the change is so slow as to be almost imperceptible. For an individual in a rapidly changing society, on the other hand, the elements of his identity definition are changing at a very rapid rate; lus identity is no longer secure because his situation is no longer secure. Tlie socially mobilized individual lias neither a secure identity, nor the limitation on goal expectations which flow fi-om the peasant's perception of limited good. It is this social situation which gives rise to the "modern" man; an individual I call tlie Insatiable Personality, llie Insatiable Personality's primary characteristic is non-directed striving, for only by striving can he develop a sense of selfness. He is, at once.

PAGE 45

35 Hagen's innovative personality, ^2 McClelland' s high n-achiever,^^ Lerner's high empathy r.ype,^^ ^^^^ p^.^.g confused identity. 1^' ^^ He has high empathy because the mobilization process has destroyed the blinders of social strictures. He is innovative and achievement oriented because he has no limitation on expectations. He is a confused identity because his social referents are in constant flux -the more he succeeds the more he has to strive. With no limitation on good expectations, his reference groups and standards constantly rise as each new goal is reached. 12 Hag en, oj^. cit. ^ David McClelland, Tlie Achievi n g Society (Princeton: D. Van Nostrand Company, Inc., 1961). 14 Daniel Lemer, The Passing of Traditiona l^Socrety^ flodgrnj^u^^ Middle East (New York: The Free Press, 1958). 15 ~ Lucian W. Pye , Politics, Personality and Na t ion-Building: B urma_^ Search for Identity (NeuHlaven: Yale University Press, 1962). Hac^en uses status -deprived groups such as tlie Antioquenos of Latin America and the Samurai of Japan as examples of highly innovative groups. Yet, all that is unique about these groups is that they were mobilized at an earlier date than other members of their societv. By virtue of tlieir m>obilization, tliey had no traditional role to which they were still comjnitted, and they were, therefore, generallv willing to take advantage of new situations as they arose. The important point here is simply that tliese individuals din not create the situations from which they profited. Their innovativeness Is-passive not active. Lerner's concept of empathy also is generally passive not active. This individual is able to consider new modes of behavior, but he does not necessarily seek them out. McClelland's high n-achievcr, on i.he other hand, is aggressive in his striving and he comes closer to the economically modernizing individualwe seek. Yet, McClelland's high n-achiever is willing to strive m whatever social context he finds liimself; he is success oriented, but not necessarily innovative.

PAGE 46

36 The Insatiable Personality exists in a social matrix wliich places no normative or value limits on wliat he may expect from life. His identity is insecure because it is defined by comparison; he sees himself as he believes others see him. He strives constantly in an attempt to obtain a more gratifying self-perception. '^ Ifliere one peasant may be envious but not threatened by anotlicr peasant's bigger harvest, one machinist will be both envious and threatened if another machinist receives a raise. In both cases, identities are defined relative to other people, but in the peasant case the relation; are fixed, while for the modern man they are variable and insecure. 17 There is an interesting parallel here between Lenski's discussion of institutional evolution of stratification systems, the cultural perception of limited good (o£. ci t_. ) , and Veblen's discussion of conspicuous consumption as a cultural artifact (Tliorstein Viiblen, The The ory of the Leisure Class (New York: Viking Press, 1912), 60-70). At both ends of Lenski's spectrum, hunting and gathering societies and mature industrial societies, uo find relatively loose stratification systems. Veblen's discussion of conspicuous consumption suggests that in primitive societies and industrial societies, conspicuous consumption is a right, whereas in traditional societies (Lenski's agrarian societies) it is a privilege. In both forms of society where conspicuous consumption is a cultural artifact of the mass, we find loose stratification systems and no perception of limited good. Given the absence of limited good and the flexibility of the stratification systems, conspicuous consumption is an obvious and probably necessary means of identity definition. V^blen argues that conspicuous consumption is a means of maintaining reputation, hence a primary component of individual identity. Reisman's discussion of tlie "other-directed" individual in modem society (David Re is man. The Lonel y Crot\fd: A Study o f the Changin g /oriorican Character (New Haven: Yale University Press, 1964)) parallels the discussion of conspicuous consumption as a ineans of reputation building. Tne other-directed individual is oriented to identity definition by continuous comparison of outward cues.

PAGE 47

57 In terms of individual economic behavior, modernization is, in large degree, a potluck process. Not all Antiqueiios, all Samurai, or all Puritans were mcdernizers. ^^ However, if enough individuals attain the characteristics of the Insatiable Personality, then it is probable that som.e of them will engage in modernizing behavior. Including a large number of Insatiable Personalities in a society is a necessary, but not sufficient, condition for modeniization; neither does the mobilization of a group guarantee their future high ecoacnic activity. As more and more individuals are socially mobilized, the process becomes largely self-sustaining. Once this point is reached, the key factor is no longer the destruction of the norm structure of perceived limited good, but its general absence. An individual socialized in modem industrial society automcitically defines his identity by comparison and his self-esteem by striving. In terms of sustained social change, there is a continuous interaction between the individual ar.d society so that the individual may, at the same time, be both cause and effect. By his very existence, the Insatiable Personality creates an unstable social situation for other individuals since he is, for them, part of society. Tne significance of the Insatiable Personality as a conceptual tool in explaining social and political change lies in its ability to help us understand, the generalized striving of mobilized individuals. 18 Ilagen, cvp_. cit.

PAGE 48

38 Nash relates an interesting anecdote about New Guinea tribesmen presented with metal knives. The resulting increase in efficiency was expended; not in increased productivity, but in increased leisure. It is liard to imagine an Insatiable Personality not putting a comparable advantage to use to increase his productivity and thereby increase his status. Productivity here is not confined to simple economic increase, but to general efficiency in any line of endeavor. Tlie New Guinea natives were secure in tlieir current situation, and even presented with an aspect of modernity took a long time to use it. The limitations of the psychologically oriented theories I'esult partially from their attempt to identify a first cause for modernization. The first cause of moderni;:ation in any society is really unimportant. From the point of view of the mass of men, tlie process of social change is always one of diffusion. In examining a modem society there is little consequence in discovering what came first, the chicken or the egg -our concern is with understanding how we got a whole flock. Vvliile the pi'ocess of social m.obilization results in Insatiable Personalities, social mobilization cannot long continue without continued economic advancem.ent . As previously stated, through most of history the only way to increase wealth was by horizontal economic expansion. In terms of any given area of population there was no significant increase in the total quantity of valued goods. Horizontal expansion 19 Nash, on. cit., 42-57.

PAGE 49

39 led to a redistribution of v;calth which couid produce an increase in the total available life chances for ineinbers of one social system. liov;ever, since the total supply of goods was inelastic, there was necessary deprivation of members of other systems. Tliis inelasticity in agrarian economic orders was the primary basis for the re-traditionalization of the norm structure of perceived limited good, since all m.embcrs of competing societies were faced with what am.ounted to a zero-sum game. (It should be noted that the situation in the United States did not result in a zero-sum game due to the availability of unoccupied land.) Increasing wealth after the industrial revolution created a non-zero-sum game. Economic expaiision became a process of largely vertical growth, increasing not only the quantity of life chances in terms of real goods (growth), but the quantity of life chances in 20 terms of available roles (differentiation). Tlie special increase in wealth provided the necessary reinforcement for a continuous process of social mobilization. Since a continuous increase in the range of available life cliances existed, the Insatiable Personality 'v-as not faced with the reassertion of the social strictures resulting from the former inelasticity of valued goods. Tliere was the possibility for the reinforcing interaction through time between the Insatiable Fcrscnality 20 Wilbert E. Moore, "Changes in Occupational Structures," in Neil J. Smelser and Seymour M. Lipset (eds). So cial Structure and Mo bility in Economi c Pevel opme nt (Chicago: Aldine Publishing Company, 1965), 194248.

PAGE 50

40 and increasing life chances which ive have called modernization.. Social modernization is not a necessaiy result of either the increase in the number of Insatiable Personalities nor tlie inception of economic grovvth. Although aspiration is enough to ensure striving for a limited period of time, ignorance can provide as effective a deterrent to continued striving as did the formerly inelastic supply of valued goods. IVe have argued that the Insatiable Personality is willing to engage in new modes of behavior and to strive to achieve in order to secure his identity. Enabling liim to pursue achievement is a primary linkage between politics and society. We have defined politics as the attempted or actual use of the autliority of the state to redistribute tlie. life chance matrix of a society in an innovative manner. Social pressure for an increased ability to satisfy current aspirations will lead to pressure for greater educational opportunities, and in terms of increased permeance, government activity will proceed in a very special direction -toward increased education. 'Tliere is no expectation that either the pressure on the government, or its activity will be successful. Success (in the sense of increased education) will depend upon the rate of change in the society ajid the rate of demand with 'vhich the political situation can copc.^^ If the pressure for greater education results in higher levels of educational attainment, then the probability of 21 Later discussion will outline tlie effect on political development of very rapid rates of miobili zat Ion.

PAGE 51

41 continued modernization is very high. If the pressure tor greater education does not result in higher educational attainment, then the probability of continued modernization is aln-ost non-existent. The basic elements for an explanation of increasing political development have new been outlined. Modernization is an interaction between increasing wealth and increasing social mobilization; each factor is dependent on the other and neither can long continue without the other. Yet, the aspirations raised by social mobilization can only be satisfied, through time, by increased individual capability arising from increased educational opportunity. If the Insatiable Personality is not to experience system de-stabilizing frustration which will ultimately end the process of modernization, widespread education m.ust become normal. Since the state is the only agency which can supply educational opportunities on a mass basis, extreme permeance growth begins. Tlie state thus becomes intimately involved in the redistribution of life chances for the society, and by being involved, becomes an iiitegral part of the process of ' modernization. Political development is one possible result of modernization if the rates of change allow balanced growth. Imbalances in econoirdc growth or social m.obilization can result in hybrid forms of political organization which may or may not become developed in time. Our concern in the following pages will be to present a specific model of the ideal-typical form of development. To this point, our theoretical concern has centered upon the general case of transition from peasant to industrial society. Peasant

PAGE 52

42 societies were used as tlie base line to insure applicability to many societies. In the specific case of the United States, \v'e are examining the transition from a society based upon agrarian enterprize to one based on industrial production. Prior to the onslaught of massive industrialization, the process of froTitier expansion, indeed the very process of colonization, had increased the range of life chances available to the average individual. /American society was already experiencing the process of social development resultiiig fi'om increased wealth production, llie AjTierican was a farmer, not a peasant. Although increased life chances were available to the average Ajiicricaji as a result of the abundance and quality of available land, as in a peasant society, the means of wealth production were still based on agriculture. Despite the increased availability of wealth, the average American was still faced with a restricted set of alternative life styles. So long as the society was based on agricultural production, it was limited in its potential for social, and hence political, development. Since political and social development are processes, not events, there is no compelling reason to begin at the beginning in order to conduct valid tests of models of these processes. Accepting the preliminary changes in the social structuie which established America as a free agrarian, rather than peasant society, it is still expected that the transition from a rural-agrarian to an urban-industrial social order would be deeply traLur.atic. Although the theoretical argument for the experiences which result in t]ie Insatiable Personality remain the same, it must be acknowledged that the average

PAGE 53

43 Amei-ican's situation in .1870 '.-.as not as restricted as the situation of the average peasant. To cover as long a time period as possible ^ it would be best to commence the rescarcli at as early a date as possible. At tlie sam.e time, extension of the time period should not result in any loss of quality in the variables used, nor too sevea^e a loss in the number of available data points. Tlie specific time period of 1870 to 1960, was chosen for the research as a result of several considerations. Prior to 1850, data on newspaper consumption and literacy are not available on a state by state basis. Tlierefore, two of the major variables are unavailable prior to 1850. Comm.encing the research with 1850 would have included the disturbing effects of the Civil War. Further, since many of the states were not, in fact, states during the period of secession, the political participation index would be truncated. Consequently, it was decided that the earliest practical year to include ia the research v;as 1870. l-igure i presents the gross total pig-iron shipments in the United States between 1820 and 1960, by census years. Pig-iron is used as an index of basic industrial capacity since it provides the raw material for all advanced industrial processing of steel and steel related products. In addition, the basic metal used in the advancement of the rail network in the United States was forged iron. From Figure 1 it can be seen that tiirough 1370, little chanve had occurred in the level of industrial activity. By beginning with 1870, we will be examining the process of transition from .agrarian to industrial .society from its earliest stages.

PAGE 54

Legend for Figure 1 Production noted in thousands of short tons. Raw data from the United States Bureau of the Census, Historical Statistics of the United States j Colonial~Timcs to 1857 (Washington, dTc.: United States Government Printing Office, 1960), and Historical Statistics of t h e United State s, Cont i nuation to 1962 and Revisions (V'ashington, D.C.: United States Government Printing Office, 1965).

PAGE 55

Total Pig Iron Shipments in the United States: 1820 1960

PAGE 56

46 Our previous discussion suggests the following linkage for the theoretical elements: increasing wealth leads to iiicreasing social niobilization, which in turn aids in increasing wealth. If increasing social mobilization leads to increased individual adaptability, then social !;iobilization and wealth will likely continue increasing." Diagrarpjiatically, we may represent tlie simultaneous occurrence of these relationships by intercoiniected circularities. V/ealth"^ 'reduction. ^~^ Social Mobilization General Adaptability Figure 2 Causal Model of Three Concepts witii Simultaneous Interactions Considering the above with.in the framework of passing time makes the entire system clearer. For any single period, T, the linkages are shown in Figure 3. V.'ealth _ Production Soci al General Mobi lization Adaptability Figure 3 Causal Model of Tliree Concepts with One-U'ay Causation For a similar, but confusing model, see I'torman H. Nie, G. Bingham T'owell, Jr., and Kenneth Prewitt, "Social Structure and Political Farcicipation: Developmental Relationships, I r, II," American iJ::^!! -J-^'^] ^"^jg^tce Revie w (Jane and September, 1969), 361-378 and 808328.

PAGE 57

If we consider tliese same relationships for a period, T to T+1, .we Kould have tiie linkages shown in Figure 4. U'ealth _ Production Social ;!obilization Wealth Pro duct ionn V Social Mobil ization-p^j^ General Adaotability 'T+1 Figure 4 Causal Model of Three Concepts with Inter-Tenporal Effects General Adapt ability^P^I^ Through time, each variable's continued growth becomes a necessary precondition for the continued growth of the other variables in the system. Tills argument assumes that in past periods, of rapid social change, the process was interrupted by the inelasticity of available wealth. The ensuing re-institutionalization of the norm structure of perceived limited good led to the demise of further social mobilization and the reassertion of the traditional agrarian social order." In considering current modernization, wealth increases have become Lhe result of industrialization, and therefore have been relatively continuous. In current modernization, all the variables in the system have become more closely interdependent since wealth production is now a lughly complex social process. The social orgc^ni zation of industrial society requir widespread social mobilization in order to continue increased industry. 23 Lenski, og_. ci_t_. 275.

