Robert A. Dahl - Polyarchy - Participation and Opposition-Yale University Press (1972) PDF

Cover
Page i
POLYARCHY
Page ii
This page intentionally left blank
Page iii
POLYARCHY
Participation and Opposition
by Robert A. Dahl
Page iv
Copyright © 1971 by Yale University.
All rights reserved. This book may not be
reproduced, in whole or in part, in any form
(beyond that copying permitted by Sections 107
and 108 of the U.S. Copyright Law and except by
reviewers for the public press), without written
permission from the publishers.
Library of Congress catalog card number: 70140524
ISBN: 0300015658 (paper)
Designed by Sally Hargrove Sullivan
and set in Times Roman type.
Printed in the United States of America by
BookCrafters, Inc., Chelsea, Michigan.
26 25 24 23 22 21 20
Page v
To the memory of
Mary
and her hopes
Page vi
Page vii
CONTENTS
Acknowledgements ix
1. Democratization and Public Opposition 1
2. Does Polyarchy Matter? 17
3. Historical Sequences 33
4. The Socioeconomic Order: Concentration or Dispersion 48
5. The Socioeconomic Order: Level of Development 62
6. Equalities and Inequalities 81
7. Subcultures, Cleavage Patterns, and Governmental Effectiveness 105
8. The Beliefs of Political Activists 124
9. Foreign Control 189
10. The Theory: Summary and Qualifications 202
11. Postscript: Some Implications for Strategies of Change 208
Appendixes
A. 114 Countries Classified According to Eligibility to Participate in Elections and Degree of
Opportunity for Public Opposition 231
B. Contemporary Polyarchies, circa 1969 246
Index 251
Page viii
Page ix
ACKNOWLEDGMENTS
I wish to express my appreciation to the Rockefeller Foundation for financial support that has helped to make possible this volume and others I have undertaken on the
subject of governments and oppositions. I am also indebted to the Concilium of Yale University for financial assistance from the Henry L. Stimson Fund for Research in
World Affairs and to Yale University for a Senior Faculty Fellowship that enabled me to complete the manuscript.
The comments of a number of colleagues who have read various drafts in whole or in part have been particularly helpful. These include Frederick Barghoorn, Robert
Dix, William Foltz, Michael Leiserson, Rajni Kothari, Juan Linz, and Gordon Skilling, all of whom are coauthors of a companion volume, Regimes and Oppositions,
and also Hans Daalder, Joseph LaPalombara, Val Lorwin, Nelson Polsby, and Stein Rokkan. I wish to thank students in my graduate and undergraduate courses for
allowing themselves to be subjected to a draft of the manuscript and providing reactions that were often extremely useful.
For the patience, skill, and speed with which they transformed my drafts into readable typescript, I wish to express both my thanks and my admiration to Mrs. Betty
Mauceri, Mrs. Miriam Swanson, and Mrs. Nancy Hoskins. And once again I have increased my debt to Mrs. Marian Ash for her editorial supervision at the Yale
University Press.
R.A.D.
Page x
Page 1
1. DEMOCRATIZATION
AND PUBLIC OPPOSITION
Given a regime in which the opponents of the government cannot openly and legally organize into political parties in order to oppose the government in free and fair
elections, what conditions favor or impede a transformation into a regime in which they can? That is the question with which this book is concerned.
Concepts
Since the development of a political system that allows for opposition, rivalry, or competition between a government and its opponents is an important aspect of
democratization, this book is ecessarily about one aspect of democratization. But the two processes—democratization and the development of public opposition—are
not, in my view, identical. A full description of the differences could lead us into a tedious exploration of a semantic bog. To avoid this detour, I hope I may be allowed
to indicate rather summarily some of my assumptions without much in the way of defense or elaboration.
I assume that a key characteristic of a democracy is the continuing responsiveness of the government to the preferences of its citizens, considered as political equals.
What
Page 2
other characteristics might be required for a system to be strictly democratic, I do not intend to consider here. In this book I should like to reserve the term
‘‘democracy” for a political system one of the characteristics of which is the quality of being completely or almost completely responsive to all its citizens. Whether such
a system actually exists, has existed, or can exist need not concern us for the moment. Surely one can conceive a hypothetical system of this kind; such a conception
has served as an ideal, or part of an ideal, for many people. As a hypothetical system, one end of a scale, or a limiting state of affairs, it can (like a perfect vacuum)
serve as a basis for estimating the degree to which various systems approach this theoretical limit.
I assume further that in order for a government to continue over a period of time to be responsive to the preferences of its citizens, considered as political equals, all full
citizens must have unimpaired opportunities:
1. To formulate their preferences
2. To signify their preferences to their fellow citizens and the government by individual and collective action
3. To have their preferences weighed equally in the conduct of the government, that is, weighted with no discrimination because of the content or source of the
preference.
These, then, appear to me to be three necessary conditions for a democracy, though they are probably not sufficient. Next, I assume that for these three opportunities
to exist among a large number of people, such as the number of people who comprise most nationstates at the present time, the institutions of the society must provide
at least eight guarantees. These are indicated in table 1.1.
I am going to make the further assumption that the connections between the guarantees and the three fundamental
Page 3
opportunities are sufficiently evident to need no further elaboration here.1
Table 1.1. Some Requirements for a Democracy
among a Large Number of People
For the opportunity to: The following institutional guarantees are required:
I. Formulate preferences 1. Freedom to form and join organizations
2. Freedom of expression
3. Right to vote
4. Right of political leaders to compete for support
5. Alternative sources of information
II. Signify preferences 1. Freedom to form and join organizations
2. Freedom of expression
3. Right to vote
4. Eligibility for public office
7. Free and fair elections
III. Have preferences 1. Freedom to form and join organizations
weighted equally in 2. Freedom of expression
conduct of government 3. Right to vote
4. Eligibility for public office
5a. Right of political leaders to compete for votes
7. Free and fair elections
8. Institutions for making government policies depend on votes and other
expressions of preference
Now from examination of the list of eight institutional guarantees, it appears that they might provide us with a theoretical scale along which it would be possible to order
1. Some of the relationships are discussed in my A Preface to Democratic Theory (Chicago: University of Chicago Press, 1956), pp. 3–81, and in Robert A. Dahl and
Charles E. Lindblom, Politics, Economics and Welfare (New York: Harper, 1953), chaps. 10 and 11.
Page 4
different political systems. Upon closer examination, however, it appears that the eight guarantees might be fruitfully interpreted as constituting two somewhat different
theoretical dimensions of democratization.
1. Both historically and at the present time, regimes vary enormously in the extent to which the eight institutional conditions are openly available, publicly employed, and
fully guaranteed to at least some members of the political system who wish to contest the conduct of the government. Thus a scale reflecting these eight conditions
would enable us to compare different regimes according to the extent of permissible opposition, public contestation, or political competition.2 However, since a regime
might permit opposition to a very small or a very large proportion of the population, clearly we need a second dimension.
2. Both historically and contemporaneously, regimes also vary in the proportion of the population entitled to participate on a more or less equal plane in controlling and
contesting the conduct of the government: to participate, so to speak, in the system of public contestation. A scale reflecting the breadth of the right to participate in
public contestation would enable us to compare different regimes according to their inclusiveness.
The right to vote in free and fair elections, for example, partakes of both dimensions. When a regime grants this right to some of its citizens, it moves toward greater
public contestation. But the larger the proportion of citizens who enjoy the right, the more inclusive the regime.
Public contestation and inclusiveness vary somewhat independently. Britain had a highly developed system of public contestation by the end of the eighteenth century,
but only a miniscule fraction of the population was fully included in
2. Throughout this book the terms liberalization, political competition, competitive politics, public contestation, and public opposition are used interchangeably to refer
to this dimension, and regimes relatively high on this dimension are frequently referred to as competitive regimes.
Page 5
it until after the expansion of the suffrage in 1867 and 1884. Switzerland has one of the most fully developed systems of public contestation in the world. Probably few
people would challenge the view that the Swiss regime is highly “democratic.” Yet the feminine half of the Swiss population is still excluded from national elections. By
contrast, the USSR still has almost no system of public contestation, though it does have universal suffrage. In fact one of the most striking changes during this century
has been the virtual disappearance of an outright denial of the legitimacy of popular participation in government. Only a handful of countries have failed to grant at least
a ritualistic vote to their citizens and to hold at least nominal elections; even the most repressive dictators usually pay some lip service today to the legitimate right of the
people to participate in the government, that is, to participate in “governing” though not in public contestation.
Needles to say, in the absence of the right to oppose the right to ‘‘participate” is stripped of a very large part of the significance it has in a country where public
contestation exists. A country with universal suffrage and a completely repressive government would provide fewer opportunities for oppositions, surely, than a country
with a narrow suffrage but a highly tolerant government. Consequently, when countries are ranked solely according to their inclusiveness, not taking into account the
surrounding circumstances, the results are anomalous. Nonetheless, as long as we keep clearly in mind the fact that the extent of the “suffrage” or, more generally, the
right to participate indicates only one characteristic of systems, a characteristic that cannot be interpreted except in the context of other characteristics, it is useful to
distinguish between regimes according to their inclusiveness.
Suppose, then, that we think of democratization as made up of at least two dimensions: public contestation and the right to participate. (Figure 1.1) Doubtless most
readers believe that democratization involves more than these two di
Page 6
mensions; in a moment I shall discuss a third dimension. But I propose to limit the discussion here to these two. For the point has already emerged, I think: developing a
system of public contestation is not necessarily equivalent to full democratization.
To display the relationship between public contestation and democratization more clearly, let us now lay out the
FIGURE 1.1 Two Theoretical Dimensions of Democratization
two dimensions as in figure 1.2.3 Since a regime may be located, theoretically, anywhere in the space bounded by the two dimensions, it is at once obvious that our
terminology for regimes is almost hopelessly inadequate, for it is a terminology invariably based upon classifying rather than ranking. The space enclosed by our two
dimensions could of course be cut up into any number of cells, each of which might be given a name. But the purposes of this book make an elaborate typology
redundant. Let me instead provide a small vocabulary—a reasonable one, I hope—that will en
3. An array of 114 countries along these two dimensions will be found in appendix A, table A1.
Page 7
able me to speak precisely enough about the kinds of changes in regimes that I want to discuss.
Let me call a regime near the lower left corner of figure 1.2 a closed hegemony. If a hegemonic regime shifts upward, as along path I, then it is moving toward greater
public contestation. Without stretching language too far, one could say that a change in this direction involves the liberalization
FIGURE 1.2 Liberalization, Inclusiveness, and Democratization
of a regime; alternatively one might say that the regime becomes more competitive. If a regime changes to provide greater participation, as along path II, it might be said
to change toward greater popularization, or that it is becoming inclusive. A regime might change along one dimension and not the other. If we call a regime near the
upper left corner a competitive oligarchy, then path I represents a change from a closed hegemony to a competitive oligarchy. But a closed hegemony to a competitive
oligarchy. But a closed hegemony might also become more inclusive without
Page 8
liberalizing, i.e., without increasing the opportunities for public contestation, as along path II. In this case the regime changes from a closed to an inclusive hegemony.
Democracy might be conceived of as lying at the upper right corner. But since democracy may involve more dimensions than the two in figure 1.2, and since (in my
view) no large system in the real world is fully democratized, I prefer to call real world systems that are closest to the upper right corner polyarchies. Any change in a
regime that moves it upward and to the right, for example along path III, may be said to represent some degree of democratization. Polyarchies, then, may be thought
of as relatively (but incompletely) democratized regimes, or, to put it in another way, polyarchies are regimes that have been substantially popularized and liberalized,
that is, highly inclusive and extensively open to public contestation.
You will notice that although I have given names to regimes lying near the four corners, the large space in the middle of the figure is not named, nor is it subdivided. The
absence of names partly reflects the historic tendency to classify regimes in terms of extreme types; it also reflects my own desire to avoid redundant terminology. The
lack of nomenclature does not mean a lack of regimes; in fact, perhaps the preponderant number of national regimes in the world today would fall into the midarea.
Many significant changes in regimes, then, involve shifts within, into, or out of this important central area, as these regimes become more (or less) inclusive and increase
(or reduce) opportunities for public contestation. In order to refer to regimes in this large middle area, I shall sometimes resort to the terms near or nearly: a nearly
hegemonic regime has somewhat more opportunities for public contestation than a hegemonic regime; a nearpolyarchy could be quite inclusive but would have more
severe restrictions on public contestation than a full polyarchy, or it might provide opportunities for public
Page 9
contestation comparable to those of a full polyarchy and yet be somewhat less inclusive.4
The need to use terms like these later on in this book testifies to the utility of classification; the arbitrariness of the boundaries between “full” and “near” testifies to the
inadequacy of any classification. So long as we keep firmly in mind that the terms are useful but rather arbitrary ways of dividing up the space in figure 1.2, the concepts
will serve their purpose.
4. The problem of terminology is formidable, since it seems impossible to find terms already in use that do not carry with them a large freight of ambiguity and surplus
meaning. The reader should remind himself that the terms used here are employed throughout the book, to the best of my ability, only with the meanings indicated in the
preceding paragraphs. Some readers will doubtless resist the term polyarchy as an alternative to the word democracy, but it is important to maintain the distinction
between democracy as an ideal system and the institutional arrangements that have come to be regarded as a kind of imperfect approximation of an ideal, and
experience shows, I believe, that when the same term is used for both, needless confusion and essentially irrelevant semantic arguments get in the way of the analysis.
At the opposite corner, hegemony is not altogether satisfactory; yet given the meaning I have indicated, the term hegemonic seems to me more appropriate than
hierarchical, monocratic, absolutist, autocratic, despotic, authoritarian, totalitarian, etc. My use of the term ‘‘contestation” in “public contestation” is well within normal
(if infrequent) English usage; in English contestation means to contest, which means to make something the subject of dispute, contention, or litigation, and its most
immediate synonyms are to dispute, challenge, or vie. The utility of the term was, however, first suggested to me by Bertrand de Jouvenel’s “The Means of
Contestation,” Government and Opposition 1 (January 1966): 15574. Jouvenel’s usage is similar to my own, as is the identical French term he used in the original,
meaning: débat, objection, conflit, opposition. In the same issue of this journal, however, Ghita Ionescu (“Control and Contestation in Some OneParty States” pp.
240–50) uses the term in its narrower but currently quite common meaning as “the antisystem, basic and permanent postulates of any opposition on the grounds of
fundamental, dichotomic differences of opinion and ideologies” (p. 241). Clearly this is a more restricted definition of the concept than the one I use here and that, I
believe, Jouvenel uses in his essay.
Page 10
The Question Restated
The question with which this chapter opens can now be restated as follows:
1. What conditions increase or decrease the chances of democratizing a hegemonic or nearly hegemonic regime?
2. More specifically, what factors increase or decrease the chances of public contestation?
3. Even more specifically, what factors increase or decrease the chances of public contestation in a highly inclusive regime, that is, a polyarchy?
Qualifications
This book, then, is about the conditions under which systems of public contestation are likely to develop and exist. Because public contestation is an aspect of
democratization, this book is necessarily to some extent about democratization, as I noted at the beginning of this chapter. But it is important to keep in mind that the
focus here excludes a number of important matters that would be considered in an analysis of democratization.
It is convenient to think of democratization as consisting of several broad historical transformations. One is the transformation of hegemonies and competitive oligarchies
into nearpolyarchies. This was, in essence, the process at work in the Western world during the nineteenth century. A second is the transformation of nearpolyarchies
into full polyarchies. This was what occurred in Europe in the three decades or so that spanned the end of the last century and the First World War. A third is the
further democratization of full polyarchies. This historical process can perhaps be dated to the rapid development of the democratic welfare state after the onset of the
Great Depression; interrupted by the Second World War, the process seems to have renewed itself
Page 11
in the late 1960s in the form of rapidly rising demands, notably among young people, for the democratization of a variety of social institutions.
This book is concerned with the first and second of these transformations but not the third.5 Whether it prospers or fails, the third wave of democratization will surely
prove as important as the others. Since it will take place only in the most “advanced” countries and will help to shape the character of life in the “advanced” countries in
the twentyfirst century, to many people in these countries the third wave may well seem more important than the others. Yet most of the world still lies beyond the
possibility of this particular transformation. Of the 140 nominally independent countries existing in 1969, about two dozen were highly inclusive and had highly
developed systems of public contestation: they were, in short, inclusive polyarchies. Perhaps another dozen or fewer were nearpolyarchies within reasonable reach of
full polyarchy. It is in these three dozen countries that the third wave must occur. Whether some nonpolyarchies can overleap the institutions of polyarchy and arrive
somehow at a fuller democratization than now exists in the polyarchies, as ideologues sometimes promise, seems remote, in the light of the analysis that follows. For
most countries, then, the first and second stages of democratization—not the third—will be the most relevant.
The focus of this book is, in fact, even narrower than an analysis of the first two stages of democratization. I have referred to “regimes” and ‘‘systems of public
contestation.” But so far I have not specified the level of the polity at which regimes and public contestation may be effective. Let me then emphasize at once that the
analysis here deals with national regimes, that is, regimes taken at the level of the
5. I have dealt with some aspects of the third in After the Revolution? Authority in a Good Society (New Haven: Yale University Press, 1970).
Page 12
country, or, if you will, the legally independent state, or, to use less appropriate terms, the nation or the nationstate. Doubtless some of the analysis could be applied to
subordinate levels of political and social organization, such as municipalities, provinces, trade unions, firms, churches, and the like; perhaps some of it might even be
relevant to the polities that are emerging at more inclusive levels—international organizations of various kinds. But the argument is specifically developed only with
respect to national regimes.
Again, this would be a grave omission in a book about democratization. Even from the perspective of public contestation, the omission is important. For casual
observation suggests that countries differ in the extent to which they furnish opportunities for contestation and participation in the processes not only of the national
government but of various subordinate governmental and social organizations as well. Now to the extent that gross differences in the general characteristics of
subnational units appear to be associated with differences in the nature of the national regime (for example, whether it is a polyarchy or not), I shall try to take these into
account in the analysis.
Yet it might seem reasonable to insist that the analysis ought to go a good deal further. A full description of the opportunities available for participation and contestation
within a country surely requires one to say something about the opportunities available within subnational units. The extraordinary attempt in Yugoslavia to grant a large
measure of selfgovernment in subnational units means that the opportunities for participation and contestation are greater in that country, despite the oneparty regime,
than, let us say, in Argentina or Brazil. An inclusive view of the matter, then, would require one to pay attention to all the possibilities suggested in figure 1.3. Indeed a
number of recent critics of incomplete democratization in polyarchies contend that while polyarchies may be competitive at the national level a great
Page 13
many of the subnational organizations, particularly private associations, are hegemonic or oligarchic.6
Important as the task is of moving beyond the description of the national regime to the subnational units, at present the attempt to examine a fairly large number of
countries would I think require an analysis so complex and would encounter problems of data so overwhelming as to make the
FIGURE 1.3 A Hypothetical Ordering of Countries
According to the Opportunities Available
for Contestation
enterprise highly unsatisfactory. In principle, to be sure, subnational organizations could be located along the two dimensions illustrated in figures 1.1 and 1.2. Yet the
problem is not simply to locate countries in the hypothetical space suggested by the figure 1.3. For one thing, that space has to do with only one of the two main
dimensions: contestation. Ob
6. Cf. in particular Grant McConnell, Private Power and American Democracy (New York: Knopf, 1966); Henry S. Kariel, The Decline of American Pluralism
(Stanford: Stanford University Press, 1961); and to some extent also Robert Paul Wolff, The Poverty of Liberalism (Boston: Beacon Press, 1968).
Page 14
viously a similar procedure would be required for the other main dimension: participation. What is more, even within a country, subnational units often vary in the
opportunities they provide for contestation and participation. For example, in many modern countries these opportunities are much greater in municipal governments
than in trade unions, and greater in trade unions than in business firms. Consequently, one would have to break subnational units into a number of categories: business
firms, trade unions, municipal governments, churches, educational institutions, etc.7 At this stage, these requirements are, unfortunately, little short of utopian, and it is
for this reason—pragmatic rather than theoretical—that I have decided to restrict my attention to the national level.
Assumptions
When hegemonic regimes and competitive oligarchies move toward polyarchy they increase the opportunities for effective participation and contestation and hence the
number of individuals, groups, and interests whose preferences have to be considered in policy making.
From the perspective of the incumbents who currently govern, such a transformation carries with it new possibilities of conflict as a result of which their goals (and they
themselves) may be displaced by spokesmen for the newly incorporated individuals, groups, or interests.
The problem of their opponents is the mirror image of the problem of the incumbents. Any transformation that provides opponents of the government with greater
opportunities to translate their goals into policies enforced by the
7. The already classic study by Seymour Martin Lipset, Martin A. Trow, and James S. Coleman, Union Democracy (Glencoe: The Free Press, 1956), concentrates on
the deviant case of a trade union in which contestation and participation are high. To describe and explain that deviant case within the context of a single country was a
very sizable undertaking.
Page 15
state carries with it the possibility of conflict with spokesmen for the individuals, groups, or interests they displace in the government.
Thus the greater the conflict between government and opposition, the more likely that each will seek to deny opportunities to the other to participate effectively in policy
making. To put it another way, the greater the conflict between a government and its opponents, the more costly it is for each to tolerate the other. Since the opposition
must gain control of the state in order to suppress the incumbents (at which point opposition and government have changed roles), we can formulate the general
proposition as an axiom about governments tolerating their opponents:
AXIOM 1. The likelihood that a government will tolerate an opposition increases as the expected costs of toleration decrease.
However, a government must also consider how costly it would be to suppress an opposition; for even if toleration is costly, suppression might be very much more
costly and hence obviously foolish. Therefore:
AXIOM 2. The likelihood that a government will tolerate an opposition increases as the expected costs of suppression increase.
Thus the chances that a more competitive political system will emerge, or endure, may be thought of as depending on these two sets of costs:
AXIOM 3. The more the costs of suppression exceed the costs of toleration, the greater the chance for a competitive regime.
Axiom 3 can be illustrated graphically as in figure 1.4.
The lower the costs of toleration, the greater the security of the government. The greater the costs of suppression, the
Page 16
greater the security of the opposition. Hence conditions that provide a high degree of mutual security for government and oppositions would tend to generate and to
preserve wider opportunities for oppositions to contest the conduct of the government.
FIGURE 1.4
The question posed a moment ago can therefore be restated:
What circumstances significantly increase the mutual security of government and oppositions and thereby increase the chances of public contestation and polyarchy?
But before I try to answer that question, let me first consider a prior one: does polyarchy matter?
Page 17
2. DOES POLYARCHY MATTER?
Some readers might be inclined to think that differences in national regimes do not matter much. For example, one might share the view of those like Gaetano Mosca
who argue that every regime is, after all, dominated by a ruling minority. As an astringent challenge to the belief that portentuous consequences for the people of a
country must necessarily follow a transformation of the regime, Mosca’s skepticism has a good deal to be said for it. Moreover, what appear superficially to be
changes of regime are sometimes not really changes in regime at all, but simply changes in personnel, rhetoric, and empty constitutional prescriptions.
Yet few people seem able to adhere consistently to the view that differences in regimes—for example, differences between polyarchy and inclusive hegemony—are at
base negligible. In fact, I have the impression that this view is most often espoused by intellectuals who are, at heart, liberal or radical democrats disappointed by the
transparent failures of polyarchies or nearpolyarchies; and that, conversely, intellectuals who have actually experienced life under severely repressive hegemonic
regimes rarely argue that differences in regime are trivial. Perhaps the most telling examples are furnished by Italian intellectuals like Mosca and Croce who spent their
lives attacking the sorry and patently defective parliamentary regime that existed in Italy before Fascism.
Page 18
Although during the seventy years between unification and Fascism the Italian polity traversed the classic path from competitive oligarchy to inclusive polyarchy, the
defects of trasformismo in political affairs and the general incivismo of Italians in public life were too glaring to enable the parliamentary regime to win much support. Yet
even this highly defective regime was, Mosca saw, different in important essentials from Fascism—and what is more, unworthy though it may have been, better than
Fascism. In his last speech to the Italian Senate in 1925, Mosca confessed that he spoke
with a certain emotion because, let us be frank, we take part in the funeral rites of a form of government. I should not have thought it possible that I would be the one to
deliver the funeral oration on the parliamentary regime … I, who have always taken a harsh attitude toward it, I am today obliged to lament its departure … One may
say in all sincerity: the parliamentary regime was better.
But he was not permitted to drink the bitter potion in one grand gesture of regret, for he lived until 1941 and thus witnessed all but the final disintegration of that
wretched new order. As for Croce, who first welcomed Fascism, he was finally to admit that all during the time he had been pouring out his contempt for the
parliamentary regime,
it had never seemed to him even remotely possible that Italy could let herself be robbed of the liberty that had cost her so much and that his generation had considered
a permanent acquisition.
And by 1945, Gaetano Salvemini, who as an intellectual of radical persuasion had been a fierce critic of Giolitti’s Italy, had no doubt that for all its defects the
parliamentary regime was far better in actuality and in potential than what came later. “As for the results of the Fascist dictatorship in contrast with those of Italian
democracy in the making,” he
Page 19
concluded, “they are here before our very eyes. Let us hope that the Italians will not be the only ones to learn from that frightful experience.”1
Although eyewitness testimony of this kind does not prove the point, it cautions against accepting the easy notion that changes of political regime do not matter very
much. To analyze in a responsible way the extent to which, and the conditions under which, changes of regimes ‘‘matter” would, I fear, require a book, and I intend to
forgo that effort in this one. Moreover, if theory and data are far from satis
1. The quotation from Mosca is in James Meisel, The Myth of the Ruling Class (Ann Arbor: University of Michigan Press, 1958), pp. 225–26. That from Croce is in
Giovanni Sartori, Democratic Theory (Detroit: Wayne State University Press, 1962), p. 37; Croce’s early acceptance of fascism is discussed in Sartori, Croce Etico
Politico e Filosofo della Liberta (Florence: Universita degli Studi, n.d.), pp. 191 ff. Salvemini’s statement is from the introductory essay to A. William Salamone, Italy in
the Giolittian Era: Italian Democracy in the Making, 1900–1914 (Philadelphia: University of Pennsylvania Press, 1945, 1960). Salvemini’s brief essay in effect argues
that representative government as it was emerging in Italy compared not too unfavorably with England and the United States. His judgment is summed up by his
statement that “Italian democracy would have needed still another generation of trial and error before becoming not a ‘perfect democracy’ but a ‘less imperfect
democracy.’ The crisis that followed the First World War, however, was fatal to the democratic process.” (p. xx) Salvemini goes on to say:
Looking back at the work of the crusader after thirty years, I find that I have nothing to regret. I must acknowledge, however, that I would have been wiser had I been
more moderate in my criticism of the Giolittian system. My knowledge of the men who came after Giolitti in Italy as well as of countries in which I have lived during the
last twenty years has convinced me that if Giolitti was not better, neither was he worse than many nonItalian politicians who followed him. For while we Italian
crusaders attacked him from the Left accusing him of being—and he was—a corrupter of Italian democracy in the making, others assailed him from the Right because
he was even too democractic for their taste. Our criticism thus did not help to direct the evolution of Italian public life toward less imperfect forms of democracy, but
rather toward the victory of those militarist, nationalist, and reactionary groups who had found even Giolitti’s democracy too perfect.
Page 20
factory for determining the conditions most favorable to the development of different regimes, they are in an even more deplorable state with respect to the differences
in the consequences (in the recent argot of political science, “outputs”) of different regimes. Nonetheless, there are good reasons for thinking that a transformation of a
regime from a hegemony into a more competitive regime or a competitive oligarchy into a polyarchy does have significant results.
1. To begin with, there are the classic liberal freedoms that are a part of the definition of public contestation and participation: opportunities to oppose the government,
form political organizations, express oneself on political matters without fear of governmental reprisals, read and hear alternative points of view, vote by secret ballot in
elections in which candidates of different parties compete for votes and after which the losing candidates peacefully yield their claim to office to the winners, etc. In the
wellestablished polyarchies, these freedoms have long since lost the attraction of a new cause, let alone any revolutionary appeal. Familiar, imperfectly achieved,
clearly insufficient to insure a good society, trivialized over many generations by rhetorical overkill, they are easily taken for granted as an inheritance of quite modest
significance. Their value no doubt appears greater to those who have lost them or have never had them. It was liberties of this kind that critics of the preFascist
parliamentary regime in Italy like Mosca, Croce, and Salvemini took so much for granted that they failed to foresee how oppressive Italy would become under a new
regime. It was largely to expand freedoms of this kind that the liberalizing forces were moving in Czechoslovakia before their revolution was halted and reversed by the
Soviets. To gain liberties like these for Spain is the one goal that many of the oppositions to Franco’s dictatorship have shared.
2. Broadened participation combined with political competition brings about a change in the composition of the po
Page 21
litical leadership, particularly among those who gain office by means of elections—mainly, then, members of parliament. As new groups are granted the suffrage,
candidates closer in their social characteristics to the newly incorporated strata win a greater share of elective offices. Thus when the narrow suffrage of a competitive
oligarchy has been extended to the middle classes, the number of party leaders and members of parliament drawn from the middle classes has increased. Something of
the same kind has occurred when the working classes have been enfranchised, particularly in countries where labor or socialist parties have acquired a large share of
workingclass votes.2 When Reconstruction provided southern Negroes with the suffrage after the American Civil War, black Southerners for the first time began to
hold office; when Reconstruction came to an end, blacks dis
2. There is a wealth of evidence about these changes, but so far as I am aware no comparative analysis. Systematic, longrange studies include Matteei Dogan,
“Political Ascent in a Class Society: French Deputies 1870–1958”, in Dwaine Marvick, ed., Political DecisionMakers: Recruitment and Performance (Glencoe: The
Free Press, 1961), pp. 57–90; and W. L. Guttsman, The British Political Elite (London, MacGibbon and Key, 1963). For the changes brought about in Britain after
1832, the evidence is ample but apparently unsystematic. However, compare Sir Lewis Namier’s data on members of Parliament from the boroughs in 1761 in The
Structure of Politics at the Accession of George III, 2d ed. (London: Macmillan, 1961), pp. 84 ff., with W. Ivor Jennings, Parliament (Cambridge: Cambridge
University Press, 1939), table II, p. 38, and with Guttsman, The British Political Elite. For changes in the occupational and social class composition of the Italian
parliament from 1909 to 1963 (universal suffrage was introduced in 1913 and proportional representation in 1919), see S. Somogyi, L. Lotti, A. Predieri, and G.
Sartori, Il Parlamento Italiano, 1946–1963 (Naples: Edizioni Scientifiche Italiane, 1963), pp. 160–62, 168–69, and 197–200.
For differences in Argentina between the socioeconomic levels of parliamentarians from the conservative parties that had dominated parliament before universal suffrage
in 1911, and from the Radical and Socialist parties that had a majority of seats after the elections of 1916, see Darío Canton, “Universal suffrage as an Agent of
Mobilization” (paper presented to the VIth World Congress of Sociology, Evian, France, September 1966), p. 24.
Page 22
appeared from public life. When they began to regain the suffrage after the passage of the Civil Rights Act of 1964, they once again began to win public offices.3
This is not to say that political leadership and parliaments ever become a representative sample of the various socioeconomic strata, occupations, or other groupings in
a society. They never do. In contemporary legislative bodies, middleclass and professional occupations are numerically overrepresented; blue collar occupations are
numerically underrepresented (even among representatives of labor, socialist, and communist parties) as are many other categories—farmers and housewives, for
example.4 Even if the “political
3. Negro suffrage and its results under Reconstruction are discussed in C. Vann Woodward, The Burden of Southern History (New York: Vintage Books, 1960), pp.
98–103. For the recent period, data from the Voter Education Project, Southern Regional Council, show that by the summer of 1968 the percentages of Negroes
registered to vote had enormously increased. The percentages, with corresponding percentages of whites registered shown in parentheses, were: Alabama, 56.7
(82.5); Arkansas, 67.5 (75.2); Florida, 62.1 (83.8); Georgia, 56.1 (84.7); Louisiana, 59.3 (87.9); Mississippi, 59.4 (92.4); North Carolina, 55.3 (78.7); South
Carolina, 50.8 (65.6); Tennessee, 72.6 (81.3); Texas, 83.1 (72.3); Virginia, 58.4 (67.0). For these states as a whole the totals were 62.0 (78.1). Southern Regional
Council, Voter Education Project, ‘‘Voter Registration in the South, Summer, 1968” (Atlanta: Southern Regional Council, 1968). By the summer of 1969, some 473
black officials had been elected in southern states, including 17 mayors and 200 city councilmen. “Black Elected Officials in the Southern States,” a memorandum to
selected members of the American Political Science Association, August 12, 1969, from Emory F. Via, Director, Labor Program, Southern Regional Council, Inc.
4. On Britain, see W. L. Guttsman, “Changes in British Labour Leadership,” in Political DecisionMakers, pp. 91–137. For data on candidates for and members of the
House of Commons in the 1950s and 1960s, see J. Blondel, Voters, Parties, and Leaders (Baltimore: Penguin, 1963), pp. 135–45, and Peter G. J. Pulzer, Political
Representation and Elections, Parties and Voting in Great Britain (New York: Praeger, 1967), pp. 67 ff.
For the postwar Italian parliament, see Sartori et al., Il Parlaimento Italiano, pp. 93–97. For Belgian members of parliament in 1964, see F. Debuyst, La Fonction
Parlementaire en Belgique: Mecanismes d’Access et Images (Brussels: CRISP, 1966), pp. 90–
Page 23
class” is never a fair sample of a country’s social and economic categories—and many advocates of representative democracy would argue that it need not and should
not be5 —a broadening of the suffrage together with political competition does nonetheless make parliaments in particular and political leadership in general
considerably less unrepresentative in the purely statistical sense.
3. As a system becomes more competitive or more inclusive, politicians seek the support of groups that can now participate more easily in political life. The response of
politicians to the existence of new opportunities for participation and public contestation are manifold and have farreaching effects. I have just described one of these:
to offer candidates whom the voters feel are in some sense “closer” to themselves. Another is to adapt rhetoric, program, policy, and ideology to what are thought to
be the desires or interests of the groups, segments, or strata not hitherto represented. Thus the rise of socialist and labor parties in Western Europe is intimately tied to
the grant of the suffrage to urban and rural working strata. When, as was true in many countries that are now polyarchies, political parties were relatively free to
organize before the suffrage had been broadened, among the first demands of socialist and labor parties was universal suffrage. Once the working classes had the vote,
naturally these parties initially directed most of their efforts to mobilizing these strata.
Competition and inclusiveness bring about changes in the party system itself. The most drastic and visible changes occur, of course, when a oneparty hegemonic
regime is rapidly replaced by a polyarchy; the hegemony of the single
109. Debuyst also has tables comparing the professional backgrounds of members of the national legislatures in Belgium, France, Britain, Italy, and the U.S. (Senate)
(p. 110), and the percentages in various European socialist and communist parties of M.P.’s from lower middle or workingclass occupations or party functionaries (p.
113).
5. Eg., Hanna Fenichel Pitkin, The Concept of Representation (Berkeley: University of California, Press, 1967), chap. 4, pp. 60–91.
Page 24
party suddenly gives way to two or more competing parties, as in Italy, Germany, and Japan at the end of World War II. Countries in which opportunities for
participation and contestation expand over a lengthier period of time display somewhat similar developments in slow motion. When the suffrage moves beyond the
notables and their clients, the old parties and factions based mainly on the social connections among the notables—on ties of family, class, residence, life style, and
tradition—are displaced or supplemented by parties more effective in appealing to the middle classes. The process is repeated again when the working classes are
granted the suffrage. In Britain, the old Whigs gave way to the Liberals after the Reform Act of 1832; the Reform Acts of 1867 and 1884 facilitated the formation and
growth of the Labor party. In Norway, the struggle over the mobilization of the peasantry in the 1860s and 1870s led to the development of electoral and parliamentary
coalitions of the Left and the Right. The struggle over manhood suffrage and its achievement in 1900 produced new parties. While the old Right became the
Conservative party, the old Left alliance fragmented into its main components of Liberals, rural Christian fundamentalists, and farmers, while the Labor party acquired a
large share of the working classes.6 Although the details vary from country to country, a similar pattern seems to emerge wherever polyarchy has evolved over a
considerable period of time.
The parties also change in structure and organization. As has often been pointed out, the need to mobilize a bigger electorate triggers off the development of “modern”
party organizations. For as the electorate grows, the traditional, mainly informal arrangements that worked well enough with a tiny group of voters (many of whom were
in any case
6. Cf. Stein Rokkan, ‘‘Norway: Numerical Democracy and Corporate Pluralism” in Robert A. Dahl, ed., Political Oppositions in Western Democracies (New Haven:
Yale University Press, 1966), pp. 70–115, esp. pp. 75–81.
Page 25
under the thumb of the notables) are simply inadequate. If a party is to survive in the new competition, it must reach out to its members, followers, and potential voters
with organizations at the level of ward, section, cell, and the like. Many of these now familiar forms of party organization were initially developed in the country where
mass suffrage was first established—the United States—but they rapidly appear wherever political competition takes place in the midst of a broad suffrage. In Britain,
for example, the formation of local Conservative and Liberal associations, and the famous Birmingham Caucus followed hard on the heels of the broad suffrage created
in 1867 and the introduction of the secret ballot in 1872.7
The change in the organization of parties and their increasing penetration of urban and rural areas trips off still further changes in political life. Political competition and
participation are both heightened. As the nationally organized parties reach out to mobilize their voters, the number of uncontested or nonpartisan elections declines.
And the competition for members, adherents, and voters increases the politicization of the electorate, at least in the initial stages; participation in elections, for example,
is likely to be higher in constituencies where there are competing parties.8
7. See, for example, Pulzer, Political Representation. On the origin of the Conservative party’s “handmaid,” the National Union, as an organization for wooing the
newly enfranchised urban workers, see R. T. McKenzie, British Political Parties (London: Heinemann, 1955), pp. 146 ff. On the Liberal associations and the
Birmingham Caucus, see Sir Ivor Jennings, Party Politics, vol. 2, The Growth of Parties (Cambridge: Cambridge University Press, 1961), pp. 134 ff.
8. Again, comparative crossnational data are lacking. In Britain, the number of constituencies without a contest in parliamentary elections was 57% in 1835, 43% in
1868, and 23% in 1880. Pulzer, Political Representation, pp. 61–62. In Norway, as the Labor party established units to mobilize voters at the local level (communes),
opponents found it necessary to do likewise; thus from 1900 onward the number of rural communes with nonpartisan plurality elections declined (from 78% in 1901 to
2% in 1959), while the number with two or more competing party lists increased. Turnout at elections was markedly lower in communes with nonpartisan plurality elec
Page 26
4. In any given country, the greater the opportunities for expressing, organizing, and representing political preferences, the greater the number and variety of preferences
and interests that are likely to be represented in policy making. In a given country at a given time, therefore, the number and variety of preferences and interests
represented in policy making are likely to be greater if the political regime is a polyarchy than if it is a mixed regime, and greater under a mixed regime than under a
hegemony. Hence in any given country the transformation of a hegemony into a mixed regime or a polyarchy, or a mixed regime into a polyarchy, would be likely to
increase the number and variety of preferences and interests represented in policy making.9
5. The consequences for government policies of lower thresholds for participation and public contestation are, unfortunately, obscure. Crossnational studies confront
extraordinary difficulties in this area. Even studies of variations among the fifty American states in policies, politics, and socioeconomic variables have not so far
produced unambiguous findings on the extent to which variations in policies are related to variations in political competition and participation—though of course the
range of variation on all these variables must be markedly narrower than among countries.10 Because of the powerful impact on govern
tions than with proportional representation and party lists. Cf. Stein Rokkan and Henry Valen, “The Mobilization of the Periphery: Data on Turnout, party Membership
and Candidate Recruitment in Norway,” in Stein Rokkan, ed., Approaches to the Study of Political Participation (Bergen: The Chr. Michelsen Institute, 1962), pp.
111–58, esp. tables 2, 2.1, and 2.2, pp. 144–45. See also Torstein Hjellum, “The Politicization of Local Government: Rates of Change, Conditioning Factors, Effects
on Political Culture,” Scandinavian Political Studies 2 (1968): 69–93, tables 1 and 2, pp. 73–74.
9. I have developed this point at greater length in the introduction to Regimes and Oppositions (New Haven: Yale University Press, forthcoming 1971).
10. The first statistical analyses found that political variables like voter participation and party competition had little relation to state policies; the most powerful
explanatory factor was the level of socioeconomic development as indicated, for example, by per capita
Page 27
mental policies of such factors as a country’s level of socioeconomic development, the characteristics of its social and economic systems, and its traditions, it may well
be that the character of the regime has little independent effect on most governmental policies.
We probably need to look elsewhere to find the impact of regime on policy, in particular, on the extent to which the government adopts policies that involve severe
physical coercion for relatively large numbers of people. The lower the barriers to public contestation and the greater the proportion of the population included in the
political system, the more difficult it is for the government of a country to adopt and enforce policies that require the application of extreme sanctions against more than
a small percentage of the population; the less likely, too, that the government will attempt to do so.
The evidence on this point is impressionistic. However, so far as I know, no polyarchy has ever undertaken policies involving anything like the degree and extent of
coercion used during the forced collectivization of farming in the USSR in 1931–32, when millions of people were deported to Siberian labor camps or died from
execution or starvation. Stalin’s purges in the thirties sent many more millions to prison, torture, and death.11 Hitler’s policy of extermination
income. Thomas R. Dye, Politics, Economics and the Public (Chicago: Rand McNally, 1966), and Richard E. Dawson and James A. Robin son, ‘‘Interparty
Competition, Economic Variables and Welfare Policies in the American States,” Journal of Politics 25 (1963): 265–89. See also Ira Sharkansky, The Politics of Taxing
and Spending (Indianapolis: BobbsMerrill, 1969), pp. 121–45. More recently analysis indicates, however, that political variables do have effects. Charles F. Cnudde
and Donald J. McCrone, “Party Competition and Welfare Policies in the American States,” American Political Science Review 53 (September 1969): 858–66; Ira
Sharkansky and Richard I. Hofferbert, “Dimensions of State Politics, Economics, and Public Policy,” ibid., pp.867–78; and Brian R. Fry and Richard F. Winters, “The
Politics of Redistribution,” ibid., 54 (June 1970): 508–22.
11. We shall probably never have reliable data on these matters. In his famous letter, the Russian physicist Andrei D. Sakharov indicates the figure of up to 15 million
deaths attributable to Stalin is accepted
Page 28
of Jews and all political opponents is too well known to need emphasis. Changes of leadership and basic policies in hegemonic regimes frequently entail considerable
bloodshed. When Indonesia shifted from a procommunist to an anticommunist dictatorship in October 1965, it is estimated that at least a quarter of a million people
lost their lives over the space of a few months.12 In late 1969, some 116,000 persons suspected of communist sympathies had been incarcerated. 13
I do not mean to argue that such massive coercion inevitably occurs in hegemonies nor, certainly, in mixed regimes, but only that the risk is significant, whereas it does
not incur in polyarchies. The seeming exception that most readily comes to mind actually supports the point. In chapter 6 I shall argue that in order for the white people
to coerce Negroes in the American South, the South had to develop a dual system, a kind of polyarchy for whites and hegemony for blacks. It is important to keep the
distinction in mind, not for the sake of logic chopping, definitional purity, or “saving” polyarchy at all costs, but precisely because of the
by many Russian intellectuals. New York Times, July 22, 1968, p. 15. In his meticulously detailed but hostile account, Robert Conquest estimates that collectivization
produced “about 5frac12 million deaths from hunger and from the diseases of hunger,” while “some three million seem to have ended up in the newly expanding labour
camp system.” Conquest cites ‘‘the most careful estimates” of the labor camp population as “mainly at the 5 million level” in 1933–35, and at 6 million in 1935–37; he
accepts “a figure of about 8 million purgees in the camps in 1938.” Of those in the camps in 1936–38, he estimates that “about 3 million” died. The Great Terror,
Stalin’s Purge of the Thirties (New York: Macmillan, 1968), pp. 23–24, 333, 335–336.
12. Donald Hindley, who conducted interviews in Indonesia from May to December 1967, states that, “In all, perhaps 250,000 persons were killed, an equal number
herded into prisons and hastily constructed concentration camps.” In a footnote, however, he states that “informed foreign observers have estimated the number of
dead as high as one million.” “Dilemmas of Consensus and Division: Indonesia’s Search for a Political Format,” Government and Opposition 4 (Winter 1969): 79.
13. New York Times, June 22, 1970, p. 8.
Page 29
empirical generalization that it reinforces: if the freed Negroes had been allowed to participate in the system of public contestation in the South, they could not have
been subjected to systematic repression by coercion and terror, I believe, for they were much too large a minority. It was only by excluding them forcibly from the
polyarchy that the system of coercion and terror could be maintained in the South. And precisely to the extent that black people were excluded, polyarchy in the United
States was not fully inclusive.14 It was, in fact, less inclusive than most other polyarchies after the First World War, for following the general adoption of universal
suffrage no other country with a polyarchal regime (with the exceptions of Switzerland and the transitory polyarchy in Argentina) contained an excluded group of
comparable size. (It would not be entirely unreasonable to define polyarchy as requiring a degree of inclusiveness greater than that met by the United States, in which
case this country would have to be classified as a near polyarchy.)
The example of the United States suggests a final point about the consequences of regimes for policy. I do not believe that polyarchies are more considerate than other
regimes toward people who are effectively excluded from the rights of citizens. Among these excluded groups were (and to some extent still are) black people living in
the American South, but for every polyarchy foreigners living outside the boundaries of the particular country are excluded. Though there is no reason to think they are
worse, countries with polyarchal regimes are probably no better than other countries in responding to the interests of people beyond their boundaries.
6. One could speculate about other possible consequences of differences in regimes. It is possible, for example, that
14. Southern Negroes were 10.3% of the total population of the United States in 1900, 8.4% in 1920, and 6.8% in 1950. U.S. Bureau of the Census, Historical
Statistics of the United States, Colonial Times to 1957 (Washington, D.C.: Government Printing Office, 1961), pp.7 and 12.
Page 30
over long periods of time differences in regime may have effects on beliefs, attitudes, culture, and personalities. As we shall see in chapter 8, these are usually treated as
intervening or independent variables affecting regimes. But it is also reasonable to suppose that there is a reciprocal interaction between factors of this kind and the
character of a regime: if these factors affect the chances of a particular type of regime, over time the nature of the regime influences beliefs, attitudes, culture, and
perhaps even personalities that are likely to develop in a country. There are fascinating and important possibilities along these lines, but so many alternative hypotheses
are plausible and yet impossible to appraise in the light of satisfactory evidence that I shall not take the matter any further in this book.
The thrust of the argument is, however, clear enough. It seems reasonably evident that different regimes do have different consequences. Although some people may
deny the importance of these consequences, at least the advocates of polyarchy and their opponents both agree that the consequences are significantly different and
important. If the consequences of polyarchy were no different from those of nonpolyarchy, or if the consequences were unimportant, there would be no reason to
advocate a polyarchy rather than a oneparty dictatorship—or the converse. Probably most readers will also agree that the consequences—particularly the first—are
important.
The source of the controversy over the relative value of polyarchy versus hegemonic or mixed regimes may not be so much in the results to be expected from public
contestation and inclusion discussed above as in the consequences for other values. For example, it has been argued that oneparty regimes are desirable in most
African countries because a oneparty regime expresses a natural consensus or solidarity or is necessary to achieve economic development, to build a nation out of the
country’s diverse subcultures, or to secure political stability. As S. E. Finer has convincingly shown,
Page 31
some of these arguments are self contradictory—one cannot logically defend the single party as an expression of “natural” consensus and also argue that it is required in
order to build national solidarity out of tribal diversity and disharmony; all the alleged advantages of oneparty regimes seem to be belied by the facts.15
But it is not my purpose here to make a case for polyarchy. It is enough if I have shown that important consequences will follow from reducing the obstacles to public
contestation and increasing the share of the population entitled to participate. A great many people will agree, I think, not only that these consequences are important
but that they are also desirable, that the benefits often (if not always) outweigh the adverse consequences, and that the net gain in such cases is well worth striving for.
The conceptual scheme I employ in this book reflects a commitment (or as others might see it, a bias) in favor of polyarchy as against less democratized regimes. (What
may not be apparent, since it is less relevant to the subject of this book, is a bias also in favor of greater democratization of polyarchies.) Nonetheless, I do not assume
that a shift from hegemony toward polyarchy is invariably desirable. Let me make clear at once my conviction that a shift from hegemony toward polyarchy is frequently
desirable; my belief furnishes one of the motives for examining the subject of this book and for formulating the central questions and the concepts as I do. Strictly
speaking, however, one could deal with the questions posed in this book and employ the concepts set out here with no assumptions at all as to the desirability of any
particular direction of change. In fact, even one who held the extreme position that a shift from hegemony toward polyarchy is never desirable would want to
understand, I should think, the conditions required to prevent such a change. In this sense, the analysis is intended
15. S. E. Finer, “The OneParty Regimes in Africa: Reconsiderations,” Government and Opposition 2 (July–October 1967): 491–508.
Page 32
to be independent of my commitments or biases in favor of polyarchy—though given the difficulties of data analysis at this stage I may not have wholly succeeded.
Finally, I want to make clear that I make no assumption that a shift from hegemony toward polyarchy is historically inevitable. Just as the outcome of the third wave of
democratization remains in doubt and could even lead to a regressive narrowing of the opportunities for public contestation now available in polyarchies, so it would be
absurd to suppose that some sort of historical law of development imposes on societies an inevitable transition away from political hegemony to public contestation—
or, for that matter, in the opposite direction. Since modern nationstates have displayed movements in both directions, a few wellknown cases are enough to falsify any
simple law of unidirectional development. One might reflect, for example, on the histories of Argentina, Brazil, Germany, Italy, Russia, Czechoslovakia, and Japan. One
of the implications of the analysis in this book is, as we shall see, that the conditions most favorable for polyarchy are comparatively uncommon and not easily created.
Returning now to the question posed at the end of the last chapter—what conditions significantly increase the chances of public contestation and polyarchy?—I shall
explore in the chapters that follow the consequences of seven sets of conditions: historical sequences, the degree of concentration in the socioeconomic order, level of
socioeconomic development, inequality, subcultural cleavages, foreign control, and the beliefs of political activists.
Page 33
3. HISTORICAL SEQUENCES
One can conceive of historical processes as having two aspects relevant to our central question: the specific path or sequence of transformations of a regime and the
way in which a new regime is inaugurated.
The Path to Polyarchy
Does the sequence matter?1 Are some sequences more likely than others to lead to mutual security and thus to facilitate the shift toward a more polyarchal regime? The
two figures introduced in the last chapter to represent the two dimensions of democratization with which we are concerned allow, of course, for an infinite number of
paths. History has traced out some of these. But even if one were to limit his imagination by history and common sense, he would surely discover and invent more paths
than anyone could deal with. A modest concern for a reasonably parsimonious and manageable theory impels me to try for a narrower
1. This is also the central question in Barrington Moore, Jr., Social Origins of Dictatorship and Democracy: Lord and Peasant in the Making of the Modern World
(Boston: Beacon Press, 1966). However, as the subtitle suggests, Moore is concerned with different variables and longer historical sequences. Moreover, he chooses
to ignore the experience of smaller countries on grounds I find unpersuasive (p. xiii).
Page 34
focus. Let me begin, then, by considering only three possible paths to polyarchy:
I. Liberalization precedes inclusiveness:
A. A closed hegemony increases opportunities for public contestation and thus is transformed into a competitive oligarchy.
B. The competitive oligarchy is then transformed into a polyarchy by increasing the inclusiveness of the regime.
II. Inclusiveness precedes liberalization:
A. A closed hegemony becomes inclusive.
B. The inclusive hegemony is then transformed into a polyarchy by increasing opportunities for public contestation.
III. Shortcut: A closed hegemony is abruptly transformed into a polyarchy by a sudden grant of universal suffrage and rights of public contestation.
FIGURE 3.1 Some Paths to Polyarchy
These three paths are represented in figure 3.1. The first is a fair approximation of the paths taken by England and by Sweden.2 The second is roughly the path taken
by Germany from the Empire to Weimar. The third is roughly the
2. Their paths were of course not nearly as schematic as the diagram suggests. For example, until the effects on the suffrage of the Reform Act of 1832 began to be felt,
in many constituencies the candidates
Page 35
path taken in France from 1789 to 1792 (although given various restrictions on voting and the freedom to organize the terminus would perhaps be more accurately
described as nearpolyarchy).3
put forward by the notables won without being contested in the election. ‘‘Of the 22 towns with over 1000 voters, eleven went to the poll in 1761; of the 22 towns
with 500–1000 voters, twelve; while of the remaining 201 English constituencies only 18: i.e., more than half of the larger boroughs were contested, and about one in
ten of all the other constituencies.” Sir Lewis Namier, The Structure of Politics at the Accession of George III (London: MacMillan, 1961), p. 83. As late as 1830,
elections in the counties usually went uncontested: “In the forty counties of England and Wales there were in 1830 only nine contests, the same number as in 1820; and
in 1831 there were eleven, one more than in 1826. In most counties the great landowners nominated the candidates, very often by agreement among themselves so as
to avoid the expense of a contested election and the disturbance of the peace of the country.” Sir Ivor Jennings, Party Politics (Cambridge: Cambridge University
Press, 1961), p. 81. Even in 1833, when the first election under the Reform Act occurred, there were no contests in nearly one third of the constituencies. Ibid., p. 84,
n. 1.
3. About 60% of the adult males had the right to vote under the electoral law of 1789. Under a system of indirect election these “active citizens” chose delegates who in
turn elected the deputies. Although the figure is in dispute, probably not more than 45% of the adult males could qualify as delegates. Cf. R. R. Palmer, The Age of the
Democratic Revolution: The Challenge (Princeton: Princeton University Press, 1959), appendix V, pp. 522 ff., and Peter Campbell, French Electoral Systems and
Elections, 17891957 (London: Faber and Faber, 1958), pp. 50–57. The electoral law of 1792, under which the Convention was elected, introduced manhood
suffrage, though it retained indirect elections; the Constitution of 1793 provided for universal manhood suffrage, but that constitution was never applied. “At all elections
in this period large numbers of electors did not vote. In 1792 only 700,000 of the national electorate of 7,000,000 voted. In the referenda on the successive
constitutions between onethird and fivesixths of the electorate abstained. Under the Republic corruption, fraud, intimidation, and violence were practised by the
candidates of all factions and their supporters … electors with the wrong views were prevented from voting; citizens who might have the wrong views were
disfranchised.” Campbell, p. 57. Moreover, the Le Chapelier Law prohibited economic organizations of workers (and also in principle if not in practice, of
entrepreneurs and merchants). Val R. Lorwin, The French Labor Movement (Cambridge: Harvard University Press, 1954), p. 4.
Page 36
Probably the commonest sequence among the older and more stable polyarchies has been some approximation of the first path, that is, competitive politics preceded
expansion in participation.4As a result, the rules, the practices, and the culture of competitive politics developed first among a small elite, and the critical transition from
nonparty politics to party competition also occurred initially within the restricted group. Although this transition was rarely an easy one, and party conflict was often
harsh and bitter, the severity of conflict was restrained by ties of friendship, family, interest, class, and ideology that pervaded the restricted group of notables who
dominated the political life of the country. Later, as additional social strata were admitted into politics they were more easily socialized into the norms and practices of
competitive politics already developed among the elites, and generally they accepted many if not all of the mutual guarantees evolved over many generations. As a
consequence neither the newer strata nor the incumbents who were threatened with displacement felt that the costs of toleration were so high as to outweigh the costs
of repression, particularly since repression would entail the destruction of a welldeveloped system of mutual security.
The other two paths are more dangerous, and for the same reason: to arrive at a viable system of mutual security is a
4. Obviously this capsule description ignores variations that in another context would be vital for explaining differences in contemporary European systems—the party
systems, for example. The most extensive analysis of which I am aware of the different historical paths of European countries and their political consequences is to be
found in the work of Stein Rokkan. Cf. his “The Comparative Study of Political Participation,” in A. Ranney, Essays on the Behavioral Study of Politics (Urbana:
University of Illinois Press, 1962), pp. 45–90; “Mass Suffrage, Secret Voting, and Political Participation,” Arch. Eur. Sociol. 2 (1961): 132–52 “Cleavage Structure,
Party Systems, and Voter Alignments’’ (with S. M. Lipset), in Stein Rokkan and Seymour Martin Lipset, eds., Party Systems and Voter Alignments (New York: The
Free Press, 1967), pp. 1–64; and “The Structuring of Mass Politics in the Smaller European Democracies: A Development Typology” (Paper presented to the
International Political Science Association, Brussels, September 1967).
Page 37
difficult matter at best; the greater the number of people and the variety and disparity of interests involved, the more difficult the task and the greater the time required.
Tolerance and mutual security are more likely to develop among a small elite sharing similar perspectives than among a large and heterogeneous collection of leaders
representing social strata with widely varying goals, interests, and outlooks. This is why the first path is more likely than the other two to produce stable transformations
away from hegemony toward polyarchy. The third path drastically shortens the time for learning complex skills and understandings and for arriving at what may be an
extremely subtle system of mutual security. The second path requires that the system of mutual security be worked out, not within a small and relatively homogeneous
elite, but among spokesmen who reflect the whole spectrum of social strata and political perspectives in the society, or at least in a broad part.
There seem to be few if any unambiguous cases in which the shortcut has been successfully taken.5 To be sure, in
5. The case of Denmark seems to be somewhat anomalous, though I know too little about it to make a valid appraisal. Under the constitution of 1665, the power of the
monarch was absolute, and during the next two centuries the country was ruled by a highly centralized administration under the king. The July revolution of 1830 in
France persuaded the king to establish four provincial assemblies for consultative purposes. Under the stimulus of the revolution of 1848, the monarch proclaimed a
constitution that entrusted the legislative power to the Rigsdag. Suffrage was granted to all men 30 years of age or older, except those working as servants and farm
helpers not having their own household, and those receiving or having received poor relief. In this sense, Denmark did indeed take a shortcut. However, voting for the
lower house was in public, by show of hands; that for the upper house was indirect, and under the constitution of 1866 the landowners and highly taxed citizens were
given preponderant influence in the upper chamber. Moreover, the monarch refused to accept the principle that his ministers were responsible to Parliament; after 1901
responsibility existed de facto and after 1915 de jure. The constitution of 1915 also established universal suffrage for men and women 29 years of age and over, and
abolished the privileged suffrage for the upper chamber. Thus Denmark took a shortcut in 1849 to a broad suffrage and a considerable increase in opportunities for
public contestation but delayed for half a century
Page 38
Italy, Germany, and Japan an existing hegemony was destroyed by military conquest in the Second World War, and the hegemonic regime was replaced during the
occupation that followed defeat by an inclusive polyarchy. But these are highly ambiguous cases. For in all three countries, a transition to competitive politics had
already been made before the dictatorial seizure of power, and some of the older traditions of competitive politics reappeared after the destruction of the dictatorship.
In Japan, the preservation of the monarchy also helped to convey some traditional legitimacy to the new regime of competitive politics. Moreover, in each case the
dictatorship was not destroyed from within but from outside by overwhelming military defeat; the occupying forces at least temporarily banned the spokesmen of the
old dictatorship from public life, and for a few years they decided all the crucial questions. For all these reasons, and doubtless others, the new regimes were not beset
by fatal conflicts over legitimacy arising out of counterclaims set forth by spokesmen for the old regime. Nonetheless, these three cases do show that under certain
highly unusual conditions an abrupt shift from hegemony to polyarchy may result in tolerably stable regimes. The conditions may, however, prove to be historically
unique.
The second path is also risky. When the suffrage is extended before the arts of competitive politics have been mastered and accepted as legitimate among the elites, the
search for a system of mutual guarantees is likely to be complex and time consuming. During the transition, when conflict erupts neither side can be entirely confident
that it will be safe to tolerate the other. Because the rules of the political game are ambiguous, and the legitimacy of competitive politics is weak, the costs of
suppression may not be in
the final transition to the eighth institutional guarantee listed in table 1.1. The Danish political experience has been subjected to so little systematic analysis that I am
unclear as to how it bears on the argument of this section.
Page 39
ordinately high. The danger, is then, that before a system of mutual security can be worked out among the contestants, the emerging but precarious competitive regime
will be displaced by a hegemony ruled by one of the contestants.
Although the first path seems to be the safest of the three, it is not likely to be followed in the future, for as we have already seen most countries with hegemonic regimes
are already inclusive. Only a rather small minority of countries deny the suffrage to more than 10 percent of their male citizens, and probably no more than a half dozen
traditional monarchies or dictatorships have refused to grant the suffrage at all. Moreover, the suffrage seems to be more easily expanded than contracted; historically
the process has typically been in one direction: once granted, it is rarely taken away. In this respect, the oscillations in France from 1789 to 1848 between a wide or
universal manhood suffrage and a restricted electorate seem to be unusual. Even the few regimes now existing that have not yet granted their citizens the suffrage will
probably not pursue the first path. For if demands for inclusion and liberalization begin to threaten the regime, the leadership will doubtless be tempted to make the
cheapest concession possible: by granting the suffrage they can clothe the hegemony with the symbols and some of the legitimacy of “democracy”—at little cost,
initially, to the leaders.
The argument thus far can then be summarized in four propositions:
1. The first path is more likely than the others to produce the degree of mutual security required for a stable regime of public contestation.
2. But the first path is no longer open to most countries with hegemonic regimes.
3. Hence the liberalization of nearhegemonies will run a serious risk of failure because of the difficulty, under conditions of universal suffrage and mass politics, of
working out a system of mutual security.
Page 40
4. The risks of failure can be reduced, however, if steps toward liberalization are accompanied by a dedicated and enlightened search for a viable system of mutual
guarantees.
Inaugurating the Competitive Regime
Does it matter how a competitive regime is inaugurated? By inauguration I mean the application of power, influence, or authority to introduce and to legitimize a
regime—in this case a competitive regime. In this sense inauguration emphasizes transitional processes that are, conceptually speaking, somewhere between the paths
to polyarchy that we have just been concerned with and the maintenance of the regime after it has been inaugurated. Although the distinctions between paths,
inauguration, and maintenance blur at the edges, the concept of inauguration6 helps us to focus on an important element in the development of competitive regimes.
One way of deciding whether inauguration matters is to consider some of the important ways in which polyarchies or nearpolyarchies have been inaugurated in the
past. The chief forms seem to be:
I. Within an already independent nationstate
A. The old regime is transformed by evolutionary processes: the new regime is inaugurated by incumbent leaders, who yield peacefully (more or less) to demands for
changes and participate in the inauguration of polyarchy or nearpolyarchy.
B. The old regime is transformed by revolution: the new regime is inaugurated by revolutionary leaders, who overthrow the old regime and install a polyarchy or near