PAGE 58

Social i;iob i .1 i z.at ion can be broken down into two major, interrelated processes, urbanization and mass communication, each more specific and more easily operationalized. In the following discussion, urbanization will be considered only in terms of intense demographic concentration. Industrialization as it occurred under past levels of technology required extreme demographic concentrations. Large scale industrial production requires masses of workers concentrated in close proximity to the place of work, lliroughout the liistory of industrialism, tlie growth of cities has been the primary result of stripping the hmterlimd of agrarian population. Urbanization did not result merely from natural population increase in pre-existing cities. Indeed, as Deutsch^^ has noted, in all industrializing societies, new cities sprang up as new industries were form.ed. ^ Again, tliere is p. cycle; industrialization makes large scale urban populations possible, and increasing urban populations are necessary for further industrial 24 Rather than treat urbanization as a syndrome, it was considered wiser to examine the most basic characteristic of urban life and then test for associated relationships. 25 Karl W. Deutsch, Nationalism and Social Communication. An Inquiry into the F ound ations of Nationality (New York: Wiley, 1953). 26 One must, of course, also consider the effect of urbanization on peasants who are intermittently exposed to city life. The extraordinary urban slum sprawl of many cities of the third world is an apparent reflection of the effect of urban opulence on the unprepared rural population. Sjc also, Philip M. Ilauser, "The Social, Economic and Technological Probleias of Rapid Urbanization," in B.F. Hoselitz and V/.E. Moore (eds). Indus trializa t ion and S ocie ty (UNESCO: Mouton, 1963), 199-217.

PAGE 59

49 expansion. Tlie by-product of this process is an ever increasing quantity of Insatiable Personalities v.-ho, by their activities, will tend to reinforce both industrialization and urbanization. Urbanization necessitates tlie second process of social mobilization, mass communication. Mass communication is essentially the process of large scale coirjnunity integration. If cities were merely organized in small enclaves, industry by industry, then integra-cion \sould be less important. Ilov.-ever, even in a relatively small city, the extreme complexity of the logistical system makes widespread communication necessary. To create an urban society , mass conununication is an absolute essential. In earlier epochs, communication required literacy since the media were print. This has led some theorists to presume that literacy is necessary for mass communication. There is, in reality, no reason for mass literacy to precede mass communication. Prior to the development of electronic media, as long as the nodal personality in a sociometric network v.'as literate, the other persons in the net could effectively be considered 'part of the mass communications system. A society cannot support urbanization without the community integration carried on by the processes of mass coirmiunication. Demographic concentration can continue for some time, but the process of modernization will break down under the pressure and confusion of a nonintegrated environment. A cycle is again apparent. Industrialization leads to urbanization which in turn is necessary for further industrialization. Yet, continued urbanization is not possible without the integrating fur;cticns of mass communication. Diagramatically, in terms of two time periods, the relatio:

PAGE 60

50 are shown in l-Mgure 5. Indus t ri al i z at ion V Urb an i z at i on T ^ T ^ Conununication Industrialization, T+1 Urbanization >" Communication,, Figure 5 Inter-Temporal Effects of Three Operational Indicators r+ 1 Mass communication has a significant impact on the development of the Insatiable Personality. Uliere urbanization acts as the first stage of social mobilization, mass communication acts as the second stage by increasing the individual's aspirations, making him aware of currently available alternative life styles. At this juncture, social stress becomes reflected in the political sphere. Ay^pirant individuals are faced with the unpleasant realization that their abilities are not comjp.ensurate with their aspirations. It has been argued that social mobilization destroys tlie individual's links with liis pievious social position and identity. In order to exist in his new situation, the individual must find some new form of identity in which to secure himself. He must, in short, take on new roles. Yet, his ability to fill new roles is contingent upon his level of general adaptability. Education m.ay be learning through experience, or formal classroom schooling. If the individual is educated through experience, it will be a process of trial and error.

PAGE 61

Repeated exposure to ir.odem life styles may simply inure the individual to the harassinents of modernity. If the rate of mobilization is slow, then learning from experience will be sufficient to perpetuate modernization for a time. Hie more rapid the original rate of mobilization, the more necessary an efficient form of education will be. Insecurity and maladjustment are conmion products of repeated failures to cope in the face of an incomprehensible social order. As the nuFiber of maladjusted individuals increases, the ability of the remaining mobilized individuals to sustain the process of modernization decreases. Hence, the process of social mobilization creates a situation in wliich education becomes increasingly necessary. In terms of. an entire society, mass fox-mal schooling becouios necessary if tlie process of modernization is to continue. Diagrammatical ly, the relations explicated so far are shown in Figure 6, first for one time period, T, and then for a time span, T to T+1. Fo r Time T Industrialization >• Urbanization > Comjnunication -^ Education F or Time T+1 Industrialization ^o Urbanization -> Communication -5.. Education T ^ T Industrialization > Urbanization ^ Comiiiuni cation > Education T+1 T+1 T+1 T+1 Figure 6 Inter-Tem.poral Effects of Four Socio-Econcmic Indicators

PAGE 62

52 While iiici-easing communication produces the need for increased ediication at a given time, T, continued increase in the level of communication at time, T+1, is contingent upon prior increases in education Formal schooling provides the only method of increasing individual adaptability in the mass of the mobilized populace. Tlie necessity for increasing the level of individual adaptability generates the first extreme permeance growth of the state. Government activity is obviously necessary to the process of social modernization, yet it is not a necessary link in the chain of causality. Mass education is a social service which has no immediate return. Govennnent activity is thus a necessary intervening variable betv;een the pressures resulting from mass mobilization and the increased need for education. If it were possible to increase the level of mass education 'vitliout government intervention, then the process of modernization would continue v.-itliout the state. It is precisely the process of m.odernization which renders non-political solutions to the problem impossible. As the rate of social change increases, political activity becomes more and more necessary. These relations can be diagrammatically represented by Figure 7, for any time, T. Industrialization I!edia Consumption _____ — / Education. IJrbani?.ation Government Activity Figure 7 Causal Relations of Four Variables and Gcvornment Activity

PAGE 63

It is possible, to extrapolate this argument one step further. Having acquired some measure of education, the average individual will 27 be more willing to engage in political activity. In a rapidly changing society many issues will have become politicized. The average individual's interest in political activity will vary with the number of issues which he perceives as affecting him. Previous exposure to the activities of mass media and government will arouse his awareness of the importance of political participation. In general, however, the tendency will be against attempting to participate until a certain measure of self-confidence is attained. For the average citizen, tliat self-confidence will be the result of the process of education. A generalized disposition toward atter.ipting to affect one's fate will be the result of education. Tlie actual opportunity I'or participation will be determined by the rate of social change which the exta);t political institutions can acconmodate. Presumably, if the current political institutions have been able to provide a satisfactory level of education, then the probability is high that they will be able to accommodate mass political participation. In other words, it is expected that if the developmental process continues through the level of creating widespread education, then increased political participation will follov; as a matter of course. The expected final model will follow the path indicated in Figure 8. 27 For the general evidence already relating education to participation see, Lester W. Milbrath, P o 1 i t i c a 1 P a r t i c i p a 1 1 on . Ijow_an_d_Miy_Do_ p£^^RlP-gp>-J^^'^^^'^^ ^'"^ P o'^itics? (Chicago r Rand, McMally and Company, 1965),^ 48-89.

PAGE 64

54 Industrialisation '>!edia roiisun-.ption rinal Representation of a Six-Variable ik)del of Political Development The ensuing discussion will describe the metaodolog)'' and operational indicators used to examine the adequacy of c"''ie above model.

PAGE 65

CH.^PTER III INDICES AND METHODOLOGY In the preceding chapters, some objections were raised to the metliods used to test models of development. One of those objections referred to the cross-sectional nature of the tests. Development is a dynamic process, and it cannot be tested by an examination confined to one point in time. In fact, it is all too easy to come to false conclusions about a concept simply from examining its behavior at only one point in time. As an example, let us examine the relationship between the concept of inter-party competition iind education. Interparty competition is used as an example because it was not included in the model of development. Table I presents a list of correlations between an indicator of inter-party competition and the level of education in tlie United States, by state, from 1870 to 1960. Table I Decennial Correlations of Inter-Party Competition and Education^'^ Year 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 IPC with -.20 .52 .33 .A4 .76 .06 .64 .71 .25 -.05 education The indicator of inter-party competition used was tlie losing party's percentage of the wiiming party's vote, b Raw data from Sven Petersen, A Statistical Hist ory of the Amer ican Pre siden tial Elections (New Y'oik : Ungir'Tub 1 i sh e rs , 1963), ar/d U'.S. Bureau of the Census, U.S. C ensus of Population CWashington, D.C.: U.S. Government Printing Off iceV 1872, 1883, 1S95, 1901, 1913, 1922, 1933, 1943, 1952, 1962). 55

PAGE 66

56 It is obvious that obsei'ving tliis correlation at only one point in time cou] d lead to erroneous conclusions. Observing the i-olation in only 1940 would lead us to conclude that a very high relationship exists betv.een education and inter-party competition. It", liowever, we confined our observation to 1920, '\'e would conclude the exact opposite. In o r der for a generalization to be valid, it mu st hold for more than one point in time. To avoid falling prey to this same criticism, the current research uses data drawn from a period from 1870 to 1960. Observations were made at ten year intervals beginning in 1870. Tlie ten year interval was used primarily for th.e sake of convenience. Data for the socio-demographic variables were readily available at the ten year intervals of the census. It would have been possible to generate the necessaiy data for shorter intervals by interpolation. However, since interpolated data would have been based on the decennial census figures, it was felt tliat little would be gained by increasing the number of observation points. A second objection raised against previous woi'k was the use of cross-societal samples as a data base for creating intra-societal generalizations. Tliis research used the United States, by state, as a data base. Table II presents the number of data points used for each 1 United States Bureau of the Census, U.S. Census o f Population riVashington, D.C.: U.S. Government Printing Office, 1972, 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962).

PAGE 67

57 census period. Table II Number of Observations for Each Decennial Census Period Year 1S70 1880 1390 1900 1910 1920 1950 1940 1950 1960 38 44 45 48 48 48 48 48 A state became a data point if, at the time of the census, it >vas legally a state. During the 1870's, several of the southern states h.ad still not been readmitted to the Union. Although tlie sociodemographic data were available for these states, no election figui'es '.s'ere available, and, therefore, tl^e last stages of the model could not be tested. The same situation prevailed in regard to the various territories which had not yet been admitted to the Union. Data Sources All socio-demographic data used in this analysis were taken froiii official government publications. The primary sources of infor2 mation \;ere the various volumes of the Census of the United Sta t es. At times, certain series of statistics were more readily available in 2 Ibid.

PAGE 68

58 the United Sta tes Statistical Abstr act. Uliile the figures in the abstract arc generally based en the data reported in the Census, the abstract figures are often presented in a inore usable form, especially in r-egard to state summary totals. In certain cases, notably early school attendance figures, it v;as necessary to use special reports. Data on school attendance were retrieved from the R eport of the United States Cor .nn issioner of liducation . Figures for postal receipts, by state, v.'ere obtained from the Report of th e Postmaster General . Beginning in ISSO, the Bureau of the Census was directed to collect figures on social statistics, and there are data on literacy, newspaper circulation, church attendance, etc. starting with 1850. Oddly enough, the figures on Presid-ential and Congressional elections were not available en a state by state basis from official publications before 1920. Data for the Presidential elections were drawn fi'ora Petersen, A Statistical History of the .'\merican P residential Elect ions , and were checked against the figures presented by '.'/.D. 3 U.S. Bureau of the Census, Statistical Ab stract of the United States_ (Washington, D.C.: U.S. Government Printing Office, 1883, 1893,' 1903, 1915, 1923, 1933, 1943, 1953, 1965). 4 U.S. Comn.issioner of Education, Annual Report of th e Co inmission er of Education (Washington, D.C.: U.S. Government Printing Office, 1871, 1S81).' 5 U.S. Postmaster General, Aimual Rej^ort of th e Postmaster General (Washington, D.C.: U.S. GoveiTnneirt Printing Office, 1871, 1881). 6 Svcn Petersen, A St at istica l History of the Arrierican Presiden tial Elections (New York: Ungar Publishers, 1963).

PAGE 69

59 rely on newspaper accounts of the various elections in those states 'vhere no special totals v.'cre kept by a state agency. Data for the Congressional elections v.'ere taken from an archival record supplied by o the Inter-lJniversity Consortium for Political Research. The tape presented only congressional district totals, but for the purposes of this research the district totals were aggregated to the state level. Variables A series of fourteen socio-demographic variables v.-as collected for the entire period. This series included data on governnent einplo>':iient and government cxpcnditutes v.liich were available from census reports. Table III pi'esents a list of the variables originally selected for examination as possible operational indicators of the six major variables: industr ialization, ui'ban izati on, educati on, media consum.ption , government act ivity aiid po litical part ic ipation . First it was necessary to select a series of variables, appropriate as indicators, both in terms of variable behavior through time, and in terms of the relationships among the several variables. Second, it was 7 iVi'.lter D. Burnham, Presid ential Ballots; L83&-iS92 (Baltimore: Jolms Hopkins University Press ^ 1955) ." 8 Tnter-University Consortium for Political Research. Historical Data Archives, Sta tistics of the American Congressi ional El ections, 1324-1963 (CojppuTer te;pe) .

PAGE 70

60 jiccessary to clioose either a single operational indicator for each concept, or to create an index of the concept. Table III Original Variables Selected as Possible Indicators 1. Per capita value added from manufacturing 2. Percent of the total population resident in cities greater than 2,500 3. Percent of the total population resident in cities greater than 25,000 4. Average daily newspaper circulation per 1,000 population 5. Telephones per 1,000 population''^ 6. Radios per 1,000 population 7. Postal receipts per capita 8. Total horsepower in use in manufacturing per capita 9. Percent of the working age population in manufacturing 10. Percent of the population who are literate 11. Percent of the school age population in elem.entary school 12. Percent of the school age population in high school 13. Percent of the age group 19-24 enrolled in college*' 14. Total state and local government expenditures per capita 15. State and local government employees as a percent of working age population 16. Turnout in Presidential elections as a percent of population 17. Turnout in Congressional (off-year) elections as a percent of population ^Available from. 1880 on b /Available from 1930 on •^Available from 1910 on

PAGE 71

61 Of the seventeen variables presented in Table III, four were found unsatisfactory and were not included in the analysis. Two variables, government emp loyees and eleinent ary school attendance , were discarded due to the extremely erratic nature of their nam values through time. Both of these variables were anticipated to demonstrate a relatively consistent growth pattern through time. However, this was not the case. The difficulty probably lies in the sources of the statistics. Tlie elementary school attendance figures were often incomplete during the early period covered in the research. Tlie employment figures were probably correct, but erratic due to constantly shifting definitions of the census category during the early period. Perhaps reflecting tlie stable nature of American industrialism, or more probably an artifact of the census categories, the percent of working age populatio n in manufacturing has remained remarkably steady over the past hundred years. If a variable does not change, it cannot bo exajnined. Since the mean value of this variable remained steady at around twelve percent' of the working age population, it could contribute no additional information to the analysis. Tlie fourth variable, horsepower J._n us e in manuf acturing, presented a different problem, also related to its variance. Tliere was no effective means of combining this variable with the other usable indicator of industrialization, value added per c apita. If the total horsepower were m.easured against a per capita base, then the value added variable, being measured in dollars, would dominate the index. If horsepower vrc-re measured against some other base, then it would most probably dominate

PAGE 72

62 9 tlic index. Since these tu'o variables v^e'^e highly correlated (0.S5), It was decided to use just one and not try to combine them. Since the primary theoretical concern is with industrialization as a source of wealth production, it was decided to use value added per capita as the ip.ajor indicator of industrialization. Figure 9 presents the mean values of the index of industrialization from 1870 to 1960. In this research, urbanization is viewed simply as a measure of demographic concentration, and does not represent any complex of expected associated traits. Two indicators, perce nt of p opula tion in cities over 2,500 and percent of pop ulation i n cities over 75,000 , were combined into an additive index of urbanization. Figure 10 presents the mean values for the two indicators from 1870 to 1960. It is necessary to refer to the concept of increasing scale to explain the manner in wliich several of the indices, including urb g n i z at i on , \-;c re constructed. At any one time, T, a variable will have some variance, S, Through time, the value of S may alter. Uliile we refer to the same conceptual .variable at many times, T, T+1,...T+N, the variance of that variable at different times may be different. If the variance of tlie variable changes in a steady and predictable manner due to increasing scale, it may be necessary to build into the index a meajis of compensating for tlie alteration in the variance of one of th sub-indicators. Specifically, if the component variables of the index have a naturally occurring saturation point, then it will bo necessary 9 For my objections to the use of standard scoring techniques of index construction, see the. Appendix,

PAGE 73

Legend for Figure 9 Units of measure in dollars Raw data from the U.S. Bureau of the Census, U.S . Census of Pop ulati on (U'asi i i n gt on , D . C . : U.S. Government Printing Office, 1872, 1883, 1895, 1901, 1913, 1922, 1953, 1943, 1952, and 1962), and the U.S. Bureau of the Census, Star tistical Abstrac t of the Unite d States^ (Washington , D.C.: U.S. Government Printing Office, 1883, 1893, 1903, 1913, 1923, 1933, 1943, 1953, and 1963). Means computed using Norman H. Nie et al . , Statistical Package for the SocialSciences '" "(New York: McGraw-Hill Book Company, 1970).