polyarchy.
6. I am indebted to my colleague, Juan Linz for his insistence on the relevance of the way in which a competitive regime is inaugurated.
Page 41
C. The old regime is transformed by military conquest: after a military defeat, victorious occupying forces help inaugurate a polyarchy or nearpolyarchy.
II. In a hitherto dependent country subject to another state
D. The old regime is transformed by evolutionary processes: the new regime is fostered among the local population, whose leaders inaugurate polyarchy or near
polyarchy without a national independence movement or serious struggle against the colonial power.
E. The old regime is transformed as a part of the struggle for national independence, in the course of a “revolution” against the colonial power: the new regime is
inaugurated by leaders of a national independence movement, who install polyarchy or nearpolyarchy during or after a succesful struggle for national independence.
Examples of the inauguration of polyarchies are given in table 3.1.
Although the examples in table 3.1 show that there has been no uniform process of inaugurating polyarchies, they also suggest that the various alternatives may not be
equally auspicious. A disproportionately large number of the stable highconsensus polyarchies seem to have come about in the first way, by peaceful evolution within
an already independent nationstate, or the fourth, by peaceful evolution within a dependent country. The reason is probably that peaceful evolution is most likely to
result in a polyarchy supported by a widespread sense of legitimacy. As the incumbents yield peacefully (on the whole) and participate in the changes, their consent is
won, the legitimacy attached to the previous regime is transferred unbroken to the new regime, and the
Page 42
Table 3.1. Processes of Inaugurating Polyarchies
I. Within an already independent nationstate
A. By evolutionary processes
l Britain
l Belgium
l Chile
l Costa Rica
l Denmark
l Japan (Meiji Restoration to the 1930s)
l Netherlands
l Norway
l Sweden
l Switzerland
l Uruguay
B. By collapse or revolutionary displacement of old regime
l France (1789–92, 1848, 1870)
l Germany (1919)
l Austria, First Republic (1918)
l Spain (1931)
C. By military conquest (all following World War II)
l Austria, Second Republic
l German Federal Republic
l Italy
l Japan
II. Within a subject state
D. By evolutionary processes
l Australia
l Canada
l Iceland
l New Zealand
l Philippines
E. By a national independence struggle
l Finland
l India
l Ireland
l Israel
l United States
process of peaceful change, so important to polyarchy, gains in legitimacy.
By comparison with the first process, the second—inauguration after the abrupt collapse or revolutionary overthrow of the old regime—is infrequent: in the three most
notable
Page 43
cases—the French Revolution, Weimar Germany, and the Spanish Republic—revolution or collapse was followed by an unstable regime that soon regressed to
hegemony. Was this reversal accidental? Probably not, for where peaceful evolution cannot or does not take place and revolution occurs, the legitimacy of the new
regime is more likely to be contested. A sudden collapse of the old regime leaves the new without a legacy of legitimacy; a revolutionary inaugural by the new
legitimates the use of revolution against itself. The most critical years, then, are likely to be the early ones, when the legitimacy of the new regime is still in question, and
loyalties to the old regime are still alive.
The third process has proved so far to lead to surprisingly stable polyarchies in the only four countries where inauguration of polyarchy by conquest has occurred in
recent times. Some possible reasons for the stability of the polyarchies inaugurated under the Allies after World War II have already been suggested; I also suggested
that these may be historically unique cases.
The fifth process is the one with which Americans are most familiar and, at least rhetorically, most sympathetic. As in the United States, so too in Finland, Ireland,
Israel, and India the independence movement blended nationalism with the ideology of representative government and political liberalism. Thus the ideology of
democracy was reinforced by the ideology of nationalism: to attack representative democracy was to attack the nation. The success of the movement for national
independence largely liquidated the principal contenders for the legitimacy of the old regime. Mainly agents of the colonial power, they either returned to the home
country or permanently exiled themselves from the new nation, as in the case of the Tories who moved to Canada after the American Revolution. Or, as with Ireland
and Ulster, those who would have constituted a disaffected minority in the new country remained a part of the old.
In the future, however, stable polyarchies are unlikely to
Page 44
be inaugurated by the fifth process. For one thing, in many of the new states where the sense of nationhood is weak, the leaders of nationalist movements, who during
the struggle for independence proclaimed the goal of democracy, later on as leaders of the new and fragile nation see organized opposition as a threat to the integrity of
the country. In the new states, then, nationalism does not so much encourage toleration of dissent and oppositions as it provides a ready and acceptable justification for
intolerance and repression.7 (It is worth keeping in mind that in the United States throughout our national history fears about nationhood and loyalty have resulted in
attempts—sometimes successful—to repress dissent. The relationship between nationalism, loyalty, and fear of dissent in the United States is nicely symbolized by the
original name of the official organization that has most fully displayed hostility to dissent: the UnAmerican Activities Committee of the U.S. House of Representatives.)
What may be an even more important limitation on the fifth strategy is that world developments have made it obsolescent. With the disappearance of colonial empires,
most of the world now consists of nominally sovereign states. In a world of independent states there are no longer many opportunities for movements of national
independence to inaugurate more competitive regimes.
In fact, the options seem to be even more narrowly restricted. The disappearance of colonial empires also reduces the opportunities for the fourth process of
inauguration. If it is true that the third process—via military conquest—is unlikely, then the most likely alternatives are reduced to the first two: in existing hegemonic
regimes a more competitive system will have to be inaugurated either by evolution or by revolution. Simply because the revolutionary process carries a high risk of
failure does not mean that it will not be tried,
7. Edward Shils, “Opposition in the New States of Asia and Africa,” and Hans Daalder, ‘‘Government and Opposition in the New States,” Government and
Opposition 1 (January 1966): 175–226.
Page 45
but revolutions will probably saddle new regimes with serious conflicts over legitimacy and hence create from the start a high probability of regression toward
hegemonic rule.
In the future as in the past, then, stable polyarchies and nearpolyarchies are more likely to result from rather slow evolutionary processes than from the revolutionary
overthrow of existing hegemonies.8
If this interpretation seems unduly restrictive, it is worth recalling that among most of the wellestablished polyarchies today, where there exists a high tolerance for
oppositions of all kinds, the transformation was exceedingly slow. In Britain a “formed” opposition was still illegal and illegitimate at the end of the seventeenth century.
A century later the idea of a more or less organized but “loyal” opposition in Parliament to His Majesty's government had gained a considerable legitimacy.9
But still another century elapsed before Britain evolved its present system of highly organized parties competing for the support of a broad electorate. Elsewhere, as in
France, attempts to shortcircuit this slow evolutionary process by revolution sometimes produced lasting oppositions to the new regime. It is also worth recalling that in
1968 the USSR celebrated the 50th anniversary of the Bolshevik Revolution. While the extreme hegemony of the Stalinist period has been replaced, the USSR has not
yet been transformed into a nearhegemony, and the inauguration of a nearpolyarchy appears even to optimistic observers to be considerably more than a generation
away.
That polyarchies and oneparty hegemonies have de
8. Moore's emphasis on the vital importance of the violent revolution as a stage along the road to democracy is, I believe, misleading, particularly if it is applied to the
process of inauguration. Moore stresses heavily the English Civil War, the French Revolution, and—a very doubtful case—the American Civil War. Social Origins of
Dictatorship and Democracy, passim. His argument here is, I think, weakened by his belief that the experience of the smaller countries is somehow irrelevant. The
question is: irrelevant to what?
9. See Archibald Foord, His Majesty's Opposition (Oxford: Oxford University Press, 1964).
Page 46
veloped in the twentieth century means that neither evolution nor revolution can occur in complete isolation from existing models that were quite unknown and would
have been incomprehensible in the eighteenth century. A loyal opposition, a twoparty system, a oneparty dictatorship do not have to be reinvented as if there were no
models to copy. No country needs to grope through centuries of experience with no clear idea of the elementary institutions required for a highly liberalized regime:
competitive parties and uncoerced elections are not merely a goal, but a fact. Likewise when a nearpolyarchy collapses, antidemocratic leaders do not have to fumble
for the oneparty formula.
The consequences of this for the inaugural process are unclear. An evolutionary process leading to polyarchy in the future need not, and probably cannot, consume the
centuries it took in Britain, Sweden, and elsewhere. Already ‘‘tested” models are now available that offer modern societies such radically different regimes as a unified
hegemony in its most totalitarian form or an inclusive polyarchy with an extraordinary tolerance for oppositions of all kinds. The availability of these “tested” models,
which do not have to be reinvented, may sometimes facilitate rapid transformation of regimes and even oscillations from one extreme to the other within a very short
historical period: witness Italy, Germany, and Japan from, say, 1919 to 1950.
The argument of this section can be summed up in the following propositions:
1. The process of inauguration most auspicious for a polyarchy is one that transforms previously legitimate hegemonic forms and structures into the forms and structures
suitable for political competition and thus produces no lasting cleavages or widespread doubts about the legitimacy of the new regime.
2. The inaugural process most likely to lead to this result is peaceful evolution within an independent
Page 47
nationstate or within a quasiindependent nation that is granted independence without a national independence movement.
3. The process of inauguration least auspicious for a polyarchy is one that leaves a large segment of the citizen body opposed to the legitimacy of competitive politics.
4. This result is likely when a polyarchy is inaugurated by a civil war or revolution in which a large segment of the people who uphold the legitimacy of the old regime or
deny the legitimacy of the new are defeated but nonetheless incorporated as citizens in the new regime.
5. The decline of colonial empires and the unlikely repetition of the circumstances facilitating the introduction of polyarchy in the defeated countries by the Allies at the
end of World War II mean that in the future the main options available are evolution or revolution within an already independent nationstate.
6. The presence in the world of functioning models of polyarchy and oneparty hegemony probably has an impact on the process of inaugurating regimes, but the
effects are uncertain. At a minimum, their presence probably raises expectations that regimes can be rapidly transformed in either direction.
7. Yet in countries without a recent legacy of experience in the operation of competitive politics, the transformation of hegemonic regimes into polyarchies is likely to
remain a slow process, measured in generations.
8. The length of the process can probably be reduced and the prospects of a stable transformation increased if inaugural processes are accompanied by a search for an
internal system of mutual security.
Page 48
4. THE SOCIOECONOMIC ORDER:
CONCENTRATION OR DISPERSION?
What difference does the social and economic order make? Are the chances that a hegemonic regime will be transformed into a more competitive regime higher under
some socioeconomic orders than others? Are the chances that a polyarchy will be maintained dependent on the socioeconomic order?
Assumptions
In the first chapter I introduced a more or less selfevident axiom which asserted that a government is more likely to tolerate an opposition as the expected costs of
suppression increase and as the expected costs of toleration decrease. Since the costs of toleration or suppression are in turn dependent on the relative resources
available to the government and to the opposition, it is obvious that:
AXIOM 4: The likelihood that a government will tolerate an opposition increases as the resources available to the government for suppression decline relative to the
resources of an opposition.
Now the key resources that governments use to suppress oppositions are of two broad types: violent means of coercion,
Page 49
persuasion, and inducement, typically wielded by military and police forces; and nonviolent means of coercion, persuasion, and inducement, or, as they will be called
here, socioeconomic sanctions, chiefly in the form of control over economic resources, means of communication, and processes of education and political socialization.
Hence
AXIOM 5: The likelihood that a government will tolerate an opposition increases with a reduction in the capacity of the government to use violence or socioeconomic
sanctions to suppress an opposition.
Two very general kinds of circumstances can reduce the capacity of a government to use violence or socioeconomic sanctions against an opposition. First, these factors
sometimes cease to be available as political resources. This possibility is particularly relevant to violence against opponents of the government by police or military
forces, for the police and military may actually be very small, or, what amounts to very nearly the same thing, they may become so depoliticized that they can no longer
be used by political leaders for internal political purposes. Second, these (and other) political resources may be so widely dispersed that no unified group, including the
government (or a unified group of leaders in the government) has a monopoly over them.
Thus, in the eighteenth century, the “professional” military and police forces of Britain, other than the Navy, were not only dispersed throughout the countries, where
they were subject to control by the local gentry, but they were in fact very nearly nonexistent. Britain’s greatest instrument of organized violence was the Navy, over
which the government enjoyed a monopoly control, but it was not an effective instrument for domestic coercion. The United States developed into a polyarchy without
a standing army or a national police force and with a very wide distribution of firearms among its citizens. Even if the police in Americancommunities were sometimes
involved in politics, control over the police was
Page 50
widely dispersed among innumerable local governments throughout the country. In Switzerland, providing for the defense of the country by universal military service has
made for a microscopic professional standing army.
Where the military is relatively large, centralized, and hierarchical, as it is in most countries today, polyarchy is of course impossible unless the military is sufficiently
depoliticized to permit civilian rule. Why highly organized military forces intervene in politics in some countries but not others has been the subject of an enormous
amount of inquiry, controversy, and puzzlement. The crucial intervening factor, clearly, is one of beliefs. But why beliefs in political neutrality, constitutionalism, and
obedience to civil authority are developed and sustained among military forces only in certain countries (not all of which are polyarchies), raises problems of so vast a
scope that I cannot examine them here, crucial though they may be. The point to be made here is simple and obvious: the chances for polyarchy today are directly
dependent on the strength of certain beliefs not only among civilians but among all ranks of the military. Thus polyarchy has been possible in Chile, where the military
has traditionally been reluctant to intrude into the political arena; while in neighboring Argentina, polyarchy is impossible so long as the military responds to the belief that
its leaders have a right and duty to set aside the results of elections whenever the outcome, in their view, bodes ill for the country.
Although it is obvious that when a government has a monopoly over violence and socioeconomic sanctions and is free to use these resources to suppress oppositions
the chances for competitive politics are practically nonexistent, it does not follow that the mere absence of a governmental monopoly over these key resources
necessarily favors competitive politics. For in some circumstances the lack of these key resources may produce only a weak and unstable competitive regime. Table
4.1 will help to clarify the point.
Page 51
The circumstances most favorable for competitive politics exist when access to violence and socioeconomic sanctions is either dispersed or denied both to oppositions1
and to government. The least favorable circumstances exist when violence and socioeconomic sanctions are exclusively available to the government and denied to the
oppositions. But what of the remaining case when these key resources are a
Table 4.1. Relative Access to Violence and Socioeconomic Sanctions:
Government and Opposition
monopoly of the opposition? The pure case would hardly exist, since under these conditions a “government” would lack the definitional characteristics of a government.
However, the situation may exist temporarily in a country where economic resources are monopolized by a small group of local or foreign owners and managers or
where the military forces are politically committed to the defense of specific social strata or ideologies. Confronted by situations of this kind, a government is bound to
be weak and unstable, for whenever its conduct displeases the opposition, the government can easily be overthrown.
A number of Latin American countries provide a rough approximation to the circumstances I have in mind, not so much because socioeconomic sanctions are
monopolized but
1. To simplify the theory and exposition at this point, I treat “government” and “opposition” each as single, unified actors. Clearly this is rarely the case.
Page 52
because of a tradition of military intervention. Where the military forces are prone to intervene in political life in defense of special interests or their own conception of
the country’s interest, then any government that pursues policies of which they disapprove is likely to be shortlived, as in Argentina.
It would be misleading to conclude, however, that violence and socioeconomic sanctions are necessarily distributed in the same way. Consider table 4.2.
Table 4.2. Distribution of Violence and
Socioeconomic Sanctions
Access to violence is:
Dispersed
or Monopolized
Nuetralized
Access to Dispersed or A B
socioeconomic Nuetralized
sanctions is: Monopolized C D
Clearly the most favorable situation for competitive politics is A, which I shall call a pluralistic social order.
It is equally obvious that the situation least favorable for competitive politics and most favorable for hegemony is D, which I shall call a centrally dominated social order.
The other two situations are more ambiguous. Both are less favorable to political competition than is a pluralistic social order, but both are less favorable to a
hegemonic regime than is a centrally dominated social order. Contemporary Spain, Portugal, and Argentina roughly approximate B, which might be called a quasi
pluralistic social order with repressive violence. The remaining possibility, C, which might be called a quasidominated social order without repressive violence seems to
be rare, perhaps because a governing elite with such great resources for dominance would have no incentives for allowing all the major instruments of violence to be
dispersed or politically neutralized and would probably
Page 53
possess enough resources (for example, legal authority, promotions, pay, and wealth) to prevent it.
Agrarian Societies
Since so many countries in the world today are still predominantly agrarian or barely emerging into the industrial stage, the tendencies of agrarian societies have more
than purely historical interest. Historically, agrarian societies seem to have fallen roughly into two extreme types, which of course, many variations. The most prevalent
type, which might be called the traditional peasant society, has a very high propensity for inequality, hierarchy, and political hegemony.2
The other, which I shall call a free farmer society, is considerably more egalitarian and democratic. Although the free farmer society is often ignored in discussion of
agrarian societies, it furnishes too many important historical examples to pass it by: Switzerland, the United States,3 Canada, New Zealand, and Norway, to take the
leading cases.4
It would be a large and fascinating enterprise to try to unravel the causes leading to one or the other. Tocqueville is an obvious point of departure for the ambitious
theorist. But that effort is beyond the scope of this essay, and I shall do little more here than provide a descriptive summary.
Three underlying conditions seem to be particularly rele
2. Gerhard Lenski, Power and Privilege (New York: McGrawHill, 1966), chaps. 8 and 9; Kaare Svalastoga, Social Differentiation (New York: David McKay,
1965), chap. 3.
3. The absence of a feudal past has strongly been emphasized by Louis Hartz as an explanation for the development of a liberal democracy in the United States. See his
The Liberal Tradition in America (New York: Harcourt Brace, 1955).
4. There are other relevant examples, though these would require greater qualification: e.g. Australia, Chile (nineteenth and early twentieth centuries), Ireland (twentieth
century), and, provided one is prepared to ignore the existence of slavery, Athens in the fifth and fourth centuries B.C. Historically, Sweden was perhaps at the margin
between the two types. Costa Rica is perhaps the closest approximation in Latin America.
Page 54
vant and will help to give our description a bit of dynamics. Tocqueville was perhaps the first to point out how difficult it is to explain the political development of the
United States (as compared, let us say, with other countries to the south) unless one gives considerable weight to the independent effects of beliefs, including naturally
beliefs about “equality.’’5 For a second explanatory factor one might look, as Tocqueville did, to the degree of equality in distribution of land. Because in an agrarian
society the possession of land or a right to the produce of the land is the main source of status, income, and wealth, inequality in land is equivalent to inequality in the
distribution of political resources. To put it differently, in an agrarian society inequalities will be cumulative, not dispersed, and (as Harrington, the seventeenthcentury
English philosopher, argued) power will be highly correlated with landed property. A third factor, one to which Tocqueville paid less attention, is the state of military
technology, that is, the bearing of technology on the capacity of individuals to employ coercion. During some periods, military technology reinforces inequalities by
facilitating a monopoly over instruments of coercion among a small minority, as in the familiar case of the expensively armed and mounted knight before whom the
unarmed or lightly armed medieval peasant was relatively ineffectual. Or to offer another example, the initial monopoly over horses and muskets held by the
Conquistadores allowed a handful of Spaniards to conquer and subjugate the advanced Indian civilizations of Mexico and Peru. In other periods, military technology
reinforces equality by dispersing the most effective instruments of coercion widely over the population, as in the case of the
5. This masterly demonstration, which superbly exhibits his capacity for comparative analysis, will be found in Democracy in America (New York: Vintage Books,
1955), vol. 1, chap. 17, “Principal Causes Which Tend to Maintain the Democratic Republic in the United States,” pp. 298 ff. The main references to Latin America
are on pp. 331–33. He uses Latin America as a kind of “control” for his mental experiment on pp. 331–33.
Page 55
relatively cheap but efficient musket and rifle in eighteenthand and nineteenthcentury America.
In the traditional peasant society, these three factors all operate in the same direction. Cumulative inequalities of status, wealth, income, and means of coercion mean a
marked inequality in political resources, an inequality that is reinforced by prevailing beliefs. A small minority with superior resources develops and maintains a
hegemonic political system (often headed by a single dominant ruler) through which it can also enforce its domination over the social order and hence strengthen the
initial inequalities even more. Limits on this potentially runaway cycle of everincreasing inequalities are set by dangers of mass starvation, passive resistance, and even
sporadic uprisings among the peasants, a decline in agricultural output, and, because of wide disaffection, vulnerability to foreign invasions. But for the great bulk of the
population life is one of hardship, deprivation, dependence, repressed dissent, and comparative ignorance,6 while a tiny minority enjoy exceptional power, wealth, and
social esteem.7 The dynamics of the traditional peasant society might then be represented crudely as in figure 4.1.
In the contrasting society of free farmers, land is more equally distributed, even though it is always a far cry from perfect equality. If the norms are egalitarian and
democratic,
6. Mehmet Beqiraj, Peasantry in Revolution (Ithaca, N.Y.: Center for International Studies, Cornell University, 1966), passim.
7. “For example, recent research indicates that in nineteenth century China, the gentry or degree holders, who formed the governing class, totaled about 1.3 per cent of
the population in the first half of the century and about 1.9 per cent toward the end. In midnineteenth century Russia the nobility constituted 1.25 per cent of that
nation’s population. In France, on the eve of the Revolution, the nobility of all ranks and grades constituted only 0.6 per cent, despite the recent influx of many wealthy
mercantile families. During the last days of the Roman Republic, the governing class is estimated to have included about 1 per cent of the capital’s population. Finally, in
seventeenth century England, peers, baronets, knights, and esquires combined constituted roughly 1 per cent of the total population.” Lenski, Power and Privilege, p.
219.
Page 56
FIGURE 4.1 Dynamics of the Peasant Society
as Tocqueville insisted they were in the United States, then the one reinforces the other. Finally, in a number of cases both of these tendencies toward equality (or
toward a lower limit on inequality) are strengthened by certain aspects of military
FIGURE 4.2 Dynamics of the Free Farmer Society
technology. In the United States, the musket and later the rifle helped to provide a kind of equality in coercion for over a century. In Switzerland the mountains, in
Norway and New Zealand the mountains and fjords, the continental proportions and the enormous length of Chile—all reduced the prospects for a successful
monopoly of violence by any one stratum of the population.8 The way these factors interact in a society of free farmers is represented by figure 4.2.
8. Nor should one underestimate the effects on the use of violent coercion of beliefs and norms toward law, order, and individual violence. Of the two Englishspeaking
countries in North America, Canada has traditionally been the more lawabiding and the less violent. See, for example, Seymour Martin Lipset, Revolution and
Counterrevolution (New York: Basic Books, 1968), pp. 37–39. A Canadian writer has argued that these two cultures even displayed themselves during the Gold
Rush:
Canadian and U.S. mining camps grew up with varying legal customs, which to a considerable extent point up the very real difference between the Canadian and
the U.S. character. The
Page 57
Commercial and Industrial Societies
Historically, commercial and industrial societies have been more hospitable than agrarian societies to competitive politics. Orthodox liberal doctrine explained this by
establishing a connection between a pluralistic social order and a privately owned competitive economy: competitive politics requires a competitive economy. In effect
classic liberal doctrine set forth the following equation:
Competitive politics a pluralistic social order
a competitive economy private ownership9
American, freed by his own will of what he considered colonial bondage, has always insisted on running his own affairs from the ground up—especially on the
frontier. The Canadian, who never knew the blood bath of revolution, has more often preferred to have law and order imposed from above rather than have it
spring from the grass roots.
In the three British Columbia gold rushes, constabulary and courts of justice enforced a single set of laws in the British colonial tradition. Mining law was the same
everywhere, and the gold commissioner in charge had such absolute power that the lawlessness so familiar to American mining history was unknown in the B.C.
camps.
But in the Rocky Mountain camps of the U.S.A., and later in Alaska, each community had its own customs and its own rules made on the spot. Authority was
vested in the miners themselves, who held town meetings in the New England manner to redress wrongs or dispense justice … On Alaskan territory, during the
hectic days of 1897–98, there was no organized machinery of government, to speak of: rule was by local committee, sometimes wise, sometimes capricious,
always summary. On the Canadian side there was, if anything, too much government, as the graft in Dawson City was to demonstrate; but there were also, at
every bend in the river, the uniformed and strangely comforting figures of the Mounted Police.
Pierre, Berton, The Klondike Fever (New York: Knopf, 1958), pp. 23–24.
9. The double arrow can be read as “implies” or “requires.” Reading in the reverse direction from right to left, the symbol means “is a necessary condition for.”
Page 58
For, it was argued, just as toleration of oppositions and the existence of a competitive representative government require a pluralistic social order, so a pluralistic social
order in turn requires a competitive capitalist economy. At the same time, classical liberal thought argued that in order for a socialist economy to exist—and socialism
was understood to be the only modern alternative to capitalism—there would have to be a completely centralized social order with access to social, economic, and
physical sanctions concentrated in the hands of central authorities; obviously such a social order would require (and make possible) a hegemonic regime. The twin
equation, then, was:
A socialist economy a centrally dominated social order
a hegemonic regime
Thus classic liberalism rigidly premised the existence of competitive politics and later of polyarchy on competitive capitalism: it said, in effect, that you cannot logically
choose to have the freedoms associated with competitive politics unless you also choose to have a competitive capitalist economy; if you choose to have a socialist
economy, logically you are also opting for a hegemonic regime and the destruction of political liberties. After the Bolshevik Revolution, the Soviet Union could be cited
as proof of these two equations, for there a highly hegemonic political system maintained a centrally dominated social order, a key element of which was its completely
centralized socialist economy.
This analysis, though superficially persuasive, did not really demonstrate, however, that the equations were correct, and other historical developments have revealed
their inadequacy.
Classical liberal economists like Adam Smith knew from the history of mercantilism that private ownership is not a sufficient condition for economic competition; the
first equation only specifies that it is a necessary condition. Further experience with numerous dictatorships—in Italy, Germany, Japan, Spain, and elsewhere—has
shown that private owner
Page 59
ship is indeed no guarantee of a competitive economy or a political order that permits public contestation, much less polyarchy. The extreme cases—as in Italy,
Germany, and Japan—demonstrated that a kind of private ownership10 could coexist even with a centrally dominated social order.
Since the equations speak of necessary and not sufficient conditions, strictly speaking these developments left the argument intact. But other developments have actually
falsified the equations. One is the persistence of inclusive polyarchy in countries with mixed (not strictly competitivecapitalist) economies that employ an endless variety
of techniques and controls which, in combination, preserve and may even strengthen a pluralistic social order. One thinks of Sweden as the archetype. In 1959
expenditures of government, social security, and public enterprises were 53 percent of Sweden’s GNP.11 But practically all industrialized countries with polyarchal
regimes have displaced pure competitive capitalism with mixed systems, and in the process they have managed to maintain pluralistic social orders.
Where the equations of classic liberalism went wrong was in supposing that any alternative to competitive capitalism necessarily required a centrally directed economy,
whereas in fact competition among privately owned firms is by no means the unique method of decentralizing an economy.
10. Of course, this kind of ownership might not be equivalent to some definitions of ‘‘private” ownership. The term might be defined so that a centrally dominated social
order would by definition exclude the existence of private ownership of the means of production and distribution.
11. Other figures were: U.K. 45%; Austria (with not all public enterprises included), 44%; and New Zealand, 43%. Russett et al., World Handbook of Political and
Social Indicators (New Haven: Yale University Press, 1964), table 15, p. 63. It has been estimated that in Austria “75% of the total corporate capital is directly or
indirectly in the public domain.” Alexander Vodopivec, Wer Regiert in Osterreich? 2d ed. (Vienna: Verlag für Geschichte und Politik, 1962), p. 255, cited in Frederick
C. Englemann, “Austria: The Pooling of Opposition,” in Robert A. Dahl, ed., Political Oppositions in Western Democracies (New Haven: Yale University Press,
1966), p. 270.
Page 60
Indeed, in recent years some of the communist regimes of Eastern Europe have been moving away from central direction; among these, Yugoslavia has gone farthest in
decentralizing controls over economic enterprises. If decentralized socialist economies prove capable of handling major economic problems with a fair degree of
success, then there is no inherent reason why socialism cannot produce and sustain a highly pluralistic social order, and hence competitive politics.
The correct equations, in short, seem to be:
Competitive politics pluralistic social order
decentralized economy
Highly centralized economy centrally dominated social order
hegemonic regime
The argument of this chapter can thus be summed up as follows:
1. A competitive political regime, and therefore a polyarchy, is unlikely to be maintained without a pluralistic social order. A centrally dominated social order is more
favorable to a hegemonic than to a competitive regime (and therefore to a polyarchy).
2. A competitive regime cannot be maintained in a country where the military or police forces are accustomed to intervening in politics, even if the social order is
otherwise pluralistic and not centrally dominated.
3. Agrarian societies seem to approach two extreme types, the traditional peasant society characteristically associated with a hegemonic political regime and the society
of free farmers characteristically associated with a competitive regime and evolution toward an inclusive polyarchy. The main factors determining the direction an
agrarian society takes
Page 61
seem to be: norms about equality, the distribution of land, and military techniques.
4. Private ownership is neither a necessary nor a sufficient condition for a pluralistic social order and hence for public contestation and polyarchy.
5. A pluralistic social order, and hence public contestation and polyarchy, can exist in a country with a decentralized economy, no matter what the form of ownership.
6. But public contestation, and hence polyarchy, is unlikely to exist in a country with highly centralized direction of the economy, no matter what the form of ownership.
Page 62
5. THE SOCIOECONOMIC ORDER:
LEVEL OF DEVELOPMENT
It is widely assumed that a high level of socioeconomic development not only favors the transformation of a hegemonic regime into a polyarchy but also helps to
maintain—may even be necessary to maintain—a polyarchy. To what extent is this assumption correct?
Earlier answers that leaned heavily on impressionistic interpretations and a case study or two have recently given way to more ambitious efforts to exploit a rapidly
increasing supply of crossnational data and easy access to suitable computer programs for handling these data. Even if the data, the measures used, and the
interpretations do not yield unambiguous or unchallenged answers, recent studies1 do give impressive support to several propositions.
1. Seymour Martin Lipset, “Economic Development and Democracy,” in his Political Man (New York: Doubleday, 1960), pp. 45–76; Phillips Cutright, “National
Political Development, Its Measurement and Social Correlates,” in Nelson W. Polsby, Robert A. Dentler, and Paul A. Smith, Politics and Social Life (Boston:
Houghton Mifflin, 1963), pp. 569–81; Everett E. Hagen, ‘‘A Framework for Analyzing Economic and Political Changes,” in Development of the Emerging Countries:
An Agenda for Research (Washington, D.C.: The Brookings Institution, 1962), chap. 1; Irma Adelman and Cynthia Taft Morris, Society, Politics and Economic
Development: A Quantitative Approach (Baltimore: The Johns Hopkins Press, 1967); Deane E. Neubauer, “Some Conditions of Democracy,” American
Page 63
Some Settled Questions
First, various measures of socioeconomic level are all highly intercorrelated. Thus Russett writes that
In a study of over 100 countries and colonies there was found to be a high correlation between G.N.P. per capita and such other indicators of economic and social
development as the percentage of the population in cities of over 20,000 (r = .71), the percentage of adults literate (.80), the proportion of the population enrolled in
higher education (.58), the number of radios per 1000 population (.85), and the number of hospital beds per inhabitant (.77). These variables, along with others like the
percentage of the labor force employed in agriculture, the percentage of wage and salary earners in the population, and other indices of health and mass
communications, form a cluster that tends to vary according to the level of economic development.2
If a country is relatively poor or relatively rich, then, its poverty or affluence shows up in all sorts of ways in addition to per capita income. The commonsense notion
that different countries are at different “stages” of economic and social development is amply confirmed by the data.3
Second, there is unquestionably a significant association between socioeconomic level and “political development.”
Political Science Review 61 (December 1967): 1002”09; Bruce M. Russett, et al., World Handbook of Political and Social Indicators (New Haven, Yale University
Press, 1964), pp. 293–303.
2. Bruce M. Russett, Trends in World Politics (New York: Macmillan, 1965), pp. 125–26.
3. For example, nine indicators applied to 107 countries yield five “stages” of economic and political development in Russett, Trends in World Politics, p. 127. Factor
scores used as indicators yield three levels of socioeconomic development for 74 “less developed countries” ranging from Niger with a 1961 per capita GNP of $40 to
Israel with a 1951 per capita GNP of $814 in Adelman and Morris, Society, politics and Economic Development, p. 170.
Page 64
Expressed in terms of the questions raised in this essay, there can no longer be any doubt that competitive politics and socioeconomic level do tend to run together. The
data show rather conclusively that:
The higher the socioeconomic level of a country, the more likely it is to have a competitive political regime.
The more competitive a country’s political regime, the more likely the country is at a relatively high level of socioeconomic development.
Thus Russett finds that among countries in the two most “advanced” stages of economic development (which he calls Industrial Revolution Societies and High Mass
Consumption Societies) the proportion of “competitive” and “semicompetitive” systems is much higher than among the countries in the three less advanced stages,
where “authoritarian” regimes predominate. (Table 5.1)4
4. Adelman and Morris in their factor analysis of 74 ‘‘less developed” countries find that per capita GNP is correlated with both a social factor and a political factor
(table IV, p. 151). The political factor is essentially the extent and strength of competitive political institutions. “An increase in this factor may be interpreted to represent
a movement along a scale that ranges from centralized authoritarian political forms to specialized political mechanisms capable of representing the varied group interests
of a society and aggregating these interests through participant national political organs … Thus a positive change in Factor II is composed of (1) increase in the
effectiveness of democratic institutions, freedom of political opposition and press, competitiveness of political parties, and strength of the labor movement; (2) a
movement from political parties emphasizing considerations of national unity toward those stressing ideological platforms; and (3) decreases in the strength of the
military and in the extent of centralization” (pp. 155–56). Additional evidence is also provided by Arthur S. Banks and Robert B. Textor, A CrossPolity Survey
(Cambridge: M.I.T. Press, 1963): see, for example, the characteristics of “polities where the electoral system is competitive, rather than noncompetitive” versus
“polities where the electoral system is noncompetitive, rather than competitive” (FC 104). See also FC 101, FC 107, FC 139. And see also Cutright, “National
Political Development: Its Measurement and Social Correlates,” table 1, p. 577.
Page 65
A relationship between socioeconomic development and inclusive or nearly inclusive polyarchy shows up perhaps even
Table 5.1. Economic Development and Political System
“Stage” of development
Political system I II III IV V
Competitive 13% 33% 12% 57% 100%
Semicompetitive 25 17 20 13 0
Authortitarian 63 50 68 30 0
N (8) (12) (25) (30) (14)
Source: Bruce M. Russett, Trends in World Politics (New York: Macmillan, 1965), table 8.2, p. 140.
Note: Percentages do not always add to 100% because of rounding.
more sharply in the data. (Tables 5.2 and 5.3) As with competitive regimes, so with polyarchy:
The higher the socioeconomic level of a country, the more likely that its regime is an inclusive or nearpolyarchy.
If a regime is a polyarchy, it is more likely to exist in a country at a relatively high level of socioeconomic development than at a lower level.
The major propositions advanced so far are, on the evidence, pretty much beyond dispute: it makes sense to speak of “levels” or “stages,” for all the major
socioeconomic indicators are highly interrelated and tend to change together. And not only competitive politics in general but polyarchy in particular are significantly
associated with relatively high levels of socioeconomic development. The answer to the question posed at the beginning of this section is, then, reasonably clear: The
chances for political competition do indeed depend on the socioeconomic level of the society.
Yet even though the argument up to this point is well supported by all the evidence now available, it does not take us
Page 66
Table 5.2. Distribution of 29 Polyarchies
by Levels of Socioeconomic Development
Polyarchies
Percent of
GNP countries at the
Total per capita same socio Percent of all
N Range Mean N economic level polyarchies
“Traditional primitive” societies 11 $45–64 $56 0a 0 0
“Traditional civilizations’’ 15 70–105 87 1b 6.7% 3.5%
“Traditional societies” 31 108–239 173 1c 3.6 3.5
“Industrial Revolution” societies 36 262–794 445 13 36.0 45.0
“High mass consumption” societies 14 836–2577 1330 14 100.0% 48.0
Totals 107 29 100.0%
Note: Polyarchies classified by the author; see appendix table A3. The six nearpolyarchies in table A.3 are excluded from the polyarchies above. Levels of
socioeconomic development are from Bruce M. Russett et al., World Handbook of Political and Social Indicators (New Haven: Yale University Press, 1964), p. 294.
The number of countries in this table is slightly larger than in table 5.1, which is from Russett”s Trends in World Politics. Russett’s five stages use six socioeconomic
indicators and three political indicators (percent voting, percent military personnel, and expenditures of the central government). While relationships among the six
socioeconomic indicators are strong, their relation to these particular “political” variables is quite weak, and the location of a country among the five “stages” seems to
be independent of the three political indicators. Consequently, in spite of the slight contamination of “percent voting” with our measure of polyarchy, the relationship
between polyarchy and the five “stages” given by Russett does not appear to be spurious.
a Somalia, which had maintained an inclusive polyarchy from independence in 1960 to 1969, was not included by Russett et al. in this classification. Somalia’s per
capita GNP is given as $57 (ibid, p. 157). In October 1969, the Somali Army seized power, arrested all government ministers, and dissolved the National Assembly.
b India.
c The Philippines.
very far. In fact, it leaves a number of crucial questions unanswered about the nature and strength of the very general relation between competitive politics and
socioeconomic “level.”
Page 67
Table 5.3. Polyarchy and Socioeconomic Development
Polyarchies
and
Socioeconomic All countries nearpolyarchies
indicator N Mean N Mean
Population in localities over 20,000 120 23% 31 38%
GNP per capita, 1957 122 $377 32 $822
Wage and salary earners 79 35% 31 42%
Labor force in agriculture 98 50% 27 19%
Nonagricultural employment as % of workingage pop. 77 36% 31 46%
Employment in industry as % of working pop. 78 15% 31 20%
Enrollment in higher education per 100,000 pop. 105 281 33 499
Enrollment in primary and secondary education as % of pop. aged 5–19 125 43% 33 62%
Literate, 15 years and over 118 52% 33 82%
Note: Figures for all countries are from Russett et al., World Handbook. Figures for polyarchies and nearpolyarchies are computed from data supplied by Yale
Political Data Program. Variations in number of countries occur because of missing data. Classification of polyarchies by the author; see appendix table A3.
Some Unsettled Questions
Are there “thresholds”?
One such question is whether there are “thresholds” below or above which the chances for competitive politics or polyarchy do not change enough to matter. To pose
the problem in a different way, is the relationship linear or curvilinear?
An inspection of the evidence (for example, in table 5.2 and 5.3) strongly suggests that the relationship is not linear, and instead that:
There exists an upper threshold, perhaps in the range of about $700–800 GNP per capita (1957 U.S. dollars),
Page 68
above which the chances of polyarchy (and hence of competitive politics) are so high that any further increases in per capita GNP (and variables associated with such
an increase) cannot affect the outcome in any significant way.
There exists a lower threshold, perhaps in the range of about $100–200 GNP per capita, below which the chances for polyarchy (although not necessarily other forms
of competitive politics) are so slight that differences in per capita GNP or variables associated with it do not really matter.
What about the deviant cases?
Yet even if we accept the idea of “thresholds,” it is definitely not true that competitive regimes or even polyarchies exist only in countries at a high level of
socioeconomic development. Nor is it true that all countries at a high level of socioeconomic development have polyarchies, or even competitive regimes. Any ranking
of a considerable number of countries along the dimensions of economic or socioeconomic development and political competition or polyarchy invariably displays a fair
number of deviant cases.5 Discrepancies and deviant cases are furnished, for example, by India, a
5. For examples see Hagen’s “Classification of Asian and African Countries by Type of Political Structure and Rank in Economic Development’’ and also his similar
classification of Latin American countries in “A Framework for Analyzing,” Tables 1.1, 1.2, pp. 2, 4; and figure 1, “Relationship of Political Development to
Communications Development: 71 nations,” in Cutright, “National Political Development, Its Measurement and Social Correlates,” pp. 572–73. See also James S.
Coleman’s “Composite Rank Order of Latin American Countries on Eleven Indices of Economic Development” and the similar rank ordering of Asian and African
countries in Gabriel A. Almond and James S. Coleman, The Politics of the Developing Areas (Princeton: Princeton University Press, 1960), pp. 541–42. Also
Adelman and Morris, “Scatter Diagram Relating per capita GNP and Country Scores on Factor Representing the Extent of Democracy,” Society, Politics, and
Economic Development, p. 262.
Page 69
competitive regime—indeed a polyarchy—in a country with a 1957 GNP per capita of about $73; by the USSR and East Germany, both hegemonic regimes at high
socioeconomic levels (1957 GNP: $600); by the puzzling political contrasts among four Latin American countries at comparatively high levels of socioeconomic
development—Argentina ($490), Chile ($379), Cuba ($431), and Uruguay ($478); among “transitional” societies by the existence of competitive regimes in the
Philippines ($220), Turkey ($220), and Ceylon ($128) but not in Paraguay ($114), Indonesia ($131), Egypt ($142), and Portugal ($224).
If the contemporary world furnishes a large supply of discrepant examples, history also provides its share. How do we account for the early appearance of an inclusive
polyarchy in the United States? Goldsmith has estimated that at about the time Tocqueville wrote Democracy in America, American GNP per capita was around
$350–400.6 But competitive politics was established in the United States well before Tocqueville’s visit. The country already had created an inclusive polyarchy (for
whites) by 1800, when per capita GNP would have been considerably lower than in 1840. What is more, according to the usual indicators of socioeconomic level, the
United States in 1800 was definitely agricultural, premodern and nonindustrial. By 1820, there were only five cities larger than 50,000. Only about 3 percent of the total
population lived in cities over 25,000. As much as 93 percent of the population lived in strictly rural territory. Of the total working force, some 70 percent were
6. In 1957 prices. Raymond Goldsmith, “Long Period Growth in Income and Product, 1939–1960” in Ralph Andreano, ed., New Views on American Economic
Development (Cambridge, Mass.: Schenkman, 1965), chart II, p. 357. Goldsmith calculated GNP in 1929 prices. I have converted to 1957 prices using U.S.,
Department of Commerce, U.S. Income and Output (Washington, D.C.: Government Printing Office, 1958), table VII2, “Implicit Price Deflators for Gross National
Product or Expenditure 1929–1957,’’ pp. 220–21.
Page 70
engaged in agriculture.7 And of course no families had telephones, radios, or automobiles. A social scientist armed only with the data examined so far—and the
theories often used to explain these data—might justifiably conclude that in the early nineteenth century there was scarcely a chance for the development of democracy
in America, yet I suspect that most of us continue to find Tocqueville’s interpretation more convincing.
What holds for the United States holds not only for Australia, New Zealand, and Canada, but in some degree for Britain, Norway, Sweden, and a number of other
European countries where competitive politics (though not inclusive polyarchy) existed in the nineteenth century. By the indicators applied to the contemporary world,
these countries were then at low levels of socioeconomic development.
That competitive politics is undoubtedly associated in some way with socioeconomic development is not, it seems, a very satisfactory—nor perhaps even a very
interesting—conclusion. What is more tantalizing is the fact that the association is weak, that the conclusion ignores a number of critical deviant cases, and that the
relationship of one to the other is unexplained. Among other mysteries in the relationship is the direction of causation.
What about the causal direction?
Do high levels of socioeconomic organization and productivity “cause” competitive politics? Does competitive politics, conversely, induce socioeconomic development?
Do competitive politics and socioeconomic development interact and reinforce one another? Or, finally, are both caused by something else?
As the authors of the studies cited earlier frequently warn, to demonstrate that a relationship exists tells nothing about
7. The data are from U.S., Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1957 (Washington, D.C.: Government Printing Office,
1960), pp. 14 and 72.
Page 71
causes.8 Causes can be teased out of data only with the aid of theory.
One thing seems clear, however: whatever the causal relationships may be, they are not simple and onedirectional.
Causal theory that could account for both the general tendency and the deviant cases would have to be complex. For the evidence simply does not sustain the
hypothesis that a high level of socioeconomic development is either a necessary or a sufficient condition for competitive politics nor the converse hypothesis that
competitive politics is either a necessary or a sufficient condition for a high level of socioeconomic development.9
I do not believe it is possible at this time to advance acceptable causal theory that will account for all the cases. All I hope to de here is to offer some explanations that
will help us to understand both the general tendency and the deviant cases. I do not offer what follows as in any sense a complete theory.
Are preindustrial societies inherently unsuitable
for competitive politics?
The relation between competitive politics and preindustrial societies poses a paradox. In the nineteenth century, in certain preindustrial societies, competitive politics and
in some cases even polyarchy flourished:the United States, Australia, New Zealand, Canada, Norway, Sweden, to name a few. But in the contemporary world, few
countries in the preindustrial stages have polyarchal regimes; in fact they are much more likely to have hegemonic or authoritarian regimes.
8. See, for example, the disclaimer of Adelman and Morris, Society, Politics, and Economic Development, p. 148.
9. For a similar critique, see Dankwart A. Rustow, “Democracy, Consensus, and the New States,” presented to the International Political Science Association,
Brussels, September 1967; and his ‘‘Transitions to Democracy: Toward a Dynamic Model,” Comparative Politics 2 (April 1970): 337–64.
Page 72
The solution to this paradox is to be found in the fact that the European and Englishspeaking preindustrial, agrarian societies of the nineteenth century (not to mention
the Athens of Pericles) were in many respects very different from the preindustrial societies of the modern world. Doubtless preindustrial United States would contrast
most sharply with a contemporary preindustrial society, which is ordinarily marked by widespread illiteracy, a traditionbound, preliterate, prescientific culture, weak or
fragmented systems of communication, severe inequalities of wealth, status, and power, a tiny or nonexistent independent middle class, and frequently a tradition of
autocratic or authoritarian rulership. One has only to read Tocqueville to see how radically different on each of these characteristics was the America he saw.10
Added to all these structural differences there is frequently an important difference in the role assigned to the state in socioeconomic development. To be sure, the state
was never a negligible force in the complex processes of transforming agricultural societies into urbanized industrial societies in
10. Even making due allowance for exaggeration by Tocqueville, it is obvious that no contemporary would describe a preindustrial society today in these terms: “The
social condition of the Americans is eminently democratic; this was its character at the foundation of the colonies, and it is still more strongly marked at the present day
… Even the germs of aristocracy were never planted in [New England] … But wealth circulates with inconceivable rapidity, and experience shows that it is rare to find
two succeeding generations in the full enjoyment of it … I do not believe that there is a country in the world where, in proportion to the population, there are so few
ignorant and at the same time so few learned individuals. Primary instruction is within the reach of everybody; superior instruction is scarcely to be obtained by any …
In America there are but few wealthy persons … In America most of the rich men were formerly poor … In America the aristocratic element has always been feeble
from its birth … We can scarcely assign to it any degree of influence on the course of affairs … Men are there seen on a greater equality in point of fortune and
intellect, or, in other words, more equal in strength, than in any other country of the world, or in any age of which history has preserved the remembrance.” Democracy
in America (New York: Vintage Books, 1955), 1: 48–55.
Page 73
the nineteenth and twentieth centuries; governments, even in the United States, generally played a key part. But ordinarily they did not have a commanding role;
economic development was more “autonomous” than “induced.”11 By contrast, in many preindustrial countries today political leaders are committed to using all the
means of inducement and coercion available to the state in order to transform or displace the traditional and often stubbornly resistant institutions of the older society.
Thus wherever competitive politics was highly developed in preindustrial societies during the nineteenth century, political leaders were essentially commited to an
outlook and strategy that left most of the initiative for development to nongovernmental groups. Among preindustrial societies today, the leadership is much more prone
to the outlook and strategy of dirigisme. The first perspective helped to produce a considerable measure of autonomy and decentralization in the social order. In the
perspective more characteristic of today’s world, autonomy and decentralization merely perpetuate the traditional society and impede the transformations needed for
economic growth. Hence the strategies of change employed by leaders in contemporary preindustrial societies are more likely to emphasize a compelling need for
centralization and hegemony.
If some deviant cases in this century make perfectly clear that industrialization and urbanization are not sufficient conditions for competitive politics (for example, the
USSR or Germany in the 1930s), the historical deviant cases demonstrate that industrialization and urbanization are not even necessary conditions for competitive
politics.12 Preindustrial,
11. Bert F. Hoselitz, Sociological Aspects of Economic Growth (Glencoe: The Free Press, 1960), pp. 74 and 97 ff.
12. Ignoring the historically important deviant cases in which competitive politics did develop in a preindustrial setting leads, I think, to an exaggerated emphasis on the
importance of urbanization in the development of democratic systems. For an example, see Donald J. McCrone and Charles F. Cnudde, “Towards A Communications
Page 74
rural, agrarian societies are definitely not inherently unsuitable for competitive politics or even for polyarchy. For some preindustrial, rural, agrarian societies have had
competitive political systems—have sometimes, indeed, provided a marvelous foundation for an inclusive polyarchy.
If a preindustrial society is a poor setting for competitive politics or polyarchy in the modern world, surely this is a consequence of social characteristics such as
illiteracy,poverty, a weak middle class, and an authoritarian political culture. Today these characteristics are associated with a weak industrial and urban base. But they
are not—or at any rate were not—inherent features of preindustrial societies.
Explaining the Relationship
Assuming, now, that there is a relationship between public contestation (and polyarchy) and level of socioeconomic development, that there are important exceptions,
and that there may be thresholds below and above which the chances for polyarchy or public contestation do not change significantly—how can we account for all this?
A very general hypothesis will help, I think, to establish the connection between political system and socioeconomic level:
The chances that a country will develop and maintain a competitive political regime (and, even more so, a polyarchy) depend up on the extent to which the country’s
society and economy
(a) provide literacy, education, and communication,
(b) create a pluralistic rather than a centrally dominated social order,
(c) and prevent extreme inequalities
among the politically relevant strata of the country.
Theory of Democratic Political Development: A Causal Model,’’ American Political Science Review 61 (March 1967): 72–79.
Page 75
I shall examine briefly the first two conditions in this chapter. Because of its importance and complexity, the next chapter is devoted entirely to the third.
Literacy, education and communication
It is surely unnecessary to argue here that whenever the citizen body is large the chances for extensive participation and a high degree of public contestation depend to
some degree on the spread of reading, writing, literacy, education, and newspapers or their equivalents. I do not propose to explore the exact nature of this
dependence and the possibilities of compensating for illiteracy in various ways, as in India or Turkey. The relevant point is that the extent of literacy, education,
newspapers, and other forms of communication is—as we saw earlier—related to urbanization and industrialization. The development of cities, commerce, industry,
and the professions not only requires but also fosters these elemental requirements.
However, a moderately educated people with a generous supply of newspapers (or, today, access to radios and television sets) does not require a highly industrialized
or urbanized society. After all, as Tocqueville pointed out, most white Americans were literate in the early nineteenth century; opportunities for a modest education
were fairly widespread (though perhaps less than Tocqueville thought), newspapers were generally available, and political communication seems to have been
moderately efficient despite the great area over which Americans were spread. In other countries, too, widespread literacy and education have preceded extensive
industrialization, the growth of cities, and high average per capita incomes: New Zealand, Australia, Canada, Norway, Iceland, and Finland are all examples. For the
costs of providing a general access to education and news media are not so high that they cannot be borne by moderately prosperous agrarian societies.
Page 76
It seems reasonable, then, to conclude that:
The hypothetical lower threshold for competitive politics could be accounted for, in part, by the difficulty countries below this level have in mobilizing the resources
required for widespread literacy, education, and news media.
Nonetheless, these minimal needs for competitive politics, and particularly for polyarchy, can be met by countries above this lower threshold, even though these
countries are predominantly agricultural, rural, and unindustrialized.
Thus the need to meet these minimal requirements helps to account for the lower threshold but does not account for the general relationship.
A pluralistic social order
Consider now what a relatively “advanced economy” both makes possible by its performance and requires for its performance. An advanced economy not only can
afford but also requires the reduction of illiteracy, the spread of universal education, widespread opportunities for higher education, and a proliferation of means of
communication. Not only can it afford to produce an educated labor force, but it needs one: workers able to read and write, skilled workers who can read blueprints
and respond to written directions, engineers, technicians, scientists, accountants, lawyers, managers of all kinds. Not only does it produce but it must have speedy and
reliable systems of communication, including systems that transmit a vast amount of public or quasipublic information. It not only makes possible but at the same time
requires a multiplicity of durable and highly specialized organizations manned by strongly motivated staffs who are loyal to the goals of the organization: factories, banks,
stores, schools, universities, hospitals, mass transit systems, and
Page 77
thousands upon thousands of other types of organizations.
Because of its inherent requirements, an advanced economy and its supporting social structures automatically distribute political resources and political skills to a vast
variety of individuals, groups, and organizations. Among these skills and resources are knowledge; income, status, and esteem among specialized groups; skill in
organizing and communicating; and access to organizations, experts, and elites. These skills and resources can be used to negotiate for advantages—for oneself, for a
group, for an organization. Groups and organizations develop a thrust toward autonomy, internal and parochial loyalties, complex patterns of cohesion and cleavage.
When conflicts arise, as they inevitably do, access to political resources helps individuals and groups to prevent the settlement of the conflict by compulsion and
coercion and to insist, instead, upon some degree of negotiation and bargaining—explicit, implicit, legal, alegal, illegal. Thus systems of bargaining and negotiation grow
up within, parallel to, or in opposition to hierarchical arrangements; and these systems help to foster a political subculture with norms that legitimate negotiating,
bargaining, logrolling, give and take, the gaining of consent as against unilateral power or coercion.
Even within ostensibly hierarchical organizations, leaders learn that compulsion and coercion are often damaging to incentives. In an advanced economy, longrun
performance under threat or coercion is less productive at all levels than a more willing performance based upon voluntary compliance. Thus the fear of punishment for
bad performance is not merely supplemented but in many respects displaced by the expectation of rewards for successful performance. Just as slave labor is in general
less efficient than free labor, so badly paid, discontented workers are less productive in the long run than highly paid, contented workers. For technicians, executives,
scientists, and intellectuals the need for a measure of willing performance, based on their “consent,”
Page 78
is even greater. And a large measure of autonomy and discretion is also found to produce better results than rigid, overcentralized supervision.
Thus an advanced economy automatically generates many of the conditions required for a pluralistic social order. And as a pluralistic social order evolves, at least in an
elementary form, some of its members make demands for participating in decisions by means more appropriate to a competitive than to a hegemonic political system.
If we use a single arrow with a C to suggest the direction of causation, the argument might be represented as follows:
Advanced economy — C pluralistic social order — C
demands for a competitive political system
The argument as it stands is, to be sure, oversimple. It requires at least three qualifications:
First, even if an advanced economy creates some of the conditions required for a pluralistic social order, it does not create all the conditions required: witness the
USSR and East Germany, which combine rather advanced economies with centrally dominated social orders.
As we have already observed, the fit between economic “level” and political system is loose; yet just as the fit is better at very low levels (polyarchies rare), so it is
better at very high levels (hegemonic systems rare). Our argument implies—rightly, I think—that as countries with hegemonic systems move to high levels of economic
development (for example, the USSR and the Eastern European countries) a centrally dominated social order is increasingly difficult to maintain. For if our argument is
correct, economic development itself generates the conditions of a pluralistic social order. The monopoly over socioeconomic sanctions enjoyed by the hegemonic
leaders is therefore undermined by the very success of their economy: the more they succeed in transforming the economy (and with it, inevitably, the society) the more
they are threatened with political failure.
Page 79
If they allow their monopoly over socioeconomic sanctions to fragment and yet seek to retain their political hegemony by exploiting their monopoly over violence—a
transformation from a centrally dominated social orderto what was called earlier a quasipluralistic social order with repressive violence—then they confront the
enormous limitations, costs, and inefficiencies of violence, coercion, and compulsion in managing an advanced society where incentives and complex behavior are
needed that cannot be manipulated by threats of violence. 13
The tensions within a hegemonic regime in a society at a high level of development might be indicated as follows, where the jagged, doubleended arrows represent
conflict.
13. For a similar view see Alexander Eckstein, “Economic Development and Political Change in Communist Systems,” World Politics 22 (1970): 475–95. George
Fisher argues that the Soviet authorities may be able to contain pressures for social pluralism and liberalization without imparing economic incentives. If I understand his
argument correctly, he foresees a “limited monism’’ in which “autonomy from state control” is denied to “major social groups in the society” but permitted “in
professional and personal life, in technical activities (including the economy), and in some parts of law and public affairs.” The Soviet System and Modern Society
(New York: Atherton, 1968), pp. 14–18, 142–53. However, I am unable to see how the evidence in his book on the emergence of “dual executives” in the USSR
supports his argument.
Page 80
Second, while economic “success” may threaten hegemonies by generating demands for political liberalization, economic success has not threatened polyarchies, but
economic failure has. For economic difficulties, particularly when they take the form of severe unemployment or rapid inflation, generate demands for a hegemonic
regime and a centrally dominated social order.
Third, it may be that the sharpness of this difference is already being blunted, for it is now becoming evident that affluent societies generate their own frustrations and
discontents. Although affluence may increase the pressures for competitive politics in countries now ruled by hegemonic regimes, it is far from clear that affluence will
continue to strengthen allegiance to democracy in countries that already have inclusive polyarchies.
Page 81
6. EQUALITIES AND INEQUALITIES
Ever since Aristotle, and probably since the preSocratic philosophers, it has been commonly held among political theorists that extreme inequalities help to produce
hegemonic regimes and that nonhegemonic systems of a more egalitarian sort must contain a preponderant middling group of citizens more or less equal and therefore
must avoid extreme differences (among citizens) in status, income, and wealth. Advanced industrial societies harbor powerful tendencies toward extreme inequalities,
and yet—a phenomenon the Greeks could not foresee—inclusive polyarchies have flourished most in advanced industrial countries. This seeming contradiction has
given rise to many explanations. Some seek to solve the puzzle by explaining away the inequalities in these countries, while others solve it by explaining away the
“democracy”: their outwardly “democratic” regimes are really hegemonies in disguise.
The bearing of the social order and the economic level on equality has already been touched on at a number of points in our discussion, but it is time now to bring
certain inconclusive matters into sharper focus.
Equalities and inequalities in a society seem to affect the chances for hegemony or political competition by way of at least two different sets of intervening variables: the
dis
Page 82
tribution of political resources and skills and the creation of resentments and frustrations.
The Distribution of Political Resources and Skills
In allocating income, wealth, status, knowledge, occupation, organizational position, popularity, and a variety of other values, every society also allocates resources with
which an actor can influence the behavior of other actors in at least some circumstances. These resources then become political resources. Who receives what and how
much political resources is not, however, simply an inert output of socioeconomic institutions. Actors who influence or control the state may use the various powers of
the state to rearrange the initial distribution of political resources that results from the processes of the socioeconomic institutions: by income taxes, for example, or by
imposing limits on campaign contributions; or they may actually create and allocate new political resources, such as the suffrage.
Extreme inequalities in the distribution of such key values as income, wealth, status, knowledge, and military prowess are equivalent to extreme inequalities in political
resources. Obviously a country with extreme inequalities in political resources stands a very high chance of having extreme inequalities in the exercise of power, and
hence a hegemonic regime. So much for the axiomatics.
In agrarian societies, I suggested earlier, such key values as knowledge, wealth, income, status, and power are strongly intercorrelated: the welloff are welloff in all
these respects while the badlyoff—who are the bulk of the population in many agrarian societies—are badlyoff in all respects. Political resources then, are strongly
cumulative: if actor A outranks actor B with respect to one political resource, say wealth or income, then A also outranks B with respect to all other political resources,
knowledge, say, or status. Yet as we have seen there are two major variants of the agrarian
Page 83
society. To put it simply: in the traditional peasant society there is extreme inequality in the distribution of values, hence in political resources, and consequently in the
exercise of power. But in the free farmer society there is a considerably greater degree of equality in the distribution of values, hence in political resources, and so in the
exercise of power. If the relatively greater condition of equality in a free farmer society is also associated with a larger measure of political equality allocated through
suffrage, competitive parties, elections, and responsive leaders, then the accumulation of inequalities is even further inhibited. By accumulating popularity, followers, and
votes, leaders may offset some of the potential effects of differences in wealth and status, and use the regulatory powers of the state to curtail these differences or their
consequences for political life.
Some interesting evidence on this point is furnished in a study by Russett on the relationship between regime and inequality in the distribution of land in 47 countries
around 1960. Inclusive polyarchies are more common among countries with the greatest equality in land distribution. Conversely, countries with the greatest inequality
tend to be nonpolyarchies. (Tables 6.1 and 6.2) Of the 23 countries with greater than median equality in land, 17 are inclusive polyarchies. To look at the data in
another way, of the 24 inclusive polyarchies, 17 are above the median. Of the 24 countries with median equality or less, only 7 are inclusive polyarchies and 15 are
nonpolyarchies. Because polyarchal countries are more developed economically on the average, in these countries less of the labor force is employed in agriculture;
hence equalities and inequalities in land distribution have a smaller impact on political life. For example, among the 7 inclusive polyarchies with median equality or less, 5
have less than a third of their labor force in agriculture. By contrast, among the hegemonies, which have a relatively high proportion of their labor force in agriculture, the
consequences of land inequality for political life are enhanced.
Page 84
Table 6.1. Polyarchies, NearPolyarchies, and Nonpolyarchies,
by Degree of Inequality in Land Distribution,
47 Countries, circa 1960
Gini index Inclusive
of inequalitya polyarchiesb Nearpolyarchiesb Nonpolyarchiesb
43.7 Yugoslavia
45.0 Poland
45.8 Denmark
47.0 Japan
49.7 Canada
49.8 Switzerland
52.2 India
56.4 Philippines
57.7 Sweden
58.3 France
58.7 Belgium
59.8 Ireland
59.9 Finland
60.5 Netherlands
63.8 Luxembourg
65.2 Taiwan
66.9 Norway
67.1 South Vietnam
67.4 West Germany
70.0 Libya
70.5 United States
71.0 United Kingdom
73.7 Panamad
74.0 Austria
74.0 Egypt
74.7 Greece
75.7 Honduras
75.7 Nicaragua
77.3 New Zealand
78.0 Spain
79.2 Cuba
79.5 Dominican Republic
80.3 Italy
81.7 Uruguay
82.8 El Salvador
83.7 Brazil
84.9 Colombia
86.0 Guatemala
86.3 Argentina
86.4 Ecuador
87.5 Peru
88.1 Iraq
Page 85
Table 6.1 (continued)
Gini index Inclusive
of inequalitya polyarchiesb Nearpolyarchiesb Nonpolyarchies
89.1 Costa Rica
90.9 Venezuela
92.9 Australia
93.8 Chilec
93.8 Bolivia
Data Source: Bruce M. Russett, “Inequality and Instability: The Relation of Land Tenure to Politics, World Politics 16 (april 1964): 442–54, table 2.
a The Gini index is a generally accepted measure of inequality. See, e.g., Hayward Alker, Jr., and Bruce M. Russett, “Indices for Comparing Inequality,’’ in Richard L.
Merritt and Stein Rokkan, eds., Comparing Nations: The Use of Quantitative Data in CrossNational Research (New Haven: Yale University Press, 1966), pp. 349–
82.
b The classification here differs from Russett’s, who classifies the countries as stable democracies, unstable democracies, and dictatorships. For the classification used
here, see appendix table A3.