PAGE 74

64 ?. u 1 50 100. 50 70 80 50 1900 10 20 30 kO 50 Figure 9 Mean Per Capita Productivity from Manufacturing

PAGE 75

Legend for Figure 10 Percent of population in places over 2^500 population, 1870-1960. _ Percent of population in places over 25,000 population, 1870-1960. Raw data from the U.S. Bureau of the Census, Statis tical A bstract of the United States , 1S83, 1893, 1903, 1913, 1923, 1933, 1943, 1953, 1963; and U.S. Bureau of the Census, U.S. Census of Popu l ation , 1872, 1883, 1895, 1901, 1913, 1922, 1953, 1943, 19"52, 1962. Means computed using the S tatistical Package for the Social Sciences.

PAGE 76

66 167C Figure 10 leaa Percent cf Population in Cities over 2,500 and over 25,000 by Census Years

PAGE 77

67 to continue adding cc^'.icnent variables to maintain the variance of the index. For exar;:ple, -during the early phases of urbanization in a society, residence in a place cf more than 2,500 people may be an adequate index of arbmiization. Ilov/ever, as the percentage of the population resident in places of more than 2,500 approaches 100, then tlie index is no longer an adequate basis for differentiating what is urban. In short, the variable, through time, will cease to vary. Examine tlie data presented in Table IV, presenting the means and standard deviations of the tv;o componer.t variables of the index of urbanization. Table IV Means and Standard Deviations of Indicators of Urbanization by Census Years^'^ 1870

PAGE 78

68 Had we depended on tlie single variable, percent in cities over 2,500, as iin index of urbanization, our index would liavc begun to saturate, i'/e would have been led to infer that the importance of the conceptual variable, urbanization, was decreasing, wlien in fact \vhat was occurring was the saturation of a society where approximately the same proportion of the populace was resident in places over 2,500. In short, intertemporal research requires that operational indicators take account both of increasing scale and, thus, decreasing variance in component variables. The index of urbanization was created by summing the percent of t!ie population resident in cities greater than 2,500 and the percent of tlie population in cities greater tlum 25,000. Tlie majority of the variance of this index from 1870 to 1930 was contributed by the percent in cities greater than 2,500. Since 1930, the percent in cities greater tlum 25,000 ];as contributed the greater portion of the variance. It should be clear that to continue the research for a time period beyond 1960, a furtlier variable would have to be added to tlie index. Figure II presents the mean values of the index of urbanization from 1870 to 1960. An index of media cons'jinp'tion was created following clie same procedure as used for urbanization. For 1870, the index of media consumption was created by summing the per capita postal receipts and 10 Postal receipts per capita were used as an index of media consumption following the argument that postal correspondence is a measure of the level of community integration. Since our primaiy concern was with media consumption as a means of community integration as well as a means of ir.odei^n behavior pattern trajismission, the level of postal activity was included.

PAGE 79

Legend for Figure 11 Raw data from the U.S. Bureau of the Census, U_.S. Ce nsus of Population , 1872, 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962; and Bureau of the Census, Statisti c al Abstract o f the United States , 1883, 1893, 1903, 1913, 1923, 1933, 1943, 1953," 1963. Computations performed using the Statistical Package for the Social Sciences. "

PAGE 80

70 19 10 Figiu-e 11 Mean Values of the Index of Urbanization by Census Years

PAGE 81

1iie average daily nev.'spaper circulation per 1,000 population. Figures 12 and 13 present the mean values of the newspaper and postal receipts for 1870 to 1960. In 1S80, the number of telephones per 1,000 population were added to the newspaper and postal receipts to create the index. This same configuration was used for 1890, 1910 and 1920. Figvire 14 presents the mean values of the telephone variable from 1880 to 1960. Beginning in 1950, the number of radios per 1,000 population was added to the index of media consumption. Tlie composition of the index from 1930 to 1960 was the sum of the newspaper, postal receipt, telephone and radio variables. Figure 15 presents the mean values for the total index of media consumption from 1870 to 1960. Government activity was indexed by per capita state arid local government expenditures. It was unnecessary to use more than one variable for this index since it had no saturation point during the period under study. Table V presents the means and standard deviations of the index of government activity for 1870 to 1960. Figure 16 presents the mean values of the index for the same period. Table V Means and Standard Deviations (S.D.) of the Index of Government Activity by Census Years Year 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 Mean 6.5 5.9 9.1 14.5 15. S 33.9 74.0 92.9 69.4 330.7 S.D. 3.6 3.5 5.2 7.5 11.4 13.5 25.3 29.7 19.3 70.2

PAGE 82

Lescnd for Figure 12 Quantities in real numbers. Raw data from the U.S. Bureau of the Census, U.S. Ce ns us of Popu lation, 1872, 1883, 1895, 1901, 1913, 1922, 1955, 1945, 1952, 1962; and Bureau of the Census, Statistica l A bstract of the United States , 1885, 1895, 1905', 19r57~l9'23. 1963. 1933, 1943, 1953, Computations performed using the S tatistica l Package for the Social Sciences.

PAGE 83

9 19 1 FicTjre 12 Moan Daily Newspaper Circulation per 1,000 Population by Census Years

PAGE 84

Legend foi' Figure 13 Quantities in dollars. Raw data from the U.S. Bureau of the Census, Statistica l Abstract of the Unit ed States, 1883, 1893, 1903, 1913, 1923, 1933, 1943, 1953, 1963; and U.S. Postmaster General, Annual Re port o f the Postmaster General (Washington: U.S. Government Printing Office, 1871, 1881). Computations performed using the Statistical Package for the Social Sciences.

PAGE 85

75 1870 80 90 1900 10 20 30 4 50 6( Figure 13 ^Ieari Per Capita Postal Receipts by Census Years

PAGE 86

Legend for Figiu'e 14 Quantities in i.-eal numbers, Raw data from the U.S. Bureau of the Census, U.S. Census of Pop ulation, 1872, 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962; and Bureau of the Census, Statistical Abstract of the United States , 1883, 1893, 1903, 1913, 1923, 1933, 1943, 1953, 1963. Computations performed using the Statistical Package for the Social Sciences.

PAGE 87

77 Mean Nvmter of Telepliones per 1,000 Population bv Census Ycai:s

PAGE 88

Legend for Figure 15 Quantities in index numbers. Raw data from the U.S. Bureau of tlie Census, U.S. Census of Populatio n, 1872, 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962; and Bureau of the Census, Statistical Abstrac t of the Unite d State s, 1883, 1893, 1903, 19137"T923, 1933, 1943, 1953, 1963. Computations performed using the S tatistical Package for the Social Sciences.

PAGE 89

79 900 350 300' 250i 2 01370 80 90 1900 10 20 30 i* 50 Figure 15 Mean Values of the Index o? Media Cons uinpL ion by Census Years

PAGE 90

Legend for Figure 16 Quantities in dollars. Raw data from the U.S. Bureau of the Census, U.S. Census of Population , 1872. 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962; and Bureau of the Census, S tatistical Abstract of the United S tates, 1883, 1893, 1903, 1913, 1923, 1933, 1943, 1953, 1963. Computations performed using the Statistical Package for the Social Sciences.

PAGE 91

81 Mean Per Capita Government Expenditures by Census Years

PAGE 92

82 For tlie years 1870 to 1910, the index of education was created by smiming the percent of the population who were literate and the percent of the school age population attcndinji high school. Beginning in 1910, the percent of persons, age 19 to 24, attending college was added to the index. Figure 17 presents the mean values for the entire index of education for the years 1870 to 1960. Figui-e IS presents the rr.ean values for the index for the same period of time. Electoral participation is the final conceptual variable in the model. It was indexed by two different series of electoral statistics. The first series was turnout in Presidential elections as a percent of the total populace. Turnout in Presidential elections is an obvious measure of participation; however, the base used to compute a percentage tux-nout may occasion some comment. The decision to ignore the obvious differences in electoral laws regulating the suffrage was deliberately made. Tlie most compelling argiiment for using the total population, rather than the legally eligible electorate, as a data base lies in tlie definition of political development on w]iich this woi'k is based. Political participation is understood as one element in the process of development. Participation is seen as a political response to a change in the social environment. It is, of course, possible to disenfranchise certain groups in the society. Tlie basis for disenfranchiscment may be color, language, religion or sex, but, whatever the reason, exclusion of a social group from political participation does not exclude them from tlie society. There are reasonable bases for exclusion, such as age or mental inca.pacity, but

PAGE 93

Legend for Figure 17 Percent literate. Percent attending High School Definition of literacy changed in 1950, to functional literacy. Raw data from the U.S. Bureau of the Census, U.S. Census of Population , 1872, 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962; and Bureau of the Census, Statisti cal Ab stract of th e United Stat es, 1883, 1893, 1903, 1913, 1923, 1933, 1943, 195"3, 1963; iwd U.S. Commissioner of Education, Annual Repor t of the Commissioner of Education (Washington , D . C . : U.S. Government Printing Office, 1871, 1881). Computations performed using the Statistical Package for the Social Sciences.

PAGE 94

84 100' / / / 1870 80 90 1900 10 20 30 40 Figure 17 Moan Percent of Population \frio Are Literate and Mean Percent of Population Attending High School by Census Years

PAGE 95

Legend for Figure 18 Quantities in index numbers. Raw data from the U.S. Bureau of the Census, U. S. Census of Population , 1872, 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962; and Bureau of the Census, Statistical Abstract of the United States , 1883, 1893, 1903, 1913, 1923. 1933, 1943, 1953, 1963; and the Repo rt of the Commissioner of Education , 1871, 1881. Computations performed using the Statistical Package for the Social Sciences.

PAGE 96

BO 90 Figure 18 Mean Values of the Index of Education by Census Years

PAGE 97

87 even here, tiie criteria may yield to a changing social milieu. We may be forced to argue that the cighteen-yeai'old vote is more developed than the twent}'-one-year criterion. If it becomes clear that the desire for participation of the eigliteenyearold population is high, then their exclusion from political decision making indicates a lack of development. The argument can be raised that in a caste system, it is possible for one caste to advance while a lesser caste remains backv.'ard -perhaps. But basically, such an argument seems to claim that tlie upper caste constitutes tlie society. For this argument to be valid, we must be willing to equate political nonexistence with social non-existence. It is precisely the increasing self-awareness of social existence upon which we have predicated our argument of increasing pressure for political growth on both the permeance and participation dimensions, Hence it was deemed permissible to use the total population as a data base. Tliis base was also thought most reliable, since an enLuneration of the population would tend to.be more valid than an estimate of the eligible electorate. 11 11 V/alter Bumham argues (Walter D. Burnham, "The Changing Shape of the American Politica.l Universe," American Political Science Review (March, 1965), 7-2S3 that the Ainerican political universe altered from a predominantly high participatory state to a less participatory state around the turn of the century. This change can partly be attributed to a decline in participation rates by eligible white males.

PAGE 98

88 Operating on the assumption that alternative electoral arenas might reflect different developmental patterns, it was deemed advisable to check the model by using another series of electoral statistics. The second series used was turnout in off year Congressional elections. Congressional elections were chosen since the constituency represented is at the sub-state level and the pressures in tliis type of election would seem to be the most unlike the Presidential race. Yet, the office is significant enough to merit general attention. Off year elections were used to minimize the pulling effect that a Presidential contest is believed to exercise on the electorate. Ilius, the pi'essures to participate were not expected to come from the effects of a possibly exciting Presidential contes't. To further minimize the deviations inherent in single election results, the electoral data were aggregated around the appropriate decerinial census periods. For example, the elections for the Presidency for 1SS8 and 1892 \;ere averaged to give one figure per state, per decade. In this way, minor variations in turnout were smoothed over, and the effects of particular campaigns on the level of participation were partially controlled. Figures 19 and 20 give the mean values of the participation indices for the period 1870 to 1960. Before discussing the miethodology used, a summary statement of the variables and their operational indicators, as used in the research, is Dresented in Table VI.

PAGE 99

Legend for Fiaure 19 Raw data from Sven Petersen, A Statistical History of the American Presi dent ial Electio ns (New York: Ungar Publishers, 1963). Computations performed using the Statistical Package for the Social Scienc es.

PAGE 100

30 90 1900 10 20 3 Figure 19 '•lean Values of Turnout in Presidential Elections by Cen'sus Years Tl ^V

PAGE 101

eg end for Figure 20 Raw data from the Inter-University Consortium for Political Research. Statistics of the American Congre ss ional Election s, 18241968 (Computer tape). Computations performed using the Statistical Package for the Social Sciences.

PAGE 102

92 1370 SO 90 1900 10 20 30 40 Figure 20 Mean Values of Congressional Election Turnout by Census Years

PAGE 103

93 Table VI Variables Included in Each Final Index u'ith Years of Inclusion Variable Index Components Years INDUSTRIALIZATION Per capita value added in manufacturing 1870-1960 URBANIZATION Percent of population in cities larger than 2,500 1870-1960 Percent of population in cities larger than 25,000 1870-1960 :^!EDIA CONSUMPTION Average daily newspaper circulation per 1,000 population 1870-1960 Postal receipts, per capita 1870-1960 Telephones par 1,000 population 1880-1960 Radios per 1,000 population 1950-1960 GOVERN^qENT ACTIVITY Per capita state and local government expenditures 1870-1960 EDUCATION Percent of population who are literate 1870-1960 Percent of the school ago population attending high school 1870-1960 Percent of the 19 to 24 age group enrolled in college 1910-1960 POLITICAL PARTICIPATION Voter turnout in Presidential elections as a percent of total population; or 1870-1960 Voter turnout in off year Congressional elections as a percent of total population 1870-1960

PAGE 104

94 Met hodology Causal infereiice analysis using the Simon-Blalock method was the pi'imaiy technique For testing the model. Since causal inference has become a generally used technique in the discipline there will be no extended discussion of the teclmique itself. However, some subsidiary questions have been raised in recent i;ork wliich should be dealt with. Blalock has argued that comparison across populations is best pursued using the unstandardized regression coefficient. Blalock suggests that the correlation coefficient has limited generalizability . Since the correlation coefficient is standardized by the saianle, or population standard deviation, the coefficient is inextricably linked to tlie population parameters. If we compare coefficients fi-om two samples of the sam.e population we are correct, since the sample standard deviations are unbiased estimates of tlie population parameter. Comparison of coefficients based on differeirt 12 Hubert M. Blalock, Jr., Causal Inf erences in .-.'onExperimental Research (Chapel Jiill, N.C. : University of North Carol:}na Press, 1964). 13 Authors H-ho have used the tec]\nique include Done Id J. McCrone and Charles F. Cnudde, "Toiv'ard a Commujii cations l>.eory of Democratic Political Development: A Causal Model," ^J32^\53^n _Poj.itica] Sci?nc_e_ Review (March, 1967), 72-79; Arthur K. Smith", Jr., ""Socioecouomic Development and Political Democracy: A Causal Analysis," MicIw8St_ Jounial of Political Sci ence (February, 1969), 95-125; Deane K. Neubaucr, "Some" Conditions of Democracy," Am:erican Political Scie nce Review (December, 1967), 1002-1009; and Gilbert R. Wi'nham, "Political Development and LeiTier's 'Pneory: Further Test of a Causal Model" Ajueri can Political Science R eview (September, 1970), 810-818. Hubfrrt M. clalock, Jr., "Causal Inference,, Closed Populations and Measures of Association," .American Political Science Review (March, 1957), 150-136. ; ~~

PAGE 105

95 populations is incorrect, according to Blalock, since the standard deviations are now unbiased estimates of differe nt population parameters, If time is considered as a variable, then the ten year span between observations would suggest that v/e are comparing different populations. In one sense, in fact, each set of observations has been considered a population. ?\o tests of statistical significance were performed, since each series of observations constitutes an enumeration of the population at that point in time. Hence, statistical significance has no bearing in its normal sense. Tliis does present the problem raised by Blalock; are we, in fact, performing an incorrect comparison by using correlation coefficients? I do not think so. A correlation coefficient is a suinmary statement of the 'relationship between the operational indicators of two conceptual variates. Blalock himself argues tliat what we consider to be a population is determined by theoretical considerations. Tliis places us in a paradoxical situation. If we have X number of societies, and decide to consider all societies as constituting a population, and each individual society as constituting a sample, then we may compare correlation coefficients generated from data on each society. If, on the other hand, we consider each society to be a population, then we may not compare the coefficients. In both cases we have generated exactly the sa'ne set of numbers, but in one case Blalock suggests we may compare them, and in the other case we may not. Blalock 's suggested alternative is no more useful than the . correlation coefficient. Tlie unstandardized regression coefficient is

PAGE 106

obviously not standardized by the combined standard deviations of the comparison variables. If the core of Blalock's argument is that coefficients based on, and highly biased by, the populations from which they are drawn should not be compared, then the regression coefficient is prey to the same criticism as the correlation coefficient. By their . very nature, the correlation and regression coefficients are too intimately related to be differentiated in this fashion, A second critique in vogue in the discipline is the question of multi-collinearity. Tlie condition of multi-collinearity occurs in research situations where some of the indicators are so liighly correlated as to be indistinguishable. In causal inference analysis this means the loss of one or more terms in the equations and a consequent inability to test the model. Multi-collinearity is fundamentally a question of reliability. If you can theoretically differentiate between variable x and variable y, then you can tolerate very high levels of correlation between them and still include them in the same equation. If, on the other hand, you are dealing with variables possessing high levels of eri'or variance, you cannot tolerate a high level of correlation betv.-een them since you cannot put much confidence in the coefficients. Insofar as the research has been based on the enumeration of the population, the problem of sample variance does not enter. Hence, one possible source of error of estimate is eliminated. 15 Hugh D. Forbes and Edward R. Tufte, "A Note of Caution in Causal Modeling," A mericaji Politi cal Scie nce Revi ew (December, 1968), 12581264.

PAGE 107

97 The method of. computing the coefficients used to test the model increases their reliability even furtlier. For each decennial census period, a Pearson Product-?!oment Correlation coefficient matrix was generated. Tnese individual matrices were then aggregated for varying lengths of time hy a technique known as Z transformation. To average two correlation coefficients we would convert each r into an equivalent z. V,'e would then weight each coefficient by its respective N and sum. Tlie base for averaging is N^ + N2-2. Tluis we have the formula: Zr l(Np . Zr^CN^) A com.posite correlation coefficient of this sort has the combined reliability of both N and N . Tlie matrices of composite correlation coefficients referred to in the following chapters have been constructed using this technique. In addition to the zero-order correlations used to test the model, a series of multiple correlation coefficients was generated to test the substantive significance of the model. Miile the model could be satisfactory from the point of view of its internal relations, if the amount of explanatory power were m.inimal , then the model itself would be merely precious. Tlie mean multiple R for the five socio16 Tinman L. Kelley, Fundament al s of Statistic s (Cambridge, Mass.: Harvard University Press, 1947), 353-365.

PAGE 108

98 dcmograpliic variables regressed on turnout was 0.80 for the period 1S70 to 1960. llie following chapters will report on the actual tests of the model.

PAGE 109

CHAPTER IV MODELS OE POLITICAL DEVELOPMENT Basically, there are three ways to test the proposed developmental model. First, we may include all the observations, disregarding time as a variable. In this method, all the data are entered into the computation of one regression equation and the attendant correlations. From this series of coefficients we may check to see if the proposed model is operative for the entire time period, ignoring the effects of different levels of the variables as a function of the time examined, -mis test attempts to ascertain whether certain levels of one variable are generally related to certain levels of another variable, no matter when the measurement was made. Second, we may test a series of cross-sectional data sets aggregated for varying lengths of time. Tliis approach allows develop-m.ent of a first approximation of the model. From this test, we can ascertain whether or not a particular path is operative; however, we cannot be certain of the direction of causation. This approach is more reliable than single time period tests, but still leaves the possibility of mutual causation. Finally, we can test using a conl^ination of both the preceding methods. Here we correlate variables from aii earlier time period with variables for a later time period. In this manner we may establish temooral uriority for hypothetical causal variables. Tliis method 99

PAGE 110

100 eliminntes the possibility of laatual causiition inherent in tiie preceding two methods of testing. In the following pages all thi'ee methods of testing will be applied to establish as higli a level of I'cliability for the resulting model as is possible with the available data. B asic Hypotheses To prove the proposed model valid, two conditions must be met. First, all the relations among tlie six variables must exist as suggested. Second, the m.odel must hold for the entire time period from 1870 to 1960. Testing the proposed model will require a series of tests consisting of six seperate hypotheses: 1. Industrialization causes urbanization, which in turn causes increased media consumption. Tlierefore, if we control for the effects of urbani;-ation, the relationship between industrialization and media consumption will become negligible. 2. Urbaaization causes increased media consumption, which in turn causes increased government activity. ITierefore, if we control for the effects of increased media consumption, the relationships between urbanization and goveiTiment activity will become negligible. 3. Media consumption causes increased government activity, wliich in turn causes increased education. Therefore, if we conti'ol for the effects of government activity, the relationship between m,edia consumption and education will becom.e negligible.

PAGE 111

4. Increased government activity causes increased education, which in turn causes increased political participation. Therefore, if we control for the effects of education, the relationship between government activity and political participation will become negligible. 5. Urbanization and media consumption are intervening variables in the causal sequence leading from increased industrialization to increased government activity. Tlierefore, if we control for the effects of both urbanization and media consumption the relationship between industrialization and government activity will become negligible. 6. Government activity and education are intervening variables in the causal sequence leading fi'om media consumption to political participation. Tlierefore, if we control for the effects of government activity and education, the relationship between media consuirrption and political participation will become negligible. To determine the model's validity, the series of prediction equations shown in Table VII and representing the six hypotheses, will be tested. Table yil Causal Inference Prediction Equations for Each Hypothesis^ Hypothesis Equation Expected Value #1 IM.U lU X UM = . #2 UG.M = UM X MG = #3 ME.G. = MG X GE = #4 GP.E = GE X EP = #5 IG.UM = lU X UM X MG = #6 MP. EG = MG X GE X EP = IIndustrialization, U= Uibani^ation, M= Media Consumption, G= Government Activity, E== Education, and P= Political Participation

PAGE 112

102 Tlie first test will examine tlie possibility that the model is operative in all its relations for the entire period, 1870 to 1960. Table VI 11 presents the correlation matrix c"reated by aggregating the correlation 'natrices for each of the decennial census periods from Table VIII jgregate Correlation Matrix of the Six Indices for the Period 1870 to 1960^ I U M

PAGE 113

10: Ke may tentatively conclude that the path I ^U ^ M is operative. Hie second l:ypothesis suggests that U ->M -> G, represented by equation 2. [2] uG.M = b"l X MG = actual expected difference .7705 X .7319 .5709 .5639 = .01 We may tentatively conclude that the path U -^ M -> G is also operative. Tlie preceding two equations suggest the path I -> U ^M -> G, a contention represented by hypothesis 5. Tl-ie ne.xt equation tests this argument . [5] IG.u:i = lU X UM X MG = actual exi^ected difference .8044 X .7705 X .7319 = .4453 .4563 -.01 Tliese first three equations provide support for a causal path, I->U -> M ^G as operating during the entire period, 1870 to 1960. Tlie next link in the path is ex-pected to be M^ G ->E, and is represented by equation 4. [4] ME.G = MG X GE = actual expected difference .7319 X .6554 .5946 .4797 = .11 In equation 4, the difference betv.'cen the actual and expected values shown is larger than the arbitrary cut-off criterion of .10 suggested by Blalock. ',Ve shall, for the moment, conclude that the path M ->G ->E is not operative as expected. Tlie next link is represented by the path G ^ E -> P, and is presented in equation 5. ^Hubert M. Blalock, Causal_ Inferences in Xon-Experin-ental Research [Chapel Hill: University of North Carolina Press, 1964).

PAGE 114

[5] GP.E = Gl: \ EP = actual expected difference .6554 X .7213 = .5392 .4727 .06 V.'e may tentatively conclude that the path G ->E -> P is operative as expected. Equation 6 presents the relationship between medi a cons umption and politic al participation , wlien we control for the intervening effects of government activity and education . [6] MP. EG = MG X GE X EP = actual expected difference .7319 X .6554 X .7213 = .5140 .3460 .17 The high difference is, of course, expected due to the already revealed difference presented in equation 4. The causal paths which we now expect to operate are shown in Figure 21. Industrialization Media Consumption— — — — -^ Education -^ \' Urbanization GoveiTiment Activity Political Participation Figure 21 Empirically Identified Causal Model of Development in the United States for the Period 1870 to 1960 In Figure 21, the solid lines are the paths expected due to the theoretical argument presented earlier. The dotted lines are the paths added as a result of the two correlations exceeding our arbitrary minimum. Although some paths were found invalid for the entire period, they might still be valid for some period less than the total. If we examine the preceding equations, we can see that the primary difficulty lies with the unexpectedly large direct correlation

PAGE 115

105 between medi a consumption and education . We are, however, dealing with a correlation created by averaging together a total of ten separate cort-elations. If, over the entire one hundred year span, the correlations are of approximately the sam.e magnitude, then we would have to conclude that the extra path is continuously operative. If, on the other hand, the correlation changes in some consistent and systematic manner, then we may be justified in continuing our examination of shorter time periods. It is clear from Figure 22 that a continuous increase in the magnitude of the correlation between media consumption and education has occurred from 1890 to 1940. During the period 1870 to 1930, there has been a generalized decrease in the magnitude of the correlation between media consumption and g overnment activity . During the entire period, 1870 to 1960, the correlation between education and govemment activity has remained relatively stable around a mean of approximately 0.6500. At tlie same time, for the period 1870 to 1940, there has been a consistent increase in the magnitude of the correlation between media consumption and political participation . During the early 1920 's, all these coefficients had begun to converge and, in fact, to vary in much the same pattern. Based on the behavior of these coefficients, it was decided to dichotomize the time period covered. Phase I would cover the period 1870 to 1910, while Phase II would cover the period 1920 to 1960. Each of the six hypotheses would be retested using correlation matrices generated from data confined to these two separate phases. Table IX presents the correlation matrix covering the

PAGE 116

Legend for Figure 22 Media consijjnption with Government activity Education with Government activity Media consumption with Education Media consumption with Political participation Raw data from tlie U.S. Bureau of the Census, U.S. Census of Populat ion (U'ashington, D.C.: U.S. Government Printing Office, 1872, 1S83, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962): and U.S. Bureau of the Census, Statis ti cal Abstract o f th e United States (Washington, D.C.: U.S. Government Printing Office, 1383, 1893, 1903, 1913, 1923, 1933, 1943, 1953, 1963). Computations performed using the Statistical Package for the Social Sciences.

PAGE 117

107 1370 80 90 1900 10 20 30 ifO Figure ?2 Plot of Correlation Coefficievits of Media Consumption, Govemvnont Activity, Education, and Political Participation Versus Time: 1870 i960

PAGE 118

108 period 1870 to 1910. Table IX Aggregate Correlation Matrix of Six Indices for the Period, 1870 to 1910^ 1.0000 iM 0.8875 1.0000 0.6230 0.7670 1.0000 0.5635 0.6280 0.7578 1.0000 0.4542 0.4349 0.4986 0.6545 1.0000 0.1204 0.1665 0.3621 0.5075 0.7160 1.0000 ^Computations performed using the Statistical Package for the Social Sciences. Again, the expected fii-st link in the causal path is I -> U -> M. Equation 7 gives the appropriate prediction equation. [7] IM.U = lU X UM _ = actual expected difference .8875 X .7670 = .6230 .6807 =-.05 Tlie results of equation 7 are comparable to those of equtition 1, and again lead us to tentatively conclude that the path I -> U •> M is operative. The second link in the path is U ^^ M -> G, represented by equation 8, [8] UG.M = m X MG = actual expected difference .7670 X .7578 = .6280 .5812 .05 Tlie third hypotliesis in this series, I -> U -> M -> G is represented by equation 9.

PAGE 119

109 [9] IG.LM ^ lU . X WA X MG = actual expected difference .8875 X .7670 X .7578 = .5635 .5158 = .05 Tlie preceding three equations again lead us to the same results as the tests using a data matrix aggregated for the entire one hundred year period. Tlie causal path through social mobilization to increased government activity is operative. In the previous test of the model, the primary difficulty was encountered in the next series of equations. Our theoretical argioment suggests that "the link M -> G -> E should be operative. [10] ME.G =MG X GE = actual ex-pected difference .7578 X .6545 .4986 .4959 = .002 Tlie result of equation 10 is radically different from the result given previously in equation 4. From equation 10 we may tentatively conclude that the causal path M -^ G -> E is operative from 1870 to 1910. Following the expected chain, the next link is G -^ E -v P, represented by equation 11. [11] GP.E = GE X EP = actual expected difference .6545 X .7160 = .5075 .4686 = .03 As in equation 5, we again find the link G ^ E ^ P to be operative. Finally, it will be necessary to test the link M -> G -> E -> P for both the intervening variables. [12] MP. EG = MG X GE X EP ^ actual expected difference .7570 X .6545 X .7160 = .3621 .3551 = .007 Figure 23 presents the diagraiamatic representation of the model of political development for the period 1870 to 1910.

PAGE 120

no Industrialization Media Consumption Education Urbanization Government Activity Political Participation Figure 23 Empirically Identified Model of the Development Process in the United States: Phase I 1870 to 1910 For the moment, no attempt will be m.ade to consider or interpret the significance of these results; rather, the second series of tests will be presented, using the correlation matrix covering the years 1920 to 1960, Table X presents the relevant coefficients. Table X Aggregate Correlation !'atrix of Six Indices for the Period, 1920 to 1960 a U M 1.0000 0.7675 0.5626 0.3346 0.1850 0,3690 1.0000 0.7833 0.5140 0.3304 0.3772 1.0000 0.6882 0.6992 0.6876 1.0000 0.6636 0.5946 1.0000 0.7614 1,0000 ''Computations performed using the Statistical Package for the Soc ial Sciences. A link, I -> U ->• M is again expected. This link is represented by equation 15.