c Because of suffrage limitations based on literacy, in 1960 Chile was much less inclusive than the others, though high on the dimension of public contestation.

d Subsequently a nonpolyarchy.
Among the hegemonies with median equality or less, all except Argentina have more than a third of their labor force in agriculture; many in fact have more than half in
agriculture.
Table 6.2. Polyarchy and Inequality in Land
Greater than Median equality
median equality or less
(N) (N) Total
Inclusive polyarchies 17 7a 24
Nearpolyarchies 1 2 3
Nonpolyarchies 5 15 30
Total N 23 24 57
Data Source: Russett, “Inequality and Instability.”

a Includes Chile.
As an agrarian society becomes industrialized a profound change takes place in the nature of the equalities and inequalities among the citizens or subjects.
Industrialization reallocates rewards and privileges in a drastic way. To be
Page 86
sure, these new allocations are often highly unequal. Yet as I suggested earlier the needs of an advanced industrial society and the aspirations it helps to create and to
satisfy, disperse many political resources that in traditional peasant societies are the monopoly of tiny elites—literacy, education, technical knowledge, organizational
skills, access to leaders, and the like. If industrial societies do not eliminate inequalities they significantly reduce many of them.1 As average income rises with advancing
technology and growing productivity, more and more advantages hitherto arrogated to small elites come within reach of an expanding proportion of the population.
Even inequality in incomes (before taxes) probably decreases.
Some interesting evidence on this point is provided by a study of income and inequality in the United States in 1959–60. Among the 51 political units consisting of the
50 states and the District of Columbia, there is a rather strong negative correlation (–.78) between the median income of each unit and the degree of inequality in the
distribution of income in the unit. As might be expected, the more unequal the distribution of incomes in a state, the more unequal also the distribution of housing and
education.2 (Table 6.3)
In loose language, then, one might say that as a country approaches high levels of industrialization, extreme inequalities in important political resources decline; while this
process does not produce equality it does produce greater parity in the distribution of political resources.
1. See Lenski’s conclusion (Gerhard Lenski, Power and Privilege [New York: McGrawHill, 1966]), p. 437.
2. David I. Verway, “A Ranking of States by Inequality Using Census and Tax Data,” Review of Economics and Statistics 48, no. 3 (August 1966): 314–21. In
Verway’s study (p. 320) the Gini index of inequality in incomes correlates as follows with other standard measures:
Share of top 20% 0.98
Page 87
Moreover, industrial societies, I suggested earlier, change the pattern of inequalities in still another way: although they do not wholly prevent the accumulation of
values—particularly wealth, income, and status—in comparison with a traditional peasant society they drastically reduce the accumulation of political resources and
create instead a system of dispersed inequalities under which actors badly off with respect to one kind of political resource stand a good chance
Table 6.3 Income and Inequality for the United States,
by States and the District of Columbia, 1959–60:
Coefficients of Correlation
Gini index of inequality

in incomes
Median income –.78
Gini index of inequality
Housing
Owneroccupied 0.65
Renteroccupied 0.73
Education 0.76
Use of servants: percentage employed in households 0.77
Source: David I. Verway, “A Ranking of States by Inequality Using Census and Tax Data,” Review of Economics and Statistics 48, no. 3 (August1966): 314–20.
of having access to some other and partly offsetting political resource. If the political regime itself is a polyarchy, then the system of dispersed inequalities is strengthened
even more.
Whether the development of an industrial society increases or decreases equality depends, then, on the type of agrarian society within which industrialization takes
place. Introduced into a traditional peasant society, industrialization is, sooner or later, an equalizing force: it transforms a system of cumulative inequalities into a system
of greater parity with respect to some key resources, and with respect to political resources in general it disperses (but does not eradicate) inequalities. Introduced into
a free farmer society, however,
Page 88
industrialization may actually increase inequalities in political resources, even though these inequalities are dispersed rather than cumulative.3
FIGURE 6.1 Industrial Society and Inequality
The argument of this section is summarized schematically in Figure 6.1.
The Creation of Resentments and Frustrations
If inequalities foster resentments, then resentments over inequality must exist even in industrial societies, for although inequalities may be less extreme and more
dispersed in an industrial society than in a traditional peasant society, the
3. In Who Governs (New Haven: Yale University Press, 1961), I interpreted historical evidence drawn from New Haven (which would be typical in this respect, I
think, of most old cities of the Eastern seaboard) as indicating a shift between the eighteenth and twentieth centuries from cumulative to dispersed inequalities. This is not
inconsistent with the view that in agrarian America, as distinct from these few old urban centers, the shift from the early eighteenth to the twentieth century was from
relative equality to dispersed inequalities. Michael Zuckerman’s arresting interpretation of town life in colonial Massachusetts as highly consensual and relatively
egalitarian would lend support to this view. See his Peaceable Kingdoms: New England Towns in the Eighteenth Century (New York: Knopf, 1970).
Page 89
remaining inequalities are very far from negligible. Do they not then weaken the allegiance to the regime among the less welloff strata? If so, how can we explain the
fact that most polyarchies, presumably the very regimes most endangered by inequality, actually developed in the midst of severe and widespread inequalities? What is
more, many polyarchal regimes exist even now in societies with enormous inequalities of some kinds, for example, with respect to incomes, wealth, or the chances for
higher education. How, in these circumstances, can polyarchy endure?
I hope that every reader will understand that in trying to explain these things, as I am about to do, I do not wish to justify them. That polyarchies have tolerated a good
deal of inequality does not imply that they should. Yet the fact that considerable inequality in the distribution of incomes, wealth, education, and other values can persist
in polyarchies without stimulating enough opposition to bring about a change either in the governmental policies that permit these inequalities or in the regime itself is a
matter that needs explaining.
The explanation is, I think, in two parts:
When demands for greater equality arise, a regime may gain allegiance among the deprived group by responding to some part of the demands, though not necessarily
all of them.
But a great deal of inequality does not generate among the disadvantaged group political demands for greater equality.
Responses by governments
A situation of objective inequality may give rise to a demand that its causes be removed, but it may not. If demands do arise, they may or may not be directed to the
government. The inequality may be reduced or eliminated as a
Page 90
result of actions taken by the government, but unequal conditions may also be reduced even if the government takes no positive action—or in some cases even in spite
of wrongly directed actions of the government. In some cases, even though the government’s actions are misdirected, they may reduce further demands simply because
these actions symbolize to the disadvantaged group that the government is concerned. Indeed, it seems at least theoretically possible that sometimes a government’s
wrongheaded but seemingly wellintentioned policies may completely fail to reduce inequalities, yet the very fact that the government demonstrates its concern may be
enough to hold, perhaps even to win over, the allegiance of the deprived group.
Among the various possibilities, however, two seem particularly relevant to the questions of this chapter. One moderately familiar path leads from an inequality to
government responses that reduce the inequality (or the sense of inequality) and hence strengthen the allegiance of the deprived group to the regime. Here one thinks of
the successful efforts of the Swedish government to reduce unemployment in the 1930s, or, during the same decade, various actions taken by the government of
Franklin Roosevelt to provide more economic security in the United States.
In fact, in a number of European and Englishspeaking countries that now have inclusive and seemingly quite stable polyarchies, liberalized regimes responded in the last
century and this to demands for a reduction in inequalities. Typically, these demands at first emphasized and resulted in the extension of political rights to strata who
were excluded from legal participation in the political system; this process was substantially completed in these countries by about 1920. Then these democratic regimes
became more responsive to demands, which had been made earlier, for an expansion of “social rights” to security, welfare, education, and the like. This process is still
under way, though in a number of countries it has slowed down as a result of ex
Page 91
tensive reforms already carried out. By responding to these demands for greater political and social equality, a number of countries seem to have won the long battle for
the allegiance of hitherto disadvantaged groups, particularly, of course, the working classes. This is illustrated by the path on the left in figure 6.2. Because this process
has been adequately analyzed elsewhere, I do not think it is necessary
FIGURE 6.2 Some Possible Effects of Inequality
on Allegiance to a Regime
to discuss it further here.4 Path A′ represents the case of a government which is perceived among a disadvantaged group as having reduced inequalities, even though its
actions have failed to get at the causes and the inequalities remain more
4. See, among others, T. H. Marshall, Citizenship and Social Class (Cambridge: Cambridge University Press, 1950) and Robert A. Dahl, ed., Political Oppositions in
Western Democracies (New Haven: Yale University Press, 1966), pp. 359–67. Though it is still too early to judge, it is possible that a new phase began in the late
1960s that will emphasize the democratization of a variety of social, economic, and political organizations in which authority has been predominantly hierarchical.
Page 92
or less unchanged. In time, of course, the fact that the government has not succeeded in reducing the inequalities may, and indeed probably will, become apparent. Yet
it is quite possible that by the time this happens, the government may have acquired a considerable reserve of allegiance.
A second possibility, however, is that by its acts of commission or omission the government maintains the inequality and becomes the target for the hostilities of the
disadvantaged group. Perhaps no proposals for government action arise, or at least are not vigorously pressed in public. Or, for whatever reasons, the government fails
to act despite the demands made on it. Or, if it acts, its policies are misdirected. Or it may even be that policies deliberately chosen by the government are perceived as
a principle cause of deprivation (for example, racial discrimination imposed by state governments). These possibilities are represented in figure 6.2 by paths B, C, and
D. If, as a result, the inequality persists, then the allegiance of the disadvantaged group to the particular regime is likely to be weakened.
Because any type of regime is endangered by widespread alienation and disaffection, an extreme state of inequality resented by a large segment of the people of a
country presents a threat to any regime, no matter whether hegemonic or competitive. Yet there is reason for thinking that hegemonic regimes can tolerate “more”
inequality than competitive systems, particularly polyarchies. For hegemonic regimes, especially those with centrally dominated social orders, have at their disposal
much more comprehensive means of coercion which they can employ to suppress the expression of discontent; the frustrations and aggressions of the disadvantaged
can thus be contained and perhaps even turned inward on themselves or deflected into despair, apathy, and hopelessness. Competitive political systems have fewer
resources at their disposal for coercing their people, because the essential conditions for a competitive regime, and particularly for a polyarchy, include a more or less
pluralistic
Page 93
social order and a variety of effective legal and constitutional constraints on governmental coercion.
The difference is, I believe, a real one, and vital. But it should not be exaggerated. Athenian and American democracy each furnish us with an example of a competitive
system that was inclusive with respect to one part of the population but hegemonic with respect to another (slaves, and in the United States also exslaves).
The American case is not only too salient to be ignored, but it helps to clarify the complex relationships between inequality and regime. As everyone knows, a
competitive polyarchy in the United States enforced a state of extreme inequality on its black population both during slavery and after. In the North, for almost a
century a rather inclusive polyarchy was not threatened in any significant way by the inequalities imposed on the Negroes in its midst, probably because until the end of
the Second World War Negroes were a comparatively small minority, impoverished in political resources and lacking in militancy. In the South, on the other hand,
Negroes were a much larger minority, and in two states even a majority.5 To enforce a longstanding condition of extreme deprivation upon the black people of the
South, southern whites (like the free Athenians) had to develop two political systems, one superimposed on the other: a more or less competitive polyarchy in which
most whites were included and a hegemonic system to which Negroes were subject and to which southern whites were overwhelmingly allegiant. To maintain its dual
system, the South developed repressive violence within a quasipluralistic social order: violence and terror were directed at Negroes in
5. In 1860, the percentage of slaves in the total population among the leading slave states was: South Carolina, 57%; Mississippi, 55%; Louisiana, 47%; Alabama,
45%; Georgia, 43%; Virginia, 40%; and North Carolina, 33%. Richard B. Morris, Encyclopedia of American History (New York: Harper, 1953), p. 516; U.S.,
Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1957 (Washington, D.C.: Government Printing Office, 1960), p. 13.
Page 94
general and at the handful of deviant whites who from time to time openly opposed the hegemonic system. The stability of the South’s dual system depended not only
on the effects of past, present, and threatened coercion in creating and reinforcing a mood of resignation and hopelessness among Negroes (and as the ‘‘Uncle Tom”
image suggests, even a sense that white supremacy was legitimate), but also on the strength of an understanding with the main political strata of the North not to
interfere with the southern system. By midtwentieth century both of these conditions began to decay, and as each weakened, its decay reinforced the destruction of the
other. While the older hegemony is by no means totally displaced, for the past decade it has visibly been in profound crisis. By the end of the 1960s the entry of
southern Negroes into the competitive political system was well on its way.6 And the political hegemony based on coercive violence was disintegrating.
Does the case of Negroes in the United States challenge the hypothesis that competitive regimes can sustain less relative deprivation than hegemonic regimes? The
example, I believe, argues quite the contrary: to inflict severe relative deprivation on a large segment of the population is likely to require, as it did in the South, a
hegemonic political system and a social order—whether quasipluralistic or centrally dominated—with repressive violence.
Yet the example does suggest two important qualifications: even an inclusive polyarchy can impose a relatively high state of deprivation on a small minority if that
minority, like Negroes in the North until after World War II, is, for whatever reasons, weak not only in numbers but in all political resources, skills, and demands. And
as Athens and the American South both illustrate, historically it has been possible to develop and even to sustain over a very long period a dual system that is
competitive with respect to the dominant group and hegemonic with respect to a deprived minority.
6. For some data, see footnote 3, chap. 2, above.
Page 95
Responses by the disadvantaged
Between a condition of objective inequality and the response of a disadvantaged person lie the perceptions, evaluations, expectations—in short, the psyche—of the
individual. To the dismay and astonishment of activists who struggle to rouse a disadvantaged group to oppose its lot, the human psyche does not invariably impel those
who are deprived of equality to seek it, or sometimes even to want
FIGURE 6.3. A Hypothetical Path from Objective
Inequality to Demands for Greater Equality
it.7 This phenomenon deserves a fuller—and better—treatment than it can have here, for I must limit myself to a few aspects of it.
Consider, then, a hypothetical path leading from objective inequality to demands for greater equality among the disadvantaged group. (Figure 6.3) To complete the
circuit, “Yes” answers are required to all the questions. Yet several of the linkages are prone to failure, to “No” responses.
To be sure, it is utterly fatuous to think that the connec
7. Robert E. Lane, “The Fear of Equality,” American Political Science Review 53 (March 1959): 35–51.
Page 96
tion between objective reality and subjective awareness breaks down so completely that the disadvantaged are unaware of their lot. If the myth of the happy poor (and
the unhappy rich) still needs puncturing, Hadley Cantril provides some solid evidence that the gap between condition and aspirations decreases all over the world as
one’s objective situation improves. In a 14nation survey respondents were asked to locate themselves on a hypothetical tenstep ladder between “the best possible life
for you … and the worst possible life for you.” They were also asked to locate their country on a comparable ladder. Cantril found that the higher the socioeconomic
level of a country, the higher on the average its people locate their present and past positions on the “ladder’’ of satisfaction, as well as their country’s present and past
positions. He found further that people who are objectively better off with respect to education, income, or occupational status expressed more satisfaction with their
condition, on the average, than did those who had low education, low incomes, or lower status occupations.8
Nonetheless, even if they are relatively more disatisfied with their condition, the members of a disadvantaged group may not complete all the connections along our
hypothetical path. For example, at A, perceptions of inequality are sometimes clouded by changes in the conditions of one’s “own” group. Thus Runciman concludes
that in Britain “after the Second War, relative deprivation was damped down by the achievement of some gradual amelioration of the class situation of manual workers
and the conviction that a greater distribution was taking place than in fact did occur.”9
8. Hadley Cantril, The Pattern of Human Concerns (New Brunswick: Rutgers University Press, 1965), pp. 194, 258–59. The countries, by rank on Cantril’s
socioeconomic index, were the United States, West Germany, Israel, Japan, Poland, Cuba, Panama, Yugoslavia, Philippines, Dominican Republic, Brazil, Egypt,
Nigeria, and India.
9. W. G. Runciman, Relative Deprivation and Social Justice (Berkeley: University of California Press, 1966), p. 94.
Page 97
The connection at B often breaks down because people who are objectively disadvantaged generally do not compare themselves with the most advantaged groups,
whose good fortune they simply do not consider to be relevant to their own condition, but substitute other comparisons instead. For one thing, a person in a
disadvantaged group almost certainly will compare the present situation of his group with its own past (or its myths about the past); if the group believes that it is now
better off than it was in the past, that belief may be far more salient and relevant than the fact that certain other groups are currently very much better off. Then, too,
many people identify themselves some of the time with large collectivities like the nation or country; hence for the particular individual or group, past and present
inequalities may be attenuated by the thought that the collectivity as a whole—the country, for example—is moving toward a richer or more just condition. Finally, when
a person does compare himself or his “own” specific group with other individuals or groups, he is likely to make the comparison with others who are, socially speaking,
not distant but quite close or adjacent to him. To the skilled worker, for example, the income and privileges of the company president are likely to be less relevant than
those of semiskilled workers just “below” him or his foreman just “above” him, or skilled workers in other plants.
Because of these various comparisons, in a society with extensive inequalities an individual may have a relatively favorable judgment of his own position even though,
objectively speaking, he is seriously disadvantaged in comparison with the elites.
Some evidence directly bearing on the last possibility is provided by Runciman from his survey of some 1400 respondents in Britain. Runciman distinguished between a
“comparative reference group,” whose situation or attributes a person contrasts with his own, and a ‘‘membership reference group,” which is “the startingline for the
inequality
Page 98
with the comparative reference group by which a feeling of relative deprivation is engendered.”10 He found that the sense of “relative deprivation” engendered by
inequalities in Britain was greatly attenuated, among other things by the tendency to compare oneself with groups socially rather close to one’s own. Moreover, as I
indicated above, he ascribed the rather low level of sense of relative deprivation among manual workers to recent improvements in their situation and perhaps an
exaggerated notion of just how greatly the old inequalities had actually been reduced. Conversely, middleclass people were likely to harbor resentments not about the
very wealthy but about the rapid gains among the working classes. Runciman also found that:
There is a wide discrepancy between objective inequality and a subjective sense of relative deprivation.
Irrespective of whether or not they feel their position is worsening, middleclass people are more likely to feel the prosperity of others a grievance.
Those at the top of the manual stratum are least likely of all to think of any other sorts of people as doing better than they.
Manual workers and their wives who have reached the top level of incomes for manual occupations are likely not to feel relatively deprived because in terms of
comparisons natural to them they have done well.
A fair proportion of the respondents give a figure as a desired income that is little or no higher than their actual income, and an even higher proportion say that they are
satisfied with what they are getting.
Within each onethird of the overall income distribution, manual (“working class”) respondents are considerably more likely than nonmanual (‘‘middleclass”)
10. Ibid., p. 12. Italics mine.
Page 99
respondents to give a figure as a “proper” income that is more or less equivalent to what they say they can earn at present.
Not only are comparative reference groups not chosen in accordance with the facts of inequality but such a correspondence with the facts is least likely of all among
those who are in fact most unequally placed.11
Cantril’s crossnational survey also provides some interesting, although more tangential, evidence on the connection at B. His results support the view that people often
judge their own situation by what they perceive to be the direction of change for a larger collectivity with which they identify, in this case the country. He found that:
How highly people rated their own present condition—where they stood on the “ladder”—was significantly related to how highly they rated the condition of their own
country, i.e., where their country stood on the ‘‘ladder.”
The extent to which people felt that their condition had improved over the previous five years was significantly related to the extent they felt their country had improved.
If people expected that their personal standing would rise over the next five years, they were likely to expect that their country would also improve.12
These correlations do not, of course, tell us anything about the direction of cause, which doubtless flows both ways. But some comparisons of how people in specific
countries rated themselves are at least suggestive. For example, in rating their present condition the two highest groups were a sample of members in 10 kibbutzim in
Israel—where ob
11. Ibid., pp. 197–210.
12. Cantril, The Pattern of Human Concerns, pp. 184–94.
Page 100
jective equality is exceptionally high—and a sample of adults in the United States, where the average income is exceptionally high but inequalities in income distribution
are also great. The country with the third highest ratings by its people of their personal situation was Cuba, which had undergone a revolution that greatly reduced (but
by no means eliminated) previous inequalities (probably at some cost to average incomes) and Egypt, which had also undergone a revolution but still suffered from
extraordinary poverty and inequalities. The highest personal expectations for the future were held (in decreasing order) by the people surveyed in Cuba, Egypt, the
kibbutzim, the United States, Nigeria, Brazil, and Panama.13 The country that led all the rest in the size of the change in personal ratings from past to future (that is, in
expectations of improvement) was Nigeria, one of the poorest countries in the world and one destined, at the time of the survey, for a catastrophic civil war. Nigeria
was followed by Cuba, the Dominican Republic, Egypt, Brazil, and Panama.14 Thus hopes for the future seemed to be so much influenced by the immediacy of
revolutions, national independence, or other dramatic political changes that there was virtually no relationship between the socioeconomic level of a country and the
expectations its citizens expressed either for their own future or that of their country.15
The effectiveness of the connection at C depends on the prevailing ideas in the culture or subculture into which an individual is socialized. A deprived group may well
believe that its present inferior condition is an inherent part of the order of things, justified by religion and cosmology (as Hinduism gave legitimacy to caste), subject
only to change through some ultimate and perhaps apocalyptic redemption. A worldview justifying and “rationalizing” inequality does
13. Ibid., table IX:4, p. 187.
14. Ibid., table IX:6, p. 188.
15. See the rank order correlations in ibid., p. 194.
Page 101
not persist only because it is to the advantage of the elites who benefit from the status quo. Among the disadvantaged groups themselves, such a selfdenying world
view may help to make a miserable and often humiliating existence more bearable and understandable. A group confronted over a long period of time by seemingly
ineradicable inequalities may learn to keep its demands low and thus bring them more into line with the harsh limits of the possible.16
The crossnational survey by Cantril furnishes some interesting illustrative evidence on this process of “attenuation.” By means of a rather openended but carefully
designed set of questions, people were asked to talk about their hopes and fears. One might suppose that people in the more highly developed countries would express
a larger number of hopes than people in pooreer countries, but fewer fears. Cantril did in fact find a ‘‘positive but modest” relationship between a country’s level of
socioeconomic development and the volume of hopes expressed by its people, both for the self and for the nation. But contrary to what one might expect, he also
found an even stronger positive corre
16. Thus over against “the comples of ideas of ascendance which always stress the rights of the individual” in peasant societies, Mahmet Bequiraj describes the ideas of
“attenuation” that
support the existing design of the system … It is reasoned that persisting socila problems are not so much caused by existing forms of social organization as by the
faults of man; man is the reason why the presumably faultless design of social organizaiton fails … The ideas of ascendance … endanger the balance by suggesting
new wants. In contrast the ideas of attenuation supply the individual with the strength to delay gratifications and thus permit the minimization of needs … Given the
low level of knowledge, the ideas of attenuation are functionally more significant than the ideas of ascendance… . They become “ends” in themselves. As virtues,
they aim to solidify the existing froms of social organization … Detached from the influence of the ideas of ascendance and hardened into virtues, the ideas of
attenuation block further development or advance in knowledge for the system.
Peasantry in Revolution (Ithaca: Center for International Studies, Cornell University, 1966), pp. 31–33.
Page 102
lation between the socioeconomic level of the ocuntry and the volume of fears expressed by its people: in short, the better off the country, the more hopes and fears
both. The two countries in which people expressed the fewest fears for both self and naiton were India and Brazil. Improbable as it may seem, the country with the
lowest volume of references to health—whether hopes or fears—was India, “where health conditions are worse than in any of the other countries studied.”17 There,
the process of “attenuation,” supported by Hinduism, has been carried to exhausting lengths.
The connection at the end of the path (at D and E) may also break down. Frustration, resentments, and anger may not stimulate demands for greater equality but
instead may turn into resignation, apathy, despair, hopelessness, selfdenigration, fantasy, millenial dreams, pious acceptance, fatalism, the mentality of the “Uncle Tom,”
etc. Terms like these are used, in fact, to describe the psyche of the peasant in the acutely deprivational peasant society.18
17. Cantril, The Pattern of Human Concerns, p. 164. The rank order correlations (p. 199) between volume of “concerns” (hopes and fears) and the socioeconomic
index for the country was:
Personal hopes 0.24
National hopes 0.25
Personal fears 0.46
National fears 0.51
In reporting the volume of concerns, Japan and Poland are omitted. For total concerns, see p. 155.
18. See, for example, Bequiraj’s remarkable portrait of the peasant, Peasantry in Revolution, especially pp. 143. For example:
Both the pessimistic philosophy and the trend toward piety are universal among peasant groups … Even in his own eyes the peasant is the “dolt,’’ the “brute,” and
sometimes the “villain.” He has learned to live with this selfabasement with the help of a pessimistic philosophy based on an assumption of his group’s inadequacies,
which he uses to explain personl failures. This explanation … enables him to adjust to a social world which he correctly recognizes as being unjust, intemperate, and
cruel. [pp. 11–12] … Fastening upon the utopian ethical precepts of the mors, the peasant sustains himself against the effect of the constant failures which make up
his performances. His faith
Page 103
The argument of this chapter can be summarized in the following propositions:
In a country with a jegemonic regime, extreme inequalities in the distribution of key values reduce the chances that a stable system of public contestation will develop.
In a society that already has a regime with public contestation, extreme inequalities increase the chances that competitive politics will be displaced by a hegemony.
Polyarchies are particularly vulnerable to the effects of extreme inequalities.
Extreme inequalities in the distribution of key values are unfavorable to competitive politics and to polyarchy because this state of affairs:
is equivalent to extreme inequality in the distribution of key political resources and is likely to generate resentments and frustrations which weaken allegiance to the
regime.
However, systems of competitive politics and even polyarchies manage to survive a considerable measure of inequality because:
in the eventual advent of the “good order” induces in him reverence for the good and sacred, directing him in this way to peity. The projection of his place in this
“good order” sustains his belief in his own worth, counterbalancing the debasement he experiences daily. [p. 19]
Of course it would be wrong to assume that peasants always and everywhere adhere to these views, and a Bequiraj points out, they may also stimulate insurrection and
revolution of his faith in the good order is destroyed. Peasant insurrections have been, in fact, faily common, as can be seen from the sizable number of peasant
insurrections in Sorokin’s list of 173 disturbances in France from the year 531 to 1908. Pitirim A. Sorokin, Social and Cultural Dynamics (New York: American Book
Co., 1937), vol. 3, appendix. See also George Rude, The Crown in History, 1730–1848 (New York: Wiley, 1964), chap. 1, “The French Rural Riot of the Eighteenth
Century
Page 104
A great deal of inequality does not generate among the disadvantaged group political demands for greater equality or a change in the regime.
When demands for greater equality do arise, a regime may gain allegiance among the deprived groups by responding to some part of the demands, though not
necessarily all of them, or by responses that do not reduce the objective inequalities but do reduce feelings of relative deprivation.
Page 105
7. SUBCULTURES, CLEAVAGE PATTERNS,
AND GOVERNMENTAL EFFECTIVENESS
Obviously any system is in peril if it becomes polarized into several h ighly antagonistic groups. Conronted by severe polarization, competitive regimes are prone to
collapse, to a coup d’etat, to civil war: for example, Italy from 1919 to 1923, the first Austrian Republic virtually throughout its brief existence, the Wiemar Republic
rom aobut 1929 to its demise, the Spanish Republic from 1934 to 1936, and the United States in the decade preceding our Civil War.
In the United States the victors were committed to a more or less inclusive polyarchy, a commitment that for a decade after the end of that war inlcuded even the freed
slaves. But in other cases, victory went to antidemocratic movements that successfully introduced a hegemonic regime into the country.
There are conflicts, then, that a competitive political system does not manage easily and perhaps cannot handle at all. Any dispute in which a large section of the
population of a country feels that its way of life or its highest values are severely menaced by another segment of the population creates a crisis in a competitive system.
Whatever the eventual outcome may be, the historical record argues that the system is very likely to dissolve into civil war or to be displaced by a hegemony or both.
Page 106
Thus any difference within a society that is likely to polarize people into severely antagonistic camps is a cleavage of exceptional importance. Are some countries, then,
less lkely than others to have competitive regimes and more likely to have hegemonic regimes because, for whatever reason, they are unusually subject to cleavages that
are particularly favorable to acute polarization?
Answers to this question are, unfortunately, obscured by the massive impact on social thought of certain dramatic aspects of Marxist thought. For ever a century,
reflections about polarization and civil war have been dominated, even among nonMarxists, by Marx’s conception of polarization around the node of economic
classes—working class and bourgeoisie. Yet in the 120 years since the Communist Manifesto was published, no country has developed according to the Marxist
model of conflict, nor has any regime, whether hegemonic or competitive, fallen or been transformed because of a clearcut polarization of working class and
bourgeoisie.
A preoccupation with class conflict and often an unarticulated assumption even among sophisticated social theorists that classes are somehow the “real” basis of
differences in an industrial society, to which all others are “ultimately” reducible, has tended to deflect attention from other differences that give rise to durable
subcultures into which individuals are socialized: these are differences in religion, language, race, or ethnic group, and region.1
1. Subcultures may of course be formed around economic or social ‘‘classes” or “statuses” (as these are variously defined) or around occupations, educational levels,
or other characteristics that need not be strictly related to ethnic, religious, or regional characteristics. As used here, however, the term subcultural pluralism refers to the
presence of ethnic, religious, or regional subcultures. One can think of subcultural pluralism as a hypothetical dimension for which a measure can be constructed in order
to compare the relative “amount” of subcultural pluralism in differenct societies.
One might protest that it is redundant to include the term “ethnic group” in the list of characteristics on which an ehtnic subculture can
Page 107
Differences along these axes, differences that frequently reinforce one another, have obvious, important, and persistent consequences for political life in a great many
countries—indeed, it is no exaggeration to say, in most countries in the world. Yet differences of this kind have often been ignroed or deprecated as “realy” nothing but
class differences in disguise, or somehow less “real” than class, or if not less real then surely less enduring, bound to disappear rapidly under the impact of
industrialization, urbanization, and mass communications. Yet these differences and the conflicts they engender do not always disappear and may even sharpen with the
passage of time, as in contemporary Belgium, Canada, and quite possibly in Britain.
This is not to argue that ‘‘class” differences are unimportant. It is to say that economic class is only one factor, often less important than others that can and quite
evidently do yield distinct subcultures—ways of life, outlooks, norms, identificaitons, loyalties, organizations, social structures. What is more, these subcultures are often
extraordinarily longlived not only in the life of an individual (who may change his class identification more readily than his mother tongue or his religion) but also in the
life of a society; for over a thousand years, classes and empires have risen and
be based, since it is hard to imagine an “ethic” identification not based, at least remotely, on one of the other differences, i.e. religion, language, race, or physical stock.
There are problems of precise definition here that I am deliberately trying to skirt. However, the terms ethnic subculture and subcultural pluralism are intended to be
broad enough to extend to groups for whom objective differences in language, religion, race, or physical stock have been attenuated by time but whose identification
with the “historical” group nonetheless remains rather strong. Examples might be nonreligious Jews in the United States or certain other groups with distinct regional
roots whose attachments are at present not merely to a region but to a historically developed identity, such as Englishspeaking Scots, Welsh, or Protestant Irishmen in
Great Britain or Scandinavians during periods when one of the Scandinavian peoples was subject to the rule of another without yielding its own separate identity as a
people.
Page 108
fallen while the linguistic boundaries within what are now Belgium and Switzerland have barely changed.
Presumably because an ethnic or religious identity is incorporated so early and so deeply into one’s personality, conflicts among ethnic or religious subcultures are
specially fraught with danger, particularly if they are also tied to religion. Because conflicts among ethnic and religious subcultures are so easily sen as threats to one’s
most fundamental self, opponents are readily transformed into a malign and inhuman “they,” whose menace stimulats and justifies the violence and savagery that have
been the common response of ingroup to outgroup among all mankind. The junctions of ethnic group or religion with regional subcultures creates an incipient nation
whose spokesmen clamor for autonomy, and even for independence. Consequently many students of politics have agreed with John Stuart Mill that the boundaries of
any country with a representative government must coincide with the boundaries of nationality; a considerable body of experience with multinational states lends
impressive support to their argument.2
That subcultural pluralism often places a dangerous strain on the tolerance and mutual security required for a system of public contestation seems hardly open to doubt.
Polyarchy in particular is more frequently found in relatively homogeneous countries than in countries with a great amount of subcultural pluralism. Perhaps the best
evidence for this is to be found in a study by Marie R. Haug that classifies 114 countries according to an indes of pluralism3 and data
2. Cf. Walker Connor, “SelfDetermination, The New Phase,” World Politics 20 (October 1967): 30–53. Connor summarizes the dialectical development of the
argument in Britain over the viability of representative institutions in multinational countries. Against Mill stands Acton, who held that the pluralism of multinational states
provides an antidote to despotism and is a civilizing influence besides; Ernest Barker in turn supprted Mill agains Acton; finally, Alfred Cobban aligned himself with
Acton against Mill and Barker. Connor himself strongly supports Mill and Barker against Acton and Cobban.
3. “Social and Cultural Pluralism as a Concept in Social System Analysis,” American Journal of Sociology 73, no. 3 (November
Page 109
derived from A CrossPolity Survey by Banks and Textor. Although there have been some changes of regime since the early 1960s, when the data were collected,
these can hardly have been so great as to invalidate the general relationships. Comparing the 26 countries where cultural pluralism (in the numerical sense) is negligible
with the 34 where it is extreme reveals that:
Among countries in which subcultural pluralism is negligible, more than half are also classified by Banks and Textor as integrated and homogeneous polities with little or
no extreme opposition, communalism, fractionalism, or political nonassimilation (high in political enculturation).
Among these countries, none is classified as a relatively nonintegrated or restrictive polity with a majority or nearmajority in extreme opposition, communalized,
fractionalized, disfranchised, or politically nonassimilated (low in political enculturation).
Conversely, among countries in which subcultural pluralism is extreme, only 10 percent are high in political enculturation; almost twothirds are low in political
enculturation.
Among coutnries in which subcultural pluralism is negligible, 60 per cent are also classified by Banks and Textor as having an effective allocation of power to
functionally autonomous legislative, executive, and judicial organs (significant horizontal power distribution).
1967): 294–304. The construciton of the index is shown in table 2, p. 297. The variables in the index are: language, race, religion, sectionalism, and “interest articulation
by nonassociational groups,’’ which include kinship, lineage, ethnic, regional, and religious groups. This purely numerical index unfortunately canot register the depth of
sharpness or subcultural cleavates, as in the Netherlands. See the discussion below, p. 213.
Page 110
as having complete dominance of government by one branch or by an extragovernmental agency (negligible horizontal power distribution).
Conversely, among coutnries in which subcultural pluralism is extreme, only onethird have a significant horizontal power distribution, while in 57 per cent one branch of
government is without genuine functional autonomy or two branches have only limited functional autonomy.4
Our own classification of political systems (also based largely on Banks and Textor) yields about 34 polyarchies and nearpolyarchies as of the early 1960s.5 These
make up a relatively large portion of countries with low subcultural pluralism (measured by Haug’s index of subcultural pluralism) and a relatively small proportion of
countries with marked or extreme subcultural pluralism. (Tables 7.1)
4. The definitions are from Arthur S. Banks and Robert B. Textor, A CrossPolity Survey (Cambridge: M.I.T. Press, 1963), pp. 88 and 106. The number of countries
on which data were available varied. The percentages, number of countries, and level of significance for the chi square values are:
PLURALISM: CHI SQUARE
NEGLIBIBLE EXTREME
Political enculturation:
High: integrated polity 52.6% 0.0% 31.92 (d.f.6)
(19) (31)
Low: nonintegrated polity 10.5% 63.3% p < .001
(19) (31)
Horizontal power distribution
Significant 60.0% 17.9% 16.04 (d.f.6)
(25) (28)
Limited 32.0% 57.1% p < .02
(25) (28)
Source: Haug, “Social and Cultural Pluralism,” tables 7, p. 303
5. See the appendix for the way the classification of polyarchies and nearpolyarchies was derived. The classification in table 7.1 is as of the early 1960s. Appendix
table A3 is a classification as of 1969.
Page 111
It is revealing to consider the polyarchies and nearpolyarchies (as of the early 1960s) that are classified in table 7.1 as having marked or extreme pluralism. Of the
three countries with marked pluralism, in Belgium subcultural conflicts have intensified since the early 1960s and in the Philippines there has been a revival of guerrilla
activities in the countryside (the third is Lebanon). Among the six polyarchies with extreme pluralism, Sierra Leone
Table 7.1. Polyarchy and Subcultural Pluralism
Amount of subcultural pluralism
Low Moderate Marked Extreme
Total N 26 28 27 33
Polyarchies and nearpolyarchies 15 10 3 6
Percent 58% 36% 11% 18%
Note: Low defined as 0 or 1 on Haug’s index of pluralism.
Moderate defined as 2 or 3.
Marked defined as 4 or 5.
Extreme defined as 6,7, or 8.
has ceased to be even a nearpolyarchy; Malaya has divided into two countries; India has had profound conflicts over language; in Ceylon, communal conflicts and
restraints on oppositions made it doubtful for a time whether Ceylon even met the standards of nearpolyarchy; and Canada has witnessed the eruption of French
Canadian nationalism. Of the six, only Switzerland has been relatively serene; even there demands for a separate canton have recently risen among the Frenchspeaking
citizens of the Jura.
Nonetheless, if a competitive political system is less likely in countries with a considerable measure of subcultural pluralism, it would be going too far to say that it is
impossible or that subcultural pluralism necessarily rules out an inclusive polyarchy.
One fact to keep in mind is that subcultural pluralism tends to be greatest, at present, among the less developed countries. Countries with extreme subcultural pluralism
are predomi
Page 112
nantly new nations: 70 per cent have achieved independence since 1945. Thus they often bear all the typical disadvantages of the new nations: low per capita GNP,
high employment in agriculture, low urbanization, high illiteracy, low newspaper circulation. They are, besides, relatively large in area. (Tables 7.2 and 7.3)
Table 7.2. Subcultural Pluralism and Socioeconomic Development: I
Pluralism
Negligible Extreme Chi square
Date of Independence
1913 or before 72.0% 24.2% 27.33
(d.f.6)
After 1945 8.0% 69.7% p < .001
N (25) (33)
Per capita GNP
$600 and above 50.0% 6.1% 42.08
(d.f.9)
Under $150 3.8% 75.8% p < .001
N (26) (33)
Pop. in agriculture
Over 66% 3.8% 72.7%
33% or below 50.0% 9.1%
N (26) (33)
Source: Marie R. Haug, “Social and Cultural Pluralism as a Concept in Social System Analysis,” American Journal of Sociology 73, no. 3 (November 1967), tables 5,
6, 7, pp. 301–03.
Note: Percentages are calculated vertically.
As we have seen, low levels of socioeconomic development are themselves inimical to competitive politics. In fact, in the early stages of nationbuilding, typically a
variety of factors interact to undermine the chances for a competitive regime and to produce hegemony instead; subcultural pluralism is only one of these. Hence some
of the association between subcultural pluralism and hegemony can reasonably be attributed to other factors, such as the socioeconomic level.
Page 113
Yet competitive politics can exist even in countries with a very considerable degree of subcultural pluralism. Indeed, Belgium, Canada, and India, among others, have
managed to develop and sustain polyarchies. There is, in addition, the compelling example of the Netherlands. There, differences in language, race, physical stock,
ethnic identification, or
Table 7.3. Subcultural Pluralism and Socioeconomic Development: II
Pluralism
Negligible Extreme Total N Chi Square
Urbanization
20% or more in cities over 20,000 41.1% 12.5% 56 40.92
Less than 20% in d.f.3
cities over 20,000 0.0% 51.0% 49 p < .001
Literacy
90% or above 52.0% 8.0% 25 37.68
Under 10% 0.0% 57.7% 26 d.f.9
p < .001
Newspaper circulation
300 or more per 1,000 pop. 57.1% 7.1% 14 49.57
d.f.9
Under 10 per 1,000 pop. 0.0% 57.1% 35 p < .001
Area
300,000 sq. mi. or more 9.7% 48.4% 31 15.73
Less than 75,000 sq. mi. 29.8% 12.8% 47 d.f.6
p < .02
Source: Haug, “Social and Cultural Pluralism,” tables 4 and 5, pp. 300–01.
Note: Percentages are calculated horizontally.
region play no significant part; hence, numerically speaking, subcultural pluralism is not extreme. Yet religion is a plane of cleavage that has divided the country into
three great spiritual families—(Catholic, Orthodox, Protestant, and the remainder). These three subcultures have been voluntarily segregated to a degree unknown in
any other inclusive polyarchy; yet the Netherlands has managed to sustain, and
Page 114
sturdily, a representative democracy.6 Finally, among the countries with a high degree of subcultural pluralism in the numerical sense is the impressive case of
Switzerland.
At least three conditions seem to be essential if a country with considerable subcultural pluralism is to maintain its conflicts at a low enough level to sustain a polyarchy.7
6. Hans Daalder, “The Netherlands: Opposition in a Segmented Society,” in Robert A. Dahl, ed., Political Oppositions in Western Democracies (New Haven: Yale
University Press, 1966), and Arend Lijphart, The Politics of Accommodation (Berkeley: University of California Press, 1968). There is, however, a good deal of
evidence that subcultural segregation in the Netherlands is rapidly declining. Daalder mentions as examples: “Catholic University of Nijmegen will appoint a Calvinist as
the new professor of political science; a merger of the separate unions is almost imminent; all groups are deemphasizing labels at breakneck speed.’’ (Personal
communication, 1969)
7. My analysis here has been strongly influenced by the experience of the Netherlands, described by Daalder and Lijphart in the works just cited, and Belgium,
described by Val R. Lorwin, “Religion, Class and Language in National Politics,” in Political Oppositions in Western Democracies. In particular, see the sections in
Daalder, pp. 216–20, Lijphart, pp. 197 ff. and Lorwin, pp. 174–85. I have also profited from lengthy discussions with all three scholars and an opportunity to read
works in process by Daalder and Lorwin, including Lorwin’s “Segmented Pluralism: Ideological Cleavage and Political Cohesion,” a paper presented to the Torino
Round Table of the International Political Science Association, September 1969. An invaluable source is also Lijphart’s “Typologies of Democratic Systems,”
Comparative Political Studies 1 (April 1968), pp. 3–44, where he sets out four “prerequisites for consociational democracy” having to do with elite attitudes and
behavior (pp. 22–23):
1. Ability to recognize the dangers inherent in a fragmented system
2. Commitment to system maintenance
3. Ability to transcend subcultural cleavage at the elite level
4. Ability to forge appropriate solutions for the demands of the subcultures
In addition, he sets out six “conditions conducive to consociational democracy” (pp. 25–30):
1. Distinct lines of cleavage between subcultures
2. A multiple balance of power among the subcultures
3. Popular attitudes favorable to government by grand coalition
Page 115
First, conflict is more likely to be maintained at moderate levels if no ethnic, religious, or regional subculture is “indefinitely” denied the opportunity to participate in the
government, that is, in the majority coalition whose leaders form the “government” or the administration.
This requires in turn that among a sufficient number of members of each subculture, and particularly among the leaders, there exist desires for cooperation that cut
across the subcultures at least some of the time. There are several common sources of incentives toward cooperation. One is a commitment to the preservation of the
nation, its unity, its independence, and its political institutions. A second is the recognition on the part of each subculture that it cannot form a majority capable of
governing except by entering into a coalition with representatives of other subcultures. This condition is satisfied if every subculture constitutes a minority.
The prospects appear to be worse for a country divided essentially into two subcultures than for a country divided into more than two. For if a country is split into two
subcultures, then one will be a majority and the other a minority. Indeed, even if a country is divided into more than two subcultures, one might still comprise a majority.
Strictly speaking, then, in order for each subculture to be a minority,
4. External threats
5. Moderate nationalism
6. A relatively low total load on the system
Jurg Steiner, who has used Swiss experience to test sixteen hypotheses about conditions favorable to peaceful resolution of conflicts in ‘‘subculturally segmented
democratic systems” finds that the experience of Switzerland casts doubt on the first of Lijphart’s six conditions just named, though not on the others. Bedingungen fÜr
Gewaltlose Konflictregulierungsmuster in Subkulturell Segmentierten Demokratischen Systemen: Hypothesen Entwickelt am Beispiel der Schweiz (Thun, Switzerland:
mimeo, 1968), pp. 434–37. For his comments on Lijphart’s other conditions, see pp. 432–34, 439–41, and 446–50.
Page 116
it is a necessary but not a sufficient condition that there be more than two subcultures.
If a country is divided into majority and minority subcultures, then members of the majority have less need to be conciliatory toward the minority, since they can form a
majority coalition among themselves. As a consequence, members of the minority may see no prospect of ever freeing themselves from the political domination of the
majority; hence they, too, have little incentive to be conciliatory. Doubtless this is one of the factors that now contributes to the explosiveness of racial relations in the
United States. If a minority subculture is also regionally concentrated, like Frenchspeaking Canadians, then the dangers are even greater. For as militance grows
among members of the minority, they will almost certainly begin to demand greater autonomy—and in the extreme case outright independence. In Belgium, where the
two subcultures are regionally based, tendencies toward national separation have been contained not only by Belgian nationalism but by several other factors as well.
For one, the Walloon minority has, with justification, never felt itself to be an oppressed minority, since it has been dominant socially, culturally, economically, and for
many decades even politically. Recently, as the Flamands have begun to assert themselves and the Walloon minority has begun to feel more threatened, outright
polarization is somewhat reduced by the existence of Brussels as a kind of third element. With 16 percent of the population of the country, the capital city contains both
subcultures—a majority of Walloons and a Flemishspeaking minority. Although Brussels is itself a key issue in the relations of the two subcultures, it also serves as a
force for conciliation.8
8. Lorwin stresses both aspects of Brussels, as issue and conciliator. As issue: “Brussels and its environs supply the leading issue of FlemishWalloon contention. … A
pole for Flemish immigration and the center of the nation’s Frenchspeaking dominance, Brussels became for Flamingants (militant proFlemish elements) both an
obstacle to overcome and a terra irredenta to recover.” As conciliator:
Page 117
To produce more than two subcultures, and hence the possibility that each is a minority, does not require more than one major line of cleavage. To this extent, the
number of different types of cleavage forms an inadequate measure of the extent of subcultural pluralism in a society. In the Netherlands, as we have just seen, religion
forms a single plane of cleavage that has created and sustained the three zuilen—Catholics, Orthodox Protestants, and the remainder —none of which is a majority.9
More than two subcultures can also be created if two (or of course more) planes of cleavage intersect to partition a population into, for example, four groups. Thus in
Switzerland the Germanspeaking majority and the Frenchspeaking minority are both divided into Catholics and Protestants, so that none of the resulting four groups is
a majority; these four are subdivided further by region.10
In India, language, caste, and region generate a fantastic panoply of subcultures, each of which is a relatively small minority. (In a nation of 500 million a relatively small
minority is often larger in absolute terms than most of the countries of the world.) As Kothari argues, India’s extreme pluralism is not merely a source of difficulty but is
also in some ways one of the strengths of the Indian polyarchy, for it now compels the leaders of every group to learn and
“Brussels has a ‘national vocation.’ Its very existence as a metropolis depends upon the performance of national political, administrative, commercial, financial, and
cultural services. Out of interest and sentiment, Brussels wants (even if it fails to show the necessary linguistic tolerance) to maintain the unitary state. Conversely,
advocates of regional separation À deux find Brussels a stumbling block.” Lorwin, “Religion, Class, and Language in National Politics,” pp. 172, 177.
9. In 1960, Roman Catholics constituted 40.4% of the population, the two major Protestant groups 37.6%, ‘‘No Church” 18.4%, and other Protestant sects, Jews,
and “others,” 3.6%. Daalder, “The Netherlands: Opposition in a Segmented Society,” appendix table 6.5, p.425.
10. Steiner. Bedingungen fÜr GewaltlÖse Konflictregulierungsmuster, pp. 6–7, 432–33.
Page 118
practice the arts of conciliation and coalition formation and prevents any single unified group from even approaching a monopoly of political resources.11
The condition that no subculture be indefinitely denied the opportunity to participate in the government can be met in two ways—by a system oriented toward
unanimity, in which, as in Switzerland, every major party is represented in the government of the Gemeinde, the canton, and the federation, or as in Belgium and the
Netherlands by a system of shifting coalitions that, over time, allows each group to shift out of opposition into the government—a system for which Lorwin has coined
the “short and catchy name” AllgemeinkoalitionsfÄhigkeit.12
The second requirement for reducing conflict in a country with considerable subcultural pluralism is a set of understandings or engagements, not always codified into
formal constitutional provisions, that provide a relatively high degree of security to the various subcultures.
Among the most prevalent forms of mutual security arrangements are guarantees that the major subcultures will be represented in parliament in some rough
approximation to their numerical weight, a guarantee that is frequently secured by using one of the various types of proportional representation for electing candidates.
This kind of guarantee may extend even to the executive, as in the allparty Federal Council in Switzerland, or under the Proporz arrangement introduced in Austria
under the Second Republic. Where participation in the executive is assured, the arrangement typically requires unanimity or (to put it differently) permits
11. Rajni Kothari, “India,” in R. A. Dahl, ed., Regimes and Opposition (New Haven: Yale University Press, forthcoming 1971).
12. Lorwin, “ Religion, Class and Language in National Politics,” p. 178.
Page 119
each minority to exercise a veto over decisions affecting major subcultural concerns. Where the subcultures are more or less regional, mutual security may also be
provided by federalism, as in Canada, India, and Switzerland; in Switzerland, considerable local autonomy provides still further protection to the subcultures.13 Finally,
mutual guarantees may be provided by specific constitutional provisions, pacts, or understandings that impose limits on the constitutional authority of any parliamentary
coalition to regulate certain matters important to one or more subcultures, for example, language guarantees in Switzerland, India, and Canada or the guarantees and
understandings in the Netherlands that concede a large measure of autonomy to the three zuilen with respect not merely to religion, newspapers, political parties, trade
unions, and farm organizations, but also statesubsidized schools, social security programs, and the stateowned radio broadcasting stations and television network.14
The third requirement is not only more conjectural but also more difficult to state precisely. The chances for a polyarchy are higher if the people of a country believe that
a polyarchy is effective in responding to demands for coping with the major problems of the country, as these problems are defined by the population, or at least by the
political stratum.15
13. Steiner points out that an important source of discontent in the Bernese Jura is that the Catholic, Frenchspeaking inhabitants constitute a minority in these important
respects not only within the country but also within the canton, where the majority are Germanspeaking speaking Protestants. In the perceptions of the Jurassiens,
therefore, the majority appears to be completely dominant. Steiner, Bedingungen fÜr GewaltlÖse Konflictregulierunsmuster, pp. 433–34. Given this unusual situation, it
is easy to see why the inhabitants of the Jura would want to form a separate canton, a solution that (to the outsider) appears to be quite in the Swiss scheme of mutual
guarantees.
14. Even the revenue from the state tax on radio and television sets is distributed to the broadcasting companies of the zuilen, which have their own studies, staffs, and