PAGE 121

Ill [13] IM.U = lU X UM = actual expected difference .7675 X .7833 .5626 .6000 = -.03 Ke may tentatively conclude that the link I -> U ^ H is operative for the period 1920 to 1960. T\\e next expected link is presented in equation 14. [14] UG.?' = UM X MG = actual expected difference .7833 X .6882 .5140 .5286 = -.01 We may also conclude that the path U -> M -> G is operative from 1920 to 1960. These two links suggest that the path I-> U -> M ^ G will also prove operative during this period. We may test that contention through an examination of equation 15, presenting the prediction equation controlling for both intervening variables. [15] IG.UM lU X UM X MG = actual expected difference .7675 X .7833 X .6876 = .3346 .4100 -.07 For the third time we may tentatively conclude that the path I -> U ^n -> G is operative. Recalling the results of testing for the total time period, we may expect that it will be necessary to add direct links from media consumption to ed ucation and pomjxal_partJ^ation for the period 1920 to 1960. Equation 16 presents the prediction equation for the link M -> G -V E. [16] ME.G = MG X GE --= actual ex-pected difference .6882 X .6636 = .6992 .4435 = .25 As expected, there is a direct link between media consumpt ion and education during the period, 1920 to 1960. As we have already examined

PAGE 122

112 one sub-period v;hich did not shov; this pattern, it is obvious that the total effect was primarily a result of the characteristics of the period 1920 to I960. The next link in the expected chain is G ^ E -> P, represented by equation 17. [17] GP.E = GE X EP = actual expected difference .6636 X .7614 = ,5946 .5053 .08 For the third time, we may conclude that the path G^ E -> P is operative. The preceding two equations would lead us to expect a direct I'elationship between media consumption and political participation . Equation IS presents the relevant prediction equation. [18] MP.GE MG X GE X EP = actual expected difference .6882 X .6636 X .7614 = .6876 .3391 = .32 Equation IS clearly indicates the necessity for directly linking media consumption and politi ca l participation during the period 1920 to 1960. 71iis result was expected since total period examination showed this same pattern. As a 'check on the point of dichotomization, the data were also aggregated to cover the period 1870 to 1920. Calculations from these coefficients indicated that the Pliase I model was not operative. We may, therefore, conclude that the point of change between Phases I and II did occur between 1910 and 1920. Table XI and Figure 24 are summary presentations of the characteristics of the t \?o successive Phases of tlie developmental process elucidated thus far. Since the model for the period 1870 to 1910, fulfilled our expectations so nicely, there is little need to discuss this model.

PAGE 123

113 Table XI Prediction Equations for the First and Second Models of Development Phase I 1870 to 1910 Actual Expected Difference IM.U = lU X UM .8875 X .7670 .6230 .6807 -.05 UG.M = UM X MG .7670 X .7578 .6280 .5812 = .04 ME.G = MG X GE .7578 X .6545 .4986 .4959 = .002 GP.E = GE X EP .6545 X .7160 .5075 .4686 = .04 IG.UM = lU X UM X MG .8875 X .7670 X .7578 .5635 .5158 = .05 MP. EG = MG X GE X EP .7578 X .6545 X .7160 .3621 .3551 = .007 Phase II1910 to 1960 IM.U = lU X UM .7675 X .7833 .5626 .6000 = -.04 UG.M = UM X MG .7833 X .6882 * .5140 .5286 = -.01 ME.G = MG X GE .6882 X .6636 .6992 .4455 .25 GP.E. = GE X EP .6636 X .7614 .5946 .5053 = .09 IG.UM = lU X UM X MG .7675 X .7833 X .6876 .3346 .4100 = -.07 MP. EG = MG X GE X EP .6882 X .6636 X .7614 .6876 .3391 = .32

PAGE 124

114 24a Phase I 1870 to 1910 Cndustrialization Media Consumption F.ducation Urbanization Government Activity Political Participation 24b Phase II1920 to 1960 Industrialization Media Consumption Urbanization -^ Education 7 Government Activity Political Participation Figure 24 Summary Presentation of Both Phases of the Causal IModol On the other hand, the model for the period 1920 to 1960 is neither as parsimonious, nor as pleasing. It is clear that throughout the entire period, 1 S70 to 1960, me dia consumption h as been gaining prominence as an explanatory variable in regard to both education and political p arti c ipati on. Our prime concern at this point will be attempting to

PAGE 125

115 understand the reason for this increasing prominence. During the entire period, 1870 to 1960, the fii-st part of the theoretically expected path, I -)U > M -> G, is operative. Let us review the theoretical expltmation for the second path, M -x G -v E -> P. Tlie second stage of social mobilization causes increased pressure for individual adaptability. The primary method of increasing the general level of adaptability in a society was stated to be mass education. It v/as also argued that this need occassioned widespread government intei'vention into the social order, due to the inability of existing social institutions to fill tlie need for educational opportunity. Tlie effect of increased individual adaptability due to increased education should be increased political participation. Tlais inci-eased pai'ticipation is expected to be a result of both the increased equality of opportunity resulting from education, and an increased feeling of political self-confidence. An educational system is obviously an infrastructure.^ Miile in need of maintenance and upgrading, it is not a constant object of massive energy input from the political structure. After a period of time, the infrastructure can operate as an ongoing institution rather than as a newly risen social agency. If the primary purjjose of education is an increased level of individual adaptability, then that function has been fulfilled between 1870 and 1910. The definition of 3 Warren F. Ilchman and Norman T. Uphoff, I1ie_ Polit ical Economy of Change (Los /jigeles: University of California Press, 1969), 35.

PAGE 126

116 political development offered earlier ei:ip]iasi2ed a continuous increase in equality of opportunity. Quite obviously, increasing educational opportunity is the primaiy component of increasing equality of opportunity for only a portion of tlie developmental process, and is not the single cause during the second phase of the model. Tlie second phase of the model may, therefore, be understood as a result of restricting the concepts of government activity and equality of opportunity as represented by education, to too narrow a base. It was suggested earlier that a retest, using an alternative indicator of political participation, would be advisable. The retest, using data from Congressional elections, provided confirmation of the results already reported. Equation 19 presents the prediction equation for the link G ^ P -> E for 1870 to 1910. [19] GP.E = GE X EP = actual expected difference .6545 X .6632 = .5012 4341 .06 Equation 20 presents the prediction equation for the link M -> G -> E ^ P. [20] MP.GE = MG X GE X EP = actual expected difference .7578 X .6545 X .6632 .3625 .3290 = .03 [Since the preceding links in the chain are based on the same set of data, they of course, remain the same.) Using off year Congressional election turnout as an indicator of political participation, the same model is operative for the period 1870 to 1910, as that represented in Figure 24a. IVith one minor variation, retesting the model for the period 1920 to 1960, replicates the model shown in Figure 24b.

PAGE 127

117 Equation 21 presents the pi-ediction equation for the link G -> E -> P. [21] GP.E = GE X EP actual expected difference .6636 X .6294 = .5690 .4177 = .15 This result differs from that shown in Figure 24b; tliere is now the possibility of a direct link between govei'nme nt activity and poli tic al participation. Tliat link will, however, be shown to be spurious. Examination of equation 22 illustrates the effect of the two intervening variables, government activity and education , in the path M ->• G •> E -> P. [22] MP.GE MG X GE X EP actual expected difference .6882 X .6636 X .6294 = .6383 .2875 = .35 Tlie result of equation 22, as well as tlie previously presented result of equation 16, leads us to expect the I'clation betv.'een government activity and poli tical participation to be a product of the combined relationships of government activit y with education and media consumption . Since these relationships do not represent a straight causal path, it will be necessary to t6st the relationship using the true second order, partial correlation. Equation 23 compares the expected correlation after partialing with the actual correlation. [23] GP.ME = partial zero-order .1305 .5690 From equation 23 we may tentatively conclude that the path G -> P is a spurious result of the combined intervening effects of media consumption and education. Tlierefore, the retest of the model for 1920 to 1960, using off year Congressional elections as an indicator of po liti cal pa rtic ipation, replicates the results presented in Figure 24b,

PAGE 128

118 For the model to continue in tlic patli presented in Figure 24a, it would be necessary to expand the indices of gove rnment act ivity and education to include more recent forms of behavior, especially tliose forms of behavior which represent qualitatively different forms of government activity. Therefore, before concluding that the theoretical argument is invalid for the period 1920 to 1960, further analysis will have to be performed using more sophisticated indices of government activity and education. One consistent path of change, the path from industrialization to urbanization to media consumption to government activity , has been evident througliout all three tests of the model. It can be concluded fi'om these tlirce tests that there exists a relatively consistent relation between increased rates of social change and increased political permeance gi'owth. Vi'e have defined political development as simultaneous growth on botli the political permeance dimension and the political participation dimension. Our m.odel has clearly indicated an intervening variable during a large part of the process of development. Based on the existence of this intervening variable, educati on, we must allow for a time lag in the growth of both relevant dimensions. Table XII presents the correlations between government activity in one census period with polit i cal participation in the succeeding census period. Table XII indicates an increasing agreement between the dimensions of permeance and participation, an agreement which we have termed political development. This increasing agreement throughout

PAGE 129

119 the entire liundred year period lends additional credence to the contention that the Tuodel presented represents the causal path of political development in the United States. Table XII Inter-Censal Correlations of Government Activity at T, with Political Participation at T2 for the Period, 1870 to 1960 Years

PAGE 130

120 included in the model have provided a consistently high level of explanatory power in regard to the level of political participation. Table XIII Multiple Correlation Coefficients of Political Participation with Indices of Education, Media Consumption, Government Activity, Urbanization and Industrialization by Census Years Year Multiple R

PAGE 131

included in the model are susceptible to very high levels of explanation using the socio-deraographic variables. In addition, our reformulation of the relations between the political variables indicates that the political system, in response to external pressure, created its own future support. \'/hen seen within the framework of passing time, education, conventionally considered an output of the political system, is obviously also a very important input. Table XIV Multiple Correlation Coefficients of Government Activity with Media Consumption, Urbanization and Industrialization by Census Years ^ Year 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 Mult

PAGE 132

122 To tl\is juncture we have tested the hypothetical model by aggregating data across time for a series of cross-sectional tests. However, this metliod of testing is not as longitudinally oriented as we might wish. An alternative method of testing is suggested by the data presented in Table XV. Table XV presents the correlations of each variable with itself at succeeding time periods. Table XV Cross Year Correlation Matrix for the Period 1870 to 1960 ^ I U 1870 \.lth 1880 0.9444 0.9728 0.8976 0.9147 0.9859 0.7971 1880 with 1890 0.9655 0.9661 0.8495 -0.9406 0.9819 0.8210 1890 with 1900 0.9475 0.9784 0.9080 0.8254 0.9902 0.6977 1900 wich 1910 0.9709 0.9827 0.8305 0.8835 0.9864 0.9649 1910 with 1920 0.9587 0.9882 0.8905 0.9195 0.7608 0.9204 1920 with 1950 0.9537 0.9894 0.9271 0.7874 0.7882 0.9554 1930 with 1940 0.9770 0.9968 ' 0.9684 0.9393 0.9544 0.9766 1940 with 1950 0.9512 0.9753 0.9495 0.8919 0.9529 0.9821 1950 with 1960 0.9593 0.9383 0.9480 0.6943 0.9475 0.9591 'Computations performed using the St atistical Pac kage fo r the Social Sciences. Figure 25 indicates the basis for the drop in the correlations for the educational variable between 1910 and 1920 and between 1920 and 1930. The period witnessed a particularly rapid, and apparently non-uniform, shift in the level of education. The drop in the level

PAGE 133

Legend for Figure 25 Quantities in index numbers Raw data from tlie U. S. Bureau of the Census, U .S. C en sus of Population , 1872, 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962; Bureau of the Census, Statistical Abstract of the United States , 1883, 1893, 1903, 1913, 1923, 1933, 1943, 1953, 1963; and U.S. Commissioner of Education, Annu al Report of the Commissioner of Educat ion (Washington, D.C.: U.S. Government Printing Office. 1871, 1881). Computations performed using the Statistica l Package for the Social Sciences.

PAGE 134

L24 1370 80 90 1900 10 20 30 "+0 Figure 25 Mean Values of the Index of Education: 1870 to 1960

PAGE 135

125 of correlation of the participation index between 1890 and 1900 is probably a reflection of the disenfranchisment of Blacks in the South under the impact of the Jim Crow laws. Tlie Inter-temporal correlations presented in Table XV indicate an extremely high level of stability in transition from period to period. Change is obviously occurring, but the rate of change is not disturbing. For instance, the level of government activity may increase very rapidly from year to year, and in fact does. However, if the rate of change is uniform throughout the society, then the year to year correlations will not be greatly affected. Table XV indicates that the pattern of change in American society has been consistent for the last one hundred years. As the pattern of change has been consistent, it should be possible to treat the entire period from 1S70 to 1960 as one data set. Rather than generating a series of correlation matrices and then aggregating, we may generate one matrix directly for the entire period. The figures of Chapter III which plot the variables' mean values through time, indicate that some variables do, in fact, experience rapid alterations in their rate of change. It will, therefore, be necessary to take into account the varying rates of change as we proceed to generate the single correlation matrix. Since these figures indicate the possibility of curvilinear rates of change, it was decided to control for alterations in rates of change through the use of logarithimic transformations. All variables were transformed into base ten logarithims. Correlations were then generated between

PAGE 136

126 the untrans formed variables and their transformed countei-parts . Table XVI presents that series of correlations. Table XVI Correlation Matrix of Original Variables with Logarithmic Transformed Counterparts ^ Log I 0.7945 Log U 0.9049 Log M 0.8427 Log G 0.8171 Log E 0.9654 Log P 0.9481 a ~~~~ Computations performed using the Statistical Package for the Social Package for t he Social Sciences. A correlation of less than 0.9000 was adopted as indicating significant deviation from linearity. Using that criterion, three variables, industrialization , media consumption and government activity , were used in the matrix as logarithms. Figure 26 presents these variables plotted as the mean logaritlunic value from 1870 to 1960. Comparing these figures with the corresponding figures in Qiapter III vv'ill indicate the general reduction in non-linearity obtained by the use of log transforms. Ha-i/ing transformed these three variables into their logarithmic counterparts, a correlation matrix for the entire period was generated.

PAGE 137

Legend for Figure 26 Media Consumption . Industi'ialization Government Activity Computations performed using the Statistical Package for the Social Sciences.