of course programs.
15. See the discussion in the following chapter, pp. 144–49.
Page 120
For if demands on the government for ‘‘solutions” to major problems go unmet year after year, allegiance is likely to give way to disillusion and contempt, particularly
when a “problem” involves extensive and acute deprivation among some considerable part of the population—galloping inflation, for example, extensive unemployment,
acute poverty, severe discrimination, woeful inadequacies in education, and the like.
Now this requirement is not unique to polyarchy, nor is it even unique to countries with a considerable amount of subcultural pluralism; it is very likely a general
requirement for all regimes—surely for all polyarchies. The reason for emphasizing it here, however, is that it may sometimes be inconsistent with the first two
requirements. For to contrive a political system that works by unanimity and minority veto, or by shifting coalitions, and guarantees that no majority coalition will act
adversely toward any of its subcultural minorities may be a perfect recipe for governmental immobility, for a system, that is to say, in which major problems, as these
are defined by the political stratum, go unsolved because every possible solution is vetoed by some minority whose leaders feel that its interests are threatened.
Although, as the experience of Switzerland and the Netherlands shows, immobility is not inherent, it is a serious complaint in some systems of this kind.16
Systems with marked subcultural pluralism may sometimes confront, then, a set of unhappy and even tragic choices: (a) a polyarchy that provides mutual guarantees to
its minorities but cannot respond to demands for solutions to major problems sufficiently well to hold the allegiance of the people; (b) a hegemony that tries to meet
these problems by
16. On Lebanon, for example, see the comments of Michael Hudson, The Precarious Republic: Political Modernization in Lebanon (New York: Random House,
1968), pp. 11–12, 87–88, 328–31, and Ralph E. Crow, “Religious Sectarianism in the Lebanese Political System,” Journal of Politics 24 (August 1962): 489–520.
On the general point, see Lorwin, “Segmented Pluralism,” p. 16.
Page 121
coercing, if need be, members of one or more subcultures; or if the subcultures are also regional, (c) separation into different countries. Only the last may enable
polyarchy to survive among the dissenting minority. Thus the price of polyarchy may be a breakup of the country. And the price of territorial unity may be a hegemonic
regime.
Political Institutions and Governmental Effectiveness
The extent to which political institutions are merely an outcome of all the other factors that have been discussed here, or are themselves in some degree an independent
factor, is one of many questions that political scientists have not satisfactorily resolved. Although for about a generation political institutions have been more and more
widely interpreted as mere epiphenomena, this period of reductionism in political science may now be coming to an end. However that may be, in order to assay the
problem of governmental effectiveness in competitive regimes, some emphasis must, I think, be given to political institutions, even though I intend to skirt completely
around the formidable problem of “explaining” how countries with competitive systems come to their specific governmental arrangements.
In polyarchies, two kinds of institutional arrangements seem to have important consequences for the “effectiveness” of the government. One of these has to do with the
relationship between the executive and the other major political forces in the country, including a relationship that has often been the source of major problems in
competitive regimes —between the executive and the legislature. It seems to be true that all the nineteenthcentury competitive regimes that have managed to survive as
polyarchies in the twentieth century have developed strong executives armed with extensive capacities for action. All, including France, have rejected the classic model
of assembly government. Whether de facto or de jure, everywhere responsibility for coordinat
Page 122
ing policies and establishing priorities, and much of the responsibility for innovation, has shifted toward the executive.
The other significant institution is the party system. It is a defensible, if, alas, vague hypothesis that the costs of toleration are reduced where party systems appear to
contribute more to integration and action than to fragmentation and paralysis. Although it is exceptionally difficult to cast this proposition in operational terms, it seems
clear that highly fragmented multiparty systems (Sartori’s “extreme” or “polarized’’ pluralism)17 can lead to unstable or weak coalitions that are unable to cope with
major problems and thereby exaggerate in the eyes of the public and even the political elites the partisan or game aspects of political life. These results in turn may
stimulate a loss of confidence in representative democracy and the willingness to tolerate political conflicts.18
About a third of the contemporary polyarchies meet the problem of party fragmentation with twoparty systems of one sort or another. Among the remaining twothirds
a substantial number have successfully avoided extreme fragmentation by:
A party system with two large, dominant parties and one or several small parties as in the German Federal Republic.
17. In various works, Giovanni Sartori has carried out an extensive comparative analysis of the causes and consequences of highly fragmented multiparty systems. See
his “European Political Parties: The Case of Polarized Pluralism,” in Joseph LaPalombara and Myron Weiner, Political Parties and Political Development (Princeton:
Princeton University Press, 1966) pp. 137–76; his Partiti e sistemi dipartito (Florence: Editrice Universitaria, 1965); and his “The Typology of Party Systems,” in Stein
Rokkan and Erik Allardt, eds., Cleavages, Parties and Mass Politics (New York: The Free Press of Glencoe, 1970).
18. Some data and argument supporting this interpretation will be found in Jack Dennis, Leon Lindberg, Donald McCrone, and Rodney Stiefbold, “Political
Socialization to Democratic Orientations in Four Western Systems,” Comparative Political Studies 1 (April 1968): 71–101, esp, 91–92.
Page 123
A party system with one normally dominant government party that gains close to 50 percent of the popular votes and seats in the parliament, with the opposition
fragmented into three or more parties, as in India, Japan, and the Scandinavian countries.
A special coalition of the two major parties, as in Austria until recently and in West Germany from 1966 to 1969.
A highconsensus, unanimityoriented system as in Switzerland, where all major parties are included in the executive.
But countries that experiment with competitive politics do not always succeed in preventing fragmentation of the party system. In any country where competitive politics
is accompanied by a highly fractionalized party system (which in a parliamentary system is likely also to produce a weak executive) the chances for a shift toward a
hegemonic regime are rather high. In the past decade, nearpolyarchies in both Greece and Argentina foundered in part, it seems, because party fragmentation made for
a regime that because of its inability to deal effectively with the country’s problems was unable to win or to hold the allegiance of enough of the population to prevent a
coup d’etat from establishing a dictatorship. And in France, the classic home of assembly government, the Fourth Republic foundered because of its failure to solve the
excruciating problem of Algeria.
Page 124
8. THE BELIEFS
OF POLITICAL ACTIVISTS
To the extent that the factors described in the preceding chapters propel a country toward a hegemonic regime or toward public contestation and polyarchy, they must
operate, somehow, through the beliefs of the people, particularly those who are most involved in political life. At the very least, then, the beliefs of the political activists
are a key stage in the complex processes by which historical sequences or subcultural cleavages, for example, are converted into support for one kind of regime or
another.
Assumptions
The relationships examined in this chapter might be represented as follows, where the simple arrow means “explains” or “accounts for,” and the arrow — P means
‘‘affects the probability of”:
I II III IV
Factors determining Political Political
— P Regimes
beliefs Beliefs Actions
The question is, then, the extent to which beliefs (II) determine the political actions (III) of people in a country and
Page 125
thus affect the chances for a particular kind of regime (IV), defined here according to the extent of hegemony, public contestation, and polyarchy. To the extent that
beliefs do affect regimes, we shall want to know what factors determine beliefs (I). There are many problems with this paradigm, but the paradigm itself helps us to get
at these problems.
Let me begin by making clear that I use the term “beliefs” in a broad sense. I am not going to distinguish between beliefs and knowledge. By knowledge we usually
mean beliefs that seem to us well grounded in truth, perhaps incontrovertible. Most of us are willing to call a physicist’s beliefs about physics his “knowledge” of
physics; yet many of us would be less charitable toward his beliefs about politics. In politics one man’s “knowledge’’ is often another man’s disbelief.
So far as I am aware, no thoughtful person denies the relevance of beliefs to action. Beliefs guide action not only because they influence or embody one’s more distant
goals and values—one’s religious salvation, for example, one’s security in old age, or the independence of one’s country—but because beliefs make up our
assumptions about reality, about the character of the past and the present, our expectations about the future, our understanding of the hows and whys of action: in short,
our “knowledge.”
The importance of the purely cognitive element in our beliefs, whether beliefs about politics or about other things, can hardly be overestimated. Knowledge of the rules
of a game, for example, has not only normative but also cognitive significance, as many a foreigner has discovered in the United States upon attending his first game of
baseball. The foreigner may not give a hoot whether the rules of baseball are “good” or ought to be observed, but because he does not know what the rules are he is
totally unable to make sense of what is happening before his eyes. An American watching a game of cricket is equally lost. As in baseball or cricket, so too in politics:
what we believe influences not only what we want to happen but what we think actually
Page 126
happens. In its emphasis on “is” rather than “ought,” Marxism may be an extreme case, but it serves nicely to make the point. While Marx’s writings unquestionably
have moral import and appeal, their manifest content is almost exclusively cognitive, not moral. To the Marxist, Marx’s work explains what has happened, is happening,
and will happen; only by implication does it specify what ought to happen.
It seems evident, too, that individuals’ beliefs influence collective actions and hence the structure and functioning of institutions and systems. It is difficult to see how a
polyarchy could exist if a majority of the politically active strata of a country believed strongly that a hegemonic regime was more desirable and could be achieved by
supporting antidemocratic leaders and organizations. In a hegemonic regime, presumably the leaders, at least, must prefer hegemony to any single alternative.
As these examples suggest, the way in which different beliefs are distributed among the people of a country is also important. Probably a polyarchy requires a much
more widespread belief in the general desirability of the system than is required to sustain a hegemonic regime. But if it is true that in all systems different people have
unequal impact on political outcomes, then it is important to know who holds what beliefs. Surely Lenin’s beliefs, and later Stalin’s, had very much more influence on
developments in Russia in this century than the beliefs of, say, two randomly selected Russian peasants. The way the institutions of government have operated in the
United States since 1787 has been determined much more by the beliefs of the fiftyfive men at the Constitutional Convention of 1787 than by the beliefs of fiftyfive
average citizens of Philadelphia that year. I realize that my assumption runs the risk of opening up the old controversy over the role of the Great Man in history, a
question I do not wish to pursue here. But I hope that most readers will share my assumption. In any case, in this chapter I am going to be mainly concerned with the
beliefs of the people
Page 127
most involved in political action, such as activists, militants, and in particular those with the greatest manifest or implicit power, actual or potential, the leaders or
potential leaders. This concern does not mean that the beliefs held among the less influential strata are irrelevant, but only that a stronger case can be made for treating
the beliefs of the politically most active and involved strata as an important explanatory factor.
The case is strengthend by evidence on several other aspects of political beliefs. Within a particular country, the more rudimentary and unorganized an individual’s
political beliefs are, the more uncertain the relation, if any, between belief and action. To take some hypothetical examples, a person whose belief system in effect
contains the essentials of Mill’s argument on liberty is very likely to oppose a regime that threatens censorship of the press or the suppression of political parties. One
whose beliefs contain in effect the central arguments of the Communist Manifesto is likely to favor attempts by workingclass organizations to contest a bourgeois party
or regime. But it is very difficult to predict the kind of regime, movement, or party that would be opposed or supported by one whose whole political credo can be
summed up as ‘‘the times are bad and people like me are badly off in this country.”1
Most people, it appears, have quite rudimentary political beliefs. Rich and complex systems of political belief are held only by small minorities. On the present evidence,
it is reasonable to think that while this is true in all countries, the lower the average level of formal education in a country, the smaller the minority is that has complex
beliefs about politics. In general, the likelihood that one holds a complex set of interrelated beliefs about politics increases with one’s level of education and with one’s
interest and involvement in
1. Philip E. Converse, “The Nature of Belief Systems in Mass Publics,” in David E. Apter, ed., Ideology and Discontent (New York: The Free Press of Glencoe,
1964) pp. 213 ff.
Page 128
politics. It seems obvious, in fact, that the purely cognitive component—knowledge of a variety of aspects of political life, including the rules of the game—must be
greater among leaders and activists than among, say, politically apathetic inhabitants of a country.2
To sum up, it is reasonable to think that activists and leaders are more likely than most other people
1. to have moderately elaborate systems of political beliefs
2. to be guided in their actions by their political beliefs
3. and to have more influence on political events, including events that affect the stability or transformation of regimes
Several questions arise:
First, what kinds of beliefs among activists crucially affect the chances for public contestation and polyarchy?
Second, how do these beliefs come about?
Third, can an explanation of how political beliefs come about account for them so fully that beliefs can be treated simply as intervening variables? To put the question
the other way round, is it useful to treat the political beliefs of activists as an independent explanatory factor, comparable in significance to those described in earlier
chapters?
Some Crucial Beliefs
It is an unhappy fact that no one can answer these questions satisfactorily at present. To be sure, it is possible to say something about how people in different countries
vary in their attachment to certain beliefs, that is, about variations at II in the paradigm presented a few pages back. But satisfactory evidence on variations in beliefs
among (and within3)
2. Ibid., pp. 218 ff.
3. Since beliefs are rarely if ever uniformly held in a given country, we are concerned with “significant” variations or differences in “modal types.” The conceptual and
methodological problems in
Page 129
countries is still limited to a few countries, most of which, as might be expected, are polyarchies. Evidence about beliefs in hegemonic regimes is slight; often we use
indirect evidence from Italy or Germany as a substitute, on the assumption that beliefs of Italians or Germans may help to account for their experiences with fascism and
nazism.
Because of the small numbers of countries studied and the concentration on polyarchies, it is impossible at present to demonstrate conclusively that variations in beliefs
(at II) significantly affect the chances for polyarchy or hegemony (at IV). The problem is compounded by uncertainties about the intervening relationships with behavior
(at III).
Strictly speaking, then, assertions about the impact of beliefs on the character of a regime must be treated as hypotheses that cannot yet be tested satisfactorily against
reliable and relevant data. Nonetheless, the possible role of beliefs is too great to be ignored. For there are plausible reasons for thinking that certain beliefs do affect
the chances for hegemony or polyarchy. Since the evidence is so fragmentary, I intend to do no more than to offer a few illustrations.
Legitimacy of polyarchy
Let me begin by offering three propositions.
First, the greater the belief within a given country in the legitimacy of the institutions of polyarchy, the greater the chances for polyarchy.
Although such a statement may seem so nearly selfevident as to require no discussion, a few comments will serve to
volved are analogous to those in the study of “the modal personality” of a country or a culture. Cf. Alex Inkeles and Daniel J. Levinson, “National Character: The Study
of Modal Personality and Sociocultural System,” The Handbook of Social Psychology, ed. Gardner Lindzey and Elliot Aronson, 2d ed. (Reading, Mass.: Addison
Wesley, 1969), 4: 418–506.
Page 130
round it out. To believe in the institutions of polyarchy means to believe, at the very least, in the legitimacy of both public contestation and participation. In practice,
these two dimensions of polyarchy are somewhat independent not only historically, as I pointed out in chapter 1, but also as beliefs. Around the time of the reform of
the British Parliament in 1832, many Whigs and Liberals affirmed by word and deed their belief in the legitimacy of the institutions of public contestation that had been
evolving in Britain for more than a century. If the legal basis for some of these forms of contestation was less clear than it is today—the freedom of the press to criticize
the government, for example—in practice the boundaries were fairly broad, and these defacto freedoms were supported by most Whigs and Liberals. What most
sharply distinguished Whigs and Liberals from Radicals, however, was the Radical commitment to a very broad, even universal, suffrage. Probably most Whigs and
Liberals in the 1830s believed that the Reform Act went quite far enough; yet that reform left about six adult males out of seven without the franchise—a degree of
exclusion the Radicals found unacceptable. It is not too much to say, then, that in the 1830s most Whigs and many Liberals did not believe in the legitimacy of
polyarchy. They believed firmly instead in the legitimacy of what I have called a competitive oligarchy.
It is also possible to believe in the legitimacy of inclusion but not of public contestation. Peron’s dictatorship in Argentina sought to do what no previous regime had
done, to incorporate the working strata into Argentine life—economic, social, and political. Paradoxical as it may seem, since 1930 the only elections in Argentina that
have been reasonably honest, fair, and widely participated in by the electorate, the outcome of which has not been set aside by the military, occurred during Peron’s
dictatorship. Not that Peron believed in or supported polyarchy; under his rule, oppositions were increasingly suppressed. Yet Peronism stood and even today stands
for the full inclusion of the working strata in the politi
Page 131
cal system, and although it may grant legitimacy to dictatorship it denies legitimacy to any system that excludes or discriminates against the working strata or their
spokesmen.
As with other beliefs, the views of activists and leaders are likely to be more crucial than those of other people. Yet since even inactive or excluded strata may
sometimes be mobilized, their beliefs are very far from irrelevant. As with other beliefs too, the complexity and richness of beliefs about the legitimacy of polyarchy no
doubt increase with education, political interest, and involvement. Even in wellestablished polyarchies, very young children can hardly be expected to believe in
“democracy’’ the word and the concept have no meaning, after all, to most fouryearolds. But evidence indicates that in the United States, Italy, and Germany, by the
time they finish school a substantial majority of young people agree that “democracy is the best form of government.” Nor is this belief devoid of concrete detail; they
have also begun to learn something about the content of that commitment. Thus a substantial proportion of American children in the early grades appear to believe that
political conflict does more harm than good, but only a small minority of older students express this belief. To many American children, competing political parties at
first seem unnecessary and undesirable or downright confusing, but as they develop they tend to acquire a belief in the desirability of competing parties. The idea that
people should be allowed to make speeches “against our kind of government” is unfamiliar and uncongenial to young children, but a great many older children have
learned to accept it.4
With all that is now known about political beliefs, it would be wildly unrealistic to expect, even in a country like the United States where a democratic ideology has been
the domi
4. Jack Dennis, Leon Lindberg, Donald McCrone, and Rodney Stiefbold, “Political Socialization to Democratic Orientations in Four Western Systems,” Comparative
Political Studies 1 (April 1968): 78, 86, 89.
Page 132
nant belief system for generations, that many people would possess an elaborately workedout democratic theory. Scratch the average democrat, and you will not find
a Locke, a Rousseau, a Jefferson, or even a Lincoln. Nonetheless, the legitimacy of polyarchy or the belief that “democracy is the best form of government” probably
does not exist in a total vacuum. As a belief, it is not ordinarily insulated completely from other beliefs. Thus it is a reasonable conjecture, supported by a certain amount
of evidence, as we shall see in a moment, that an individual is more likely to believe in the legitimacy of polyarchal institutions if he also holds certain other beliefs. But
before we examine these, let me first add two more propositions:
Countries vary a great deal in the extent to which activists (and others) believe in the legitimacy of polyarchy.
These variations are to some extent independent of the social and economic characteristics of a country: two countries with a great manysimilarities in their social and
economic orders may vary significantly in the extent to which activists (and others) believe in the legitimacy of polyarchy.
Argentina: the notsodeviant case.5 To illustrate and to lend
5. The argument and data in this section are derived from the following sources: “Political Oppositions in Argentina” an unpublished seminar paper by Guillermo
O’Donnell, who first suggested to me some of the main lines of interpretation that follow; Carlos Alberto Floria, “El Comportamiento de la Oposicion en la Argentina”
Natalio R. Botana, “Las Crisis de Legitimidad en Argentina y el Desarollo de los Partidos Politicos’’ Mariano Grondona, “Algunas observaciones sobre la evolución
reciente del sistema político argentino” and Rafael Braun, “La representatividad de los partidos políticos y la interpretación del interés publico por parte de las fuerzas
armadas: un dilema argentino” (all papers presented at the Primer Encuentro International de Ciencia Politica, Buenos Aires, August, 1969);Carlos Alberto Floria,
“Una Explicación Política de la Argentina,” CIAS, Revista Mensual del Centro de Investigación y Acción
Page 133
some support to these propositions, particularly the last, it is instructive to compare Argentina with, let us say, Sweden. Outwardly, Argentina’s political development in
the late nineteenth century followed a path to polyarchy very similar to that of Sweden, Britain, and a number of other stable polyarchies. To be sure, Argentina’s early
history is hardly analogous to Sweden’s; at the very least, Argentina squeezed into a few decades what occurred over centuries in Sweden: a decade of revolution and
struggle for independence from 1810 to 1820, a decade of what Germani has called anarquía, caudillismo, guerras civiles, followed by two decades of a unifying
autocracy.6 For the next eighty years, however, Argentina had all the external appearance of a developing polyarchy. It operated under the constitution of 1853 (as it
has continued to do, formally, down to the present day). It is true that participation was restricted to a small minority, but participation in elections does not appear to
have been more restricted in Argentina than in Britain until the Reform Act of 1867 or in Sweden or the Netherlands until late in the century. Toward the end of the
century, the rapidly expanding middle class began to make itself felt in political life. Organized parties came into existence as powerful electoral forces —particularly the
Radicals who, as spokesmen for the middle classes, insistently demanded entry into the system. In
Social 16 (November 1967); Mariano Grondona, La Argentina en el Tiempo y en El Mundo (Buenos Aires: Editorial Primera Plana, 1967); Gino Germani, Politica y
Sociedad en una Epoca de Transicion (Buenos Aires: Paidos, n.d.).; Dario Canton, “Military Interventions in Argentina, 1900–1966” (Paper presented to the
Conference on Armed Forces and Society Working Group, International Sociological Association, London, September 1967), and his “Universal Suffrage as an Agent
of Mobilization” (Paper presented to the VIth World Congress of Sociology, Evian, France, September 1966); and James W. Rowe, “The Argentine Elections of
1963” (Washington, D.C.: Institute for the Comparative Study of Political Systems, n.d.).
6. Politica y Sociedad, p. 196, and in his “Hacia una Democracia de Masas,’’ in Torcuato S. Di Tella et al., Argentina, Sociedad de Masas (Bueno Aires: Editorial
Universitaria de Buenos Aires, 1965), p. 211.
Page 134
1911, about the same time as Sweden and a few years before the Netherlands, Argentina effectively guaranteed universal male suffrage. As a result, election turnout
took a large jump—from 21 percent of all male citizens over eighteen in 1910 to 69 percent in 1912. Although turnout fell off during the next dozen years, it was
roughly comparable to that in the United States and not much lower than it was in Sweden during the 1920s. Moreover, it rose again to 81 percent in 1928 and 75
percent in 1930. Thus in electoral participation Argentina was also comparable with a number of other polyarchies. What is more, with the victory of the Radical
party’s presidential candidate in 1916, Argentina appeared to manage without violence the dangerous transition from competitive oligarchy to a polyarchy based on
universal suffrage.
There were still other factors favorable to polyarchy. According to the criteria so frequently used today to predict the chances for “stable democracy,” Argentina
provided solid grounds for optimism. Its population was highly urbanized, literacy was low, education was widespread, the country was moderately prosperous (its
GNP per capita was probably the highest in Latin America), and it had a large middle class. On all these counts, Argentina was at least as well off as the three countries
where polyarchy has proved to be more durable than in Argentina—Costa Rica, Chile, and Uruguay. Nor does that deus ex machina of much theorizing about
democracy help us greatly: landholdings were concentrated, to be sure, but apparently less than in Chile, possibly less even than in Costa Rica, and very likely less than
in Australia.
As it did for most other countries, the onset of the Great Depression in 1929 created serious problems for Argentina. But other polyarchies were also hit by economic
crisis. Some that were also highly dependent on international trade, like Sweden, and even some that were heavy exporters of agricultural products, like Australia and
New Zealand, none
Page 135
theless met the crisis with actions that retained, restored, perhaps even enhanced the confidence of their citizens in the effectiveness of their governments. In Argentina,
things went differently. Although the notables lost few of their privileges under Radical governments, their parties steadily lost votes and representatives. In the late
1920s, conservatives had increasingly attacked the institutions of polyarchy in Argentina—universal suffrage, political parties, the Chamber of Deputies, the
ineffectiveness and personalismo of the President. Then came economic crisis. A military coup brought the Argentine experiment with polyarchy to an end. Since 1930,
polyarchy has never been effectively restored.
It is puzzling that polyarchy collapsed in Argentina but not in Sweden, Australia, and New Zealand, or for that matter in Costa Rica or, if we take the whole period into
account, in Chile and Uruguay. I am not aware of any wholly satisfactory explanation for the failure of polyarchy in Argentina. Yet one thing seems clear: the differences
in regimes cannot be explained by appealing to the usual explanatory factors—the level of socioeconomic development, urbanization, education, size of the middle
class, per capita income, and so on. Although a full explanation would surely be very complex, one crucial factor does emerge with striking clarity: Argentinians appear
never to have developed a strong belief in the legitimacy of the institutions of polyarchy. Consequently, when the regime confronted serious difficulties it was easily
swept aside by dictatorships that were supported first by conservative elites, the socalled oligarchy, and later under Peron by the working classes—neither of which
had ever acquired a very deep conviction about the legitimacy of polyarchy.
As every student of Argentine history is aware, despite outward appearances Argentine’s polyarchy did not develop from 1853 to 1930 according to the standard
Western European model. The description presented a moment ago is correct—yet it is also highly misleading, for it ignores
Page 136
two interrelated factors deeply implicated in the failure of polyarchy to acquire legitimacy in Argentina: (1) throughout their period of dominance, the notables openly
rejected elections as the legitimate foundation of their power to govern; and (2) a large part of the population remained isolated and alienated from the political system.
Thus, as it traversed the path from competitive oligarchy to polyarchy, Argentina failed to accomplish what some of the European countries achieved during their
transition: neither the Argentine elites, the middle strata, nor the working classes were ever converted to a belief in the legitimacy of polyarchy.
1. The notables failed utterly to develop and transmit a belief in the legitimacy of certain central institutions of polyarchy. Although the more remote factors are elusive,7
one thing stands out clearly: By their own conduct, the notables taught the Argentinians that elections need not be binding on the losers or the potential losers. Thus they
denied legitimacy to a central institution of polyarchy. For throughout the long period of seemingly “constitutional” government from the 1860s to the electoral reform of
1911, the notables regularly and publicly flaunted both constitutional and legal requirements for universal suffrage. According to law and constitution, ‘‘voting was
universal, for men only, with no qualifications regarding literacy, wealth, property, etc.”8 The notables evidently assumed, no doubt correctly, that if these requirements
were ever enforced, their own governing position would be weakened or destroyed. As in Spain during the same period (and with similar consequences) elections were
uniformly conducted with massive fraud and violence. The
7. Floria mentions as derivatives of the Spanish tradition “the common disrespect for the law,” the practice of “caciquismo,” which “organized illegality under the slogan:
‘for one’s friends even when they are in the wrong, and against one’s enemies even when they are in the right,‘” and the practice of resorting to force when no legal
means are available. He also refers to a nineteenthcentury liberal tradition that reconciled “contempt for the common man with faith in democracy.” “El
Comportamiento,” p. 6.
8. Canton, “Universal Suffrage,” p. 4.
Page 137
notables not only excluded the mass of the population from participating in political life; they were unwilling even to permit conflicts among themselves to be settled
through elections. They interpreted differences among themselves (such as the exent to which Buenos Aires should dominate the provinces) as “zerosum” conflicts with
consequences far too grave to be determined by mere elections.
Thus from the beginnings of the “constitutional” regime, the leaders of the country denied that elections were a legitimate way of displacing a government: as a practical
matter, they provided no alternative to revolution.
It is true, of course, that in other countries where polyarchy later acquired considerable legitimacy, notables had resorted to illegality, corruption, even violence, to win
elections. Elections in Britain in the eighteenth century were hardly a model of fairness. But Argentina was different in three respects. First, in sheer mass and extent,
fraud and violence were by all accounts extraordinary—not far from those famous elections in Spain in the nineteenth century when the outcome of the vote was
announced in advance.9 Second, as in Spain, by openly flouting the very constitution and laws that ostensibly gave their regime its legitimacy, the notables helped to
undermine the legitimacy of any regime based on elections. Finally, unlike Britain, in Argentina there was no transition period between rule by notables and full suffrage
during which the political strata came to accept the legitimacy of honest elections. The lesson the notables transmitted, then, was this: when elections turn out badly, the
losers are not obliged to accept the results.
This was the legacy the middle and working classes inherited when they were finally allowed into electoral politics by the reform of 1911. That reform, which came
about under pressure from the Radicals (pressure that included abortive revolts in 1893 and 1905) effectively enforce an already
9. Gerald Brenan, The Spanish Labyrinth (Cambridge: Cambridge University Press, 1962), p. 5.
Page 138
existing constitutional provision for universal male suffrage, provided for a secret ballot, and created electoral courts of federal judges to oversee the elections. But
where it is the norm that those who lose the election may reverse the results by force, polyarchy rests on fragile foundations. And such was the case in Argentina. That
little has changed since 1930 is vividly demonstrated by the brief interlude of nondictatorship from 1955 to 1966. In 1962, when the Peronistas were allowed to
present candidates for Congress and some provincial governorships, they gained the largest share of the votes. Thereupon President Frondizi annulled the election; even
so, he was arrested and thrown out of office by antiPeronist military forces discontented with the election results and fearful of a repetition of Peronism. When
elections were scheduled in 1963 for a president to replace Frondizi, Peronist candidates were prohibited. However, when it began to appear in 1966 that the
Peronistas might win many provincial governorships in the forthcoming elections, instead of waiting to annul the elections the military took over, evidently with very wide
support among notables and middle classes fearful of a return to Peronism and tired of a system of fragmented parties. The new dictatorship under General Onganía
sought to put an end to these evils by suppressing all political parties and elections and ruling by decree.
2. Excluded from politics by notables who showed open contempt for electoral processes, the great bulk of the population remained isolated and alienated from the
political system—hardly likely to rush to its defense to prevent a military takeover. But isolation and alienation were vastly amplified by another factor: the enormous
number of noncitizens, who remained psychologically and legally outside the electoral process. The rate of immigration to Argentina far exceeded that to the United
States.10 Yet where quasiuniversal suffrage and competitive parties in the United States en
10. Germani, Politica y Sociedad, pp. 247, 265.
Page 139
sured that immigrants would soon be made citizens in order that they might vote, in Argentina the incentive of the immigrant to acquire citizenship was reduced by the
fact that until 1911 even if he were to become a citizen he would still be excluded from politics, and by a variety of other factors. In any case, because most immigrants
did not become citizens, a large part of the population remained legally and psychologically outside the Argentine polyarchy. In 1914 nearly a third of the population of
Argentina were foreigners; in Buenos Aires, which dominates the political life of the country, half the people were foreigners. Among adults, the proportions were even
higher—72 percent in Buenos Aires, 51 percent in five of the most thickly settled provinces, 20 percent elsewhere. Among male adults (who as citizens would have
been entitled to vote) the percentages were higher yet: perhaps four out of five adult males in Buenos Aires were noncitizens and thus automatically excluded from
participating in the political life of the country. The immigrants were, of course, disproportionately concentrated among the working classes. Thus Argentina had a large
proletariat and even a sizable middle class with not great attachment to the Argentine political system.
The upshot of Argentina’s political development, then, is that belief in the legitimacy of the institutions of polyarchy—and particularly in the binding character of
elections—is shallow, especially among the elites, but perhaps also among the people as a whole. For example, the belief in free elections among key segments of the
population is too weak for the country to engage regularly in elections and abide by the results. A survey taken in greater Buenos Aires in March 1966, shortly before
the military took over, reveals a good deal. (Table 8.1) It is important to keep in mind that the Peronistas could be expected to win a third or more of the votes: thus to
ban their participation was to prohibit the largest group of voters in Argentina from having its own candidates. Nearly a fifth of greater Buenos Aires thought
Page 140
it was right to do so; almost as many more expressed no opposition to the idea. But what is interesting, though hardly surprising, is the extent to which the question of
banning Peronists was a class issue. The table shows that the idea of banning Peronist candidates was strongest among the upper strata, where opposition to Peronism
was strongest, and weakest among lower strata, where support for Peronism was
Table 8.1. ‘‘Would you consider it good or bad if the government
were to ban Peronist participation in future elections?”
Don’t know,
Wrong Right no answer
64% 19% 17%
By education
Illiterate or
primary education 65% 16% 19%
Secondary education 63% 24% 13%
Academic education 50% 35% 15%
N = 1004
Source: Encuestas Gallup de la Argentina, March 1966, reported in Polls 2 (Spring 1967), p. 22.
greatest. Only half the more educated group opposed the ban; this was also the segment that would support the military takeover. The fact that the least educated
opposed the ban can hardly be read as a commitment to free elections on their part; it would be interesting to know how many might have supported a total ban on
elections of any kind under a Peronist regime.
Thus a somewhat cynical but not overly exaggerated formulation of a basic norm in Argentine politics would be this: I believe in elections as long as I can be sure that
my opponents will not win.
Authority
A number of writers have stressed the importance of beliefs about or attitudes toward authority.11 Thus in the theory
11. Inkeles and Levinson, “National Character,” p. 448 ff.
Page 141
that he has been developing for some years, Eckstein assigns central importance to the patterns of authority that prevail within a country. Eckstein has hypothesized that
democracy would be more stable if the authority patterns in the government were “congruent” with the patterns in other institutions and associations in the country.12
Presumably “congruent” authority patterns would also make it easier for hegemonic regimes to carry on.
The idea is highly plausible, certainly, that beliefs about the nature of authority relationships between government and the governed are crucial to the chances for the
emergence of different kinds of regimes. To make the point in extreme form: if most of the inhabitants of a country believe that the only proper relation of people to their
government is one of complete hierarchy, of rulers to subjects, of command and obedience, the chances that the regime will be hegemonic are, surely, high. Given such
views, there can hardly be a place for public contestation. And it may also be true, as Eckstein can be read as saying, that people find it hard to reconcile markedly
different beliefs about authority; it is stressful to be required to believe in the validity of hegemonic authority in one relationship and democratic authority in another.
Eckstein, who chose Norway for a limited test of his theory, has argued that Norwegians have been singularly successful in handling their conflicts and cleavages
because they almost uniformly believe in (and adhere to) democratic authority relationships.
Since democratic norms pervade Norwegian life in the family, in schools, in economic and friendly societies, pressure groups and parties, local and national
government, one need not wonder that Norwegians regard it as both natural and moral, a general “way of living,”
12. Harry Eckstein, Division and Cohesion in Democracy: A Study of Norway (Princeton: Princeton University Press, 1966), app. B (1961), and passim.
Page 142
not merely a form that has merit in national governance.13
Norway may represent one end of a scale. Yet polyarchy seems to be consistent with considerable variation in beliefs about authority.14 In fact since polyarchy
requires both open conflict among organizations and a capacity for inventing and accepting compromise solutions, it is arguable that this combination of conflict and
conciliation may be facilitated if a considerable measure of hierarchic authority is acceptable within organizations. Lijphart has emphasized the strong tendency of the
Dutch “to be obedient and allegiant—regardless of particular circumstances,’’ a quality without which the strongly organized cleavages in the Netherlands might long
since have destroyed polyarchy in the Netherlands.15 Even the democratic Norwegians are described by Eckstein as deferential toward “specialized functional
competence,” which “leads to a particularly high valuation of the expert and the specialist.”16 At least since Bagehot, Englishmen have been characterized as deferential
toward their “social betters” and toward governmental authority.17 Americans are thought to be a good deal less deferential toward government and social elites than
Englishmen18 and may be less deferential toward expertise than Norwegians.
13. Ibid., p. 173.
14. Ibid., p. 189.
15. Arend Lijphart, The Politics of Accommodation: Pluralism and Democracy in the Netherlands (Berkeley and Los Angeles: University of California Press, 1968),
pp. 144 et seq. See also the comments of Hans Daalder, “The Netherlands: Opposition in a Segmented Society,” in Robert A. Dahl, ed., Political Oppositions in
Western Democracies (New Haven: Yale University Press, 1966), p. 197.
16. Ibid., p. 166.
17. Ibid., pp. 156, 183. See also Richard Rose, Politics in England (Boston: Little, Brown, 1964), pp. 38–41.
18. Some interesting fragmentary evidence on the tendency of English children to see the world as more hierarchial than American children is in Fred I. Greenstein and
Sidney Tarrow, “An Approach to the Study of Comparative Political Socialization: the Use of a Semi
Page 143
Nonetheless, it is most unlikely that polyarchy is equally consistent with all beliefs about authority. Certainly some extreme patterns are much more favorable to
hegemony. The beliefs of the ruling tribal group in Ethiopia, the Amhara, provide an example. According to Levine:
The complex of beliefs, symbols, and values regarding authority constitutes a key component of Amhara political culture. Throughout Amhara culture appears the
motif that authority as such is good: indispensable for the wellbeing of society and worthy of unremitting deference, obeisance, and praise. … Obedience is the
prime objective in the socialization of Amhara children … This experience in the family is continuous and consistent with the rest of Amhara culture … Family
patriarch, parish chieftain, wealthy landlord, ecclesiastical dignitary, political dignitary, military officer—all are perceived in the imagery of fatherhood; all are the
objects of comparable attitudes regarding the obeisance which is due them and the benefits which may be expected from them … All of these themes are realized
in the most complete form and on a grand scale in connection with the highest authority figure of all, the emperor. The incumbent of this role has been the recipient
of the most extreme forms of obeisance. It has been customary, for example, for Ethiopian subjects to prostrate themselves and refrain from lifting their eyes in his
presence.
Views like these are hardly conducive to public contestation.
Amhara culture proscribes direct and honest public criticism of any authority. In authoritarian relationships —and again, all political interaction among the Amhara
Projective Technique” (Paper presented to the meetings of the American Political Science Association, September 1969).
Page 144
is contained within authoritarian relationships—there are only three alternatives: complete deference, acquiescence, and flattery; criticism by devious and covert means;
or outright rebellion.
In these circumstances, it is hardly surprising that ideas favoring the institutions of polyarchy fall on totally barren soil:
Despite the existence of Westerneducated individuals who have promoted the idea of parliamentary representation in Ethiopia for more than sixty years, the idea
runs so counter to the authoritarian cast of Amhara political culture that electoral procedures are simply not taken seriously.19
Effectiveness
Another set of relevant beliefs, as I suggested in the last chapter, are those having to do with expectations about the effectiveness of different regimes in dealing with
critical problems.20 Since we are dealing here primarily with the beliefs of activists and leaders, a “problem” exists and is “critical” if it is perceived as such by a
significant proportion of activists or leaders. (Naturally this way of defining a ‘‘critical problem” might be unsatisfactory in a different context.)
Expectations about governmental effectiveness can be a more or less fixed element in the political culture of a country: the young may be regularly socialized into
believing that their government is, on the whole, highly effective or chronically ineffective. Thus by the time they reach the eighth grade in school a high proportion of
white Americans
19. Donald N. Levine, “Ethiopia: Identity, Authority, and Realism,” in Lucian W. Pye and Sidney Verba, eds., Political Culture and Political Development (Princeton:
Princeton University Press, 1965), pp. 250–51, 253.
20. Cf. Seymour Martin Lipset, Political Man (Garden City: Doubleday, 1960), p. 78.
Page 145
acquire the belief (or at least they did until recently) that their government is effective. In one major study, only 2 percent of the eighthgrade students believed that the
American government very often makes mistakes; almost 60 percent believed that the government makes mistakes rarely or almost never. Almost threefourths
believed that the government “knows more than anyone” or “knows more than most people.”21 Given the enormous impact of this socialization process, it is not
surprising that in response to the question in Almond and Verba’s fivenation survey, “Speaking generally, what are the things about this country that you are most
proud of,” 85 percent of the Americans mentioned their governmental and political institutions. Americans not only mentioned pride in political institutions far more often
than anything else, but the American proportion far exceeded that of other countries.22 And political activists in the United States were even more likely than other
people to express pride in American political institutions.23
Although I have not been able to find strictly comparable crossnational data on expectations about governmental effectiveness, there can hardly be any doubt that in
some countries a significant proportion of the population, including the activists, do not have much confidence in the capacity of the government to deal effectively with
the country’s prob
21. David Easton and Jack Dennis, Children in the Political System: Origins of Political Legitimacy (New York: McGrawHill, 1969), p. 133. See also their article,
‘‘The Child’s Image of Government,” in The Annals of the American Academy of Political and Social Science 361 (September 1965): table 6, p. 54.
22. The figures were: U.S., 85%; Britain, 46%; Germany, 7%; Italy, 3%; Mexico, 30%. Gabriel A. Almond and Sidney Verba, The Civic Culture (Boston: Little,
Brown, 1965), table III.1, p. 64.
23. Even with education controlled, American respondents classified as high or medium in “subjective competence” expressed pride in political institutions much more
frequently than those low in “subjective competence.” Ibid., table VIII.4, p. 199. “Subjective competence,” in turn, is highly related to political activity and involvement.
Cf. table VI.2, p. 144; VIII.1, p. 189, and VIII.2, p. 193; and chap. X, passim.
Page 146
lems.24 And just as American children are taught to believe in the effectiveness of their government, so in some countries children learn quite the opposite.25
Nonetheless, even if beliefs about governmental effectiveness are often acquired through early socialization, it does not follow that they must remain unchanged. The
performance of a government may itself change—or may be perceived to change. The skepticism of villagers in Italy, France,
24. For example, interviews in 1964–65 of “167 members of the Iranian political elite identified through a two stage reputational analysis,” to the question, “On the
whole, do you think that the Iranian political system is working very well, not too well, or badly?” the responses were:
Very well 15.6%
Fairly well 31.7
Not too well 26.3
Badly 15.0
Don’t know, etc. 11.4
Marvin Zonis, “Political Elites and Political Cynicism in Iran,” Comparative Political Studies 1 (October 1968): 363.
In a survey by DOXA in Italy in 1967, in response to the question, “Do you feel that things in Italy go well or badly with regard to the performance of the
government Administration (Ministries, public officials, etc.)?” the answers were:
Very well 2%
Well enough 19
Neither well nor badly 22
Rather badly 26
Very badly 8
Don’t know 22
In comparison with other countries, the performance of the Administration was judged as “Better than in other countries’’ by 16%, worse by 34%, and don’t know by
49%. A similar response was given to a question about the honesty and devotion of Italy’s statesmen. Polls 3, no. 4 (1968): 62, #13, #15, #18.
25. In the village in Provence studied by Wylie, the children “constantly hear adults referring to the Government as a source of evil and to the men who run it as
instruments of evil. There is nothing personal in this belief. It does not concern one particular government composed of one particular group of men. It concerns
government everywhere and at all times.” Laurence Wylie, Village in the Vaucluse, rev. ed. (New York: Harper and Row, 1964), p. 208.
Page 147
or India is rooted in centuries of experience with their own governments. Yet as the performance of the government changes, villagers in southern France or in India
also change; they come to believe that the government need not be merely an inept or evil force, but that it can help in dealing with problems.26 Or confident
expectations like those held by Americans can be reversed by failure. Given the unbelievable faith of young Americans in their government described a moment ago, it
is scarcely to be wondered at that when young people were confronted by failures of the government in dealing with the crucial problems of race, poverty, and war, a
number lost confidence in “the system” and became alienated, cynical, or radical.
Where beliefs about the effectiveness of government are uncertain or shallow, as is often true with new regimes, inept performance is even more dangerous. Thus the
inability of newly established polyarchies in Italy, Germany, and Spain to cope effectively with palpable problems—even to maintain public order—generated doubts
about the effectiveness of polyarchy and spurred the shift toward dictatorship. In the personal life histories that Abel collected in Germany in 1934, he found ample
evidence that opposition to the republic was, after 1929, “reinforced by the belief that the growing number and severity of the problems confronting the German nation
were largely due to the inefficiency of the government.”27
Beliefs about governmental effectiveness may be influenced not only by the performance of one’s own government but also by the perceived successes or failures of
other gov
26. To some extent, this may have occurred in Wylie’s village. See his comments on “Peyrane Ten Years Later,” in ibid., pp. 364–65. See also the changes reported
by Mrs. Wiser in the village of Karimpur in India in William and Charlotte Wiser, Behind Mud Walls, 1930–1960 (Berkeley: University of California Press, 1963), p.
224.
27. Theodore Abel, The Nazi Movement (New York: Atherton Press, 1966 [1938]), p. 121.
Page 148
ernments, ancient or contemporary. Philosophers, publicists, propagandists, ideologues, and many other people all exploit the experiences of other governments to
convince their fellow citizens of the relative effectiveness of various kinds of regimes. Like many later writers, Machiavelli and Rousseau used republican Rome as a
standard. Montesquieu made Britain his model—and influenced Americans. Like Britain in the eighteenth century and the United States in the nineteenth century, in the
twentieth century a variety of performances has influenced perceptions—particularly among intellectuals—of the relative effectiveness of different kinds of regimes. In
the 1930s the successes of dictatorships in Italy, Germany, and the USSR were in competition with the achievements of a “middle way” in Sweden and the New Deal
in the United States. In Latin America the influence of authoritarian models has often been great: PRI in Mexico, Peron in Argentina, Castro in Cuba, and most recently
the reformist military junta in Peru.
The evidence indicates, then, that beliefs about governmental effectiveness are strongly influenced both by political socialization and by the way the performance of
different regimes is perceived.
Beliefs about governmental effectiveness may reinforce, weaken, or alter prevailing beliefs about authority.28 If a government is perceived as effective, its successes are
likely
28. Eckstein’s formulation is different from but not inconsistent with this one. He hypothesizes that “congruence between [the government’s] authority pattern and the
authority patterns of other social units in the society’’ is a necessary but not sufficient condition for “high performance by a government.” In Eckstein’s formulation,
governmental performance is broader than the concept of effectiveness used here; it includes not only output efficiency and strife avoidance, which would be roughly
equivalent to my concept of effectiveness, but also durability, legitimacy, and permeation, which is defined as “the ability of a polity to derive resources from, and to
carry out its directives in the various segments of its social space.” Harry Eckstein, “Authority Relations and Governmental Performance: A Theoretical Framework,”
Comparative Political Studies 2 (October 1969): 283–87
Page 149
to enhance the prestige of the authority patterns it embodies; the converse is true if it fails. (Figure 8.1)
FIGURE 8.1
Since all governments fail some of the time, socialization builds up a reservoir of confident expectations that function as a reserve during periods of adversity. In new
regimes, these reservoirs are low or even empty; in old regimes with a record of considerable achievement, the reservoirs are likely to be high. In Italy from 1919 to
1923, in Weimar Germany and in Republican Spain, new polyarchies had few reserves to draw on, whereas a reservoir of confidence in the capacity of the government
to deal with problems helped polyarchy in Britain and the United States to survive the challenge of mass unemployment in the 1930s.29
29. Although this is conjectural, the evidence is extremely persuasive. I am unable to find relevant survey evidence earlier than a Fortune Poll taken in the U.S. in March
1940. At that time, about 15% of the labor force, or over eight million persons, were still unemployed. The poll asked:
Which one of the following most nearly represents your opinion of the American form of government: (1) Our form of government, based on the Constitution, is as
near perfect as it can be, and no important changes should be made in it. (2) The Constitution has served its purpose well, but it has not kept up with the times and
should be thoroughly revised to make it fit presentday needs. (3) The systems of private capitalism and democracy are breaking down, and we might as well
accept the fact that sooner or later we shall have to have a new form of government.
Page 150
Trust
Another element of belief that is often emphasized is “the extent to which members of a political system have trust and confidence in their fellow political actors.”30
Thus LaPalombara has stressed the importance of distrust in the Italian political culture: “the general atmosphere is one of fear, suspicion, distrust, and hostility.” “I am
compelled to assert,” he writes,
that on the basis of the evidence at my disposal the attitudes Italians maintain are essentially Hobbesian and that the history of the country suggests that, when the
state of nature becomes too intolerable, it is also a Hobbesian solution concerning government to which Italians turn. The fact that republican government remains
relatively intact … [does not mean] that the Republic can survive many more years on its culturally uncertain and problematical foundations.31
Levine has also called attention to the high levels of distrust among the Amhara in Ethiopia:
Unbelievable as it may seem in retrospect, even the poor preferred the first and overwhelmingly rejected the third:
STATEMENT #1 STATEMENT #2 STATEMENT #3 DON’T KNOW