PAGE 138

128 _ J -^ / 870 80 90 1900 30 40 50 Figure 26 Logarithmic Transformations of the Mean Values of Media Consujnption, Industrialization and Government Activity: 1870 to 1960

PAGE 139

In this matrix the N'-is 452. Each state, at each census period, was used as one observation. Table XVII presents the correlation matrix for the entire period. Table XVII Correlation Matrix of Six Variables Using All Observations ^ M I

PAGE 140

130 of a community iiitcgr.ating function in an increasingly urban and, t!:erefore, increasingly complex society. In a cross-sectional analysis we are dealing with a scries of physical necessities which make the path I > U -^ M e.xpected. However, this path is not necessarily a completely accurate representation of the process of change. At any time T, we would expect the path to operate. However, in terms of our theoretical groimding, urbanization and media consumption are decomposed elements of the phenomenon of social mobilization. In an inter-temporal sequence, as we are testing in this case, industrialization is actually the cause of both urbanization and media consumption. 'A'e may test the hypothesis of common causation using the predictioii equation given in equation 25. [25] MU.I = IM X lU = actual expected difference .8284 X .7782 = .7001 .6447 = .05 Hie results conform with the argument tliat through time, industrialization is the mutual cause of both urbanization and media consumption. We would still expect that increased media consumption would lead to increased government activity. Since industrialization is now seen as a direct cause of increased media consumption, we would expect the path I ->M -> G to be operative. Equation 26 presents the relevant coefficients for the hypothetical path I -> M -> G. [26] IG.M = IM X MG = actual expected difference .3284 X .8782 = .8072 .7275 = .07 Given the evidence of equation 26, we may tentatively conclude that the path I -^ M -> G is operative. The next expected link corresponds to the

PAGE 141

previously discerned path M -> G -> E. Equation 27 presents the prediction equation. [27] ME.G = MG X GE = actual expected difference .8782 X .8945 = .8300 .7855 .04 The lesult is again expected from previous test evidence. Equation 28 presents the prediction equation for the next expected link, G -> E ^ P. [28] GP.E = GE X EP = actual expected difference .8943 X .7648 = .7261 .6840 = .04 This leads us to tentatively conclude that the path G -> E ->P is operating. The final expected link is based on the existence of both government activity and education as intervening variables between medi a consumpti on and political participation . [29] MP.GE = MG X GE X EP = actual expected difference .8782 X .8943 X .7648 = .6566 .6007 = .05 Figure 27 presents a diagrammatic summary statement of the series of paths represented by tlie preceding equations. Media Consumption Education Industrialization Urbanization Government Activity Political Participation Figure 27 Empirically Identified Model of the Development Process in the United States Using All Observations as Data Base In general, all test results for the model of developm.ent have provided confirmation of our basic expectations as to the path of

PAGE 142

132 development. Model I, covering the period 1870 to 1910 is obviously the clearest statement of the fundamental process. Model 11, covering the period 1920 to 1960, is not as parsimonious, but is basically the same. Model III, the inter-temporal test resi'lt, again provides confirmation of the basic theoretical argument, and substantiates the results of the earlier tests. Despite some differences in the internal relations between the models, all three produce very similar predictive results in regard to the level of political participation. Table XVIII presents a comparison of tlie multiple correlation coefficients between political participation cind the five socio-demographic or political variables. Table XVIII Multiple Correlation Coefficients for the Three Models of Development ^ Model Multiple R R-Square I 0.7429 55% II 0.8478 72% III 0.7798 61% Computations performed using the St atistical Package for the Social Sciences. The ensuing section will further consider the significance of these findings and discuss some additional ramifications of the model. It will be particularly concerned with the inter-temporal verification of the findings of the two preceding tests of the model.

PAGE 143

133 An I nter-Tc ;np or.al Test of the ::odel of Dcvelopinent The final method of model testing is a combination of the two previous tests. Time is explicitly introduced into the model, yet the cross-sectional characteristics of the first tests are retained. Tlie technique for generating the test matrices is as follows. Each variable at time T, is correlated with its expected succeeding variable at time T2 . This series of matrices, one for each inter-censal correlation, is then aggregated to the appropriate sub-set of the total period from 1S70 to 1960. For example, industrialization at time T^^, say 1880, would be correlated with urbanization , m.edia consumption , government activity , education and political participation in 1890. Urbanization at time T, would be correlated with media consumption , government activity, educati on and political participation at time T2, and so on, until all the correlations were generated. Tliis set of correlations would provide the matrix for the inter-censal period 1880 to 1890. In all, nine sets of inter-censal correlations v.ere generated. Tliese sets of correlations were then aggregated to correspond to the periods 1870 to 1910 and 1920 to 1960. Tliis method of testing introduces inter-temporal correlations into the analysis, and sets up a series of asymmetrical relations. In this test, the variable which is seen as the cause in the chain has occurred a full ten years before the effect variable. Tliis eliminates the possibility of mutual causation which is inherent in the cross-sectional tests of the model. Table XIX presents the aggregate matrix of inter-censal correlations for the period 1370 to 1910.

PAGE 144

134 Table XIX Aggregate Correlation Matrix of the Inter-Temporal Correlations: 1870 to 1910 ^

PAGE 145

135 [52] IG.UM = lU X UM X MG = actual expected difference .8805 X .7414 X .6718 = .4935 .4386 = .05 Tiie next expected link is the series from M ^ G ->E, represented by equation 33. [33] MG.E = MG X GE = actual expected difference .6718 X .6669 = .4715 .4480 .02 Tliese results lead us to expect the link G -)E -v P, presented in equation 34. [54] GP.E = GE X EP = actual expected difference .6669 X .7613 = .4926 .5077 = -.01 Finally, we must test the link M -> G -> E -)P, shown in equation 35. [35] MP. EG = MG X GE X EP = actual expected difference .6718 X .6669 X .7613 .5656 .5406 = .02 Tlie results of this series for the period 1870 to 1910, are summarized in Figure 28. Industrialization Media Consumption Education Urbanization Government Activity Political Participation Figure 28 Empirically Identified Model of the Development Process in Phase I, 1870 to 1910, Using Inter-Temporal Correlations Clearly the results shown in Figure 28 replicate exactly the results of the tests summarized in Figure 24a of the preceding chapter.

PAGE 146

136 The results of both tests confirm our expectations. Table XX presents the aggregate correlation matrix covering the period 1920 to 1960. Table XX Aggregate Correlation Matrix of the Inter-Temporal Correlations: 1920 to 1960^

PAGE 147

[57] UG.M = ITA X MG = actual e.-q.iected differsnce .7974 X .6223 = .4710 .4962 = -.02 ITiis result confirms our exi:)ectations, and leads us to test the final link of this section, I > U ^ M -> G, which is represented by equation 38. [38] IG.UM = lU X UM X MG = actual e.xpected difference .7323 X .7974 X .6223 .3950 .3634 .03 The next section in the model is the link M -> G -> E, which is represented by equation 39. [39] ME.G = MG X GE == actual expected difference .6223 X .6637 = .6969 .4130 = .28 The result is expected given the results o-f previous tests of the model for the second period. Tlie next link in the path, G -> E -^ P, did operate in one test, although the difference was close to being excessive. This path is represented by equation 40. [40] GP.E = GE X EP actual expected difference .6637 X .7454 '= .6937 .4947 = .20 In this test, as in another previous test, the link does not operate. Given the preceding two equations, we can be assured that M -> G ^E -> P does not operate. [41] MP. EG = MG X GE X EP = actual expected difference .6223 X .6637 X .7454 .7060 .3079 = .39 As indicated in the pre\ious test, it is still expected that the link between government activity and political participation^ is a spurious result of the path G ^ E ^P, and the mutual causation effect of media consimption on both government activ ity and political participation .

PAGE 148

138 l\e may test this h)'potIiesis by generating the true partial correlation coefficient between governm ent activity and political participation, controlling for media consumption and education. Equation 42 presents the true partial, as well as the zero-order correlation. [42] GP.EM = partial zero-order .1747 .6937 From the above result we may conclude that the relation G ->. P is, in fact, spurious. Partialling out the combined effect of media consumption and educatio n reduced the correlation from explaining 48 percent of the variance to explaining less than 4 percent of the variance. This reduction clearly indicates that the original relation is spurious. Figure 29 summarizes the paths of the preceding equations in diagrammatic form. Industrialization Media Consumption. — Urbanization Government Activity Political Participatic Figure 29 Empirically Identified Model of the Development Process in Phase 11^ 1920 to 1960, Using Inter-Temporal Correlations The results of this test correspond to the preceding tests of the m.odel in all respects. V,'e may, therefore, conclude that tlie model is operative in the United States, in two phases, for the period 1870 to 1960.

PAGE 149

139 The series of tests conducted in the preceding pages provide a fundamental confirmation, in the United States, for the theoretical arguments upon which the research was based. For the earliest period of .^nericaji history examined, the path of development was a straight chain beginning with increasing wealth and ending with increased political participation. Later, a second period was uncovered in which a direct link between media consumption and participation was found to exist. In the ensuing chapter an attempt will be made to revise the theoretical expectations to include this change as a function of the establishment of a new mfrastructural element. In sum, the preceding tests indicate that it is possible to develop a model of the socio-demographic and political forces that give rise to the phenomenon we have termed political development. Ihis model allows us to assign causal priorities to the relevant variables, and provides an empirical platform from which to engage in further theoretical activities. It was noted in Chapter II that the research began in a tim.e period where the mass of the populace was not of peasant background. In research conducted on the Philippines (Jesse F. Marquette, "Social Mobilization and the Philippine Political System," Comi^arat ive Politica l Stu dies (in press), an obviously peasant society, the model of development advanced in this work was found to operate in virtually the same manner. Hence, it is possible to suggest that the model is operative for societies beginning the process of development from an agrarian level far more extreme than that of the United St-Jtes circa 1870.

PAGE 150

CHAPTER V DISCUSSION A.ND CONCLUSIONS It has been argued that the pattern of social change is a primary causal element in determining the pattern of political change. This research has attempted to uncover the particular pattern of social change which leads to the pattern of political change labeled political development. It was argued that the first steps in the process are based on an increase in the total disposable wealth of the society. As this wealth increases, a further process of social mobilization is expected to occur as a consequence of increasing wealth. Tlie social stresses occasioned by the two interacting processes are expected to lead to an increasing involvement of the state in the affairs of the society. This increasing involvement is termed permeance growth. Tlie set of relations between the three variables is expected to follow a causal path which is represented by: Wealth-> > ^ ^ Social Mobilizatlon-> -> -> -y -s-Premeance Growth The concept of social mobilization was decomposed into two distinct processes, urbanization and media consumption. The causal path resulting from the conceptual decomposition is expected to be: V,'ealth-> > ^^Urbanization-> -> ->l^!edia Consumption^ ~> ^Permeance Growth Tnis sot of conceptual variables was operationalized by a series of four indicators: industrialization, urbanization, media consumption, and 140

PAGE 151

141 goveriiiTient acti"\'ity. In the actual research, the empirically examined path of change was: Industrialization^ -> Urbanization-*-> Media Consumption-^ ^ Government Activity The research results reported in the preceding pages substantially confimn the expectations derived from the model. For the period 1870 to 1960, in two alternative tests of the model, the path of causation did, in fact, correspond to the path: Industrialization-> ^ Urbanization-*-> Media Consumption-*->Government Activity Anotlier method of testing, treating the period 1870 to 1960 as one unit of time, produced somewhat dissimilar results in that the model was found to follow the path shown in Figure 30. Industrial! zation ^I'edia Consumption -> Government Activity Urbanization Figure 30 First Section Causal Model Resulting from Total Obsei-vations Test Tliis result again substantially confirms the basic theoretical expectations. Tlie varying paths of tlie model can be explained by reference to the time period used. In the cross -sectional and inter-temporal tests, the path I -> U -> M -> G was found to be operative Tlie basic

PAGE 152

142 expectation in i-cgard to this path v;as related to the necessity, for media consuruption as a force for social integration in an increasingly complex urban environment. Industrialization was expected to cause both media consumption and urbanization. At any time T, however, urbanization was expected to be an intervening variable between industrialization and media consumption due to the social stresses arising from the process of urbanization. Over long periods of time, as represented by the test in which the period 1S70 to 1960, was treated as one unit, industrialization was expected to cause both urbanization and media consumption. The apparent cause of increasing social complexity, giving rise to the need for increased social integration, was industrialization. Over time, as the .results of this test suggest, urbanization can be imderstood as an epiplienomcnon of industrialization, lliere is a lag between increasing productive complexity and the integrative social response because the social response is actually a result of the most immediately occurring form of increasing complexity^ urbanization. In the full time. period test, this lag is no longer apparent. Political development is defined as the activity of the state in insuring equality of opportunity over ever larger ranges of life chances. Hypothetical ly, in a rapidly changing social environment, the major activity of the state in response to the pressures for equality of opportunity would be in the. direction of incre.-.sing individual adaptability through education. This major theoretical premise is based on the concept of the Insatiable Personality. As this

PAGE 153

personality type becomes more prevalent as a result of social mobilization, the demand for more equal opportunity is expected to increase. General increases in opportunity are expected to be contingent upon the ability of the individual to adapt in an unfamiliar environment. In order to meet this social pressure, the activity of the state necessarily centers on educational expansion. As a resultant by-product of the adaptability and self-confidence instilled by education, there was expected to occur an increase in the level of participation. This section of the model was expected to follow the path: Goveniment Activity-^ -> Education^ -* Political Participation In general, the results of the researcli have coiifirmed the existence of this process in the United States. During the period 1870 to 1910, this path existed exactly as specified. For this period the model of the six variables followed the path shown in Figure 31. Industrialization Media Consumption Education Urbanization Government Activity Political Participation Figure 31 Sumipary Presentation of Phase I Model, 1870 to 1910 Tiiis path confirms the theoretical arguments advanced. For the period 1920 to 1560, the model shov.'n in Figure 32 indicates a direct linkage betvveen education and media consumption , and between poli tical participation and med ia consumption .

PAGE 154

144 Industrialization -.Media Cons luiipti on' — — —y nducation Urbanization Government Activity Political Participation Figure 32 Summary Presentation of Pliase II Model, 1920 to 1960 Tne path shown in Figure 32 indicates an alteration in the internal relations of the model not accounted for by the theory. Before considering this alteration, note that the model for the full time period does not indicate this alteration. Industrialization Media Consumption Education Urbanization Government Activity Political Participation Figure 33 Summary Presentation of Model Generated from Total Observations Matrix Figure 33 indicates that over the entire period, the general tlieoretical expectations have been met. Hov;ever, the results of the model covering 1920 to 1960, in the cross-sectional tests, indicate that in recent decades a sliift has been occurring. Tlie apparent reason for this shift lies in the nature of the educational institution. In any society, education is an infrastructural element, it is necessary for other ends, but not an end in

PAGE 155

145 itself.^ During the period 1870 to 1910, the infrastructure building process was still under way, while in the second period the infrastructure was largely established. By 1920, well over 90 percent of the populace was literate. This suggests that there would be a shift in the last stages of the model, making education a correlated but not causal variable. IVere this the case, we would expect the remaining variables in the last section of the model to assume a new pattern of causality. The theory leads us to expect that the social pressures arising in the earlier section of the m.odel give ris*^ to tlie necessity for increased government activity, which, in turn, answers these pressing social problems by increasing the level of adaptability through education. The level of political participation of the citizenry is raised by tlie increase in adaptability. If the process of education is an intervening variable in the pi'ocess of social change only so long as it is developing as an infrastructural elem.ent, then we would expect the immediate causal process to reassert itself when tlie infrastructure is complete. Given our earlier theoretical expectations, the immediate factor in this section of the model is media c onsumpti on. If there has been a decrease in the effect of education, we would expect media consumption to be the cause of both government activity and political participation. The path from government activity to education 1 IVarren F. Ilchman and Norman T. Uphoff, The Political Econ o my of Change (Los Angeles: University of California Press, i969)7"35^

PAGE 156

146 to participation was necessary during the period wlien general political self-confidence was being established. Once the general populace achieved a minimal level of adaptability, the future cause of their political activity should be the general state of social integration. Attendant problems of a complex society should also impact more directly. We would, tlierefore, expect a path in wliich these three variables are related as shown in Figure 34. Media Consumption -;>. Political Participation Government Activity Figure 34 Model of Expected Third Phase Relations It is possible to excunine this final hypothesis using only the data for the most recently available period, 1960. Table XXI presents the correlation matrix for these three variables in 1960 only. Table XXI Correlation Matrix of Media Consumption, Government Activity, and Political Participation, 1960 ^ M M 1.0000 G P 0.6857 1.0000 0.6821 0.4440 1.0000 Computations performed using the Sta tistical Package for the Social Sciences .