National total 64.2% 19.2% 5.2% 11.4%
Prosperous 79.9 14.2 2.4 3.5
Upper middle class 71.4 19.8 3.7 5.1
Lower middle class 64.1 20.5 5.2 10.2
Poor 58.4 18.6 7.6 15.4
Hadley Cantril, and Mildred Strunk, Public Opinion, 1935–1946 (Princeton: Princeton University Press, 1951), p. 980.
30. Sidney Verba, “Conclusion: Comparative Political Culture,” in Pye and Verba, Political Culture, p. 535.
31. Joseph LaPalombara, ‘‘Italy: Fragmentation, Isolation, and Alienation,” ibid., pp. 290, 297.
Page 151
The particular features which dominate the Amhara’s image of generic human nature are man’s inherent aggressiveness and his untrustworthiness … Related to this
view of man’s powerful latent hostility is the notion that man is untrustworthy.32
In at least three ways, mutual trust favors polyarchy and public contestation while extreme distrust favors hegemony. In the first place, polyarchy requires twoway or
mutual communication, and twoway communication is impeded among people who do not trust one another. Thus Almond and Verba found much higher levels of
trust among their American and British respondents than in Germany and Italy. Probably authoritarian regimes in Germany and Italy sharply eroded mutual trust; as
LaPalombara’s comments quoted above suggest, the transformation of polyarchy into hegemony may also have been facilitated by a relatively low level of mutual trust.
At any rate, Almond and Verba also found that the percentage reporting that they did not feel free to discuss politics with anyone was markedly higher in Germany (32
percent) and Italy (34 percent) than in britain (12 percent) and the United States (18 percent).33
Second, a certain level of mutual trust is required in order for people to join together freely in order to promote their goals. Organizations based on command, with
authority downward, may be possible (though not necessarily efficient) with mutual distrust; organizations based on reciprocal influence are difficult to form and maintain
in an atmosphere of distrust. Thus Levine argues that the extreme distrust prevailing among the Ethiopians helps explain “the lack of organizations for the articulation and
aggregation of interests in Ethiopia.” This lack, he points out,
32. “Ethiopia,” pp. 257, 258.
33. The Civic Culture, table III.8, p. 83. For responses on trust and distrust, see table IX.2, p. 213. As in many other tables, the data on Mexico are less easily
explained. Thus the percentage of Mexican respondents who did not feel free to discuss politics was 21%—only slightly above the United States.
Page 152
is the result not only of the authoritarian character of the regime, supported as this is by traditional Amhara attitudes toward authority. It also reflects the difficulty
Ethiopians have in undertaking any sort of concerted action, particularly in the political sphere. Here again the Amhara conception of human nature is a factor of more
than academic significance. The mutual distrust and lack of cooperation which inform the political climate of the country are directly related to a very low regard for
man’s capacity for solidarity and consensus … The idea that it is possible to transcend the prevailing atmosphere of anxiety and suspicion by trusting one another …
has been slow to appear and extremely rare.34
Finally, conflicts are more threatening among people who distrust one another. Public contestation requires a good deal of trust in one’s opponents: they may be
opponents, but they are not implacable enemies. “A sense of trust,” Richard Rose writes of Britain, “is pervasive in the political culture. … At the level of government,
trust is important between colleagues and partisan opponents because it reassures all that the particular group in office will not take advantage of the absence of
constitutional restraints upon the powers of government. To do so would not be a violation of law, but of trust, and political leaders value their reputation for
trustworthiness.”35
Cooperation
Trust is obviously related, as these remarks make clear, to the capacities of a people for engaging freely and easily
34. “Ethiopia,’’ pp. 277–88.
35. Politics in England, p. 43. Rose has also written, “This depth of trust is evidenced by the maintenance for centuries of a stable system of government without a
written constitution.” “England, The Traditionally Modern Political Culture,” in Pye and Verba, Political Culture, p. 96.
Page 153
in cooperative actions.36 As with trust, beliefs about the possibility and desirability of cooperation vary a great deal, and, as Levine’s comment suggests, there are
good reasons for thinking that an inability to cooperate reduces the chances for polyarchy.
The relevant units for cooperation or conflict need not be individuals but aggregates—factions, parties, social classes, regions, and the like. What we are concerned
with, then, are the beliefs that people have about the prospects of cooperation and conflict among relevant actors in political life, whether these are individuals,
organizations, or other actors. Keeping this in mind, it will be useful to distinguish three different ways in which someone might view cooperation and conflict among the
relevant political actors. At one extreme, relations among actors may be viewed as a strictly competitive (zerosum) game, where the central rule is: what you gain, I
lose, and what I lose, you gain. Since we have everything to lose and nothing to gain by cooperating, we believe that the best strategy for an actor to follow is strict
competition: never compromise or cooperate but try to win completely on every issue. According to LaPalombara, this is how Italians tend to view politics:
The fact is that the typical Italian feels that elections are contests among mutually and fundamentally antagonistic groups—between the “we” and “the enemy.” It is
assumed that the winners will take advantage of and exploit the losers. For many the election is described
36. As some social psychologists have shown. Thus an attitude survey among college students found that a person who scored high on trust in people is likely to believe
that international conflict can be reduced or settled by cooperation and mutual understanding; low scores are likely to feel that the only way to handle international
affairs is through coercion, force, and power. Morris Rosenberg, “Misanthropy and Attitudes Toward International Affairs,” Journal of Conflict Resolution 1, no. 4
(December 1957): 340.
Page 154
as essentially a lifeanddeath struggle between opposed and irreconcilable ways of life.37
Political activists in the two dominant parties probably hold this view even more strongly than people less involved in politics. A recent study of Communist and
Christian Democratic party activists shows that a preponderant majority of both view the world essentially in Manichean terms as a conflict between dichotomous
forces of good and evil—the good represented by their own party, of course, and evil by the other.38
At the other extreme, relations among actors may be seen as strictly cooperative. Here the central rule is: not only is there no conflict between us but our interests are
identical or so intertwined that we stand to gain or lose together. Hence the best strategy is to cooperate completely and avoid all conflict. Weiner suggests that some
such view as this is an important element in the beliefs of Indian elites:
For reasons which must lie deep in the psychology of Indians, India’s national party leaders, intellectuals, and bureaucrats look upon conflict within their society as
intrinsically undesirable. While the notion of competition and conflict is central to American political life and thought, notions of cooperation, harmony and, to
37. “Italy,” pp. 290–91.
38. Francesco Alberoni, Vittorio Capecchi, Agopik Manoukian, Franca Olivetti, and Antonio Tosi, L’attivista di partito (Bologna: Il Mulino, 1967), pp. 381–87. The
108 activists in the Christian Democratic (DC) and Communist (PCI) parties were classified as follows according to their ‘‘vision of the world”:
DC PCI Total
Absolutely dichotomous 3 17 20
Predominantly dichotomous 14 21 35
Conceives of reality as a struggle or fight among positions but has a more
28 14 42
articulated sensitivity to the alternatives
In no way dichotomous 9 2 11
54 54 108
Page 155
use a favorite Indian word, “synthesis” are central to Indian thought. This is not to say that there is more cooperation and less conflict in India than in the United States,
but simply that while in general Americans easily adapt to situations of conflict, the Indian elite generally finds such situations intolerable.
… Indian historical writings tend to stress the unities of the past and minimize the conflicts and the struggles over power. Indian intellectuals generally envision the pre
British village as a harmonious social and political unit. And contemporary historical writings on the struggle for independence tend to underplay the great internal
dissensions which existed.
Most of the national leadership believes that every effort should be made to eliminate conflict within rural India. No progress, so the argument goes, can be achieved in
the villages without a spirit of harmony and cooperation. Similarly, conflict should be minimized or eliminated at all levels of society, for it is only as the country is united
that it can move forward.39
A third view sees relationships as cooperativecompetitive. Conflict, competition, and cooperation are all viewed as normal aspects of social relationships which
contribute to a healthy, vigorous, progressive society. In this view there are great gains to be secured from cooperation, but because one cannot agree with everyone on
everything some conflict is inevitable. Nor is conflict necessarily bad; it is often a part of a larger process in which the actors in conflict all end up better off. The
important strategy in a conflict is to search for mutually beneficial solutions. Far from being a betrayal of principle, a compromise is essentially a good thing, and a spirit
of compromise vital.
Belief in the virtues and possibilities of cooperativecompetitive relationships is central to English liberal thought. It
39. Myron Weiner, “India: Two Political Cultures,” in Pye and Verba, Political Culture, pp. 235–36.
Page 156
was at the heart of both Adam Smith’s revolutionary notion of the beneficent results of competition and J. S. Mill’s classic defense of the benefits to be derived from the
liberty to express clashing opinions.
What can we conclude about the consequences of these beliefs for hegemony and polyarchy? It would be astounding if the first view did not seriously impede
cooperation among the various actors. Experiments by social psychologists have in fact found this to be the case.40 And this is of course pre
40. One wellknown game used in experiments is the Prisoner’s Dilemma. This is a twoperson game with four possible outcomes: (1) A’s best alternative, which is the
worst for B. (2) B’s best alternative, which is the worst for A. (3) The worst alternative for both. (4) A satisfactory alternative, better for each than the worst but
somewhat less satisfactory for each than his best. An example would be:
B chooses alternative
x y
(4) (1)
A chooses p A wins $9 A loses $10
alternative B wins $9 B wins $10
(2) (3)
q A loses $10 A loses $9
B wins $10 B loses $9
Communication between the players is forbidden. Clearly if A trusts B and tries to be cooperative by choosing p, B might betray him and chose y, as in (1). So too if B
trusts A and chooses x but is betrayed by A, as in (1). If each distrusts the other and they refuse to cooperate, both end up losing, as in (3). Thus they can both avoid
losing and win, as in (4), only if both trust one another and try to cooperate. In one experiment with a game of this kind, Morton Deutsch found that players who were
each told to act as if their opponent’s welfare as well as their own was important, and were also informed that their opponent had received the same instruction, were
more likely to take the risk involved in trusting their oponents and cooperating in order to achieve outcome (4), even in the absence of communication and personal
knowledge of their opponent. When each was told to do as well for himself as possible, and to do better than the other, they were more likely to end up with the worst
outcome (3). Morton Deutsch, “Trust and Suspicion,” Journal of Conflict Resolution 2, no. 4 (December 1958): 265. In a later experiment by Scodel et al., when 22
pairs of players were told to win as much as possible over 50 trials, only 2 of the 22 pairs collaborated, and collaboration did
Page 157
cisely what many observers of Italian politics conclude. LaPalombara, for example, concludes that
one consequence … at an organizational participatory level is extreme polarization of feeling which is in part reflected in the proliferation of political parties … The
fragmentation that grows out of this mutual suspicion and antagonism is dramatically portrayed when we look at Italy’s secondary associations. In the field of labor,
for example, [the labor organizations] … reproduce and intensify in the arena of organized labor the intense feelings that characterize interpolitical party relationships
… [whether] one is speaking of agricultural organizations, professional associations, athletic clubs, youth groups, women’s federations, university student
movements—really the entire gamut of voluntary associations—one is likely to find them fragmented into at least Communist, Socialist, Catholic, and Fascist
factions.41
If the strictly competitive view impedes the degree of cooperation and trust required for polyarchy, the opposite view, that all relationships are or ought to be strictly
cooperative is by no means free of difficulties for the operation of a polyarchy. For it tends to undermine certain key institutions in a polyarchy of their legitimacy,
particularly the political parties. Political parties necessarily engage in conflict; they may even exacerbate conflict. But if political conflict is an unredeemed evil, then
surely competing political parties are an evil. This is, in fact, the public ideology in many countries ruled by hegemonic regimes; in communist countries, certain aspects
not increase as the game went on. According to the experiments, the fact that many players saw the game as competitive prevented them from giving their opponent any
chance to beat them by trying to establish cooperative relations. Alvin Scodel et al., “Some Descriptive Aspects of TwoPerson, NonZeroSum games,” Journal of
Conflict Resolution 3, no. 2 (June 1959): 114.
41. “Italy,’’ pp. 291–93.
Page 158
of Marx’s thought provided a logical and persuasive foundation on which Lenin and his successors were able to erect a theoretical and practical justification for
repressing all parties other than the single ruling party.42
Even among those who believe strongly in democracy, a belief in strict cooperation and the evils of conflict is likely to undermine the legitimacy of political parties.
Given the beliefs about cooperation and conflict in Indian culture, it is not surprising that a number of Indian intellectuals, particularly those of Gandhian inspiration, are
advocates of partyless democracy. The most distinguished advocate of this view is Jayaprakash Narayan, “a strong exponent of what he calls a ‘communitarian’
society, in which party politics would be eliminated, and where popular elections would be replaced by a system of indirect elections from village councils up to the
national parliament.”
During the past few years [Myron Weiner wrote in 1965], Jayaprakash and many of his followers have carried his critique of parliamentary democracy to many
parts of India. Parliamentary democracy, he has argued, with its system of political parties, has intensified conflict at all levels of Indian society, divided India’s
villages, aggravated caste and communal tensions, and encouraged men to seek power for its own sake. He fervently believes that it is possible to increase the
powers of local government (a position he strongly advocates) without at the same time increasing local conflict if only political parties would voluntarily abstain from
participating in village life.…
Even those who reject Jayaprakash’s solution as Utopian are prepared to accept his criticisms of the parliamentary system and his basic assumption that harmony is
a prerequisite for national development … Several … versions of this view … find support
42. Cf. Robert A. Dahl, “Marxism and Free Parties,” Journal of Politics 10, no. 4 (November 1948): 787–813.
Page 159
among some intellectuals and national leaders. One … is that the present system of party government should be replaced by a “national” government which would
consist of all of India’s political parties. The argument that … political conflict would be transferred from outside of government to within government is countered
with the reply that in a “national” government all the political parties would have to give up their own political ambitions in order to serve the common goal of national
unity and national development.43
Obviously it would be wrong to label such views antidemocratic, but it is equally obvious that beliefs of this kind are hostile to polyarchy. Conceivably, beliefs of this
kind, if acted on, might produce a system more democratic than polyarchy (though I think this is highly unlikely); meanwhile they weaken the legitimacy of polyarchy.
Some Indians may, of course, hold both views: they may see politics as strictly competitive, given the current state of things, and yet they may ardently desire that
politics should actually be strictly cooperative. That many Italians may feel much the same way is suggested by the fact that 30 out of 54 Christian Democratic activists
viewed society as a meeting of forces that ideally could work together harmoniously.44 An interesting feature of Leninism is the way that it combines both views: in
bourgeois societies, politics in its most elementary aspect is a strictly competitive, zerosum game played between bourgeoisie and proletariat, but as soon as the
proletarian revolution eliminates classes, politics must necessarily become strictly cooperative. It is not surprising, then, that in Italy activists in the PCI overwhelmingly
perceive social reality as a collision between “antagonistic groups and interests, a class struggle between bourgeoisie and proletariat, an opposition between capital and
43. ‘‘India,” pp. 236–37.
44. Alberoni et al., L’attivista di partito, p. 356.
Page 160
labor.”45 It would be more surprising if many of them, particularly older and more orthodox militants, did not also believe, with Lenin, that a successful proletarian
revolution would bring about a society ruled by harmonious cooperation, in which, therefore, competing parties would be unnecessary.
While extreme beliefs in either strict competition or strict cooperation probably generate an unfavorable environment for polyarchy, the range of beliefs about
cooperation and conflict that are compatible with polyarchy may be quite broad. The operation of polyarchy and public contestation require both cooperation and
conflict in highly visible institutions such as elections, parties, and parliaments. Hence one would expect polyarchy to be favored by beliefs that emphasize the possibility
and the desirability of both conflict and cooperation, particularly, perhaps, where political conflict can be seen as an element in, and restrained by, a higher order of
cooperation.
I am not aware of any crossnational data that bear directly on this question, but scattered evidence on attitudes toward compromise may be relevant, for in the general
sense of the term, compromise is a normal and desirable consequence of conflicts bounded by cooperation. Although the evidence is weak, there do seem to be
significant differences among countries in beliefs about compromise. Certainly in some countries with long and sturdy traditions of polyarchy and public contestation,
compromise is honorable in both word and deed. For example, “Compromise is traditionally considered the Swedish technique in politics, and the word ‘compromise’
evidently sounds pleasing to Swedish voters.”46 Or consider the Netherlands:
45. Ibid., p. 347.
46. Nils Stjernquist, “Sweden: Stability or Deadlock?’’ in Robert A. Dahl, ed., Political Oppositions in Western Democracies, p. 139. It is worth nothing that the well
known book on Sweden by Dankwart A. Rustow is titled Sweden: The Politics of Compromise (Princeton: Princeton University Press, 1955). Rustow comments: “In
Sweden
Page 161
Dutch politics is a politics of accommodation. That is the secret of its success. The term accommodation is here used in the sense of settlement of divisive issues and
conflicts where only a minimal consensus exists.47
The Dutch, and particularly the leaders, take the existence of “ideological differences as basic realities which cannot and should not be changed.” This imposes an
obligation on leaders to search for (and nonleaders to accept) compromises that can be agreed to by all the blocs.48
That a belief in the virtues and possibilities of compromise prevails in many polyarchies seems obvious. Such a belief has been central to American politics. The Swiss
system could not operate without it. In Britain, where the parties follow strictly competitive strategies in elections and cabinet formation, cooperation is facilitated by a
spirit of compromise.49
It would be interesting to have more evidence about countries in which “compromise” is condemned or disdained. As one possible instance, the Spanish cognate
compromiso covers a variety of meanings rather different from those of the English word, and the meaning closest to English usage is generally said to carry with it a
somewhat unfavorable connotation (as is the term “to be compromised” in English). To the extent that these differences in language reflect differences in outlook, as
seems to be the case, then parties and politicians may gain respect for their compromises in Sweden but contempt in Spain and Latin America. Where
perhaps more than anywhere else the technique of compromise has been an essential ingredient in the art of proposing the terms on which men can live together. Many
of her outstanding politicians… have been accomplished masters of that technique” (pp. 230–31).
47. Lijphart, The Politics of Accommodation, pp. 103, 124, 125.
48. Ibid., p. 124.
49. “The practise of settling political differences peacefully has existed for so long that it has come to be expressed in basic cultural norms concerning compromise, self
restraint, and national duty before party duty.” Rose, Politics in England, p. 157.
Page 162
compromises are disdained, cooperation is difficult and conflicts are more likely to go unresolved. If conflict is unbearable, then hegemony may be seen as the best way
to compel cooperation.
Explaining Beliefs: A Misleading Paradigm
As I warned, the beliefs I have described are no more than illustrations. I have not touched at all upon the possible consequences of more coherent bodies of belief that
may unite the specific fragments just described into a more comprehensive and unified outlook, theory, or ideology. If the evidence at hand is inadequate to support
hypotheses about variations in discrete beliefs of the kind just mentioned and their consequences for regimes, to link belief systems or ideologies with regimes would be
even more speculative.
It might be argued that the attempt to isolate beliefs as a factor in accounting for differences in regimes is a vain enterprise, since beliefs themselves are best interpreted
as merely intervening variables. After all, it might be said, if we can explain how beliefs come about, and if beliefs help to explain how regimes come about, we can
account for variations in regimes by explanatory factors more ‘‘basic” than beliefs. The point can be clarified by the paradigm suggested earlier.
If it is true that:
A. Factors determining
beliefs Political beliefs
Political
Actions — P Regimes
then it seems reasonable to argue that:
B. Factors determining
beliefs ——————— P Regimes
Page 163
The simplified paradigm B is, however, thoroughly misleading. In the first place, it assumes that we know a great deal more than we do. In actual fact, existing theory
and data are quite unable to demonstrate the validity or even the plausibility of B. It therefore represents a kind of reduction that is unwarranted on scientific grounds.
No doubt the most popular explanation for beliefs is to attribute them to selfinterest. The difficulty with this particular form of reductionist explanation is not so much
that it is wrong as that it is nearly meaningless. It may very well be true that wholly altruistic beliefs are rare; indeed, it is difficult to imagine how anyone could adhere to
a belief that ran counter to the “interests” of the self at all levels, conscious and unconscious. But that is exactly the problem: to say “selfinterest” is, despite the illusion
of precision, to say very little.50 What one identifies as the “self” varies from one individual to another, from one situation and role to another, and from one culture or
subculture to another. In the culture of ‘‘amoral familism” of South Italy described by Banfield, for relationships outside the family the “self” is the family, and the
villagers act as if they were following this rule:
Maximize the material, shortrun advantage of the nuclear family; assume that all others will do likewise.
The villagers also act according to certain implications of this rule:
1. In a society of amoral familists, no one will further the interest of the group or community except as it is to his private advantage to do so …
50. “The self comprises all the precious things and persons who are relevant to an individual’s life, so that the term selfish loses its original connotation, and the
proposition that man is selfish resolves itself into the circular statement that people are concerned with the things they are concerned with.” Gardner Murphy, “Social
Motivation,” in G. Lindzey, ed., Handbook of Social Psychology (Reading, Mass.: AddisonWesley, 1954), 2: 625.
Page 164
2. … Only officials will concern themselves with public affairs, for only they are paid to do so. For a private citizen to take a serious interest in a public problem will be
regarded as abnormal and even improper.51
Yet there is nothing universal about “amoral familism.” For example, in the kibbutz in Israel studied by Spiro, the identification of a member with his family is weakened
both by his identification with the kibbutz as a whole and, for the sabra who has been reared in the kibbutz, with the group of sixteen children with whom one is reared
from kindergarten through high school, the kevutza. The strength of these emotional ties is indicated by the fact that for the sabras the incest taboo extends to the
kevutza and even to the whole kibbutz and by a wealth of evidence that the sabras have a strong sense of responsibility for the kibbutz and its members.52
If the “self” is somewhat elastic, one’s ‘‘interest, what one should do to serve the self,” is even more so. For one thing, conceptions of interest depend on cognitive
beliefs, and these are shaped by many different factors: because of cultural influences, for example, a person with a burst appendix might rub sheep fat on his belly or
call a doctor. Consequently, conceptions of selfinterest or group interest vary a good deal. Thus it is not particularly helpful to explain
51. Edward C. Banfield, The Moral Basis of a Backward Society (Glencoe: The Free Press, 1958), pp. 85–87. Italics in the original.
52. “In not one instance has a sabra from [the kibbutz studied by Spiro] married a fellow sabra nor, to the best of our knowledge, has a sabra had sexual intercourse
with a fellow sabra … In no instance have sabras from the same kevutza had sexual intercourse with each other. The reason given by the sabras for their exogamy is
interesting: they view each other, they say, as siblings.” Melford E. Spiro, Children of the Kibbutz (New York: Schocken Books, 1965), p. 347. See also his Kibbutz,
Venture in Utopia (New York: Schocken Books, 1963), p. 219. Although Spiro is critical of many aspects of the kibbutz and the discrepancy between ideal and
reality, he leaves no doubt about the strength of the sabra’s attachment to his kevutza and to the kibbutz. See, e.g., Children of the Kibbutz, p. 365.
Page 165
beliefs as expressions of selfinterest, for what is seen as selfinterest depends on one’s beliefs.53
Probably the commonest form of the explanation by selfinterest interprets beliefs as a reflection of an individual’s position in the society and economy. Innumerable
studies, both systematic and casual, reveal tendencies of this kind, but they also show that the correlation between an individual’s beliefs and his socioeconomic
characteristics is almost always quite weak. The significant finding is not that there exists some relationship, but how weak and uncertain the relationship proves to
be.54
It is tempting to think that if “selfinterest” and socioeconomic position are psychologically too naÏve to serve as explanatory factors, a more sophisticated
psychological explanation will suffice. Can the beliefs of a political activist be adequately explained as a reflection of his personality structure? One of the leading
investigators of the relations between politics and personality cautions against reductions of this kind, arguing that:
Personality structures ≠ political belief ≠ individual political action ≠ aggregate political structures and processes
where ≠ means “does not necessarily predict.” He goes on to say:
I do not mean to imply an absence of empirical connection across these various links … Rather, what needs to be emphasized is that the connections are empirical …
and that the relationships are neither
53. Cf. Clifford Geertz’s comment on the “interest theory’’ as a determinant of ideology: “The main defects of the interest theory are that its psychology is too anemic
and its sociology too muscular.” “Ideology as a Cultural System,” in Apter, ed., Ideology and Discontent, p. 53.
54. For example, see Donald D. Searing, “The Comparative Study of Elite Socialization,” Comparative Political Studies 1 (January 1969): 471–500, particularly 488–
89.
Page 166
necessarily strong nor positive, especially those that cross more than one of the links.
The weakness of the first of these links is that
As is often pointed out, persons with similar underlying personality characteristics are capable of holding different political beliefs, and those with similar beliefs may
differ in underlying personality characteristics. The psychic elements of underlying personality and belief are capable of independent variation.55
Political Culture as an Explanation
If a spirit of reductionism led social scientists for a time to play down the effects of beliefs on the character and behavior of different kinds of regimes, more recently a
focus on the importance of political culture has indirectly given renewed emphasis to the importance of political beliefs.
Political culture has been defined as “the system of empirical beliefs, expressive symbols, and values which defines the situation in which political action takes place. It
provides the subjective orientation to politics.”56 It is now rather widely recognized that differences in the political cultures of various countries help to account for the
differences in the nature of their political systems.
The relationship between political culture and beliefs, as I use the term here, is fuzzy. Certainly many of the beliefs held by political activists can be regarded as a part,
or a product, of a country’s political culture. But the focus in this chapter on the beliefs of political activists may be distinguished in two respects from the recent concern
for political culture. First, studies of political culture tend to concentrate exclusively on stable and persistent outlooks produced by socialization into an ongoing system
of beliefs; hence the
55. Fred I. Greenstein, Personality and Politics (Chicago: Markham, 1969), pp. 124–26.
56. Sidney Verba, in Pye and Verba, Political Culture, p. 513.
Page 167
investigation of political cultures may easily neglect sources of change in beliefs. Because alien, esoteric, or unorthodox beliefs in one period may become part of the
political culture at a later time, the processes that make for changes in beliefs are as important as those processes, like socialization, that make for stability. Second,
although students of political culture call attention to differences between elite and mass political cultures, they have mainly paid attention to the more widespread and
diffuse aspects of the widely shared political culture of a country; up to now, they have paid a good deal less attention to the beliefs of political elites and activists. As a
result, we know little about differences in the cognitive beliefs of activists in various countries.
Nonetheless, by taking advantage of the insights and evidence provided by the recently reviving interest in political culture, it may be possible to develop a better
understanding of some of the possible causes of both stability and change in the beliefs of political activists.
Acquiring Political Beliefs
Virtually all moral and religious training is founded—correctly, I think—on certain simple assumptions, to wit:
Most people acquire their beliefs during a period when they are particularly receptive. Typically, a person is highly receptive during, and only during, the first two
decades of his life. At the end of this period, one’s outlook becomes fixed or crystallized. Thereafter, one’s beliefs are likely to remain rather stable.
When social scientists concerned with political culture place such a heavy emphasis on the importance of early socialization, they tread the ancient path of Aristotle on
ethics and the churches on religion.57
57. “The Virtues then come to be in us neither by nature, nor in despite of nature, but we are furnished by nature with a capacity for receiving them, and are perfected
in them through custom … So
Page 168
It would be wrong, of course, to assume that political beliefs are fully formed and crystallized beyond change after the early period of youthful socialization is finished.
Changes may occur later, either gradually or, in the more infrequent case, abruptly. Ordinarily, it appears, when beliefs change in later life the change is gradual, as with
the classic transformation of the youthful radical to the middleaged conservative. A gradual change of this kind need not involve the kind of crisis one associates with “a
loss of belief.” Instead, one might say that new beliefs fill up the space that has been left by the erosion of the old. With a rapid change of beliefs in a mature person,
however, there is more likely to be an interlude between the old and the new when one suffers from a sense of having lost his beliefs. Loss of beliefs is, it seems,
accompanied by a painful state too unpleasant to sustain. Hence, loss of belief is likely to coincide with a period of high receptivity—even an active search for new
beliefs to replace the old. The urgency of the search insures that the period of receptivity is short. Ordinarily it is soon followed by the acquisition of new beliefs, which,
because the threat of loss is unbearable, may be even more rigidly held than the old.
Although a young person may also suffer a sudden loss of belief, typically the youthful period of receptivity seems to be a process of gradual or intermittent acquisition,
reality testing, relearning, and crystallization as the youth is initiated into the beliefs that prevail in the major institutions of socialization in his society or subculture—the
family, school, church, workplace, peer group. “Very young children in Britain first learn to think of the Queen as the effective ruler of Britain … To the young child the
Prime Minister is merely a helper of the Queen … This view of the rela
then, whether we are accustomed this way or that straight from childhood, makes not a small but an important difference, or rather I would say it makes all the
difference.’’ The Ethics of Aristotle (London: Dent, 1911), pp. 26–27
Page 169
tionship between the Queen and the Prime Minister erodes with increasing age, but it diminishes much more rapidly among middle class children than among working
class children.” The average adult Englishman, then, has acquired the cognitive belief that the Queen reigns but does not rule; the normative belief that, as one twelve
yearold put it, “the Prime Minister is a person who rightfully sort of governs the country and he’s got men under him” and quite possibly an affective or emotional
warmth toward the monarchy and neutrality visÀvis the Prime Minister.58
It may be useful therefore to distinguish two different kinds of periods of relatively high receptivity to the acquisition of beliefs: the normal period of youthful socialization
and the more abnormal situation that occurs when, having already acquired a belief, one loses it and feels a sense of loss. What factors influence the content of the
political beliefs an individual acquires during his periods of receptivity?
Exposure
One obvious determinant is the extent to which one is exposed to a particular political outlook, which depends in turn on a number of things. The chance of being
exposed to a particular idea depends, of course, on whether the idea is present in one’s own environment. Two necessary conditions are, clearly, that the idea has been
formulated and that it has been diffused to the individual’s environment. So much seems obvious enough.
Invention and diffusion provide a person with available models. Schemes of representation, for example, existed well before the eighteenth century, but the model of
representative government for the state did not. That model had still to be articulated—invented, if you like. After it was formulated during the eighteenth century, it
struck many sensitive
58. Fred I. Greenstein, V. M. Herman, Robert Stradling, and Elia Zurick, “Queen and Prime Minister—the Child's Eye View,” New Society 14 (October 23, 1969):
365–68.
Page 170
political observes—Jefferson and Madison in America, James Mill in England—as a brilliant new discovery that solved a hitherto unsolvable problem: how to combine
democracy with the large state. To take another example, the idea of a constitutional convention was wholly novel, as Palmer has pointed out, until the American
Revolution stimulated the calling of conventions to write or rewrite the various state constitutions and, of course, the Constitutional Convention of 1787 in Philadelphia.
What most impressed the French was the very act of constitutionmaking itself, the constituting or reconstituting of government through the principle of the people as a
constituent power. What they learned from America was the possibility of having a constituent assembly or convention. The very word “convention” in this sense …
came into the French language through translation of the American state constitutions. For the subjects of government to repudiate and dismantle their governments,
revert to a state of “nature,’’ and then by deliberate planning to constitute governments anew, to invent and delimit new offices and authorities and endow them with
written grants of power, was at least in a juridical sense the very essence of revolution, the practical acting out of the social contract, and the assertion of the sovereignty
of the people.59
The extent to which one is likely to be exposed to a particular belief during youth, the normal period of high receptivity, depends on the amount of influence that the
bearers of the belief have on the process of socialization, particularly in the family, church, school, and university.60 A notable
59. R. R. Palmer, The Age of the Democratic Revolution: A Political History of Europe and America, 1760–1800, The Challenge (Princeton: Princeton University
Press, 1959), p. 267.
60. “Just as what is taught is vital in the socialization effect of education, so is who teaches. In Chile, the BonillaSilvert survey of primary and secondary school
teachers indicated that their numbers
Page 171
instance was the extraordinary influence of English or Englishestablished schools, colleges, and universities on the political perspectives of many Indian intellectuals,
scholars, journalists, teachers, administrators, and political leaders who helped to gain acceptance for liberal democratic ideas and institutions before and after
independence.61 The effect of the English universities was amplified by their indirect influence on Indian universities. The perspectives of the Westernized, or
Anglicized, Indian were often profoundly at odds with the character and actual development of India. If, as a result, the net impact of English education on India’s
political development can never be precisely determined, surely it is no exaggeration to say that if Indian intellectuals and other leaders had been as deeply exposed
during their higher education to antidemocratic or illiberal ideas, the chances for polyarchy in India would have been significantly reduced.
Prestige
As this example suggests, the particular beliefs that an individual acquires during his periods of receptivity, whether in youth or later on, depend not only on the amount
of exposure but also on the relative prestige of the ideas and beliefs to which he is exposed. The prestige of a particular political belief seems to depend among other
things on the prestige of its advocates and antagonists and on the successes and failures of the people, organizations, and institutions that symbolize the belief. Like
Benjamin Franklin in Paris, the bearer of a belief (in this case a belief in the virtues and potentialities of representative government and political equality) may himself
acquire esteem as the representative
included more downwardmobile than upwardmobile people, and that these downwardmobile teachers’ attitudes rated very high on traditional values.” Alfred
Stepan, “Political Development Theory: The Latin American Experience,” Journal of International Affairs 20, no. 2 (1966): 232.
61. Weiner, “India,” pp. 239–40.
Page 172
of a new idea and at the same time enhance the esteem in which the idea itself is held because of his own special prestige—in Franklin’s case, as savant, natural genius,
wit, and friend of the great, the famous, and the fashionable. Gandhi’s prestige in recent decades has many sources other than his advocacy of nonviolent resistance, but
the prestige of that idea has doubtless been enhanced by its association with the saintlike Gandhi. So too with his martyred follower, Martin Luther King.
In lending prestige to a belief or detracting from it, intellectuals and scholars once again occupy a strategic position; to those who do not independently investigate the
validity of a belief, the intellectual, scholar, scientist, or other accredited expert often provides the needed validation. And it is, of course, the relative prestige of writers
and scholars within the educational system of a country that gives them a relatively high degree of influence over political socialization. In the United States, for example,
critics of the “Liberal Establishment” from both Right and Left contend that the content of the curriculum from kindergarten through the universities is far more “liberal”
than is the country as a whole. The charge is very likely true, for American intellectuals and scholars are not only more ‘‘liberal” than the general population, but they
have more influence on the content and practices of the educational institutions, though much less than many of them would like.
The relative prestige of an idea also depends on what are thought to be its achievements. These are embodied in people, institutions, or organizations. Although
systematic data on the relative prsetige of belief systems are lacking, some wellgrounded impressions may suffice to make the point. Palmer writes of the impact of the
American Revolution on Europeans:
There were many in Europe, as there were in America, who saw in the American Revolution a lesson and an
Page 173
encouragement for mankind. It proved that the liberal ideas of the Enlightenment be put into practice. It showed, or was assumed to show, that ideas of the rights of
man and the social contract, of liberty and equality, of responsible citizenship and popular sovereignty, of religious freedom, freedom of thought and speech, separation
of powers and deliberately contrived written constitutions, need not remain in the realm of speculation, among the writers of books, but could be made the actual fabric
of public life among real people, in this world, now.
Thus was created an American myth, or mirage, or dream, “the first of these great movements of secular mysticism,” to quote a recent author, “which modern man has
been experimenting with for the last two hundred years.”62
Later, as Tocqueville testified, the simple fact that democracy appeared to be working successfully in the United States weakened the arguments of antidemocrats.
Throughout the nineteenth century, the relative prestige of democratic beliefs seems to have been high throughout Europe and the Americas.
Because polyarchal institutions were historically associated with capitalism, the onset of the Great Depression doubtless reduced the prestige not only of capitalism but
of polyarchy as well. Although the apparent achievement of reformist governments in a number of countries with polyarchal regimes—the New Deal in the United
States, the reforms of Labor governments in Sweden and elsewhere—probably restored some of the prestige to polyarchy, the seeming triumphs of oneparty
hegemonic regimes in the USSR, Italy, and Germany in the 1930s undoubtedly enhanced the prestige of these regimes and the ideologies they purported to re
62. Age of the Democratic Revolution, pp. 239–40.
Page 174
flect.63 However incredible it may seem in retrospect, the USSR under Stalin was regarded by some distinguished Western intellectuals and writers of liberal, left, or
socialist persuasions as the embodiment of what Sidney and Beatrice Webb unhappily chose to call a New Civilization.64 If the defeat of the Axis powers in World
War II was a severe blow to the prestige of fascism and nazism, through its sacrifices and military victories the Soviet Union emerged in a heroic and creditable light,
while in Italy and France the dedicated and effective participation of communists in the resistance helped the party, particularly in Italy, to acquire very considerable
prestige. In Latin America, the failure of competitive political systems either to bring about structural reforms or to endure, the solid achievements of PRI in Mexico,
Castro’s revolution in Cuba, reformist juntas like that of Peru—all have had an impact on the relative prestige of polyarchy and hegemony. To offer a final example, the
involvement of the United States in the Vietnam war seems to have drastically lowered the prestige of polyarchy (though not necessarily of democratic beliefs) among
young people in the United States and probably elsewhere.65 In fact, this final example may serve to illustrate the general argument: a young person growing up in the
United States in the late 1960s in all likelihood would not only be exposed to beliefs different from those of a young person in, say, 1950 or 1900, but the relative
prestige of polyarchy, particularly as embodied in American institutions, would be lower. Hence the proportion of young people who believed in something other
63. For Argentina, see Mariano Grondona, La Argentina en el Tiempo y en el Mundo (Buenos Aires: Primera Plana, 1967), pp. 146–47.
64. Sidney and Beatrice Webb, Soviet Communism: A New Civilization? (London: Longmans, Green, 1935).
65. Kenneth Keniston, Young Radicals (New York: Harcourt, Brace and World, 1968), p. 123. JeanMarie Domenach, “L’idéologie du Mouvement,” Esprit, 36e
Annee (August–September 1968), pp. 39 ff.
Page 175
than polyarchy, at least as practiced in the United States, was probably greater than in certain earlier periods.
Consistency with previous beliefs
A third factor determining the chances that an individual will acquire a particular political belief is the extent to which he feels it to be consistent with his present
beliefs.66 To be sure, even if a new idea is inconsistent with an already accepted belief, if one finds the new belief attractive he may reject the old or conclude that the
new and the old are not really inconsistent. Nonetheless, if the new view is thought to conflict with one’s existing beliefs, obviously the chance that one will adopt it is
reduced. This is what happens, in fact, when a period of high receptivity comes to an end and one emerges with a more or less crystallized set of beliefs. Thereafter,
one consciously or unconsciously tests new beliefs for consistency with the old and, ordinarily, rejects the new ones when they are felt to be inconsistent. When new
ideas conflict with old, firmly held beliefs, an enormous weight of contradictory perceptions is required in order to bring about a loss of belief; and it sometimes appears
as if the level of contradiction behind the dam of old beliefs had been secretly rising, until finally the dam itself is swept away and nothing remains except disbelief. It is
easy to see, too, why the general culture and the specific political culture into which an individual is socialized are so important in explaining political life, as social
scientists have recently reemphasized. For the culture consists of those very beliefs, perspectives, and habits of mind that are successfully transmitted to the young and
therefore provide the conscious and unconscious assumptions against which new beliefs are tested—and usually found wanting.
It is this need for subjective consistency in beliefs that
66. Robert E. Lane and David O. Sears, Public Opinion (Englewood Cliffs: PrenticeHall, 1964), pp. 44 ff.
Page 176
helps to give ideas a certain life of their own, an autonomous development not wholly under the control or manipulation of social interests. Although there is
unquestionably more determinism in social and political beliefs than in highly abstract belief systems like those of mathematics and physics, once a social or political
principle becomes firmly embedded in the political culture, it can serve quite literally as an axiom from which new unanticipated, and seemingly inescapable conclusions
can be drawn. The “functions” originally fulfilled by a principle, the ‘‘interests” initially served by it, can no longer control or dominate the conclusions to be drawn from
it.
Beliefs about political equality and inequality often take on this axiomatic quality. Wherever principles of political equality were used by spokesmen for the middle
classes to justify their entry into the political system, it was not long before spokesmen for excluded groups drew the logical consequences, showed the inconsistency of
the exclusion, and insisted that middleclass governments had to honor their own sacred principles or lose their legitimacy. In fact, certain clusters of political belief so
regularly show up in different circumstances that they appear almost to reflect the necessity with which certain conclusions follow from certain premises. To believe in
political equality among a large association of people seems to carry with it, as if by implication, a number of principles; to make the same point more cautiously, the
attempt to devise principles consistent with political equality narrows the choice down to a very much smaller subset.67
6. It is difficult to believe, for example, that democratic ideas were simply invented once at a given time and place and that subsequent repetition is nothing more than a
result of diffusion. Consider the following case: In 1781 “a reforming committee at Westminster … drew up a report that went beyond any of the new American state
constitutions in its democratic theory of representation. The report, drafted by John Jebb, demanded universal manhood suffrage, the use of the ballot rather than oral
voting, the annual election of Parliament, representation of voters in proportion to numbers in equal electoral districts, payment of wages to elected representatives, and
Page 177
Consistency with experience
“Internal” consistency, whether according to familiar principles of logic or by more subjective and personal criteria, is not, of course, the only kind of consistency one
requires. A fourth factor bearing on the likelihood that one will adopt a particular political belief is the extent to which one feels that the belief is consistent with one’s
own experiences. Experience is particularly important for the credibility of the cognitive aspects of belief, for if one’s own perceptions of reality conflict with those
maintained in the new belief to which one is exposed, then in order to reduce the strain one must reject either one’s own perceptions, the existence of a discrepancy, or
the validity of the new idea. While each of these evidently occurs, the last is without doubt the most common. Here again, it is obvious how important is the culture in
which the individual is reared, for he is conditioned to interpret the events of his life in a particular way. What one experiences is partly shaped by one’s own culture;
thus people from different cultures may experience the “same” event in quite different ways. So too, people with different political perspectives—whether weakly
organized, fragmented belief systems or coherent ideologies—may perceive the same happening in very different ways, and thus each person may accumulate
experiences that confirm his existing beliefs.68
removal of all property qualifications for election to the Commons. Here were all six points of the People’s Charter to be famous in England over fifty years later.”
Palmer, Age of the Democratic Revolution, pp. 208–09.
68. The phenomenon is, of course, familiar to psychoanalysts, psychiatrists, and psychologists. “All day long we unconsciously selectively perceive the world about us.
Man can prevent unpleasant perceptions by varying his attention and by wishful perceiving or thinking. … The simplest form of denial is blotting out the perception by
withdrawing our attention from it … People frequently ‘fill in the blanks’ with wishfulfilling perceptions.” Irving L. Janis, George F. Mahl, Jerome Kagan, and Robert
R. Holt, Personality Dynamics, Development, and Assessment (New York: Harcourt, Brace and World, 1969), p. 354. Italics in the original.
Page 178
Nonetheless, except for psychotics and to a lesser extent young children, selective perception is limited by reality.69 Among the views to which he is exposed, the
youth selects those that seem to conform most closely with his own experiences. In spite of selective perception, this process continues throughout life, and dramatic
discrepancies between personal experiences and beliefs can lead to a change in beliefs. Personal experiences have led to the collapse of armies: the Russian in 1917,
French in 1940, German in 1945. German soldiers on the Western front surrendered individually, then in small groups, and finally in vast straggling droves throughout
April 1945 rejected the desperate propaganda from Berlin urging them to fight on, for they were convinced by battlefield experience that further resistance was futile.
Yet only a few months earlier, many of the same soldiers had been fighting doggedly. During the 1960s in the United States, a significant number of young people, after
experiencing what Keniston has called a “confrontation with inequity,” lost their confidence in liberal reformist politics and adopted more radical views:
Whether it was working with the unemployed in the inner city, in voterregistration drives in the South, with Negro families in the slums, or in a detailed study in
American policies in Vietnam, these young radicals were forced into an immediate personal confrontation with the injustices of American life and policies.70
Some experiences are widely shared, others are idiosyncratic. Differences in historical experience are one source of
69. ‘‘The price of this defensive gain [denial] can be high: an inability to live in the real world. But denial usually stops short of the hallucinatory extreme of psychosis: in
the adult, wishful thinking and even the tendency to it are usually restricted by being tested against reality. Not so with the child, however, whose behavior conforms
more to the pleasure principle than to the reality principle.” Ibid., p. 354.
70. Keniston, Young Radicals, p. 127.
Page 179
the famous generation gap, since those whose outlook is shaped by a special historical experience find it increasingly difficult with the passage of time to convey to
oncoming generations the meaning and “reality” of that experience. Even if “old soldiers never die,” they often lose their audiences. It has often been observed that by
the late 1960s the ‘‘meaning” and the risk of totalitarian rule began to seem very different to people under thirty in Western Europe and North America and to older
persons, particularly Europeans, who had lived through the high period of totalitarian triumph of the 1930s and to whom the risks of a country regressing to barbaric
despotism would always remain great. To many of these older people, polyarchy seemed both more fragile and more precious than it did to younger people who had
never experienced extreme despotism but only polyarchy, with all its shortcomings when measured against democratic ideals.
Accidental or idosyncratic experiences may also play a decisive role. When Lenin was seventeen, his older brother, Alexander, was hanged for taking part in a plot by
university students to kill the tsar. According to his sister, Lenin was “hardened” by Alexander’s execution and began thinking seriously about revolution.71 In trying to
account for the development of Gandhi and his doctrine of Satyagraha (literally “truth force,” variously translated as nonviolent resistance or militant nonviolence),
Erikson places special emphasis on Gandhi’s experiences in South Africa and a strike in Ahmedabad in 1918.72 The way an American president handles his job—his
“style”—is, according to Barber, “a reflection of the ways of performing which brought him success at the time, usually in late adolescence or early adulthood, when he
emerged as a personality distinctive from his
71. Edmund Wilson, To the Finland Station (New York: Harcourt, Brace, 1940), p. 361.
72. Erik H. Erikson, Gandhi’s Truth (New York: W. W. Norton, 1969).
Page 180
family heritage, in a role involving relatively intensive participation in a socially organized setting.” The early “success” that triggers the development of the later style (and
the career that leads to the presidency) is often rather accidental.73 In his interviews with young people who participated in the “Vietnam Summer” in 1967, Keniston
found it necessary to emphasize not only the “psychological meaningfulness of the development of radicals” but also ‘‘the importance of what, from a psychological
point of view, appear to be ‘accidents.’”
The development of these young radicals was often profoundly affected by things that “happened” to them through no plan, motive, or design of their own: a major
family illness; the psychological problems of brothers or sisters; a chance to move to another school in another city. No one can weigh exactly the impact of such
events upon a developing individual: much depends on his stage of development and his sensitivities at the time. In several instances, for example, major family
upheavals in early adolescence deprived young radicals of parental attention, forcing them to become independent long before most adolescents do. In another case,
the “accident” of going to a private school reinforced psychological tendencies that might not otherwise have been strengthened.74
Another Paradigm
I have been emphasizing some factors—exposure, prestige, consistency with previous beliefs and experience—that help determine whether a particular individual will
acquire a particular belief during his particular period (or periods) of
73. James David Barber, “Classifying and Predicting Presidential Styles: Two Weak Presidents,” Journal of Social Issues 24 (July 1968): 52 and passim.
74. Young Radicals, p. 226.
Page 181
receptivity. But these same factors may also be conceived of as influencing a number of people during the same period of time; thus they help to produce those broad
and decisive changes in outlook that historians describe when they write about the Renaissance, the Enlightenment, or the growth of democratic ideas.
One can hardly exaggerate how badly off we are as we move into this terrain. If it is difficult to account satisfactorily for the acquisition of individual beliefs, it is even
more difficult to account for historical shifts in beliefs. Theories about individual beliefs explain, at best, why some people hold certain beliefs some of the time; they do
not account for anything like all individuals or all beliefs. Yet historical explanations leave even more of the phenomena unexplained. If the data are inadequate for
testing theories about how individuals acquire their beliefs, the data and even the concepts needed for explaining historical changes are far worse. One can identify an
individual, and it makes some sense to say, for example, that “Robinson acquired democratic beliefs as an adolescent.” But what is the equivalent of Robinson when we
speak of historical changes? Concepts like “European society” or “the American people’’ are much too broad or ill defined. Clearly one is referring to some subset of
persons, but the subset is extremely difficult to specify with much precision. Frequently what one seems to have in mind are those people in a country, or in some set of
countries, who make up what might be called the ideational elites, the creators and special bearers of ideas—intellectuals of various sorts, philosophers, poets,
ideologues, pamphleteers, publicists, journalists, scientists, and so on. One may also have in mind the political elites, whose beliefs are of exceptional importance in
political life because of the exceptional influence of their bearers. But the boundaries of these elites are imprecise, and we lack information about their members. It is not
surprising, then, that because of their concern with rigor and their dissatisfaction with the “softness” of historical
Page 182
description, generalization, and explanation, most social scientists have turned away from the historical movement of ideas. As a result, their own theories, however
“rigorous” they may be, leave out an important explanatory variable and often lead to naïve reductionism, as if one tried to explain the collective behavior of the players
in a game of football by attributing it mainly to each player’s toilet training in infancy.
In a rough way, the elements discussed in the preceding section can be used in trying to account for changes in the frequency or intensity of a particular belief among a
group of political activists. In a country, even in a world region such as Europe, periods of relative stability in certain political beliefs give way to periods of instability,
breakdown, receptivity, which are then followed by crystallization of new beliefs and, for a time, relatively stable beliefs once again. Thus Paulson writes of the Italian
village near Rome that he calls Castelfuoco:
At the end of World War II the people of Castelfuoco entered a period of popular ferment never before equalled. Many hoped that the war had finally crushed
the old order. The aristocrats faced political death and final economic collapse. The old rural civilization … approached extinction. The crudities and coercion of
fascism had been unmasked before the world.
Into this moment of intense search for a society more just, more clean, looking more to the future than to the past, the communists stepped vigorously …
Along with a new classe dirigente, the communists offered a meaningful commitment for those who, in the postWorld War II period, faced the shock of deep
spiritual crisis. The crisis had been maturing for a generation but suddenly, with the collapse of fascism and the quest for a new order, people had no place to turn
…
The communists … have found a commitment in
Page 183
the ideal of a better society. This commitment is contagious and inevitably makes its holders “missionaries” … In the postwar spiritual and political vacuum, facing
an opposition more or less tied to the status quo, they have an immense power to penetrate the minds of the masses.75
However, it would be misleading to insist too narrowly on the analogy with individuals. Although an aggregate of individuals may be thought of for some purposes as if it
were a single actor, some important differences must be kept in mind. Even during periods of relative stability in ideas, some activists and members of the ideational
elites are receptive to new ideas. Conversely, during periods of receptivity, not all activists are open to the new belief—and certainly not the whole population. Even in
Castelfuoco, which has one of the highest percentages of Communist votes in Italy, only about half the votes go to the Communist party; from a third to a quarter go to
the Christian Democrats, a tenth to Socialists, and the rest are scattered among other parties.76 Hence crystallization does not necessarily lead to a single dominant
view, as one would expect with the individual actor. On the contrary, conflicting beliefs may crystallize as with religion during the Reformation or the political cleavages
between communists and noncommunists that hardened in Italy and France after 1945. A period of “stability” in beliefs may not entail political stability or tranquillity; it
may mean religious wars, ideological conflict, violence.
Moreover, individual beliefs may change without affecting the aggregate distribution, which seems to be approximately the case in some countries, at least in the short
run, with party loyalties. Individuals, in fact, may die and be replaced;
75. Belden Paulson with Athos Ricci, The Searchers: Conflict and Communism in an Italian Town (Chicago: Quadrangle Books, 1966), pp. 321–26.
76. Ibid., p. 112.
Page 184
because of the processes of socialization, when beliefs are embodied in an institution, like a church or political party, the distribution and even the relative intensity of
beliefs may remain substantially unchanged in spite of individual turnover. As Rokkan and Lipset point out:
The party systems of the 1960s reflect, with few but significant exceptions, the cleavage structures of the 1920’s. This is a crucial characteristic of Western competitive
politics in the age of “high mass consumption”: the party alternatives and in remarkably many cases the party organizations, are older than the majorities of the national
electorates. To most of the citizens of the West the currently active parties have been part of the political landscape since their childhood.77
This crystallization into beliefs that with the passage of time are increasingly out of joint with perceptions of reality prepares the way for a loss of faith among the mature
and receptivity to new alternatives among the young.
Finally, because people die and are replaced by young persons who may have different beliefs, the distribution of beliefs among the politically active or the ideational
elites may change even if every individual keeps the beliefs he acquires early in life. In fact, it may be this process of replacing the old cohorts by new ones that accounts
for most largescale historical change in beliefs. The shift in the United States during the 1930s from a normally Republican to a normally Democratic majority seems to
have been of this kind.78 If, as in Castelfuoco, the youthful cohorts enter with new views at the same time as the aging cohorts suffer a loss of belief,
77. Seymour M. Lipset and Stein Rokkan, ‘‘Cleavage Structures, Party Systems, and Voter Alignments: An Introduction,” in Lipset and Rokkan, eds., Party Systems
and Voter Alignments (New York: The Free Press of Glencoe, 1967), p. 50. Italics in original.
78. Angus Campbell, Philip E. Converse, Warren E. Miller, and Donald E. Stokes, The American Voter (New York: Wiley, 1960), pp. 153–54.
Page 185
then the total amount of change is, of course, greatly magnified. Indeed, each may reinforce the other: weakening belief among the old encourages new beliefs among
the young, which in turn leads to loss of belief among the old. Despite a common romantic view that the process always represents a healthy rejuvenation, it is in fact
morally and politically neutral. If it strengthened democratic as against aristocratic views in the United States during the Jeffersonian period (as seems likely), the
process may also have weakened beliefs in democracy in Athens between Demosthenes and Pericles. And some such process very likely took place in Germany
between the First World War and Hitler’s seizure of power.79
Thus, while the notion of receptivity, crystallization, and stability in beliefs can be applied to aggregates, the analogy with individuals must be treated with caution.
Nonetheless, whether we deal with an individual or an aggregate of individuals, the chance that the actor will acquire a particular belief during a period of receptivity
seems to depend on:
1. The amount to which the actor is exposed to the belief, which in turn
1. requires that the belief has been formulated and diffused to the actor’s environment; and
2. depends on the amount of influence that the bearers of the belief exert on processes of socialization.
2. The relative prestige of the belief, which depends on
1. the personal prestige of its advocates and antagonists; and
2. the successes and failures of the people, organizations, and institutions that symbolize the belief.
3. The extent to which the new belief is consistent with the actor’s perceptions of reality as these are shaped by
79. Abel, The Nazi Movement, pp. 172–74.
Page 186
1. the actor’s present beliefs; and
2. the actor’s experiences.
This paradigm suggests several observations.
First, all of the conditions described in previous chapters that affect the chances for hegemony, public contestation, and polyarchy must do so by influencing beliefs via
the connections listed above. Thus the historical path to the present taken by a country helps to determine the successes and failures that symbolize a belief, the relative
prestige of that belief, and so the chance that it will be acquired by an individual who is exposed to it during a period of receptivity. Again, in a country with several
distinct subcultures, members of a particular subculture acquire beliefs and experiences that help to shape their perceptions of reality and thus the acceptability of
additional beliefs to which they may be exposed.
Second, while the conditions discussed in earlier chapters partly determine the extent to which activists are receptive to beliefs affirming the relative desirability of public
contestation and participation, they do not fully determine the content of such beliefs. For example, no satisfactory explanation for polyarchy in India is possible, I think,
that fails to take into account the peculiar impact of AngloAmerican political ideas on Indian political elites during the critical period of national development preceding
and following independence, an impact that came in part from the salient role of English norms in the political socialization of key leaders. Or again, the triumph of
nazism in Germany is hard to explain without taking into account the independent effect of nazi ideology. Obviously there were other factors, but there is force in the
argument of Theodore Abel that the importance of ideology “as a determinant is at least equal to that of the general discontent which existed at the time.”80
Third, there are accidents and uncertainties both in the acquisition of beliefs and in the acquisition of power. To
80. Ibid., p. 174.
Page 187
gether, they enormously compound the difficulties of explanation. The preceding analysis reveals some of the sources of uncertainties in the process of acquiring beliefs.
As to the acquisition of power, that process is, except to an insistent determinist, obviously full of accidental and uncertain elements.
When we say that an event is an “accident” we do not necessarily imply that it is without a cause. It does mean, however, that the theory with which we operate does
not enable us to predict or explain such events. Our theory can only take them into account after they occur: it cannot subsume them under some more general law,
hypothesis, or conjecture. To be sure, some kinds of accidents are subject to fairly accurate statistical prediction; they are, in this sense, “lawful.” But many accidents
are not: assassinations, for example. But even where statistical predictions are reasonably accurate, they may furnish little or no help to a theory that seeks to account
for certain kinds of historical happenings in which ‘‘accident” plays a salient role.
Consider again the case of Lenin. Unless one assumes that his arrival in Russia in April 1917 had no effect on history, then history was subject to chance. What if the
German government had not wanted him to return? Or if his train had been blown up? These accidental events—accidental because no theory makes them lawful,
predictable events—would then have changed the course of history by whatever amount Lenin’s presence in Russia after April 1917 changed the course of history.
One must be a fanatical determinist, I think, to insist that there were no accidental elements in Lenin’s acquisition of power.
Aside from the accidents (or nonaccidents) that affect the chances that a particular individual will gain power, highly uncertain trains of events may affect the chances for
a whole regime. For example, unless one assumes that the course and outcome of the Second World War were completely determined, one must reckon with the
possibility that
Page 188
polyarchal regimes would not have been imposed in Italy, Austria, Germany, and Japan, nor hegemonic regimes in Poland, Czechoslovakia, Hungary, and elsewhere.
Suppose that the Axis powers had won the war? Or secured a negotiated peace? Or that the Soviet armies had not occupied Czechoslovakia?
When I try to take into account the complexities, uncertainties, and gaps in our understanding of the processes by which the politically influential acquire their beliefs and
gain power, I am driven to two conclusions. Neither will seem the least bit surprising to anyone who already takes for granted the importance of “ideas” in history. Yet
both conclusions run directly counter to the reductionism that prevails in much of contemporary social science:
At present and for an indefinite future no explanatory theory can account satisfactorily for the beliefs of political activists and leaders.
Consequently, just as any theory that attempts to account for variations in regimes in different countries must take as major independent variables such factors as the
socioeconomic level of a country, the nature and extent of inequalities, the extent of subcultural cleavages, and other factors discussed in previous chapters, so too
must such a theory, as a practical matter, treat the beliefs and “ideas” of political activists as a major independent variable.
Page 189
9. FOREIGN CONTROL
The destiny of a country is never wholly in the hands of its own people. In some cases, domination imposed by people from outside the country can be so decisive as to
override the effects of all the other conditions that have been discussed so far.
Every country exists in an environment that contains other countries. Under every regime, policy makers must take some account of the possible actions and reactions
of policy makers in other states. In this sense even the most powerful states are in some degree limited by the influence, control, or power of other states. Moreover,
most countries participate to some extent in a multinational economy; consequently policy makers usually take into account the actions and reactions of people outside
their own country who may affect the local economy. Countries heavily dependent on international trade or foreign investments—often small countries—are particularly
subject to the actions of foreigners.
Because the influence of the international environment on the development of hegemony or polyarchy is extremely complex, I wish to deal here with only one form of
external influence, foreign domination. To distinguish that form more sharply, let me specify three ways in which the actions, reactions, or expected actions of foreigners
may affect the chances of polyarchy or hegemony in a particular country.
Page 190
First, the actions of foreigners may and almost certainly will have some impact on one or more of the conditions discussed in previous chapters. The beliefs of political
activists, paths to the present, the level of socioeconomic development, the degree of economic concentration or dispersion, inequalities, even the extent of subcultural
cleavages—all are open to influences by foreign actors.
Second, the actions of foreigners may drastically alter the options available to a regime without necessarily altering the form of the regime. As I have already said,
existence in an international environment alters and reduces the options available to any regime in every country. However enormous is the might of world powers like
the United States and the Soviet Union, neither can escape the consequences of actions by the other. For that matter, the power of these superpowers is limited not
only by one another but by a great many other actors on the international scene. The day’s news will invariably furnish any number of examples: the inability of the
United States to defeat the forces of the National Liberation Front in Vietnam or of the Soviet Union to drive the Western powers out of Berlin, the difficulty of the
major powers in agreeing on a plan for controlling nuclear weapons, etc., etc.
Nor are the effects of foreigners confined only to foreign affairs. In making decisions about domestic economic policies, American policy makers, for example, have
had to consider the deficit in the balance of payments, the role of the dollar as an international currency, the impact on the American economy of currency devaluations,
actual or threatened, by other countries, and so forth.
If the options of great powers are sometimes reduced by the actions of foreigners over whom they have little or no direct control, the situation is even more acute for
countries with more limited power resources, such as small or less developed countries, particularly small and less developed countries. Thus where the postwar Labor
government in
Page 191
Britain did not shrink from nationalizing several major industries out of fears about the impact of nationalization on foreign investment, President Frei’s Christan
Democratic government in Chile had to consider the very real possibility that nationalization of the copper industry might drastically reduce foreign investments and
access to world markets.
To the extent that the actions of foreigners reduce the options available to a country, the people lose the capacity to govern themselves. Nonetheless, a polyarchal
regime might exist in a country whose options were narrowly constricted by the actions of foreigners. Indeed, internally the political system might be highly egalitarian,
consensual, highly participatory, strongly protective of personal and political liberties, and thus highly “democratic” by the usual criteria. Yet the scope of “democracy”
might be very narrow simply because so few alternatives were available. One thinks of the difference between a local government in a polyarchy and the Greek
conception of the autonomous citystate: the political system of a small town in a modern country governed by a polyarchal regime may itself be a polyarchy, may even
be highly “democratic,’’ but the town has very little autonomy. The form of polyarchy is unchanged, but its content is drastically altered.
Third, people in one country may deliberately seek to use their resources to impose a particular kind of political regime on another country: outright foreign domination.
They will probably affect the content of politics, too, but the point I wish to emphasize here is that by their actions foreigners may massively affect the chances for
hegemony or polyarchy, pretty much independent from any of the conditions discussed in the previous chapters.
Each of these kinds of external influence would be an appropriate subject for an entire volume. Consider the first. To unravel all the ways in which the international
environment significantly influences the conditions described in previous chapters would obviously be a major task, and I do not intend
Page 192
to undertake it here. The second kind of international influence is of great importance in the contemporary world, particularly for small countries. With increasing
international integration, the Scandinavian countries, for example, will lose more and more of their autonomy. One can even foresee the time when they may be local or
regional governments in a larger political system. But this topic is also too complex for adequate treatment in this book.
The third is probably the most directly relevant to the purposes of this book. For in estimating the chances for hegemony or polyarchy, it would obviously be important
to know whether foreign powers were attempting to impose or intended to impose a particular regime. In the period immediately following the Second World War, a
large number of European regimes were to some extent imposed: hegemonies in Eastern Europe, polyarchies in Germany, Austria, and Italy. A purely autonomous
political development over a long period has never been possible in Czechoslovakia; local circumstances favorable to a competitive political regime have too often been
overridden by outside domination. Much the same has been true in Poland.
The interaction between local conditions and foreign domination is complex. How is it that Finland is a polyarchy? A part of the Swedish realm for six centuries, a part
of the Russian Empire (the Grand Duchy of Finland) from the Napoleonic Wars to the Russian Revolution, twice at war with the Soviet Union, first in 1939–40 and
then from 1941 to 1944, twice defeated, a cobelligerent with Germany in the Second World War, how is it that this small nation of five million people exists as a
polyarchy alongside a country with fifty times its population, a superpower ruled by a hegemonic regime? If the Russians had been prepared to pay the costs it hardly
seems open to doubt that they could have imposed a hegemonic regime on Finland in 1944–45. If in the Soviet view the expected costs exceeded the expected gains,
this was not only because the gains were lower than in Poland and
Page 193
Czechoslovakia but perhaps partly because the Finns, long accustomed to living under a liberalized regime, could be counted on to make the costs high over the long
run.
These examples suggest that concrete historical statements, or predictions based upon a particular configuration of international forces at a specific time, may be more
fruitful than theoretical generalities about the interplay between foreign domination and polyarchy. Nonetheless, if we focus on overt rather than covert domination, and
if we confine our attention exclusively to relatively direct consequences for polyarchy (and thus ruthlessly exclude from consideration a number of consequences that
would be important from other points of view), it is possible to advance a few general statements that seem to be supported by the available evidence.
In the first place, a high proportion of the countries in which polyarchy existed in 1970 had been occupied or otherwise subject to foreign military intervention at least
once since achieving independence. As table 9.1 indicates, among the older countries more polyarchies than nonpolyarchies have been subject to overt foreign military
domination. As an inspection of table 9.2 will quickly reveal, the reason, of course, is that most European countries that were polyarchies in 1970 had been overrun and
occupied as a result of the Second World War (one of these, Belgium, had also suffered the same fate during the First). However, it should also be kept in mind that
four of these countries, though governed by polyarchies as of 1970, had been ruled by hegemonic regimes at the time of capitulation.
Second, it does not appear to be true, as is sometimes thought, that a period of overt foreign domination by a hegemonic power inevitably wreaks irreparable damage
on a polyarchy. As tables 9.1 and 9.2 indicate, the damage certainly need not be fatal. In fact, a period of foreign domination may strengthen national unity, foster a
climate of reconciliation among hostile groups, and speed the incorporation of strata struggling for more recognition and power.
Page 194
Table 9.1. Occupation and Military Intervention,
Polyarchies and Nonpolyarchies
Political regime in 1970
Independence gained: Polyarchy Nearpolyarchy Nonpolyarchy Total

Before 1900
Occupied or subject to military intervention at least once since
independence 8 1 6 15
Never occupied or subject to military intervention since
independence 7 3 22 32

1900–1945
Occupied, etc. 3 — 5 8
Never occupied, etc. 5 — 7 12
Other

After 1945
Occupied, etc., 1 1 51
59
Never occupied, etc. 5 1
Total N 126
Note: The basic data on independence and occupations are from Bruce M. Russett, J. David Singer, and Melvin Small, “National Political Units in the Twentieth
Century: A Standardized List,” American Political Science Review 62 (September 1968): 932–51. Under their classification “Dependent,” they omit some overt
military interventions of short duration (e.g., the USSR in Romania in 1944, the U.S. in Lebanon in 1958 and the Dominican Republic in 1965) and certain special
cases (e.g., Italy, 1943–46) that I classify as occupation or military intervention. The classification of polyarchies, nearpolyarchies, and nonpolyarchies is from
appendix table A3. A few microstates are omitted.
In Belgium, the German invasion and occupation during the First World War brought Socialists into the government for the first time; they remained even after the war
was over. The war also led directly to universal suffrage in Belgium. Under the old system of plural voting, as Lorwin writes,
Workers and peasants who had suffered in the trenches or in deportation could have cast one ballot each, while
Page 195
Table 9.2. Independence among Present Polyarchies
and NearPolyarchies
Full Near
polyarchies polyarchies
I. Independent before 1900
A. Independence never interrupted by occupation or military intervention Chile Colombia
Costa Rica Turkey
Sweden Venezuela
Switzerland
United Kingdom
United States
Uruguay
B. Occupied or military intervention Belgium Dominican Republic
Denmark
France
West Germany
Italy
Japan
Luxembourg
Netherlands
II. Independence gained 1900–1945
A. Independence never interrupted Australia
Canada
Iceland
Ireland
New Zealand
B. Occupied or military intervention Austria
Finland
Norway
III. Independent 1945 or later
A. Independence never interrupted Jamaica Malaysia
Trinidad
Israel
India
Philippines
B. Military intervention Lebanon* Cyprus
* At request of own government.
war profiteers could have cast two or three. Instead, the King and the government called the voters to the polls on a one man, one vote basis. The parliament they
elected legitimized the change by constitutional amendment.
Page 196
Lorwin goes on to describe the impact of the Second World War:
If the First War marked the full acceptance of the Socialist Party, the Second marked that of the labor unions. Under the Nazi occupation, clandestine personal
contacts among leaders of industry and Catholic, Socialist, and Liberal trade unions produced a "pact of social solidarity" of symbolic and practical importance. The
pact was implemented after the Liberation by wide advances in social legislation and collective bargaining … Union and industry representatives came to exercise
powers of administration in a number of quasipublic social welfare agencies.1
In the Netherlands, the first Socialist ministers came into the government in 1939 under the threat of nazism, and in that country, Belgium, and Norway, invasion and
occupation also produced allparty governmentsinexile.
Needless to say, it would be utterly fatuous to draw the conclusion that overt foreign intervention is a good thing for a polyarchy. Even if one were to put to one side all
the other costs, which have sometimes been enormous, not only is polyarchy superseded temporarily by hegemonic rule, but the longerrun consequences are by no
means all beneficial. Some cleavages may be intensified. Lorwin writes that in Belgium during the Second World War “the Germans encouraged Flemish separatism,
leaving a train of political booby traps for the Belgian nation after the war.”2 If reconciliation and national unity may sometimes facilitate the acceptance of hitherto
excluded strata and greater generosity toward their demands, opposition leaders may be so fully drawn into coalition building that they unduly soften their
1. Val R. Lorwin, “Belgium: Religion, Class and Language in National Politics,’’ in Robert A. Dahl, ed., Political Oppositions in Western Democracies (New Haven:
Yale University Press, 1966), pp. 158, 165.
2. Ibid., p. 161.
Page 197
demands, leaving a legacy of relative inequalities and resentments as a cause of later crises.
Despite these qualifications one must not lose sight of the simple proposition that overt foreign intervention is not necessarily fatal to an already existing polyarchy and
may actually strengthen it in some respects.
In the third place, it is patently untrue that polyarchy has
Table 9.3. Polyarchy, Dependence, and Independence:
29 Present Polyarchies
Polyarchy or nearpolyarchy inaugurated Polyarchy inaugurated after independence but during Polyarchy inaugurated “autonomously”: after independence and not
before independence overt foreign domination during overt foreign domination
Australia Austria Belgium
Canada West Germany Costa Rica
Iceland Italy Chile
India Japan Denmark
Jamaica France
Lebanon Israel
New Zealand Luxembourg
Norway Netherlands
Trinidad and Tobago Sweden
Philippines Switzerland
United Kingdom
Uruguay
Special cases:
Finland
Ireland
United States
come about only through strictly autonomous developments within an already independent country. As table 9.3 indicates, of the 29 countries with polyarchal regimes in
1970, in only 12 was polyarchy inaugurated after independence and not during a period of overt foreign domination. In 4 countries, as we have seen, polyarchy was
inaugurated during military occupation or intervention after the Second
Page 198
World War. In addition, in 10 countries, including some where polyarchy is now very deeply rooted, substantial strides toward polyarchy were taken while the country
was at least nominally subject to a foreign power.
Fourth, foreign domination frequently produces a boomerang effect. There can hardly be any doubt that an important consequence of nazism among countries where
polyarchy was temporarily set aside by the occupiers was temporarily to strengthen attachments to democratic ideas and to increase hostility to the antidemocratic
ideology of nazism. There is good reason to think that the Soviet occupation of Czechoslovakia has made many Czechs hostile to orthodox Soviet Marxism and
enhanced the attractiveness of at least some aspects—especially the libertarian features—of polyarchy. The boomerang effect may radiate far beyond the victims of
intervention. The actions of the United States in Guatemala, Cuba, Vietnam, and the Dominican Republic, together with a widespread perception of the U.S. as
economically aggressive and domineering, has probably helped in many parts of the world to discredit the institutions of polyarchy and belief in its capacity for social
justice.
Fifth, the circumstances in which it was possible for polyarchy to be inaugurated during a period of overt dependence (or to be preserved for subsequent revival) were
historically unusual and seem much less likely to reoccur in the foreseeable future. To see why this is so, consider the rather schematic set of possibilities in table 9.4.
The first situation is roughly comparable to that of Belgium, Denmark, the Netherlands, Norway, and France un
Table 9.4. Some Circumstances of Foreign Domination
In the country subject to domination:
The previous regime was: The conditions favor: The dominant country seeks to inaugurate:
1. Polyarchy Polyarchy Hegemony
2. Polyarchy Hegemony Hegemony
3. Hegemony Polyarchy Polyarchy
4. Hegemony Hegemony Polyarchy
Page 199
der nazi domination. An important aspect of that domination was its comparatively short duration and limited scope. Most of the political leadership survived. Among
the general population, the boomerang effect was powerful: nazism came to be widely loathed; hegemony was the rule of the enemy and occupier, the cause of defeat,
privation, humiliation, suffering.
If there is good reason for thinking that in a country where it has existed for some time and the underlying conditions are favorable to it, polyarchy is likely to be rapidly
restored after a relatively brief period of overt domination, the outcome is bound to be more uncertain if the period is long and the foreign power systematically wipes
out everyone who holds democratic beliefs.
In the second situation envisaged in table 9.4, the restoration of polyarchy would be very much less likely. The foreign power would find it easier not only to impose a
hegemonic regime but also to withdraw in due time from overt domination, leaving behind a hegemonic ally. By going native, the hegemony would accommodate itself
to the spirit of nationalism. Although there are no clearcut cases, Poland provides at least a partial example. A country with only the briefest interludes of near
polyarchy during the French Revolution and its aftermath, dominated and divided for a century thereafter by Russia and Prussia, Poland’s latest experience with
polyarchy had already terminated in presidential absolutism3 before the country was overrun by the German and the Russian armies. Unlike the situation in the occupied
countries of Western Europe, in Poland polyarchy is simply not there waiting to be revived, and the conditions for creating it anew are certainly far from favorable.
Turning now to situations in which the dominant power seeks to inaugurate polyarchy in a country where the previous regime has been hegemonic, the third case in
table 9.4 pretty
3. The term is Gordon Skilling’s: The Governments of Communist East Europe (New York: Crowell, 1966), p.28.
Page 200
clearly poses much less of a challenge than the fourth. Even if conditions were by no means wholly favorable for polyarchy in Austria, Germany, Italy, and Japan in the
wake of the Second World War, on a comparative basis each was a prime candidate for polyarchy. Comparatively speaking, reestablishing polyarchy in these
countries was an easy job. In fact, the very smoothness with which polyarchy was reinaugurated probably encouraged American policy makers to adopt simplistic and
overoptimistic assumptions about the prospects of inaugurating polyarchy elsewhere, a delusion that was reinforced by the seeming success of American aid to Greece
and Turkey beginning in 1947.
On a comparative basis practically all of the countries that have achieved independence since 1945 have highly unfavorable conditions for polyarchy. Nonetheless, in
1970 Jamaica, Trinidad, India, the Philippines, and Lebanon were all governed by polyarchal regimes, and Malaysia and Cyprus by nearpolyarchies. Although it
would be folly to predict much stability for these regimes, their existence does raise a question as to the circumstances under which foreign intervention may facilitate the
development of polyarchy in a country where many of the conditions are highly unfavorable to it.
Several factors seem to have helped the development of polyarchy in these countries. For one thing, in all these countries a substantial proportion of the political
activists were predisposed to favor the institutions of polyarchy. In India, for example, in the years preceding independence there was an astonishing degree of
agreement among leaders and activists on the proposition that India must be a democracy in the Western sense. As the last chapter has shown, to explain why the
political activists of a country hold the beliefs they do during a particular historical period is highly complex. So complex is the process by which political activists
acquire their beliefs that an outside power, particularly when it is itself a polyarchy, can have only limited success in generating support for a particular ideology; the
foreign power is caught
Page 201
in a tough network of historical and cultural forces that frequently it can do very little to manipulate.
Moreover, in these countries prolonged and massive coercion was not employed by the foreign power against the subject population. Thus the boomerang effect was
to some extent avoided. At any rate polyarchy was not discredited by association with the brutalities of a hated outside force. Thus it was possible in these countries for
the rudimentary institutions of polyarchy or nearpolyarchy to be introduced and to function over a period long enough for the political activists to acquire skill in
operating them and a vested interest in maintaining them.
These were historically uncommon circumstances, and today they seem to be absent in most countries where the underlying conditions are unfavorable to polyarchy.
The end of formal colonialism means that the outside power today must move into a nominally independent country where nationalism is probably strong and the
boomerang effect is likely to be powerful. A substantial proportion of political activists are likely to favor a hegemonic regime of some sort. Public contestation, which
may allow deadly enemies to enhance their following, will seem to be a luxury at best, at worst downright pernicious. Even if the outside power intervenes at the behest
of the local government, it will be invited in precisely because that government cannot, unaided, put down its opponents. Thus the outside power is drawn into massive
coercion …
Page 202
10. THE THEORY: SUMMARY
AND QUALIFICATIONS
If the argument of the previous chapters is correct, then the chance that a country will be governed at the national level for any considerable period of time by a regime
in which opportunities for public contestation are available to the great bulk of the population (that is, a polyarchy) depends on at least seven sets of complex
conditions. These are summarized in table 10.1, which necessarily ignores the subtleties and qualifications in the argument.
In principle it would be possible—and as better data become available no doubt it will be possible—to rank the various countries of the world according to these
variables. For the sake of exposition let us suppose that countries were ranked in deciles. If about one country in five is governed by a polyarchy, we should expect that
in the 1960s and 1970s a very high proportion of the countries in the upper deciles would be polyarchies and negligible proportions in the last two or three deciles.
Thus a country with a profile like that of A in figure 10.1 would almost certainly be a polyarchy, and probably a typical polyarchy would have a profile rather like that of
A. Conversely, one would predict with complete confidence that a country with a profile like B would not be a polyarchy; very likely it would be a hegemony.
Doubtless too, most countries with profiles like that
Page 203
Table 10.1. Conditions Favoring Polyarchy
Most favorable to polyarchy Least favorable to polyarchy
I. Historical sequences Competition precedes inclusiveness Inclusiveness precedes competition Shortcut:
from closed hegemony to inclusive polyarchy
II. The socioeconomic order:
A. Access to
1. Violence Dispersed or neutralized Monopolized
2. Socioeconomic sanctions Dispersed or neutralized Monopolized
B. Type of economy
1. Agrarian Free farmers Traditional peasant
2. Commercialindustrial Decentralized direction Centralized direction
III. The level of socioeconomic development High: GNP per capita over about $700–800 Low: GNP per capita under about $100–200
IV. Equalities and inequalities
1. Objective Low, or Parity and dispersed inequalities High: Cumulative and extreme
2. Subjective: relative deprivation Low or decreasing High or increasing
V. Subcultural pluralism
1. Amount Low High
2. If marked or high None a majority One a majority
None regional Some regional
None indefinitely out of government Some permanently in opposition
Mutual guarantees No mutual guarantees
VI. Domination by a foreign power Weak or temporary Strong and persistent
VII. Beliefs of political activists
1. Institutions of polyarchy are legitimate yes no
2. Only unilateral authority is legitimate no yes
3. Polyarchy is effective in solving
yes no
major problems
4. Trust in others high low
5. Political relationships are:
strictly competitive no yes
strictly cooperative no yes
cooperativecompetitive yes no
6. Compromise necessary and desirable yes no
Page 204
of C in figure 10.2 would be polyarchies, and most polyarchies would have a profile like A or C or some combination of the two. Again, a country with a profile like D
would almost certainly not be a polyarchy, and probably a high proportion of countries like B and D, or some combination of B and D, would be hegemonies.

FIGURE 10.1 FIGURE 10.2
The analysis in previous chapters suggests, however, that we would have a number of important deviant cases with profiles very different from these. Thus some
polyarchies would not be in the upper deciles on all variables; the outstanding exception would surely be India, which would probably fall in the lower deciles on
conditions IV and V, and (because of the traditional peasant society in which about 80 percent of the Indians live) fairly low on condition III. India's profile might look
something like that in figure 10.3. Moreover, some hegemonies would be very high on one or more characteristics; East Germany, for example, would rank high on
both IV and VI, and perhaps fairly high on V; it might be approximately like that shown in figure 10.4.
Page 205

Denmark might be characterized by figure 10.5; although the shortcut from a traditional hegemony to a polyarchy should have caused difficulties, all the other factors
are highly favorable. As a final illustration, Argentina would perhaps look like figure 10.6.

Page 206
Some Qualifications
Why are these profiles, the reader may enquire, allowed to stand as hypothetical or, better, impressionistic representations of the data? In fact, why have I not tested
the theory by actually undertaking the rankings that would provide an actual profile for every counry or at least for a sufficiently large number of countries to provide a
moderately good test of the theory? The answer points directly to one of the severe limitations of the theory: I have not done so because, given the kinds of data now
available, the results would, I believe, be misleading and illusory. To be sure, it is possible to find satisfactory data on some of the variables—notably on the level of
socioeconomic development. No doubt one reason why so much attention has been given to the relationship between regime and socioeconomic level is simply that
reasonably acceptable (if by no means wholly satisfactory) ‘‘hard” data are available from which to construct indicators. This is a perfect example of how the
availability of data may bias the emphasis of theory. As I indicated in chapter 5, to focus on socioeconomic levels exaggerates the importance of this factor and
obscures the importance of others. For example, data on objective and subjective inequalities are very poor, and as I indicated in chapter 8, only the most fragmentary
comparative information is available on the beliefs of political activists in various countries, particularly in countries with hegemonic regimes. The poverty of data may
well be overcome in the next several years as a result of various studies now under way. Meanwhile, to construct profiles for a large number of countries would, I think,
lend a spurious validity to a theoretical exploration that necessarily rests at present on a rather modest foundation of data.
The weaknesses in the data impose another serious limitation having to do with the formulation of the theory itself: I have made no attempt to assign weights to the
various
Page 207
explanatory variables. Clearly this is a matter of great theoretical (and practical) significance, and it cannot be gainsaid that the defect is a grave one. Once again, it
would be possible and indeed easy to create a mask in order to put a strong face on a slackjawed theory. It is tempting, for example, to interpret the theory as a
multiple regression equation. But in the absence of satisfactory data from which to construct the predictor variables, the values of the constants would be fictitious. The
sophisticated reader would know that we had remained where we had started out, while the innocent reader might be deceived into believing that the theory is more
precise than in fact it can be.
Page 208
11. POSTSCRIPT: SOME IMPLICATIONS
FOR STRATEGIES OF CHANGE
To the extent that I have been able to answer it, the last chapter completes the answer to the question posed at the end of the first chapter. The book really ends there.
Yet the argument of the book seems to me to have some implications for strategies of change. Although the problem of how to transform hegemonies into polyarchies
would be a whole subject by itself, it may be useful to tease out of the general argument a few conclusions, general as they may be, that are more directly relevant to
action.
One can hardly reflect on the various conditions that seem to account for differences in regimes without concluding that a country that has had little or no experience
with the institutions of public contestation and political competition and lacks a tradition of toleration toward political oppositions is most unlikely to turn into a stable
polyarchy in the space of a few years. It is also true that countries with a long history of toleration, competitive politics, and broad participation rarely turn into
hegemonies.
It is unrealistic to suppose, then, that there will be any dramatic change in the number of polyarchies within a generation or two. Some hegemonic regimes may be
transformed into mixed regimes, some nearlyhegemonic mixed regimes may become nearpolyarchies, and some nearpolyarchies
Page 209
may become polyarchies. Doubtless, too, there will be some movement in the opposite direction, but short of extensive conquest by imperialistic hegemonies the
number of polyarchies is unlikely to diminish much. As with a great many things, the safest bet about a country’s regime a generation from now is that it will be
somewhat different, but not radically different, from what it is today.
This view of regimes is sobering, perhaps, and some readers may feel that it is unduly pessimistic. It is far removed, certainly, from the boundless optimism of much
democratic thought, particularly among Americans. Yet it is worth recalling that the Age of the Democratic Revolution, as Palmer has called the last third of the
eighteenth century, ended without any enduring “democracies”—or, in the language of this book, polyarchies—except in the United States. Even in the United States, it
is no great exaggeration to say that what is called the American Revolution only legitimized or, in some cases, speeded up a bit the processes of democratization that
had already been taking place in the colonies, processes that were by no means completed by the end of the struggle for independence. I am not arguing that the
revolutionary movements had no longrun effects, nor that these effects were unimportant or undesirable; in the long run the revolutionary movements of the eighteenth
century helped to create some of the conditions for polyarchy, particularly in the matter of beliefs and the dispersion of inequalities. But they failed pretty completely in
their major objective, which was to achieve lasting representative republics based upon popular suffrage—that is, polyarchies. The revolutions that swept over Europe
in May 1848 also had longrun consequences, but they too failed to build durable polyarchies.
Page 210
Implications for Foreign Assistance
It follows that policy makers in a country like the United States who may wish to transform the government of another country from a hegemony or mixed regime into a
polyarchy face formidable and complex problems, not the least of which is our lack of knowledge about the long causal chain running from outside help to internal
conditions to changes of regime.
The most promising situation would exist in a country where the seven sets of conditions described earlier produce a highly favorable profile (for example, a country like
C in figure 10.2), but where for some reason an unpopular dictatorship has imposed itself on the country. In spite of the illusions of some American policy makers and
publicists, countries like this are a statistical rarity. The least promising situation would exist in a country where the seven sets of conditions produced a decidedly
unfavorable profile. This, of course, is precisely the case in a preponderant share of the countries that most desperately need economic aid.
If the most promising situation is rare and the least promising commonplace, it is nonetheless true that some countries have a mixed profile—more favorable for
polyarchy in some respects, less favorable in others. Radically different as they are in many ways, Argentina, Cuba, Czechoslovakia, Greece, Yugoslavia, Spain, and
Taiwan illustrate the point.
It is worth asking, therefore, how policy makers in, say, the United States might facilitate the transformation of countries like these into polyarchies. There appear to be
three main strategies: (1) to invade and occupy the country, forcibly displace the government and replace it with a polyarchy, and then protect the new government as
long as may be required; (2) to support democratic—that is, revolutionary—movements within the country, with funds, weapons, or other resources; (3) to support the
existing government with
Page 211
funds, weapons, or other resources and bring pressure on it to change.
The first strategy has had a few striking successes and many failures. In the form of colonial domination and tutelage over a lengthy period, it has helped to create
polyarchy in India, the Philippines, Jamaica, and Puerto Rico. Military defeat followed by occupation and the restoration of polyarchy helped to create polyarchies in
Italy, Germany, Austria, and Japan. Over against these successes there is a long string of failures. The revolutionary governments established with the aid of the French
in the later stages of the French Revolution all collapsed. Though many of the new excolonial countries began their independence in recent decades with an experiment
in polyarchy, the number now governed by dictatorships far exceeds the number governed by polyarchies. Moreover, the first strategy has probably been rendered
obsolete by developments since World War II. Colonialism in its classic form of direct domination has all but disappeared, while “neocolonialism” in the form of
economic and political penetration of nominally independent and sovereign countries has not so far been conducive to the establishment of polyarchy. Attempts to
establish polyarchy by direct military intervention of the kind following the defeat of the Axis dictatorships in World War II may prove to be, and one devoutly hopes
will be, historically unique. The tragedy of Vietnam has demonstrated how illusory is the belief that the United States can on its own volition establish polyarchy in
another country by force.
For a great many reasons the second strategy also runs a high risk of failure. The government of one country attempting to involve itself in the revolutionary political
movements of another is likely to be clumsy, inept, ignorant, politically insensitive, and trapped against its will in the factional disputes of the revolutionaries. In
particular, to suppose that the American government could skillfully manage such complex relationships is, since the disaster of the Bay of Pigs, highly
Page 212
implausible. Then, too, in dealing with revolutionaries, dictatorships have most of the advantages in resources. It is not too much to say that if revolutionary forces within
the country have the popular base necessary for a polyarchy, they do not need much outside support, and if they do not have a popular base then no amount of outside
support short of outright invasion and occupation (and not necessarily even that) will enable them to build a polyarchy. Then, too, throughout most of the world
democratic movements are no longer revolutionary, and nowadays revolutionary movements are rarely democratic. On the contrary, in much of the world, where the
profile of conditions is highly unfavorable to polyarchy, revolutionaries who overthrow a hegemonic regime are very likely to replace it with another. That the new
hegemony may be better or worse than the old is highly relevant in judging the desirability of revolutionary action, but it is irrelevant to the success of the second
strategy. Finally, only a new government willing to pay the price of hostile relationships with regimes different from its own is likely to find the strategy acceptable. Even
governments that emerge out of a revolution tend to abandon the strategy because of the advantages of stabilizing their relationships with other countries, irrespective of
regime. Needless to say, governments of wellestablished polyarchies are unlikely to risk much in behalf of revolutionaries; no country came to the aid of the Hungarian
revolution in 1954 or the Czechs in 1968.
The third strategy has no more prospect of success than the others. It is one thing to aid a beleagured polyarchy to survive, as in India, Israel, or Chile. But it is quite
another to aid a hegemonic or mixed regime on the assumption that aid can be used as a lever for transforming the regime. Dictators and oligarchs are not easily
beguiled by foreign assistance into destroying their own regimes. In this respect, whatever the longrun effects may be on the conditions for polyarchy in various
countries, as a strategy for transforming
Page 213
nonpolyarchies into polyarchies the American foreign aid program must be adjudged a total failure. So far as I am aware, it has not a single such success to its credit.
Suppose then that in a country governed by a polyarchy the political leaders, activists, or citizens in general wish to advance the prospects for polyarchy in other
countries. What can they do? In particular, what about economic and technical aid provided with the cooperation of the country in need of assistance?
It is important to distinguish among the grounds on which economic and technical aid might be given. In the first place, aa powerful case can be made entirely on moral
and humanitarian grounds. In addition to altruistic considerations, a powerful case can also be made that over the long run it is in the rational selfinterest of the wealthy
to eradicate misery, frustration, and suffering among the poor, that any welloff country should provide assistance to any less welloff country where the aid would be
used for the reduction of misery. If assistance were rendered on these grounds, the nature, ideology, and foreign policies of the regime would be irrelevant. And unless
international organizations proved to be less efficient in allocating and administering funds, there would be no reason why aid should not be administered by international
or multilateral organizations rather than purely national ones.
Second, no doubt something of a case can also be made (in the end it is perhaps a special instance of the arguments on moral and humanitarian grounds or from self
interest) that in the very long run socioeconomic development facilitates polyarchy. But for the reasons discussed in earlier chapters, the socioeconomic level of a
country is only one of a number of variables bearing on the chances of polyarchy, and the relationship between the one and the other is very far from simple, direct, or
well understood. Because of the tenuous, uncertain, and longrun nature of this relationship, assistance on these grounds to any particular coun
Page 214
try that is not already close to a full polyarchy is a gamble with a highly uncertain payoff. It may be reasonable to demand that citizens of the most affluent polyarchies
should undertake the gamble. But given the enormous uncertainties in the lengthy causal running chain from assistance to polyarchy, with rare exceptions it would be
virtually impossible to show that assistance to underdeveloped country X is more likely to lead eventually to polyarchy than assistance to underdeveloped country Y.
Once again, the conclusion seems to follow that aid should be granted on the basis of a country’s need and capacity to use it, without regard to the nature or ideology
of the regime. And here, too, there is no selfevident reason for channeling aid through a national agency rather than through international organizations.
Third, a strong case can be made on moral grounds that citizens of affluent polyarchies should provide assistance if it is requested by the government of a needy country
that has already established a polyarchy or even a nearpolyarchy that is close to arriving at a comparatively high degree of liberalization and participation.
Finally, then, there is the question of providing economic, technical, or military assistance to specific hegemonic or mixed regimes on the assumption that these kinds of
aid will bring about a transformation of the regime into a polyarchy, or something very close to it. The whole burden of this book, I believe, argues against the rationality
of such a policy. For the process of transformation is too complex and too poorly understood to justify it. The failure of the American foreign aid program to produce
any transformations of this kind over two decades gives additional weight to this negative conclusion.
It would be a serious misinterpretation, however, to read this argument as an objection to economic and technical aid. On the contrary, there is every reason, I believe,
why the wealthy countries should allocate a substantial part of their increasingly fabulous riches to reduce misery and suffering
Page 215
in poorer countries. But the justification for such action is morality and compassion, or rational selfinterest, and not the prospect that outside assistance can be used as
a lever to pry democratization and liberalization out of a hegemonic regime.
Implications for Political Action
Suppose, now, that we look at the problem from the perspective of someone within the country. Assume that a hypothetical Innovator in a country governed by a
hegemonic or mixed regime wishes among other things to lower the barriers to effective participation and political oppositions—wishes, that is, to move the country
closer if not necessarily all the way to full polyarchy. What if anything does the analysis contained in the previous chapters offer our Innovator?
Virtually nothing, certainly, in the way of advice about how to acquire the power needed for political innovation in the context of a specific personal, national, and
historical situation: whether to seek reform or revolution, to work within, alongside, outside, or against the existing regime, to acquire and use power as a member of the
governing class or as clandestine member of a conspiratorial opposition, etc. But beyond these crucial tactical problems of acquiring the necessary power for change,
there are questions of strategic objectives, of the specific ways in which power, once acquired, can be used in order to move closer to polyarchy in a particular country.
As always, tactics and strategy cannot be sharply separated; in principle, strategic objectives should govern the choice of tactics, and yet the choice of tactics may
prove decisive in the achievement of objectives.
Even if the argument of this book is unhelpful for matters of tactics, it does, I think, have some implications for strategy that may be useful for our Innovator. Let me
now suggest several of these.
Page 216
Taking Stock
Suppose the Innovator begins simply by taking stock. Perhaps the most obvious conclusion he might draw from the preceding chapters is that in any given country at
any given time, the possibilities are limited. In some countries, a wise and even courageous citizen might simply throw up his hands in despair and await the further
unfolding of history until the moment became more propitious for his endeavors.
Enthusiastic democrats will doubtless say that such a statement is excessively pessimistic. But I believe that more lasting harm has been done to the prospects of
democracy by ignoring the odds than by facing them squarely. Surely the political events of this century have fully vindicated the older view that democracy is not bound
to triumph irresistibly over all the obstacles placed in its path by the history and current condition of a people.
Yet nothing in the preceding pages endorses a mindless pessimism nor, certainly, the bias, so common among Americans, which holds that representative democracy
requires a unique conjunction of exceptionally rare conditions such as only a few peoples, like the Americans themselves, are fortunate enough to be endowed with. For
what stands out sharply in a world perspective is the variety of circumstances in which polyarchies, if I may revert to that term once again, are operating today.
In sum: according to the view presented here and supported at least in a modest way by the evidence available to me, the extent to which the opponents of a
government have an opportunity to contest the conduct of that government is chancy but not accidental. And if the prospects for oppositions are mainly determined by
factors beyond their immediate control, the degree of freedom oppositions enjoy can sometimes be altered by deliberate human choice.
The preceding chapters have emphasized seven kinds of
Page 217
factors which by their variation may favor either polyarchy or hegemony. To repeat, if the Innovator lives in a country where all seven elements favor political
competition, the chances for a polyarchy are of course very high, even if not a sure thing. If he lives in a country where all the elements favor hegemony, then the
chances for a competitive regime, and particularly for an inclusive polyarchy, are naturally very low, while chances for a hegemonic regime are very high.
In his choice of tactics and strategic objectives, the Innovator must of course consider the profile of his country and the limitations and possibilities it suggests. In a
country with a highly favorable profile and considerable experience with competitive politics under a mixed regime it would be reasonable to work for changes that
would allow the operation of a full polyarchy. In a country with a highly unfavorable profile, the immediate introduction of a full polyarchy would be a utopian objective;
for even if the existing hegemony were overturned or fell into dissolution and thus enabled a group of innovators to adopt a constitution prescribing the institutions of
polyarchy, that constitution could hardly be expected to be effective, and the institutions it prescribed would be empty or ephemeral. Yet even in a situation so
depressing for the democrat, an Innovator might bring about certain changes that would permit greater participation and contestation and thus over a much longer
period increase the prospects for polyarchy.
Mutual guarantees
Opponents in a conflict cannot be expected to tolerate one another if one of them believes that toleration of another will lead to his own destruction or severe suffering.
Toleration is more likely to be extended and to endure only among groups which are not expected to damage one another severely. Thus the costs of toleration can be
lowered by effective mutual guarantee
Page 218
against destruction, extreme coercion, or severe damage. Hence a strategy of liberalization requires a search for such guarantees.
The prescription stresses the search rather than the specifics, since in each country the problem is so different that it would be foolish to offer a general solution. The
first steps toward liberalization in a hitherto highly hegemonic regime are bound to be different from the next steps in a quasihegemony. In a full hegemony the first step
may be nothing more or less than some kind of understanding that in conflicts within the ruling groups the losers will not be punished by death or imprisonment, exile, or
total destitution. In this respect the change in the USSR from Stalin’s hegemony to the postStalinist system was a profound step toward liberalization.
As we have already seen, among polyarchies the nature and explicitness of the guarantee to potential antagonists vary widely; in highly homogeneous countries explicit
arrangements are less important than in countries with deep subcultural cleavages, where constitutional and institutional structures may be required to give adequate
security to the subcultures. What is more, the cleavages and the institutions for managing them are rarely static: in India the state boundaries were redrawn from 1956
onward to conform more closely to major linguistic boundaries; in the Netherlands, verzuiling is very far from an unchanging phenomenon, and since it was identified it
has been declining. What is persistent in the Netherlands or in the very different system of Switzerland, what is a cornerstone of the developing system of regional
(state) governments in India, what has been evolving with the Walloons and Flemish in Belgium, and what is increasingly demanded by French Canada is a system of
mutual guarantees that provide security to major subcultures.
In a country with subcultural cleavages and a hegemonic
Page 219
regime, the Innovator who seeks to liberalize the system also confronts the problem of mutual guarantees. In their moves toward liberalization in 1968, before the
Soviet invasion and occupation, the leaders of the oneparty hegemonic regime in Czechoslovakia had endorsed the idea of a federal system that would allow the two
territorial and linguistic groups—the Czechs and Slovaks—each to have its own province within the nation. Moreover, a proposed new constitution for the ruling
Communist party would have met the problem of subcultural pluralism not only by reorganizing the party along federal lines but also by a provision for a mutual veto
over certain decisions taken by the Central Committee. While decisions of the Central Committee would ordinarily be taken by majority vote, decisions concerning
‘‘the existence, sovereignty, or superior national or territorial interests” would have to be approved by at least half of the representatives of each subculture in the
Central Committee, each group voting separately. Since the Slovaks composed only about 18 percent of the party members, in effect the opposition of slightly more
than 9 percent of the party could, in principle, block a decision of this kind.1 Thus the Czechs and Slovaks appeared to be moving toward a “consociational” solution
to the problem of protecting a subcultural minority in a system intended to be much less hegemonic and more responsive to popular pressures.
As in polyarchies, the closer a country approaches a system of mutual veto the more it runs the risk of immobilisme. The risk of immobilisme is the price a country
with potentially hostile subcultures pays for tolerating the political expression of its cleavages. In many countries, obviously, those who opt for or against liberalization
are not prepared to pay the price. Yet in the long run the price may be lower than it first appears. For the experience of some “consociational” countries shows that a
system of mutual guarantees can
1. See the report by Michel Tatu in Le monde (Sélection Hebdomadaire, August, 814, 1968), pp. 1,4.
Page 220
gradually reduce the fears and hostilities of the subgroups, help to generate a broader and deeper support for the institutions of participation and public contestation,
strengthen a sense of common nationhood, and in time allow the political energies absorbed in subcultural conflicts to be channeled to the other important problems of
the society.
Executive authority
When barriers to public contestation and participation are lowered, interests and demands will appear which the government has hitherto ignored. If public contestation
and participation are to be effective, the authority of political institutions responsive to these new interests and demands must be increased. But to reduce the likelihood
of immobilism and deadlock, the executive must retain a considerable measure of power for rapid and decisive action, especially in emergencies. Thus the executive
must have authority that in a realistic sense is beyond the capacities of transitory majorities in parliament to curtail and yet not beyond the reach of influence of
substantial and persistent coalitions, whether minorities or majorities.
This prescription, which is necessarily vague, directs the Innovator’s attention to what has invariably been one of the most formidable problems of constitution makers
who are trying to develop a competitive political system. The formulation above is an extremely general statement of the problem and the solution—or rather the
direction in which a solution must be sought. The most familiar special case of this problem in representative governments is of course the classic question of the
relations between executive and legislature. Operating with a theory of representative democracy that emphasized the exclusive legitimacy of the elected assembly as
the supreme representative of the people’s will, nineteenthcentury constitution makers found it difficult to provide for
Page 221
an executive with independent sources of authority. In this century, however, most polyarchies have rejected as unworkable the model of assembly government with a
weak, subordinate executive; in practice governments in polyarchies and especially their executives have been armed with very great powers for vigorous and decisive
action. It is now all too clear that any attempt to emulate assembly government is likely to lead to disaster and, in time, a clamor for the rapid creation of hegemonic
controls.
In a general way, then, the movement in polyarchies has been from a dependent executive to what is de facto if not de jure an independent executive. The problem for
our Innovator is exactly the opposite. Political leaders who are too independent must be made more dependent for their authority on political institutions responsive to a
variety of interests and demands. The ancient and recurring lesson from hegemonies is that too much independence is oppressive. The lesson of polyarchies is that too
little independence can lead to governments so ineffectual as to stimulate support for hegemonic rule.
It is doubtful whether the strategic objective can be formulated much more precisely than these very general remarks. Given the enormous variations from country to
country, a more detailed strategy would obviously have to be formulated to fit the situation of a particular country.2
Preventing fragmentation
Since the costs of toleration are raised by excessive fragmentation into competing political parties, a strategy of liberalization requires a search for a party system that
avoids a great multiplication of parties.
2. An attempt to do exactly this for Argentina is contained in an article, “Del gobierno revolucionario al orden constitucional,” Criterio 61 (Buenos Aires, June 13,
1958): 371–75. Although I had not seen the article when I first wrote this chapter with its recommended strategies, there is sufficient similarity to reassure me that the
strategies recommended here are in fact relevant for at least some countries.
Page 222
The Innovator must consider with very great care the probable pattern of cleavages that might emerge under various kinds and degrees of liberalization. In a highly
homogeneous country, one could give serious consideration to the specific institutional and constitutional changes that might ultimately produce a political system roughly
on the model of the classic twoparty parliamentary system—a kind of idealized version of the British system. But in all likelihood the Innovator’s country is not
sufficiently homogeneous to enable the classic twoparty parliamentary model to work. The chances are that his country contains a number of fairly distinctive and
potentially antagonistic subcultures—regional, linguistic, religious, racial, ethnic; for high homogeneity is, as we have seen, a rare phenomenon, while subcultural
cleavages are relatively commonplace. And wherever there are extensive subcultural cleavages, the classic twoparty parliamentary model simply cannot be achieved—
witness Canada, South Africa, India, Belgium, the Netherlands, Switzerland, and Lebanon.
The chances are, then, that our citizen will have to keep in view a very different model. It is probably unfortunate, and indeed it may be disastrous, that both the model
of assembly government and the classic twoparty parliamentary model have been widely regarded as ideal forms of representative democracy, from which all
deviations are a sign of defect. For just as the assembly model has been rejected by practically all polyarchies in the twentieth century, so the classic twoparty
parliamentary model cannot be successfully adopted in countries with subcultural cleavages, that is, in most countries. Yet the prestige of these models is high, and
alternative systems that have been developed to cope with the problem of severe subcultural cleavages are little understood, slighted in writings on representative
government, and downgraded as secondbest substitutes for the real thing. But most countries cannot move toward liberalization, and certainly not toward the degree
of liberalization
Page 223
that exists in a polyarchy, except by one of these secondbest substitutes that in some countries have contributed to an extraordinary degree of peaceful toleration of
diverse and often conflicting subcultures.
Because of the way in which party systems have evolved in the theory and practice of representative democracy, they are often thought of as “natural” institutions that
develop spontaneously and faithfully mirror the “natural’’ cleavages of the society. According to this view of parties, deliberate attempts to relate and control the nature
and number of parties in a polyarchy are either doomed to failure or must violate the democratic rules of the game. Neither of these assumptions seems to me to be
valid. The first has become increasingly suspect as the relations between various election arrangements and party systems have been carefully explored. The exact
nature of these relations is still subject to dispute, but the most systematic analysis to date establishes among other propositions that:
1. All election systems tend to award to larger parties a share of seats in the legislature larger than their share of the votes—and hence to smaller parties a smaller
share of seats than votes.
2. The degree of divergence is a function of electoral formula and the size of districts.
3. The maximum divergence results from elections in singlemember districts, and the maximum proportionality from P.R. with a large number of seats in each
district allocated according to the formula of the “largest remainder.”3
The party system, then, is not a natural, spontaneous, or inevitable mirroring of social cleavages. It is dependent to some degree on electoral arrangements. And these
can be deliberately manipulated in order to maximize or minimize
3. Douglas Rae, The Political Consequences of Electoral Laws (New Haven: Yale University Press, 1967), pp. 134–40.
Page 224
fragmentation. The manipulation of electoral procedure has of course been discredited because governments have sometimes used it to benefit their own party or
coalition at the expense of an opposition, but since electoral procedures are one of the few aspects of a political system subject to more or less deliberate alteration, to
reject this possibility of social engineering seems to me equivalent to the medical profession refusing to use antibiotics because they have been so widely abused.4
Aside from purely partisan manipulation, it is difficult to see why the deliberate attempt to reduce potential fragmentation through electoral engineering is inherently
“undemocratic,” or, even if it were, undesirable in a liberalizing hegemony. During a century of debate no one has ever satisfactorily demonstrated whether P.R. is more
“democratic” than plurality elections in singlemember districts or less. The difficulty of any demonstration seems to lie in the fact that all democratic schemes for arriving
at a majority vote require formal and informal processes of great complexity for scheduling alternatives and aggregating preferences, and thus finally reaching a decision
on a pair of alternatives remaining to policy makers. The institutions in which scheduling and aggregation take place vary in different polyarchies. In any particular
country, a twoparty system with highly unified parties would require much more aggregation and reconciliation of different interests and preferences to take place
within each party than would be the case in a twoparty system with heterogeneous parties and shifting legislative coalitions. A twoparty system with heterogeneous
parties probably would require more aggregation within each party than would a multiparty system, where more aggregation would occur in legislative and cabinet
coalitions.
4. A point on which I occur with Giovanni Sartori, “European Political Parties: The Case of Polarized Pluralism,” in Political Parties and Political Development, ed.
Joseph LaPalombara and Myron Weiner (Princeton: Princeton University Press, 1966), pp. 165–66.
Page 225
Finally, if the country had three or four significant parties more aggregation would have to take place within each party, and correspondingly less in legislative and
cabinet coalitions, than if the country had five or six significant parties. It is difficult to see how one of these systems would be more “democratic” than another, unless it
insured a more equal weighting of individual preferences in the final choice (including preferences for a satisfactory but not necessarily most preferred outcome). And
this has not been demonstrated.
In any case, whether one or the other electoral and party system is more ‘‘democratic” is virtually irrelevant to the problem of liberalization of a hegemonic regime. It
would be absurd to argue for a strategy of liberalization requiring electoral arrangements and a party system thought to be perfectly democratic if these would lead
predictably to high political fragmentation, rapid disillusionment with the game of party politics, and renewed support for stronger hegemonic controls in order to
eliminate political fragmentation.
A hegemonic regime taking its first steps toward a plural party system may find it desirable to regulate the number of parties even more directly, by making it difficult—
or in the extreme case impossible—to form more than two parties, a government party and an opposition party.
Whether in any given country the number of parties optimal for the expression and aggregation of preferences is two, three, four, or more cannot be answered even
approximately without examining the cleavage system of the particular country. Thus a strategy of liberalization might erect barriers to party formation and competition
that would effectively limit the number of parties to two, three, four, or conceivably even more. The important point, however, is that in a rational strategy of
liberalization the number of parties can and should be regulated. In principle, provided at least two parties exist, to regulate the number of parties is independent of the
regulation of freedom of dissent, criticism, and opposition in elec
Page 226
tions and legislation. A government might grant the fullest freedom for dissent, criticism, and opposition in elections and legislation, and yet restrict the number of
competing parties to two, each with equal opportunities to state its views, propose candidates, gain votes, and win elections.
Local governments
Since somewhat autonomous representative institutions below the national level can provide opportunities for the opposition to acquire political resources, help to
generate crosscutting cleavages, and facilitate training in the arts of resolving conflicts and managing representative governments, a strategy of toleration requires a
search for ways of developing subnational representative governments.
In a hegemonic regime—particularly, perhaps, in a unified hegemony—it may be prudent to take bigger steps toward liberalization at lower levels before these are
introduced at higher levels, and particularly at the national level. For example, allowing oppositions to expand their participation in municipal elections may help to
socialize both the opposition and the government. Smaller representative units provide training in solving concrete, comprehensible problems for which abstract
ideological solutions are less relevant. Confronting these problems may unite groups that would be antagonistic in national political life. Moreover, extending opposition
rights and privileges at subordinate levels is likely to be less threatening to the incumbent national leadership since it can be treated as a trial; if the trial “fails,” the step
can be reversed.
Perhaps the most noteworthy example of subnational representative governments is Yugoslavia, where not only in local governments but in the more famous case of
workers’ control a very large measure of autonomy has been extended; and while organized political opposition remains rather re
Page 227
stricted in the factory and in the municipality, these units evidently do provide a widely diffused training in the arts of participating in representative institutions, in
discussion, debate, conciliation, compromise, concrete analysis, responsibility. In Mexico, the eventual shift from a highly pluralistic hegemony to a polyarchy may well
come about slowly and not too painfully as the opposition wins an increasing number of victories at the local level and takes on the always sobering responsibility of
governing.
Unless one is prepared to dismiss the experience of polyarchies as irrelevant to the ways in which hegemonies might be liberalized, then one ought not to be overly
surprised to realize, as the attentive reader will have done by now, that the main elements in the strategy of liberalization that I have suggested—mutual security among
conflicting groups, a strong and vigorous executive dependent on institutions responsive to a variety of interests and demands, an integrating rather than a fragmentary
party system, and representative governments at subnational levels—are all familiar aspects of the most durable representative democracies.
Page 228
This page intentionally blank
Page 229
APPENDIXES
Page 230
This page intentionally left blank.
Page 231
APPENDIX A
114 Countries Classified According to Eligibility to
Participate in Elections and Degree of Opportunity for
Public Opposition
Table A1 displays 114 countries classified according to the opportunity of adult citizens to vote in elections and ranked according to the opportunities available to
political oppositions to compete for popular support and public office.
These two “dimensions” correspond approximately to the two dimensions of polyarchy described in chapter 1. The classification of the 114 countries according to
eligibility to participate in elections was done by Richard Norling in 1969, but changes since the end of 1968 were not included. The ranking of the countries according
to the degree of opportunity for public contestation was done by Mary Frase Williams and Richard Norling. It may be useful to describe the procedures they used.
The Variables
The hunch that underlies an enterprise of this kind is that there may be an underlying, hypothetical continuum that extends from the greatest to the least opportunity for
oppositions, as described in chapter 1, and that countries can be
By Robert A. Dahl, Richard Norling, and Mary Frase Williams.
Page 232
Table A1. 114 Countries Ranked by Opportunities to Participate
in National Elections and to Oppose the Government, circa 1969
Unascertained,
Opportunities Elections Percent of adult citizens uncertain, or
for political not held eligible to vote transitionala N
oppositions: Under Over
scale types 20% 20–90% 90%
Greatest opportunity 1 Switzerland Belgium 8