PAGE 157

147 Tlie prediction equation for the expected path is represented by the following equation. GP.M = MG X MP = actual expected difference .6857 X .6821 = .4440 .4677 -.02 In this final phase of the model, media consumption becomes the mutual cause of both government activity and political participation. Here, education is simply a correlated variable, its causal priority in the sequence of development being lost as soon as the vast majority of the populace acquires an education. Theoretically, education is seen as an output of the political system, v;hich, by insuring individual adaptability, also insures equality of opportunity and further political participation. Once the majority of the populace has achieved a minimal level of education, this source of insuring equality of opportunity is established, and future political outputs will have to come to grips with other problems of com.plex society. Tlie more highly educated a group, the m.ore they would be expected to participate in the political system, but v;e no longer place education in the causal chain. Rather, leaving been established as an infrastructure, education has become causally independent. Causally, po]itical participation is now directly I'elated to the level of social integration indicated by media consumption. Tlie conceptual meaning and importance of participation here corresponds

PAGE 158

148 to Poinpor's "indirect effects."^ It was suggested earlier that, political participation was understood to index congruence growth. Tlie argument is that in a complex society it becomes increasingly difficult for anyone to anticipate the needs and desires of the mass of the populace for alternative forms of social mobility. Political participation is not expected to directly insure greater equalization of opportunity, but rather to prevent current inequalities from persisting too long. Pomper quotes J.S. Mill; "Men, as well as women, do not need political rights in order that they may govern, but in order that they not be misgoverned." In this view, extended participation by the mass is necessary as indication of the directions of stress currents in the social system. IVe have indicated the process of political development by the continuing inter-temporal increase in the correlation between government activity and political participation. As these two variables become more directly related, and as the mutual causality of the level of social integration becom.es more apparent, we may argue that both the input and output processes of the political system are becomiiig more directly responsible to the needs of the social system. This dii-ect, and mutually supportive, interaction between the elements of the social and political spheres is the process we have termed political development. 2 Gerald Pomper, Elections in ^Xmerica: Cont rol and Influence in Dem ocratic_Politics (New York: Dodd, Mead and Company, 1968), 25-40. Ibid., 28.

PAGE 159

Conc l usicns One definition of political development catalogued by Pye suggests that developmeiit is the "politics of stability and orderly 4 change." This research seems to suggest that rather than being the politics , development is the result of stability and orderly change. Villi le the rates of change in the individual variables have altered during the past one hundred years. Table XVI indicates that the process itself has remained consistent throughout. In Eckstein's terms, the relationships within the American political and social systems have demonstrated a low level of tension during this period. V.liile there s obvious stress within these systems during tlie periods of study, t is possible to suggest that the consistency of the process has ended to minimize this stress. Tnis researcli indicates that increasng individual adaptability is the primary necessity for a developing political system. This increase is an extremely time consuming process. It cannot occur very rapidly, simply because the systemic capacity to produce institutions fostering adaptation is limited. At a minimum, om or two generations is required to raise the level of adaptability in the majority of the populace. If, during this period, the social and political systems undergo severe stresses, the average individual is 4 Lucian VL Pye, Asp ects of Political Developmen t (Boston: Little, Brown and Company, 1966), 41-42. 5 Alexander Eckstein, "Economic Development and Political Change in Communist Systems," World -Po lit ics (July, 1970), 475-495.

PAGE 160

150 likely to be faced with situations to which he cannot adapt and the process of development will begin to bi'cak down. Tlius, we would not expect political development in societies which experience frequent or severe social or political crises, because such crises sap the responsive capacities of tlie average citizen. Eckstein uses the concept of tension to explain the "potency of an ideology."^ Tlie higher the level of tension within a system, the more potent tlie ideology and the more closely it approximates a Weltanschauung. We may extend this argument and suggest that at different levels of social mobilization, different kinds of tensions result, and hence different form.s of ideology. Tlie level of social mobilization at a given moment may be indicated by the percent of the population engaged in non-agricultural occupations. We may argue that if the process of change is interrupted by a crisis at one level of mobilization, one form of ideology will tend to dominate, wliile if the crisis occurs at a different level, a different fonn of ideology will dominate. Specifically, if tlie process is interrupted at an early stage, where few members of the populace have been mobilized, then the likely result of this conjunction of process and event will be what Apter has 6 Ibid., 479.

PAGE 161

151 termed the "mobilization system," If the crisis occurs at higher levels of mobilization, the result will tend to be what I will terra q the "integration system."'^ Apter describes the prime characteristics of the mobilization Let us turn to the conditions of a mobilization system. Here the emphasis is on realizing potentiality. As we have been using the term, it implies a hierarcliical system of authority. Instrumental values are elevated to the level of consummatory values, with the result that the goals of the state, particularly those of m.odernization and industrialization, become sacrosanct. Tlie postponement of iimnediate gratifications in consumption is identified as social discipline and is a required feature of the individuals' orientation to the community. Tlirift, saving and other forms of abstinence are linked to the creation of a future society in which abundance, personal dignity, and natural benevolence will be the conditions of life. Quite apart from the strong Utopian element to be found in all mobilization systems, which pi-ovides the moral basis of their social discipline, they are oriented toward the future. Tliey are for the youth, not the aged; the future, not the past. 9 Die characteristics of the integration system are similar to those of the mobilization system, with some significant variations. In general, the tendency is for the state to become an end in itself 7 David Apter, The Politics of Modernization (Chicago: University of Chicago Press, 1965); for typologies relating stages of development and regime forms, see also Barrington Moore, Tlie So cial Ori gins of Dictato rship and Democracy (Boston: Beacon Press, 1966), and E.B. Haas, Beyond the Nat i on-State (Stanford: Stanford University Press, 1964) . 8 F.L. Carsten, T he Rise of Fascism (Los i\ngeles: University of California Press, 1967). 9 Apter, op. £it^. , 359.

PAGE 162

rather than a means to an end. I'/ithin the integration system, personal abstinence is valued as a contribution to the state, not as a contribution to a future society. Uliile instrumental values remain raised to the level of consuiraiiatory values, the avov/ed pui-pose is consei'vation and integration in the framework of individual submergence within the state. Both systems utilize the party of solidarity as the main means of assuring authoritative leadership in periods of social flux. Both systems have authoritarian emphases, because both systems are crisis based. Ilie difference between the systems is based on responses to radically different social crises. The mobilization system is responding to the crisis of mobi-lization of masses of unmobilized individuals. Prime examples of this form of crisis mobilization are Russia, China and Yugoslavia. In these systems, war and foreign invasion mobilized masses of individuals in response to wartime dislocation and the activities of the conquering armies. In the integration systems, notably Germany, Italy, and Japan during the thirties and forties, the crises were economic and political in nature, the responses of inadequately integrated social and political systei.is which faced crises after having achieved relatively high levels of social mobilization. Tliis form of crisis could not strike an economically backward society, simply because the level of economic 10 Apter, 02_. cit . , 359.

PAGE 163

153 interdependence u'ould not be high enough to be so drastically affected. Both systems are faced '.vith extremely high levels of tension, but the fundamental causes of the tension are different and tlie response? of equally potent ideologies are con-espondingly different. The ideology of the mobilization system is based on the necessity of dealing with both the social and political problems attendant upon any attempt to rapidly create a system capable of meeting the needs of a no longer quiescent populace. The ideology of the integration system is based on the necessity of securing a set of political and economic relationships which are pre-existing, but currently inoperative due to severe crisis or stress. Table XXII should more clearly illustrate the basis for this argument. Tne coluirai labeled percent in agriculture indicates the level of agricultural employment at approximately the period in which the crisis reached its peak. Table XXII Relationships of Mobilization Levels and System Types as Crisis Outcomes in Six Nations Percent in Res pons e Agriculture Mobilization System Integration System More than RUSSIA 50 percent CHINA YUGOSLAVIA Less thaji ITALY SO percent GERMANY JAPAN The preceding discussion reveals a fruitful topic for future research in com.bining the model of a stable developmental s>stera

PAGE 164

154 presented here, with alternative forms of crises at varying lev61s of mobilization, attempting to understand the interaction betv.-cen processes and events which produce alternative regime forms. Pye jiotes tl\at another definition suggests that political development is the political prerequisite of economic development. Tliis model of development indicates that this view is only partially correct. At any time T, the current state of the political system partially determines the possibilities for future economic growth. At the same time, liowever, the model clearly indicates tliat the current level of the economic system determines the future capacities of the political system. It is, therefore, improper to define political development in terms of its relation to economic developmient since the relationship is one of interdependence, ihe same type of criticism can be registered against tlie definition of development as the politics typical of industrial societies. Tliis definition is too encompassing. It is clear that some form of increasing wealth is necessary to development; however, it need not be industrialization. 'Aliat is significant is the general stability of the process, not the type of v-'ealth increase. Two further partial definitions of development have been 12 combined in this presentation to form one usable definition. One definition sees political developmenc as administrative and legal 11 Pye, op. cit., 33. 52 . Ibid. . 38-';0.

PAGE 165

155 development, the other as mass mobilization and participation. Tliese tv.'o definitions are analagous to the two fonns of political growth defined previously, permeance and congruence. It should again be reiterated that political development can occur without democratic political participation. Our definition of congruence emphasizes state action insuring equality of opportunity. We have indexed congruence growth by democratic political participation given the extremely low probability of any state being able to effectively anticipate demands without the information provided by effective mass participation. The effect of this definition is to render doubtful the significance of one further definition noted by Pye. This definition 13 sees political development as the building of democracy. Democracy is not necessary to development. It is likely to prove the only practical solution to the necessity for information in a complex system, but it is not the only possible means by which development could occur. Tlie central inadequacy of these alternative definitions is that they define political development primarily, if not wholly, in terms of political chai-acteristics . If political development is not defined in terms of the relationship between the political and social systems, then it is possible to define development in many different ways, and there is no method by which we may judge the adequacy of the 3 Ibid., 40-41.

PAGE 166

156 definition except by personal prefei'ence. If development is defined as a pai'ti.cular effect of the political system on the economic system, a vast range of social relationships obviously affected by the state are largely ignored. [f development is defined in tenns of the effect of political activity on the social system, tlien we may judge the level of development of political systems v;ith widely varying characteristics. This line of argument sheds some light on the recent findings 14 of Thomas Dye in the realm of public policy analysis. Tlie model indicates tliat the level of government activity is affected by the level of industrialization and the level of social integration. Systems faced witli similar problems will, of necessity, have to expend similar quantities of effort in order to resolve those problems. Tlie "political variables" utilized by Dye relate largely to organizational cliaracteristics of the j\merican state political systems. Tlie only meaningful question in regard to policy outputs (permeance growth) is whether or not the social problems were met -how they were met is largely irrelevant, as Dye's analysis indicates. If Dye had explicitly included time in his analysis and allowed for future political response to current outputs, his findings would not have indicated as great a disjunction between policy outputs and system characteristics. One fundamentally important conclusion may be drawn from this research. The process of political development is inalterably time consuming. It can be speeded up to some degroe, but true political 14 Thiomas R. Dye, Politics, Economi cs and the Public: Policy Ou tcomes in Ameri c an States ( Chicago: Rand, McNally and Company, 1966).

PAGE 167

157 development, not mere permeance growth, requires long periods of time. Even with the whole-hearted desire of an elite to foster political development, the process itself requires, at a probable minimum, two generations. In the early stages of development, as indicated by the model, the linchpin of the process is increasing individual adaptability. It is, thei-efore, clear that the time necessary to establish the social infrastructures required to promote increasing adaptability is the limiting time for increasing political development. Methods of compensating for the general lack of adaptability, such as the politicization of communal structures suggested by V.'einer, may be effective for some time. However, since we are dealing with a social process which increases social complexity, these forms of traditional stiricturcs would seem to have limited utility. More "m.odern" forms of organization such as mass parties also have limited utility if the membership is uneducated. They may be mobilized for certain forms of mass participation, but their effectiveness will depend entirely upon the abilities of the cadres. Tlie ceiitral difficulty of any organization which depends on non-adaptable individuals for mass membership is their volatility; politically mobilized but unable to personally deal with the frustrations of a complex environment, they will tend to project these frustrations into the political 15 Myron Vveiner, The Pol i tics of Scarcity: Public Pressu re and Political Response in India (Chicago: Chicago University Press, 1962).

PAGE 168

158 sphere. Deliberate mobilization of masses of individuals, not • psychically capable of dealing with an entirely new range of problems, would seem to be the least likely method for achieving political development. Tliis conclusion may seem anti -democratic, but premature "democratization" in tlie form of intense mass participation prior to general adaptability will most probably force the system into a retreat from effective participation as a consequence of its inability to meet mass demands. It is quite clear from the model of development that the political sub-units of American society effectively created their own future support by fostering mass education. Had general political participation been encouraged prior to effective general adaptability, the results of this research would have probably been very different, llie high correlation between education and participation in the early stages of the model indicate that, if left to himself, the average individual will tend not to participate unless he has the selfconfidence acquired through education. Tnere is no method for deciding which economic organization for wealth production is most likely to insure political development. The model makes clear only that the society must begin to increase disposable wealth before the process will begin in earnest. The model does indicate, however, that if an elite attempts to speed up the process of economic and social development, tliey must make haste slowly. Any form of economic oi'ganization which requires very rapid rates of change is likely to lead to very high levels of social tension.

PAGE 169

159 tlie one factor most likely to reduce the ultimate possibility of entering upon the process of political development. Neither free enterprise, nor the various types of command economies, seems to require rapid rates of change; they are methods of organization for production and distribution, but do not decree the rate at which change will take place. Hence, the rate of change seems to be fundamentally an elite decision in one case, and a result of the absorptive capacity of the society in the other. The basis of choice does not, tlierefore, really depend on the virtues of one form over the other. It is quite clear, as indicated by tlie model, and by general experience, that there is a close linkage between the general social system's capacities and the capabilities available for permeance growth. Tlierefore, there will likely be an extremely great temptation to effect social and economic development as rapidly as possible, in order to further increase political permeance. If the decision is to foster true political development, then the rate of change in the economic and social spheres must be kept to a pace commensurate with increasing individual adaptability. Balanced growth is the crux of the process. For economic and social change to lead to political development, the rate of change and the consistency of the process must allow recently mobilized portions of the populace to achieve a high enough level of adaptability to deal independently with the frustrations of their newly complex environment. For those who wish to produce "strategies" of developruent, this empirically identified model of the process clearly

PAGE 170

160 indicates the necessity, not for achieving lapid economic modernization, but for achieving a balance between the rate of social cliange and tlie developing capacity of the society to independently adapt to that change.

PAGE 171

APPENDIX Using summed standard scores to create a new index introduces an arbitrary term which is a function of the number of original variables used to create the index and the level of association between ^ 2 these original variables. It should be clear that I (Z^) is equal i=l to N. The standard deviation of the Z distribution is given by: [1] Tn addition, we know that the ^ Z 0, xo Z will be zero. Wien N is i=l N 2 laroe. S will be one, and it is, therefore, obvious that z (Z . ) must z i-1 be equal to N. We know that the standard deviation of the Z distribution is one. The standard deviation of the sum of two or more Z distributions is not one. To illustrate, we may use the two variable case represented by equation 2. [2] •1 ^2 ^ 2 E [(Z^ = Z^) -(Z^ + Z^)]' i-1 N ;'e know that Z Z = by definition, and equation 2 becomes; 161

PAGE 172

162 Evaluating equation 3,. we arrive at the result in equation 4, [4] S + = / 2[N + ( E Z Z )] 1 ^2 / i=l ^ ^ If it is the case that the original variables exhibited a very high positive correlation, then Z-. would begin to approximate Z2, due to the special nature of the Z distribution. In this case, the quantity N N E Zj^Z^ approaches the quantity E (Z,) , which we recognize as N. i-1 " i-1 Tlius, in the special case where the original variables used to form the index are perfectly correlated, equation 4 becomes: / 2(N + y N and [5] S^^ . ^^ ^ / 2(N . N) [6] S^ 1 2 Obviously, for any nimiber of standardized variables k, an index created by summing tlie variables will hhve a standard deviation which is less than, or equal to /k^ , or k times the single index standard deviation. TIius, if the standard scores used are Z scores, and two variables are used to form the index, the highest value the standard deviation of the index can assume will be two. It should now be clear that when standard scores are used in the computation of indices, and these indices are used in regression analysis, the meaning of the resulting correlation and regression coefficients are ambiguous.