Denmark
Findland
Luxembourg
Netherlands
Norway
Sweden
3b Chile Iceland 4
United States Israel 4
4 Ireland Cyprus 5
Italy Dominican Republic
United Kingdom
5 Australia 8
Austria
Canada
West German Federal Repulic
Japan
New Zealand
Philippines
Uruguay
6 France Turkey 3
Lebanon
7 India 3
Jamaica
Trinidad and Tobago
8 Costa Rica Malaysia 2
9 Ecuador Colombia Guatemala 5
Venezuela Honduras
Page 233
Table A1. (continued)
Unascertained,
Percent of adult citizens eligible to vote
Opportunities for political Elections not uncertain, or
oppositions: scale types held Under 20% 20–90% Over 90% transitionala N
10 Uganda 1
11 (Bolivia)c
[France]d
12 (Libya)c 1
13 Morocco
14 South Africa Ceylon 4
Malagasy Republic
Mexico
15 El Salvador 1
16 Nicaragua 1
17 (Somali 1
Republic)c
18 Mauritania 1
19 Brazil Ruwanda 2
20 Peru Jordan Paraguay Iran 7
South Korea
Laos
Pakistan
22b Argentina Liberia Ivory Coast 4
Burma
23 Indonesia Algeria 3
Tunisia
24 Burundi 4
Cameroun
Chad
Gabon
25 Dahomey Sudan Congo (Brazz) Afghanistan 7
Mali
Niger
Tanzania
26 Central Af. South Vietnam Ethiopia Cambodia 15
Rep. Guinea
Haiti
Ghana Poland
Greece
Page 234
Table A1. (continued)
Unascertained,
Percent of adult citizens eligible to vote
Opportunities for political Elections not uncertain, or
oppositions: scale types held Under 20% 20–90% Over 90% transitionala N
Nepal 1
Panama
Sierra Leone
Taiwan
Thailand
United Ar. Rep.
27 Syria Senegal Portugal 6
Soviet Union
Spain
Yugoslavia
28 Upper Volta Czechoslovakia 3
Hungary
29 Albania China 8
Bulgaria North Korea
East Germany
Mongolia
Romania
North Vietnam
30 Cuba 4
Least Opportunity 31 Nigeria
Saudi Arabia
Togo
Yemen
Totals 21 1 8 61 23 114
a Includes countries where a constitutional government or elections have been superseded or nullified at least once since 1960, the constitution has been suspended, a
state of siege declared, or massive civil violence has occurred.

b No countries were found in scale types 2 and 21.

c Countries in parentheses have had a coup d’étate, resulting in military rule, since the classification above.

d See text for discussion of the case of France
Page 235
located along this continuum. This assumption is illustrated with purely hypothetical countries in figure A1. The initial problem, then, is to determine whether it is
reasonable to assume that such a continuum actually exists and whether the available data will enable one to rank countries along it. In practice, a scalogram analysis of
the kind used to produce table A1 requires that both tasks be carried out together.
The first need is to find variables that will accomplish this, that is, serve as valid indicators of the underlying continuum—in this case the ‘‘degree of opportunity for
public contesta
FIGURE A1. Hypothetical Continuum of
“Opposition Opportunity”
tion” or political oppositions—and for which adequate data are available. The solution adopted here was to employ as indicators of the seven required conditions
described in chapter 1 certain variables from Arthur S. Banks and Robert B. Textor’s A CrossPolity Survey, which sought to classify 115 countries on 57
characteristics as of about 1960–62.1
The following required condition: Is indicated by the following variables:
1. Freedom to form and join organizations Freedom of group opposition [#30]2
Interest articulation by asociational groups [#33]
1. Cambridge: M.I.T. Press, 1963.
2. Number in brackets is the number of the variable in Banks and Textor.
Page 236
2. Freedom of expression Freedom of the press [#13]
3. Right to vote Representative character of current regime [#28]3
Current electoral system [#29]
4. Right of political leaders to complete for support Interest articulation by political parties [#37]
Party System: Quantitative [#41]
Freedom of Group opposition [#30]
Current electoral system [#29]
5. Alternative sources of information Freedom of group opposition
Freedom of the press
Representative character of current regime
Party system: Quantitative
6. Free and fair elections Freedom of group opposition
Freedom of the press
Representative character of current regimes3
Party system: Quantitative
7. Institutions for making government policies depend on votes and other Constitutional status of present regime [#26]
expressions of preference Interest aggregation by legislature [#40]
Horizontal power distribution [#48]
Current status of legislature [#50]
Mrs. Williams found that these elevent characteristics did form a satisfactory scale and was able to place 102 countries in 27 scale types. However, because the data in
A CrossPolity Survey had been gathered during 1960–62and resutled in some obvious misclassifications as of 1969, in that year
3. Later abandoned, for reasons described below.
Page 237
Richard Norling, using the Yale Political Data program and a number of standard sources,4 attempted to bring the data up to date. He also developed the data for a
new variable, the opportunity to participate in national elections, which became the basis for the fourfold classification of countries shown in table A1. It was then
decided to drop, as redundant, one of the eleven variables from the CrossPolity Survey (#28, Representative character of current regime). Hence the scale for
opportunities for political oppositions in table A1 is based on only ten of the eleven variables listed above. The variables are shown in table A2, with the number and
percentage of countries in each category.
Weighting
Since there was no a priori or theoretical reason for weighting the variables unequally or assigning a greater weight to one category than another, the variables were all
handled the same way. Since the underlying attribute in which we are interested is the opportunities available to oppositions, the alternatives within each of the ten
variables were ordered from the greatest to the least opportunity. The alternative that provides the greatest opportunity was given a score of 1, the next greatest a score
of 2, etc.5 The total score for each country was obtained by summing the scores for the individual criteria for that country.
4. T. E. Smith, Elections in Developing Countries (New York: St. Martin's Press, 1960), Canada Yearbook, Statesman's Yearbook, The New York Times, Christian
Science Monitor, and Carolyn K. Colwell, Foreign Affairs Division, Legislative Reference Service, Library of Congress, “The Independent Nations of Africa: Leaders
ElectionsPartiesPopulationElctorates,” August 8, 1968.
5. For every question a number of countries were coded as unascertained, unascertainable, ambiguous, and irrelevant. These answers were excluded from the analysis
and every country that had such answers for more than five of the variables was also excluded from the analysis.
Page 238
Table A2. Variables Used as Indicators of Opportunities for Oppositions
% N Var. #a Variable description and categories
13. Degree of freedom of the press
38 40 1. Complete (no censorship or governmental control of either domestic press or foreign correspondents)
18 19 2. Intermittent (occasional or selective censorship of either domestic press or foreign correspondents)
26 27 3. Internally absent (strict domestic censorship.No restraint on foreign newsgathering, or selective cablehead censorship)
18 19 4. Internally and externally absent (strict direct or indirect censorship or conrol, domestic and foreign)
100% 105
2 7. Unascertainable
7 9. Unascertained
26. Contitutional status of current regime
48 53 1. Constitutional (government conducted with reference to recognized constitutional norms)
2. Authoritarian (no effective constitutional limitation, or fairly regular recourse to extraconstitutional powers. Arbitrary exercise of power
34 37
confined largely to the political sector)
18 20 3. Totalitarian (no effective constitutional limitations. Broad exercise of power by the regime in both political and social spheres)
100% 110
3 9. Unascertained.
29. Current electoral system
44 47 1. Competitive no party ban, or ban on extremist or or extraconstitutional parties only)
4 4 2. partially competitive (one party with 85 percent or more of legislative seats)
33 35 3. Noncompetitive (singlelist voting, or no elected opposition)
19 20 4. No elections (elections not usually held, or suspended by regime in power)
100% 106
2 7. Unascertainabl
6 8. Ambiguous
30. Degree of freedom of group opposition
40 41 1. Autonomous groups free to enter politics and able to oppose government (save for extremist groups, where banned)
Page 239
Table A2 (continued)
2. Autonomous groups free to organize in politics, but limited in capacity to oppose government (includes absorption of actual or potential
19 19
opposition leadership into government)
31 32 3. Autonomous groups tolerated informally and outside politics
10 10 4. No genuinely autonomous groups tolerated
100% 102
2 8. Ambiguous
33. Interest articulation by associational groups
17 19 1. Significant
12 13 2. Moderate
26 29 3. Limited
45 49 4. Negligible
100% 110
3 8. Ambiguous
37. Interest articulation by political parties
22 22 2. Moderate
13 13 3. Limited
48 47 4. Negligible
100% 99
5 8. Ambiguous
40. Interest aggregation by legislature
16 16 2. Moderate
13 13 3. Limited
59 59 4. Negligible
100% 100
6 8. Ambiguous
41. Party system—quantitative
25 25 1. Multiparty coalition or minority party government normally mandatory if parliamentary system)
12 12 2. Twoparty or effectively twoparty (reasonable expectation of party rotation)
6 6 3. Oneandahalfparty (opposition significant, but unable to win majority)
14 14 4. One party dominant (opposition, but numerically ineffective at national level. Includes minority par
Page 240
Table A2 (continued)
ticipation in government while retaining party identity for electoral purposes)
5. One party (all others nonexistent, banned, nonparticipant, or adjuncts of dominant party in electoral activity. Includes “National Fronts”
34 35
and oneparty electoral systems)
10 10 6. No parties, or all parties illegal or ineffective
100% 102
8 7. Unascetainable
4 8. Ambiguous
48. Horizontal power distribution
29 32 1. Significant (effective allocation of power to functionally autonomous legislative, executive, and judicial organs)
20 23 2. Limited (one branch of government without genuine functional autonomy, or two branches with limited functional autonomy)
51 51 3. Negligible (complete dominance of government by one branch or by extragovernmental agency)
100% 112
1 9. Unascertained
50. Current status of legislature
26 28 1. Fully effective (performs normal legislative function as reasonably “coequal” branch of national government)
15 16 2. Partially effective (tendency toward domination by executive or otherwise partially limited in effective exercise of legislative function)
22 24 3. Largely ineffective (virtually complete domination by executive or by oneparty or dominant party organization)
36 39 4. Wholly ineffective (restricted to consultative or ‘‘rubber stamp” legislative function, or no legislature)
100% 107
Source: Arthur S. Banks and Roberts B. Textor, A CrossPolity Survey (Cambridge: M.I.T. Press, 1963), updated to 1968 and recoded.

a These numbers are those assigned to the variable by Banks and Textor.
Determining the Scale Types
For each variable, a frequency distribution was obtained (table A2). You will note that the number of countries that
Page 241
could be classified varied somewhat from one variable to another; consequently the percentages are based on different numbers.
A crucial assumption is that each of the ten variables covers the entire underlying continuum and that this continuum can be divided into various segments that
correspond to the categories. On the basis of the information available, it is impossible to determine how the various categories divide the hypothetical continuum into
segments. However, we can obtain other information that will serve the same purpose. To do so, we assume that every country is located somewhere along the
underlying continuum; that each criterion divides the continuum into segments and the countries into subgroups; and that the frequency distributions reflect the way the
questions subdivide the countries.
N variables with a total of m categories will normally cut the underlying continuum into (m − n + 1) segments. With 10 variables subdivided into a total of 40 categories,
presumably there are (40 − 10 + 1) = 31 scale patterns. The 31 perfect scale types were ultimately reduced to 29 because no countries were found to fit scale types 2
and 21.
Classification of Nonscale Types
Once the perfect scale types have been established, the next step is to assign every subject to a scale type before we can assess how well the data conform to the
hypothetical scale. This requires procedures for assigning countries with patterns that represent nonscale types. There are two alternative procedures. One is to assign
each country to the perfect scale type that has the same total score. The other alternative is to assign each nonscale type to the type it most closely resembles, that is, to
the scale pattern that will minimize the number of errors.6 Unfortunately the second procedure does
6. An error is said to occur when a country is placed in a category of one of the variables where it would not be predicted to belong simply on the basis of the rank or
scale type of the subject.
Page 242
not yield unambiguous results, since there may be more than one perfect scale type that will result in the same (and minimum) number of errors. If this happens, then
auxiliary criteria must be used to assign the subject to one of these types.
Assignment on the basis of total scores is easier and yields unambiguous results, but it is an unsatisfactory procedure for our data since so many of the countries have no
information for at least one question. This means that the raw total score cannot be used to assign subjects to scale patterns. Instead the total score must be
standardized in some way to take account of the number of criteria that go into making up that score. However, this solution—a standardized total score—proved
unsatisfactory for two reasons: (1) the standardized score depended heavily on the identity of the missing question(s) and (2) it resulted in a larger number of scaling
errors than the minimum error method. For these reasons, nonscale types were assigned according to the criterion of minimum error.
Accuracy of Fit
The coefficient of reproducibility (Rep) is the criterion commonly used to determine whether the data approximate a perfect scale well enough to treat the data as if it
were a perfect scale.
total number of responses — number of errors
Rep =
total number of responses
Thus Rep is the percentage of responses that coincide with the scale pattern to which the subject has been assigned.
The total number of errors under the minimum error method was 180. The value of Rep is therefore 0.829 compared with an expected value of 0.443. While this is
some
Page 243
what lower than one would like, it suggests that the ten variables form a moderately satisfactory scale.7
A Word of Caution by R.A.D.
In considering the rankings, one should remember that many of the variables on which the rankings are based necessarily require decisions based on judgments rather
than “hard” data. This is true of the original data from Banks and Textor as well as the revisions and reclassifications made by Richard Norling. Judgments of this kind
may be biased by many things, including the simple fact that a great deal more is known about some countries than others. There are several ways of coping with
missing data, but no solution is entirely satisfactory.
Whether for reasons of this kind or because inaedquacies in my choice of the variables, the scalograms resulted in a few anomalies. The most obvious was the location
of France, and a brief discussion of the case of France may serve as a useful example and caution on the use of the tables. In both Mrs. Williams’ scalogram based on
Banks and Textor and in Mr. Norling’ revision, France appeared in the same scale pattern as Bolivia. One does not need to be biased in favor of France to conclude
that France was badly misplaced. For even before the coup in 1969, Bolivia was hardly a polyarchy, while oppositions in France enjoyed a very high level of
opportunity to contest the government. Looking into the details, it turns out that France failed to gain high marks from Banks and Textor on a large proportion of the
indicators: #13, Freedom of the Press (intermittent); #37, Interest Articulation by Political Parties (moderate); #41, Party System: Quantitative (one party dominant);
#48, Horizontal Power Distribution (limited); and #50, current Status of Legislature (partially effective). It seems clear to me that
7. When Mrs. Williams used all eleven characteristics on 102 countries, the coefficient of reproducibility was 0.881
Page 244
standards were applied with considerably more severity to France than to Bolivia and many other countries.
It may be helpful if we look at several of these classifications with the hindsight of 1970. In 1962, Banks and Textor classified France as a country where freedom of
the press (#13) was intermittent, along with such countries as Bolivia, Brazil, Burma, Congo, Laos, South Africa, Syria, and Turkey. The higher category, where
freedom of the press is “significant,” included some 43 countries, among them the Dominican Republic, Panama, and a number of African countries now ruled by a
single party. One may well wonder whether freedom of the press in France is not considerably more “significant’’ and much less “intermittent” than in countries such as
these.
Similarly, France was classified as “limited” in “interest aggregation by the legislature” (#40), along with countries such as Bolivia, Brazil, Guatemala, Honduras, Libya,
Nicaragua, etc. Finally, with considerable justification as of 1962, FranceÙs party system was classified as “one party dominantÝ along with Bolivia, Mexico,
Paraguay, and Vietnam, and others. While tendencies toward oneparty dominance doubtless seemed strong in 1962, by 1969 it was becoming clear that France no
longer quite fitted the definition of “one party dominant” given by Banks and Textor: “opposition, but numerically ineffective at the national level.” Although France may
not precisely fit the remaining categories, its party system might be classified with greater justification as “oneandahalf party,” defined as “opposition significant, but
unable to win a majority.”
Consider what the effects of these changes in classification would be. If France were now classified on the basis of its new total score, it would fall with India, Jamaica,
and TrinidadTobago. Classified by the minimum error method used in table A1, it falls in type 6, with Lebanon and Turkey.
Although Mrs. Williams and Mr. Norling were, quite properly, unwilling to substitute their judgment on France for
Page 245
those of the experts used for A CrossPolity Survey, I am less reluctant to do so. It is true that once the process of second guessing the classification begins—except
on the basis of actual changes in the political system or the appearance of new data—it would be hard to know where to stop. And to Üadjust” all the cases that seem
out of line would undermine one’s confidence in the validity of the whole process. Nonetheless, since France appears to me to be markedly more misclassified than any
other country in Table A1, it is shown in type 6 and also, in brackets, in type 11. Its dual location will perhaps serve as a visible warning against taking the ranking in
table A1 as if it had been engraved in stone by the hand of God. Doubtless there are other errors. Nonetheless, the ranking is, I believe, useful. The time may not be
far off when criteria of the kind used here, or better and more appropriate criteria, will yield improved rankings that can be an ordinary part of the current store of
information as readily available to social scientists as data on, say, GNP, which also must be taken cum grano salis. If and when the country profiles discussed in
chapter 10 are also available, we shall be in a better position to test generalizations about the conditions for polyarchy.
Page 246
APPENDIX B
Contemporary Polyarchies, circa 1969
Using table A1, it is possible to construct a list of polyarchies as of about 1969. As a cutoff point, I have chosen scale type 8. All countries in column 4 in scale types
1–8 are included, a total of 25 countries. The three special cases are:
1. In Chile, a literacy requirement excluded a substantial percentage of the population until the 1960s. Increasing literacy and perhaps some relaxation in the
enforcement of the requirement have led to such a rapid diminution of the unenfranchised minority that Chile may by 1970 have entered the ranks of the fully
inclusive polyarchies.
2. Switzerland does not have universal female suffrage in national elections.
3. In the United States, Negroes were systematically deprived of the franchise in much of the South until after the passage of the Civil Rights Act of 1964. (For
percentages of whites and blacks registered to vote in Southern states in 1968, see footnote 3, chap. 2.) Although the total percentage of the population
excluded was never more than 10% (by 1950 less than 7%), the fact that it was a particular
Page 247
and sharply deprived minority greatly increased the discriminatory nature of the exclusion. Residency requirements reduced the electorate even further.
The list of contemporary polyarchies in table A3 is almost identical with the list of 31 “contemporary democratic systems” provided by Rustow (table A4). Rustow’s
table includes all countries which at the beginning of 1967 had regimes based on three or more consecutive, popular, and competitive elections. The four discrepancies
are italicized. They include Greece, where as Rustow notes constitutional government was overturned by a military dictatorship in April 1967; Mexico and Ceylon,
which because of restraints on opportunities for oppositions in the 1960s falls into scale type 14 in table A1; and Columbia, which I have classified as a near
polyarchy. Moreover, my list (table A3) includes two countries omitted from Rustow’s list (Jamaica and TrinidadTobago) because they had not yet held three
popular elections since their independence in 1962. Both lists omit several microstates that would doubtless qualify as polyarchies.
The list of nearpolyarchies includes four countries in scale types 1–8 that are listed in column 5 because during the 1960s the processes of polyarchy were at some
point massively interrupted by civil violence, repression, or military intervention. The list also includes Colombia and Venezuela, where despite sporadic violence the
institutions of polyarchy emerged rather fully during the 1960s. Throughout this period Colombia was operating under a constitutional amendment (due to expire in
1975) that imposed limitations on party competition in elections. In all of the nearpolyarchies, the emergent institutions of polyarchy were still so fragile that it was
uncertain how long they would bear the strains of internal conflict.
Page 248
Table A3. Polyarchies and NearPolyarchies, circa 1969
Fully inclusive polyarchies
1. Australia
2. Austria
3. Belgium
4. Canada
5. Costa Rica
6. Denmark
7. Federal Republic of Germany
8. Finland
9. France
10. Iceland
11. India
12. Ireland
13. Israel
14. Italy
15. Jamaica
16. Japan
17. Lebanon
18. Luxembourg
19. Netherlands
20. New Zealand
21. Norway
22. Philippines
23. Sweden
24. Trinidad and Tobago
25. United Kingdom
26. Uruguay
Special cases: Electoral restrictions
27. Chile
28. Switzerland
29. United States
Nearpolyarchies
1. Colombia
2. Cyprus
3. Dominican Republic
4. Malaysia
5. Turkey
6. Venezuela
Page 249
Table A4. Contemporary Democratic Systems
Continuous
popular elections
Country since
1. United States 1788
2. Norway 1814
3. Belgium 1831
4. United Kingdom 1832
5. Netherlands 1848
6. Switzerland 1848
7. New Zealand 1852
8. Denmark 1855
9. Sweden 1866
10. Canada 1867
11. Iceland 1874
12. Luxembourg 1868
13. Australia 1900
14. Finland 1906
15. Mexico 1920
16. Ireland 1921
17. Lebanon 1926
18. Ceylon 1931
19. Chile 1932
20. Uruguay 1942
21. Austria 1945
22. France 1946
23. Greece 1946
24. Italy 1946
25. Japan 1946
26. Philippines 1946
27. Israel 1949
28. West Germany 1949
29. Costa Rica 1949
30. India 1952
31. Colombia 1958
Note: This list is from Dankwart A. Rustow, A World of Nations: Problems of Political Modernization (Washington, D.C.: Brookings, 1967), table 5, pp. 290–91 I
have included only one of the fourteen columns in his table. Other columns include data on participation, party changes, and constitutional continuity. Countries in italics
are omitted from the list of polyarchies in appendix table A3.
Page 250
Page 251
INDEX
Abel, Theodore, quoted, 147, 185, 186
Activists, political (see chap. 8): beliefs of, and regime change, 125–29; Great Man and, 126; preference of, for hegemony, 201. See also Beliefs, political
Africa: oneparty regimes in, as desirable, 30; cleavage and twoparty system, 222
Agrarian society: propensity for hegemony, 53–54; and equality, 82–83; effect of industrialization on, 85–86, 92
Algeria, as French failure, 123
Almond, Gabriel A., quoted, 145
American Revolution, impact of, in Europe. 172–73, 209
Amhara. See Ethiopia
Argentina: politicized military in 50; socioeconomic level in, 69; labor force of, in agriculture, 85; failure of party system in, 123; legitimacy of dictatorship in, 130;
development of suffrage in, 133–34; failure of polyarchy in, 135–41 passim; political exclusion of immigrants in, 138; mixed profile of, 205, 210. See also Peron,
Peronism
Aristotle: cited, 81; on importance of early training, 167
Armed forces, role of, politicized, 50
Athens (Athenian democracy): agrarian society, 72; inequalities in, 81; as dual polyarchy, 93; importance of early training in, 167; weakened beliefs in, 185; citystate
system in, 191
Attenuation, process of, 101–02
Australia: socioeconomic level in 70; as preindustrial polyarchy, 71; literacy and socioeconomic level in, 75
Austria: demise of regime in, 105; Proporz in, 118; polyarchy imposed in, 192; polyarchy reestablished in, 200, 211
Authority, beliefs and attitudes, 140–44 passim
Banfield, E. C., quoted, 163–64
Banks, Arthur S., quoted, 109–10
Barber, James D., quoted, 179–80
Belgium: polarization in, 107, 111, 113, 218, 222; polyarchy sustained in, 113; foreign domination not fatal to, 193, 198–
Page 252
99; cleavage and party systems in, 222
Beliefs, political (see chap. 8): education related to, 127; limited influence of, 127, 131–32; genesis of, 128; in legitimacy of polyarchy, 129; in children, 131; and other
kinds of belief, 132; authority and, 140, 143; effectiveness and, 144; mutual trust and, 150; cooperation and trust and, 152–53; compromise and, 161; selfinterest
and, 163–64; personality structure and, 165; political culture and, 166–67; early acquisition of, 167–68, 177; loss of, 168; axiomatic quality of, 176; exposure and,
169–70; prestige and, 171; consistency in, 175; cultural experience and, 177; reality and, 178–79; personal experiences and, 180; replacement of, 184; individuals
and, 185; ideology and, 186; accidents, uncertainties, and, 187
Brazil, personal expectations in, 100, 102
Britain: slow development in, 46; nonpoliticized navy of, 49; socioeconomic level in, 70; potential polarization in, 107; 1830 beliefs in legitimacy in, 130; fair elections
in, 137; belief in authority in, 142; as effectiveness model, 148; survival of polyarchy in, 149; mutual trust in, 150; attitude of youth toward monarchy in, 168–69;
influence of on India, 171
Canada: free farmer society, 53; socioeconomic level in, 70; preindustrial polyarchy in, 71; literacy and socioeconomic level in, 75; sharpened polarization in, 107,
111, 116, 218; polyarchy sustained in, 113, 119; cleavage and party system in, 222
Cantril, Hadley, quoted, 96, 99, 101, 102
Catelfuoco. See Italy
Castro, Fidel: follows authoritarian model, 148; prestige of, 174. See also Cuba
Ceylon: socioeconomic level in, 69; pluralism in, 111
Chile: nonpoliticized military in, 50; geographical protection of, 56; socioeconomic level in, 69; foreign influence in, 191–92; aid to, for survival, 212
Coercion: in different regimes, 27; violent (military), nonviolent (socioeconomic), 48–50; technology and, 54; Peronism and, 130–31; foreign domination and, 201. See
also Repression
Compromise, as virtuous and as evil, 161
Contestation: necessity of, to polyarchy, 20–21; factors broadening, 24–26. See also Suffrage
Cooperation: as necessary element of belief, 152–60 passim; and competition, 155, 160. See also Trust
Costa Rica, as durable polyarchy, 134, 135
Croce, Benedetto: as critic, 17, 20; on Fascism, 18
Cuba: socioeconomic level in, 69; personal expectations in, 100; U.S. control in, boomerangs, 198; mixed profile of, 210. See also Castro, Fidel
Czechoslovakia: attempt to liberalize, 20, 219–20;Soviet control in, boomerangs, 198; mixed profile of, 210
Page 253
Democracy: definition, guarantees, 1–9 passim; contestation in, 6; nationalism and, 43. See also Polyarchy
Denmark: as anomaly, 37n; foreign control not fatal to, 198–99; mixed profile of, 205
Dominican Republic: personal expectations in, 100; U.S. control in, boomerangs, 198
East Germany: socioeconomic level in, 69, 78; hegemonic profile of, 205
Eckstein, Harry, quoted, 141, 142
Effectiveness, beliefs about, 144–49 passim
Egypt: socioeconomic level in, 69; personal expectations in, 100
Erikson, Erik H., quoted, 179
Ethiopia: unilateral authority in, fixed, 143–44; distrust in, 150–52
Finer, S. E., quoted, 30–31
Finland: blended ideologies in, 43; literacy and socioeconomic level in, 75; anomalous polyarchy in, 192–93
Foreign aid: as facilitating polyarchy, 210–15 passim; morality and selfinterest of, 212–15; usefulness of, as strategy, 213
Foreign control (see chap. 9): means of effecting, 190–91; imposes regimes, 191, 192–93, 200; military aspects of, 193; as not fatal to polyarchy, 193; boomerang
effect of, 198; as neocolonialism, 211
France: 1789–92 path to polyarchy in, 34–35; regression to hegemony in, 43; founders over Algeria, 123; skepticism in, 146–47; foreign control not fatal to, 198–99
Franco, Francisco, attempted opposition to, 20. See also Spain
Franklin, Benjamin, as symbol of prestige, 171
Free farmer society: as precedent for polyarchy, 53–56; and political resources, 83
Frondizi, Arturo, overthrow of, 138
Gandhi, Mohandas K.: as symbol of prestige, 172; influence of Ahmedabad strike on, 179
Germany: hegemonic regime replaced in, 23, 38; coercion in, 27–28; historical path to Weimar in, 34; regression to hegemony in, 43; Weimar collapse in, 105; party
system in, 123; beliefs under Nazism in, 129; beliefs of youth in, 131; competition with “middle way” in, 148; eroded trust in, 148, 151, 186; effect of personal
ideology in, 186; polyarchy imposed in, 192; polyarchy reestablished in, 200, 211
Goldsmith, Raymond, quoted, 69
Greece: failure of party system in, 123; U.S. aid to, 200
Greenstein, Fred I., quoted, 165, 169
Guatemala, U.S. control in, boomerangs, 198
Harrington, James, on power and property, 54
Haug, Marie R., quoted, 108, 110, 112
Hegemony: defined, 7–8; factors necessary for change from, 10; desirability of shift from, irrelevant, 17, 30–31, 225; consequences of, important, 30; tested models
of, 41; agrarian society and, 53–55; private ownership and, 58–59; socioeconomic level and tensions of,
Page 254
78–79; political resources and, 82; disintegration of, in U.S. South, 94; belief in, 125–29 passim; authority patterns of, 141–44 passim; eroded mutual trust in, 150,
156; imposed by historical accident, 187–88; imposed by foreign control, 189, 191; favored by activists, 201; profiles favoring, 203–04; strategies for liberalizing,
225–26. See also Inequalities; Polyarchy; Prestige; Repression; Revolution; Subcultural pluralism
Hitler, Adolf, extermination policy of, 27–28. See also Germany
House UnAmerican Activities Committee, and dissent, 44
Iceland, literacy and socioeconomic level in, 75
India: blended ideologies in, 43; socioeconomic level in, 68; ‘‘attenuation” in, 102; language conflict in, 111; pluralism in, 114, 117, 119, 218, 222; party system in,
123, 222; skepticism in, 147; cooperation and belief in, 153–54, 159; British influence on, 171, 186; readiness of, for polyarchy, 200; exceptional profile of, 204;
colonial background of, 211; aid to, for survival, 212
Indonesia: coercion in, 28; socioeconomic level in, 69
Industrialization, inequalities and, 85–87
Inequalities (see chap. 6): and hegemony, 81–83, 93; toleration of, in polyarchy, 89–90; government response to, 90–91; disadvantaged groups’ response to, 95–96;
as objective and subjective, 98–99; “attenuation” and, 101–02
Ireland, blended ideologies in, 43
Israel: blended ideologies in, 43; high personal expectations in kibbutzim in, 99, 100; familism weakened in kibbutzim in, 164; aid to, for survival, 212
Italy: Mosca and Croce on preFascist regime in, 17–18; Salvemini on, 19n; hegemonic regime replaced in, 24, 38; beliefs in, and Fascism, 129; beliefs of youth in,
131; skepticism in, 146–47; competition with “middle way” in, 148; eroded trust in, 150–54 passim; amoral familism in, 163; communist missionaries in Castelfuoco,
182–83, 184; polyarchy imposed on, 192; polyarchy reestablished in, 200, 211
Jamaica: unstable polyarchy in, 200; colonial background of, 214
Japan: hegemonic regime replaced in, 24, 38, 200; party system in, 123
Jayaprakash Narayan. See Narayan, Jayaprakash
Jebb, John, and Westminster Constitution, 176n
Jefferson, Thomas: cited, 132; and representative government, 170
Keniston, Kenneth, quoted, 174, 178, 180
Kibbutz, See Israel
King, Martin Luther, Jr., as symbol of prestige, 172
Kothari, Rajni, quoted, 117–18
LaPalombara, Joseph, quoted, 150, 151, 153, 157
Lebanon: cleavage and party system in, 111, 222; as unstable polyarchy, 200
Page 255
Lenin, Vladimir Ilyich: as Great Man influence, 125; combines competition and cooperation, 159–60; early decisive experience of, 179; success of, accidental, 187–88
Levine, Donald N., quoted, 143, 150, 151, 153
Lijphart, Arend, quoted, 142
Lincoln, Abraham, cited, 132
Linz, Juan, on inauguration of polyarchy, 40n
Lipset, Seymour M., quoted, 36,184–85
Literacy, as condition for polyarchy, 75–76
Locke, John, cited, 132
Lorwin, Val R., qouted, 117–18, 195–96
Machiavelli, Niccolò, cited, 148
Madison, James, and representative government, 170
Malaya, pluralism in, 111
Marxism: irrelevance of, to rise and fall of regimes, 106; as cognitive, not moral, 126–27
Mexico: conquest of, 54; PRI follows authoritarian model in, 148; prestige of PRI in, 174; local changes in, 227–28
Mill, James, and representative government, 170
Mill, John Stuart: on multinational states, 108; influence of, on beliefs, 127; on cooperationcompetition, 155–56
Montesquieu, Baron de la Brède et de, cited, 148
Mosca, Gaetano, on preFascist Italy, 17, 18
Narayan, Jayaprakash, as advocate of partyless democracy, 158
Nationalism, and democracy, 43
Neocolonialism: not conducive to polyarchy, 211
Netherlands: sustained pluralistic polyarchy in, 113, 118, 120, 219; politics of accommodation in, 160–61; foreign domination not fatal to, 196, 198–99; cleavage and
party system in, 222
New Haven, Conn., inequalities dispersed in, 88n
New Zealand: as free farmer society, 53; socioeconomic level in, 70; preindustrial polyarchy in, 71; literacy and socioeconomic level in, 75
Nigeria, personal expectations in, 100
Norway: as free farmer society, 53; socioeconomic level in, 70; preindustrial polyarchy in, 71; literacy and socioeconomic level in, 75; authority patterns in, 141;
foreign domination not fatal to, 199
Onganía, General, ruler by decree, 138
Palmer, Robert R., quoted, 170, 172
Panama, personal expectations in, 100
Paraguay, socioeconomic level in, 69
Party systems: oneparty regimes, 30–31; in liberalization of hegemonies, 221–25 passim
Paulson, Belden, quoted, 182–83
Peron, Peronism: suppressed, 130, 138–39; inclusion of working strata in, 130–31; follows authoritarian model, 148. See also Argentina
Peru: conquest of, 54; follows authoritarian model, 148; prestige of junta in, 174
Philippines: socioeconomic level in, 69; pluralism in, 111; colonial background of, 211
Page 256
Pluralism. See Subcultural pluralism
Poland, prospect for polyarchy in, faint, 199
Political action: value of, as stimulus, 215; strategies for, 215–26
Polyarchy: terminology clarified, 4n, 9n; defined, 7–8; desirability of, irrelevant, 17, 30–31, 126; historical paths to, 31–35; ways of inaugurating, 40–42, 47; GNP
and, 69–70; profiles favoring, 203; evaluation of data on, 206–07. See also Hegemony
Polyarchy, dual: in U.S. South, 29, 93–94; in Athens, 93
Portugal, socioeconomic level in, 69
Prestige, usefulness of, 171–74
Prisoner’s Dilemma, as indicator of trust risk, 156n
Puerto Rico, colonial background of, 211
Religion, in subcultural conflict, 106n, 115, 117–18
Representative government: invention of model for, 169–70; Westminster draft of, 176n
Repression, cost of, and security, 15–16, 43. See also Coercion
Revolution: as precursor or polyarchy, 45, 208–09; foreign support and, 210; and hegemony, 212
Rokkan, Stein, quoted, 184–85
Rome, republican, as model of effectiveness, 148
Rose, Richard, quoted, 152
Rousseau, Jean Jacques, cited, 132, 148
Runciman, W. G., quoted, 96–99 passim
Russett, Bruce M., quoted, 63–64, 83–85 passim
Russia. See USSR
Sakharoff, Andrei D., data of, on coercion, 27n
Salvemini, Gaetano, on Fascism, 18–19, 20
Sartori, Giovanni, quoted, 122
Scandinavia, possible loss of autonomy in, 192
Sierra Leone, pluralism in, 111
Smith, Adam: on private ownership, 58; belief of, in competition, 156
Socioeconomic development (see chaps. 4, 5): GNP as index of, 63, 68–69; political development and, 65, 82–83; autonomous vs. induced, 73; literacy and, 75;
communication and, 76; pluralistic order and, 76–77; high GNP and tensions in, 79; subcultural pluralism and, 111
Somalia, army seizes power in, 66
Spain: regression to hegemony in, 43; conquest of New World, 54; civil war in, 105; fair elections in, 137; contempt for compromise in, 161–62; mixed profile of, 210.
See also Franco, Francisco
Spiro, Melford E., quoted, 164
Stalin, Joseph V., as Great Man influence, 125
Subcultural pluralism (see chap. 7): management of, 105; Marxism and, 106; terminology of, 106n; and multinational states, Haug data on, 108, 110; related to GNP,
111–12; and participation in government, 116; religious aspects of, 117; political systems and, 119–20; government effectiveness and, 120–21; party system and,
123–24, 221–22; mutual guarantees and, 218
Subnational units, importance of, 12–14. See also Yugoslavia
Suffrage: after U.S. Civil War,
Page 257
21, 28–29; effects of broadening, 22–26 passim
Sweden: historical path to polyarchy in, 34; as model of pluralistic order, 59; socioeconomic level in, 70; preindustrial polyarchy in, 71; response to inequalities in, 90;
earlier restricted suffrage in 133–34; “middle way” of, as model, 148; politics of compromise in, 160–62
Switzerland: suffrage in, 5; as free farmer society, 53; sustained pluralistic polyarchy in, 180, 111, 114, 222; religious pluralism in, 118, 119, 121
Taiwan, mixed profile of, 210
Terminology, clarified, 4n, 9n, 106n
Textor, Robert B., quoted, 109–10
Tocqueville, Alexis de, on U.S. in 1820s, 54, 56, 69–70, 72, 75, 173
Toleration, cost of, and security, 15–16, 48
Trinidad, unstable polyarchy in, 200
Trust: beliefs about, 150–52; risk involved in, 156n. See also Cooperation
Turkey: socioeconomic level in, 69; U.S. aid to, 200
USSR: suffrage in, 5; coercion in, 27–28; fiftieth anniversary of, 45; GNP per capita in, 69, 78; competition with “middle way” in, 148; seen as New Civilization by
Webbs, 174; move from Stalin hegemony in, 218. See also Marxism; Stalin
United States: dual polyarchy of South, 28–29, 93–94; blended ideologies in, 43; dissent in, 44; as free farmer society, 53; socioeconomic level in, 69; as preindustrial
polyarchy, 71; income inequality study in, 86; government response to inequality in, 90; personal expectations in, 100; cleavage and war in, 105; Great Man belief in,
126; beliefs of youth in, 131; belief in authority in, 142–43; diminished pride in, 145–46; New Deal as model in, 148; survival of polyarchy in, 149; Great Depression
and prestige of, 173; belief replacement in, 184; foreign influence on options in, 190; boomerang effect of attempts of to dominate, 198; and Bay of Pigs failure, 211;
and foreign aid failure, 213. See also Vietnam
Uruguay, socioeconomic level in, 69
Verba, Sidney, quoted, 145
Vietnam: war in, and U.S. prestige, 174, 196, 211; and young radicals, 180; failure to establish polyarchy in, 211
Webb, Sidney and Beatrice, on USSR, 174
Weimar Republic. See Germany
Weiner, Myron, quoted, 154, 158
Westminster Constitution. See Jebb, John
Yugoslavia: opportunities for subnational units in, 12, 226; mixed profile of, 213
Page 258
Also by Robert A. Dahl and available from Yale University Press
On Democracy
‘‘This elegantly written handbook on democracy will be widely read. It is vintage Dahl.”
—Michael Walzer, Princeton University
Democracy and Its Critics
“A work of extraordinary intelligence and, what is even rarer, a work of extraordinary wisdom.”
—Robert N. Bellah, New York Times Book Review
Winner of the American Political Science Association’s 1990 Woodrow Wilson Foundation Book Award and of the Conference for the Study of Political Thought’s
1991 Elaine and David Spitz Book Prize
Dilemmas of Pluralist Democracy
“Essential reading for anyone who takes democracy seriously.”
—Kay Lehman Schlozman, Political Science Quarterly
Cowinner of the American Political Science Association’s 1983 Gladys M. Kammerer Award
After the Revolution?
Authority in a Good Society
Revised Edition
“Dahl interweaves normative and empirical notions in an effort to develop a broadly integrative theory of democracy. … His effort is both remarkable and …
successful.”
—Annals of the American Academy of Political and Social Science
Who Governs?
Democracy and Power in an American City
“A landmark of American political science.”
—Duane Lockard, Annals of the American Academy of Political and Social Science
Winner of the American Political Science Association’s 1962 Woodrow Wilson Foundation Book Award

Robert A. Dahl - Polyarchy - Participation and Opposition-Yale University Press (1972) PDF

Uploaded by

Copyright:

Available Formats

Robert A. Dahl - Polyarchy - Participation and Opposition-Yale University Press (1972) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Robert A. Dahl - Polyarchy - Participation and Opposition-Yale University Press (1972) PDF

Uploaded by

Copyright:

Available Formats

Cover

STATEMENT #1 STATEMENT #2 STATEMENT #3 DON’T KNOW

Greatest opportunity 1 Switzerland Belgium 8

You might also like