PAGE 173

163 We may relate r, to b _ by the following equation: 5 In tlie special case v.'hei'e S • = S , the ratio x_ becomes one. In X y y this case, the above equation reduces to: [8] T^y = by^ Mien \;e use one standardized variable for x and one standardized variable for y, we automatically create this special case. This special case is often very useful and is usually referred to as the bet; weight. However, the demonstration above points out the very serious result of si-unming standard scores. In this case, the ratio x is related to the numiber of variables used to compose each index. Rather than measuring the relationship of the variance of index X to the variance of index Y, the correlation coefficient is now largely determined by the internal relations and nimiber of original variables included in each index.

PAGE 174

BIBLIOGRAPHY Alher, ilayv.ood R. , "Causal Infex-once and Political Analysis," in J. Bernd [editoi'). Mathematical Applica ti ons in Political Scie nce II (Dallas: Southern Methodist Press, r966'j~ Almond, Gabriel A., "Political Development: Analytical and Noryiiative Perspectives," Comp arative Political Studies (January, 1969), 447-469. Almond, Gabriel A., and G. Binghsm Powell, Compar ative Politics: A Developmental Appro ach (Boston: Little, Brown and Company, 1965). AliP.ond, Gabriel A., and Sidney Verba, The Civic Culture: Polit ical Attitud es and D emo crac y in Five Nations (Boston: Little, Brown and Company, 1965) . Apter, David, The Politics of Modemization (Chicago: Chicago University Press, igTsJ^ " Eendix, Pcinhard, N ation-B uilding and Citiz ensliip (Garden City, N.J. : Doubleday-Anchor, 1969). Blalock, Huliert M. , Ca usal Inferences in N o n-Gxperimental Research (Chapel Hill: University of North Carolina Press, 1964). Blalock, Hubert {•!., "Causal Inferences, Closed Population and Measures of As;~ociaticn," American Political Science Review (March, 1967), 130-135, Burnliam, V.'alter D., "The Changing Shape of the /Vmerican Political Universe," America n Political Science Review (March, 1965), 7-28. Burnham, Ivalter D. , Presid ential Ballots: 1836-1892 (Baltimore: Johns Hopkins university Press, 1955).' Burrowcs , R. , "Multiple Time-Series /uialysis of Nation-Level Data," Comparative Political Studies (January, 1970), 465-480. Carstsn, r.L., The Rise of Fascism (Los Angeles: University of Califor!ii?. Press, 1967). Cattell, R.B., H. Bruel, and H.P. Hartman, "An Attempt at More Refined Definitions of the Cultural Dimensions of Syntality in Modem Nations," .'-\ mericr.n Sociological Rev iew (August, 1952), 408-421. Coleman, J.S., "The Mathematical Study of Oiange," in H. Blalock, and A. Blalock (editor's), Methodology in Social Rese a rch (New York: McGraw-Hill Book Company, 1968)". 164

PAGE 175

Colexan, J.S. (editor). Education and Political Development (Princeton: Princeton University Press, 1965). Cutriglit, Philips, "National Political Development: Measurement and .\nalysis," in Nelson Polsby et_ al_. (editors) , Politics and Social Life (Boston: Houghton Mifflin, 1963). Davies, loan. Social Mobility and Political Change ( New York: Praeger Publishers, 1970). Davis, Harold Tliayer, Tlie .Analysis of Economic Time Series (Bloomingtonj Indiana: Tlie Principia Press , 1941) . Davis, Otto A., M.A.H. Dempster, and Aaron Wildowsky, "A Theory of the Budgetary Process," Ajnerican Poli tical Science Review (September, 1966) , 529-547! De Sch'.-einitz, Karl, Jr., "Growth, Development and Political Modernization," W orld Politi cs (July, 1970), 518-540. De Schweinitz, Karl, Jr., "On Measuring Political Performance," Com.par ative Politic al Studies^ (January, 1970), 503-511. Deutsch, Karl, "Social ;-!obilization and Political Development," Ameri can Political Science Review (September, 1961), 493-514. Deutsch, Karl, Nat ionalism and Soc ial Communication. An Inquiry into_ the Foundations of Nationality (New York: Wiley, 1953). Dye, Thomas R. , Politics, Economics and th e Public: Policy Outcomes in the American ~States (Oiicago: Rand, McNally and Company, 1966). Easton, David, The Political Sys tem: An In quiry into th e Sta te of Pol iticaT ~Science~(Tre"w~Yoi-k: Alfred A. Knopf, 19"53) . Eckstein, Alexander, "Economic Development and Political Ch.ange in CoCiP.unist Systems," World Politics (July, 1970) ,475-495. Eisenstadt, S.N., "Breakdowns of Modernization," E conomic Develo pment and C ultural Chan ge (July, 1964), 345-367. Peith, Herbert, "Indonesia's Political S\Tnbols and Tneir Wielders," World Politics (October, 1963), 79-97. Forbes, H.D., and E.R. Tufte, "A Note of Caution in Causal Modeling," -American rolitical Science Review (December, 1968), 1258-1264. Haas, Ernst B. , Beyond the Nation -S t ate (Stanford: Stanford University Press, 19647^ — -

PAGE 176

166 liagen, Everett IE., On t he Theory of Social Change (Honewood, 111.: Tlie Dorsey Press, 1962). Hauser, Philip M., "The Social, llconomic and Technological Problems of Rapid Urbanization," in B.F. Hoselitz, and W.E. Moore (editors). Industr ialization and Society (UNESCO: Mouton, 1963). Holt, Robert T. and John E. Turner, Ihe Political B asis of Economic Devel opment (Princeton: D. Van Nostrand Company, Inc., 1966). Hoselitz, Bert F. , "Economic Growth and Development: Non-Economic Factors in Economic Development," American Econom ic Review (May, 1957), 28-41. Huntington, S.P., Politi cal Order in Changing S ociety (New Haven: Yale University Press, 1968). Ilchman, Ivarren F. and Norman T. Uphoff, The Political Economy of Chang_e (Los Angeles: University of California Press, 1969). Inter-University Consortium for Political Research, Hist orical Data Archives, Statistics of the American Con g ressional Elections, 1824-1968 (Computer tape)^ Kelley, Truman L. , Fundamen tals of Statist ics (Cambridge, Mass.: Harvard University Press,~~r947)T ~~ Lasswell, H'irold D. , Politics: V.lio Gets Vvliat , I'.hen ajid How (New York: McGra'.s-Hill Book Company, 19 36). Lenski, Gerhard E. , Power and Privilege: A Theory of Social Stratifica tion (New York: McGraw-Hill Book Company, 1966). Lerner, i^iaaiel, "Communications Systems and Social Systems: A Statistical Ex|jloration in History and Policy," Behavioral Science (March, 1957) ,266-275. Lenier, Daniel, The Passing of Traditional Society: Modernizing the Middle East (New York: The Free Press, 1958). Lerner, Daniel, "Toward a Communications Tlieory of Modernization," in Lucian W. Pye (editor), Communicatioiis and Political Development (Princeton: Princeton University Press, 1965). L i p 3 e t , S e y mo u r M . , The F irst New Nati o n_: _The_ Urn. ted State s in Histo r ical aiKi Comparative Pers pect ive (Garden Citv, W.J.: Doubleday-Anchor, 196 7). '^~~ Lipset, Sev'-mour M. , "Som,e Social Requisites of Democracy: Economic Development and Political Legitimacy," A merica n Politic al Sci ence Revie-.j (March, 1959), 69-105,

PAGE 177

167 McClelland, David, The Achie ving Society (Pi'inccton: D. Van Nostrand Company, Inc . ,-1961) . McCrone, Donald J., and Charles F. Cnudde, "Towai-d a Coirimuni cat ions Theory of Democratic Political Development: A Causal Model," American Political Science Review (March, 1967), 72-79. Marquette, Jesse F., "Social Mobilization and the Philippine Political System," Co:.;p arative Political Studies (in press). Morton, Robert K. , Social Theory and Social Structure (Glencoe, 111.: V.ie Free Press "of "Clencoe, 1949). Milbrath, Lester W. , Politic a l Pai'ticipation. How and Wiy Do People Get Involved in Politics? (Chicago: Rand, McNally and Company, 1965). Moore, Harrington, Tlie Social Origins of Dictatorship and Democracy (Boston: Be aeon Press ,~1966) . " ~ Moore, Wilbert E., "Changes in Occupational Structures," in Neil J. Smelser and Seymour M, Lipset (editors). Social S tructu re and Mobility in Economic Develop ment (Chicago: Aldine Publishing ComjDany, 1966) . Nash, Manning, Pri mitive an d Peasan t Ec onomic Sy stems (San Francisco: Chandler Publislting Company, 1966). Ne tt 1 , J . P . , Political Mobi lization: A Sociologic al Analysis of Methods a;u i Concepts (New York: Basic Books, 1967). Neubauer, D.E., "Some Conditions of Democracy," American Political Scien ce R_e\2ew_('^'^ member, 1967), 1002-1009. Nie, Norman H. , G. Bingham Powell, Jr., and Kenneth Prewitt, "Social Structure ajid Political Participation: Developmental Relationsliips I," American Political Science Review (June, 1969), 361-578. Nie, Norm.an H. , G. Bingham Powell, Jr., and Kenneth Prewitt, "Social Structure and Political Participation: Developmental Relationships 11," Am erican Politic al Science Review (Septemiber, 1969), 80S-328. Nie, Norman H. et_ al.. Statist ical Pack age for the Social Sciences (New York: McGraw"Hill Book~Company , 1970) ^^ ~ Organski, A.F.K., The Stag es of Po l iti cal D evelopment (New York: Alfred A. Knopf, 1965). ~

PAGE 178

168 rackeniiain, R. , "Approaches to the Study of Political Development;," l.'o rld P o litics (October, 1964), 108-124. Parsons, Talcott, "Evolutionary Universals in Society," American Sociol ogical Review (June, 1964], 339-357. Pennock, Roland, "Political Development, Political Systems, and Political Goods," Worl d Polit ics (April, 1966), 415-434. Petersen, Sven, A Statistical History of the American Presidenti al Elections (New York: Ungar Publisliers, 1963). Pomper, Gerald M. , Elections in iVjnerica: Control and Influence in Dcnoc riitic Politics (New York: Dodd, Mead and Company, 1968). P)'e, Lucian IV., Aspects of Political Development (Boston: Little, Brown and Company, 1966). Pye, lucian IV. , Politics, Personality and Nation-Building: Burma's Sea rch for Identity (New Haven: Yale University Press, 1962). Reisman, David, Tlie Lonely Crowd: A Stu dy of the Changing American Clia racter (New Haven: Yale University Press, 1964). Riggs, F.W., "Agraria and Industria: Toward a Typology of Comparative Administration," in W. Siffin (editor). Towar d the Comparative Stud y of Public Administration (Dloomington: Indiana University Press, 1959). Rogers, Everett M. , Modernization Among Peasants: The Impact of Comijnu ni cation (New York: Holt, Rinehart and Winston, Inc., 1969) Rosto\;, Walter W. , The Stages of Economic Growth: A Non-Communist M anifesto (Cambridge, England: Cambridge University Press, 1966) Schromm, W. , M ass Media and National Development (Stanford: Stanford university Press, 1964). Sm?lspr, Neil J., and Seymour M. Lipset, "Social Sti-ucture, f!obility aiid Development," in Neil Smelser and S.M. Lipset (editors). Social Structure and Mob ility i n Eco nomic Devel opment (Chicago : Aldine Publishing Company, 1966). Smith, Arthur K. , Jr., "Socioeconomic Development and Political De-T!ocracy: A Causal /\nalysis," Midwest Journal of P olitical Science (February, 1969), 95-125. Tonter, R. , "Toward a Theory of Political Development," Midwest Journal of__Political Science (May, 1967), 145-172.

PAGE 179

169 Tneodorson, George A., "Acceptance of Industrialization and Its Attendajit Consequences for the Social Patterns of Non-Western Societies," Ainerican Sociolo gi cal Review (October, 1953), 477-484. Tumin, Melvin M. , Social Stratification: Tne Forms ana Functions of Inequality (Englewood Cliffs, N.J.: Prentice-Hall, 1967). U S Bureau of the Census, U.S. Census of Popu lation (Washington, D.C.: U.S. Government Printing Office, 1872, 1883, 1895, 1901, 1913, 1922, 1933, 1943, 1952, 1962). U.S. Bureau of the Census, Statistica l Abstract of the United States (Washington, D.C.: U.S. Govemment Printing Office, 1883, 1893, 1903, 1913, 1923, 1933, 1943, 1953, 1963). U S Commissioner of Education, Annual Report of the Commissioner of Ed ucation (Washington, D.C.: U.S. Government Printing Office, 1871, 1881). U.S. Postmaster General, Aimual Report of the Postmaster General (WashiVigton, D.C.: U.S. Government Printing Office, 1871, 1881). Vebien, Thorstein, The -n^eory of th e Leisui-^Cj^ss_ (New York: Viking Press, 1912). Weber, Max, "Class, Status and Party," in R. Bendix and S.n. Lipset (editors), Class, Status and Power (Glencoe, 111.: Tlie Free Press of Glencoe, 1953). Weiner, Myron, The Politics of Scarcity: Public Pressure and Political Re sponse in India (Oiicago: University of Chicago Press, 1962). Winham, Gilbert R. , "Political Development and Leracr's Theory: Further 'Test of a Causal Model," Ajperican Political Science Review (September, 1970), 810-818.

PAGE 180

BIOGRAPHICAL SKETCH Jesse F. Marquette vvas born on September 4, 1945, in Patorson, New Jersey. In 1963 he graduated from Central High School in >!iani, Florida, and continued his education at the University cf Florida, receiving his Baclielor of Arts degree in 1967, and Master of Arts degree in 1968, both with majors in Political Science. lie has pursued liis Master of Arts and Doctor of Philosophy degrees with the assistance of an NOEA Title IV Fellowsliip. The autiior is a member of Pi Sigma Alplia, national political science honorary, and the American Political Science Association. He also liolds memberships in the Soutliern Political Science Association and the American Sociological Association. 170

PAGE 181

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. r7l r^-rx 1\ \ 1 Alfred B. Clubok, Chairman '>.. Associate Professor of Political Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy, Manning J. jDauer Professor ; and' ChaiiTiian, Political Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosopliy. Joseph S. Vandiver Professor of Sociology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality as a dissertation "^or the degree of Doctor of Philosophy. -^ John W. Spahier Professor, of Political Science

PAGE 182

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scliolarly presentation and is fully adequate^ in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Frank J. MungeiV Professor of Pottjtical This dissertation was submitted to the Department of Political Science in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for tlie degree of Doctor of Philosophy. Dean, Graduate School August, 1971


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EDNV2ICP0_YQ7RQN INGEST_TIME 2017-07-14T21:44:36Z PACKAGE UF00097679_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES