Contemporary Political Systems: Classifications and Typologies 9781685851477

Citation preview

CONTEMPORARY POLITICAL SYSTEMS

CONTEMPORARY POLITICAL SYSTEMS Classifications and Typologies edited by

Anton Bebler and Jim Seroka

Lynne Rienner Publishers • Boulder & London

Published in the United States of America in 1990 by Lynne Rienner Publishers, Inc. 1800 30th Street, Boulder, Colorado 80301 and in the United Kingdom by Lynne Rienner Publishers, Inc. 3 Henrietta Street, Covent Garden, London WC2E 8LU © 1990 by Lynne Rienner Publishers, Inc. All rights reserved Library of Congress Cataloging-in-Publication Data Contemporary political systems: classifications and typologies [edited] by Anton Bebler, Jim Seroka Includes bibliographical references. ISBN 1-55587-147-X (alk. paper) 1. Comparative government. I. Bebler, Anton, n . Seroka, Jim. JF51.C628 1989 320.3—dc20 89-39943 CIP British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library. Printed and bound in the United States of America The paper used in this publication meets the requirements of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48-1984.

Contents

List of Tables and Figures Acknowledgments

Vll Vlll

PART 1. THEORETICAL AND HISTORICAL CONCERNS 1 Introduction Anton Bebler

3

11

2 On Classification Alberto Marradi 3 The Theory of the Mixed Regime and the Problem of Power: A Machiavellian Meditation Bruce James Smith

PART 2.

45

MODERN TYPOLOGIES

4 Democratic Political Systems A rend Lijphart

71

Commentary, Burkhard Koch

88

5 Authoritarianism Leonardo Morlino

91

Commentary, ZacharyIrwin

111

6 On the Classification of Socialist Political Systems Boris N. Topornin Commentary, Huang Zhong-Liang

117 133

7 Typologies of Socialist Systems Jim Seroka

141

Commentary, Miodrag Jovicit

153

8 The Comparative Study of Third World Political Systems Veniamin Chirkin Commentary, Iqbal Narain and Chandrakala Padia v

159 170

vi

Contents

9 Typologies of Third World Political Systems Dirk Berg-Schlosser Commentary, Iqbal Narain and Chandrakala Padia 10 A Neoinstitutional Typology of Third World Politics Fred W. Riggs Commentary, Iqbal Narain and Chandrakala Padia 11 Typologies Based on Political Parties Jerzy J. Wiatr Commentary, Paul Bacot and Claude Journis 12 Typologies Based on Civilian-Dominated Versus Military-Dominated Political Systems Anton Bebler Commentary, Henry Bienen

PART 3.

173 202 205 240 243 249

261 275

CLASSIFICATION BY GEOGRAPHIC AREA

13 African Political Systems Claude E. Welch, Jr. Commentary, Richard Dale 14 Arab Political Systems Bahgat Korany Commentary, Friedemann Buttner 15 Latin American Political Systems Carlos Huneeus

281 298 303 330 337

Commentary, George Philip

352

Commentary, Karen L. Remmer

356

16 Conclusions and Summary: A Long Journey Begins with a Single Step Jim Seroka

363

About the Contributors Index

369 375

Tables and Figures Tables 4.1 4.2

4.3 4.4

5.1 5.2 7.1 7.2 7.3 7.4 9.1 9.2 9.3 9.4 9.5 11.1 13.1 14.1 15.1 15.2 15.3 15.4 15.5

A Typology of Democratic Systems Twenty-Two Democratic Regimes Classified According to the Two Dimensions of Majoritarian Versus Consensus Democracy A Schematic Presentation of the Overlap Between Consociational and Consensus Democracy Coefficients of Correlation Between the Degree of Consensus Democracy and Pluralism, Population Size, British Heritage, and Democratic Quality Dimensions and Salient Variations in Authoritarian Regimes Models of Authoritarian Regimes Comparative Political-Demographic Data for Socialist States Comparative Socioeconomic Data for Socialist States Utility for Typology Development of Models of Socialist State Politics Suggested Representation for Socialist Nations on a Three-Dimensional Categorization, 1987 Types of Political Systems: Combination of Characteristics "Development" Performance by Dynamic System Type, 1960-1980 Resource Allocation by System Type Contributions to Growth Rate and DRR by System Type Observation of Normative Standards by Present System Type, 1980 Typology of Political Regimes by Degree of Political Differentiation and Responsiveness to Pluralism Typology of African Political Systems Main Characteristics of Arab Political Systems Typology of Political Systems in Latin America Principal Characteristics of Latin American Political Systems Examples of Each Latin American Political System Conditions of Emergence for Each Latin American Political System Emergence of Exclusionary Authoritarian Regimes: Preconditions and Development (The Cases of Argentina 1966, Brazil 1964, Chile 1973, and Uruguay 1974)

72

73 78

83 96 98 145 146 149 150 176 186 188 191 192 247 292 322 339 340 341 341

344

Figures 12.1 15.1

Spatial Distribution of Contemporary Political Systems Two-Dimensional Classification of Regimes

270 358

Acknowledgments

We deeply appreciate the help of Wendy Eidenmuller and Norma Hartner, without whose assistance this book would not have been possible. Anton Bebler Jim Seroka

viii

PART ONE Theoretical and Historical Concerns

1 Introduction Anton Bebler

Writing in an authoritative U.S. compendium in the early 1960s, Harry Eckstein stated that the field of comparative politics was "characterized by nothing as much as variety, eclecticism and disagreement" (Eckstein & Apter 1963, 29). Another well-known author in the same book, Roy C. Macridis, qualified the same field (calling it the comparative study of government) as being comparative in name only. According to him, its "traditional approach" had been essentially descriptive, parochial, static, monographic, and noncomparative (Eckstein & Apter 1963, 41-51). It was telling that about two dozen best-known Western political scientists whose works were printed in this representative volume failed to deal with, or even to raise the issue of, classification of political systems. By the late 1970s to early 1980s the situation had improved only slightly. If we look into the most widely used U.S. college textbooks on comparative politics, we are still likely to find the lack of discussion on classifying contemporary political systems, the absence of explicit classifications or of typologizing on the worldwide level, and only partial comparisons within groups of arbitrarily selected states on such mostly nonpolitical dimensions as age, size, wealth, economic inequality, cultural heterogeneity, and international dependency (Almond & Powell 1974). The book written by Gary Bertsch, Robert R Clark, and David M. Wood (1978) stands out in this respect. Its authors explicitly attempt to classify the 167 relatively self-governing states that were identified in Political Handbook of the World: 1980 (Banks 1980). Most of these states (141) are divided into three groups, respectively labeled the First, the Second, and the Third World. In spite of its overwhelming popularity, this sorting of political systems is unsatisfactory for social science purposes. Formally, the Bertsch-ClarkWood listing fails as a classification because it leaves twenty-six states outside its conceptual framework, listing them as "mixed," "uncertain," or "otherwise unclassified." This clearly violates the rules of exhaustiveness and of mutual exclusiveness of categories. The stated criteria for classification are "strategies for political change" and the relative salience, in the political 3

4

Introduction

process, of four basic human values: power; respect; well-being; and enlightenment. These criteria are nowhere operationalized and are evidently not followed through in many cases. Why should Andorra, Cyprus, Liechtenstein, Malta, Monaco, and San Marino be put, on the stated dimensions, in the same basket with the Maldives and Djibouti and not in the First World? The actual common denominators in the fourth ("unclassified") group seem to be smallness of population and area (i.e., a demographic and a geographic criteria). And why, then, should the authors put into another basket (i.e., the Third World) such states as Bahrain, Botswana, Brunei, Equatorial Guinea, Gabon, Gambia, Guinea-Bissau, Mauritius, Oman, Qatar, and Swaziland—all with populations below a million? In terms of strategies for development, what is the basis for grouping Turkey, South Korea, Taiwan, and Singapore together with Benin, Ethiopia, and Burma in the Third World, while Israel and Switzerland are in the First World? With respect to the four basic human values, why should Costa Rica be in the same class as Qatar? If the actual criterion for belonging to the First World is the presence of Western political democracy, to which group did Spain, Portugal, and Greece belong in the 1970s? The Bertsch-Clark-Wood classification is in fact the partially modified, camouflaged, and rehashed ideological division of the world into "Western democratic states" versus "communist totalitarian states." The "rest" of the world, comprising over two-thirds of the total units, is lumped together in the large, murky, and unspecified category "neither-nor" (i.e., the Third World). By giving these examples from the mass consumption North American political science literature, I do not imply that this obviously unsatisfactory state of affairs in comparative politics is particularly critical and acute on that continent. On the contrary, the situation in many other national or regional communities of political scientists is in this respect no better or worse. It is often dominated by unscientific ideological-political considerations, by parochialism and ignorance, by racial, religious, cultural, ethnic, and other stereotypes. A list of relevant quotations from Soviet, Eastern European, Chinese, Indian, Arab, Iranian, African, and other sources could be quite long and embarrassing. The retarded state of the discipline even in the largest communities of political scientists, the very uneven geographic distribution of scientific efforts in this field, and the low level of international communication became obvious to me more than a decade ago. I was gently pushed to deal with the problem when the Yugoslav Political Science Association proposed in 1983 that I organize a panel on classifications of contemporary political systems at the 1985 International Political Science Congress in Paris. The proposal was accepted by the International Political Science Association (IPSA). Two years later, our panel met in the venerable building on rue Saint-Guillaume, and it was a lively and stimulating experience. It attracted considerable attention, in

Anton Bebler

5

part because of the presence and active participation of some well-known authors, particularly the dean of French comparativists, Maurice Duverger. It occurred to me then that a book developed from our panel's work could be a useful contribution to the literature on comparative study and, it is hoped, to the development of international political science. Not wanting to follow the cliché of conference compendia, I decided to use only some of the papers presented at Paris as starters for a well-rounded, comprehensive, and consistent volume. The format of the congress and several concomitant limitations did not allow exploration of many relevant and indeed fundamental topics, while the composition of the panel was in part accidental. I therefore decided to search for competent authors who would fill in the lacunae left by the original panel's discussions. In looking for contributors, I sought, in accordance with the spirit of the IPSA, to secure the widest feasible geographic representation. In no other social science are classifications and typologies more susceptible to originating in or becoming value-laden designations for mass unscientific consumption and propaganda than in political science. In part because of this mass usage, many influential classifications and typologies wear off and become obsolete in political science faster than in other disciplines. They are also selectively used and, as a rule, given different meaning in various quarters. For these reasons, I strived to attract respectable scholars to the project from different political, ideological, and cultural environments, as well as those representing various philosophical viewpoints and different schools of social science research. The inevitable price to pay is the considerable heterogeneity of the resulting book in terms of approaches, styles, terminology and so forth. My main intention was to enhance our understanding of the complex worldwide reality through a polemical exchange of learned opinions and arguments across various borders and divisions. Instead of having a series of monologs, I wanted to infuse into the book as much dialog as might be possible. I wished also to pit "insiders" as contributors against "outsiders" as commentators (either in terms of personal origin, ideological affinity, citizenship, political affiliation, professional interest, etc.) or one "insider" against another "insider." These parts of my original design proved to be very time- and effort-consuming. In addition to personal contacts at various international gatherings for over six years, it required several hundred letters and many cables and telephone calls. My coeditor and I learned the hard way how different the levels and styles of communication culture are among social scientists in various parts of the world: from prompt substantive replies and timely honored commitments to no replies at all. Whenever the desire to achieve wider representation and to control for possible biases of various kinds clashed with the criterion of competence, we always opted for the latter. The understandable wish of any editor to attract already well-known

6

Introduction

and "catchy" names had to be, in several cases, put aside for substantive and procedural reasons. The end result of often conflicting considerations and cross-pressures were substitutions, several omissions, and compromises. For example, one planned chapter (on Asia) fell as a casualty while another (on Africa) came out differently than originally intended. The completion of the project might well have proved impossible were it not for the felicitous entry at a critical moment of my colleague and friend, Jim Seroka, as coeditor. This burden-sharing proved to be in itself a fruitful exercise in international across-the-ocean cooperation. Bonds of old friendship with the publisher Lynne Rienner also made the completion of the project in some respect a more pleasant task. Substantively, the objective of this book is to take stock and revaluate some frequently used classifications and typologies in comparative politics in order better to understand the complex reality of the contemporary world. Our level of analysis is sovereign states worldwide, and the timeframe is the post-1945 period. We also decided to examine epistemological and methodological foundations for classifications and to look back at premodern classifications of states and clarify their relationship with those we use today. In spite of more sophisticated appearances, many contemporary classifications clearly descend from much older ones. Some of them boil down to the oldest and simplest dichotomous classing of polities into "ours" versus "theirs" or, in a modified religious form, "godly" versus "ungodly." Every comparative study requires some classification, be it implicit or explicit. If properly developed and stated, classifications and typologies can help formulate research questions, establish the agenda for inquiry, and lead to the development of scientifically testable hypotheses and to theorybuilding. Without classifications and typologies, one can hardly expect further progress in comparative politics. Classifications and typologies are particularly important in this discipline because we lack a universally agreed-upon general theory of politics. The editors of this book and, we believe, also all other contributors share the view that the time has come to revaluate classifications and typologies of contemporary political systems as well as the assumptions and theoretical constructs on which they were based. From the time when Aristotle and his disciples surveyed polities of the ancient eastern Mediterranean, there have been many attempts to class known existing states. Every political system is, strictly speaking, sui generis, because its combination of features and its dynamics of change are nowhere exactly replicated. In each political system, exceptional personalities and unique events shape for a considerable duration the predominant values, sensitivities, issues for political debate, institutions, and patterns of conflict resolution. They also play an important role in creating "national styles" in politics, et cetera. Yet, all these idiosyncracies do not obviate the appropriateness of comparing systems. In order to grasp the complex reality, political thinkers have long used,

Anton

Bebler

7

implicitly or explicitly, models of political systems. Many of these models were derived by analogy from natural phenomena and manmade objects. Thus, flow, wheel, balance, weight, pendulum, and later clocks, and the human anatomy were used as heuristic tools for understanding political systems and for differentiating one from another. The famous ancient Greek trichotomies—monarchy, aristocracy, democracy and tyranny, oligarchy, demagogy—could and possibly did have a pyramid as an initial model. Among modern additions to model-building in comparative politics, one ought to mention Karl Deutsch's (1963) model of political systems based on information flows and derived from cybernetics. Another U.S. author, Morton Gorden (1972), used as a basis for classification the two contrasting "conflict management configurations" (command and pluralist). He arrived at a twenty-four-cell matrix in which four basic types of political systems (pluralist-participant, pluralist-mediated, command-mediated, command-dominated) could have numerous subtypes along the following dimensions: goals; rules of exposure; spokesmen; organization for action; bargaining; and implementation. The model of the internal combustion engine seems to be behind Gorden's scheme, which predicates conflict management as the most important feature of political systems. Milton Rokeach's (1973) social-psychological infusion into political analysis, with his model of political variation of two basic social values (equality and freedom), lends itself well to differentiating political systems, at least in a part of the world. Nevertheless, contributions to this book indicate that these and a number of other relatively recent innovations have had little impact on comparative politics. Concepts derived from, or associated with, sociology and economics have been widely used in some communities and groups of political scientists for comparing states and political systems. Marxist-oriented comparativists, mostly in socialist, communist-ruled states, still mix the criteria of ruling class and dominant ideology as the main classification tools, arriving at an improper dichotomy (capitalist versus socialist states). Many non-Marxists find useful the various measures of economic and social development when dealing with the "Second" and the "Third" worlds as well as with the NorthSouth contrasts. Italian political scientist Alberto Marradi treats, in his contribution to this book, an old controversy in comparative studies. Are classifications, typologies, and taxonomies heuristically valuable in themselves or are they merely empty boxes? The famous Mendeleev table of chemical elements has been used not only for explanation but has helped in the search for hitherto unknown chemical elements. Is it unreasonable, then, to have similar expectations that classifications of political systems may point out possible mutations in individual political systems or may indicate possible configurations of political systems that have not so far been observed in

8

Introduction

empirical reality? These expectations cannot be met without empirical v a l i d a t i o n o f underlying theories, as was so in M e n d e l e e v ' s case. Classifications and typologies are no substitutes for theory, but they are tools for detecting regularities in the observable universe. It is certainly improper to use them in order to imply, and still less to " p r o v e , " cause-effect relationships or to launch them for ideological justification. Classifications also ought not to mislead by drawing attention to formal structures of governance and away from informal power configurations and processes. The approximately 175 contemporary political systems could obviously be grouped in a large number of ways. However, a grouping could qualify as classification only if explicit criteria along dimensions of fundamental importance for the functioning of political systems were consistently used. A classification is expected to be pertinent, "natural," simple, universal (in time and space), parsimonious, and heuristically productive in facilitating discovery of new empirical entities. The rules of comprehensiveness, exhaustiveness, and mutual exclusiveness of classes have been already mentioned. Multidimensionality, mutations, and abrupt and often unexpected changes pose great challenges to comparative politics, much more so than to many other sciences. Thus, in the seventy-plus years of its development, the Soviet political system has experienced at least seven distinct political regimes even though the profile of the "ruling class," the dominant ideology, the institutional framework, and constitutional arrangements have changed only modestly. Yet, we expect from classifications that they will help us to differentiate not only between groups of political systems, but also within them. Very few classifications can actually meet all these demands or can prove themselves durable tools for political analysis. We set out in this book to review the usefulness of frequently used classifications and typologies of contemporary political systems. These reassessments were carried out either directly (i.e., by applying classifications to the entire universe of contemporary political systems) or indirectly, but not less cogently (i.e., by inquiring into fundamental properties and differences within large groups of states such as democratic, authoritarian, socialist, and "Third World"). A l l internationally influential classifications of political systems have been based on concepts that were originally developed in Europe. In order to control for the "Eurocentric" bias, we asked several contributors to review contemporary political systems in culturally, socially, and politically different groups of states (sub-Saharan African, Arab, and Latin American). Their efforts were expected to test the usefulness of originally European concepts for comparative political analysis in other environments. The design of the book, out of necessity, leaves out several areas and dimensions of classifying contemporary political systems or deals with them only in passing or indirectly. We believe that future cooperative efforts in international political science will allow us to cover the remaining grounds

Anton

Bebler

9

and to advance our knowledge in other respects as well. We hope that our book will stimulate this development.

References Almond, Gabriel A., and G. Bingham Powell, Jr. (eds.) 1974; 1980. Comparative politics today: A world view. Boston: Little, Brown. Banks, Arthur S. (ed.) 1980. Political handbook of the world: 1980. New York: McGraw-Hill. Bertsch, Gary, Robert P. Clark, and David M. Wood. 1978; 1982. Comparing political systems: Power and policy in three worlds. New York: John Wiley. Deutsch, Karl W. 1963; 1966. The nerves of governments: Models of political communication and controls. New York: Free Press. Eckstein, Harry, and David Apter (eds.) 1963. Comparative politics: A reader. New York: Free Press. Görden, Morton. 1972. Comparative political systems: Managing conflicts. New York: Macmillan. Rokeach, Milton. 1973. The nature of human values. New York: Collier.

2 On Classification Alberto

Marradi

The purpose of this chapter is to achieve a conceptual clarification of the semantic field associated with the term "classification" and to assess the proper role of classificatory procedures within the social sciences. In the first two sections, I demonstrate why a conceptual clarification is necessary, and I make a modest contribution toward that end. In the third section, I argue that the role of classification has been improperly assessed in several arguments; in particular, by those who credit it with ontological capacities and tasks, by those who see classificatory procedures as an old-fashioned activity to be abandoned in favor of more "scientific" measurement, and by those who blame the retarded development of an "explanatory" social science on the undue attention paid to classification by many of the founding fathers of sociology and its cognate disciplines.

Classification as an Operation The need for conceptual clarification stems from the fact that the term "classification" is indifferently used for several (intellectual) operations and for several different products of such operations. Let us start with the former. By analyzing the formal definitions and implicit acceptations of the term "classification" as an operation, three main families of meanings may be clearly recognized: 1. Classification as an intellectual operation whereby the extension of a concept at a given level of generality is subdivided into several (two or more) narrower extensions corresponding to as many concepts at a lower level of generality: this subdivision is obtained by stating that an aspect of the intension of each of the latter concepts is a different partial articulation of the corresponding aspect of the intension of the higher concept.1 Similar operations, which are performed simultaneously on several aspects of a concept or successively on concepts of decreasing generality, 11

12

On

Classification

belong in the same family of meanings. 2. Classification as an operation whereby the objects or events of a given set are grouped into two or more subsets according to the perceived similarities of their states on one or, more frequently, several properties: subsets may be successively grouped into subsets of wider extension and higher hierarchical level. 3. Classification as an operation whereby objects or events are assigned to classes or types that have been previously defined: in most instances, classes have been defined through the first process; however, they may also have been defined through the second process and then brought to bear on objects or events not belonging to the original set. In order to distinguish among the kinds of classification, operations of family 1 could be named "intensional" classifications, and operations of family 2 "extensional" classifications. The difference between the two is apparent: 1 is a prerequisite for 3 and, in most cases, family 1 operations will have been performed with the aim of making possible one or more operations of family 3. In clearer terms, classes or types are usually formed in order to have empirical instances assigned to them. The difference between operations in families 2 and 3 is also apparent, in that classes or types are the starting point of 3 while they are the final target of 2. Intensional

Classification

The intellectual operations we have just characterized as belonging to 1 are almost universally labeled "classifications," with the exception of Scheffler (1967/1983, 44ff), who uses the term "categorization." Sometimes, the activity is labeled "classifying" in order to distinguish it from activities of 3, which are labeled "classing" (Riggs 1979,180). From a semantic point of view, intensional classification may be seen as a process of conceptual elaboration (Glaser 1978; Turner 1981). The concept whose intension is articulated in one of its aspects is "explicated" or "unpacked"; hence, it is enriched and clarified (Sartori 1970). Concepts corresponding to individual classes are either formed or clarified by the definition of their boundaries with contiguous concepts. Different terms or expressions are allocated to each class concept, and the concept-term ties are fastened, as for all linguistic phenomena (Saussure 1916), by the implicitly oppositive nature of any systematic allocation. 2 In this sense, Hempel is perfectly correct in stating that classification is "a special kind of scientific concept formation" (Hempel 1967, 139; Sandri 1969, 85-86). From the logical point of view, once the intension of each class concept has been defined clearly (also by opposition with the intensions of each other class concept), it specifies the "necessary and sufficient conditions of membership" in the class, "by stating certain characteristics which all and only the members of the class possess." Each class is therefore "defined . . .

Alberto

Marradi

13

as the extension o f ' the correspondent class concept (Hempel 1965,137). A warning is in order: The operation we have just begun to describe is but the most elementary (in the sense of basic) in the family of intellectual operations we proposed to designate as "intensional classification." Before covering other forms, we need to introduce a few additional terms and the associated concepts. The most important instrumental concept in classification is still largely known under its Latin label: fundamentum divisionis (literally; the basis of division [Sartori 1984, 73]), though it is sometimes given an English label (e.g., "classificatory principle" [Lazarsfeld & Barton 1955, 84]). These expressions stand for the particular property, among the many composing the intension of a concept, which is being articulated in its various states in order to create the class concepts. The concept of political system is a good example. One of its properties is the principle used to legitimate rulers. If that property is taken as the fundamentum divisionis, the classes might be theocratic, autocratic, plutocratic, democratic, and so on. However, if the degree of autonomy of a state's territorial components is taken as the fundamentum divisionis, then the classes will probably be unitary, federal, confederal, and so on. As the chapters in this book testify, there is no dearth of fundamenta divisiones for classifications of states and political systems. A second instrumental concept in classification is the level of generality of classes. Two classes (X and Y) are at the same level of generality when: (a) the extension of X (total number of concrete instances classed under X) is not part of the extension of Y, and vice versa; and (b) the extension of X is not part of the extension of a class Z that is at the same level of generality as is Y, and the same can be said for Y. While the violation of requisite a brings about a grossly incorrect classification, requisite b is violated for practical purposes in several actual classifications. Consider, for instance, the following classification of behavior at the latest elections, which could have been used in a "closed question" of a survey: (1) voted for party A; (2) voted for party B; (3) issued an invalid vote; (4) issued a blank vote; (5) did not go to vote. Assuming that only parties A and B are running, this is a perfectly workable classification. It involves, however, no less than four levels of generality (i.e., went to vote/did not; expressed a vote/did not; expressed a valid vote/did not; voted for A/for B). In this instance, the violation of requisite b will have no negative consequences, because a correct reduction of a taxonomy has been operated. A third instrumental concept in classification is "mutual exclusiveness." This must be construed as a property of any pairs of classes of a well-formed classification (the above example has shown that "same level of generality" is also a property of couples of classes, but not a strict prerequisite of being well-formed). Classes X and Y are mutually exclusive if no single instance is

14

On

Classification

a member of the extensions of both X and Y. Obviously, violation of requisite a for sameness of level also entails lack of mutual exclusiveness. Such a negative consequence does not obtain, however, if only requisite b is violated. Most violations of mutual exclusiveness are due to the adoption of more than one fundamentum divisionis for the same classification. Consider, for instance, the following classification of data relevant for political science (Deutsch, 1966): (1) aggregate data from censuses and other official gatherings; (2) data on opinions and attitudes from polls and interviews; (3) data on voting behavior of both electors and legislators; (4) data on elites (background, recruitment, and power); (5) secondary data, obtained through statistical treatment of other data; and (6) historical data. While Deutsch's list continues, it is already apparent that he is using as many different fundamenta as he has classes (i.e., the type of property to which data refer, the level of aggregation of the reference unit, the type of data-gathering process, whether the data have received statistical treatment, whether they concern a sufficiently remote period in the past or are more recent, and so on). By combining all such criteria, or a part thereof, a huge typology would have been built; Deutsch probably intended his classification as a drastic "reduction" of that typology. Whatever he meant, however, it is so sloppily conceived that more than half the possible pairs of classes (e.g., 1 and 5, 1 and 6,2 and 3,2 and 4 , 2 and 5) are not mutually exclusive. While it is not uncommon to find classification patently defective as to mutual exclusiveness of classes due to structural, so to speak, flaws in the formation of class concepts, most classifications may present serious problems on that account when it comes to the actual assignment of concrete instances (i.e., to operations of family 3). Those problems, and associated remedies, will be mentioned when operations of family 3 are discussed. A fourth instrumental concept in classification is "exhaustiveness," a property of the complex of classes. Given the particular property that is being taken as fundamentum divisionis, a classification is exhaustive if each possible state on that property has been allocated to one of the classes. In extensional terms, if we have classified forms of political systems, any actual political system must be a member of one of the classes—regardless of what fundamentum has been chosen. Taken together, the two conditions of exhaustiveness and mutual exclusiveness provide that every concrete instance is assigned to one and only one class. Mutual exclusiveness is absolutely warranted only if one of the classes is so defined as necessarily to include all the instances not included in the other classes. The intellectual operation that we are calling "intensional classification" may be performed by taking more than one fundamentum divisionis simultaneously into account. Rather than a one-dimensional set of classes, an w-dimensional set of types is then created, where n is the number of

Alberto

Marradi

15

fundamenta. A type (from the Greek typos: cast, model) is therefore a concept whose intension is the intersection (in the set-theoretical sense) of the n classes that are combined to form it. On logical grounds, it will often be discovered empirically that the fact of being the combination of n classes, each belonging to a different dimension, will endow the type with features not implied by the intension of each of the n classes. As the simplest possible example of this kind of intensional classification, take Apter's (1965) typology of regimes. It combines two fundamenta: predominant values with two classes (instrumental and consummatory) and type of authority with two classes (hierarchical and pyramidal). Each type of regime is the logical product of one class on the value dimension and one class on the authority dimension (e.g., pyramidal with consummatory values). The semantic space as structured by those intellectual operations (logical products) was first conceptualized by Hempel and Oppenheim (1936) and then labeled "property space" by Lazarsfeld (1937). The complete set of types (i.e., the product of this form of intensional classification) is often also called a classification, or a "combinatory classification" (Granger 1960). I strongly prefer the label "typology," also often used, because I consider it semantically unwise to use the same term for an operation and its product. An intensional classification can be performed, so to speak, in a chain. After dividing the extension of a concept in classes by applying a fundamentum divisionis, one, several, or all the classes can further be divided by applying other fundamenta, almost ad infinitum. This procedure was first conceptualized by Plato as diairesis (literally, dichotomous choice: each concept was divided into the two concepts of immediately lesser generality). Aristotle introduced the analytical distinction between ghene (the concept whose extension is being divided) and heide (each of the classes resulting from that division). The distinction has survived in Western culture under the Latin labels, genus and species. The terminological and other aspects can be clarified by a symbolic example. Take a concept of high generality: A. By articulating a certain aspect of its intension (the fundamentum divisionis), three less general concepts are formed: AA, AB, AC. In this passage, A is the genus, while AA, AB, and AC are species. Then, an aspect of AA's intension is articulated, and four still less general concepts are formed: AAA, AAB, AAC, and AAD. In this passage, AA is the genus, while AAA, AAB, AAC, and AAD are species. Notice that the fundamentum divisionis in this passage will usually be different from that in the first, but it may also be a finer-grained articulation of the same fundamentum. Usually, if AA is being articulated, then AB and AC are, also. The fundamentum need not be the same as that used for articulating AA (i.e., the concept at the same level of generality), while it may be the same as that used for articulating A, or a concept at a lower level of generality.

16

On

Classification

Aristotle's classification of animals used a different fundamentum for each level of generality, and the same for all classes at the same level. Albert the Great did the same in classifying plants, or, to take a more recent example, Dufrenoy (1845) in classifying minerals. On the other hand, the rigor and elegance of similar solutions may cost a reduced efficacy of the product. While the analytical distinction between genus and species is very useful, a general term for any subdivision at any level of generality (i.e., AA, AAB, AABC, etc.) is also useful. Some natural scientists (e.g., Mayr 1969) use the Greek term "taxon" (pl. taxa), meaning order or class, which is very appropriate for its genetic link with "taxonomy" (literally, the law of orders or classes), the name that is frequently given to the complex of taxa (i.e., to the product of this form of classification). Other expressions are used, such as "articulated classification" or "multilevel classification" (Lazarsfeld & Barton 1951), "Linnean classification" (Granger 1960), "hierarchical classification" (Sartori 1984), or "hierarchical classificatory system" (Sandri 1969). I believe, however, that the Greek term is most appropriate, not only to honor the Greek origin of the concept and operation, but also because the alternative Latin term (classification) is, as I stated, better preserved for the operation and not used for the product as well. In summary, intensional classification is an operation whereby the extension of a concept is subdivided into the extensions of several concepts of lower generality. This may be done by articulating a single aspect of the higher-level concept (i.e., by one fundamentum divisionis), by combining two or more fundamenta (in which case a typology, made up of type concepts, will be obtained), or by a repeated subdivision of concepts at progressively lower levels. In the last case, a taxonomy, made up of several layers of taxon concepts, will be obtained. Each taxon is a species vis-à-vis the more general concept whose intension it articulates, and a genus vis-à-vis the more specific concepts that articulate its intension. Extensional

Classification

Earlier, I identified a second family of operations that usually goes under the name "classification." An operation of this family would group the objects or events of a set into subsets according to some criterion related to the perceived similarities of their states on one or, more typically, several properties. Bunge (1967, 75) correctly remarked that both analysis and synthesis are exerted in the process, as the elements of a set are both discriminated and grouped. However, very few stress the discriminating function (Sandri 1969, 81), and most definitions concentrate on the grouping function (e.g., Blalock 1960/1970, 28; Capecchi 1964a, 292), and, in particular, on the reduction of complexity (Kemeny 1959/1972, 195; Lazarsfeld & Barton 1951; Calogero 1949). Several definitions underline the need to maximize homogeneity

Alberto

Marradi

17

within classes and heterogeneity between classes (Lenzel 1938, 32; Blalock 1960/1970, 28; Pelloux 1967, 1448; De La Vega 1970, 117). Everitt (1980, 75) singles out the final stage of the process: "In the most general terms, classification is the process of giving names to a collection of objects which are thought to be similar to each other in some respect." While Blalock (1960/1970, 28) is correct in claiming that objects may be grouped by taking only one characteristic into account, most extensional classifications are operated on the basis of several characteristics. Accordingly, extensional classifications will usually form types rather than classes. One remote intellectual origin of the procedure might be found in Aristotle's criticism of Platonic diairesis as based on one character only (Parts of Animals, 64b, 21-644a, 10). However, an extensional view of classification waited for at least twenty centuries before acquiring intellectual dignity. This might be suiprising, considering that children seem to learn the concept of classification through a series of experiences and operations that are unquestionably extensional (i.e., centered on grouping objects) rather than intensional (i.e., centered on articulating concepts) (Piaget & Inhelder, 1959). In my opinion, this belated development depends on the fact that, in order to be somehow formalized from a spontaneous activity into a respectable intellectual operation within a scientific discipline, extensional classification had to wait for the development of another intellectual tool (i.e., the idea of orderly recording of the states of a vector of objects on a vector of properties). An extensional classification can, in fact, be performed with guarantees of intersubjectivity only if the following are exactly established: a. b. c. d.

e.

/

The membership, in terms of objects/events, of the set to be subdivided The array of properties in terms of which the internal homogeneity and mutual heterogeneity of classes are to be maximized The procedures by which states on properties under b are to be identified and coded3 The criteria by which differences in states on each property are to be evaluated (are all differences to be equally evaluated, or may they be ordered in terms of magnitude, or even measured?) A logical and/or mathematical formula combining the differences on all properties considered into a single "measure" of difference between any pairs of objects/events4 A set of decision rules as to how classes are to be formed, if only in terms of the general distance function, or by also taking into account maximum allowed differences on individual properties, if there are to be minimum and/or maximum limits on absolute or proportional membership in a class; and as to how "chains" are to be treated

18

On

Classification

(where each object/event is sufficiently similar to other objects/events adjacent along the chain, but not similar enough to [some] more remote members), and so on The formula under e may include criteria for: e'.

e".

excluding from consideration some properties that are not discriminating enough or otherwise fail to meet any other requirements differentially weighting some properties based on their superior discriminating power, theoretical importance, or any other asset whatsoever

The rules under/may include criteria for linking already formed classes with other classes or isolated objects. If this regrouping is performed by levels (i.e., second-order classes are allowed only after all possible first-order classes have been formed, and the same for all higher-order classes, if any), the product will appear identical to a taxonomy produced by intensional classification. If, on the other hand, higher-order classes are allowed to form, while lower-order classes are still being formed, the product will show a closer resemblance to the actual branching of a tree. The term "dendrogram" (from Greek dendron, tree) is currently used for the graphical presentation of the latter structure. A controversial question is how many properties are to be considered under b. Considerations of parsimony and elegance would tend to keep that number down.5 On the other hand, it may be remarked that "increasing the number of variables increases the probability of correct classification" (May 1982, 43). Since the concept of "correct classification" is rather questionable, a better argument might be that, with a formalized approach, there is no way to consider the information on the properties of objects/events unless those properties are included in the matrix. Therefore, leaving properties out of the matrix entails the loss of an indeterminate amount of information. Criteria of maximal information and of parsimony can both be satisfied if the former inspires the insertion of properties in the matrix, and the latter inspires the selection of those variables that, having been found discriminating or otherwise useful, are entered into the formula for computing the distance function. By reverting to the terminology of intensional classification, many potential fundamenta divisiones must be explored, but only a few actually need be retained. Extensional classification builds a typology in which types are the logical product of each class on each fundamentum (i.e., of each coded state on the variables considered in the distance function). We do not agree, therefore, with Galtung (1967, 72) when he states that in (intensional) classification "tht fundamentum divisionis may come

Alberto

Marradi

19

afterwards, not prior to the division, as when a social scientist looks at a list of occupational data, analyzes the distributions, and decides how to divide them into strata." The example is inappropriate, because "occupational" is already a fundamentum divisionis. Galtung is correct, however, concerning those informal extensional classifications that we operate in our daily life when we instinctively first group objects and then make explicit the criteria we followed in grouping. But as a scientific procedure, extensional classification needs something like a matrix organization of information. It needs it more strictly as the operations grow more complex because of an increasing number of objects/events and/or properties. It needs it inescapably if computers are to be employed. And, as I stated above, the variables in the data matrix are nothing else than fundamenta divisiones. It may be considered more than a coincidence that, approximately contemporaneously, protostatisticians Conring and Achenwall adopted rudimentary matrix presentations of their data on the military and economic resources of rival German states, and botanist Michel Adanson stated that "all parts and qualities, or properties and faculties of plants . . . barring not even one" ought to be considered before attempting a classification (1763, clviclviii). Along with this idea, Adanson operated extensional classification and produced taxonomies based on the rate of equal states on the total of properties considered between any two plants (1763, vol. 1). We ignore whether Adanson availed himself of some form of matrix to organize his data or if he used some more painstaking procedure. Present specialists of extensional classification, who use thousands of computer iterations to produce the best classification possible of the objects in a given data matrix, correctly recognize Adanson's priority in imagining a distance function that is based on the rates of equal/different states, and even label "Adansonian" the form of classification they operate (Sneath 1957; Sokal 1958). The possibility of computing such distance functions for a large number of objects obviously depends, however, on a matrix organization of data. The data matrix as a tool is now taken for granted in several scientific disciplines; however, like many other simple and basic intellectual tools, it was no mean and no easy achievement. Now that the two main prerequisite intellectual tools (data-matrix organization and distance function) have been molded, and practical tools (i.e., computers) have become so powerful and available, algorithms for distance functions and solutions to all sorts of technical problems have proliferated.6 New labels have also been proposed for the operation over and above the old ones (i.e., classification and taxonomy). Among them are "numerical taxonomy" (Sokal & Sneath 1960), "class formation" (Capecchi & Moller 1968), and "cluster analysis." In my opinion, "numerical taxonomy" should be reserved for the product, or rather, for a particular form of product, and "cluster analysis" for a particular technical procedure. "Class

20

On

Classification

formation" has the required generality and is correctly referred to as an operation having intellectual components. However, it might very well be used for the kind of operations I dealt with at the beginning, which, on the contrary, are duly distinguished if the label "intensional classification" is assigned to them, while the label "extensional classification" is assigned to the operations dealt with in the present section. Classing The third and last family of classificatory operations identified is the assignment of objects/events to classes, types, or taxa that have been previously defined. This kind of operation is often called simply a "classification" (e.g., Frake 1969, 29; Oppenheim 1975/1985, 217); however, several authors remark upon the importance of using a label to allow one to distinguish this activity from other kinds of classification (Capecchi 1964b; Cormack 1971; Fox 1982). Other labels proposed or used are: "categorical assignment" (Scheffler 1967/1983, 49), "diagnosis" (Hempel 1965, 138; Capecchi 1964b, 294), "diagnostics" (Capecchi 1964a, 292), "class identification" (Capecchi & Moeller 1968, 63), or simply "identification" (Capecchi 1964b, 294). The kind of category (class, type, or taxon) to which assignment is made does not lead to important differences in the classing operation. Only one fundamentum divisionis needs be considered in the assignment to classes. Two or more are necessary in the assignment to types, and a series of fundamenta are needed in the assignment to taxa. Increasing complexity might be expected as one passes from the former to the latter, but, in practice, this is not necessarily true. In some circumstances, it may even be found that assignment to taxa or types is easier than is assignment to classes. Far more important are the consequences on the assignment procedure of the kind of classification through which the categories have been formed. Extensional classification requires the preliminary definition of a set of objects/events on which the operation is to be performed. Performing the classification entails the automatic assignment of those objects/events to the resulting types. If the objects/events that we want to class do not belong to the set on which the extensional classification was performed, their states on the properties of the matrix must be ascertained and coded. This preliminary step, however, involves a problem: Should we consider all the properties of the original matrix, or just those that have been retained in the distance function? This may be considered a trivial problem in itself (and no problem at all if all the properties in the matrix were considered in the distance function), but its solution depends on another problem that is unavoidable and difficult to discount: Should we use for the new objects/events the same distance function that has been established through an elaboration of entirely different objects/events? The answer may depend on a great variety of motives, including the

Alberto

Marradi

21

belief that the classification already arrived at is the "natural" one and, therefore, will be retrieved on any set of objects/events of the same domain. On the other hand, more empirically oriented researchers might be tempted—since the data matrix has been filled, and the computer is there to do the tedious work—to try to see whether the distance function comes out the same. The next question, of course, is: What if the distance function comes out significantly different? Apparently, the conclusion that different typologies are good for the two sets of objects/events, though the properties considered were the same, can hardly be avoided. At this point, a real chain of questions ensues: How has the assignment of objects/events to the first or the second set been determined? May the sets be considered random samples from the same universe? Was any form of randomization performed? Or are there significant differences, known a priori, between the two sets? And how are those differences related to the difference empirically found between the two distance functions? Most of the above questions h^ve a familiar ring for readers interested in philosophy of science or in experimental design, comparative research, data analysis, and the like. I believe that extensional classification is bound to face all of them, and probably a few others, unless it is performed as a one-shot affair (i.e., never attempting replications on different sets of objects/events). Indeed, the paucity of replications reported in the specialized literature makes one suspect that this is exactly the way extensional classification has been conceived and performed so far. Assignment of individual instances to categories formed by intensional classification stages is a quite distinct situation. The problem of classing objects/events differently on two separate occasions loses most of its drama because, after an intensional classification, states on one property (or combination of states on few properties) are classed, rather than whole objects/events. It is therefore perfectly normal that objects/events are grouped differently if different fundamenta divisiones are considered. In classing objects by color, we take into account only the states of objects A, B, C, D . . . on the property "color." We may then class the same objects by shape, and most likely the latter fundamentum divisionis will discriminate differently. The same is true when we assign objects/events to types or taxa rather than to classes. Groups resulting from classing nation-states into a typology formed by secularization and economic development need not be the same as groups of the same nation-states in a typology formed by political alignment and degree of military aggressiveness. A problem remains, of course, if the same objects/events are classed differently (by more than one classer, or by the same classer on separate occasions) into the same classes, types, or taxa. That, however, is a comparatively minor technical problem, which does not challenge the status of this kind of classification.

22

On

Classification

While an intensional classification scheme enjoys the above-described advantage over an extensional classification scheme (and the associated advantage of dealing with one or a limited number of properties only), it lacks the empirical check supplied by an at least partially automatic procedure. Everything hangs on how effectively categories have been formed a priori, and how accurately rules are provided to fit the abstract scheme to the concrete instances. Fit is nonproblematic only with such objects as logical and mathematical abstractions. For example, integers are neatly classed into prime numbers or multiples. But things are otherwise with empirical objects/events: "In scientific research, however, the objects under study are often found to resist a tidy pigeon-holing of this kind" (Hempel 1965, 151). Mutual exclusivity "is a property of the [classification] systems, and it is seldom also a characteristic of the objects to be classified" (Sandri 1969, 84). "A given member of the population may manifest traits belonging to different sets, so that classifying [it] as an instance of only one type . . . takes on an arbitrary aspect" (Tiryakian 1968, 179). In this connection, the role of careful operational definition of the properties adopted as fundamenta divisiones cannot be overrated. Such a definition should progressively incorporate and assimilate the decisions made in classing single borderline cases. Following judicial wisdom, borderline cases met earliest during the classing operation should be allocated "on the narrowest possible grounds," and early precedents should be construed narrowly, in order not to preempt subsequent decisions by rules too large in scope. Conversely, after a sizable portion of the objects/events have been processed, the microrules already established in solving individual instances need to be generalized, coordinated, and systematized. In principle, it is advisable to allow for feedback from the classing operation to the (intensional) classification stage. In other words, if some of the classes exhibit a poor fit with the empirical domain under study, the outcome of the previous intensional classification should be revised in order to improve the fit with the help of the information gathered. 7

Classifications as Products Classification Schemes The structurally simplest product obtains when only one fundamentum divisionis has been considered in an intensional (or, exceptionally, an extensional) classification. This kind of instrument (i.e., a list of classes) is almost universally called a "classification." 8 Nevertheless, faithful to our principle of not using the same term for an operation and its product, I shall adopt "classification scheme," following Berger and Zelditch (1968,448) and Fox (1982, 127). I suggest that this expression should be used whenever the

Alberto

Marradi

23

context does not make clear enough that one is speaking about the product. Classification schemes are built by articulating one fundamentum divisionis only. This is a necessary, but not a sufficient, condition of mutual exclusivity of any pairs of classes. The complementary role of operational definitions, including detailed rules for assigning borderline cases, is important; such rules, for example, play an essential part in field- or officecoding of open-ended interview materials in the social sciences (Montgomery & Crittenden 1977). When a list of preselected answer categories is provided, the requirement that only one answer be chosen is a consequence of the principle of mutual exclusivity (Galtung 1967,13). When objects to be classified are logical or mathematical abstractions, both mutual exclusivity and exhaustiveness can easily be obtained with a modicum of care in the definition of classes. In most empirical situations, however, establishing a really exhaustive classification scheme would be a formidable task in the intensional classification stage, and having to deal with inexhaustive classification schemes in the classing stage would force classers consciously to operate incorrect assignments in order not to leave instances unclassed. Such serious problems are usually avoided by inserting one or more residual categories in the classification scheme. Resort to residual categories is at least as ancient as Platonic diairesis, whereby a genus was dichotomized into two species—one characterized by the presence, the other by the absence, of a certain trait. When absence was the only common trait of organisms designated under the latter species, that species was evidently just a residual category.9 A plurality of residual categories that jointly warrant exhaustiveness may be found, for instance, among the standard formats for precoded answers to survey questions (i.e., providing for such classes as "don't know," "no answer," "other," and "inapplicable"). While the last class is there to bring structurally tautological, but necessary, information, and the first two carry some information, the "other" category is hopelessly uninterpretable, since it may group wildly different answers. Totally uninterpretable residual categories should be avoided whenever possible. For instance, often, votes for minor parties are lumped together under the category "other parties," spoiling those data for any further analysis. This could be avoided if several residual categories were supplied, (e.g., "other conservative parties," "other progressive parties," "other confessional parties"). While one cannot expect the political position of a party to be so brusquely assessed by official publications, one may ask such an assessment from a "country expert" political scientist building and circulating an "ecological file" with election data. The advantage would come whenever one analyst wants to correlate the total progressive vote with any other territorially distributed variables. If one cannot make the residual category somehow meaningful, one

24

On

Classification

should at least avoid swelling it with cases. 10 In survey research field-coding of open-ended material, the "undecidable" or similar residual categories are swollen by interviewers who throw into that dustbin any answers for which they cannot decide between two classes. It is wise to encourage them to choose either of the two eligible classes, because a possible slight distortion of the intended meaning is preferable to a total loss of information. Earlier, I remarked that classes need not be at the same level of generality, provided that a taxonomy is correctly reduced; let me show what that means, taking an example from Thurstone (1931, 252). Suppose one survey researcher wants to submit a list of foreign nationalities to disclose respondents' feelings toward them. For practical purposes, he can correctly insert a class "South Americans" on a par with "French" or "Canadians." However, this should prevent his inserting "Brazilians," lest he violate mutual exclusiveness. In addition, a "paraphyletic" device, such as inserting both "Brazilians" and "other South Americans," would have a dubious psychological effect on respondents and, therefore, a dubious legitimacy (see note 9). In general terms, a taxonomy is correctly reduced whenever a genus and one or more of its species or subspecies do not show up as classes in the same classification scheme. The grouping of several species at the same level of generality into a single class is a question of opportunity. A good reason for creating classification schemes with classes at different levels of generality is the widely differing extension (in terms of number of assignable instances) that can obtain between classes at the same level. Take, for instance, a taxonomy of religious denominations. The first division should be between A (believers) and B (nonbelievers). Then A could be split into AA (believers in divinities) and AB (believers in spirits, i.e., animists). AA could be split into AAA (believers in monotheistic religions) and AAB (believers in polytheistic religions). AAA could then be split into AAAA (Christians) AAAB (Moslems), and so forth. We would then probably have AAAAA for Protestants, AAAAB for Catholics, AAABA for Sunnites, and so on. The fewer letters in the identification tag, the higher the level of generality. In order to somehow balance the extension of classes, while retaining exhaustiveness, a classification scheme of denominations should probably include several taxa of the fifth level (e.g., Catholics, Sunnites), several of the fourth (e.g., Shintoists, Israelites), probably two residual categories at third level (e.g., other monotheistic and other polytheistic), one taxon at the second (e.g., animists), and one at the first (e.g., nonbelievers). Mutual exclusivity and level of generality are the most basic kinds of relationships obtaining between the classes of the same classification scheme. Two other possible relationships are those of order and ratio. Both are associated with particular kinds of fundamenta divisiones and operational definitions thereof. If we perceive the states on a property as ordered, we probably want that

Alberto

Marradi

25

order reproduced between the classes of the corresponding classification scheme. While that puts a constraint on the way classes are to be formed in the intensional classification stage, it makes no difference either for the classing procedure or for the characteristics of the resulting scheme. Classes must be based on one fundamentum divisionis only, and they must be mutually exclusive and jointly exhaustive. Ordered, like nonordered, classes need not be at the same level of generality. Suppose that an educational system provides for three levels—primary, secondary, and advanced—and that there are three different degrees (A, B, C) at the advanced level. One may order, in terms of some criterion, the classes as follows: primary, secondary, advanced type B, advanced type C, and advanced type A. The former two classes will be lower in order but higher in generality than the latter three, and they all will legitimately be part of the same classification scheme. Also, we may perceive the states on a property as aligned along a continuum (i.e., isomorphic to real numbers). However, since only a finite number of ciphers can be recorded, the continuum will have to be broken down by a unit of measurement, or by a scaling procedure. Finally, we may perceive the states on a property as countable and decide to count them. By this decision, as well as by adoption of either a unit of measurement (plus a rounding rule), or a given scaling procedure, we automatically form a classification scheme. Its classes may look like " 1 foot, 2 feet, 3 feet. . ." or like "fully agree, agree, uncertain . . ." or like "no son, one son, two sons . . ."; they still have—if the scheme is well formed—the basic properties of any other classification scheme. Order or ratio relationships between classes may be established in connection with certain assumptions about the nature of the property and certain characteristics of the operational definition. However, they do not replace mutual exclusivity or joint exhaustiveness. The list of permissible classes of an ordered scale or a ratio scale is, by all standards, a classification scheme. The ideal number of classes cannot be established in general terms. It is only reasonable to remark, as Kemeny (1959/1972, 195) and Lazarsfeld and Barton (1951/1967, 235) have done, that classes ought to be few enough to assure handiness, but not so few as to impair internal homogeneity. If a classification scheme is to be submitted to respondents in a survey to have them pick out the class they prefer, then a more compelling orientation comes from information theory: No more than seven items can simultaneously be kept in focus by most individuals. If asked to evaluate more, people will automatically ignore some (Miller 1956; Wallace 1961; d'Andrade 1981). Galtung (1967, 108) has claimed, drawing from a vast personal experience, that in mass surveys the limit should be five items, and the same figure is indicated by Schutz (1958) for coders of content analysis. According to my experience, even five might be optimistic for many survey

26

On

Classification

respondents. Natural human limitations are not so strict, however, if the classes are ordered, because order itself helps each respondent focus on a narrow range of classes relevant to his or her own answer. If, moreover, each class has a reduced semantic content, as is typical of classes linked by ratio relationships, then respondents' choice is easier, and information-processing limitations cease being particularly relevant. Human limitations are also less relevant when classers are not pressed for time and can proceed by progressive elimination. In that case, a fully detailed and sensitive classification scheme is to be recommended. After all instances have been classed, it will be perfectly feasible to reduce the number of classes by merging couples or subsets of them, keeping both semantic proximity and balanced extension in mind. Typologies

Typologies are produced when an intensional classification is performed by taking more than one fundamentum divisionis simultaneously into account. Types are the logical product of n classes (one per fundamentum), and logical products enjoy the commutative property of all products (Cohen & Nagel 1934, 123). This means that the order in which fundamenta are considered is irrelevant. The type of theocratic regime that also is authoritarian is the same as the type of authoritarian regime that also is theocratic. This differentiates typologies from taxonomies, which obtain when fundamenta are considered in succession. Like classes, types must be mutually exclusive and jointly exhaustive. Mutual exclusivity is violated as a consequence of any change in the list of fundamenta in passing from one type to another. If type X is "secularized and receptive political systems" and type Y is "centralized and aggressive political systems," type X and type Y cannot belong to the same typology, because an actual political system can be classed in both types. Residual categories can be used and are, in fact, used on any of the fundamenta. A type is conceivable that is the logical product of n residual categories. Although this type may be required to make the typology exhaustive, it will hardly be inteipretable in substantive terms. Since few, if any, actual instances will be assigned to it, it will be wise to merge it with other types in the reduction stage. Intensional classification is not the only way to produce typologies; extensional classification typically produces them. Typologies produced by the two types of operations can be compared, with intellectual gratification proportional to the rate of overlap in the list of fundamenta used. In an interesting comparison of that kind (Hudsoif 1982, 56-83), a cluster analysis was performed on a data matrix with 273 North American native ethnic units and a set of variables drawn from Murdock's Ethnographic Atlas (1967); a dendrogram was produced. Then the diagram was "cut" so as to obtain fifty-

Alberto

Marradi

11

five groups, and their membership was compared with Voegelin and Voegelin's (1966) fifty-five "language families"; a second cut at thirty-five groups permitted a comparison with Murdock's thirty-five "culture provinces"; a third cut at seven groups permitted a comparison with Kroeber's (1939) seven "culture areas"; and so on. A third way to produce a typology has been conceptualized by Lazarsfeld (1937), who called it "substruction of a property space." Substruction is made necessary by the fact that many scholars propose lists of types without explicating the fundamenta divisiones (as lamented by Andreski 1965, 163-174; Nowotny 1971; Pye 1968); the fundamenta have to be inferred by other students who extract them from the intensions of the types as described. Lazarsfeld (1937) performed a substruction of Fromm's four types of authority relationships; Lazarsfeld and Barton (1951/1967) "substrued" the property space underlying Durkheim's types of suicide and Kingsley Davis' types of social norms. Capecchi (1966) substrued the property space underlying Merton's five types of anomie and showed the logical possibility of four additional types, all having meaningful counterparts in actual behaviors. The importance of substruction has been correctly perceived by scholars aware of its methodological aspects. Berger and Zelditch (1968, 447) see it as coessential with classification itself: "The purpose of the classification . . . is to order the known types, showing how they are similar and different, . . . identifying the underlying system of variables that gives rise to the observed phenomena, and locating types not yet identified and studied that are generated by the same system of variables." Conversely, it must be recognized that the role of substruction is entirely dependent on the intellectual sloppiness of most scholars, who, unfortunately, often propose lists of types but do not explicate their fundamenta divisiones to build a full-fledged typology. Another intellectual operation, however, plays a more general role, which would remain crucial for typology-building even in an ideal world of widespread methodological awareness. It was first conceptualized by Hempel and Oppenheim (1936) and labeled "reduction of a property space." The crucial role of reduction stems from the fact that the number of types in a typology (sometimes referred to as its "power") is a multiplicative function of the number of classes in each of the fundamenta considered. This property of all products, including logical products, has three negative consequences for typologies: 1. The mere number of types is bound to be unmanageable, unless only two or three dichotomous fundamenta divisiones are considered 2. Some types are likely to be mere logical possibilities, devoid of conceptual interest for the discipline 3. Some types will be empirically empty; i.e., no or few instances will be

28

On

Classification

assigned to them in the classing stage (empty type sets will usually have a large overlap with the sets of uninteresting types, though they will be larger) Those drawbacks are, or might be, alleged as an excuse for sloppy typology-building, but reduction of a logically well-formed typology, rather than sloppiness, is the proper remedy for the drawbacks. As the term suggests, reduction consists in lowering the number of types, and, therefore, the intellectual complexity of the typology. The correct way to do this is not wholesale elimination of some types, which would destroy exhaustiveness, but the aggregation of two or more types into a type of wider extension and less articulated intension. Lazarsfeld and Barton (1951/1967, 267-270) distinguish among functional reduction (based on the relationships between fundamenta divisiones), pragmatic reduction (based on researcher's goals and aimed at a balanced extension), and arbitrary numeric reduction (which squeezes a typology down to a classification scheme by assigning numerical codes to types and performing mathematical operations on such codes). Capecchi (1966, 17) proposes a typology of reductions by combining two dichotomous fundamenta—whether or not mathematical instruments are used, and whether or not empirical evidence on the extension of types is considered. In my opinion, the researcher's goals are the foremost consideration. They frame the evaluation of the degree of semantic proximity between fundamenta and between classes thereof. This, in turn, controls the process of aggregation of types. Semantic considerations, however, find a limit in the desirability of balancing the extension of types. If two types comprise one third each of the total population of empirical instances, researchers are, or should be, reluctant to merge them, whatever their semantic proximity, into a single mammoth type. The recognized intension of a type may grow richer than is implied by its being the logical product of the intension of n classes. Let us examine that process through an example: The intension of the concept of "regime" is made up of a number of aspects, only a part of which are captured by explicit definitions. In operating an intensional classification of regimes, we pick up one of those aspects as the fundamentum divisionis and allocate each of its various facets to one of the classes we are forming. Let us suppose the fundamentum is the principle used to legitimate rulers, and the classes are four: theocratic; autocratic; plutocratic; and democratic. In principle, each of the class concepts formed retains all other aspects of the intension of the concept of regime except those prohibited by the mutual exclusiveness of classes (i.e., a democratic regime can have every trait any other regime has, except a theocratic, autocratic, or plutocratic legitimization). Suppose now that we want to produce a typology that combines form of legitimization with form of exercise of authority (our second fundamentum, with four classes: authoritarian; paternalistic; legal-

Alberto

Marradi

29

rational; and consensual). The typology will comprise sixteen types, each of which will, in principle, retain all the aspects of the intension of the concept of regime, except those prohibited by the mutual exclusiveness of types. However, examination of the characteristics of concrete instances of one type (e.g., a theocratic and authoritarian regime) may show that certain aspects (e.g., general elections) that are not prohibited (either by definition or de facto) by the intension of the composing classes seem to be de facto prohibited by their combination. It may, as well, show that certain features (e.g., education based on wealth) tend to characterize all the instances (or a proportion far superior to that found among the theocratic-nonauthoritarian and authoritarian-nontheocratic types). Of course, confidence in those empirical findings will be enhanced if a theoretical justification is found for it—as it usually is. The observed regular occurrence, or nonoccurrence, of a certain aspect enriches the intension of a type concept. Insofar as such enrichments are assimilated, they will tend to stimulate conjectures as to other possible developments, and that particular type concept will tend to acquire a degree of semantic autonomy from its constituent classes. In my opinion, this process supplies the link between the concept of type and the concept of "ideal type." The latter concept, proposed by Weber in a methodological essay (1904), is certainly one of the most fiercely debated in the philosophy of the social sciences (von Schelting 1922; Becker 1940; Watkins 1952; Hempel 1952a; Schutz 1954; Martindale 1959; Bendix 1963; McKinney 1966; Runciman 1972; Nowak 1976; Smelser 1976). Although most of that debate is beyond the scope of this chapter, Nowak's position is relevant here. He states that ideal-typic (or syndromatic) concepts include in their intension many characteristics beyond those necessary and sufficient to their identification. Such "surplus meaning . . . is equivalent to a set of theoretical propositions about the objects or phenomena . . . [S]yndromatic definitions (possessing . . . surplus meaning) are met in many sciences as another method of systematization of theoretical knowledge about some classes of phenomena" (1976, 140-41). We can imagine the ideal type as a type whose minimal intension, consisting of the logical product of the intensions of the constituent classes, has been so enriched that it has acquired full semantic autonomy. Another distinctive feature of ideal types, on which Weber insists, is that (in our terms) the intension has been enriched not only through empirical observation, but essentially through conceptual elaboration of observational evidence and unilateral accentuation (Steigerung) of a certain observed trait in order to build a conceptually coherent, but entirely ideal, picture (phantasie-massige Konstruktion). Taxonomies

If several fundamenta divisiones are considered in succession, rather than simultaneously, a taxonomy results. The order in which fundamenta are

30

On

Classification

considered is highly relevant. The taxonomy obtained by using property X to classify a genus and then property Y to classify its various species is by no means the same as that obtained by considering property Y first and property X afterwards. Because fundamenta are considered in succession, each single passage from a genus to its own species, at every step down the taxonomic hierarchy, is structurally similar to the building of a classification scheme. The species of the same genus must be mutually exclusive and jointly exhaustive, and residual categories may be employed to ensure exhaustiveness and to reduce the number of species. Whatever differences exist, they are due to the fact that every step is not an isolated operation but must be inserted into a hierarchical framework. As an obvious consequence, all species of the same genus must be at the same level of generality. This condition might entail both a proliferation of species and a great difficulty in achieving exhaustiveness. However, it can be relaxed by a wise use of residual categories and supplemented by the consideration that genera at the same level may be articulated each by a d i f f e r e n t fundamentum. In practical terms, this means that if the articulation of a given fundamentum produces too many species, some of them may be grouped into a residual category. This residual category will then be articulated at the next step, which would possibly retrieve the distinctions that were considered too fine at the higher level. A very elegant taxonomy is obtained when a different fundamentum divisionis is considered at each level of generality, while the same fundamentum is considered for all the taxa at the same level. That criterion, however, may impose an unbearable burden on actual operations and transform intensional classification into a formal exercise that does not fit actual phenomena. Moreover, the concept "same level of generality" is of questionable application across different branches of a taxonomy, when no relationship of inclusion, and not even an overlap, obtains between the extension of different taxa. Using the notation used earlier, we can say that taxon AB is at the same level of generality as taxon AC (as both are species of genus A) and that it is more general than taxa ABC and ABAC (which are one species and one subspecies of AB). We cannot, however, compare the generality of taxa ABC and ACAB, because they are not species of the same genus, and, thus, there is no overlap between their extensions. In the latter conditions, the concept of "rank" is more correctly used (Gil 1981, 1027). Rank means the number of steps down any branch of the taxonomy, starting from the summa genus (the concept on top of the hierarchy). We will therefore say that, in our taxonomy, ABC is higher in rank than ACAB—and we can determine this, in our notation, just by counting the number of letters in the tag. Ordinary extensional classification produces typologies; however, it may also produce structures that have various degrees of similarity to taxonomies,

Alberto

Marradi

31

if certain criteria are followed in grouping objects/events. If already formed groups are allowed to merge with other groups and/or be joined by objects/events not yet grouped, and higher-level groups are thus formed, the product will be a dendrogram. Dendrograms are considerably different from taxonomies, because, while in taxonomies all taxa of the same rank appear to split simultaneously, in dendrograms only one group splits at a time (i.e., there is a complete ordering between splits: No two splits are at the same rank). As a consequence, the display of a taxonomy will grow very rapidly in width as ranks of taxa are added, while a dendrogram will develop mainly in height. Further, there are usually no provisions for balancing the number of splits on different branches of a dendrogram, which as a consequence looks less harmonious and symmetrical than does a taxonomy. Moreover, if second-order groups are allowed to form only after all possible first-order groups have been formed, and the same for every higher order, then the product will look the same as a taxonomy produced by intensional classification. It must be remembered that in extensional classification several fundamenta are usually considered simultaneously in order to group (or regroup) objects/events. As a consequence, while the set of species of the same genus is a classification scheme, the set of groups that are merged into a higher-order group forms a typology because the lower-order groups are differentiated along several dimensions simultaneously. Only in those rare cases when groups are formed and merged by considering only one property at a time will the product of an extensional classification be a taxonomy from a structural point of view. 11

The Role of Classification The Essentialist Fallacy Building taxonomies seems to be a natural intellectual activity for mankind. Primitive classifications are not singular or exceptional . . . [ 0 ] n the contrary, they seem to be connected, with no break in continuity, to the first scientific classifications.... First of all, like all sophisticated classifications, they are systems of hierarchized n o t i o n s . . . . Moreover, these systems, like those of science, have a purely speculative purpose. Their object is not to facilitate action, but to advance understanding [Durkheim & Mauss 1963,

81]. The "primitive mind" builds taxonomies because they are believed to represent the hierarchical structure of reality (Levy-Bruhl 1910). The organization of society is the model that teaches us to think in taxonomic terms and is projected in taxonomies (Durkheim & Mauss 1963).

32

On

Classification

In classical Greece, intensional classification was practiced as an intellectual drill, but classification schemes and taxonomies, once produced, were believed to portray faithfully the inner structure of reality. Plato's five meghista ton ghenon (the greatest of all genera) are seen as an attribute of being, as well as of thinking. The same is true for the Pythagoreans' ten categories. During the Christian Middle Ages, Aristotle's detailed taxonomies of animals were believed to be a reproduction of the order of Creation. Also, the post-Renaissance taxonomies that superseded Aristotle's (e.g., Cesalpino's, Aldrovandi's, John Ray's) were conceived and received in the same spirit. Linnaeus (1735) wrote: "[T]ot numeramus species quot ab initio creavit supremum Ens [We enumerate as many species as were created by God]." While the ontological and the epistemological dimensions were regularly mingled, the axiological was often added. Linnaeus' taxonomy, for example, was also a hierarchy of degrees of perfection in which humanity was placed at the peak. Schlegel (1808) considered Indo-European inflectional languages inherently superior to affixational languages. The possibility of a "natural classification" was disputed by Buffon, Louis XVI's chief gardener at the eve of the French Revolution. His skepticism, however, was countered by Cuvier's reorganization of evidence in favor of the morphological stability of species, which had typically Aristotelian teleologic overtones. Cuvier's contemporary, Candolle, claimed that "classification... is truly natural" (1813, 91). The opposition between "natural" and "artificial" classification has been a recurring theme over the past two centuries. "Naturality" is gradually reinterpreted in terms of significant relationships with other classifications (Hempel 1965, 146-147; Kaplan 1964, 50), utility "for a wider range of inductive generalizations" (Gilmour 1940, 466), "systematic import" (Huxley 1940; Hempel 1952b; Sandri 1969), links with theory (Hempel 1952b; Bunge 1967, 83), or "projectibility of discriminating concepts" (Sandri 1969, 99ff). In short, the concept of "natural classification" has been transferred from the ontological to the epistemological domain. However, as Tiryakian (1968, 177) has remarked, "the reification of typologies is a frequent temptation and pitfall." In a typical manual of the neopositivist period, one can still read a statement as plain as: "A natural class is based on the fundamental character of things" (Lenzel 1938, 32). And one may suspect that, if the epistemological coat of paint were scraped off, substantial remnants of rusty essentialism would surface in the still widespread concern for "natural" classifications. The Scientific Fallacy The belief that genera and species of plants and animals had been handed down unchanged from the Creation to the classifier's days was buried by the theory of evolution. Phylogenesis replaced morphology as the governing

Alberto

Marradi

33

criterion, and taxonomies were reinterpreted as genealogies (Simpson 1945). That can be seen either as an epochal revolution or as an episode in the longterm swing of the pendulum between emphasis on structure and emphasis on genesis. Nevertheless, the merit of having imposed consideration for the diachronic dimension on a scientific community controlled by a static ontology can hardly be denied to evolutionists. Evolutionists, however, did not manage to avoid the same "category mistake" (Ryle 1938) of mingling the ontological and the epistemological planes. Being found guilty of an incorrect ontology, classification and taxonomy were discredited as intellectual tools and came to be regarded as dull and stuffy. Julian Huxley, the leading third-generation evolutionist, stated it was high time to cease being "bogged down in semantics and definitions" (quoted by Bethell 1986, 67). Even evaluative overtones were not missing. Supporting taxonomy and classification came to be regarded as a conservative undertaking, while a liberal attitude implied their rejection (Tiryakian 1968,182). The neopositivists initiated a subtler attack on the role of classification operations. The underlying message was that in the paradigmatic science (i.e., nineteenth-century physics), classification had long been superseded by measurement. If other sciences wanted to achieve the same status as physics, they must also abandon classificatory activities and turn to measurement. The diffusion of behavioralism and operationalism in the social sciences had already paved the way for an enthusiastic reception and amplification of this message, as indicated by an early and crude formulation in sociology: If such measurement [of social phenomena] is possible, then the path of the social science leads across the same difficult but not insurmountable terrain over which physics and other sciences have travelled to their present conspicuous triumphs. . . . The ways of science attract us so strongly both because of the results that can be achieved with science and because of its academic and public prestige [Lundberg 1938,197, 200; emphasis added].

The syllogism implied by Hempel and Lundberg may be criticized both for what it states and for what it assumes: 1. It is questionable that the achievements of physics can be ascribed to measurement any more than to a myriad of other factors (Hayek 1952/1967, 23; Zetterberg 1954, 126; Kuhn 1962/1969,49-60; Sartori 1970, 1038). 2. It can be disputed that measurement supersedes classification, because the choice of a unit of measurement automatically breaks down a continuum into classes with the same characteristics as the classes of any classification scheme, plus ratio relationships among them (Coombs 1953, 473). "Any level of measurement . . . includes, as a minimal requisite, a procedure of classification" (Blalock 1960/1970, 28).

34

On

Classification

3. Granted that the social sciences wish to achieve the same scientific status as physics (or, more empirically and cynically, that social scientists wish to achieve the same status as physicists), it must still be proved that they must, or even can, achieve it through the same means as does physics (Hayek 1952/1967, 12ff; Schutz 1958, 266-267; Radnitzky 1968, 70ff; Mackenzie & Mackenzie 1974/1979, 199-208; Phillips 1977/1981, lOff; Mokrzycki 1983, 7ff). 4. In particular, physicists measure properties they conceive as continuous after establishing a unit of measurement. But what if a property cannot be so conceived, or if a proper unit of measurement cannot be established? These situations happen to be quite common in the social sciences, but that fact did not discourage operationalists, behavioralists, and other followers of the neopositivist creed. While procedures were contrived that imitated, with various degrees of success, the characteristics of measurement (these are, at any rate, a valuable legacy from the behavioralist period in the social sciences), the residual gap was filled by terminological tricks. All sorts of code-assigning procedures (counting, scaling, classing) and their preliminary intellectual operations (intensional classification, scale-building, even the choice of indicators, etc.) were labeled "measurement," and any set of codes was labeled a "scale." "Measurement is the assignment of numerals to objects or events according to rules—any rule" (Stevens 1951, 22). "Measurement in its simplest form consists of substituting symbols or names for real objects" (Coombs 1953, 473). "A language of measurement defines classes of phenomena by providing specific criteria for deciding whether an observation can be assigned to a particular class" (Przeworski & Teune 1970, 93). In definitions such as these, the term "measurement" has been expanded to occupy the semantic spaces of intensional classification, classing, and even naming. The semantic space of a classification scheme has also been occupied. "The nominal scale represents the most unrestricted assignment of numerals. The numerals are used only as labels or type numbers, and words or letters would serve as well" (Stevens 1951, 25). "Nominal scales . . . are merely scales for the measurement of identity and difference" (Ellis 1968, 42). "A nominal scale exists if equality and diversity among the numbers can be interpreted as equivalence and non-equivalence among the units" (Galtung 1967, 73). The semantic stretching of the term "scale" reaches the point of a contradictio in adjectu, as the adjective "nominal" (meaning that differences between classes are exclusively in kind and not in degree) contradicts the noun (scale), which implies a difference in degree between steps (one has no scale if the steps are variously colored, but all at the same level). Such semantic considerations do not seem to bother behavioralists and their followers. If anything worries them, it is trespassing with too lowly feet on the semantic preserve of a more powerful and dignified talisman,

Alberto

Marradi

35

measurement. Stevens, the proponent of the expression "nominal scale," was indeed so worried that he returned at least three times to the point: "The nominal scale is a primitive form, and quite naturally there are many who will urge that it is absurd to attribute to this process of assigning numerals the dignity implied by the term measurement" (1946, 679). 1 2 The status imbalance was also perceived by Coombs: "A nominal scale . . . is so primitive that it is not always recognized as measurement" (1953,473). The generalized substitution of the self-contradictory expression "nominal scale" for classification scheme by social scientists would be astonishing could it not be explained by their lack of terminological sensitivity and by the prestige of "scientific" terms vis-à-vis "philosophical" ones. To the best of my knowledge, this expression has been criticized strongly and directly only by Sartori (1970) and indirectly only by Torgerson (1958,17). Because of its wider and less technical scope, the stretching of the term "measurement" has met with a wider range of reactions. Neopositivists Carnap (1950, 9) and Hempel (1952,58) have remarked that no measurement is possible without a proper unit. Kaplan (1964, 213) has exposed the misuse of measurement as a sham: "It often happens that a quantitative idiom is used, not only without any actual measurements having been performed, but without any being projected, and even without any apparent awareness of what must be done before such measurements can be carried out." Hayek (1952/1967, 57-58) has denounced "the blind pretension to systematically extend measurement to a field where specific conditions are not met" as responsible for "the worst aberrations and absurdities" in scientistically oriented social sciences. Lazarsfeld and Barton (1951/1967, 231) have criticized the assumption that rigor and scientific status depend on the use of a specific instrument—the measuring scale—and Rosenblith (1961) has stated that a science's maturity cannot depend on the number of quantitative concepts it employs. Even a historian of psychology rather sympathetic to behavioralism has conceded that the cult of measurement was "favored by the desire of investigators to claim the prestige of science for their research. Especially has this motivation operated among the psychologists, insecure because of their unscientific heritage from philosophy" (Boring 1961, 124). Political scientists and sociologists had an even greater insecurity regarding their scientific status. Accordingly, they avoided philosophical terms, among which "classification" was prominent. In the Encyclopedia of the Social Sciences of 1930, for example, there were no entries for classification, typology, or taxonomy. In the International Encyclopedia of the Social Sciences of 1968, there was only a single entry under typology. Other terms with a philosophical lineage (e.g., theory) were absent from both works. The downgrading of classification vis-à-vis measurement may not be a

36

On

Classification

permanent phenomenon in the social sciences. The growing emphasis in the last two decades on "theory," a term that positivists had practically banished in favor of "law," which was considered more "objective" and observational, is one sign of a change. Once again, a balanced assessment of the role of classification and related activities seems hard to attain. A statement by one of the leading sociologists of antibehavioristic orientation is highly significant: "[T]he search for beautifully comprehensive sets of categories is hardly the essence of explanatory science. The past half-century of typological sociologies should be enough to convince us of where it stands on the pole of scientific priorities" (Collins 1975,114). As early as 1968, Tiryakian (179) stated: The very success and acceptance of a typological classification may . . . freeze the level of explanation. Since typologies have much more de facto status in social sciences . . . than in the physical sciences, typological classification may share some of the responsibility for the retardation of more powerful theoretical explanation.

The idea of the superiority of theories and explanations (i.e., propositional knowledge) of concepts, classifications, and other forms of nonpropositional knowledge stems from hidden survivals of both realist and "certist" epistemologies. By "certist" epistemology is meant the conviction that the truth or falsity of propositions about the world may be ascertained. The conviction focuses attention on the dichotomy true/false, and therefore on propositions (whose main attribute is truth/falsity, while concepts are neither true nor false). By realist epistemology (in continental Europe the term would be "gnoseology") is meant the conviction that our concepts faithfully portray referents (objects, events, or "states of the world," as an archrealist, the first Wittgenstein, put it). If our concepts (including classes and types) do reflect objects/events, then we may take them for granted as nonproblematic. We can then focus on propositions or systems of propositions (hypotheses, theories, laws, models, explanations, etc.) that will tell us how objects behave, how events cause other events, how "states of the world" evolve.

Conclusions Classification should be freed from any ontological tasks and implications. Because such implications are not easily kept out of a debate on "natural" classification, it is wiser to abandon that debate, or at least the loaded term, altogether, as the conventionalists suggested (Giedymin 1982, 19). Classification schemes, typologies, and taxonomies do not make assertions and, therefore, cannot be judged true or false (Scheffler 1967/1983,

Alberto

Marradi

37

52; Swanson 1972, 193). It is misleading to consider them "particular kinds of hypothetical constructs or hypotheses" (Sandri 1969, 104), or to consider "some of them . . . as real rather than nominal definitions" (Berger & Zelditch 1968, 447). As concepts, they are tools for conferring organization and stability on our thoughts about reality. Like tools, they may be judged or found more or less useful for a particular purpose (Kemeny 1959/1972, 195). C l a s s i n g is guided b y c l a s s i f i c a t i o n s c h e m e s , t y p o l o g i e s , a n d taxonomies, plus the associated operational definitions; however, it is not determined by them (Scheffler 1967/1983, 49). A particular act of classing may be judged to be a correct or a (more or less) mistaken application of some rules of the operational definitions. It cannot be judged as true or false. The anxious question whether classification must be considered to be k n o w l e d g e or to be a preliminary to k n o w l e d g e (Gil,1981, 1024) can probably be answered without drama according to what each person would say about familiarity with the vocabulary of a language: Is that familiarity knowledge, or just a preliminary to knowledge of the assertions that can be made in that language? To claim that classification is not as germane as measurement to the scientific enterprise and, therefore, must be replaced by measurement might sound as ridiculous as claiming that French, being an "adjectival" language, should be spoken by always substituting adjectives for nouns or disguising nouns as adjectives. What is germane to the scientific enterprise is using the appropriate tool (or tools) to tackle each problem, rather than the same tool r e g a r d l e s s of t h e p r o b l e m . N o tools are s c i e n t i f i c o r u n s c i e n t i f i c in themselves, and any general hierarchy of tools in terms of scientific dignity, independent of problem and purpose, is nonsense. To claim that the attention given to classification has blocked the road toward "explanatory" science, or any other kind of science, might sound as ridiculous as claiming that one cannot say anything in French yet because one has spent all of one's time learning French words. We realize some m o d e m language teachers would subscribe to perhaps some slightly amended form of this statement. Alas, that awareness does not make the statement sound a trifle less ridiculous to me—or, I hope, to the reader.

Notes 1. Notice that in principle all other aspects of the intension of the higher concept are carried into each of the lower concepts: If we classify dogs by color, the class "black dogs" is characterized by all properties of "dogness," except those incompatible with "blackness"—if any. 2. By this we mean: If to a vector of concept A, B, C, D . . . is assigned a vector of terms a, b,c,d ..., the one-to-one relationship between A and a is strengthened by the particular form of allocation, which automatically excludes that term a be also

38

On

Classification

allocated to concepts B, C, D . . and that terms b, c, d . . . be also allocated to concept A. 3. Notice that, for each property, this entails an operation of measurement, counting, scaling, or an intensional classification plus a classing. In this sense, intensional classification is, in most cases, among the prerequisites of extensional classification. 4. This formula has come to be called a "distance function" (Sokal 1958). 5. For an application of the parsimony principle to the classification of test-takers, see Capecchi and Moeller 1968. 6. As it happens; see Cormack (1971) for a general review and Hudson (1982) for a selection of social science applications. 7. On the problem of empirical fit and the steps to improve it, see penetrating remarks by Turner (1981). 8. A generalized exception are linguists, who tend to use "typology" even when there is only one fundamentum (see Greenberg 1957 and authors cited therein). 9. The habit, rather common in natural sciences, of creating species characterized only by the common absence of a trait was later criticized as "paraphyletic" (i.e., as an improper application of taxonomic procedures). 10. We have seen a book introducing "a foreign-events data set" in which one-fifth of the variables in the file had from 94 to 99.9 percent of the cases in residual categories (Heyman, et al. 1973). 11. It must be remarked, in this connection, that in principle nothing would forbid that in intensional classification, too, a genus were divided into species by simultaneously considering a combination of fundamenta divisiones. It can by no means be excluded that this has in fact been done by several builders of taxonomies in zoology or botany, in order to solve particular difficulties and to obtain a better fit of their taxa to the phenomena at hand. On the other hand, in view of the complexities introduced by mingling the principles governing typology and taxonomy, it is not surprising that such a hybridization has never been conceptualized—as far as I know—in the methodological literature. 12. The same statement is found in Stevens (1951, 26) and a similar one in Stevens (1968, 174).

References Adanson, Michel. 1763. Families des plantes. Paris: Vincent. Andreski, Stanislav. 1965. The uses of comparative sociology. Berkeley: University of California Press. Apter, David E. 1965. The politics of modernization. Chicago: University of Chicago Press. Becker, Howard S. 1940. Constructive typology in the social sciences. American Sociological Review 5(l)(February): 40-55. Bendix, Reinhard. 1963. Concepts and generalizations in comparative sociological studies. American Sociological Review 28(4)(August): 532-539.

Alberto

Marradi

39

Berger, Joseph, and Morris Zelditch. 1968. Review of T. Parsons: Sociological theory and modern society. American Sociological Review 33(3)(June): 446-450. Bethell, Thomas. 1986. Creation and evolution. Dialogue 1: 66-71. Blalock, Hubert M. 1960. Social statistics. New York: McGraw-Hill. Quotations from the Italian translation: 1970. Statistica per la ricerca sociale. Bologna: Il Mulino. Boring, Edwin Garrigues. 1961. "The Beginning and Growth of Measurement in Psychology." In Quantification, ed. Harry Woolf. Indianapolis, IN: BobbsMerrill. Bunge, Mario A. 1967. Scientific research. Vol. 1: The search for system. New York: Springer. Calogero, Guido. 1949. "Classificazione." In Enciclopedia Italiana, vol. 10. Rome: Istituto della Enciclopedia. Italiana. Candolle, A. P. de. 1813. Théorie élémentaire de la botanique. Paris: Déterville. Capecchi, Vittorio. 1964a. Une méthode de classification fondée sur l'entropie. Revue Française de Sociologie 5(3): 290-306. . 1964b. I modelli di classificazione e l'analisi della struttura latente. Quaderni di Sociologia 13(3) (July): 289-340. —. 1966. Typologies in relation to mathematical models. Ikon 58 (July-September) supplement: 1-62. , and Frank Moeller. 1968. Some applications of entrophy to the problems of classification. Quality & Quantity 2(1-2): 63-84. Carnap, Rudolf. 1950. Logical foundations of probability. Chicago: University of Chicago Press. Cohen, Morris Raphael, and Ernst Nagel. 1934. An introduction to logic and scientific method. New York: Harcourt. Collins, Randall. 1975. Conflict sociology: Toward an explanatory science. New York: Academic Press. Coombs, Clyde H. 1953. "Theory and Methods of Social Measurement." In Research methods in the behavioral sciences, ed. Leon Festinger and Daniel Katz. New York: Dryden. Cormack, R. M. 1971. A review of classification. Journal of the Royal Statistical Society, ser. A, no. 134: 321-367. d'Andrade, Roy G. 1981. The cultural part of cognition. Cognitive Science 5: 179-195. De La Vega, W. Fernandez. 1970. Quelques aspects du problème de la determination automatique des classifications. Quality & Quantity 4(l)(June): 116-152. Deutsch, Karl W. 1966. "Recent Trends in Research Methods in Political Science." In A design for political science, ed. James C. Charlesworth. Philadelphia: American Academy of Political and Social Science. Dufrenoy, P. A. 1845. Traité de minéralogie. Paris: Carilian-Goeury et Vor Dalmont (4 vols.) Durkheim, Emile, and Marcel Mauss. 1963. Primitive classification.. London: Cohen & West. Ellis, Brian. 1968. Basic concepts of measurement. New York: Cambridge University Press. Everitt, Brian. 1980. Cluster analysis. Quality & Quantity 14(l)(January): 75-100. Fox, John. 1982. "Selective Aspects of Measuring Resemblance for Taxonomy." In Classifying social data, ed. Herschel C. Hudson. San Francisco: Jossey-Bass. Frake, Charles O. 1969. "The Ethnographic Study of Cognitive Systems." In Cognitive anthropology, ed. Stephen P. Tyler. New York: Holt. Galtung, Johan. 1967. Theory and methods of social research. London: Allen & Unwin. Giedymin, Jerzy. 1982. Science and convention. Oxford: Pergamon Press.

40

On

Classification

Gil, Fernando. 1981. Sistematica e classificazione. Enciclopedia, vol. 8. Turin: Einaudi. Gilmour, J. S. L. 1940. "Taxonomy and Philosophy." In The new systematics, ed. Julian Huxley. Oxford: Clarendon Press. Glaser, Barney G. 1978. Theoretical sensitivity: Advances in the methodology of grounded theory. Mill Valley, NY: Sociology Press. Granger, G. G. 1960. Pensée formelle et sciences de l'komme. Paris: Montaigne. Greenberg, Joseph. 1957. Typologies in linguistics. International Journal of American Linguistics 23(2). Hayek, Friedrik A. von. 1952. The counter-revolution of science: Studies on the abuse of reason. Glencoe, IL: Free Press. Quotations from the Italian translation: 1967. L'abuso della ragione. Florence: Vallecchi. Hempel, Carl Gustav. 1952a. Typological methods in the natural and the social sciences. Proceedings of the American Philosophical Association: 65-86. . 1952b. Fundamentals of concept formation in empirical science. Chicago: University of Chicago Press. . 1965. "Fundamentals of Taxonomy," In Aspects of scientific explanation, ed. Carl G. Hempel. Glencoe, IL: Free Press. , and Paul Oppenheim. 1936. Der Typusbegriff im Lichte der neuen Logik. Leyden: A.W. Sijthoff. Heyman, Charles F., et al. 1973. CREON. A foreign events data set. London: Sage. Hudson, Herschel C. (ed.) 1982. Classifying social data. San Francisco: Jossey-Bass. Huxley, Julian (ed.). 1940. The new systematics. Oxford: Clarendon Press. Kaplan, Abraham. 1964. The conduct of inquiry. San Francisco: Chandler. Kemeny, John G. 1959. A philosopher looks at science. Princeton, NJ: Van Nostrand. Quotations from the Italian translation: Il filosofo e la scienza. 1972. Milan: Il Saggiatore. Kroeber, Alfred L. 1939. Cultural and natural areas of native North America. University of California Publications in American Archaeology and Ethnology 38: 1-242. Kuhn, Thomas S. 1962. The structure of scientific revolutions. Chicago: University of Chicago Press. Quotations from the Italian translation: 1969 La struttura delle rivoluzioni scientifiche. 1969. Turin: Einaudi. Lazarsfeld, Paul Felix. 1937. Some remarks on the typological procedures in social research. Zeitschrift fur Sozialforschiing 6: 119-139. -, and Allen H. Barton. 1951. "Qualitative Measurement in the Social Sciences: Classifications, Typologies, and Indices." In The policy sciences, ed. D. Lemer and H. D. Lasswell. Quotations from the Italian translation: 1967 "Classificazioni, Tipologie e Indici." In Metodologia e ricerca sociologica, ed. P. F. Lazarsfeld. Bologna: Il Mulino. — 1955. "Some Principles of Questionnaire Classification." In The language of social research, ed. P. F. Lazarsfeld and Morris Rosenberg. Glencoe, IL: Free Press. Lenzel, Victor F. 1938. International encyclopedia of unified science vol. 1 (5): Procedures of empirical science. Chicago: University of Chicago Press. Levy-Bruhl, Lucien. 1910. Les fonctions mentales dans les sociétés inférieures. Paris: F. Alcon. Linnaeus (Karl von Linne). 1735. Systema naturae. Lundberg, George A. 1938. The concept of law in the social sciences. Philosophy of Science 5(2)(April): 189-203. Mackenzie, Brian D., and S. Lynne Mackenzie. 1974. The case for a revised systematic approach to the history of psychology. Journal for the History of the Behavioral Sciences 10(3): 324-347. Quotations from the Italian translation:

Alberto Marradi

41

1979. "Proposta per un nuovo approccio sistematico alla storia della psicologia." In Storiografia delle scienze e storia della psicologia, ed. Nicoletta Caramelli. Bologna: Il Mulino. Martindale, Don. 1959. "Sociological Theory and the Ideal Type." In Symposium on sociological theory, ed. Llewellyn Gross. New York: Harper. May, Ronald W. 1982. "Discriminant Analysis in Cluster Analysis." In Classifying social data, ed. Herschel C. Hudson. San Francisco: Jossey-Bass. Mayr, E. 1969. Principles of systematic zoology. New York: McGraw-Hill. McKinney, John C. 1966. Constructive typology and social theory. New York: Appleton-Century-Crofts. Miller, George A. 1956. The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review 63: 81-97. Mokrzycki, Edmund. 1983. Philosophy of science and sociology: From the methodological doctrine to research practice. London: Routledge. Montgomery, Andrew G., and Kathleen S. Crittenden. 1977. Improving coding reliability for open-ended questions. Public Opinion Quarterly 41: 235-243. Murdock, George Peter. 1967. Ethnographic atlas. Pittsburgh, PA: Pittsburgh University Press. Nowak, Stefan. 1976. Understanding and prediction. Dordrecht: Reidel. Nowotny, Helga. 1971. The uses of typological procedures in qualitative macrosociological studies. Quality & Quantity 6(June): 3-37. Oppenheim, Felix E. 1975. "The Language of Political Inquiry: Problems of Clarification." In Handbook of political science, vol. 1, ed. Fred I. Greenstein and Nelson W. Polsby. Reading, MA: Addison-Wesley. Quotations from the Italian translation: 1985 "Problemi di chiarificazione del linguaggio.". In Metodo scientifico e ricerca politica, ed. Domenico Fisichella. Rome: La Nuova Italia Scientifica. Pelloux, L. 1967. "Classificazione." In Enciclopedia filosofica, vol. 1. Florence: Sansoni. Phillips, Derek L. 1977. Wittgenstein and the scientific knowledge. London: Macmillan. Quotations from the Italian translation: 1981. Wittgenstein e la conoscenza scientifica. Bologna: Il Mulino. Piaget, Jean, and Barbel Inhelder. 1959. La genése des structures logiques élémentaires chez l'enfant: classifications et sériations. Neuchatel: Delachaux & Niesdé. Przeworski, Adam, and Henry Teune. 1970. The logic of comparative social inquiry. New York: Wiley. Pye, Lucian W. 1968. Typologies and political development. Il Politico 33(4): 725-733. Radnitzky, Gerard. 1968. Contemporary schools of metascience. Goteborg: Akademiforlaget. Riggs, Fred W. 1979. The importance of concepts: Some considerations on how they might be designated less ambiguously. American Sociologist 14(4)(November): 172-185. Rosenblith, Walter A. 1961. "Quantification of the Activity of the Nervous System." In Quality and quantity, ed. Daniel Lerner. Glencoe, IL: Free Press. Runciman, Walter G. 1972. A critique of Max Weber's philosophy of social science. Cambridge: Cambridge University Press. Ryle, Gilbert. 1938. Categories. Proceedings of the Aristotelian Society AA. VV.: 189-206. Sandri, Giorgio. 1969. On the logic of c l a s s i f i c a t i o n . Quality & Quantity 3(l-2)(January): 80-124. Sartori, Giovanni. 1970. Concept misformation in comparative politics. American

42

On

Classification

Political Science Review 64(4) (December): 1033-1053. . 1984. "Guidelines for Concept Analysis." In Social science concepts. A systematic analysis, ed. G. Sartori. London: Sage. Saussure, Ferdinand de. 1916. Cours de linguistique générale. Paris: Payot. Scheffler, Israel. 1967. Science and subjectivity. Indianapolis, IN: Bobbs-Merrill. Quotations from the Italian translation: 1983. Scienza e soggettività. Rome: Armando. Schlegel, Friedrich von. 1808. Über die Sprache und Weisheit der Indier. Heidelberg: Mohr & Simmer. Schuetz, Alfred. 1954. Concept and theory formation in the social sciences. Journal of Philosophy 51(9)(April 29): 257-273. Schutz, William C. 1958. On categorizing qualitative data in content analysis. Public Opinion Quarterly 22(4)(Winter): 503-515. Shils, Edward. 1958. Political development in the new states. The Hague: Mouton. Simpson, George G. 1945. The principles of classification and the classification of mammals. Bulletin of the American Museum of Natural History 45. Smelser, Neil J. 1976. Comparative methods in the social sciences. Englewood Cliffs, NJ: Prentice Hall. Sneath, P. H. A. 1957. Some thoughts on bacterial classification. Journal of General Microbiology 17. Sokal, Robert R. 1958. Distance as a measure of taxonomic similarity. Systematic Zoology 10(1): 70-79. , and P. H. A. Sneath. 1963. Principles of numerical taxonomy. San Francisco: W. H. Freeman. Stevens, Stanley Smith. 1946. On the theory of scales of measurement. Science 103(2684)(June 7): 677-680. . 1951. "Mathematics, Measurement and Psychophysics." In Handbook of experimental psychology, ed. S. S. Stevens. New York: Wiley. . 1968. "Ratio Scales of Opinion." In Handbook of measurement and assessment in behavioral science, ed. Dean K. Whitla. Reading, MA: AddisonWesley. Swanson, Guy E. 1972. "Frameworks for Comparative Research: Structural Anthropology and the Theory of Action." In Comparative methods in sociology: Essays on trends and applications, ed. Ivan Vallier. Berkeley: University of California Press. Thurstone, Louis Leon. 1931. The measurement of social attitudes. Journal of Abnormal and Social Psychology 26(2) (October): 249-269. Tiryakian, Edward A. 1968. "Typologies." In International encyclopedia of the social sciences, vol. 16. London & New York: Macmillan. Torgerson, Warren S. 1958. Theory and methods of scaling. New York: Wiley. Turner, Barry A. 1981. Some practical aspects of qualitative data analysis: One way of organizing the cognitive process associated with the generation of grounded theory. Quality & Quantity 15(3)(June): 225-247. Voegelin, C., and F. Voegelin. 1966. Anthropological linguistics: Index to languages of the world. Bloomington: Indiana University Press. Von Schelting, A. 1922. Die logische Theorie der historischen Kulturwissenschaft von Max Weber und im besonderen sein Begriff des ideal Types. Archiv für Sozialwissenschaft und Sozialpolitik 49. Wallace, A. F. C. 1961. On being just complicated enough. Proceedings of the National Academy of Sciences 97: 4 5 8 ^ 6 4 . Watkins, John W. N. 1952. Ideal types and historical explanation. British Journal for the Philosophy of Science 3(1): 22-43. Weber, Max. 1904. Die Objektivität sozialwissenschaftlicher und sozialpolitischer

Alberto

Marradi

43

Erkenntnis. Archiv für Sozialwissenschaft und Sozialpolitik 31. Wittfogel, Karl. 1957. Oriental despotism. New Haven, CT: Yale University Press. Zetterberg, Hans L. 1954. On theory and verification in sociology. Totowa, NJ: Bedminster Press

3 The Theory of the Mixed Regime and the Problem of Power: A Machiavellian Meditation Bruce James Smith The most influential contribution of ancient political science to subsequent political studies was the idea of constitution, or regime type. As many commentators have pointed out, the ancient understanding of constitution was more comprehensive than what modern theorists such as Locke have understood by " f o r m of government" (1965, 400). While the ancient understanding included the specific institutional structures of governance and practices of rulership, these were seen in relation to "a scheme of life" (Mcllwain 1983, 27). In the famous comparison of Athens and Sparta that Thucydides puts in the mouths of the Corintheans, it is the opinions, sentiments, affections, tastes, and proclivities of the two peoples, rather than their institutions, that capture the attention of these outside observers. On the other hand, in the Funeral Oration, Pericles attributes the distinctive qualities of the Athenian character to the city-states democratic " f o r m s . " The Corintheans fix their attention on what we would call political culture. Pericles is particularly interested in those characteristics that are tenacious, which (as he recognizes) often limit or circumscribe the capacity of intellect and will to shape the quality of political action. However, for Pericles (the engaged statesman, accomplished in the ways of power) form is the author of political life (Thucydides 1.68-72, 2.37-42). Political theory might be said to have originated in the effort to overcome this problem of vantage point. Otherwise put, the ancient inquiry into regime type is an attempt to understand the nature of the relation between a constitutional "form" and a people's "way of life." The shared etymology of "regime" and "regimen" suggests much about this relation as well as about the origin of that which the Greeks called the political. Political associations had distinctive or public ways of life, and only these were properly described as constitutional. Socrates insisted that tyranny was no form of constitution. Aristotle, who repeated Socrates' opinion, advised tyrants to destroy or disable the bios politikos. A significant part of Aristotle's theoretical project was to specify the parameters of the political and to give an account of the principles and circumstances upon which 45

46

The Theory

of Mixed Regimes

and the Problem

of

Power

political variety rested. In its maturity, the ancient inquiry into regimes was composed of three interrelated elements, each a response to a question. What is the full array of "natural" regime types? What are the causes and dynamics of constitutional change? What contributes to the maintenance of constitutional form? The most important answer to this last question was "the mixed form." This answer is important for two reasons: First, the theory of the mixed constitution was the most general theoretical conclusion of ancient political science; second, modern political theory established its autonomy and distinctiveness in significant ways by rejecting this conclusion. In this chapter, I will sketch the outlines of this transformation, announced, as is well known, by Machiavelli. Against the familiar sixfold classification scheme popularized by Polybius, Machiavelli (1965, 11) offered a radical proposal: "All the states, all the dominions that have had or now have authority over men have been and now are either republics or princedoms." Oddly enough, Polybius played a pivotal role in this transformation. For in attempting to formalize and simplify the constitutional reflections of "Plato and certain other philosophers" (Polybius 1923, 6.5) upon the relation of political form and substance, Polybius deprived this complex body of material of its theoretical suppleness and prepared the ground for Machiavelli's radical break with ancient method, and for his novel and far-reaching suggestions regarding regime classification. To understand Machiavelli's radicalism, however, it is necessary to get behind the Polybian reification of the ancient inquiry into regime type.

Pre-Socratics The empirical bases of ancient political science are especially evident in the earliest efforts to distinguish regime types. The original distinctions were descriptive and founded on experience and observation. This also suggests that the original mode of political reflection was comparative. The political "theorist" was one who had seen political variety. Political theory, so understood, makes its first appearance in the work of the historians. The earliest extant discussion of regime type is found in Herodotus, who distinguished three regimes on the basis of the question, "Who rules?" The traditional answer—the one, the few, and the many—is perhaps our oldest political convention.1 The one, the few, and the many are terms of quantity and suggest that the first discovery of ancient political science was that the dynamics of political life are determined in important ways by the number of people occupying or with access to "public space." In the Herodotian account, Darius and his companions, having enlarged the range of political possibility by laying low the power of the magian, debate the relative merits of rule by the one, the few, or the many. This early

Bruce James

Smith

47

discussion of regime type is of particular interest, for it gives some indication of the state of pre-Socratic political reflection and the intellectual milieu within which the Socratic reconsideration of the problem of regime took shape. Otanes, Megabyzus, and Darius speak in turn. Otanes blames monarchy and praises democracy, both blame and praise based upon an analysis of power. When power is possessed by one only, even if he is "the best man on earth," he is " p u t . . . outside the thoughts that have been wont to guide him." Monarchical power has two psychological consequences: On the one hand, the "satiety" of the all-powerful breeds "outrageousness"; on the other hand, the envy that is natural to human nature is enlarged beyond all bounds. Paradoxically, self-esteem is never more precarious than when one is "an absolute sovereign." Slandered by moderate admiration and offended by abject flattery, Otanes* monarch is a man without satisfaction or companionship. In the end, "he turns upside down all ancestral observances, forces women, and kills men without trial." Otanes' account of monarchical power suggests the degree to which Socrates' depiction of the soul of the tyrant draws upon an earlier tradition of theorizing. In contrast to monarchy, Otanes' portrait of democracy is characterized by "equality before the law" and restraint. Most important, democratic power is subject to the "check of audit" (Herodotus 3.80). Megabyzus prefers oligarchy. And while embracing Otanes' critique of monarchy, Megabyzus also condemns democracy. But, unlike that of Otanes, Megabyzus' praise and blame rests upon an analysis of the relation of power and knowledge. Otanes had focused on the power of leaders and concluded that democratic leaders are accountable. But to whom, asks Megabyzus, are the many accountable? Indeed, democracy seems to be inferior to monarchy, for though as much subject to the "outrageousness" of power as is the despot, the mob also acts without "knowledge" or "intelligent purposes." Only when power is "entrusted" to the "Best Men," says Megabyzus, will the political community receive adequate "counsel." Megabyzus attributes the excellence of the few to their knowledge of that which is "fine." Such knowledge, he suggests, is partly "innate" and partly "taught." We are left to conclude that the best men enjoy both ability and means, but Megabyzus is silent about the consequences of shared power among the few best (Herodotus 3.81). Darius is the last to speak, and he offers an alternative account of the best men. Whatever their individual excellences, the struggle for the reputation of "public virtue" engenders "private enmities," and from these grow "factions" that destroy the state. Darius advocates "despotism." Despotism's claim to superiority is timely. Taken together, the issue joined by Otanes and Megabyzus had defined the parameters of the political—those forms of rule characterized by what Aristotle (1328A) would later describe as "an association of equals." Darius' critique of democracy and oligarchy is, in fact, a critique of political life itself. Despotism is the "best of systems," Darius argues, because political life is inherently unstable, for both

48

The Theory

of Mixed Regimes

and the Problem

of

Power

democracy and oligarchy tend to "relapse" into despotism (Herodotus 3.82). Herodotus is circumspect in his assessment of these three "opinions." His narrative, however, is suggestive. Darius completes despotism's defense by claiming that as "freedom" was gained through the action of "one man," so it shall be best maintained (needless to say, a dubious proposition). The only question that remains is who shall become king. Megabyzus, who had numbered himself and his companions among the best men, joins the advocates of monarchy in the quest for kingship. Megabyzus, it seems, his preference for the companionship of excellence notwithstanding, secretly desires to be "the chief man" (Herodotus 3.82). His action appears to confirm Darius' critique of oligarchy. Otanes, on the other hand, has no desire to rule and refuses to enter the contest. Unsuccessful in his effort to make power accountable, Otanes insists that he "will not rule or be ruled" and chooses instead to become a private man, remaining "free" of public duty and agreeing only not to "overstep" the laws. The others resolve "that all of them should mount their horses in the outskirts of the city and as the sun rose whichever horse neighed first, his rider should possess the throne" (Herodotus 3.85). Darius, however, refusing to leave a matter of such importance to chance, contrives with his groom a "trick" that brings him the kingship. In this, Darius gives evidence of the political knowledge required for despotism, for he proves to be a master of fraud as well as force. 2 Herodotus offers no comment on Darius' success, but Darius' subsequent behavior—his arrogance and fear of his former companions (Herodotus 3.18-19)—suggests the truth also of Otanes' earlier critique of the rule of one. Herodotus' view of democracy is more difficult to discern. Unlike the others, Otanes cares not for power but seems rather to cherish relations of equality. Denied these, he would sooner "mind his own business." The question that remains unanswered, however, is whether Otanes is the only democratic type. I have explored these Herodotian materials in some detail because they reveal the highly developed character of ancient political science before the Socratic turn and clarify the questions implied in that turn. Three general questions emerge from these materials: First, does the tendency of power to destroy self-restraint and habit permit of any solution? Second, is it possible for power to be informed by knowledge? Third, and perhaps most important, to what extent can the inherent instability of bios politikos be overcome? In this respect, at least, Darius' critique of public life anticipates the Socratic political project, which, as Sheldon Wolin (1960, 42) has observed, put political science at odds with its subject.

Socratics For the Socratics, political life was, in Plato's phrase (Letter 7 325), a "whirlpool" whose "shifting currents" left the philosopher "dizzy." Not

Bruce James Smith

49

surprisingly, Socratic political science sought to understand "the general causes" (Aristotle Politics 1301A) of constitutional change. Herodotian political theory had isolated political form and had distinguished three regime types on the basis of a quantitative distribution of power. One of the anomalies of this analysis was the two-sidedness of political form, for each regime allowed of two rival accounts. Plato concluded that the threefold distinction based upon quantity was inadequate to political understanding (Statesman 292A; see also Aristotle Politics 129A). Perhaps animated by democratic confidence in the efficacy of power, Herodotian political theory had failed to address the relation of regime and way of life. As a result, each regime appeared to be characterized by two contrasting "dispositions" (Plato Republic 544E). It had long been recognized that regime types appeared to have corresponding ways of life. So conceived, however, the relationship between form and substance was ill suited to explain constitutional change. Socratic political science begins with a reconsideration, involving two claims, of this relationship. Thrasymachus, who shares the democrat's confidence in power, extends the pretensions of form: "A Democracy sets down democratic l a w s , . . . and the others do the same" (Plato Republic 338E). Against these pretensions, Socrates insists that regimes "arise . . . from the dispositions of men." Not the wonder of power, but these "dispositions" or "characters" (Plato Republic 544D-E) are the proper object of political study. Likewise, against the oligarchic claim that form should simply adhere to ancestral practices or the old way of life, Socrates presents political ways of life as inherently dynamic, and these dynamics are discussed in several ways. While there are as many regimes as there are dispositions, the correspondence between the "arrangements" of the city and the "soul" (Plato Republic 544E) of a citizen is imperfect. What appeared in Herodotus as rival accounts of a single regime are for Socrates measures of the dynamics of political disposition. Although Socrates shares the oligarch's preference for a state that is "hard to move" (Plato Republic 546A), he also knows that the oligarch's confidence in the fixity of political substance is no less myopic than is the democrat's confidence in the architectonics of political form. Socrates initiates the quest for justice with the maxim: The city is the soul writ large. But it is, rather, Socrates' insight into the disproportion of regime and human character that informs his political science. Socrates understands this disproportion as a problem of "rearing" (Plato Republic 561A), but the difficulty is less the failure than the limit of civic education. In the Republic, Socrates portrays this limit in generational terms. "Character" is notable for its tenacity, but its recreation in the young is problematic. While regimes shape the character of their citizens (an inward extension of the constitution), these human types are nonetheless inadequate to the maintenance of the regime, or, more exactly, to the conservation of its way of life. Aristotle (Politics 1310A) explains the difficulty: Those "habits" that give rise "to the right constitutional temper" for a particular regime type are

50

The Theory

of Mixed Regimes

and the Problem

of

Power

at odds with what brings "delight" to its citizens . A regime, Socrates insists, is what it loves, but ironically it is the "greediness" (Plato Republic 562B) for the good (as the regime understands it) that drives the "disposition" of its citizens beyond itself. It is less the decay of a regime's principle (honor, wealth, freedom) than its completion that transforms it into something else. The tension between "the influence of teaching" (Aristotle Politics 1310A) and the "delight" of citizens is, at base, a problem of power. Herodotus had linked the psychology of political power (its tendency to put its possessor "outside" his former habits) to constitutional change or the "relapse" of political forms into despotism. Socrates reformulates this problem by noting the tendency of a regime's dominant principle (including its characteristic social and political forms) to accrue ever greater power over time. As a result, oligarchies generally become more oligarchic, democracies more democratic. Socrates concludes that regime change is always the result of a failure of the ruling "part." While political rulers are of "one mind," a regime "cannot be moved" (Plato Republic 545D), but the lessons learned on the periphery of power are at war with what is fitting in those who rule. The failure of the ruling part is a failure of self-restraint, the progressive inability of rulers to "subject" themselves to the laws (Plato Letter 7 337C-D). The central question of Herodotian political science was: Who possessed power? The three possible answers emphasized political differences based upon the number participating in power. Herodotus had focused his attention on what might be described as the structure of power. Plato, however, was more interested in the psychological implications of its possession. Power was inherently dynamic, corrosive even of its own architecture. De Jouvenel (1981, 17) has argued that the threefold classification provided a schema for the several "forms of Power," but that an adequate political science must look behind these forms to power's "essence." Plato was the author of this suggestion. In distinguishing between lawful and lawless regimes, Plato emphasized power's chief attribute: its tendency to overstep and erode all boundaries. Plato's hostility to the essential character of public life, an animus to which both Wolin (I960,42) and Arendt (1974,222) call attention, was, at base, a hostility to the dynamics of power and would have lasting consequences for the classification of regimes and their evaluation. The Platonic distinction between the lawful and the lawless raised a host of interrelated questions. To the Herodotian question of who (one, few, many), Plato asks howl He develops this distinction in two quite different schemata. In the Republic, lawlessness is depicted as the increasing liberation of appetite. Socrates describes a political culture (embodied in a succession of distinctive human types: aristocratic man; timocratic man; oligarchic man; democratic man; and tyrannic man) gradually rejecting self-restraint and embracing the principle of freedom, which in reaching its most complete expression is transformed into (or indistinguishable from) slavery. In the political order of things, each succeeding regime is characterized by an

Bruce James

Smith

51

increasingly weakened spirit of law. By this account, the rule of the few (which now has three incarnations—aristocracy, timocracy, and oligarchy) is seen as earlier and superior to both the rule of the many and that of the one (which together form a dialectical whole and counterpoint to the rule of the few). In every instance (if in varying degrees), the rule of the few is characterized by self-restraint. Even oligarchic man, for all his secret prodigalities, must remain sober and "stingy" (Plato Republic 554B-555A). With the coming of democracy, however, self-restraint is experienced as a kind of despotism and is quickly thrown off. Significantly, it is only the transition from oligarchy to democracy that involves class war and the revolutionary seizure of political power. An alternative account appears in the Statesman where the Eleatic Stranger argues that each regime type (one, few, many) may be doubled on the basis of its lawful or lawless practices. In this view, each numerical type has both a good and perverse form. In the Republic, a psychological trajectory provides the unifying thread to what appears a narrative of political decay. In the Statesman, the Stranger alters the vantage point, asking "which [regime] is hardest to bear and which is most tolerable." In both the Republic and the Statesman, an analysis of the nature of power remains the primary inquiry, but while the Republic explores the implications of power for its possessors, the Statesman considers power from the perspective of those subject to it. The Stranger evaluates each regime in relation to its double in light of "the aim . . . of men everywhere . . . to secure for themselves the most tolerable life." The rule of one, it seems, is best when lawful, but worst when lawless. Likewise, the rule of the few turns out to be, alternately, second best and second worst. The standing of democracy, or the rule of the many, is the most interesting conclusion of the Stranger's analysis. While democracy is the "least desirable" of the lawful forms, it is the "best" of those forms that "flout the laws" (Plato Dialogues 302B-303B). Taken together, Plato's two efforts at regime categorization radicalize the problem of political science. Socrates' account includes a discussion of regime transition; the Stranger's does not. In Socrates' account, the several varieties of the rule of the few succeed one another only to be succeeded by the regimes of the many and the one. While Socrates is interested in power relations both within a single numerical type and among the several numerical types, the Stranger examines only the dichotomies within numerical types. According to Socrates, regime change within numerical types precedes changes from one numerical type to another. Democracy emerges from this analysis as the special focus of political science. From the point of view of the quality of human life, the transition from oligarchy to democracy represents an advance. Moreover, the two forms of the rule of many promise greater continuity in the quality of human life. Yet, behind the rule of the many and the appeal of the principle of freedom (the most fertile soil for the practice of philosophy) lurks the prospect of the worst regime and

52

The Theory

of Mixed Regimes

and the Problem

of

Power

the wholesale destruction of human life as such. Thus, it is not surprising that the philosopher, finding himself in the regime of the many, should turn his attention to political affairs. Democracy is the seed ground of the Socratic turn.

Aristotelians The Socratic analysis of political power suggested two possible responses: the improbable conjoining of power with indifference to power (the project of the Republic), and the mixing of the several "forms of Power" so that rival claims to a "share in the enjoyment of power" (Aristotle Politics 1281B) might be arranged in such a way so as to create the "proper balance" (Plato Laws 693E) of authority and liberty, thereby forestalling the tendency toward lawlessness. 3 The idea of the mixed regime made its first appearance in Plato's Laws (693D), where "the Athenian" insists that the well-founded regime must be a mixture of monarchy and democracy—the "two matrices of constitutions from which all others [are] derived." But the theoretical elaboration of the mixed regime awaited Aristotle, and for good reason. The differences between Plato and Aristotle have often been exaggerated. Aristotle embraced the Socratic critique of political power and the centrality of the distinction between lawful and lawless regimes (though Aristotle was perhaps more confident than was Plato of the capacity of wellfounded civic habits to circumscribe the prodigalities characteristic of public space). Still, the differences are real. Aristotle (Politics 1316A-B) criticized his teacher for having failed to elaborate an adequate classification of regimes or an account of regime change. Underlying his criticism are differences in both perspective and attitude. For Plato, the classification of regimes was largely a matter of understanding the "best" and the "worst" (Republic 544A). "Tyranny," says James Schall, " . . . was precisely the political possibility that made the discussion of the best form of government necessary" (1978, 101). Plato's enduring interest, from his youth onward, was the relation of democracy and tyranny, and the deep ambivalences of a constitution devoted to the principle of "freedom" ("Letter 7" 324B-325C; Laws 693D-E; Republic 557A-564A). Aristotle, on the other hand, insisted that it is the question of "equality" that "involves" the philosopher in political "speculation" (Politics 1282B). Aristotle's detailed exploration of the relation of the few and the many may be seen as a rejoinder to Plato's preoccupation with the relation of the one and the many. In Plato's view, democracy was more than one of a number of regimes. It was the completion of the political idea, the political matrix in its purest form {Laws 693D). Historically, ancient democracy was the culmination of the crisis of membership, which had originally given rise to political forms of

Bruce James

Smith

53

association (De Jouvenel 1981, 84-91). Plato's insistence that freedom was the political principle par excellence derived from his view of political life as tending toward the progressive liberation of its participants from the constraints of the ancestral. It is significant in this regard that Socrates portrays the hypothetical regime of the Republic as also the ancestral regime. The political problem, so understood, was one of increasing disproportion between the requirements of regime maintenance and the "dispositions" engendered by particular regimes. With the emergence of democracy, this disproportion becomes so pronounced that Socrates discovers that democracy "contains all species of regimes" (Plato Republic 557D). The complete dissociation of regime and regimen promoted by the democratic principle, "where each man organize[s] his life . . . privately just as it pleases him" (Plato Republic 557A), presages the destruction of the political association itself, and its replacement by tyranny. While not sharing Darius' enthusiasm for the rule of one, Plato appears to endorse his conclusions regarding the tendencies of political life. Aristotle regards Plato's account of regime change, with its emphasis upon "an exaggerated sense of liberty," as "defective." Aristotle's criticisms are of two kinds: First, Plato's unidirectional and apparently necessary descent is at odds with observation; second, and more importantly, his account of the transformation of "disposition" in public life is insufficiently attentive to the distinctiveness or integrity of political relations. Aristotle (Politics 1316A-B) finds it "curious" that Plato would attribute regime change "merely" to changes in character without paying due regard to "offence[s] against justice." While Plato does not despair of retarding regime change, he overlooks, in Aristotle's view, certain resources, intrinsic to political life, that are of consequence for the relation of regime and regimen. In the "whirlpool" of public life, Aristotle discovers certain points of fixity, or at least regularities of motion. In the Republic (422E), Socrates observes that every political association is really two cities—a city of the rich and a city of the poor. A cautionary note for Plato, the Socratic observation takes on greater significance in Aristotle's constitutional reflections. In attempting to account for political variety, Aristotle argues that every state is made up of different "parts." These parts include kinship groups, social classes, and occupational groups, as well as functional distinctions within the political structure, notably those responsible for "war, justice, and deliberation." Philosophically considered, "there must therefore be as many constitutions as there are modes of arranging the distribution of office according to the superiorities and the differences of the parts of the state." However, there is "a prevalent opinion" that there are only two constitutions—oligarchy and democracy. While Aristotle maintains that a classification based upon the several parts of the state would bring us "nearer the truth," he concedes that there is great force behind the common opinion.

54

The Theory of Mixed Regimes

and the Problem of Power

For while the same persons . . . may serve as soldiers, farmers, and craftsmen, [and] again, may act as both a deliberative council and a judicial c o u r t , . . . there is one thing which is impossible: the same persons cannot be both rich and poor. This will explain why these two classes—the rich and the poor—are regarded as parts of the state in a special and peculiar sense.

What exactly is "special and peculiar" about these parts? Their importance resides in their relation to one another. They "appear" as "opposite parts." While every part might have a distinct "interest" (one might speak of the interests of farmers, soldiers, craftsmen), the exclusivity of class membership makes such membership the axis around which political forces are most often arrayed (Aristotle Politics\2%9B, 1291A-B). In emphasizing quantitative distinctions, Herodotian political science had focused attention on the level of political participation. While not indifferent to the distinctive qualities of the one, the few, and the many, the importance of number was most prominent. In this, the Herodotian vantage point revealed its Periclean pedigree, for it was enamored of what Arendt calls "the space of appearances." This vantage point preserves something of the wonder that must have accompanied the emergence of the political association: "Look how many appear in public!" Socrates, however, had insisted that political science must look behind appearances to the "dispositions" of those in public life. Aristotle restores respect for appearances in political science, but on Socratic foundations. In classifying regimes, number, says Aristotle, is "an accidental attribute." "The real ground of difference between oligarchy and democracy is poverty and riches." One might expect differences in disposition among farmers, soldiers, and craftsmen no less than between rich and poor. Aristotle does not deny this, but differences in disposition among social classes are the most politically salient. Following Socrates' lead, Aristotle notes that the oligarchical character is more "severe," the democratic "soft and relaxed." But the exclusivity of class membership and the differing dispositions attending wealth and poverty acquire their political significance largely as a result of the public perspectives they engender (Aristotle Politics 1279B, 1289B). Differing interests and dispositions may be found in various types of human associations, but social class is "special and peculiar" because of its tendency to generate a distinctive conception of the good. A social class is characterized by an ethos. As a result, social class differs from other "parts" in its capacity to shape politically relevant appearances or "public opinion." Social classes thus bring great dangers to political associations, but also great benefits (an insight not lost on Machiavelli). Both threat and promise rest upon similarities between class and polity. As one cannot be both rich and

Bruce James

Smith

55

poor, neither can one be both Athenian and Spartan. In its consciousness of itself, its sense of boundary and belonging, its conviction of its own worthiness, social class approximates political membership. At its worst, social class may mistake itself for the political community or set itself up as a rival. Yet, these political evils are themselves measures of the way in which social classes produce and reproduce something akin to character, and tenacious in outlook. As a result, social class emerges as a principal vehicle of civic education, imparting regularity, shared habits, and, most importantly, a "hold" on a certain "conception of justice" (Aristotle Politics 1280A). We might pause here to note two conclusions that flow from Aristotle's analysis of social class, each of consequence for the theory of the mixed constitution. First, a political existence produces, above all, class formations. Despotisms may contain poverty and riches, but self-conscious membership in a class formation is a political phenomenon rooted in the distinctive "spirit of [public] intercourse" (Aristotle Politics 1280A). And second, regime types are characterized by the predominance of certain characteristic and recognizable opinions that lay claim to public (or just) consideration and grow out of such self-conscious membership. These conclusions undergird Aristotle's analysis of regime types and constitutional change. In Plato's dispositional account of regime type, constitutional change originates in "the insatiable character of the good" (Republic 555B) as it is understood by those (dominant "character" types) who rule. While Aristotle preserves the structure of this argument, his emphasis on social class modifies its content. In identifying the goods sought by the several human types through political action, Socrates makes no effort to distinguish public and private goods. This is the meaning of Aristotle's criticism, noted above, of Plato's treatment of constitutional change. Aristotle observes that the transformation of a regime of the "Spartan type" (timocracy) into an oligarchy cannot be explained simply as the result of honor lovers becoming "money lovers," but must take account of "the conviction, natural in men of greatly superior wealth, that it is an offence against justice for men without property to be put on a level in the state with the owners of property." In other words, it is not the immediate pursuit of the good, oligarchically understood (i.e., wealth), that structures the oligarchical regime (and pushes it beyond itself), but a "conviction" regarding "justice." While this "conception of justice" is "natural" to those of a particular social class, its "principle^]" character depends upon the existence of a "state" or public way of life. Aristotle presses his criticism because it involves more than a necessary elaboration of the Platonic account; it is rather a question of explanation. Oligarchies are not the result of "magistrates" becoming "profit makers" (after all, Aristotle argues, in some oligarchies, "profitmaking is forbidden"), but of the dominance of certain public claims, "natural" in a civic body in which there are some of "greatly superior wealth." This public claim is not simply a mask for similarly situated individuals pursuing a

56

The Theory

of Mixed Regimes

and the Problem

of

Power

private good (though men are admittedly poor judges "in their own case"); it is the "conviction" of a self-conscious class, convinced of its own desert. Of course, Aristotle knows that the rival claims of oligarchs and democrats "both fail to carry [justice] far enough," but each is animated by the "appearance" of a public good {Politics 1280A, 1316A-B). Plato's dispositional analysis had led him to distinguish regimes on the basis of the freedom to pursue certain goods. Aristotle's class analysis, with its emphasis upon the appearance of the few and the many as "opposites," points to different understandings of political equality (or what equality requires) as the explanatory key to both constitutional differentiation and change. In a sense, both oligarchs and democrats are egalitarians (on reflection, this is not surprising, for the political association is "an association of equals, and only of equals"), but they disagree on the criteria upon which equality is to be established (Aristotle Politics 1280A, 1328A). Aristotle was perhaps the first fully to appreciate the dynamics of equality and their importance for political ways of life. Egalitarian dynamics, precisely because of their comparative (relational) content, are more subject to disequilibrium than are hierarchical relations. Aristotle's differentiation of the political association into "parts" is a methodological prelude to a consideration of the relations of these parts. In this light, each regime type is characterized by a certain set of relations. With Aristotle, the Socratic suggestion regarding power's dynamism issues in a fully formed analytic alternative to the Herodotian emphasis on power structures and becomes a political science of power relations. While these dynamics may take a variety of forms (of which the sixfold classification is only the most general typology), the visibility, exclusivity, and ubiquity of poverty and riches most often order them along the axis of class relations. This is why oligarchies and democracies are the most prevalent regime types, and why many believe that they are the only regime types. Among its advantages, Aristotle's analysis of political association in terms of equality enjoys a theoretical economy, for it both classifies regimes and provides a framework within which constitutional changes of the widest variety might be accommodated. As we have already seen, Plato had presented regime change as the outcome of dialectical struggles within the "mind" of the ruling class, which transform its "disposition." This emphasis on the "ruling part" led to a conception of change that diminished the importance of relations between "parts," or political conflict. (The transition from oligarchy to democracy in the Republic is an important exception but appears as only one among several changes, and is perhaps even anomalous.) Aristotle's emphasis on equality, as one might expect, brings political struggle center stage. While he shares Plato's interest in the "objects which are at stake" in political life, Aristotle is especially interested in "the state of mind which leads to sedition." On closer inspection, "the principal and general cause" of the seditious "attitude" turns out to be two states of mind that, together, form what might be described as

Bruce James Smith

57

the perceptual boundaries of the egalitarian relation. As Aristotle puts it, some think "they have the worst of the bargain in spite of b e i n g . . . e q u a l ; . . . others . . . conceivfe] that they get no advantage . . . although they are really more than equal." Such perceptions (with or without "justification" commonplace in public life) are the basis for Aristotle's ironic suggestion that love of justice animates most revolutions. Aristotle is the originator of that tradition that finds envy and contempt at once the most common and the most destructive of the "passions" that fill the "minds" of those who share a public space.4 In this tradition, political science seeks especially the means by which to reduce such passions (Aristotle politics 1295B, 1302A-1303A). Aristotle's reconsideration of the bases of regime type and the causes of constitutional change contains theoretical possibilities not readily available within the parameters of the Platonic account. As we have seen, while Plato portrays regimes as the political corollaries of culturally dominant character types, Aristotle presents regimes as particular constellations of political relations ordered by specific conceptions of justice. Several implications of Aristotle's rethinking of the relation of regime and regimen are especially relevant here. His shrewd insight into the nature of political ways of life—that their fundamental characteristic is struggle over the meaning of equality—draws attention to the "spirit" of civic relations. Consequently, while regime types rest upon dominant "conceptions" or shared meanings, that which is shared or held in common must necessarily emerge from an ongoing public existence. While it is in the nature of public existence to throw up rival conceptions of what equality requires, shared meaning is the child of public "intercourse" in ways "dispositions" are not. Finally, while Aristotle preserves the Eleatic Stranger's classification scheme, with its distinction between good and perverse forms, even the perverse forms are now seen to have a "conception of justice." (Tyranny, which has no such conception, is the delimiting case.) This is particularly important because the mixed constitution will draw its materials from the perverse forms. Plato's suggestion in the Laws, that the corrosive character of power might be stayed by mixing the "matrices" of freedom and authority, while it would prove theoretically fertile, was not easily combined with the dipositional analysis of power in the Republic. The libertarian spirit was overmighty and seemed capable of destroying any balance momentarily struck with authority. Moreover, rear-guard efforts to redouble the spirit of restraint seemed only to strengthen the forces of license. Aristotle's methodological starting point underlines the theoretical advantage his analysis of social class affords, an advantage absent from Plato's analysis. "To know the causes which destroy constitutions is also to know the causes which ensure their preservation." From this "general proposition," Aristotle derives three practical maxims: 1. "[W]here the elements are well mixed,.. . keep a look-out against all

58

The Theory

of Mixed Regimes

and the Problem

of

Power

lawlessness" "[C]onfidence should never be placed in devices intended to hoodwink the masses" 3. "[T]he principle of equality . . . should extend to all who are really 'peers"' 2.

These maxims acquire their force from "the real difficulty" underlying all varieties of the mixed form—that of "combining" oligarchy and democracy (.Politics 1979 1307A-1308A). With this difficulty in mind, Aristotle begins his discussion of mixture by turning to techniques for encouraging and discouraging the participation of the rich and the poor in the law courts and assemblies, and for arriving at a "middle term" in the sharing of power. This initial emphasis on levels of participation and the appointment of magistrates has a Herodotian flavor in the significance it attaches to the legal form of the mixed regime. "A good criterion of the proper mixture," adds Aristotle, is when a constitution can be described "indifferently" as either democracy or oligarchy. However, this focus on institutional arrangements is deceptive. Their prominence is not the result of Aristotle's confidence in constitutional mechanics, but a measure of the force he attributes to political appearances, and points to the limitations of statecraft. The mixed constitution, Aristotle insists, must "look as if it contained both oligarchical and democratic elements—and as if it contained neither" (emphasis added). This formulation both acknowledges and conceals the limits of form. Aristotle's analysis of social class suggests the difficulty. The mixed form must appear to satisfy, at once, the just claims of each class while rejecting the unjust claims of its "opposite." In other words, the state must look oligarchical to oligarchs and democratic to democrats, yet avoid appearing democratic to oligarchs or oligarchic to democrats. For political "stability," Aristotle insists, does not depend merely upon the support of the majority (which even poor constitutions may enjoy), but rather on no "single section" desiring a change. And the complexities of appearance do not end here. Such is the nature of political life that should a state be "equally balanced" between rich and poor, revolution rather than stability is its likely fate. Ironically (as Hobbes would later note), nothing enhances the hope of victory more than the appearance of equality of power. Thus, Aristotle turns from the discussion of form to that of substance—the "way of life" of the mixed regime (Politics 1294A-B, 1304A-B). It is the responsibility of political form to cultivate the appearance of justice. But under what circumstances will form be able to do its work? In attempting to answer this question, Aristotle returns to the theme of social composition, but he now presents a new configuration of social classes: "In all states, there may be distinguished three parts, or classes, of the citizenbody—the very rich; the very poor; and the middle class which forms the mean." The discovery of "the man in the middle" was probably the

Bruce James Smith

59

consequence of an ethical hypothesis (indeed, Aristotle suggests as much), for while this social condition had been observed before (Aristotle cites the prayer of Phocylides), a consideration of the political significance of middling condition required putting aside the traditional distinction of the one, the few, and the many. The middle classes have several attributes that make them the social nexus of the well mixed regime. Suffering least from "ambition," the middle classes neither covet the goods of others, nor do they fear that others covet what is theirs. Where the middle class is large, political conspiracy, Aristotle reasons, will be uncommon; with less cause for fear or envy, the "mind" of the middle class is better able "to follow the lead of reason" (Politics 1295B, 1297A). Again, however, it is the public manifestation of the interests and dispositions of the middle classes that is of the greatest interest to Aristotle. Unlike his discussion of the rich and the poor, Aristotle does not attribute a distinctive conception of justice to the middle class. It may be that the middle classes are less animated by political passions of any kind (this was Tocqueville's view), which would explain their suitability for inclusion in both oligarchical and democratic constitutions. In any case, where the middle class is preponderant, its slight "ambition" and middling station alter the character of public discourse. By disposition and habit, neither the very rich nor the very poor find it easy to "listen"; by situation, neither can have "confidence" in the other. In the end, the mixed regime rests upon the possibility of arbitration, and "the man in the middle," says Aristotle, "is such an arbiter" (Politics 1296B, 1297A). While some commentators attribute to Aristotle a constitutional theory of checks and balances, Wills (1982, 100) is closer to the mark in describing the mixed regime as one in which each class would "have a say" (see also Von Fritz 1958, 82). But, as Aristotle suggests, in politics listening is more difficult than speaking. In this, one finds also a Socratic rehabilitation of public space, as the cacophony Plato heard in the streets of Athens (with its echoes inside the souls of oligarchic and democratic man) is replaced by a conversation about justice. Yet, as Aristotle was forced to acknowledge, while the mixed regime was within "the reach of ordinary men," the social configuration upon which it could be based was rare. Even this "most . . . practicable" of Socratic regimes had "never been established" (Aristotle Politics 1295A-1296A). As Machiavelli knew, the problem of the few and the many remained.

Polybius The tradition of inquiry I have been exploring received its final form from Polybius. Indeed, Polybius was quite conscious of his place in a tradition and saw his sixfold classification and theory of constitutional cycles as a

60

The Theory of Mixed Regimes

and the Problem of Power

distillation and culmination. Yet, Polybius' relationship to his predecessors (he names Ephorus, Xenophon, Callisthenes, and [emphasizing his particular importance] Plato) is complex and ambiguous. Drawing upon the sixfold classification scheme of the Statesman and the dispositional descent of the Republic, Polybius constructs a cycle of "natural" regimes, which moves through the double identities of the rule of the one, the few, and the many. The cycle begins (anthropologically) with the rule of one, moving alternately from "good" to "corrupt" forms and toward the expansion of political participation (one-few-many), and ends in extreme democracy or mob rule, only to return to the rule of one. Of particular importance for my purposes is Polybius' political naturalism. Polybius assimilates constitutional change to natural processes. Corruption is "inbred," akin to the rusting of iron or the presence of woodworms in timber (whether these analogies are themselves consistent is another matter). Corruption is "proper" (in keeping with the order of political things) to every "simple" constitutional form "and naturally follows on it." For both Plato and Aristotle, constitutional forms and their transformation resulted from the prodigality of the good. Constitutional disorder was essentially a problem of human character and social perspective, and pointed to the ethical insufficiency of political existence without philosophy or a complete understanding of the good for human beings. Polybius, however, conceives of political "decay" as an organic process originating in "growth." Descriptively, Polybius' account of regime change draws upon the earlier tradition, emphasizing the love of power and the equivocal character of civic equality, but these are now seen to arise from "success" and attendant "prosperity." Socratic constitutionalism sought to forestall decay through adherence to law (Plato) and dependence upon a social configuration that encouraged moderation and reasonableness (Aristotle). But Polybius' organic analogies (Histories 6.2, 4, 7-9, 10, 45, 57), which tie decay to growth (experienced as victory or success), hardly permit such solutions. At the first mention of regime types, Polybius insists that it is "evident" that "the best constitution" is a "combination" of the three good forms (kingship, aristocracy, democracy). The "proof' of this, Polybius argues, can be found in "actual experience," and he finds that of Sparta and Rome to be particularly instructive. The character of such a mixture, however, is not clear. Lycurgus' great discovery was the tendency of each "principle" (one, few, many) to "grow unduly." The aim of "combination" is to so arrange things institutionally that the one will be kept within bounds out of "fear" of the many, and likewise the many out of fear of the few. Like Plato, Polybius treats the one and the many as polar types, but his emphasis on institutional arrangements dramatically alters the analysis. Not training and habits, but fear, keeps the mixed regime "in a state of equilibrium." The fear of each "neutralize^] . . . the force of each." The status of the few is crucial in this balancing process. As the middle part, the "elders" (the institutional

Bruce James

Smith

61

embodiment of the few) act as a counterweight against that polar principle that would seek to "outbalance" the other. The implicit argument is that, situated in the middle, the few have reason to fear both the one and the many and, therefore, have an interest in constitutional equilibrium. It is unclear whether the few, upon whom the integrity of the mixed regime rests, are "always on the side of justice" because they are "the best citizens" or because they have the most to fear. To this ambiguity may be added another: To what extent is the "growth of the state" consistent with constitutional equilibrium CHistories 6.3, 4, 10, 57)? These difficulties are of great interest because of the foothold they provide for Machiavelli. Moreover, the "actual experience" of Sparta and Rome is more complex than Polybius first suggests. Although Lycurgus and the Romans "arrived at the same . . . foim of government," the Spartan regime was the result of "a process of reasoning," whereas the "formation and growth" of the Roman state were "due to natural causes." Briefly treating the Lycurgan constitution, Polybius passes over to his real "aim," an examination of those "peculiar political institutions [through which] in less than fifty-three years nearly the whole world was overcome and fell under the single domination of Rome." At first glance, this appears an odd theoretical maneuver. The Lycurgan regime was the creation of a theoretical intelligence; the Roman state was the result of that people's unique experience and their gift for choosing well in the face of it. But in these prefatory remarks, Polybius has revealed both his theoretical interest and his project. He is fascinated by the capacity of the Roman state to generate such enormous power ("a thing the like of which had never happened before" [Histories 6.2, 4, 10]). The similarity of Sparta and Rome is only apparent. For while the Spartans were the most martial of the Grecians, they were also (unlike the Romans) reluctant and ill-suited imperialists. Cumming (1969, 89) has suggested that Polybius may not have known "quite what he [was] up to as a political theorist." Cumming is probably right, but Polybius knew what he wanted to be up to. He wanted to understand the relation between republican forms and the generation of power. In need of explanatory categories, he appropriated the theory of the mixed regime and put it to use as a tool for discovering answers to his question. That it was a poor tool for his purposes is evident, but his attempt to modify it to suit those purposes is of the greatest interest. Polybius claims only to be simplifying the "theory" of his predecessors to make it accessible to those of "common intelligence." He gives little indication of being cognizant of the degree to which his "short summary" is in fact a radical departure. His rehearsal of the "arrogance" of power notwithstanding, Polybius' tastes are at odds with those of the Socratics. Polybius is himself a lover of power. But the differences run deeper yet. For Plato and Aristotle, a human being is a zoon politikon because lesser associations are insufficient for a fully human life. The Socratic interest in constitutional stability derived from the relation of political existence to

62

The Theory of Mixed Regimes

and the Problem

of Power

humanness. Because the polis was the only sufficient basis for the human things, its stability was essential. In Polybius' political anthropology, it is not the need for self-sufficiency but the "natural weakness" of human beings that is the origin of "political societies." From the vantage point of the weak, no proper limit to power is discernible. Plato and Aristotle knew well the ubiquity of the love of power in political life but treated it as the chronic disorder of public spaces. For Polybius (6.5), however, power is "the central political fact" (see Von Fritz 1958,45; Wolin 1960,91). The critique of power that the classical tradition undertook was framed by a certain theoretical context. I have presented this context as a dialectical relation between form and substance, regime and regimen. This relation was susceptible to several formulations or regime types. A regime was an arrangement of power in light of a particular way of life. From the vantage point of ancient political science, understanding the relation of these two elements was essential because political ways of life imparted habits, shaped perspectives, and cultivated a "spirit of intercourse" that established the propriety of claims to power and boundaries to its use. Only when there existed such a relation between the "form of Power" and a specific bios politikos could one speak of the existence of a constitution. In Polybius' work, the relation of regime and regimen is ambiguous. "In my opinion," he says, "there are two fundamental things in every state, by virtue of which its principle and constitution is either desirable or the reverse. I mean customs and laws." In this, Polybius appears to follow the classical tradition. In support of this view, he turns again to the example of the Lycurgan constitution that was so successful in inculcating "fortitude and temperance." However, this praise is followed by a reconsideration of the aims of political life: For the purpose of remaining in secure possession of their own territory and maintaining their freedom the legislation of Lycurgus is amply sufficient, and to those who maintain this to be the object of political constitutions we must admit that there is not and never was any system or constitution superior to that of Lycurgus. But if anyone is ambitious of greater things, and esteems it finer and more glorious than that to be the leader of many men and to rule and lord it over many and have the eyes of all the world turned to him, it must be admitted from this point of view the Laconian constitution is defective, while that of Rome is superior and better framed for the attainment of power [Polybius Histories. 6.46, 48, 50],

Among the customs of the Romans, Polybius emphasizes two—their military orders ("one of those things really worth . . . knowing") and religious practices. Neither, however, can be said to habituate, train, or shape the character or disposition of the Roman citizen. Rather, both are instruments of fear. While military discipline rests upon palpable horrors, religion marshals "invisible terrors." The Roman "way of life" turns out to be an extension of

Bruce James Smith

63

her institutional structure. What Polybius (6.26, 48,56) calls private life and public action become complements, collapsing the dialectic of form and substance. 5 In assessing the merit of the Roman state, Polybius takes his bearings from the defeat of the Romans at Cannae (the point of their greatest weakness) in order to measure the rate of growth of Roman power. Power is inherently dynamic and can only be understood in terms of "success" or "reverse." The extraordinary success of the Roman state, Polybius (6.2) insists, must be attributed to its "peculiar political institutions." His reconsideration of the aims of political life alters the nature of the political problem. The original aim of the theory of political mixture was to contain the corrosive dynamic of the love of power. But if power was the central political fact, and the proper aim of political societies was its maximization, then a constitutional form had to be found that could "balance" the forces unleashed in public space (the power potential of associations of equals) so that such energies might be directed to "greater things" than the maintenance of "freedom." Wolin (1960, 83-85) rightly credits Polybius with the discovery of the capacity of institutions to "distort" and redirect political conflict. Polybius' discovery was not, however, simply an acknowledgment of the reality of power; it was an attempt to formulate a political science in the service of power. The real difficulty lay in creating a power structure at once internally resilient and equilibrating, and yet capable of growth.6 As a solution, Polybius described an institutional arrangement that permitted rival forces both to "counteract and cooperate." But the theoretical logic of the threefold mixture was ill suited to these purposes. As Polybius' own presentation suggests, such a structure was far better suited to "thwart," "check," "hamper," and "counterwork" the pursuit of power than to foster its growth {Histories.6.15, 18). As Machiavelli would later suggest, Polybius had made two crucial errors: He had conceived of a well-founded power structure as an "equilibrium," and he had imagined the fundamental political dynamic to be the struggle between the one and the many. Had he had access to Aristotle, he might have reached other conclusions.

Machiavelli Machiavelli claimed to have discovered new "ways and methods" through his "continual reading" of the ancients. Machiavelli's work has about it always this two-sidedness—hunting for the "unknown" in the company of "ancient men" (1965, 10, 190, 929). His presence here is in the service of a proper denouement. I noted earlier, by way of preface, his radicalism, which makes its first appearance in the form of a regime classification. Machiavelli's famous dichotomy of princedoms and republics must have been (and was surely intended to be) arresting to an audience schooled in the tradition of the

64

The Theory

of Mixed Regimes

and the Problem

of

Power

one, the few, and the many. His classification transforms three into two by combining the few and the many in a single regime type. Behind this novel formulation, however, is an important reorientation, it has several aspects, but all revolve around questions of political development and democratization. As we have seen, the ancient classification scheme had emphasized what Socrates called "the ruling part." Machiavelli's dichotomy distinguishes, rather, subject and citizen. The traditional relation of ruler and ruled becomes, in Machiavelli's hands, that of leader and led. As contemporary commentators now recognize, Machiavelli's chief interest was in free societies. His dichotomy contains what he took to be the two most important questions about such societies: How are they most likely to be brought into being? How are they best maintained? Machiavelli's radicalism is grounded in his answers to these questions; to the first, he answers: through the rule of one; to the second: through the ongoing struggle of the few and the many. From the point of view of the tradition we have been exploring, Machiavelli reaches two conclusions: 1. The nature of the political association (the relation of the one, the few, and the many) can only be understood in terms of its development (or, the state can only "become perfect" in time) 2. Political development is the result of political struggle, and especially class conflict ("[T]hose who condemn the dissension between the nobility and the people . . . find fault with . . . a first cause") The implications of these conclusions are well beyond the parameters of this chapter, but I want briefly to examine the manner in which Machiavelli (1965, 196, 202, 218, 929) refashions the materials he finds in "ancient courts." In the Discourses, Polybius is undoubtedly Machiavelli's point of departure. Upon clarifying the focus of his inquiry ("free peoples"), Machiavelli turns directly to a consideration of the traditional account of regime types and the superiority of a regime "that partakes of them all." While he refrains from attribution, the content and manner of presentation replicate that of Book 6 of Polybius' Histories. The regime that begins to emerge in the following chapter, however, bears marks of a more complex parentage (Machiavelli 1965, 193, 199). Polybius (Histories 6.57) had concluded that states decay as the result of one of "two agcncics"—external weakness or the "growth of the state itself." The difficulty, of course, was that the most plausible response to external vulnerability was political development. Machiavelli accepts this view but finds the Polybian solution defective. For Polybius imagined that the forces unleashed by an association of equals (the substance of political life) could be preserved yet balanced by institutional forms. With Machiavelli, the ancient relation of form and substance recovers its complexity, if not its original aim.

Bruce James

Smith

65

Concluding that one must choose between balance and development (or Sparta and Rome), Machiavelli embraces the latter. Having dispensed with the need for "equilibrium," he discovers new wine in old constitutional bottles. Aristotle had been right; the fundamental conflict in political life is that between the few and the many. Indeed, such conflict is the distinctive mark of "free peoples" (Machiavelli 1965, 334-337). In every city, says Machiavelli, there are two "humors" (umori) or dispositions—that of "the people" and that of "the rich." Each humor has its characteristic desire: "The people desire not to be bossed by the rich; the rich desire to boss and oppress the people." The Aristotelian project had been the pacification of such conflict through deliberation, arbitration, and the sharing of power. But Machiavelli, emphasizing the endemic and tenacious character of such humors (a view Madison would later second), concluded that "it is not possible f o r [the few and the many] to agree." The novelty of Machiavelli's constitutionalism derives from his unflinching exploration of the consequences of allowing the natural dispositions indigenous to political existence to flourish: "As a result of these two opposed desires, one of three effects appears in the city: princely rule or liberty or license" (1965, 39, 1094). Drawing upon the political experience of the polis (and especially that of Athenian democracy), the Socratics had associated the struggle against the few with lawlessness and tyranny. Machiavelli, however, located the more serious (because more constant and self-conscious) threat to political existence in the desire of the few themselves. While the people might raise up a tyrant, this was a desperate and derivative danger—desperate because it sprung not (as the ancients imagined) from popular arrogance, but from the popular fear of oppression; derivative because it was only an extreme manifestation of the primary political vice of the many: the desire to be left alone. In the natural order of political things, the constancy of the few (in their oppressive designs) and the desperation of the many made either "princely rule or . . . license" likely. "Liberty" (and its concomitant power), on the other hand, required good foundations. In light of the constancy of the few and the inconstancy of the many, good foundations had to accomplish two things—strengthen the people's hand and prevent their flight from public life in the name of private satisfaction (Machiavelli 1965, 95, 208, 237, 1086). Reduced to aphorism, Machiavelli's constitutional advice is: "[A]rm the People!" This is the meaning of his obscure teaching: "There cannot be good laws where armies are not good, and where there are good armies, there must be good laws" (1965,47,95,208,237,1086). In emphasizing the armed citizen as the foundation for a political existence, Machiavelli greatly simplified the relation of regime and regimen. The good regime might remain exceptional, but the theory of the good regime was no longer at odds with its subject matter. The doctrine of balance had been the hub of ancient constitutionalism. In

66

The Theory

of Mixed Regimes

and the Problem

of

Power

dispensing with the ancient tradition's preoccupation, Machiavelli transformed constitutional theory. Among the advantages of a constitutional theory based upon political development and democratization was its greater applicability. In The Prince, Machiavelli ridiculed those who had "fancied for themselves republics or principalities that have never been seen or known to exist in reality." He intended to "write something useful" (1965, 57); there is much to suggest that he did. In his monumental study of the "essence" of power, Bertrand De Jouvenel (1981) has argued that the "decisive moment" in the early history of a political people is the "crisis" between the king and the aristocracy. Against the aggrandizing prerogative of the king is arrayed the patriciate, mighty in its self-consciousness and solidarity. The subsequent predominance of either the one or the few largely depends upon the allegiances of the many at this "fork in the road." Where "the patricians" bring "the plebeian horde" over to their side with the promise of citizenship, the political result is "the oligarchic republic." On the other hand, should the fear of aristocratic oppression drive "the plebeian horde" into the arms of the king and their combined strength carry the day, the resultant consolidation of centralized power is accompanied by greater social equality and democratization. The first of these scenarios, De Jouvenel argues, describes the origin of the ancient republic; the second, for the most part, that of the modern state. In time, the person of the prince gave way to a political abstraction, but this was only the extension of the prince's work—the democratization of the modem state (De Jouvenel 1981, 85, 89-90). Machiavelli was the prophet of this transformation; as one might expect, he put it bluntly: So if a prince succeeds in conquering and holding his state, his means are always judged honorable and everywhere praised, because the mob is always fascinated by appearances and by the outcome of an affair; and in the world the mob is everything; the few find no room there when the many crowd together [1965, 67].

While there were certain advantages to a constitution based upon "discord" between the few and the many, there were also weaknesses. Machiavellian constitutionalism rejected equilibrium in favor of the energies of "discord." But in the absence of balance, toward what did the power-state tend? Was there a trajectory to discord? A remark of Machiavelli's (1965, 202-203) is suggestive: "The aspirations of free peoples are seldom harmful to liberty, because they result either from oppression or from the fear that there is going to be oppression." But what would become of such "aspirations" once the many had, in fact, become "everything"? James Schall (1978, 120) has observed that the question of the best regime "necessarily involve[s] the retention of man in . . . public liberty." Yet, as Tocqueville recognized, in a world where "the many crowd together" and

Bruce James

Smith

67

"the few find no room," the question of the best regime changes its form, for the bases of public liberty are altered in novel ways. This fact must be the point of departure for any inquiry into the modem state.

Notes 1. It may well be that the threefold distinction is itself derived from an earlier twofold distinction rooted in the struggle between the few and the many that has always characterized political life and was later informed by the experience of tyranny. For Herodotus, the rule of one turns out to be Persian despotism, and Thucydides is primarily interested in taking the measure of oligarchy and democracy. In the Politics (1290A), Aristotle calls attention to what he describes as "a prevalent opinion" that there are only two kinds of constitution—oligarchy and democracy. While he rejects this view as defective, its appearance in the text suggests its standing as conventional wisdom. This twofold distinction might also be described as the most important content of the practical political knowledge the ancient citizen had acquired through familiarity with, and participation in, public affairs. Significantly, Machiavelli, who appealed to ancient practice over and against classical political philosophy, would lay the foundation of the modem state on the basis of "discord" between "the rich" and "the people" (1965, 202-203), largely by rethinking the problem of regime type. 2. Aristotle (Politics 1313A) would later identify force and fraud as the marks of tyranny. 3. The enduring interest among the Socratics in the Lycurgan constitution, which had maintained the integrity of law for centuries, was based largely on the view that the stability of the Spartan regime was the result of the superiority of its mixture. 4. It is worth observing that Aristotle's particular interest in the relation of the few and the many extends to his consideration of the rule of one. The virtue of a king is recognized in his "aim" of promoting friendship between "the owners of property" and "the bulk of the people," while the would-be tyrant is "always dividing the state in two and waging war against the rich" by fanning the flames of envy among the many (Politics 1304b-1305a, 1310a-1311a). 5. Polybius' collapsing of the dialectical poles of ancient political science is especially evident in his rejection of Thebes and Athens as instructive examples. Polybius (Histories 6. 43) dismisses them because their "successes . . . were [not] due to the form of their constitution, but to the high qualities of their leading men." Plato and Aristotle would have found such an observation quite wrong-headed, for the "qualities o f . . . leading men" are, in his view, an integral part of any constitution. In his discussion of Polybius, Cumming (1969, 172-173; see also p. 94) has something similar in mind in preferring the term structure to constitution, and in his emphasis on "Polybius' confidcnce in the causal efficacy" of such institutional structures. 6. Polybius anticipated Machiavelli's great discovery regarding power—that its growth depends upon a struggle for its possession. For while power is potentially

68

The Theory

of Mixed Regimes

and the Problem

of

Power

unlimited, it appears finite and scarce to most of those who would have it. Power's increase is directly proportional to the collective energy unleashed by its pursuit. Ironically, however, but a few pursue unlimited power; most seek power only for safety's sake.

References Arendt, Hannah. 1974. The human condition. Chicago: University of Chicago Press. Aristotle. 1979. The politics of Aristotle, ed. and tr. Ernest Barker. Oxford: Oxford University Press. Cumming, Robert Denoon. 1969. Human nature and history: A study of the development of liberal political thought. Chicago: University of Chicago Press. De Jouvenel, Bertrand. 1981. On power: Its nature and the history of its growth, tr. J. F. Huntington. Westport, CT: Greenwood Press. Herodotus. 1987. The history, tr. David Grene. Chicago: University of Chicago Press. Locke, John. 1965. Two treatises of government, ed. Peter Laslett. New York: New American Library. Machiavelli, Niccolo. 1965. The chief works and others, tr. Allan Gilbert. Durham, NC: Duke University Press. Mcllwain, Charles H. 1983. Constitutionalism: Ancient and modern. Ithaca; NY: Cornell University Press. Plato. 1964. The collected Dialogues of Plato, ed. Edith Hamilton and Huntington Cairns. New York: Pantheon Books. . 1968. The Republic of Plato, tr. Allan Bloom. New York: Basic Books. Polybius. 1923. The histories, tr. W. R. Paton. London: William Heinemann. Schall, James V. 1978. The best form of government: A perspective on the continuity of political theory. Review of Politics 40(l)(January): 97-123. Thucydides. 1934. The Peloponnesian war, tr. Richard Crawley. New York: Modern Library. Von Fritz, Kurt. 1958. The theory of the mixed constitution in antiquity. New York: Columbia University Press. Wills, Garry. 1982. Explaining America: The Federalist. New York: Penguin Books. Wolin, Sheldon S. 1960. Politics and vision. Boston: Little, Brown.

PART TWO Modern Typologies

4 Democratic Political Systems Arend

Lijphart

My purpose is to elaborate on the typology of democratic political systems proposed in my book Democracies (1984), in which I contrast the majoritarian and consensus types. After a brief summary of this typology, I shall discuss its relationship with my earlier typology in which consociational democracy was the main element, assess the question of how well the individual countries fit it (with an emphasis on the problematical cases), and explain both the causes and the consequences of the different types. The last of these tasks—which is the theoretically most important and challenging one—entails the treatment of the degree of majoritarianism versus, consensus, both as a dependent variable (with the extent of social and cultural pluralism, population size, and the influence of the British Westminster model as the independent or explanatory variables) and as an independent variable (with the quality and stability of the democratic system as the dependent variables to be explored). Table 4.1 presents the typology in its simplest form: a two-dimensional typology in which each dimension is based on the contrast between the majoritarian and consensual models of government. I constructed it partly on logical and partly on empirical grounds. My first step was to formulate nine majoritarian characteristics of democratic government and politics that are logically based on the principle of concentrating as much power as possible in the hands of the majority. Similarly, the nine consensual characteristics are based on the principle of sharing, dispersing, and limiting power in a variety of ways. With one exception, they can also be seen as the logical opposites of the list of majoritarian characteristics. The exception is purely representative democracy—that is, democracy without such devices of direct or semidirect democracy as the referendum, initiative, and recall—because majoritarian and consensus democracy have this trait in common. My second step was to apply the remaining eight pairs of contrasting characteristics to the empirical world of democratic nation-states: the twentyone countries that have been continually democratic since approximately the end of World War II. These twenty-one countries became twenty-two cases 71

72

Democratic

Political

Systems

Table 4.1 A Typology of Democratic Systems Dimension II (Federal-Unitary) Majoritarian Majoritarian

Consensual

Majoritarian

Majoritarian-Federal

Consensual-Unitary

Consensus

Dimension I (Executives-Paities) Consensual

because I treated the French Fourth and Fifth Republics, based on completely different constitutions, as separate cases. The initial conjecture in my 1984 book was that, because the eight majoritarian characteristics are derived from the same principle and hence are logically connected, they would also occur together in the real world, and that, likewise, the eight consensual traits would be found together. This turned out not to be the case. I did find very strong relationships among the eight variables, but in the form of two separate clusters instead of one cohesive cluster. This meant that two separate dimensions of the majoritarian-consensual contrast had to be distinguished. The first is based on five variables that have to do with the party and electoral systems, and government coalitions; I shall call it the executives-parties dimension. The second is based on three variables and may be conveniently labeled the federal-unitary dimension. The juxtaposition of these two dimensions produces the four basic types shown in Table 4.1. The classification of the twenty-two cases in this two-dimensional scheme is presented in Table 4.2. Each dimension is divided into three categories, yielding a matrix with nine cells. This division is arbitrary; in principle, each dimension is a continuous variable, ranging from maximum majoritarianism to maximum consensus. Hence, I could also have chosen a fourfold, fivefold, or even sixfold division of the dimensions—which would have produced sixteen, twenty-five, and thirty-six cells respectively—but very soon the number of cells would have exceeded the number of cases. Because one of the purposes of typologizing is to group sets of similar cases, the choice of nine cells appears to be a reasonable compromise. The grouping of the eight variables into two dimensions, and the assignment of scores on each dimension to the twenty-two cases, were based on a factor analysis of which, for the sake of brevity, I shall omit the details here (see Lijphart 1984; 211-220). Let me now turn to the characteristics of majoritarian and consensus democracies, beginning with the five variables that make up Dimension I (the executives-parties dimension). I shall describe the eight contrasting characteristics—as well as the ninth that the two models have in

Arend

Lijphart

73

Table 4.2 Twenty-two Democratic Regimes Classified According to the T w o Dimensions of Majoritarian Versus Consensus Democracy Dimension II Majoritarian

Intermediate

Consensual

New Zealand (5) United Kingdom (4)

Ireland (4)

Australia (5) Austria (5) Canada (5) Germany (5) United States (3)

Iceland (3) Luxembourg (1)

France V (6) Norway (1) Sweden (1)

Italy (4) Japan (5)

Denmark (1) Israel (3)

Belgium (1) Finland (1) France IV Netherlands (1)

Switzerland (1)

Majoritarian

Dimension I Intermediate

Consensual

Source: Adapted from Arend Lijphart, Democracies (New Haven, CT: Yale University Press, 1984), p. 219. Notes'. The plural and semiplural societies are italicized. The intermediate category encompasses countries with factor scores between -.50 and +.50. The figures in parentheses are Dahl's democratic scale types (see Robert A. Dahl, Polyarchy [New Haven, CT: Yale University Press, 1971], p. 232).

common—as succinctly as possible, illustrate them with the help of the British and Swiss examples, and explain how I operationalized them for the purpose of determining their interrelationships.

Dimension I: Definitions and Operationalizations 1. Concentration of executive power versus executive power-sharing: the typically majoritarian executive is a one-party, bare-majority cabinet—exemplified by virtually all British cabinets since World War II—whereas the ideal-typic consensual executive is a power-sharing coalition consisting of all the major parties in the legislature—exemplified by the Swiss Federal Council. In operationalizing this variable, I gave predominant weight to the question of whether cabinets are bare-majority cabinets—"minimal winning" cabinets in the terminology of the coalition theorists—or more inclusive "oversized" cabinets in which one or more parties are represented that are not necessary to give the cabinet a parliamentary majority. Minority cabinets form an intermediate category. British cabinets were minimal winning ones in almost the entire 1945-1980 period. The Swiss Federal Council was oversized 100 percent of the time. 2. Executive dominance versus executive-legislative balance: this

74

Democratic

Political

Systems

contrast is self-explanatory, but it is not easy to operationalize. The best available method is to measure average cabinet durability. Cabinets that are d u r a b l e — t h o s e that do not c h a n g e f r e q u e n t l y in t e r m s of p a r t y composition—tend to be much more powerful vis-à-vis their legislature than are less durable executives. The British score is eighty-one (months). Switzerland is a special case because of the formally independent status of the executive; it was assigned a score of thirty. 3. Two-party versus multiparty system: the best way to operationalize this variable is the "effective number of parties" measure proposed by Laakso and Taagepera (1979). It counts the number of parties weighted by party size. The mean effective number of parties, based on their shares of legislative seats in the 1945-1980 period, was 2.1 for Britain and 5.0 for Switzerland. 4. One-dimensional versus multidimensional party system: in the m a j o r i t a r i a n m o d e l , the two m a j o r p a r t i e s d i f f e r f r o m e a c h other programmatically on only one dimension: socioeconomic policy. The consensus model assumes differences among the major parties not only on the left-right issue dimension but also on one or more of the following: the religious, cultural-ethnic, urban-rural, regime support, foreign policy, and postmaterialist dimensions. I gave one point to a dimension of high salience and half a point to those of only medium intensity. The British score is one (one issue dimension, the socioeconomic, with high intensity), while Switzerland's score is three (high-salience socioeconomic and religious dimensions and medium-salience cultural-ethnic and urban-rural ones). 5. Plurality elections versus proportional representation (PR): because a few countries have electoral systems that are neither plurality nor PR and because, in practice, not all PR systems are equally proportional and not all plurality systems are equally disproportional, I operationalized this variable in terms of the degree of disproportion of the actual electoral outcomes: the average deviation between the vote and seat shares of the two largest parties in all elections between 1945 and 1980. The British score is 6.2 percent, the Swiss 1.5 percent.

Dimension II: Definitions and Operationalizations 6. Unitary and centralized versus federal and decentralized government: because, in practice, federations tend to be considerably more decentralized than are unitary states, I operationalized this variable simply in terms of a measure of centralization: the central government's share of total central and noncentral tax receipts in the 1970s. These shares were 87 percent in unitary Britain and 41 percent in federal Switzerland. 7. Unicameralism versus strong bicameralism: strong bicameralism entails a two-chamber legislature in which the houses are roughly equal in

Arend

Lijphart

75

power (symmetrical) and elected by different methods (incongruent). Legislatures in which the two chambers are asymmetrical (the second being inferior to the first) or congruent (the two being virtual carbon copies as a result of being elected by the same methods) are intermediate cases between unicameralism and strong bicameralism. Britain is not the best example here because, though Parliament is bicameral, it obviously belongs to the highly asymmetrical category. New Zealand's unicameralism provides a better example. The two Swiss chambers exemplify strong bicameralism very well: They have equal powers, and the second is a federal chamber in which the cantons have equal representation. 8. Unwritten versus written and rigid constitutions: the optimally majoritarian constitution is unwritten because an unwritten constitution does not impose any formal limitations on the power of the parliament (i.e., of the parliamentary majority). At the other extreme is a written constitution that is protected by judicial review and that is difficult to amend; for instance, approval by a popular referendum or, better yet, by extraordinary majorities may be required in order to change it. Written constitutions that are flexible (relatively easy to amend) or that are not protected by judicial review are in intermediate positions between the two extremes. Britain is one of very few democracies with an unwritten constitution. Switzerland happens to be an imperfect example of the opposite: It has a written and relatively rigid constitution, but it lacks judicial review. Because decentralization, strong bicameralism, rigid constitutions, and judicial review are all associated with the concept of federalism in the political science literature, the second dimension may also be called the federal-unitary dimension.

Representative Versus Direct Democracy The ninth characteristic of the majoritarian model is exclusively representative democracy. Elements of direct democracy, such as the referendum, might interfere with the supremacy of the parliamentary majority and are therefore incompatible with majoritarianism. However, this ninth majoritarian characteristic does not differentiate the model from the consensus model of democracy—which is also based on the concept of representative instead of direct democracy. In practice, direct democracy is a rare phenomenon: Of the various devices of direct democracy, such as the referendum, initiative, recall, and primary elections, only the referendum is fairly frequently used, and it is used by both majoritarian and consensus systems. Britain has held only one national referendum: on the issue of Common Market membership in 1975. Consensual Switzerland is a very frequent user of the referendum, but majoritarian New Zealand is also among

76

Democratic

Political

Systems

the fairly frequent users. For the purpose of operationalizing this variable, I constructed five categories ranging from the greatest frequency of referenda, with Switzerland as the only occupant, to the group of nine countries in which no referenda occurred between 1945 and 1980. As indicated earlier, my empirical analysis confirmed that this ninth variable did not distinguish between majoritarian and consensus democracy and that it was completely unrelated to Dimensions I and II. Before I leave this topic, however, I should like to point out that, in one special circumstance, the referendum can probably be regarded as a consensual instead of a neutral characteristic. This argument runs counter to the conventional wisdom, which tends to think of the referendum as more majoritarian than consensual; in fact, referenda are often considered even more majoritarian than is representative-majoritarian democracy, because elected legislatures offer at least some opportunities for minorities to present their case in unhurried discussion and to engage in bargaining and logrolling. Butler and Ranney (1978, 36) state: "Because they cannot measure intensities of beliefs or work things out through discussion and discovery, referendums are bound to be more dangerous than representative assemblies to minority rights." But the referendum is not inevitably a blunt majoritarian instrument. When it is combined with the popular initiative, as in Switzerland, it gives even very small minorities a chance to assert a claim against the wishes of the majority of the elected representatives. Even if the effort does not succeed, it forces the majority to pay the cost of a referendum campaign. Hence, the potential calling of a referendum by a minority is a strong stimulus for the majority to be heedful of minority views. In Switzerland, as Lehner (1984, 30) has forcefully argued, "any coalition with a predictable and safe chance of winning has to include all parties and organizations that may be capable of calling for a successful referendum." The special prerequisite for making the referendum into a consensual instrument, of course, is its combination with the initiative. Of the twelve Western democracies that have used the referendum since 1945, only two provide for the popular initiative: Switzerland and Italy. It is hazardous to generalize on the basis of so little evidence, but both the logic of the referendum-plus-initiative and the Swiss example support the suggestion that it may greatly strengthen the consensual character of a democratic system.

Consociational and Consensus Democracy How does consensus democracy compare with the concept of consociational democracy that I first formulated in the late 1960s (Lijphart 1968; 1977)? The two are very closely related in the sense that they are both nonmajoritarian, or even antimajoritarian, forms of democracy. The

Arend

Lijphart

77

differences between them can largely be explained in terms of how they were derived. I proposed the concept of consociation in order to show how it is possible for societies that Almond (1956) called "culturally fragmented" to enjoy a high degree of political stability. In contrast with Almond's assumption of competitive elite behavior in such societies, I argued that elite behavior should be treated as a variable: Cooperative and coalescent elite behavior can turn a potentially unstable political system into a stable one. Next, I analyzed the four main examples of countries in which the elites behaved in such a cooperative and stabilizing manner—the Netherlands, Belgium, Austria, and Switzerland—and I discovered four essential practices: grand coalition; segmental autonomy; proportionality; and mutual veto. These became my four consociational principles. In Democracies (1984), I reversed this procedure, enumerating first the characteristics of majoritarianism and subsequently defining each trait of nonmajoritarian democracy as a contrast with the corresponding majoritarian trait. I intentionally labeled this non-majoritarian model "consensus democracy" in order not to confuse it with the similar but not identical concept of consociational democracy. Table 4.3 tries to clarify the exact relationship between the two models by juxtaposing the nine characteristics of consensus democracy and the four consociational principles. The first difference that stands out is the different numbers of defining characteristics—nine compared with four. However, this difference is not as great as it appears, because a few of the consensus characteristics are also implicitly part of consociation. For instance, consociational democracy is defined as a form of democracy that is a response to the challenge of a culturally fragmented or plural society (that is, a society that is sharply divided along religious, ideological, linguistic, cultural, ethnic, or racial lines into virtually separate subsocieties with their own political parties, interest groups, and media of communication). Thus, a plural society is a precondition of consociation, and it might have been more accurate to define consociational democracy in terms of five instead of four characteristics. Moreover, if the plural society contains three or more distinct segments (subsocieties), these are likely to manifest themselves as a multiparty and multidimensional party system—thus exhibiting a parallel with two features of consensus democracy. Also, an implicit assumption of consociation is elite predominance, which corresponds roughly with the principle of exclusively representative democracy. The second difference is that the four basic principles of consociational democracy are broader than the corresponding traits of consensus democracy. The grand coalition principle may take the form of executive power-sharing in a multiparty coalition cabinet, but it may also be implemented in the form of various informal advisory arrangements. Moreover, for consociational democracy, it is the inclusion of all segments rather than all parties that is important when segments and parties do not coincide. Federalism is merely

78

Democratic

Political

Systems

Table 4.3 A Schematic Presentation of the Overlap Between Consociational and Consensus Democracy Consociational Democracy

Consensus Democracy I.

n.

m.

l. 2. 3. 4. 5.

Executive power-sharing Executive-legislative balance Multiparty system j. Multidimensional party system» Proportional representation

1. Federalism and decentralization 2. Strong bicameralism 3. Rigid constitution

Representative democracy

1.

Grand coalition

(5. Plural society) 3. Proportionality 2. Segmental autonomy 4. Minority veto (6. Elite predominance)

one way of establishing segmental autonomy. When the segments of a plural society are geographically intermixed, segmental autonomy can be instituted in the form of autonomous cultural councils and educational associations; Friedrich (1968) has called this "corporate federalism." The most important aspect of the consociational principle of proportionality is proportional representation, but it also includes proportional appointment to the civil service and proportional allocation of public funds. The minority veto is a broader concept than the mere requirement of extraordinary majorities to amend the constitution; however, it is narrower than the idea of a rigid constitution in one respect: It does not imply the need for judicial review. The first and second differences can be summarized as follows: Consociational and consensus democracy overlap considerably, but neither is completely encompassed by the other. This means that consociational democracy cannot be seen as a special form of consensus democracy or vice versa. A third difference emerges from the discussion of the first two: Where consensus and consociational democracy differ, the former tends to emphasize formal institutional devices whereas the latter relies to a larger extent on informal practices. In order to state the fourth and final difference, one similarity needs to be discussed. Both consociational and consensus democracy are highly suitable forms of democracy for divided societies. Another way of expressing this is that homogeneous societies, in which a high degree of consensus naturally exists, can afford majoritarian and competitive government; however, in societies that are not naturally and spontaneously consensual, the political system needs to be arranged in such a way as "artificially" to introduce as much consensus as possible. Both consociational and consensus systems are designed to do this. The difference between them is that consociation is the stronger medicine: While consensus democracy provides many incentives for

Arend

Lijphart

79

broad power-sharing, consociation requires it and prescribes that all significant groups be included in it; similarly, consensus democracy facilitates but consociational democracy demands segmental autonomy. Thus, for a plural society with very deep and intense cleavages, such as Lebanon or South Africa, a consociational rather than mere consensus government would be recommended (see Lijphart 1985a). It should also be noted that Switzerland, which has served as my prime example of consensus democracy, is a thoroughly consociational system, too. For instance, the Swiss federal executive is not only a grand coalition of political parties but also of the major religious and linguistic groups. In addition, proportionality is the basic norm, not only for political representation, but also for civil service appointments. The Swiss example makes clear that consociational and consensus democracy can easily be combined. This means that if we prescribe consociation for a deeply plural society, the consensual features that are not also part of consociation, such as strong bicameralism and judicial review, can be added to the recommendation. Although there are several differences between these two forms of democracy, these differences do not entail any conflict. Consociational and consensus democracy are perfectly compatible (see Lijphart 1985b).

Problematical Cases I have found that most experts on the twenty-two democratic systems that are classified in Table 4.2 agree that these systems appear to be correctly placed. Where doubts have arisen, these entailed suggestions that slight adjustments were necessary; no radical moves from one side of the matrix to the completely opposite side have been proposed. Let me review the placement of some of the individual cases with an emphasis on those that seem to pose the most serious problems. Three of the four countries that served as the initial examples of majoritarian and consensus democracy—the United Kingdom, New Zealand, and Switzerland—are classified, as expected, in the diagonally opposite top left-hand and bottom right-hand pure cells. It should be emphasized that this did not simply happen by definition but as the result of a factor analysis of the various operational indicators of the characteristics of majoritarian and consensus democracy. Belgium is not in the same cell with Switzerland, but it is a near miss: Its factor score of -.48 on Dimension II narrowly misses the arbitrary -.50 cutoff point. Moreover, its placement reflects the entire 1945-1980 period; contemporary Belgium would certainly go into the bottom right-hand cell as a result of the federalization that began around 1970. I should like to add a note on the United Kingdom. I have also called the majoritarian model the "Westminster model," and, in fact, I have tended to

80

Democratic

Political

Systems

use the two terms as synonyms. This does not mean, however, that the majoritarian model is simply abstracted from the institutions and practices of British democracy. Instead, as stated at the beginning of this chapter, the characteristics of the Westminster-majoritarian model are derived logically from the principle of majority rule. For instance, Britain has a bicameral parliament, but this does not mean that bicameralism is characteristic of the Westminster-majoritarian model. Unicameralism is the logical way of arranging the legislative function from the perspective of majority rule, and, in this respect, Britain deviates from the Westminster model. In spite of this and a few other small deviations, the factor analysis still assigns Britain to the pure majoritarian category. It must be stressed that this is the result of an empirically discovered pattern and not because British democracy and the Westminster model coincide by definition. The six countries that have formally federal constitutions—Australia, Austria, Canada, Germany, Switzerland, and the United States—are, as expected, all classified in the consensual category of Dimension II, which can also be called the federal-unitary dimension. The only surprise is that unitary Italy and Japan are also classified as "federal." The Japanese case is certainly not a misclassification: Japan has a quite decentralized government—more so than Australia and Austria—as well as a rigid constitution, judicial review, and bicameralism. The Italian case is much more doubtful because, while Italy shares the latter three features with Japan, it is not at all decentralized. This problematical case suggests that the degree of centralization should be given greater weight in Dimension II. Perhaps the greatest surprise of Table 4.2 is that the consociational countries (Austria, Belgium, Luxembourg, the Netherlands, and Switzerland) and the semiconsociational ones (Canada and Israel) do not all cluster in or near the pure consensual category, of which Switzerland is the lone occupant. One explanation is that the consociational principle of segmental autonomy takes the form of corporate federalism in these countries more often than that of territorial federalism. The main exceptions are Switzerland and Canada, and these countries are indeed consensual on Dimension II. Austria is also consensual on this dimension, but this is a coincidence because its territorial federalism is only tangentially related to segmental autonomy. The remaining four countries are all characterized by nonterritorial segmental autonomy; hence, they should not be expected to be consensual on Dimension II. The second explanation is that some of the consociational countries—Austria, the Netherlands, and Luxembourg—became less consociational during the 1945-1980 period covered by Table 4.2. The change occurred in the late 1960s, and particularly Austria and the Netherlands can be regarded as not more than semiconsociational during the last decade of the period. Third, even when we take the above two factors into consideration, we would still expect the consociational and semiconsociational countries to be

Arend

Lijphart

81

consensual on Dimension I—which turns out to be the case for only four of the seven. How can we explain the deviant cases of Austria, Canada, and Luxembourg? The explanation is relatively easy in the Canadian case. Canada is only semiconsociational, and its consociationalism has been weaker with regard to the grand coalition than the segmental autonomy principle. Thus, we should expect Canada to be more consensual on Dimension II than Dimension I—as it is. However, there is a real problem of measurement, too. Canada's one-party cabinets, especially the Liberal ones, have tended to be linguistic grand coalitions—a kind of power-sharing that is not captured by my classification of cabinets into minimal winning, oversized, and minority cabinets, which focuses exclusively on the party composition of cabinets. Luxembourg is not a sharply deviant case because it is in the intermediate instead of the majoritarian category on Dimension I. Its political parties are segmental in nature, but almost all cabinets have been minimal winning ones. What is important here, however, is that the cabinets have included all three large parties on a kind of rotating basis so that, over time, the minimal winning cabinets can actually be called grand coalitions (Lijphart 1984,65-66). The explanation in the much more deviant Austrian case is that its plural society consists of two large segments that naturally manifest themselves as two large parties—making the Austrian party system superficially similar to the majoritarian model's two-party system. But here, too, we have a considerable measurement problem. Austria had a grand coalition of these two large segmental parties from 1945 to 1966, but during most of this twenty-one-year time span, neither party had a parliamentary majority, so that these coalitions have to be classified as minimal winning in spite of their genuinely "grand" nature. Perhaps a minimal winning coalition with more than about 80 percent parliamentary support should be counted as oversized. Such a different criterion of measurement would move the Austrian case quite a bit in the direction of the consensual category on Dimension I. At first blush, it also seems surprising that the French Fourth and Fifth Republics are in adjacent cells, though the two Republics are often described as completely different and though the Fifth Republic was engineered by political leaders who detested the Fourth. When we take a closer look, however, it becomes clear that no great changes occurred on Dimension II at all. On Dimension I, the extreme multiparty system of the Fourth Republic changed to a simpler two-bloc system, which, however, was still a multidimensional multiparty system. And there continued to be a tendency toward oversized coalitions. In short, the differences between the Fourth and Fifth Republics are great when we compare them with each other but much less great in comparison with the other twenty democratic systems. In my opinion, they are correctly placed in Table 4.2 (for a more extensive discussion, see Lijphart 1985c.)

82

Democratic

Political

Systems

The case about which I have the most doubts is the United States. I believe that it is correctly classified on Dimension II: It is a strong federal system in every respect. But its classification as majoritarian on Dimension I seems odd in view of its balanced executive-legislative relationship. Moreover, the U.S. two-party system consists of two uncohesive parties, quite unlike the British or New Zealander two-party systems. It might be more accurate to call it a two-bloc multiparty system—not a typically majoritarian characteristic. This is again mainly a problem of measurement. If I would disregard my operational indicators and the result of the factor analysis, I would probably place both the United States and Austria in the intermediate category of Dimension I and Italy in the intermediate category of Dimension II. But I would not make any other changes in Table 4.2: The remaining nineteen democratic systems appear to be placed where they properly belong.

Causes: Pluralism, Population Size, and Political Traditions How can we account for the way in which the twenty-two democratic systems are distributed in Table 4.2? The figure suggests three causal explanations. First, there is a clear relationship between the degree to which the countries are plural societies and their type of regime. The plural societies (Austria, Belgium, Israel, Luxembourg, the Netherlands, and Switzerland) and those that can be considered semiplural (Canada, Finland, France, Germany, Italy, and the United States) are italicized in Table 4.2. As we move from the upper left-hand cell to the lower right-hand cell, we encounter plural and semiplural societies with increasing frequency. There is another, perhaps even more convincing, way of interpreting the evidence in Table 4.2. The relationship between pluralism and democratic type would be perfect if we found only nonplural societies in the three cells in the upper left-hand corner (that is, the cell in the extreme corner plus the two adjoining cells), only plural and semiplural societies in the three lower right-hand cells, and a mixture of plural, semiplural, and nonplural societies in the three in-between cells (that is, the three cells on the diagonal from the lower left-hand to the upper right-hand corner). The empirical evidence comes very close to such a perfect fit: Twenty of the twenty-two countries are in the expected pattern, and only Japan and Luxembourg are deviant cases. Second, population size is linked with the second dimension of the majoritarian-consensual contrast. If the twenty-one countries are divided into eleven small and ten large countries, with a population of 10 million as the dividing line, we find that the left-hand column contains only small countries, with one exception (Britain), that the right-hand column contains only large

Arend Lijphart

83

countries, with two exceptions (Austria and Switzerland), and that the middle column contains small and large countries in equal proportions. Britain is a strikingly deviant case, but Austria and Switzerland are among the larger of the small countries. Third, the strong influence of the Westminster model in countries with a British cultural heritage is very clear with regard to Dimension I. Table 4.2 shows an almost perfect dichotomy between Anglo-American countries, in the top row, and all other countries, in the middle and bottom rows. The only exceptions are Austria and Germany, but, especially in the latter case, we should remember the strong influence on postwar politics of Britain and the United States as occupying powers. These relationships are summarized in the first three columns of Table 4.4. It presents the correlation coefficients between the degree of consensus democracy (the factor scores on Dimensions I and II as well as the sum of these scores) on the one hand and the three causal variables on the other hand. Because the populations vary widely in size, I followed the conventional procedure of using the logarithms of the population sizes. For the pluralism and British heritage variables, I used threefold classifications into plural, semiplural, and nonplural societies, and into those with British traditions, an intermediate category consisting of Germany, Austria, and Japan, and those having no Anglo-American traditions. As expected, the strongest correlations are between British heritage and Dimension I—obviously a negative correlation—between population size and Dimension II, and between pluralism and the two dimensions together. However, pluralism correlates especially strongly with Dimension I. It should also be noted that these correlations are very strong indeed by normal social science standards: .46, .53, and -.73. And it is especially noteworthy that the British or Anglo-American heritage is clearly the strongest determinant.

Consequences: Democratic Stability and Quality In the previous section, I regarded the types of democracy as the dependent variable, and I tried to determine the independent or explanatory variables. In Tabic 4.4 Coefficients of Correlation Between the Degree of Consensus Democracy and Pluralism, Population Size, British Heritage, and Democratic Quality Consensus Democracy Dimension I Dimension II Dimensions I and II

Pluralism

Population (log)

+.42 +.21 +.46

-.13 +.53 +.30

British Heritage -.73 +.11 -.44

Democratic Quality +.61 -.15 +.30

84

Democratic

Political

Systems

this section, I shall turn this question around and treat the types of democracy as the independent variable. (That is, does it make a difference whether a country is a majoritarian or a consensus democracy?) Traditionally, typologies of democratic systems—such as Almond's (1956) distinction between "Anglo-American" and "continental European" democracies and the contrast between two-party and multiparty systems—have been concerned with two consequences: the stability of the political system and democratic quality (Lijphart 1968). These two dependent variables are also important in the analysis of majoritarian versus consensus democracy. With regard to system stability, my analysis does not offer any direct evidence. The reason is that the twenty-one democracies covered in my book Democracies have all shown a high degree of stability; in fact, the criterion of selection was their having been continuously democratic from about 1945 until the present. This means that system stability is a characteristic that the majoritarian and consensus democracies treated in Democracies have in common, not something that sets them apart. I should also like to point out that cabinet stability is not a valid empirical indicator of system stability. It is more accurate to call cabinet "stability," cabinet durability. As I have mentioned earlier, durability does indicate strength, but strength only in the cabinet's relationship with the legislature —not necessarily with respect to policy effectiveness. That is why I have used cabinet durability as the operational indicator of executive dominance, the second characteristic of majoritarian democracy. Consensus democracy is characterized by a lower degree of cabinet durability, but this merely indicates a higher degree of executive-legislative balance—and no conclusions concerning system stability are warranted. However, we do have some indirect evidence with regard to system stability. As indicated above, consensus democracy can be regarded as especially suitable for plural societies because they lack a natural consensus. This line of reasoning leads to the highly plausible hypothesis that, for plural societies, consensus democracy is a more stable and effective form of government than is majoritarian democracy. The indirect evidence that supports this hypothesis is the strong link between consensus democracy and pluralism among our twenty-two cases, which we discovered earlier. It also seems plausible that this link would be even stronger if it were not for the interference of British political traditions. As far as democratic quality is concerned, the conventional argument is that a two-party system, characteristic of majoritarianism, is more democratic than is the multiparty system of consensus democracy, because it gives the voters the opportunity to choose the government they want and to hold this one-party government strictly accountable for its actions. On the other hand, since we can define democracy as government by and for the people, majority rule seems less democratic than consensus democracy because it excludes the losing party from participation in the government (Lewis 1965,

Arend Lijphart

85

64-65). This means that only if there is a regular alternation in office, and if the majority and minority parties are not too far apart in their policy objectives, can majoritarianism claim a sufficiently democratic character. The most objective way to test the democratic quality of our types of democracy is to use Dahl's (1971, 231-245) "scale types." All but one of the democracies in Table 4.2 are rated by Dahl as of 1968; the French Fourth Republic, which ended in 1958, is the exception. For the others, Dahl's scale types are indicated in parentheses, and they all fit one of his six highest categories. Scale type 1 indicates the greatest degree of democracy. Especially with regard to Dimension I, there are very large differences: Dahl gives his highest rating to almost all of the consensual democracies, a more mixed judgment to the intermediate category, and much lower ratings to all of the majoritarian systems. Table 4.4 presents the same conclusion. The correlation coefficient between the second (federal) dimension of consensus democracy and democratic quality is insignificant, but between the first dimension and democratic quality it is .61—reflecting a very strong relationship. Before we accept Dahl's evidence as definitive, however, we should take a closer look at how his scale types were formulated and whether his criteria entail the introduction of any promajoritarian or proconsensus biases. In one respect, there appears to be a promajoritarian bias: Dahl (1971, 237) states that the ten variables on which the scale types are based were selected to represent "the opportunities available to oppositions" as the "underlying variable"—a characteristic that fits the competitive and adversarial tendencies of majoritarianism more closely than does the coalescent and collaborative style of consensus democracy. On the other hand, the rating categories of two of the variables favor the consensus type to some extent. First, Dahl uses a party system variable in which multiparty systems are rated above two-party systems. The second proconsensus variable concerns the status of the legislature: A "fully effective legislature" that performs its legislative function as a "reasonably ' c o e q u a l ' branch of national government" is rated more highly than a "partially effective" legislature that tends to be dominated by the executive (Dahl 1971,239-240). The remaining variables are genuinely evenhanded: freedom of press; freedom of group association; a competitive party system without party bans; strong parties and interest groups; and so on. Overall, in my judgment, Dahl's criteria do favor the consensus over the majoritarian type of democracy at least slightly. Thus, some skepticism regarding the evidence cited above is clearly justified. But I should like to make two further observations. The first is a rhetorical question to democrats inclined to, or mainly accustomed to, majoritarian democracy: Is there not a good case to be made, in terms of pure democratic quality, for a strong instead of a submissive legislature and for a flexible party system that is not effectively closed to challengers to the existing parties? My second point is

86

Democratic

Political

Systems

that I am much less interested in proving the superiority of the consensus type than in disproving the facile assumption, prevalent in Anglo-American dominated political science, that majoritarian democracy is the best possible form of democracy. The evidence is certainly strong enough to support this negative conclusion. It is worth stressing that these findings concerning stability and quality partially contradict the main propositions of the earlier typologies. Two-party systems and Almond's Anglo-American systems coincide roughly with my majoritarian type, and multiparty and continental European systems with my consensual type. But whereas two-party and Anglo-American systems were rated more highly on stability and quality, I rate the consensus type of democracy more highly in both respects, but with three important qualifications: (1) As far as system stability is concerned, consensus democracy is superior to majoritarian democracy for plural societies; (2) the democratic quality of consensus democracy is superior insofar as its first dimension is compared with the same dimension of majoritarian democracy; (3) my conclusions are tentative and should be treated with caution because of the indirect and ambivalent nature of the evidence. The negative case can be stated with much greater force: There is no good reason to regard consensus democracy as inferior to majoritarianism with regard to either system stability or democratic quality.

Note This chapter is a slightly revised version of an article originally published in the Journal of Theoretical Politics 1 (l)(January): 33—48.

References Almond, Gabriel A. 1956. Comparative political systems. Journal of Politics 18 (3) (August):391-409. Butler, David, and Austin Ranney. 1978. "Theory." In Referendums: A Comparative Study of Practice and Theory, ed. D. Butler and A. Ranney. Washington, DC: American Enterprise Institute. Dahl, Robert A. 1971. Polyarchy: Participation and opposition. New Haven, CT: Yale University Press. Friedrich, Carl J. 1968. Trends of federalism in theory and practice. New York: Praeger. Laakso, Markku, and Rein Taagepera. 1979. "Effective" number of parties: A measure with application to West Europe. Comparative Political Studies 12 ( 1 ) (April):3-27. Lehner, Franz. 1984. Consociational democracy in Switzerland: A political-economic explanation and some empirical evidence. European Journal of Political Research 12 l(l)(March):25^2. Lewis, W. Arthur. 1965. Politics in West Africa. London: Allen & Unwin.

Arend

Lijphart

87

Lijphart, Arend. 1968. Typologies of democratic systems. Comparative Political Studies 1,1 (April): 3-44. . 1977. Democracy in plural societies: A comparative exploration. New Haven & London: Yale University Press. . 1984. Democracies: Patterns of majoritarian and consensus government in twenty-one countries. New Haven & London: Yale University Press. . 1985a. Power-sharing in South Africa. Policy Papers in International Affair, No. 24. Berkeley: Institute of International Studies, University of California. . 1985b. Non-majoritarian democracy: A comparison of federal and consociational theories. Publius 15 (2) (Spring):3-15. . (1985c). From the Fourth to the Fifth Republic—and to the sixth? A note on regime change and regime continuity. French Politics and Society 11 (September) :24-31.

El Commentary Burkhard Koch

Viewed from a Marxist position, the democratic political systems typologized by Lijphart have proved to be effective forms of political organization of society under capitalist socioeconomic conditions. Lijphart has devised and continuously refined criteria for a typology of democratic political systems in the Western style and has thus provided a theoretical instrument for differentiated categorization and assessment of the various types of bourgeois democratic systems. The criteria applied in Lijphart's typology are helpful in judging the degrees of effectiveness obtainable from the various types of Western democracy under differentiated political conditions. First, the criteria for classification of Western democracy by majoritarian and consensus types are based on the existence of both common and differentiated interests in the given political mechanism and institutions of democracy. With regard to political organization, this is a theoretically relevant and practicable set of instruments for classification. Second, by making reference to social and cultural pluralism, size of population, and effects of the "Westminster model," some social, ethnic, cultural, and historical criteria have been included for typoligization and have been related to the institutional criteria. Third, socioeconomic or sociostructural conditions have hardly been taken into consideration. It is, however, a Marxist perception that these conditions represent major causes for a specific form of political organization of society and democracy. It should be worthwhile, in my opinion, to work on some criteria for classification along this line of thinking, which might prove helpful in theoretical enrichment of the typology. Some criteria that might be suitable for the purpose are: the roles played by private or state sectors in manufacturing and service industries; the contribution made by governmentsponsored investment or stimulation or control of investment; size of the armament sector and its contribution to gross national product; and dimensions of the tertiary sector. Lijphart's typology proves suitable for assessment of processes actually 88

CommentarylBurkhard

Koch

89

taking place in the development of the political systems in the countries he appraises for classification. A trend to the majoritarian type of democracy was implied in the long-time predominance of polarization by right-left stereotypes in Western European states or between liberals and conservatives in the United States, but those trends toward polarization are, at present, supplemented or even partially overshadowed by increasing diversification in Western societies. Growing flexibility, the emergence of new social movements, emancipation efforts undertaken by various groups of the general public, and sociostructural changes have been conducive to a trend toward the consensus type of democracy. It is for that reason that the relationship between divided societies and consociational and consensus democracy, on the one hand, and homogeneous societies and majoritarian democracy, on the other, appears to be of particular theoretical interest. I should like to discuss this relationship between society and type of democracy by making reference to some of Lijphart's criteria, especially under Dimension I, which I feel are relevant in this context. 1. Concentration of executive powers versus executive power-sharing: there is a certain trend toward power sharing even in typical majoritarian executive administrations. This seems to reflect mounting pressures to have former single-party cabinets replaced by coalition cabinets and to admit independent experts. 2. Two-party versus multiparty system: the effective number of political parties is rising even in relatively stable majoritarian democracies. This can be seen in the United Kingdom, even though Liberals and Social Democrats are temporarily losing ground. This general trend also may be seen in the appearance and parliamentary activity of environmental or right-wing extremist parties in Western European continental democracies. 3. One-dimensional versus multidimensional party system: differentiation, in reality, is growing between parties as well as within majoritarian parties. This occurs against the background of diversification of Western societies under parliamentary rule. The right-left pattern is increasingly superimposed on, and enlarged by, cultural-ethnic, urban-rural, foreign policy, and postmaterialist issues. Consociational and consensus democracy, no doubt, are similar, but they are not identical. I can subscribe to Lijphart's view, according to which consociational democracy is a reflection of pluralistic, diversified society. His criteria (religion, ideology, language, culture, ethnicity, and race) might be expanded to include some sociostructural parameters. This would add to the theoretical weight of the approach, which might then be more effectively applied to an evaluation of the processes of increasing flexibility and diversification in Western societies. Two of the consequences that Lijphart identifies deserve special

90

Democratic

Political

Systems

comment: namely, system stability and democratic quality. I am inclined to support Lijphart's differentiation between cabinet stability and cabinet durability. Cabinet durability, considered by Lijphart as an indicator of executive dominance to characterize a majoritarian democracy, is not necessarily a manifestation of political effectiveness with regard to system stability. On the other hand, I strongly feel cabinet durability to be an important indicator of consensual democracy as regards system stability. Frequent government crises and cabinet reshuffles are, in my opinion, indicative of relative instability, at least of the political system. Unstable power relations then have to be rebalanced. Majoritarian democracies are at least temporarily capable of pitting government and cabinet stability against processes of destabilization or instability in the domestic balance of forces. However, the price that has to be paid is some sort of disqualification of larger sections of the population, who will lack proper representation for some time. This may entail the risk that sizable political and social groups can no longer be integrated, resulting in emergence of an uncontrollable potential that may endanger political stability in the long run. Democratic quality, I feel, is part of the problem. It is against the background of factual diversification of Western societies that growing validity is assumed by Lijphart's argument that in a majoritarian system the losing party is excluded from involvement in government. It may then happen that almost 50 percent of the electorate are indifferent or hostile toward the government in office because they are unrepresented and expelled from political decisionmaking. Remarkable abstention from voting has become an indicator. Under the two aspects of system stability and democratic quality, I should like to second Lijphart's cautiously formulated notion that for pluralistic Western societies consensus democracy is superior to majoritarian democracy, as an obviously more stable and more effective form of government and, let me add, of public involvement. This postulation, perhaps, is additionally confirmed if one takes into account the variety of sociostructural and other diversification processes actually taking place in Western societies. Accompanying pluralization is a compelling reason for enhancing the effectiveness of political organization for the purpose of system stability. This, in my opinion, is a point of departure for an objective trend toward the type of consensus democracy.

5 Authoritarianism Leonardo Morlino

Definition Of the approximately 175 independent countries that exist in the world today, the governments of more than 130 can be considered nondemocratic. Therefore, the first difficulty in defining authoritarianism and other nondemocratic regimes arises from the fact that the empirical referent to be included is very broad, but, at the same time, the definition must not be so general that it loses meaningfulness and explanatory power. There have been several attempts to define authoritarianism (Bayart 1976). The main problem with these is that they tend to build general models by extrapolating from a single historical experience. For example, Marx formulated the category of Bonapartism and Gramsci that of Caesarism. More recently, other authors adopt such expressions as neo-Bismarckism (e.g., Hermet 1975) or fascism. The definition that best succeeds in reconciling the need for meaningful generality with the requirement of the widest possible empirical applicability seems to be that put forward by Linz (1964). Any attempt to provide an adequate definition of "authoritarian regime" must, therefore, take this definition as its starting point. According to Linz's definition, proposed more than twenty-five years ago and widely accepted today, authoritarian regimes are those political systems with limited, non-responsible political pluralism without an elaborated and guiding ideology, but with distinctive mentalities, without extensive or intensive political mobilization, except at some points in their development, and in which a leader or, occasionally, a small group exercises power within formally ill-defined, but actually quite predictable limits [Linz 1964, 255].

There are five important features in this definition. The first, limited pluralism, concerns the political community and the actors who play a decisive role in the structures and policies of the regime. The second, a 91

92

Authoritarianism

distinctive mentality, concerns the way the regime is justified on an ideological level. The third, absent (or low) political mobilization, again concerns the political community. The fourth, the presence of a leader or small group that exercises power, refers to the authorities in the regime. Finally, the fifth, formally ill-defined limits, indicates a general aspect of the rules and procedures intrinsic to authoritarian regimes. It is essential at this point to expand further on the meaning of these dimensions. On the level of the political community, the most important aspect is the level of mobilization, or, to be more specific, the characteristics and amount of mass participation, induced and controlled from above. The political community is accorded neither autonomy nor independence. In the periods of greatest authoritarian stability, the rulers will implement policies aimed at keeping the civil society outside the political arena. However, some participation, which is neither extensive nor intense, can be desired and controlled by the leadership. This situation implies the presence of at least two factors at regime level: first, the existence of effective repressive apparatus, such as the police (either autonomous or internal to the military structure), which can carry out policies of demobilization; second, the weakness or absence of mobilization structures, such as a single party or similar state institutions (i.e., of structures that could bring about and, at the same time, control participation). Obviously, there is another implicit aspect that should not be forgotten: namely, the absence of real guarantees with regard to political and civil rights. Limited and irresponsible pluralism is also a central feature. This expression refers to the most salient actors in the regime and in the political community, and brings out three further aspects of authoritarianism. First, it is necessary to establish who the important actors in each authoritarian regime are. In doing this, we can consider the second aspect: We can distinguish between institutional and politically active social actors. Institutional actors are, for example, the army, the bureaucracy (or a part of it), and the single party (if one exists). Social actors are the church, industrial or financial groups, landowners and, in some cases, even trade unions or transnational economic groups that have important interests in the country. It is very clear, however, that in every authoritarian regime, there are many, or at least more than one, actor to be considered at one time or another. Finally, the third aspect is that these actors are not held politically responsible by means of the mechanism typical of mass liberal democracy (i.e., free, competitive, fair, and recurring elections). If there is "responsibility," it is meaningful only on the level of "invisible politics," in the bargaining and negotiations among, for example, the military, economic groups, and the landed oligarchy. Furthermore, the elections or other forms of democracy that may exist, such as direct consultations through plebiscites, have no real significance and do not have that quality of effective competition characteristic of democratic systems. They have, above all, a symbolic

Leonardo

Morlino

93

significance of legitimation, as an expression of consensus and support for the regime on the part of a controlled and nonautonomous civil society. This does not, in fact, mean that the elections in authoritarian regimes are of no importance. On the contrary, they have various functions, of which that indicated above is only one of the most important (Hermet, Rouquid & Linz 1978). The notion of limited pluralism has been extensively criticized by scholars who have analyzed Linz's definition. The principal argument used is that the term "pluralism" implies that a variety of actors may enjoy some kind of legitimacy and that, because of this, the definition narrows the gap between the authoritarian and the democratic regime, while it neglects the overall reality of suppression and the impossibility of expressing either political or civil rights or the real demands of the civil society, which are intrinsic characteristics of democracy. For this reason, the argument goes, the notion distorts the reality of authoritarianism. But this objection can be overcome once it has been made clear that the pluralism in question is not the unlimited pluralism of mass liberal democracy, which is characterized by competition and proper guarantees for civil and political rights. Limited pluralism simply means that more than one actor of the elites is important for the authoritarian regime. Finally, it should not be forgotten that the adjective "limited" itself implies the existence of control and repression on the part of the ruling class. The objection, indeed, was more the result of a misunderstanding than anything else, and than it has been overcome is evident from the second most important definition of authoritarianism (Rouqui6 1975). Rouqui6, in fact, takes a further step precisely in the direction of the identification and specification of a plurality of important actors. Starting from the analysis of the concept of Bonapartism in Marx, Rouquid (1975,1098) goes on to say: [A] semicompetitive system of the Bonapartist type—for want of a better term—is a regime that is supported by a civil and military bureaucracy, that is relatively independent of the dominant social groups, and that tries to resolve the conflicts that paralyze the ruling elites by depoliticizing all the social classes, in a nonviolent (or nonterrorist) way.

A careful reading of this definition shows that there are no substantial differences between it and that of Linz, though Rouquid does pay more attention to the policies carried out than to the institutions and, above all, arrives at the identification of a more specific type of authoritarian regime: His definition is more useful for identifying bureaucratic-military regimes than for analyzing authoritarianism, in general. Furthermore, the notion of limited pluralism is flexible enough—and herein lies its greatest advantage—to be applied to situations that differ on the level of socioeconomic development and in terms of the specific actors who

94

Authoritarianism

are salient in each case. In addition, the notion suggests how crucial the identification of the important actors is for a better understanding, both of the structure of the regime and of the policies that it implements. It enables us, then, to bring in another very useful concept, that of the dominant coalition. By dominant coalition, I mean the actors, institutional and otherwise, that form, if only on a de facto basis, the alliance that, to varying degrees, maintains each specific model of authoritarianism. The concept of dominant coalition will be dealt with more specifically below. Suffice it to say, for the moment, that a precise empirical identification of the form of limited pluralism, and the type of dominant coalition involved in each specific case, facilitates a less trivial analysis of the internal dynamics of authoritarian regimes. Moreover, if we define pluralism in terms of the presence of a number of actors in the political arena, compared to a situation where there is only one such actor who is capable of monopolizing all the politically relevant resources, it is easier to understand how an objective space for the opposition can exist within such an authoritarian regime. Both Linz (1973) and Germani (1975) have analyzed the different types and forms of opposition, semiopposition, and pseudo-opposition that can exist within this kind of regime: active or passive; legal, "a-legal," and illegal opposition. In fact, it can be very convenient for the regime to tolerate a certain degree of opposition or to maintain a pseudo-opposition that gives a liberal façade to its authoritarianism. The third characteristic concerns the extent to which an ideological justification of the regime has been elaborated. Authoritarianism is distinguished by the fact that its legitimation takes place on the basis of "mentality," to use an expression borrowed from the German sociologist Geiger and adopted by Linz; that is, simply on the basis of "intellectual attitudes" and general, more or less ambiguous, values, on which the various actors who have different characteristics and interests can agree fairly easily (Linz 1975, 266-269). These are values such as fatherland, nation, order, hierarchy, or authority. There are, therefore, no clearly articulated, complex ideologies that justify and support the regime. On the level of political structures, Linz refers to the limits within which the authoritarian rulers exercise their power, limits that are, "formally, illdefined, but quite predictable in practice." The fact that these limits are defined only approximately and do not have the prescriptive "certainty of the law" characteristic of democratic regimes gives the rulers a greater degree of discretion in the exercise of their power. In other words, if, as a wellestablished theoretical tradition suggests, the law is the main guarantee of the citizen and the very foundation of Western democracy, the absence of well-defined norms means that the civil society can be easily controlled, and that the citizen is deprived of certain spheres of guaranteed autonomy vis-à-vis the authoritarian power.

Leonardo

Morlino

95

The last dimension in Linz's definition refers to the "authorities" and, more precisely, to the leader or small group in power. In effect, authoritarian regimes are characterized by a notable personalization of power, to the visibility of the sometimes charismatic leader or of the few people who, in fact, do hold the reins of power and are members of the main decisionmaking bodies.

Models of Authoritarianism and Transitional Regimes The definition outlined above is a good point of departure for analyzing changes either within an authoritarian regime or toward a different political arrangement. From this perspective, the next step is to identify the principal forms of authoritarianism on the basis of empirical observation of actual regimes from the post-1945 period onward. In this way, it is possible to identify several forms of authoritarianism on the basis of the dimensions pointed out by Linz. It should, however, be clarified that the last two aspects—the presence of a single leader or small group at the summit of power and unclearly defined but predictable limits to the authoritarian power—are not useful for distinguishing between different types of authoritarianism. With regard to the latter, ill-defined limits are a characteristic of all regimes of this type; and insofar as there is any variation along this dimension, it depends on the juridical traditions of the country and the length of time the regime has been in force. Even these variations, however, are of limited use in distinguishing between one subtype and another. One can add a fourth dimension to the remaining three, one very relevant in an analysis that focuses on political aspects; (that is, the degree to which the regime is structured). In other words, it is necessary to establish the extent to which the authoritarian regime creates and, eventually, institutionalizes various, new, characteristic political structures, such as a single party, unions (including vertically organized unions), distinctive forms of parliamentary assembly, characteristic electoral systems, or any specific organs that differ from those of the previous regime. The other three principal dimensions also require some clarification. The mere number of actors who make up the so-called dominant coalition is not important in itself, but it is crucial to indicate who these actors are: What is the dominant coalition involved, and of whom is it composed—institutional (bureaucracy, military), political (parties and trade unions), and/or socioeconomic actors (oligarchical groups of landowners, different groups of entrepreneurs, the commercial bourgeoisie)? Nor is it enough to try to establish the degree of ideologization. It is equally essential that the values that are used to justify and legitimate the regime—be they traditional,

96

Authoritarianism

modern, or otherwise—be identified and understood. The extent of mobilization from above (and its possible institutionalization) should not be treated as the most important aspect; the characteristics of this mobilization, particularly in relation to the other three aspects mentioned above, must be identified and examined. All the dimensions are, to a greater or lesser extent, interrelated. It is, therefore, possible to suggest two "polar" models of authoritarianism to which the extremes of each dimension correspond. The first pole could be labeled "perfect authoritarianism," characterized by a marked pluralism of important actors, the absence of ideology, the absence of mobilization, and little characteristic structuring of the regime. The second pole corresponds to "quasitotalitarianism," and it is distinguished by the key role played by the single party that occupies a dominant position, a high level of ideologization, high mobilization, and the existence of institutions that are characteristic of the regime. Between the two poles, there is a whole range of specific multidimensional configurations of "authoritarianism," arising from possible variations in the actors involved, in mentalities, in extent and form of mobilization, and from various possible combinations of these factors. One Table 5.1 Dimensions and Salient Variations in Authoritarian Regimes 1. Dominant Coalition (number and identity of actors) Military

Bureaucracy

Single or hegemonic party

Other Church institutional actors

Landowning oligarchy

Industrial, commercial, and transnational bourgeoisie

2. Mentality/Legitimating Ideology (identification and extent of articulation) Nation

People

Order authority

Doctrine Rationalist Marxismof national technocratic Leninism, security Maosim

African socialism, humanism, Negritude

3. Mobilization from Above (characteristics and extent) Depolitization

Elitist mobilization (limited)

Mass mobilization (controlled)

4. Structure of the Regime (degree of innovation from none to high) Head of State Juntas, (e.g., councils, president) committees

Corporative parliament, vertical unions

Single party Bureaucracy articulation, ancillary organizations

Intelligence services

Leonardo

Morlino

97

could, for example, have a case in which a low number of important actors combines with low ideologization, low mobilization, and nonexistent structuring. The range of possible variation between one form of authoritarianism and another is, therefore, so vast and multifaceted that it is necessary to outline the most important models and to describe their characteristics in greater detail. Table 5.1 gives a first idea of the variations and combinations.1 It is not possible, on this occasion, to analyze the various models of authoritarianism and, above all, the related typology. It is enough to emphasize that several authors have formulated excellent typologies moving from similar perspectives (Linz 1975) or from different ones (e.g., Lanning 1974; see also references below). Without the aim of being exhaustive, but by way of example, Table 5.2 presents the main models and types of authoritarianism grounded in the previous discussion. The first suggestion is to look at the present role of the main political actors in the actual decisionmaking process (e.g., limited pluralism and dominant coalition). Regimes wherein the army is the main actor can be distinguished according to the number of incumbent leaders (tyranny or oligarchy) and to the role of the military organization in terms of control and penetration of political structures (governing-military regimes have stronger military penetration, repression, and goals of change) (Nordlinger 1977; Finer 1975). Regimes wherein the dominant coalition is formed by civilian groups and the army can be characterized simply by the prominent role of the bureaucracy and the army, or by a more complex political organization and ideology called "organic statism" or "corporatism," aimed at the inclusion or exclusion of some social groups from the controlled political system (Stepan 1978). A common requisite of civilian regimes is the existence of mobilizational structures and their durability. The successful or unsuccessful establishment of those regimes is related to different geopolitical areas in different periods; and the ideology is a very characteristic feature, be it the communist one in Eastern Europe or Asia, the nationalist one mainly in Africa, or the fascist ideology in Italy. While authoritarianism is the central category of reference, there are also other empirically important nondemocratic regimes to be considered: totalitarian, traditional, and transitional regimes.2 There is a vast amount of literature on the first two models, but much less on the third. As yet, it cannot really be considered a model because the level of its conceptual articulation is too low. Thus, because of restrictions of space, I shall not deal with the first two models here but shall pay closer, albeit brief, attention to transitional regimes.3 First of all, one could refer to these regimes as institutional hybrids (i.e., regimes that are no longer completely authoritarian, but that do not, as yet, fit fully into the democratic genus). Alternatively, one could adopt terms that are well known in the Iberian world, as dictablandas and democraduras

98

Authoritarianism

(Rouqui6 1975; Schmitter 1985). These regimes are mentioned here precisely because of the potential spread of this phenomenon in the context of a process characteristic of the 1980s, which have seen an everincreasing affirmation—at least on the symbolic level and on the levels of propaganda and of verbal acceptance—of Western-style democracies. The phenomenon concerns or concerned certain areas of Central America, Africa, and Asia. Institutional hybrids or, rather, transitional regimes in the narrow sense of the term are all those regimes that were preceded by an authoritarian experience, followed by a certain amount of opening up, of liberalization, and of the partial collapse of the limitations on pluralism. This means that, as one result of (inter alia) an increased respect for civil rights, a clear opposition emerges alongside the old actors who formed the now no longer dominant or cohesive coalition of the preceding authoritarian regime. The forces of opposition are allowed to participate in the political process but are substantially excluded from every possibility of acceding to government. There are, therefore, several parties, but one usually remains dominant in semicompetitive elections. At the same time, there is intense, real competition among candidates of the dominant party. The other parties are not well organized, are new or newly reestablished, and do not have a large following. Real participation does exist and not only at election time, but the level of this participation is not very high. An extremely biased electoral law ensures that the dominant party, which is very similar to a bureaucraticclientelist structure, retains its enormous advantage in the distribution of seats. This means that, already, there is no justification for the existence of the regime, even on the basis of comprehensive and ambiguous values. Authoritarian mobilization, if it ever, in fact, existed, is only a memory of the past. Forms of police repression have also almost completely disappeared, and, because of this, the political importance of the respective apparati has declined. On the whole, there is little institutionalization and, particularly, little organization of the "state," if not, indeed, a process of actual deinstitutionalization. The military can, however, continue to play an obvious Table 5.2 Models of Authoritarian Regimes Transitional Regimes Military Regimes

Military-Tyranny Military-Oligarchy Guardian-Military Governing-Military

Civil-Military Regimes

Bureaucratic-Military Corporative Exclusive Corporativism Inclusive Corporativism

Civil Regimes

Nationalist-Mobilizational Communist-Mobil izational Fascist-Mobilizational

Leonardo

M or lino

99

political role, even though this will be decreasingly direct and explicit. To understand this model better, one could add that the transitional regime emerges from an at least provisionally successful attempt by the moderate sectors of the ruling coalition in the previous authoritarian regime to resist the internal and external pressures on the dominant coalition, to continue to maintain order and the existing arrangements for cost-benefit distribution, and to satisfy in part—or at least to be seen to satisfy—the demand for transformation in the direction of democracy demanded by other actors, whose participation is successfully limited. It is necessary to specify further that there can be as many variations of transitional regimes as there are models of authoritarianism (i.e., this model should be included in the processes of transformation of the authoritarian regime, which will be dealt with in the second part of this chapter). There are many cases that could fit into this model, and this is indicative of its potential importance. 4

Transitional Dynamics, Persistence, and Crisis In order to show in the best way how the above-illustrated definition and the sketched typology are useful and relevant, I will now look at the dynamics of the passage to an authoritarian regime (transition and installation) and at those of its internal transformation (consolidation, persistence, and crisis). For lack of space, here I shall give only the definitions of the first four processes, and I shall pay more attention to the crisis and the possible following breakdown, which are the most interesting processes in the last decade or so. 5 Following the breakdown of the previous regime, "transition" refers to a phase wherein two or more opposed political coalitions confront each other, but neither (or none) succeeds in imposing itself completely and immediately on the other. The extreme example of the transitional situation is found in civil war. The second characteristic of the transition is the dismantling of the old structures, the institutional fluidity and the absence of new political structures, or rather of the norms and procedures that will, eventually, form the new regime. No political coalition has successfully asserted its dominance and, at the same time, its own characteristic political structures, its "own" regime. The period of transition closes when one coalition of actors succeeds in imposing itself on the alternative coalition and gains monopoly or at least sufficient control of the coercive resources. The installation of a new regime occurs, therefore, when a coalition of political actors succeeds in creating and implementing the norms and structures characteristic of the new regime, and when the new leaders have a monopoly or, at least, control of the coercive structures. There are, therefore, two distinctive processes involved in the establishment of an authoritarian regime. The first concerns the creation of

100

Authoritarianism

new, and the elimination or maintenance-adaptation of old, institutions. This process is, to a large extent, influenced by the ideology and political programs of the leaders in power. The second process is at the level of the political actors that form the founding coalition of the regime and concerns the development of the particular, internal dynamic that this coalition evolves as a result of the various substantive problems it has to face and of all that was involved in the first process. In providing some further clarification of the notion of the coalition of actors that found the new regime (i.e., of the dominant coalition), suffice it to note that by this term is meant, broadly, the complex of politically active social groups who support the regime in the installation phase and in the following periods or, as one could also say, the social base of the regime and the elites, who are the direct or indirect expression of those groups who participate in the government of the authoritarian regime insofar as they occupy commanding positions in the key political structures. The important point is that these groups and their elites form a coalition that is sometimes simply a de facto arrangement and, at other times, is the result of a conscious, explicit agreement on the concrete ways of resolving political conflicts, such as, for example, class conflict and all the specific questions that arise from it. The coalition is dominant in terms of coercive resources, of influence, and of status, which the actors present in the political arena use concretely to attain their own goals. It is, therefore, dominant at the time, above all, when the regime is being installed. This does not eliminate the possibility that other resources of the same kind exist, but they are not used at certain crucial moments. It should also be added, to avoid ambiguity, that dominance in terms of resources also involves the relative consideration of the field of the possible or actual opponents of the regime and of the coalition that supports it. In this sense, it is always a relative concept as far as the resources used in the political arena are concerned. Once the regime has been installed, the coalition may gradually modify itself: Certain minority actors may be emarginated, and some actors may acquire greater dominance than others, either as a direct result of the process of installation itself or as an internal effect of external events. Installation ends when a coalition of political actors has a monopoly of the coercive arena, when its members occupy the principal roles of authority, and when it has built the principal institutions of the new regime. In short, establishment consists in the takeover and shaping of government, and the process is at an end when this has occurred. The next phase is the takeover of power—that is, consolidation. This second term can be defined as the process by which the coalition of political actors, dominant in the new regime, organizes, orders, enlarges or constricts, and reinforces itself by: completing the destruction of some of the structures of the old regime and/or by transforming others in accordance with the modes, policies, and ideologies desired by the new leaders; expanding and perfecting the control of the

Leonardo

Morlino

101

coercive resources; achieving "autonomization" and, eventually, the decisional efficacy and efficiency of the new structures; and broadening and increasing the degree of internal and international legitimacy of the new regime. In a sense, consolidation is a reversal of the process involved in moving from crisis to collapse, transition, and establishment: From recourse to violence and the almost exclusive reliance on coercive resources, there is a renewed emphasis on the importance of other aspects, such as legitimacy, efficacy, and effectiveness. Authoritarian consolidation can be considered substantially at an end when the institutions have been set up, and when the dominant coalition and the institutions of the regime have acquired a definite structure. There are three possible outcomes of consolidation: crisis; unstable persistence; or stable persistence. The second and third possibilities are linked to two conditions: (1) the degree of systematic organization, strength, and breadth of the dominant coalition; and (2) the degree of legitimacy and, above all, decisionmaking efficacy, even on the level of repression, acquired by the regime. If there are civil disorder, a strong opposition, and unresolved crucial problems within the regime, and, above all, if there is conflict within the dominant coalition about the ways in which basic conflictive issues should be settled, and if, in addition, the coercive and influencial resources that the political actors have at their disposal are barely sufficient to maintain the regime, this is a definite case of weak consolidation. Its outcome will be unstable persistence. The opposite case is that of strong consolidation and stable persistence. A crisis in the authoritarian regime, which indicates the failure of consolidation, can occur immediately after the installation process or at a distance of a few years. The principal differences between these two possibilities lie, on the one hand, in the various reasons why the crisis occurs and, on the other, in the different ways in which it develops. Apart from other hypotheses, which are not considered here, the central, initial hypothesis is that the conditions for authoritarian crisis exist when the dominant coalition at the basis of the regime begins to, and then does, fall apart; in other words, when the agreement on which the authoritarian regime rests—and which can vary in explicitness but always refers to substantive problems—weakens or collapses. This general hypothesis must be clarified and broken down further. First of all, one should pay particular attention to the connections among the various actors in the coalition, to the process of institutionalization that may have come about in the regime, and to the consequent autonomization of the institutions of the state vis-à-vis some actors in the coalition, particularly the politically relevant social groups. It is obvious that the political institutions try to consolidate themselves, but this often involves the institutional military and/or bureaucratic elites separating themselves from the other actors in the coalition, which is thus made less cohesive. Decreased cohesion has limited consequences if the regime has really been

102

Authoritarianism

institutionalized and has reached complete control of the coercive arena and, possibly, some legitimation or at least passive support on a mass level. This situation does not, however, occur very frequently (e.g., in Latin American countries). What happens more often is that the coalitions themselves are fluid, not very cohesive, and weak on the level of resources; that objective conflicts of interest emerge within them, and that these make the substantive agreements at their base weak and ambiguous; and that, above all, there are divisions and potential conflicts within the institutional actors, the military in particular, about the substantive policies to be adopted. This is more likely to be the case if the regime finds it difficult to achieve its own autonomous mass legitimation because the social groups of the middle and working classes have already experienced political organization and participation of a democratic type. If this is the case, however, the authoritarian rulers often pursue pseudodemocratic legitimation by means of rigged elections, compulsory referenda, or puppet parliaments. What are the "causes" of the authoritarian crisis in such a situation? If the regime has been consolidated, then it will be necessary to look at more long-term, as well as more short- and medium-term factors. If, however, consolidation has not taken place, the empirical analysis will focus, necessarily, on the short and medium terms. With regard to long-term factors, the following three hypotheses are the most important: First hypothesis: if there are changes in the structure and nature, and in the choices and preferences, of the social and economic groups that form (or support) the dominant coalition, then these will tend to modify that coalition. Second hypothesis: these modifications can lead to the exit from the coalition of some actors who will then become active or passive opponents of the regime; and/or to internal tensions, conflicts, and demands that the regime itself be changed through adaptation. Third hypothesis: socioeconomic transformation can give greater resources of influence and coercion to new actors, who have been excluded from the dominant coalition, enabling them to mobilize against the regime. It is necessary to look at the principal socioeconomic changes that occur and to identify all the consequences that these have for the dominant coalition and the political structures. In the middle and short runs, however, the breakdown or gradual, progressive erosion of the dominant coalition can be due to at least four phenomena. First, divisions may arise within the armed forces, which are due simply to personalized struggles for power, to a change in leadership, to ideological differences or differences in objectives that have emerged within the military institution, or, finally, to concrete difficulties that the implementation of certain policies involves (Dix 1982). More specifically, Finer (1975, 176-177) correctly emphasizes the potential for divisions, tensions, and problems brought about by the fact that some members of the military (i.e., those in power), hold positions that, in the long run, can become

Leonardo

Morlino

103

very different from those held by their colleagues who remain in the barracks. This potential for division is, obviously, considerably less in the case of a regime in which a large part of the military is involved, on different levels and with different functions, in the "administration" of the regime itself. It does, however, exist in other military and civil-military regimes. Finally, if economic policies are a substantial failure, or if any other situation that threatens the unity, cohesion, hierarchy, or reputation of the military arises, it may proceed to disentangle itself simply to safeguard those "values" that strongly condition and limit its political involvement and modus operandi. Second, divisions can arise between the armed forces and the civil actors in the coalition if the military wants to affirm its dominance within the coalition; if the policies imposed by the civilian elements become unacceptable because, for example, they do not succeed in establishing the order and stability required by the armed forces; or if those objectives already seem to have been achieved, and the military is no longer ready to put up with and make others put up with the economic and human costs of these policies. These factors can, in certain situations, lead the military to reconsider the costs and the problems involved in its presence in politics and to withdraw its support for the regime or to disentangle itself from direct responsibility for government under certain conditions that satisfy its substantive demands and guarantee immunity from retaliation for its repressive activities when it was in power. Third, the civil elites may leave the dominant coalition because the policies, particularly the economic policies, launched by the regime become bankrupt, or because they feel the time has come to implement other policies that go beyond the sometimes excessively nationalistic outlook of the military. This specific "cause" is obviously more likely to arise where there is a more developed industrial structure and a stronger group of capitalist bourgeoisie that, at first, participated in the coalition in support of the regime. In fact, this can be seen as the basis of the crisis in some Latin American countries, such as Brazil, Argentina, and Uruguay and, in less recent times, in Peru and other countries (e.g., the fall of Pérez Jiménez in Venezuela in 1958).6 Fourth, defeat in a local war could revive divisions within the armed forces or lead to the detachment of the civil elites from the military, which would then be very isolated at the mass level, where passive support or opposition or, indeed, apathy could be transformed into open and intense opposition. In this case, it is often another sector of the military that, in order to maintain or reconstitute the unity, cohesion, and prestige of the armed forces, compels the military rulers to withdraw and calls civilians to government. As far as this short-run set of factors is concerned, one should emphasize that the international factor is of the greatest importance. To put it briefly, it may be the decisive intervening variable in explaining both the crisis and the

104

Authoritarianism

collapse. Before leaving this point, it is important to stress that, in a given empirical situation, there will be not one, but several interrelated and mutually reinforcing "causes." In any case, since the coercive arena and the results of the conflicts that happen in it usually play a part of the greatest importance in determining the results of the crisis, everything that concerns the military—the divisions that exist within it and its possible isolation from the other actors and from the society—"weighs more heavily" than do the other "causes." In analyzing the ways in which the crisis can develop, then, it is important to note that, in regimes that have been installed and consolidated for some time, a pattern of recurrent manifestations and reactions emerges as soon as the crisis begins. In fact, these regimes are better equipped on the institutional level to overcome both the difficulties that arise within their own dominant coalition and the pressures from the civil society, which are more limited and less likely to be voiced. The attempts to overcome the crisis that results from these difficulties go in two directions: continued repression of the civil society and, at the same time, the creation of democratic openings that turn out to be only apparent and a facade. For example, the regime reacts to strikes, peaceful demonstrations, terrorist acts, the emergence of clandestine trade unions and parties, and to all the propaganda against it (i.e., to the overall political mobilization of the civil society and, particularly, of students and workers) with dismissals, arrests, the creation of special courts and the holding of political trials, special antiterrorist laws, proclamations of a state of emergency on the local or national level, and rigorous censorship of the press. At the same time, a whole series of apparently liberalizing laws are passed, but these are only a facade and do not change the repressive substance of the regime. Some further observations on the modalities of the authoritarian crisis are necessary. First of all, it is necessary to distinguish between the cases in which, despite everything, the crisis continues for a long time and those in which crisis leads to the collapse of the regime and to its transformation. Although it is not necessary to dwell on the obvious distinction between the process of crisis and that of collapse, it is important to emphasize other aspects and manifestations of the crisis and the additional, but indispensable, conditions that lead to the collapse of the authoritarian regime. Regardless of why it occurs, the breakdown of the dominant coalition that had supported the regime has discernible consequences on the level of visible policymaking. The intensification of the conflict among the elites involves major decisionmaking difficulties, if not utter and complete immobility, perhaps in the very sectors (e.g., political economy) over which the coalition first disagreed and then broke up. If this occurs, the divisions become so deep that they are insurmountable. The legitimation of the regime on the level of elites decreases further, while its legitimacy on the mass level

Leonardo

Morlino

105

has, presumably, always been low and limited. Another possible manifestation and consequence of the divisions within the coalition, particularly if these directly concern the military, is a reduction in the repressive capacity of the regime or an increase in the cost of repression in decisionmaking terms, especially if, at the same time, social groups that were previously halfway proregime are mobilized or activated, and there is also an increase in, and organization of, opposition. To summarize briefly what has been said up to now, and bearing in mind the definition of authoritarianism, the crisis is manifested and concretely expressed in the breakdown of the no longer limited pluralism, and in increased mobilization that the authoritarian regime can no longer control and that can come about sooner or later, depending on whether the regime collapses or moves toward a more continuous and substantially agreed-upon transformation. The authoritarian crisis can last for a long time without giving rise either to the collapse of the regime or to its gradual and continuous transformation into a different regime, if certain conditions do not exist. I refer, essentially, to three particularly important and interrelated aspects that are connected with the crisis of the dominant coalition. First, sectors of the dominant coalition detach themselves from the regime and assume positions of disengagement and, then, even of active opposition. The second is that some groups, which were previously undecided or indifferent (e.g., organized groups belonging prevalently to the middle classes), become actively opposed to the regime. These then enlarge and reinforce the more clearly antiregime opposition that already exists, which the regime has never succeeded in eliminating completely, or give rise to new and different organized movements, which are also partially or completely opposed to the authoritarian regime. The third is that the opposition, which can be clandestine and/or extremist, to varying degrees, and which has succeeded in resisting authoritarian repression, regains strength and becomes capable of a greater amount of activity. This is more likely to happen if solid social bases for the opposition exist (e.g., the church or a relatively strong working class, with previous traditions of organization). In this second phase of the crisis that lays the foundation for the transformation of the regime, the varying capacity f o r activity of the opposition groups, their various attitudes and the distribution of coercive resources of influence and pressure (even if these are only of an economic kind), are not particularly important. They will become so at a later stage when they can influence the success or failure of the installation of the new regime, which will probably no longer be authoritarian, but democratic. What is important at this stage is that opposition to the regime exists, grows, and is capable of modifying the relative strengths of the different groups within the political arena. In short, it can, potentially, provide a political alternative without which the crisis could last indefinitely. At the

106

Authoritarianism

same time, however, there is another condition that is essential in cases of crisis and transformation without collapse. Mobilization must be such so as not to frighten the actors who belong to the previously dominant coalition, above all the military. The military must, however, be made aware of all the disadvantages of remaining in government (in terms of the increasingly profound divisions that are likely to arise within its organization, if not, indeed, the complete disintegration of its main hierarchy) and, conversely, of all the advantages of returning to barracks (in terms of decreased political responsibilities and the possibility of increased social prestige, and even transformation into the protectors and guarantors of change). Alternatively, what could happen is that the divisions that exist within the dominant coalition are covered over, the conditions that gave rise to the installation of the authoritarian regime are, substantially, recreated, and the coalition is reunited.7 Thus, on the one hand, the basis for a political alternative must exist, and, on the other, a delicate balance must be created and the transformation piloted through a very difficult situation. It is on this level that the capabilities of the leaders, who are guiding the process of change, can be evaluated. At this point, one could ask: If the general conditions for the transformation of the regime, which have been outlined up to now, exist, what else is required so that this change can effectively begin? Must there be some accelerative event that can precipitate the situation? Or, alternatively, is a progressive, initially tacit but conscious emergence of an agreement or compromise among some actors in the authoritarian dominant coalition and the new and old sectors of the old opposition enough? There are no definitive answers to such questions. However, one can point out that, for able and firm leaders, an initial agreement of this kind, which they also helped to create, may be enough to bring about the change. Such leaders will be better able to profit from further favorable situations that may arise. In several cases, therefore, there is an accelerative event of an internal or international kind. In nearly all cases of crisis and change without collapse, however, it is an actor or a faction that previously belonged to the authoritarian coalition who takes the initiative in changing the regime. This fact remains, even though one can notice great differences between the political programs of such an actor and the resulting democratic regime. In none of the European or Latin American cases in the last fifteen years has the democratic opposition assumed a direct and leading role in initiating change, either because it is insufficiently organized or mobilized or for other reasons of time and opportunity. Moreover, in most of these crises, the democratic and civil opposition has played a secondary role in the downfall of the regime. At any rate, the regime finally falls or is transformed because the initiative of a group of actors or of a single actor, endowed with sufficient resources to overcome whatever residual defenses the authoritarian regime may have, has created an alternative capable of standing on its own two feet.

Leonardo

Morlino

107

In the case of the abdication of the armed forces in military or civilmilitary regimes, or in the other cases of transition to a civilian democratic regime, the next problem to arise is that of democratic installation, which is also conditioned by the modalities of, and other factors in, the crisis-collapse or the crisis-transformation of the authoritarian regime.8 There are, however, at least two other possible outcomes that must be mentioned here by way of conclusion. The first is the passage from a military or civil-military regime to a civil regime that is still authoritarian. This is a relatively rare occurrence. It is characterized, first, by the consolidation of the preceding authoritarian regime and, then, by a successive, gradual crisis and transformation to another regime in which, however, the military elites continue to be a powerful, authoritative pressure group. There are at least two essential conditions for this kind of transformation: the institutionalization of the regime, whose dependence on the support of the military gradually declines, insofar as the regime relies less and less on repression and force; and the political will of a leader capable of carrying out the changes that will liberate him, to a great extent, from "military protection" that is sometimes too oppressive and demanding. Another change from a military or civil-military regime to a civil-military or military (respectively) regime is, very often, simply the result of the division of the dominant coalition and of the consequent transformations within it, in terms both of the actors directly involved in government and of those who sustain it. It can be, for example, another sector of the military that, in the aftermath of a coup d'état, either aligns itself with civilian groups or, alternatively, expels them from government. If a transformation of this kind has been gradual, rather than clearly signaled by an event such as a coup d'état, then it is more difficult to pinpoint this change, and it could be confused with a simple alternation among the elites or within the ruling elite itself.

Notes 1. Neither the policies that the regimes implement in various sectors (industry, agriculture, education, etc.) nor the degree of repression that distinguishes them are considered among the most important dimensions here. In fact, the policies are to a large extent connected to the coalition and to the ideological attitudes that characterize a certain authoritarian regime. The degree of repression can be deduced from the two characteristics shown in Figure 5.1. If, on the other hand, this relationship does not exist, as some recent studies have shown (e.g., King 1981), it is all the more reason for not considering these policies among the distinctive elements of a regime. Also, policies and forms or degrees of repression can change more easily than can the other dimensions indicated, and are therefore less useful for understanding the structural, more stable, aspects that are intrinsic to the regimes to be analyzed.

108

Authoritarianism

2. The end of the 1970s saw the emergence of another completely new phenomenon insofar as these regimes are concerned. If, in the authoritarian and totalitarian models mentioned above, the principal structure of mobilization remains the party, and if, in some military or bureaucratic-military regimes, such as those of Lybia, Sudan, and Pakistan, the Moslem religion is an important aspect as far as both the legitimation of the regime and the identification of a politically important normative basis are concerned, the regime that has been installed in Iran is completely unprecedented. It is a monistic system, in terms of the number of actors present, with a complex ideology; an enormous capacity for, and a high degree of, actual mobilization; and interesting institutional innovations at the local and national levels. In this sense, the regime could be called a "religious neototalitarianism." 3. The idea of a transitional regime has recently come to the fore in political analysis, but it has not been developed yet. One of the first authors to mention it was Finer(1975,163). 4. From the institutional point of view, this is what Finer (1970, 441—531) seems to mean when he refers to "facade democracies" and to "quasi democracies." But a more careful analysis reveals that the first of these two models can be reduced to the category of traditional regime and the second falls into the large genus of authoritarianism. In fact, Mexico before 1978 or some African countries ruled by single parties are typical examples of "quasi democracy." 5. In this regard, the empirically most relevant situations remain those of Southern Europe and Latin America. 6. Among the very few authors who have analyzed the breakdown of authoritarian regimes, Dix (1982) is the only one who has tried a systematic comparison of Argentina (1955), Peru(1956), Colombia (1957), Venezuela (1958), Cuba (1959), and the Dominican Republic (1961). The conclusions that Dix reaches differ in part from those presented here; he emphasizes economic failure, the delegitimation of the regime, the erosion of the coalition that initially supported it, the crucial role of the military and its political disintegration, the cohesion of the opposition,the international dimension, and the guarantees of security for the supporters of the regime in crisis. 7. For example, the way the fear of popular demands can give rise to a bureaucratic-authoritarian regime has been shown, explicitly, by O'Donnell (1981, 112) and, implicitly, by R. R. Kaufman (1981, 388ff). 8. For such an analysis, see Morlino (1987), and O'Donnell and Schmitter (1986).

References Albright, D. E. 1980. A comparative conceptualization of civil-military relations. World Politics 32. Bayart, J. F. 1976. L'Analise des situations autoritaires: Etude bibliographique. Revue Française de Science Politique 26. Collier, David (éd.). 1979. The new authoritarianism in Latin America. Princeton, NJ:

Leonardo

Morlino

109

Princeton University Press. Dix, R. H. 1982. The breakdown of authoritarian regimes. Western Political Quarterly 35. Feit, E. H. 1978. The armed bureaucrats. Boston: Houghton Mifflin. Finer, Samuel E. 1970. Comparative government. Harmondsworth: Penguin Books. . 1975. The man on horseback: The role of the military in politics. Harmondsworth: Penguin Books. . 1980. Militari e politica nel terzo mondo. Revista Italiana de Scienza Politica 10. Germani, G. 1975. Fascismo, autoritarismo e classi sociali. Bologna: II Mulino. Handelman, H., and T. G. Sanders (eds.). 1981. Military government and the movement toward democracy in South America. Bloomington: Indiana University Press. Hermet, G. 1975. Dictature bourgeoise et modernisation conservatrice. Revue Française de Science Politique 25. . 1978. Les fonctions politiques des" organisations religieuses dans les régimes à pluralisme limité. Revue Française de Science Politique 23. - , Rouquié, A., and J. J. Linz. 1978. Des élections pas comme les autres. Paris: Presses de la Fondation Nationale des Sciences Politiques. Herspring, Dale R., and Ivan Volgyes. (eds.). 1978. Civil-military relations in communist systems. Boulder, CO: Westview Press. Huntington, Samuel P. 1957. The soldier and the state: The theory and politics of civil-military relations. New York: Vintage Books. -. 1968. Political order in changing societies. New Haven, CT: Yale University Press. , and C. H. Moore, (eds.). 1970. Authoritarian politics in modern society. New York: Basic Books. Jackman, R. W. 1976. Politicians in uniform: Military governments and social change in the third world. The American Political Science Review 70 (December): 1078-1097. Janowitz, Morris. 1977. Military institutions and coercion in the developing nations. Chicago: University of Chicago Press. , and J. Van Doom. (eds.). 1971. On military intervention. Rotterdam: Rotterdam University Press. . 1971. (eds.). On military ideology. Rotterdam: Rotterdam University Press. Kaufman, R. R. 1981. Liberalizzazione e democratizzazione negli stati burocraticoautoritari. 11 Mulino 30. King, D. Y. 1981. Regime type and performance: Authoritarian rule, semi-capitalist development and rural inequality in Asia. Comparative Political Studies 13. Klieman, A. S. 1980. Confined to barracks: Emergencies and the military in developing societies. Comparative Politics 12. Lanning, E. 1974. A typology of Latin American political systems. Comparative Politics 6. Linz, Juan J. 1964. "An Authoritarian Regime: The Case of Spain." In Cleavages, ideologies and party systems, ed. Erik Allardt and Yrjo Littunen. Helsinki: Westermarck Society. . 1973. "Opposition in and under an Authoritarian Regime: The Case of Spain." In Regimes and oppositions, ed. Robert A. Dahl. New Haven, CT: Yale University Press. - . 1975. "Totalitarian and Authoritarian Regimes." In Handbook of Political Science, vol. 3: Macropolitical Theory, ed. Fred I. Greenstein and Nelson W. Polsby. Reading, MA.: Addison-Wesley. Luckman, A. R. 1971. A comparative typology of civil-military relations.

110

Authoritarianism

Government and Opposition 6. Malloy, James M. (ed.). 1977. Authoritarianism and corporatism in Latin America. Pittsburgh, PA: University of Pittsburgh Press. McKinlay, R. D., and A. S. Cohan. 1976a. Performance and instability in military and non-military regime systems. American Political Science Review 70. . 1976b. The economic performance of military regimes. British Journal of Political Science 6. Morlino, Leonardo. 1987. "Democratic Establishments: A Dimensional Analysis." In Comparing new democracies: Dilemmas of transition and consolidation in Mediterranean Europe and the Southern Cone, ed. E. Baloyra. Boulder, CO: Westview Press. Mosse, G. L. (ed.). 1979. International fascism. London: Sage. Mouzelis, N. 1986. On the rise of postwar military dictatorship: Argentina, Chile, Greece. Comparative Studies in Society and History 27. Nordlinger, Eric A. 1977. Soldiers in politics: The military coups and government. Englewood Cliffs, NJ: Prentice Hall. O'Donnell, Guillermo. 1973. Modernization and bureaucratic-authoritarianism: Studies in South American politics. Politics of Modernization Series No. 9. Berkeley: Institute of InternationallStudies, University of California. . 1981. Le tensioni dello stato burocratico-autoritario e il problema della democrazia. Politica Internazionale 2. , and Philippe C. Schmitter. 1986. "Political Life after Authoritarian Rule: Tentative Conclusions about Uncertain Transitions." In Transitions from authoritarian rule: prospects for democracy, 4 vols. Ed. G. O'Donnell, P. D. Schmitter, and Laurence Whitehead. Baltimore & London: Johns Hopkins University Press. Perlmutter, Amos. 1977. The military and politics in modern times. New Haven, CT: Yale University Press. . 1981. Modern authoritarianism. New Haven, CT: Yale University Press. , and V. Piave Bennett (eds.). 1980. The political influence of the military. New Haven, CT: Yale University Press. Philip, George. 1984. Military authoritarianism in South America: Brazil, Chile, Uruguay, Argentina. Political Studies 32: 1-20. Pike, F., and T. Stritch. (eds.). 1974. The new corporatism. Notre Dame, IN: University of Notre Dame Press. Ritter, A. R. M., and D. H. Pollock, (eds.). 1983. Latin American prospects for the 1980s. New York: Praeger. Rouquié, A. 1975. L'Hipothèse bonapartiste et l'émergence des systèmes politiques semicompctitifs. Revue Française de Science Politique 25. . 1982. L'État militaire en Amérique Latine. Paris: Editions de Seuil. Schmitter, Philippe C. 1985. "Le vie della democrazia nelle società modernizzanti." In I limiti della democrazia, ed. R. Scartezzini et al. Naples: Liguori. Stepan, A. 1978. The state and society: Peru in comparative perspective. Princeton, NJ: Princeton University Press. Thompson, W. R. 1973. The grievances of military coup-makers. Beverly Hills, CA: Sage.

S Commentary Zachary Irwin

Morlino's insightful discussion of the definition, typology, and dynamics of authoritarian regimes deserves special credit. In a discipline that seems at times to make a competitive sport of definitions, I find Morlino's discussion of Linz's definition to be an example of clear thinking, particularly because he is keenly aware that the concept of authoritarianism must retain "meaningfulness and explanatory power." Unfortunately, a well-reasoned definition and an examination of the consequences that flow from it are not guarantees that the most relevant questions will be raised about a state or group of states we deem "authoritarian." My criticism lies less with what the concept includes than with its utility f o r c o m p a r a t i v e a n a l y s i s . I s u s p e c t that m a n y s c h o l a r s w h o s t u d y authoritarian regimes would find Morlino's discussion of society's relations with a ruling coalition to be procrustean or simply inappropriate. While many authoritarian regimes share the characteristics he notes, I believe the objectives of comparativists will continue to be better served by comparing more limited spans of cultural or geographic traditions. Morlino recognizes the distinguishing "mentality/legitimating ideology," which, for example, leads us to expect different political outcomes in Chile, Poland, and Iran. But, a "distinctive mentality" conceals so great a source of political significance that it can hardly be considered the equivalent of other c o n c e p t s such as " l i m i t e d p l u r a l i s m . " C h i l e ' s r e l a t i v e l y " p u r e " authoritarianism under General Pinochet subordinated ideology to limiting pluralism with a vengeance. For Iran, religious (ideological) claims condition all other features of the regime, though the rule of the late Ayatollah Khomeini appears to have been hardly less dominating than that of General Pinochet. Both Pinochet and Jaruzelski of Poland may represent extremes of "bureaucratic-military rule," but such a definition leaves begging such important questions as the origins of regimes, their patterns of development, and prognosis for the future. Lingering behind the tremendous intellectual effort to discover enduring patterns of authoritarian rule is an expectation that the concept may provide 111

112

Authoritarianism

similar generalizations about (inter alia) modernization, legitimacy, and mobilization that one associates with the concepts of "democracy" or "totalitarianism." If this is a serious objective for "authoritarianism," however defined, I think it is doomed to frustration. Democratic regimes presuppose relatively comparable answers to modernization, legitimacy, and mobilization, but authoritarianism leaves these questions moot. Morlino comes close to acknowledging this problem by recognizing the "transitional" character of authoritarianism, which gives us some useful hypotheses about "hybrid" regimes. One may turn to some of the recent literature to demonstrate the vitality of region- or problem-specific approaches to authoritarian regimes. The example of "redemocratization" in Brazil is useful. Superficially, Brazil conforms to Morlino's "hybrid" model. A relatively weak array of parties existed before 1985 behind the reality of military rule, corporatism, clientism, and a central bureaucracy. According to the categories of authoritarian decay, the regime-inspired MDB (Movimento Democrdtico Brasileiro) should not have played the role that it did, which, as Mainwaring explains, "forced the military to change its strategy and ultimately undermined the military government" (1988, 97). The MDB defeated the governing party in 1974, was banned in 1979, and emerged victoriously as a new party in 1985. Authoritarianism, like Marxist analysis, is not especially receptive to acknowledging more than a formal role for political parties, but the question of redemocratization in Latin America, successful or not, is not very intelligible without a comparison of the relative position of parties with respect to patterns of authoritarian rule. Despite Brazil's tradition of relatively weak parties, the example of the MDB helped stimulate the formation of a Workers' party on the left. Similarly, weak parties have been a feature of Argentina's politics, and the absence of a left-wing party abandoned the interests of the country's working class to the Pcronists. Whether Argentina can avoid another cycle of military intervention depends less on Morlino's structures of authoritarianism than on the creation of representative institutions for which Peronism has had little interest. Finally, Pinochet's Chile displays all the features we would expect in a model of authoritarian decay—mass mobilization against the regime, division within the ruling coalition, and the breakdown of the capitalist economy. Remmer (1989) finds the resiliency of the dictatorship to lie in Pinochet's relations with the armed forces. Authoritarian generalizations not only have hardship accounting for Pinochet's survival but tell us nothing about a key question: What has been the impact of years of authoritarian rule on Chile's entrenched and polarized party system? Latin America raises some challenging questions for Morlino's analysis. Each case of authoritarian transition shows, in varying degrees, the "breakdown of the no longer limited pluralism, [an] increased mobilization

CommentarylZachary

Irwin

113

that the authoritarian regime can no longer control." For Morlino, the key to transformation lies in three conditions: the decay of the dominant coalition; the transformation of "undecided or indifferent groups" into active opponents; or the reactivization of an illegal opposition. I find it difficult to interpret the decay of Latin American authoritarianism without attention to a more discriminating analysis provided by the traditions of political parties in each state. Political parties are crucial to understanding the different prognoses for political life in Chile, Argentina, Brazil, and Uruguay. The emphasis on elite dynamics and composition leaves little means to account for the Brazilian abertura (opening) to democracy, or for the different social classes demanding universal suffrage and civilian rule in Chile. Instead of approaching Latin American authoritarianism from the viewpoint of the elite, it is more interesting to determine how authoritarian rule affects "party strength, electoral realignment and/or the rise of new partisan alternatives" (Remmer 1985, 269). The cases of redemocratization in Portugal and Spain present different examples, but ones no less discouraging for those who seek generalizations about authoritarian rule. In both cases, the style of transition—abrupt for Portugal and gradual for Spain—is basic to understanding the respective strength of the democratic left and right in each country. Authoritarian typologies and the character of the ruling coalition are of limited value. Compared with the examples of redemocratization I have mentioned, authoritarian regimes in the Middle East present an entirely different set of problems for the concept of authoritarianism. A recent study by Anderson focuses on the role of the Middle Eastern state independent of the armed forces. The work is of value especially in view of the abundant literature on authoritarian rule by the armed forces in the region. Anderson (1987, 8) finds the common feature of military rule to be superficial, for it conceals utterly different purposes in each state; e.g., Turkey's army as a "guardian of the nation," Libya's as a "vanguard of the revolution," and Syria's as a focus for "ethnic and sectarian struggle." Middle Eastern states may also exercise a level of autonomy from the society made possible by sources of external revenue. Here, the degree of popular support or acquiescence depends on the role of subsidies and the state's role in distribution. Finally, the state's history in promoting development, its origins in colonialism, and its administrative capability offer a more region-specific source of continuity and comparison than do the often turbulent factions that authoritarianism emphasizes. The level of foreign penetration and the place of nonstate actors further complicate what authoritarian rule in the Middle East can tell us about authoritarian rule elsewhere. Sub-Saharan Africa, at first, might prove to be a counterexample. Perhaps no single group of states better suggests the typologies and successive failures we associate with authoritarian rule as an ideal type. Nevertheless, Bratton (1989, 429) points out a significant qualification:

114

Authoritarianism

Because the state institutions are weak, they are less able to penetrate the society, and, as a result, "there is a relatively larger realm of unoccupied political space in Africa than anywhere else in the world." In suggesting that we turn our attention to "civic organizations" (e.g., churches, unions, and tribal associations), Bratton seeks to "treat the degree of conflict or congruence between state and civil objectives as an empirical variable rather than as a theoretical assumption." To this I might add that kinship and tribal traditions of rule add a distinguishing character to African politics, which, like civic organization, is apt to be considered incompatible with the generalizations about authoritarianism or neglected entirely. I do not think it impossible to develop theories of authoritarian rule but, rather, that the emphasis on authoritarian coalitions cannot express the problem of "conflict or congruence" in state society relations. Authoritarian typologies are unlikely to be very helpful in explaining levels of civil violence in Africa or in distinguishing mobilization in the capital city from penetration of the countryside. Finally, I wish to consider what is a more obvious and traditional limitation of authoritarianism in political science—its use in analyzing communist regimes. Marxism-Leninism, after Stalinism and Maoism, has endeavored to create a specific political culture that needs to be accommodated to typologies of authoritarianism. We might, as Huntington (1968) has done, consider communist rule as a more politically developed form of authoritarianism, less susceptible to "political decay." From the perspective of the present, however, the relevance of established authoritarian theory seems to confirm most of the warnings that area study scholars raised against comparativists two decades ago. Regardless of how unique "communist politics" may be, the rise of opposition to communism has taken forms that have little to do with Morlino's scenarios of confrontation, but reflect country-specific forms of adaptation. I am referring to what Scruton (1988) calls the "new right" in Central and Eastern Europe or the currents of Russian nationalism in the USSR. The possibility of a synthesis of the "new right" with communism has created a welter of participatory forms that are neither examples of "rcdcmocratization" nor of authoritarian decay. Again, a regional emphasis is appropriate. Can "authoritarianism" be saved? Possibly, but not without a conceptual breakthrough. Much of authoritarianism in political science is an artifact of ambition without the knowledge to accompany it, and greater empirical research may bring authoritarianism back to a level of relevance it once enjoyed.

References Anderson, Lisa. 1987. The state in the Middle East and North Africa. Comparative Politics 20(1) (October): 1-18.

Commentary/Zachary

Irwin

115

Bratton, Michael. 1989. Beyond the state: Civil society and associational life in Africa. World Politics 4\(3) (April):407-430. Huntington, S a m u e l P. 1968. Political

order in changing

societies.

N e w H a v e n , CT:

Yale University Press. Mainwaring, Scott. 1988. Political parties and democratization in Brazil and the Southern Cone. Comparative Politics 21(1) (October): 91-120. Remmer, Karen L. 1985. Redemocratization and the impact of authoritarian rule in Latin America. Comparative Politics 17(3) (April): 253-276. . 1989. Neopatrimonialism: The politics of military rule in Chile, 1973-1987. Comparative Politics 21(2) (January): 149-170. Scruton, Roger. 1988. The new right in Central Europe II: Poland and Hungary. Political

Studies

37:638-652.

6 On the Classification of Socialist Political Systems Boris N. Topornin Comparative study of political systems has already received full official recognition in the structure of modem political science. Historians of science recall the time when it was necessary to substantiate and prove both the theoretical and practical effectiveness of the comparative method and its usefulness in examining the main features of political development and the major institutions connected with the exercise of power in society and its management. Today, moreover, one can observe certain "growing pains" manifested in the obvious overestimation of the comparative method and, accordingly, in less attention paid to other scientific methods. Without attempting to replace the subject of this chapter by a discourse upon the comparative method, I still consider it useful to emphasize some points of departure. First, in Marxist-Leninist theory, the comparative method is regarded as one of the most important components of a dialectical materialist approach to the study of processes of social development. Comparison as such, however, has never been understood as an end in itself. It is not a mere description of various social phenomena. The meaning and purpose of comparison is to reveal and make clear the general features and regularities of the phenomena under consideration; to underscore their peculiarities; and to find out the causal relationships determining both similarities and differences. Second, Karl Marx, Friedrich Engels, and V. I. Lenin pointed out from the very start that the comparative method must be used only when studying comparable phenomena of a typical character. The efficiency of the comparative method will be low and conclusions inconcrete if it is employed in incidental combinations, regardless of occasional specific historical conditions. Third, the comparative method should not be seen as the only one and opposed to other methods. In political science, no less than in other social sciences, success depends on an appropriate approach (i.e., on the ability to combine all methods of research, both traditional and new ones—analysis, synthesis, generalization, etc.). In recent years in the socialist countries, many interesting works that represent comparative studies of political systems have appeared. Because 117

118

On Classification

of Socialist

Political

Systems

political systems in these countries are studied and taught, as a rule, in very close connection with research and courses in constitutional law, the majority of such works have been prepared by lawyers. This close connection between comparative politics and constitutional law in socialist states has a positive impact for the comparative study of political systems. Divorcing the theories of the state and law, which is peculiar to Western political science, diminishes the possibilities for analysis and limits the very understanding of a political system, its structural institutions, and their designation and forms of activity. It goes without saying that the origins of essential differences between the socialist and bourgeois political theories lie much deeper, in the very vision of social life, but the above fact, too, is symptomatic.

On the Concept of Political System The concept of political system as such is now firmly established in MarxistLeninist theory. It has not only been accepted as a scientific term, but it is also used in official documents. Significantly enough, Chapter 1 of the 1977 Constitution of the USSR is titled "The Political System." The new edition of the program of the Communist Party of the Soviet Union (CPSU), adopted by the Twenty-Seventh Congress, contains a special section on the development of the political system. The political system of Soviet society can be defined (though this is not the only definition) as a complicated, ramified but also dialectically interconnected and integral complex of state and public organizations and institutions, as well as a complex of norms and traditions linking them and determining their activity, through which political power is exercised and the management of the affairs of society as a whole is effected. The political system comprises relations among classes and social sections, nations and nationalities, society, the collective, and the individual. The components of the political system, along with the party, the state, public organizations, and work collectives, also include party directives, the legal system, the norms of public organizations, political traditions and customs, and morals and ethics of political life. Other researchers find it necessary to add to the concept of political system such components as political culture, and political regime. If the currently existing points of view are to be summed up, it can be said that a political system is understood as a complex of several subsystems: organizational; normative; communicative; and sociocultural (insofar as it is connected with political culture). Special attention is paid in socialist science to using to the utmost the potentialities of the system qualities, of the interaction and mutual enrichment of the subsystems within the framework of a single political system.

Boris N. Topornin

119

The concept of political system has limits. These do not imply, of course, the limits of development of the political system itself, appearance of new components in it, and improvement of the organization and activity of the operating ones. Something different is intended; namely, that a political system embraces not classes and social sections, but rather their organizations and institutions. Moreover, a political system does not include all organizations and institutions, but only those that participate in political life. In other words, it is necessary to draw a distinction between the notion of political system and such notions as political life, political superstructure, and so on. Finally, the role and importance of the political system in the life of a socialist society continuously expands. This is because, above all, the present stage of development requires a high level of organization of society, puiposefulness of its efforts, and mobilization and effective use of all its resources. Special attention is given to the human factor. A socialist society cannot function effectively without seeking new ways to develop the creative activity of the masses in all spheres of social life. The larger the scale of historical objectives, the more important becomes the dedicatedly interested, responsible, and active participation of the masses of people in their attainment.

The General and the Particular in the Progress of Political Systems The key to understanding the nature of political systems under socialism, and to their classification, is provided by a correct perception of the correlation between the basis and superstructure of the system. Marxist-Leninist theory has long proved and demonstrated that a political system is part of the superstructure of society and, consequently, its class character, main functions, and trends of development are predetermined by the basic factors. A socialist basis presupposes a socialist political system. This means, first of all, that there is a concentration of power in society in the hands of the working people. The consolidation of socialist relations of production and the development of productive forces are the main prerequisites for the development and improvement of the socialist political system. It is important to mention that Marxism-Leninism does not imply that the political system automatically and passively follows from changes in the means of production. The political system is an active, effective instrument, a factor in social and economic transformations. Historical experience shows that the political system can assist in satisfying pressing social needs and can add vigor to the socialist nature of the basis. But this is possible only through a scientific approach that considers the real potentialities of the political system. Miscalculations and errors in the appraisal of a political system, the

120

On Classification

of Socialist

Political

Systems

hope that "power can do everything," only obstruct the progress of socialist society. All political systems under socialism have in common a socialist economic system and public ownership of the means of production. In this respect, there are no differences within this group of political systems. The prevalence in society of socialist property, having as a rule two main forms—state property (belonging to all the people) and cooperative (group) property—determines the socialist character of the working class and the class of peasants, with the leading role played by the working class. It should not be ignored, however, that the economies of some socialist countries are based on small-scale commodity production. In Poland, for example, the individual sector accounts for more than half of agricultural production. Individual peasant farms and various artisan workshops exist in other countries as well. This, naturally, leaves its imprint on the development of political life. Nevertheless, even in countries with persisting multiplicity of economic forms, the most important factors to note are that the socialist system is the dominating one and that the state sector is the leading sector in the entire national economy. The social base of the political system under socialism is the alliance of the working people, with the working class at the head. The nature of Soviet society as a society of working people is especially underlined in the USSR's 1977 Constitution. At the present stage, the convergence of the working class, the collective-farm peasantry, and the intelligentsia; and the establishment of a classless structure of society, with the working class playing the decisive role in that process, constitute a law of development of social relations in the USSR. As is stated in the new edition of the program of the CPSU, overcoming differences between the classes and establishing a classless society in the USSR will take place primarily in the historical framework of the initial, socialist phase of the formation of communism. Since socialist countries vary in their stages of social development, they also have significant differences in their social-class structures. These differences can be noted in the relationship of classes and social strata to the existence, in particular, of larger or smaller groups of peasantry, artisans, and handicraftsmen outside the social sector. Among socialist states, there are variations in the consciousness of individual groups of the population, as for example, in their attitudes toward religion. Nevertheless, of the greatest and most decisive importance is the fact that the uniformity of relations of production and the growing power of Marxist-Leninist ideology determine the general tendency toward society's increasing class unity. Historical experience teaches us that each major stage in social progress confronts socialist countries with new complex tasks and problems. Their solution involves struggle and searches and the overcoming of contradictions and hardships. The progress of socialism presupposes the accumulation, generalization, and concretion of scientific ideas concerning the most diverse

Boris N. Topornin

121

aspects of social life. The objective laws of social development, and the interconnection and interdependence among economic, social, and ideological factors, on the one hand, and political ones, on the other, must be thoroughly explored. The evolution of a socialist political system also has its own regularities. These methods for building socialism have been tried and tested internationally and applied to a large group of countries with very diverse conditions. The dynamism of these regularities can be clearly seen in two circumstances: First, socialism has been adopted, not only by the USSR, but by many Latin American, other European, and Asian countries. Socialism, in short, has become a global system. Second, the successes of socialism—in economics, politics, social development, and culture—have been achieved within extremely brief historical periods and in trying historical circumstances, both internally and internationally. What are, then, the natural features and characteristics inherent in the political system of socialist society? If we sum up the results of social development to the present time, it can be said that in all countries that have adopted scientific socialism, the political system has the following characteristics: • Power in society belongs to the working people, rallied around the working class. • With the victory of the socialist revolution, power assumes the form of the dictatorship of the proletariat. It relies on the support of the broad masses of working people and on the alliance of the working class with the peasantry. With the achievement of developed socialism, the dictatorship of the proletariat grows into a political organization of all the working people. • The leading and guiding part in society and its political system is played by the Communist party. Through creative application and development of Marxism-Leninism, the Communist party defines the guidelines of society's progress and provides scientific guidance to the creative efforts of the working masses and to all elements of the political system. • Socialist democracy is being continually extended and deepened. Citizens are ever more actively drawn into the management of state and public affairs. Ever increasing scope is given to the elements of socialist self-government. • The organizational forms and methods of democracy are being improved. The role of elected representative bodies of authority is being enhanced; the independence of public organizations is being extended; the prestige of democratic forms of management of production—work collectives, first of all—is growing; and under the direction of the Communist party, the interaction of various elements

122

On Classification

of Socialist

Political

Systems

of the political system is gaining strength. • On the basis of Marxist-Leninist ideology, the level of political culture of the broadest masses of people is rising. They demonstrate ever greater interest in the affairs of society, the state, the territory, and the collective. The experience and traditions of citizens' participation in the management of state and public affairs become more deeply rooted. • The equality and friendship of all nations and nationalities become more firmly established. Socialist internationalism determines the inner progress of society and understanding of international problems • The defense of revolutionary gains from encroachments of class enemies is the center of attention of society. The use of general laws and regularities in specific conditions of each of the socialist countries is the foundation of successful development of the political system of society. Only in this way is it possible to solve the increasingly complicated and crucial problems of social progress, to surmount the difficulties of growth, and to resolve in timely fashion the contradictions that arise. Undoubtedly, this requires taking into consideration the specific historical conditions of one country or another, such as the stage of socialist construction attained, the specific features of the country's economic life, social-class relations, and the level of ideological and intellectual development. Experience has confirmed the worldwide historical importance of the Soviet experience, the experience of the country that has blazed the trail in socialism and communism. The universality of basic characteristics of the Soviet example is also stressed by the fraternal socialist countries. But simultaneously, as V. I. Lenin foresaw, the world revolutionary process is distinguished by a diversity of forms and methods of establishment and exercise of political power. Even prior to the establishment of the Soviet government, V. I. Lenin, (69-70) wrote: All nations will arrive at socialism—this is inevitable, but all will do so in not exactly the same way, each will contribute something of its own to some form of democracy, to some variety of the dictatorship of the proletariat, to the varying rate of socialist transformations, in the different aspects of social life.

As a result of the Great October Socialist Revolution, a new government—the dictatorship of the proletariat—first appeared as a republic of soviets. In other countries, the form of government following revolution was somewhat different. The second half of the 1940s saw the rise of the "people's democracy" in a number of European and Asian countries. Their systems reflected the distinctive characteristics of the countries and the time.

Boris N. Topornin

123

It will be recalled that in many of the countries that had a people's democracy (excluding Bulgaria and Yugoslavia), the revolution passed through two developmental stages: (1) the general democratic and antiimperialist stage and, in a number of cases, an antifeudal stage; and (2) a socialist stage. The nature of government also varied according to the stage. The first stage represented a broad-based class alliance, which included in some countries the national bourgeoisie. Later, when the revolution became socialist in all respects, the dictatorship of the proletariat was established. In other words, the embryonic characters of the new government (i.e., the hegemony of the working class in close alliance with the working peasantry) ensured, with the evolution of the revolutionary process, its growth into the dictatorship of the proletariat. The revolution and, consequently, both the essence and forms of government had their own idiosyncracies in the Asian countries of people's democracy and in Cuba. The people's democracy is a concept that is accepted not only in the literature of political science, but also in constitutional legislation. Thus, Article 2 of the Constitution of the Mongolian People's Republic states: "The Mongolian 'People's Republic' is a socialist state existing and developing in the form of a 'People's Democracy.' " The concept of people's democracy was sufficiently broad; comprising countries that were developing under different conditions and at different historical stages. It was used for a long time to single out a distinct new form of political system that was different from the Soviet system. What were the features and institutions that characterized a people's democracy? There are many noteworthy views on this subject in the scientific literature in the USSR and in other socialist countries. Overall, however, it must be recognized that a people's democracy was not substantially and radically different from the Soviet political system; the two shared many similarities. When describing a people's democracy, specialists, as a rule, gave special attention to the specific structure of the political system. They noted the existence of such all-embracing movements as the national front; the multiparty system that occurred in many countries; and the more broadbased, "bloc" policy of people's government that characterized the early period of the system. At the first stage of growth of a people's democracy, some of the state institutions and legislative acts, including constitutions, inherited from the past remained in force. These were essentially bourgeois democratic elements, and, until the adoption of new constitutions and a complete restructuring of the political system, these institutions and legal acts could be used in the interests of promoting the revolutionary process. In comparing the Soviet and people's democracy varieties of socialist political systems, some political scientists tried to argue that one or the other form was better. That was, at best, a waste of time. The people's democracy and the Soviet form have been equally expedient forms of political power and

124

On Classification

of Socialist

Political

Systems

of managing social affairs. The choice of form was dictated by historical necessity. Subsequently, both the Soviet political system and people's democracy have undergone quite important changes. The changes were necessitated by new social conditions and needs, and by social progress. Political institutions and other components of the political system that have outlived their usefulness have gradually been dropped. New, more perfect forms and methods of regulating political relations have appeared. Thus, the specificity called into being by temporary, transient circumstances has ceased to exist. These changes do not imply that diversity has diminished; in a number of cases, differentiation has increased. This is not accidental. Socialist countries sometimes follow their own way after starting almost simultaneously to grapple with new problems, seeking solutions that have no approved patterns. The experience of world socialism has not confirmed forecasts that, with the convergence of economic, political, and ideological foundations of society, unification of political systems would gradually occur. The fact is that there are no rigid couplings among these phenomena. Socialist countries today are converging because this is a social necessity. Whereas in the capitalist world, the law of uneven economic, sociopolitical, and cultural development operates, socialism creates conditions for raising less developed countries to the level of the advanced. The higher and nearer the levels of social development of socialist countries, the faster and more profound their cooperation, and the better organized is the process in which this occurs. Against this background, however, the process of seeking optimal forms and methods to organize modern society, far from decreasing, is, on the contrary, expanding. Activation of scientific thought, simultaneously in each of the socialist countries, serves the common interests of world socialism. Another point also needs to be stressed: The experience of the past few decades has shown that the former division of political systems of socialist society into two types—the Soviet system and the system of people's democracy—has lost its former foundation and retains primarily historical significance. There is every reason to speak today about the diversity of concrete forms of political power under socialism, which are manifesting themselves and developing on the basis of common laws and principles. This is the diversity of both state political structures (forms of government, forms of national-state organization) and nonstate political structures (multiparty and one-party systems, national fronts, etc.). Mention should be made of yet another circumstance. Comparative study of political systems in the world today shows that the mobility and dynamism of development of political systems are noticeably higher and more intensive under socialism than under capitalism. There is no "rigid" modeling of political institutions under socialism, such as that peculiar to many capitalist countries. It is significant that there has been a noticeable growth in the USSR, in the past fifteen years, in the number of public organizations. In

Boris N. Topornin

125

addition to the trade unions, Komsomol, cooperative societies, and other traditional associations, there have been created literary clubs, societies for the protection of monuments, and new societies for friendship with the peoples of foreign countries. The establishment of a temperance league has just been completed, and two other societies—of women and pensioners—will soon make their appearance. Changes in state organization occur not only in the center but also locally. Whereas the concepts "form of government" and "political regime," used by bourgeois political science, in most cases proceed from the structure and functions of the central authority, and discussions concentrate, as a rule, around the questions of classifying one country or another as a presidential or parliamentary republic, a federation or a unitary state, socialist political thought always centers its attention on the problems of improving the entire state mechanism. Under socialism, a wealth of transitions and interconnected solutions can be observed.

The State in the Political System Before considering the varieties of state forms, it will be useful to make a few preliminary remarks. In the present conditions, the state in socialist countries appears as the main and most important instrument of social transformation. It is through the state that the working people, headed by the Communist party, regulate social relations and ensure the accomplishment of the most complicated and enormous tasks connected with building and improving socialism. This conclusion fully takes into account the character and specificity of the state, as compared with the other elements of the political system. The experience of world socialism shows that "stateless socialism" has not practically existed and will apparently never be seen. It does not seem possible to single out any countries according to this criterion. "Unconventional" views, amounting sometimes to the demand for premature rejection of the state and artificial curtailment of its functions, though they did influence practice to some extent, have not brought concrete alterations in socialism's general tendencies. The way to the withering away of the state, of which the founders of scientific socialism wrote, lies not through the negation and disregard of its inherent potentialities, but through their consistent discovery and utilization. The state is an objcct of primary attention on the part of the Communist parties of the socialist countries. Emphasizing that this is a matter of key importance in its policy, the CPSU especially points out in the new edition of its program that to develop the Soviet socialist state means, first of all, increasingly to reveal its democratic nature as a state of the whole people. The supreme goal of the Soviet state is unalterable—the building of

126

On Classification

of Socialist

Political

Systems

communism in the USSR, of a highly organized society of free and conscious working people in which public self-government will be established. The attitude of the CPSU and the other Communist parties of the socialist countries to the ideas of self-government should be mentioned separately. The fact is that futile attempts have been made in Western political science to divide socialist political systems according to their being "for" or "against" self-government. Moreover, both theoretical views and practices of the ruling Communist parties were being misrepresented. Some bourgeois Sovietologists tried to prove that the CPSU and many other parties allegedly negate self-government (i.e., they have departed from the ideas of the classics of Marxism-Leninism). The reality is, of course, different. At the very initial stage of socialist construction, the Marxist-Leninist idea of self-government by the working people was not only accepted but was also developed in the documents of the Communist party of Soviet Russia. Taking into consideration its concrete historical conditions, the party placed special emphasis, in the 1920s, on the promotion of local and regional self-government, which was then especially necessary. Simultaneously, the party emphasized that self-government was of paramount importance in socialist construction and attached, and would continue to attach, increasing significance to its every possible development. It is easy to see that nowadays the idea of self-government is being carried into effect in considerably more favorable circumstances than it was in the first decade of Soviet power. Self-government is virtually unrestricted and embraces practically all levels and all spheres of social development, such as politics, economics, culture, and social progress. The new edition of the CPSU program states: The CPSU b e l i e v e s that at the present stage, the strategic line of development of the political system of Soviet society consists in advancing Soviet democracy and increasingly promoting socialist self-government by the people on the basis of day-to-day, active, and effective participation of working people, their collectives and organizations, in decisionmaking concerning the affairs of state and society.

The concept of socialist self-government will be enriched and developed on the basis of the socialist countries' collective experiences. These experiences will no doubt introduce much that is new and interesting in both theory and practice. Even now it can be said that self-government, above all, means that the work of administration must not only be carried out in the interests of working people but also must become naturally, step by step, a direct concern of the working people themselves, who, to use Lenin's words, "know no authority except the authority of their own unity." The main elements of socialist self-government by the people are the representative bodies of state authority: for example, the soviets of people's

Boris N. Topornin

127

deputies in the USSR; the National Assembly and the people's councils in Bulgaria; and the Federal Assembly, the republican national councils and the national committees in the Czechoslovak Socialist Republic. Although, as a rule, the highest representative institutions differ in name from the local institutions, this does not imply a recognition of a gap between them, as is so typical of the bourgeois state system. The principle of unity of the representative system has become firmly established everywhere in socialist countries. This helps to take into account the interests of both society as a whole and its component parts in working out and carrying through state policies. The general tendency of the socialist countries' development is a growing role for popular representation. This applies to all institutions from the highest to the local levels. At the same time, special features and new tendencies in their development have occurred. The highest representative institutions, along with improving legislation, have begun devoting more attention to key problems of the country's home and foreign policy and have increased control over the work of the state organs. Local representative institutions focus their attention on matters pertaining to the complex development of the entire territory, independent decisionmaking concerning affairs of local importance, and control over all enterprises and institutions—including those of central subordination—situated on their territory. Differences in the organization and activity of the representative bodies can be observed in the length of the sessions held, the procedure for preparation and consideration of legislative and other legal acts (the number of readings, for example), the system of permanent and temporary committees and commissions, and the forms of the deputies' activity in the constituencies. Socialist countries have recently been paying much attention to improving the representative institutions. In the perfection of electoral systems, attention is focused on giving working people more influence over the selection of candidates and control over their work. In some countries, measures have been taken to enable the free and thorough discussion of the personal and other qualities of the deputies and, thus, to ensure that only the worthiest and most respectable persons are elected to the representative bodies. In other countries, competitiveness among candidates is encouraged so that there are always more candidates nominated in the constituencies than there are parliamentary seats. Along with candidates who are elected by individual constituencies, some countries have the so-called national or territorial lists of candidates who are common for all election districts. On the whole, socialist countries are characterized by the institution of the collective head of state (e.g., the Presidium of the Supreme Soviet of the USSR, the Council of State of the German Democratic Republic). At the same time, a number of countries have introduced the office of president.

128

On Classification

of Socialist

Political

Systems

Such offices exist today in Vietnam, the Korean People's Democratic Republic, Rumania, and Czechoslovakia. According to the Constitution of Cuba, the president of the Council of State is simultaneously the head of government. A peculiarity of state organization in Yugoslavia is the structure of the republic's skupstinas (assemblies). The skupstinas have three chambers made up of delegates from basic work collectives, territorial communities, and social organizations, and are formed on the basis of a delegation system. Certain specificity of state organization in the Polish People's Republic is revealed in the organization of its local executive bodies. Although the latest law on the people's councils has increased the latters' influence on the local administrative bodies, nevertheless, compared with other socialist countries, there remains in Poland a certain detachment of the administrative apparatus, which is based on the principle of personal leadership. As we see, state organization rests, in principle, on similar foundations in all socialist countries. The existing differences cannot be characterized as distinctive types or varieties. Nevertheless, the diversity of concrete structural and procedural solutions underlines the fact that general laws manifest themselves and operate so as to take careful account of, and reflect as fully as possible, the precise social conditions and needs.

Political Parties and Public Organizations When considering political systems from the perspective of the status and role of parties and public organizations, it is necessary to bear in mind that, in all socialist countries, the leading role belongs to the Communist party; whether we are speaking of the CPSU, the Hungarian Socialist Workers' party, the Socialist Unity party of the German Democratic Republic, the Communist party of Cuba, or the Communist party of Czechoslovakia. The different names, which reflect the peculiarities of the historical process, do not alter the essence of the party, which, in each case, is a Marxist-Leninist party and constitutes the nucleus of society's political system. All other parts of the political system—the state, the trade unions, youth leagues, women's associations, cooperative and other public organizations, and work collectives—work under its leadership. The Communist party guides and coordinates their activities and sees to it that each element of the political system fulfils its functions completely. The experience of world socialism shows that the leading role of the party is steadily growing. This is due to many reasons: the growing scale and complexity of the tasks of economic, social, political, and cultural advance of society; the profound need to comprehend practical phenomena further and to enrich Marxist-Leninist theory; and foreign policy factors. It should especially be mentioned it is extremely important to pool and direct efforts at this time,

Boris N. Topornin

129

when the independence of all elements of the political system are extending and increasing their activity. Many of the socialist countries have a one-party system. Communist parties alone f u n c t i o n in the U S S R , Hungary, M o n g o l i a , R u m a n i a , Yugoslavia, Cuba, and Vietnam. Political systems in a number of other countries, however, contain many parties. In the People's Republic of Bulgaria, in addition to the Bulgarian Communist party, there is the Bulgarian Agrarian People's Union. In the German Democratic Republic, besides the Socialist Unity party, there are four others: the National Democratic party, the Liberal Democratic party, the Democratic Peasants' party, and the Christian Democratic Union. The same number of noncommunist parties are active in Czechoslovakia: two in Czechia (Socialist and People's parties) and two in Slovakia (the Slovak Freedom and the Slovak Reconstruction parties). History has thus divided political systems into one-party and multiparty systems. A socialist multiparty system, however, does not diminish the leading role of the Communist party. We can only say that the multiparty system requires devising additional political mechanisms and procedures to ensure coordinated action among the parties and their joint work within the d i f f e r e n t s e c t o r s of social l i f e . T h e u s e f u l n e s s of s u c h a d d i t i o n a l sociopolitical components is clear from the example of the Polish People's Republic, wherein the Polish United Workers' party forms a close alliance with the United Peasants' party and the Democratic party. The multiparty system in socialist countries owes its existence to the operation of a number of social factors. It must be borne in mind, for instance, that the social-class structure of society comprises different classes and social sections. There are, as well, peculiarities in the world outlook of separate groups of the population. Multiparty traditions in some countries also have a role to play. Of special significance, however, is the position of the parties themselves, particularly their attitude to revolutionary gains and to socialist construction. Other things being equal, party's participation in the building of the new society under the leadership of the Communist party acquires decisive importance. Noncommunist parties take part in the work of the state and in the entire political life of society. They have their deputies in the highest and local representative institutions and are represented in the government and other state bodies. In a number of countries, the status of noncommunist parties in the political system is confirmed in the constitution (for example, see Article I, Paragraph 3 of the Constitution of Bulgaria; Article 3, Paragraph 2 of the Constitution of Poland). The Communist parties attach much importance to their cooperation with other political parties. Naturally, the multiparty pattern in socialist countries develops in accordance with specific historical conditions. The number of political parties varies according to the differences among the functions of individual parties and their program components. Nevertheless, my general conclusion is that

130

On Classification

of Socialist

Political

Systems

the multiparty system, wherever it exists, has stabilized, and society's advance along the road of socialist construction does not imperil its future. It has also been proved that in countries with only one party, there are no prerequisites for a return for the multiparty system or to the creation of new parties. The political system of most of the socialist countries incoiporates many universal public organizations and movements, such as a people's or national front. There is no such organization in the USSR, where the bloc of Communists with nonparty people, resulting from political practices, has not acquired legal organizational foims. A people's or national front embraces, as a rule, all public organizations and political parties that exist in the given country, under the leadership of the Communist party. The fronts actually represent an effective form for attracting to socialist construction the broadest sections of the population, both united in public organizations and political parties and remaining outside them. National fronts can, with good reason, be called an organizational form of mass people's power. The subdivision of political systems in socialist countries, according to the presence or absence of a national front, must be developed and specified further. On the one hand, national fronts do not all have the same character and structure. In some cases, the front is a mass political organization as is, for example, the Socialist Alliance of the Working People of Yugoslavia. In others, the front is a sociopolitical movement (e.g., the Patriotic People's Front in Hungary). The Fatherland Front in Bulgaria combines the features of a sociopolitical organization and a mass movement. In Cuba, the Committees in the Defense of the Revolution function much like a national front.

Political System and the Development of National Relations In most socialist countries, the population is nationally homogeneous. In a number of countries, however, different nations and nationalities live together. According to the latest ccnsus, the USSR is inhabited by more than 1 hundred nations and nationalities. Yugoslavia is a multinational country as well, and the population of Czechoslovakia consists of two nations, Czechs and Slovaks, and some national minorities. Distinct peoples and national minorities also live in other countries, such as Romania and China. The multinational character of society finds its reflection in its political system. This applies not only to the organizational structure, but also to the policy pursued in society. The Communist parties give constant attention to the development of national relations and to their improvement. In the USSR, the nationalities question inherited from the past has been successfully solved, and the CPSU stresses the necessity of solving new problems arising in the process of perfecting socialism. In other words, an end has been put,

Boris N. Topornin

131

not only to the de jure, but also to de facto inequality of nations and nationalities in the USSR. The progress of national relations is characterized, on the one hand, by the further flourishing of the nations and, on the other, by their steady convergence. State organization in the USSR, the Czechoslovak Socialist Republic, and the Socialist Federal Republic of Yugoslavia is based on the principles of socialist federation. All the federations are formed on a national-territorial basis. Their exact composition, however, is not identical and is determined by specific conditions. The USSR has fifteen union republics; the Socialist Federal Republic of Yugoslavia, six; and the Czechoslovak Socialist Republic, two. Socialist federation symbolizes the political unity of a multinational society. It serves to take account of, and combine the interests of, nations and peoples in a common state policy. Political unity is understood and realized in different countries on differing scales and in differing forms, but everywhere it embraces all the main spheres of social life, including the economy. The interests of improvement of national relations require further consolidation of the unity of a federal state and enrichment of the forms of relations among the nations. The CPSU has declared in the new edition of its program that it will continue strengthening the USSR, consistently combat any manifestations of parochialism and national narrow-mindedness, simultaneously make efforts to enhance further the role of the republics, autonomous regions, and autonomous areas in accomplishing the common tasks of all nations, and ensure active participation of working people of all nationalities in the work of the bodies of authority and administration. The importance of observing the Leninist principles of socialist federation is stressed in Yugoslavia and Czechoslovakia. Federal organization finds its reflection in the structure of the state organs. Thus, according to the 1977 Constitution, the Supreme Soviet of the USSR consists of two chambers: the Soviet of the Union and the Soviet of Nationalities; the former is to represent common interests of the entire people, and the latter is to give more consideration to specific national interests. The union republics are represented in the federal bodies: the Presidium of the Supreme Soviet, the Council of Ministers, and the Supreme Court. The peculiarity of the Czechoslovak federation consists in a rule having been introduced in the Federal Assembly to prevent "majorization" on the part of one of the nations. The Czech and Slovak republics are represented on an equal footing in the Presidium of the Federal Assembly. If a citizen of the Czech Socialist Republic is elected chair of the assembly, then a citizen of the Slovak Socialist Republic is elected first deputy. Representatives of the republics and autonomous provinces in the Socialist Federal Republic of Yugoslavia are included in the Presidium the Federal Executive Vece (Council), and other bodies. The republics forming

132

On Classification

of Socialist

Political

Systems

the federation represent states in the full sense of the concept. The multinational composition of the population is reflected not only in the state organization; the structure of many public organizations is based on the principles of a national-state system.

Conclusion The experience of world socialism teaches us that political systems under socialism are of the same type. They are characterized by c o m m o n f o u n d a t i o n s , m a i n s t r u c t u r e s , and t e n d e n c i e s of g r o w t h . But t h e s e commonalities, far from excluding diversity, encourage diversity of specific forms and methods of exercising political power and developing socialist democracy. A concrete correlation between the general and the particular is not something that is programmed permanently. It is the result of a creative approach of the Communist parties and society as a whole to the problems of improving the mechanism of sovereignty of the people. The way followed by the socialist countries is predetermined by general laws. It is impossible to ensure social progress without considering these laws and without taking into consideration the concrete historical specificity of each country.

Reference V.I. Lenin. 1 9 6 0 - 1 9 7 0 . (45 vols.). Collected Languages Publishing House.

works,

vol. 23. Moscow: Foreign

H Commentary Huang

Zhong-Liang

In the 1980s, political system reform spread rapidly among the socialist countries. As the prelude to, and theoretical prerequisite for, these reforms, the concept "political system" was widely applied by scholars, and there was considerable theoretical study of the political system. Topomin's study of the classification of political systems under socialism is an example of that research, and it gives us considerable insight on this subject. Today, in the upsurge of reform, no theoretical area of study has more practical meaning than do attempts to classify and typologize political systems. Topomin points out: "The meaning and purpose of comparison is to reveal and make clear the general features and regularities of the phenomena under consideration; to underscore their peculiarities; and to find out the causal relationships determining both similarities and differences." The comparative study of political systems will help us to summarize the objective rules and positive and negative experiences of the development of the socialist political system, and, therefore, it will be helpful to the development of political system reform. In summarizing the experiences of socialist practice, we certainly should avoid overstressing peculiarities and overlooking the common rules and experiences of the different paths of socialist development. Simultaneously, we should avoid the tendency to overstress similarities and ignore differences. We have a similar problem with respect to classification. Overstressing the peculiarities and overlooking the similarities would necessarily create obstacles to generalization about the common rules that affect every political system. It also would prevent countries from drawing lessons from each other and improving upon the experiences of others. Overlooking the peculiarities of different political systems, overstressing the general features and regularities, or exaggerating a certain model or peculiar experiences of one country in order to create a general and common rule would ignore a country's reality, would indiscriminately, inappropriately, and mechanically apply the experience of others. Our socialism has suffered considerably from such blunders already. 133

134

On Classification

of Socialist

Political

Systems

One of the urgent reasons for many socialist countries to reform today is that t h e y still regard certain p o l i t i c a l s y s t e m s , w h i c h w e r e f o r m e d in a particular historical time, as the sole model that every socialist country must apply in every development period. It is the task o f the present reforms to break through these rigid and outdated ideas, to combine Marxism with the particular country's reality, and to shape a vigorous political system that w i l l be in harmony with one's o w n country, that will be welcomed by the people. T o d a y , the cause o f socialist r e f o r m s has an o b v i o u s characteristic. S o c i a l i s t countries are inspiring e a c h other, i n f l u e n c i n g e a c h other, and learning from each other. This illustrates the potential for vigorous advance and the b r i g h t future o f o u r s o c i a l i s t c a u s e . L e a d e r s o f m a n y s o c i a l i s t countries, i n c l u d i n g the great p o w e r s , pay c o n s i d e r a b l e attention to the research and experiences of the other countries. In such a situation, theorists need to study further the e x p e r i e n c e s and the concrete c i r c u m s t a n c e s o f particular countries. A s far as Topornin's classification is concerned, what w e should first p r e v e n t is the t e n d e n c y to o v e r l o o k the p e c u l i a r i t i e s o f e a c h c o u n t r y . Topornin states that socialist political systems are characterized by common foundations, main structures, and tendencies of growth. But these commonalities, far from excluding diversity, encourage diversity of specific forms and methods of exercising political power and developing socialist democracy. . . . The way followed by the socialist countries is predetermined by general laws. It is impossible to ensure social progress without considering these laws and without taking into consideration the concrete historical specificity of each country.

This shows that Topornin has considered both aspects, but it is obvious that the basic t e n d e n c y o f his chapter stresses the generality o f socialist systems, not their specifications and peculiarities. W e first need to define the concept "socialist political system." There are numerous definitions of a socialist political system, but none is universally accepted. T h e main difference among the various definitions lies in that some have a broader application while others have a narrow one. S o m e identify socialist political systems with political organization. Some define the main component of the political system as including political organization, political relations, political norms, and political culture. It is not sufficient for classification to g i v e socialist political systems a general definition. It is important to identify the most necessary components and the main characteristics of the system. It is m y opinion that the most n e c e s s a r y c o m p o n e n t s o f a s o c i a l i s t p o l i t i c a l s y s t e m are p o l i t i c a l o r g a n i z a t i o n s and the i n t e r r e l a t i o n s h i p s a m o n g t h e m . T h i s is m a i n l y characterized b y the fact that political p o w e r is exercised, and the public affairs of the society as a whole are managed, by the laboring people, led by

Commentary/Huang

Zhong-Liang

135

the working class, through various political devices mainly in the form of state power. Classification becomes easier when we grasp the main components and the main characteristic of the political system, rather than its subordinate, unessential parts. We should focus on the study of the role, impact, and interactions of the political organizations of various socialist countries in exercising political power and management of society as a whole. Through this approach we can judge the generalities and peculiarities of the different socialist political systems. The common features of socialist political systems are, of course, its essential aspects (i.e., "the natural features and characteristics inherent in the political system of socialist society"). Topornin summarized these common features thoroughly and correctly: "All political systems under socialism have in common a socialist system of economy and public ownership of the means of production." This is the foundation of all other generalities of the socialist political system. Topornin enumerates these common features: Power in society belongs to the working people, rallied around the working class. . . . [P]ower assumes the form of the dictatorship of the proletariat [relying] on the support of the broad masses of working people and on the alliance of the working class with the peasantry. . . . The leading and guiding part in society and its political system is played by the Communist party.... Socialist democracy is being continually extended and deepened;... The organizational forms and methods of democracy are being improved.... On the basis of Marxist-Leninist ideology, the level of political culture of the broadest masses of people is rising.

It is obvious that each country's socialist political systems are of the same type, differing from capitalist political systems and becoming the most advantageous type of political system in human history. The proposition of the classification of a socialist political system itself is not complete if we conclude that socialist political systems do not have different subtypes. Such a conclusion is not in conformity with reality. Because of the differences in each country's historical conditions, national traditions and norms, level of socialist construction, and degree of economic and cultural development, there exist differences in the role, impact, and interactions among various social and political organizations in exercising political power and management of social affairs. These create the differences in the form, method, and degree of socialist democracy. I suggest that socialist political systems can be classified into the following three subtypes: (1) the Soviet democratic political systems of the USSR; (2) the People's democratic political systems of China and Eastern European countries; (3) the Self-governing democratic political system of Yugoslavia. Should the Soviet democratic political system and people's demopolitical

136

On Classification

of Socialist

Political

Systems

system be so differentiated? According to Topornin, both the Soviet political system and the people's democracy have undergone quite important changes, . . . [T]he specificity called into being by temporary, transient circumstances has ceased to exist. . . . [T]he former division of political systems of socialist society into two types—the Soviet system and the system of p e o p l e ' s d e m o c r a c y — h a s l o s t its former foundation and retains primarily a historical significance.

Topornin argues that "a people's democracy was not substantially and radically different from the Soviet political system. The two systems shared many similarities." There is little question that the two political systems are substantially and basically identical; both belong to the socialist political system. But, we must add that they belong to two different subtypes of the socialist political system, and the people's democratic political system and Soviet democratic political system differ in two critical areas. First, in many countries, such as China, Poland, East Germany, Bulgaria, and Czechoslovakia, there are democratic parties other than the Communist party, and they practice a system of multiparty cooperation under the leadership of the Communist party. It is also worth noting that these democratic parties play a very important role in their country's politics. Many democratic parties or groups in these countries have a certain portion of seats in the parliament, and their leaders often assume the position of president or vice-president of the parliament. They also occupy some positions in the government, and their leaders generally serve as vice-premier or minister. Some democratic parties in these countries (e.g., the United Farmers' and Democratic Parties of Poland and the Farmers' Union of Bulgaria) are, in fact, also the parties in power. The Polish United Workers' party proposed that the "united government" be composed of three parties; the constitution of Bulgaria provides that the Communist party lead the country's socialist construction, in fraternal cooperation with the Farmers' Union. Although both the one-party system and the multiparty system belong to the socialist party system, their differences and their impact on the political system cannot be neglected. Second, there exist institutions such as the Chinese People's Political Consultative Conference and a people's front in many Eastern European countries. These institutions are the largest and the most representative political organizations in many of these countries. They are important political forms, through which the masses can participate in the management of affairs of the state and society and exercise their democratic rights. The front in some countries has the power to prepare for and organize the election of the representative body, has the right to sponsor acts in the parliament, and has the right to suggest that a national discussion and referendum be held on

Commentary/Huang

Zhong-Liang

137

certain matters. These organizations, under the rubric of front or political consultative conference, play a great role in the political life of the nation, and they become a peculiar feature of these countries, compared with the Soviet democratic political system. Of course, since their founding in the democratic revolutionary period, the multiparty system, the people's front, and so on, in these countries have undergone some changes. But compared with the period from the late 1940s to the early 1950s, they have developed further rather than diminished. Hence, it is not true to say that such a "peculiarity" no longer exists and has merely historical significance. In the current period, the role of the democratic parties and groups and people's fronts has improved considerably, and they are highly valued. The further improvement of these organizations is the context of the political system reform in many socialist countries. Yugoslavia pursues a system of delegation, nongovemmentalization of the party, and a self-government principle in defining the relationship among political organizations at each level. The lower levels are not subordinated to the higher level. In addition, Yugoslavia applies the principle of negotiated agreements among the member states in the union, rather than rule by, for example, the majority or equal power principle. All these features show that the political system of Yugoslavia differs from the Soviet democratic political system and the people's democratic political system. Apart from the classification of the total political system, we can also classify various subsystems or different aspects of various subsystems. Under the broad category party system, as mentioned earlier, there are one-party systems, as found in the USSR, Mongolia, Romania, and Cuba, and multiparty cooperative systems under the Communist party, as found in China, Czechoslovakia, East Germany, Poland, and Bulgaria. Under state systems, there are federal systems, such as in the USSR, Czechoslovakia, and Yugoslavia, and unitary systems, found in the other countries. There are bicameral or multicameral legislative systems (e.g., in the USSR, Czechoslovakia, and Yugoslavia) and unicameral systems (e.g., in China and many other countries). Under the state executive, there are collective-head systems (e.g., in Yugoslavia and the USSR), and individual-head systems (e.g., in Romania, Korea, and Cuba). Thus, while it stands to reason that "state organization rests, in principle, on a similar foundation in all socialist countries," it is untenable to say that "the existing differences cannot be characterized as distinctive types or varieties." Take another instance of the role and function of the Communist party, especially the instance of party-government relations, which is a pressing problem in the construction and reform of the political system. As Topornin said, in every socialist country, the Communist party is the nucleus of the political system and masters the leadership that is provided by the constitution and that is being strengthened steadily. We must note, meanwhile, that this kind of strength is realized gradually in the process of

138

On Classification

of Socialist

Political

Systems

political system reform, and in the long-term effort of improving the party's leadership. In both theory and practice, however, there are several different subtypes of socialist political systems with regard to the role and function of the party, and the relationship between the party and government. Division between the party and government under the leadership of the party: in this subtype, while the leadership of the party is stressed, the power and function of the party and the government are clearly differentiated according to their different natures. The institutions of the party do not manage affairs of state and society directly. They only give political instructions regarding the policies and direction of the state, and the appointments and removals of the cadres. China, the USSR, Hungary, and Bulgaria primarily belong to this subtype. "Destatism" of the League of Yugoslavia Communists: the League of Communists of Yugoslavia maintains that it is a political and ideological organization, but that it does not ask for a "monopolized" position in governmental institutions and does not intend to exercise leadership by using the power of the government. Also, the Yugoslav league contends that it should not become the real leader and direct manager of political, social, and economic life; that various social and political organizations are, within their own discretion and domain, independent and exempt from interference by the league. Functions of the party proposed by the Polish United Workers' party: the resolution of the tenth conference of the Polish party declares that the party serves the working class, laborers, and people, and functions to instruct the society and lead the country. The party maintains that its leadership is the most important principle in the political system, but that the party should not rise to be the "super unit," replacing the administrative institutions. Further, the party does not enjoy the power of appointment and removal of personnel in organizations outside itself. In K o r e a and R o m a n i a , t h e r e is a u n i f i e d s y s t e m of p a r t y and government (i.e., leaders of party and government hold concurrent posts, as the leaders of every level of the party also are the leaders of the same level of government). In summary, though all socialist countries uphold the principle of party leadership, the concrete devices and practices to implement this principle are not identical. Only when we distinguish different situations and subtypes, can we understand deeply and accurately the peculiarities of various political systems, and only on such a basis can we grasp their common features and general laws. While stressing the divisibility of the socialist political system, we should simultaneously realize that, as the political system of each socialist country develops, general laws of behavior will become progressively more clear. It is an important fact that the self-managing democratic political

Commentary/Huang

Zhong-Liang

139

system and the two other subtypes of socialist political system are converging from their two extreme poles of overconcentration and overdivision. Although I stress the divisibility of the socialist political system, I do not intend to evaluate these subtypes and decide which is best. Also, I do not intend to imply that a socialist country without an organization such as a national front should establish one, or that a country with a one-party system should restore a multiparty system. Each nation has its own peculiar situation and history, and only by relying on its people can the problems of each nation be resolved in practice.

7 Typologies of Socialist Systems Jim Seroka

The Problem After over seventy years of study and examination, our perceptions of world socialism are contrasting and extremely limited. Our viewpoints are analogous to the problem encountered in the fable about a group of blind men in India who had never seen an elephant. Each was asked to describe the whole animal after touching one part of it. Each man, having felt a different part, came to a different conclusion about the nature of the beast. One, who had felt the tail, likened the elephant to a rope; another, who had touched the trunk, characterized it as a type of snake; a third, who had come into contact with the elephant's flank, perceived it to be a wall; and still another, who had felt its leg, thought that the elephant was a kind of tree. Each blind man had identified an important feature of the elephant, but the totality of their experiences failed properly to describe the creature. In general, attempts to characterize the socialist state in the modern world have fallen into mutually exclusive patterns. Separately, each characterization or model has some significant strengths. Each also has major inadequacies. One by one, each model provides only one portion of the picture to describe the esscnce of the socialist political system. Combined, the models provide an incoherent and mutually antagonistic description of political reality in the socialist world. Unlike that of the men in the fable, our blindness is self-induced and often is motivated by political considerations. One characterization of the world socialist political community is that it is a monolithic bloc. Among some Western scholars and others, monolithism is equated with an all-encompassing Stalinist-totalitarian system. The socialist bloc is perceived to be expansionist in its intents and evil in its values and goals. U.S. President Ronald Reagan's description of the Soviet Union as the "Evil Empire" provides some suitable imagery to this viewpoint. An equally misleading, but essentially similar, monolithic view is the somewhat naive premise of the world socialist community behaving as a 141

142

Typologies

of Socialist

Systems

united internationalist brotherhood and fraternity of nations. Again, world socialism is perceived to be monolithic, though its activities are friendly and its goals and methods constructive and benign. The second major characterization of socialism appears to be a favorite among many country specialists and historians. It views the socialist nationstates as distinctive nations that are, at best, only marginally influenced by the adoption of socialism. In this formulation, there is a Russian, Polish, Hungarian, Yugoslav, Albanian, Chinese, and Vietnamese brand of socialism; and each nation's personality or its political culture asserts itself and determines that nation's political future (Brown 1984). Thus, Soviet behavior can be described as part of an innate Russian drive for warm water ports, Chinese actions are solely motivated by the historical legacy of the Middle Kingdom, and the German Democratic Republic is simply an updated extension of the Prussian state or Order of Teutonic Knights. A third descriptive framework views the socialist nation-states as essentially indistinguishable from other nations. Like the political cultural view, this viewpoint denies any significant impact to ideology. All nations, regardless of ideology, are essentially alike f o r better or f o r worse. Convergence models are one form of this conceptualization, and the convergence is either toward a fatalistic political future reminiscent of George Orwell's 1984, or toward a Utopian and peaceful world community working in harmony and solidarity, without the trappings of national interests or the backwardness of state ideologies. As this brief review suggests, adoption of any of the above-mentioned seriously impedes comparative inquiry about socialist nations. Comparison would be meaningless because: all socialist states are the same; or national culture, not socialism, explains all national behavior; or socialist states are not significantly different from nonsocialist states. My position is that all three viewpoints are incorrect, and that the truth lies somewhere within the pyramid of extremes charted by these constructs. In this chapter, I describe the goals of a typology for socialist state comparisons, define the universe of socialist states, and discuss inter- and intra-systemic variation among socialist states. In addition, I present a critique of previous models of socialist state behavior, and I propose an alternative approach for further development and testing.

Purpose of a Typology Much of the efforts involved in the study of comparative socialism in the West, and even within socialist states, has focused on the need to develop a common and consistent model that is applicable to all socialist systems. This quest is analogous to the discovery and testing of theories in the natural sciences with the hope of articulating firm and immutable laws. Paraphrasing

Jim Seroka

143

Karl Deutsch (1973; see also Orescu 1983), such a model should fulfil the following functions: an organizing function that joins interrelated facts; a heuristic function that recognizes the dynamics in a course of events; and a prognostic function that allows forecasts concerning the evolution of the system. The achievement of all three functions is necessary for the development of a comparative socialist political theory that would explain the behavior of these systems today and in the past, and their probable behavior in the future. Unfortunately, an appraisal of the state of science regarding the comparative study of socialism indicates that the goals of scientific inquiry have not yet been satisfied. In fact, I argue that the necessary precursor to theory development—the construction of an adequate typology in the field—has not yet been achieved. The study of comparative socialism must, at a minimum, identify those variables that can help lead to the construction of a useful typology and eventually to an appropriate and meaningful social theory. In other words, our current scientific inquiries should be more modest than are the development and postulation of models and theories. We first need to identify and catalog regular patterns of behavior and identify and catalog meaningful intrasystemic variation. At this point in our logical development, a philosophical judgment must be made. If all socialist political systems are observed or perceived to be fundamentally alike, the development of a typology would not be very useful. If political behavior and political outputs of the socialist nation-states are similar, then we can assume that legal, institutional, and political cultural variation is not very significant in explaining what these nation-states do. A contrary judgment, and one that I will develop further, is that there is considerable differentiation in the political behavior and politics of socialist nation-states. If such a judgment is accepted, then those legal, institutional, ideological, and political characteristics that are essentially similar across socialist states most probably do not contribute to the interstate variation in behavior. Our attention, therefore, must focus on those areas where considerable variation across states exists. In scientific fields such as this, where there are relatively few (sixteen) cases, the construction of political typologies must place considerable emphasis upon the explanation of the deviant case or cases (Linden & Rockman 1984). Thus, for example, though Albania or Kampuchea may be relatively small and insignificant socialist states, their deviant behavior is inherently significant. Deviance permits social scientists to speculate about which variables contributed most to the atypicality and to identify those variables that are most meaningful to the development of a typology that explains socialist state behavior. To summarize, in those political arenas where we judge interstate behavior among socialist nations to be very similar, we pay particular attention to variables that are intersystemically similar. If, on the other hand,

144

Typologies

of Socialist

Systems

we judge interstate socialist behavior to be dissimilar, we must transfer our attention to those variables that are intersystemically different.

Variation Among Socialist States For purposes of this analysis, I will confoim to conventional practice and define the universe of socialist states as those independent nation-states in which the Communist party holds a monopoly of power, the state controls the means of production, and Marxism-Leninism is the avowed ideology of the state, and as those states that subscribe to a foim of socialist internationalism (White 1983). Using these criteria, sixteen nations emerge as socialist: China; Mongolia; the USSR; Democratic People's Republic of Korea; Vietnam; Kampuchea; Laos; Poland; German Democratic Republic; Czechoslovakia; Hungary; Romania; Bulgaria; Albania; Yugoslavia; and Cuba. While some strong arguments could be made for including nations such as Ethiopia, Afghanistan, or South Yemen within this catalog, the existence of, or relatively high probability of, civil war, very low commitment to socialist internationalism, or a weak state presence in the economy would make an even stronger counterclaim against inclusion. Even within this limited definition, the socialist group of nations encompasses a large proportion of the world's land mass and one third of its total population. There are members in both hemispheres and on three continents. The socialist international community includes very different cultures and some of the least and most developed nations on Earth. Membership is held by the world's most populous and largest nations as well as by some of the smallest and least populous nations. Table 7.1 provides a comparative listing of some of the basic political and demographic differences among socialist states. Total population varies between under 2 million to over 1 billion inhabitants. Party membership ranges from under 50 thousand to an estimated 44 million, and the percentage of the population who are party members ranges from 1 percent in Laos to nearly 16 percent in Romania. Even Western indices of political democracy note major differences among the socialist nation-states. As Table 7.1 shows, the political democracy index scorcs range from a low of five in Cuba to 51 in Yugoslavia. Variation among socialist states is even more extreme when we consider some of the basic measures of social and economic development as represented in Table 7.2. For example, nearly 100 percent of the population of the Soviet Union is reportedly literate, compared to 28 percent in Laos. Male life expectancy ranges from a low of forty-two in Kampuchea to seventy-one in Cuba. The infant mortality rate, a generally accepted comparative health measure, shows a range from 175 per 1,000 births in Laos to only 8.5 in the Soviet Union. Per capita income differentials are even more

Jim Seroka

145

Table 7.1 Comparative Political-Demographic Data for Socialist States

Nation

Population

Party Membership

Party Members Political (% Population) Democracy Index

Albania Bulgaria Cuba Czechoslovakia East Germany Hungary Kampuchea Laos Mongolia North Korea People's Republic of China Poland Romania Vietnam USSR Yugoslavia

3,020,000 8,990,000 10,221,000 15,542,000 16,692,000 10,624,000 6,288,000 3,679,000 1,942,000 20,543,000

147,000 932,055 523,639 1,675,000 2,304,121 870,992 3,000 47,000 88,150 2,500,000

5 10 5 11 14 8 insignificant 1 5 12

18.2 37.5 5.2 20.5 18.1 11.6 na na 16.2 21.0

1,045,537,000 37,546,000 22,830,000 61,994,000 279,904,000 23,284,000

44,000,000 2,125,762 3,557,205 1,700,000 18,500,000 2,168,000

4 6 16 3 7 9

16.4 22.1 16.0 33.1 18.2 50.8

Sources: Population and party membership figures are derived from Richard F. Staar (ed.), 1987 Yearbook in International Communist Affairs (Stanford, CA: Hoover Institution Press). The political democracy index is for 1965 and is found in Kenneth Bollen, "Issues in the Comparative Measurement of Political Democracy," American Sociological Review 45 (June 1980): 370-390.

profound, with a 47 to 1 range in the per capita income levels from East Germany to Laos. The variation among socialist states summarized in Tables 7.1 and 7.2 demonstrates conclusively that within the world socialist community of nations, states are socially and economically progressing at different rates and are facing different problems and challenges in their proclaimed attempts to construct new forms of socialist society. Variation in economic, social, and political development must, of necessity, affect the capabilities of the party to manage and direct society. It must also influence the goals and needs of society as interpreted by the party, and it would conclusively shape the political climate and political culture of a state. We can also expect that the quantitative changcs would underscore the range of qualitative differentiation in the politics and society of these sixteen nation-states. To illustrate this point, we can observe the family-based oligarchy in Romania as it contrasts sharply with the more developed bureaucratized political culture found in neighboring Bulgaria or the relatively open pluralist system in Yugoslavia. An additional contrast may be Poland's former semianarchistic political climate when it is pitted against the well-structured and disciplined form of socialism found in East Germany, the rational market-oriented socialism in Hungary, and the conservative practices of Czechoslovakia. In Asia, North Korea's, Vietnam's, and China's national socialist styles have little in common, and the differences in Kampuchea's

146

Typolog ies of Socialist

Systems

Table 7.2 Comparative Socioeconomic Data for Socialist States Measure

Nation Albania Bulgaria Cuba Czechoslovakia East Germany Hungary Kampuchea Laos Mongolia North Korea People's Republic of China Poland Romania USSR Vietnam Yugoslavia

Infant Mortality Per Capita Income 1976 (per 1,000 (in 1976 U.S. dollars) Live Births)

Male Life Expectancy

Literacy Rate

490 2,100 800 3,500 4,000 2,100 90 85 750 450

86.8 21.8 22.3 18.7 13.0 23.7 na 175 na 70.0

69 68 71 67 70 66 42 48 60 65

(1985) (1980) (1978) (1984) (1984) (1984) (1985) (1985) (1985) (1985)

na 91 78 100 100 98 42 28 99 90

380 2,500 1,600 2,600 160 1,540

56.0 21.3 30.3 8.5 115.0 32.2

67 67 67 64 57 68

(1985) (1984) (1978) (1972) (1985) (1981)

40 98 100 100 60 83

Sources: Per capita income from The World in Figures (New York: Facts on File, 1980); infant mortality from The New Book of World Rankings (New York: Facts on File, 1984); life expectancy from Demographic Yearbook 1985 (New York: United Nations, 1985); literacy rate from The World in Figures (New York: Facts on File, 1980).

socialist models during recent decades belie further any claim for a standardized socialist model in the world.

Competing Models for Comparative Socialist Study Although there have been few, if any, acceptable typologies of socialist nation-states, research in the socialist and nonsocialist scholarly communities has created a large number of competing models or general descriptive orientations for discussion about the nature of politics in a socialist society. Some of these models are very useful in presenting us with a clear picture of the socialist community of nations at one point in time. Some are amenable to empirical testing and verification. None, however, are capable of dealing with more than a single explanatory variable for a full range of political behavior. None meet simultaneously the organizational, heuristic, and prognostic functions of a fully developed theoretical model. These models include the legal-institutional, monolithic-totalitarian, modernization-convergence, and competitive-bureaucratic models. The legal-institutional comparative model is the most prevalent form of comparative analysis and has been used within socialist countries as well as

Jim Seroka

147

outside them (Brown 1985). The model forms the basis of international datagathering organizations such as the International Parliamentary Union. It compares and contrasts constitutional provisions; parliamentary procedures, structures, and institutions; and other formal political characteristics. Its major strength is that it is relatively uncontroversial. Its weaknesses, however, include the neglect of behavioral considerations, temporal stasis, and absence of cross-cultural differences. Legal-institutional models have rarely identified significant differentiating factors among socialist states, and they often accept without proof the presence of a causal relationship between structure and political policy outputs (Atamanchuk 1985). Often, this mechanical causal orientation between structure and process leads to the development of conclusions that do not reflect political realities. It may also lead political experts to lay the blame for unrealized political goals upon the weaknesses of political institutions, rather than upon more substantial social, political, and economic characteristics. As is well recognized in historical studies, institutions may often be irrelevant to the broad sweep of historical progress. Similarly, it may be mistaken to believe that reforms or changes in political institutions are a sufficient remedial solution for a major social problem. Legal-institutional models, in short, confuse change in form with change in substance. Another major weakness of the legal-institutional form of analysis in comparative socialist studies is its inability to predict change. Changes in institutional structures often follow major social changes; they rarely contribute to them. As can be seen from the recent historical developments in China or Poland, institutional reform follows social changes and does not precede them. Structural functionalism, a type of legal-institutional analysis, is only meaningful in a static world, lacks a perspective on developmental change, and can only give, at best, ex post facto explanations of sociopolitical phenomena. Finally, legal-institutional models have great difficulty accommodating themselves to cross-cultural differences. National legislatures, for example, may have entirely different social functions and reference points in one society as compared to another. The legislature in Poland or Yugoslavia, for instance, may have a more important impact on political processes and politics in their respective countryies than is the case in Albania or North Korea. The second group of models conceive of the socialist international community in the context of the totalitarian-authoritarianism that was characteristic of the Stalin era (Friedrich & Brzezinski 1965). The system that these researchers describe concentrates all power in the hands of a single leader who exercises complete control over society in an attempt to reconstruct the national character and build a "new socialist man." Deprivation of personal freedom, complete centralization of power, arbitrary exercise of power, and neglect of rule of law are some of the characteristics

148

Typologies

of Socialist

Systems

of this system. These totalitarian models possess an extremely static perception of society and deny any relevance to differentiation among socialist states. Smaller and weaker socialist states are labeled puppets or satellites of leading socialist nations; they have no autonomy or independence. Little attempt is made to compare or explain political processes because politics in socialism is interpreted to be erratic or irrational. Internal politics is limited to the question of succession or maintenance of the police state. The totalitarian model, in short, meets few, if any, of the necessary prerequisites of a workable typology for comparative socialist study. The modernization-convergence models (Kautsky 1973) basically argue that i d e o l o g y is l a r g e l y i r r e l e v a n t in c o m p a r a t i v e study, and t h a t socioeconomic development forces socialist and nonsocialist states to merge their political agendas and develop common interests and policies. Some of the evidence that has been presented to verify the existence of this trend include the lack of differentiation between budgetary and policy priorities, comparing socialist and capitalist states and controlling for the level of socioeconomic development (Pryor 1968), and the appearance of similar forms of interest group articulation in both types of societies. Interest group models and empirical comparative leadership/elite studies are two additional forms that modernization-convergence models have taken. Fleron, Johnson, Kanet, and Skilling are just four of the numerous Western scholars who have examined the issue from these perspectives (Heron 1969; Johnson 1970; Kanet 1971; Skilling & Griffiths 1971). Unlike the totalitarian models, modernization-convergence models are both dynamic and prognostic. They also employ a primitive expression of M a r x i s t theory in which e c o n o m i c and social d e v e l o p m e n t s h a v e a correspondingly incremental impact upon political development. In these models, all states and societies are marching in rhythm (albeit at different rates) toward an identical political future. Pessimists foresee the future in Orwellian terms, while optimists project a more benign, peaceful future for human society. A major weakness of the modernization-convergence models is the r e l a t i v e i n s i g n i f i c a n c e assigned to h u m a n e f f o r t and i d e o l o g i c a l considerations. In reality, though both socialist and capitalist states are undergoing profound changes, they are responding to these changes very differently. In addition, the responses within each group of nations vary considerably. The goal—a communist society—may be similar for each socialist nation, but the precise vision of what constitutes that goal will differ. Leadership succession questions in the Soviet Union, China, Poland, Albania, Vietnam, and Yugoslavia have renewed interest among Western scholars in bureaucratic models of socialist politics. Unlike the models that virtually ignore the role of the Communist party in formulating and implementing policy, bureaucratic models assign extreme importance to

Jim Seroka

149

factional conflict and competition within the party and state bureaucratic apparatus; issues regarding political legitimacy become very important (Rigby & Feher 1982). Nearly all politics is explained by intraorganizational maneuvering (Hough 1969), and no one acts for anything other than individual careerist or organizational interests. Compared to the modernization models, bureaucratic models lack objective scientific rigor and long-term forecasting capacity and are open to second-guessing and constant reinteipretations about the "true" meaning of various events. Bureaucratic models are also reactive, rather than proactive. In other words, they can explain after the fact what has happened but are very weak at predicting results. The most significant weakness of the models, however, is that bureaucratic conflict analyses are not, by nature, comparative. Each socialist state possesses its own unique mix of bureaucratic actors, and each nation-state alters the mixture over time with no visible formula for change. This brief review of competing models of the study in comparative socialism indicates a paucity of true scientific tools for comparative analysis. Some of the models lack predictive capability; some are static; some are not comparative; and some do not even organize the facts. None of the models are simultaneously predictive, dynamic, and comparative. None serve simultaneously the organizational, heuristic, and prognostic functions of a true typology. Table 7.3 summarizes in encapsulated form the strengths and weaknesses of each of the models with respect to the development of an adequate typology for comparative socialist study. An inevitable conclusion is that more needs to be done.

Toward a Workable Typology for Comparative Socialist Study A typology for socialist nation-states should be based upon criteria that are most relevant for differentiating these socialist states from one another. The criteria should take into account variation within the socialist community, the essential characteristics of a socialist political system, deviant cases, and the organizational, heuristic, and prognostic functions of a workable typology. Table 7.3 Utility for Typology Development of Models of Socialist State Politics Model Criterion Organizational Comparative-Heuristic Prognostic

LegalInstitutional Yes No No

MonolithicTotalitarian No No No

ModernizationConvergence Yes No Yes

CompetitiveBureaucratic Yes No No

150

Typologies

of Socialist

Systems

Earlier in this chapter, I identified four factors that are instrumental in distinguishing socialist nations from other nations. Those four factors—party monopoly, state control of the means of production, Marxism-Leninism, and socialist internationalism—may also be useful in setting up a comparative organizational framework. In addition, the level of socioeconomic development of a society would be instrumental in establishing the parameters of change and progress toward socialism. Thus, the essential characteristics of state socialism, when combined with societal modernization, may form the basis for development of a comparative socialist typology. Conveniently, the definitional aspects of socialism and socioeconomic development correspond to the essential characteristics of the major models that have been used for comparing socialist states. The totalitarian model, for instance, identifies a socialist nation largely with respect to the monopoly of power held by the Communist party. The bureaucratic models are grounded in the experience of state control over the means of production. Legalinstitutional models rely heavily on application of socialist internationalism and Marxism-Leninism to the political process; and modernization models visualize socialist states as composed of mixtures of socioeconomic variables. As was the case in the Indian fable of the blind men and the elephant, each model has pinpointed an essential characteristic of state socialism. Perhaps, when the models are combined, a relatively more complete and accurate picture can emerge. My proposition is to introduce a typology for comparing socialist states based on three major dimensions: the continua of monopoly by the Table 7.4 Suggested Representation (or Socialist Nations on a Three-Dimensional Categorization, 1987 Dimension

Albania Bulgaria Cuba Czechoslovakia East Germany Hungary Kampuchea Laos Mongolia North Korea People's Republic of China Poland Romania USSR Vietnam Yugoslavia

Party Monopoly

State Control

Modernization

High Moderate High High Moderate Moderate Low Low High High

High High High High High Moderate Low Low High High

Low Low Low High High High Low Low Low Low

Moderate Low High Moderate Moderate Low

Moderate Moderate High High High Low

Low Moderate Moderate Moderate Low Moderate

Jim Seroka

151

Communist party (i.e., the Stalinist-totalitarian model); state control over the m e a n s of p r o d u c t i o n ( b u r e a u c r a t i c m o d e l ) ; and c u r r e n t l e v e l s of socioeconomic development (modernization-convergence models). While a complete typology should include both socialist internationalism and Marxism-Leninism, practical difficulties in accurately measuring and comparing these phenomena justify temporary exclusion. The classification listed in Table 7.4 illustrates the remarkable diversity among socialist states. With this measurement instrument, nations in the socialist world today divide themselves into twelve different patterns out of a total possible eighteen. Each pattern is distinctive; each generates different behavior; each leads to different policies and politics. Conceptually and practically, the patterns make considerable sense. The position of each nation in the typology can be visualized as a point in a three dimensional space. Each point is moving relative to its previous position and to the positions of other socialist nations. Nations would cluster differently over time, and we would expect that the changed position would alter a nation's behavior and politics. All assignations in Table 7.4 are relative, and no interval implication should be made for any of the dimensions. In addition, the party monopoly and state control assignations are currently based on subjective interpretations and could be questioned by others with more precise information. The assignations are largely based on the data listed in Tables 7.1 and 7.2, but some exceptions exist. More work, in other words, must be undertaken to make the typology a more precise instrument for measurement. The advantages of the typology are numerous: For social scientists, it can be empirically tested, empirical measures can be constructed, division points can be established, categorization of socialist types can be formed, and the theoretical and practical implications can be empirically examined and verified. We have the beginning of a useful tool with which to compare, group, and contrast socialist states. We can also examine changes in their behavior over time and forecast policy changes as socialist nations alter their positions in the typology. This sort of typology permits social scientists to focus their area of inquiry and to ask relevant questions. This particular typology can also serve as a bridging mechanism between scholars from socialist and nonsocialist states who are concerned with politics in state socialism. The typology does not assign normative values to the characteristics nor apply politically symbolic terminology. It also examines characteristics of socialist states that are generally agreed upon by many scholars from both groups as the primary identifying characteristics of socialism. If we are prepared to acknowledge that world socialism today is not monolithic, then we should be excited by the challenge of determining what differences the variation among socialist states may make. To illustrate the capability of the typology, we can consider the cases of

152

Typologies

of Socialist

Systems

Albania and Poland. If Albania were to rapidly expand its base of social and e c o n o m i c modernization in the coming decade and retain its other characteristics unaltered, would its predicted behavior pattern more closely resemble that of contemporary Romania, as suggested by the typology? If the Albanian party were gradually to relax its monopoly over societal affairs, would its internal political behavior begin to resemble that of contemporary Vietnam? If Poland's party were to reassert its involvement in public life in the coming decade, would Polish politics radically change? Through this typology or a similar arrangement, these questions and myriad others can be tested, and general social theory can be further developed.

References Atamanchuk, Grigori. 1985. The social content of administration. Paper presented at the World Congress of the International Political Science Association, Paris, July 15-20. Brown, Archie (ed.). 1984. Political culture and communist studies. London: Macmillan. . 1985. The concept of political system and soviet theory of the state. Paper presented at the World Congress of the International Political Science Association, Paris, July 15-20. Deutsch, Karl W. 1973. Politische Kybernetik. Modelle und Perspectives Freiburg: Rombach Verlag. Fleron, Frederic (ed.). 1969. Communist studies and the social sciences. Chicago: Rand McNally. Friedrich, Carl J., and Zbigniew K. Brzezinski. 1965. Totalitarian dictatorship and autocracy. Cambridge, MA.: Harvard University Press. Hough, Jerry. 1969. The soviet prefects. Cambridge, MA.: Harvard University Press. Johnson, Chalmers (ed.). 1970. Change in communist systems. Stanford, CA: Hoover Institution Press. Kanet, Roger (ed.). 1971. The behavioral revolution and communist studies. New York: Free Press. Kautsky, John H. (1973). "Comparative Communism Versus Comparative Politics." In Studies in comparative communism, 6:135-170. Linden, Ronald H., and Bert A. Rockman (eds.). 1984. "The study of Communist Elites and the Contribution of Carl Beck." In Elite studies and communist politics. Pittsburgh, PA: University of Pittsburgh Press. Orescu, Serban. 1983. A structural-functional model for the comparative study of communist systems. Studies in Comparative Communism 16(4):265. Pryor, Frederick L. 1968. Public expenditures in communist and capitalist nations. Homewood, IL: Richard D. Irwin. Rigby, T. H., and Ferenc Feher (eds.). 1982. Political legitimation in communist states. London: Macmillan. Skilling, Gordon, and Franklyn Griffiths (eds.). 1971. Interest groups in soviet politics. Princeton, NJ: Princeton University Press. White, Stephen. 1983. What is a communist system? Studies in Comparative Politics 16(4):247-264.

B Commentary Miodrag Jovicic

All attempts at establishing a typology of socialist systems must take as their starting point the fact that the socialist countries are somewhat different from other countries, and that they, of course, do not represent a monolithic bloc. These countries were not monolithic even thirty or forty years ago, and there is considerable diversity among them at present. Thus, it is necessary to try to establish the criteria that will make it possible to form a typology of socialist countries that would classify, into several distinct groups on a scientific basis, all the socialist systems. For the time being, I think that we must agree with Seroka that there does not exist any satisfactory typology of socialist systems. Those who tried to provide such a typology always started from a more or less excessively narrow basis, taking as relevant for their typology only one or another property of socialist systems. Therefore, the results, which could be considered satisfactory if it were only a question of the legal-institutional aspect or the party structure (i.e., a one-party system or a multi-party system) of socialist countries, cannot be accepted as sufficient for a comprehensive typology of socialist systems. The fundamental reason for this lies in the fact that a partial observation of socialist countries may, in general, call into question the possibility of distinguishing socialist countries as a particular group. For example, if the relevant criterion were legal-institutional considerations, there would be considerable similarities to the systems of other, nonsocialist countries. This would make worthless the attempts at establishing the typology of socialist systems. On the other hand, the application of this or some other criterion as the exclusive one would lead to the presentation of a typology of socialist countries that would not present the actual picture of similarities that these countries, considered in totality, have with one another. Seroka's determination to identify the characteristics that would classify individual countries into the category of socialist states is correct only in principle. This is true for two reasons. First, each of the four characteristics he has identified are not equally applicable to all the socialist countries. For 153

154

Typologies

of Socialist

Systems

instance, the last one, which affirms that all the countries are loyal to socialist internationalism, is not valid for one group. Second, there have been conspicuous, gradual changes in many socialist countries for each of the characteristics Seroka employs, and, thus, their relevancy changes over time. Flexibility is necessary at present, and it will become even more necessary in the future, particularly in the near future. Thus, without going so far as to predict the convergence of all the systems (which will, in all probability, occur at some time), it would be indispensable to make a serious attempt now to define the socialist system in a new way that is closer to reality. I cannot, of course, provide an alternate framework here and now because the success of such an undertaking would necessitate the joint effort of entire teams of experts. However, I have several comments regarding the individual competing models for comparative socialist studies that were cited by Seroka. The legal-institutional comparative model is, undoubtedly, the easiest and most comfortable model for comparison. The ideas of the legal organization and the institutions of the system are acquired in a comparatively simple way, and they can be specified by examining the constitutional texts and the body of legal acts. This is an advantage in relation to other models. This approach must not be overlooked and must be one of the elements of a typology. However, it is not sufficient by itself. As is well known, all over the world there exists a discrepancy between legal norms and political reality. This is particularly true in those socialist countries wherein constitutions are often written not only for domestic use, but also for "external application." Thus, they often serve to embellish reality. On the other hand, a series of institutions established by the constitution either do not function at all as regulated by the constitutions, or function in a different way. In a word, the legal-institutional expression of the systems of socialist countries is not completely adequate. Although it cannot be denied that there were elements of totalitarianism and authoritarianism in socialist countries (traces can be seen in some even now), they are not typical of socialist countries only. Totalitarianism has been, and continues to be, present elsewhere in the world. Overall, as Seroka states, totalitarianism does not supply a basis for a workable typology for comparative socialist study. The modernization-convergence model is indispensable to distinguish the theoretical-ideological aspect from the practical aspect of reality. It is undoubtedly possible to notice important analogies among modernization processes all over the world. Both politics and the economy have a similar objective everywhere: achieving material wealth and raising the standard of living for citizens so that they will reach a degree of socioeconomic development that will, in short, make life in organized societies in the twentfirst century better than in the nineteenth and even the greater part of the twentieth centuries.

Commentary/Miodrag

Jovicic

155

Important differences, regarding the means used to achieve this objective, between the socialist countries and the nonsocialist ones must be recognized. These means differ among themselves chiefly on the basis of ideology. Marxist collectivism is one choice, and liberal individualism is the other. Hence, the use of this modernization-convergence model is subject to criticism because the ideological frameworks are not taken into consideration. It is obvious that conflicts and compromises between the state leaders and the party leaders, which do occur and are the essence of the bureaucratic model are not specific to socialist countries alone. Such maneuvers are probably characterized by the fact that they occur chiefly behind closed doors. This bureaucratic leadership component of political systems cannot be exactly measured and, in addition, provides only a marginal impact for the typology of socialist systems. In my opinion, Seroka rightfully criticized each of the individually cited models and rejected the possibility of any one of them being used as the basis for a typology of socialist systems. This does not mean at all that the factors that these models identify should not be taken into consideration in trying to make a scientifically acceptable typology. Seroka's attempt at a typology of socialist systems is worthy of attention. He concentrates on only three relevant elements: the existence and the degree of party monopoly; state control over the means of production; and level of socioeconomic development (modernization). These elements are, in my opinion, well selected. Each of them, no doubt, exerts an influence on the determination of similarities and differences that exist among individual socialist countries (i.e., each can contribute to the creation of a usable typology). They are, in addition, empirically verifiable to a sufficient degree. They are not static, but dynamic, which enables us to test them over time and to make corresponding corrections in evaluations. I consider Table 7.3 a good illustration of the fact that the use of three criteria is well founded. There is, certainly, the possibility of individual disagreements in evaluating the particular systems on each of the elements of the typology. Seroka, undoubtedly, has made a serious contribution to the search for "the best" typology of socialist systems. It is a difficult and complex task, however, to reach this goal. Each effort, as well as this one, is one step further on the path. In fact, attempts at building a complex typology of socialist systems should be intensified. For this reason, I take the liberty to suggest the possibility of taking into consideration some other elements relevant to this typology. Social-historical conditions of the system: it is known that the introduction of a socialist organization in individual countries occurred under very different conditions and with different degrees of socioeconomic development and historical preparedness for the new organization. The

156

Typologies

of Socialist

Systems

differences here are enormous, and they are well known. Some of these differences have lessened over time, but they undoubtedly have left some traces on the present state. This condition must be taken into consideration in the elaboration of the typology. Class structure: great differences, not only in the past but also at present, regarding the class structure of individual socialist countries are of pronounced relevance f o r determining the character of the socialist organization. The framework of the proclaimed dictatorship of the proletariat in all the socialist countries (with some specifics in Yugoslavia) is available. It is obvious that the number of, and the role played by, the working class, on the one side, and of other classes, most importantly the peasantry, on the other side, represent one of the elements for the typology. This variable must not be static. Forms of property: here, there exist considerable differences among the individual socialist countries, not only regarding the normative establishment of forms of property but, of particular importance, in identifying their real existence and relative strength. This domain is subject to changes in a number of socialist countries, and, consequently, it is necessary to take this into account as well. Organization of the economy: the differences that exist in this dimension are, to a certain extent, related to the different roles of the forms of property. T h e y are i m p o r t a n t and vary from the strictly centralized f o r m s of organization of the economy, which are based on state property and on detailed, obligatory planning, to a comparatively liberal organization of economic subjects, with the participation of other forms of property (in addition to state property) and with elements of a market economy. This is an especially important element for the development of socialism in each particular country. System of power: beginning from the well-known fact that there are no two countries in the world that have identically organized power, it must be noticed that among the socialist countries, without taking the details into c o n s i d e r a t i o n here, there are very important d i f f e r e n c e s . T h e s e are manifested, for instance, in the application of federal or unitary organization, in the varying degrees of autonomy of local organs of authority, in different forms of application of the so-callcd assembly system (a form of horizontal organization of power), and so forth. All these points, though they probably do not exert a substantial influence on the character of socialist organization, represent elements of the system and should be included. Forms of political organization: without contesting the decisive part played by the Communist party in all the socialist countries, there are also other forms of political organization that occur in some of these countries and that are of some interest. As is known, there is a great variety regarding the part and the place occupied by other political parties in the system. Although they frequently represent a decorative accessory, we should not disregard the

Commentary/Miodrag

Jovicic

157

fact that they sometimes exist as an element relevant for the typology. The ultimate goal of the efforts to elaborate a typology of socialist systems should be, in my opinion, the establishment of several particular groups within which all the socialist systems could be classified according to their mutual relationship, on the basis of predetermined criteria. To do so, it is indispensable to establish beforehand a list of these criteria and to rank their importance. Current efforts are obviously oriented in this direction, and Seroka's discussion of these criteria is very useful. When this first phase is completed, it will be possible to undertake, with greater scientific precision, the second one—i.e., the elaboration of a complex typology of socialist systems, with a suitable division into several groups and, possibly, subgroups. In addition, it is obvious that a typology established in such a way should be submitted periodically to revision, in conformity with the changes that occur in the world of socialism.

8 The Comparative Study of Third World Political Systems Veniamin

Chirkin

The political systems of the more than 120 developing countries display a great variety and exhibit major differences, which are, at times, of fundamental significance. It is, above all, the nature of political power and the trends in sociopolitical development that determine these differences. At a number of levels, there are superficial resemblances among political systems: For example, both the People's Republic of the Congo and neighboring Zaire have one-party systems. However, the political system of the Congo and other countries of socialist orientation (Angola, the People's Democratic Republic of Yemen, Nicaragua, and others) differ fundamentally from that of Zaire and other countries with a capitalist orientation (these include a number of tropical African and Oceanian states: Ruanda, Niger, Western Samoa, Papua New Guinea, and others). The latter are following a capitalist path of development, but the local bourgeoisie is weak and has not concentrated all power in its own hands. The political systems of countries with a socialist orientation are also not uniform. Both capitalist and socialist orientations differ from the situation in developing countries that have a bourgeois (e.g., India), bourgeoislandowning (e.g., Paraguay), or semifeudal character (e.g., Saudi Arabia). There is one further difficulty in analyzing and classifying those states developing along the capitalist path. They have adopted Western political institutions, but these have often become no more than an external shell. The political processes in such countries (e.g., Malaysia, Kenya, El Salvador, and Kiribati) are following quite different models. Then, there are phenomena of a different kind, which could place insuperable barriers in the way of the assessment and classification of these political systems—if, that is, we approached them only from a normative point of view. Thus, certain countries that are actually following a capitalist path of development have proclaimed themselves to be socialist states in their constitutions (e.g., the 1976 Constitutional Amendments of India, and the 1978 Constitution of Sri Lanka). Such declarations have more to do with the power and influence of socialist ideas among the masses than with the real situation in these 159

160

The Comparative

Study

of

Third World

Political

Systems

countries. The ruling parties in certain countries of socialist orientation also include, in their constitutions and programs, clauses about the socialist state and socialist democracy when, in fact, their political systems are still only at various and sometimes very early stages in the transition to socialist aims. In a number of countries, this transition has not developed any further. In some cases, there has been a retrogression, as indicated by military coups that were organized by internal reactionary forces (e.g., in Ghana in 1966). Elsewhere, right-wing nationalist forces gained the upper hand in patriotic alliances, as occurred in Somalia in the late 1970s. At times, this was the result of direct imperialist intervention (e.g., the U.S. invasion of Grenada in 1983). The political systems of developing countries, therefore, exhibit a variety and range of distinctive features. Some of them are fundamental and distinguish them from both socialist and developed capitalist states. This poses new tasks for political scientists and demands that a well-thought-out methodology be devised for their study. Bourgeois political science has not had great success in explaining the distinctive nature of these systems and has proved unable to draw up a classificatory system based on substantive criteria. Many of the schemes it has proposed are useful, but they are mainly based on formal aspects and ignore the reality of the variety of political systems in socialist-oriented countries. Furthermore, the failures of bourgeois political scientists have, in great measure, been linked to their erroneous understanding of the essence of a political system. A genuine methodology for such studies is provided by dialectical materialism and the scicnce of Marxism-Leninism, which has been liberated from dogmatism and has taken into account the variation that occurs in the real world.

Bourgeois and Marxist-Leninist Concepts of the Political System Two scholarly categories underlie the concept of a political system: the philosophical category of "system" and the sociological category of "politics." In their interpretation of the former, bourgeois political scientists rely more and more on the propositions of cybernetics and informatics (direct links and feedback, "bytes of information," receptors, input-output, etc.). Such approaches can be useful in studying how political systems function, but only if they are not transformed into bare abstractions that ignore the social nature of political institutions. When they examine politics and political relations, bourgeois scholars usually begin with operational ideas about politics. Structuralist political scientists have put forward an all-embracing scheme of the political process, but their understanding of political relations is deprived of the chief quality of

Veniamin

Chirkin

161

such relations; namely, their class character. Policy is abstractly defined as a set of decisions that are taken by a political actor or group and concern the selection of goals and the methods of attaining them within a specified situation (Roberts 1971; Ponton & Gill 1982). Many Western political scientists adopt a personalized approach to the definition of politics. For them, it derives from, for example, human nature, conflicts between people, or human behavior in society (Houriou 1974; de Lespinasse 1981). Alongside this personalized approach, the group approach is also used. Here, politics is defined as the "expression of the group's being" (Debbasch et al. 1983). This helps one group obtain something in preference to another group and aids in controlling conflicts between them (Kourevartis & Dobrats 1980; Coulter 1981). It must be pointed out, however, that these scholars are not discussing classes at all. Groups ("interested groups" or, in the latest interpretation, "the structures for expressing interests") are defined according to the most varied criteria, but the definitions ignore the chief, class-based division of society. As a rule, bourgeois political scientists begin by stating that politics and political relations can be found in individual personal relations and in the smallest human communities—whether they be tribe, village, or family. "Politics," writes English author Leftwich (1983), "is a universal feature of all societies and all institutions and groups within them, not just of some of them." This unsound approach to the key concepts of politics and political relations leads Western political scientists to their erroneous conceptions about the political system. In different Western countries, distinctive approaches can be found to the study of the political system. The U.S. approach is pragmatic, in France it is constitutional and institutional, while in West Germany it is, to a great extent, philosophical. There are a great many definitions. They can range from the most narrow to those that embrace everything that has anything to do with politics (Entin & Entin 1986). Yet, for all this variety, the dominant concepts in bourgeois political science are based on behavioral, structural-functionalist approaches, combined with elements of cybernetics. Hence, the political system is seen, before all else, as a political process or as political behavior that takes place within the framework of one or another community. By "framework" is meant the "societal systems" (e.g., trade unions, industrial firms, clubs, cities), while the significance of political institutions is usually downgraded (Dahl 1963; Calvert 1983). The political system is understood to be a particular structure that is made up of behavioral roles. There is an input of information to the receptors, and then an output of political decisions that distribute resources (e.g., wealth, power, education) in society. In French political science, and in the countries under French influence, there is a more institutionalized understanding of the political system. Yet here also, it is often identified with the political regime (Duverger 1970) or interpreted as the interaction of political forces (Lavroff

162

The Comparative

Study

of

Third

World

Political

Systems

1975). A detailed review of the major Western concepts of the political system can be found in von Beyme (1980). The dominant view in bourgeois political science is that a great many political systems exist within each country. Parties, trade unions, tribes, and others, all have their own systems. Bourgeois academics assert that such systems are not necessarily linked with the state. Every society has a political system, writes English political scientist Bottomore (1979): To say this, he continues, does not mean, however, that every society has a "state." There are only a few authors who characterize the "systems" of party, trade union, and industrial firms' organization as parapolitical. Not many Western political scientists (Debbasch is one [Debbasch & Daudet 1974]) approach things from the standpoint of a unified political system. In spite of these reservations, all bourgeois conceptions of the political system share a common failing. They all neglect the central feature of such systems: their class nature and the possession of power by one or another social class. Bourgeois authors take the immemorial existence of the political system as their starting point; at the dawn of humanity, they claim, it was already "redistributing resources." They do not see that this formation is intrinsic only to class society, that it arose at a certain stage in social development and is able to exist for a only a definite period in social evolution. In their comparative studies of political systems, or in studies of individual countries, Western political scientists frequently analyze specific data. These works are academically sound and interesting, and their data and conclusions are useful. In their theoretical works, on the other hand, the treatment of the political system, and of political relations and the political process, is formalist and abstract. In practice, it is not abstract "bytes of information" that underlie such approaches, but voluntarist attitudes to class, which, in the final analysis and in antagonistic societies such as one finds in all developing countries, are linked with the struggle among classes for social advantages and, as the first step in obtaining such advantages, with their battle for the political and state power with which to secure their interests. Finally, we must note that the "distributive" theory of the political system ignores the decisive level of production. Before social benefits can be distributed, they must be created and produced. Marxist-Leninist political science takes a different approach to the political system. It begins with a fundamental philosophical distinction among the aggregate, the political system as a whole, and the total social system. The last of these is distinguished by the struggle between opposites that is its internal source of development. The scientific method of MarxismLeninism also makes a rigorous academic definition of politics and does not use the term in its everyday sense. Lenin, as we know, characterized politics as being, above all, the sphere in which all social classes work out their mutual relations with regard to state power. Moreover, he understood class

Veniamin

Chirkin

163

from a historical-materialist point of view, unlike many bourgeois academics who use the concept in a variety of ways and sometimes enumerate more than a dozen classes within society. In his Collected Works (vol. 5), Lenin defined the political system as "the sphere of relationships of all classes and strata to the state and the government." Politics, emphasized Lenin, is "participation in state affairs, in the direction of the state, and in determining the forms, tasks and contents of state activities" (Complete Works, vol. 33). Thus, from the Marxist point of view, politics and political relations are not linked to the differences in interests among members of a trade union or a tribe, or among the employees of a particular firm. They are, instead, bound up with the class and social strata divisions of society. Consequently, they are linked to the political power wielded by one or another class (or alliance of classes and social strata), and which is exercised, first and foremost, in the form of state power. The first institution of the political system, historically speaking, was the state. It arose with the appearance of private property and the splitting of society into antagonistic classes. Several other processes were taking place as the state emerged. Political norms, including legal noims, were established; and a certain political regime took the place of the former apolitical order of the primitive-communal system. The dominant and the repressed political ideologies took shape and both a political culture and a political process appeared in society. By political culture, we mean the totality of political conceptions and models of political behavior. The development of the political system proceeds through differentiation, growing complexity, and the appearance of new levels. Thus, in slave-owning societies we find the institution of oriental despotism; in feudal societies, the guild organization; and in bourgeois societies, the political party. The formation of trade unions and of workers' parties provided the means for working people to oppose the oppression of the exploiters, and they exist alongside the mechanisms for exercising bourgeois domination (e.g., bourgeois parties, employers' associations, bourgeois law and political ideology). The system's striving toward totality, wrote Marx (1974), is expressed in its efforts "to subordinate all elements of society to itself, or to create from them the agcncies that it itself still lacks. Thus in the course of historical evolution the system is transformed into a totality." The political and socioeconomic organizations of classes (i.e., parties, trade unions, peasant unions, and others) speak for large social entities and represent their interests. Varying types of relations arise among these entities and consequently among their representative organizations, both in form (i.e., cooperation, conflict, or neutrality) and in character (e.g., economic, ideological). Political relations are always linked in one way or another with issues of state power. When the question of who is in power has been resolved, and the state apparatus has been mastered, then the way is open for the organization and reorganization of society in accordance with the interests

164

The Comparative

Study

of

Third

World

Political

Systems

and political ideology of a particular class. Political relations, as was noted earlier, may take different forms. However, to establish the concept of a political system, it is important to appreciate the difference between primary and secondary political relations. Classes, nations, and certain other social groups act as the primary participants in politics and political relations. Issues relating to the power of one class or another, and to their influence over the state and participation in state affairs, form the basis of these relations. Then, there arise secondary "reflected" political relations among the organizations representing these social entities. It is precisely the latter that are important in establishing the concept of the political system. They are derivative and thus, in the final analysis, also concentrate on issues of power. They are the links between different levels in the political system that, during cooperation or conflict, support the continued existence of the system itself. Of course, links in the political system of an exploiting society differ in substance from those in socialist society, especially after the elimination of antagonism between classes has excluded class conflict from the political process. In the same way, these links are different in countries of socialist orientation and in states following a capitalist path of development. However, issues of state power are always crucial for any political system, whether we are referring to the power struggle, participation in that struggle, or the degree to which it takes the interests of the various classes and strata into account. In principle, the components or basic constitutional elements of such a system are always the same. In different countries, however, they may be combined in different ways and their levels may vary in number. The institutional component includes the state, together with its entire apparatus, political parties, and the mass public organizations that participate in political life. The normative component embraces the political norms, and particularly law in its political role as the regulator of social relations. By the functional component, we mean the political regime that takes shape as a result of the activities of all the system's levels. The ideological component covers political ideology. There is a particular communications element that covers political relations and ties and is an inseparable part of the system. Without it, the system could not exist. Certain authors distinguish the press, political culture, the political process, and other levels as independent elements of the system. We consider that they are subsumed under the above-mentioned components. By analyzing these large elements of the system, we gain an idea of the distinctive features of the political system in one or another country of the "Third World." Making allowance for the social character of the system, such an analysis forms the basis for a generalized classification within the framework of the major divisions between one or another type of system. Thus, in our study of developing countries, we begin with the following

Veniamin

Chirkin

165

propositions: First, the political system is the universal governing system in class society; second, the system's components (i.e., institutional, normative, functional, and ideological) are united by "secondary" political relations into a dialectically contradictory, but total, formation. In the last analysis, it is the political system that regulates production and the distribution of social benefits among the large social entities through their access to state power.

The Classification of the Political Systems of Developing Countries Western political science, for the most part, adopts a global approach in its analysis and classification of developing countries' political systems, which relies on the dichotomy between the "traditional" and the "modern." Yet, within this same approach, different methodologies are used (e.g., behaviorism, institutionalism, and normativism), and different research orientations. Although it was short-lived, the first dominant concept was that of decolonization, which focused on only one group of developing countries: those that had become free of direct colonial dependence. It concentrated attention on the introduction of the new institutions that marked the country's acquisition of political independence. These countries were usually compared, moreover, only with the former metropolis. Supporters of this approach (the English jurist Jennings and others) were distinguished by their normativism. Of longer duration was the concept of political modernization. This dominant view was developed by many authors, particularly in the United States, but perhaps its best-known expression can be found in Apter's Politics of Modernization (1967). Supporters of this approach began with the assertion that in all the developing countries the political system was "predemocratic." It would be transformed into a democratic system, they stated, by imitating the models of "industrial society" of which, in their view, the United States represented the ideal. Experience, however, did not confirm the concept of political modernization. As they developed, many countries set up military regimes or resorted to archaic institutions. Many political scientists, as a consequence, have now abandoned the concept of political modernization. Yet, they do not exclude the possibility that as the economy and culture develop and capitalist relations are extended, there will be political modernization of bourgeois models in the future. The proimperialist essence of this concept is understood in the developing world. At the thirteenth Congress of the International Political Science Association in 1985, the Indian researcher Varma (1985) commented that the idea of Western modernization was a means of drawing the "Third World" into the world capitalist orbit and was not influential in developing countries.

166

The Comparative

Study

of

Third World

Political

Systems

In certain countries today, especially Moslem countries, traditional ideas and appeals for a return to the political institutions of the distant past are being used, and leftist-radical views (which bourgeois researchers term "neoMarxist") are being developed. However, the most widespread concept now is that of "development." U.S. author Huntington's Political Order in Changing Societies was one of the first major works to start elaborating this concept. There are today a great many liberal and authoritarian versions of "development theory." The latter hold that democracy should be sacrificed for the sake of economic development: It is not important what kind of government a country has, so long as it is a stable one. Moreover, certain supporters of this concept consider that it is not obligatory to repeat the path followed by the West in its development. Proponents of each of the approaches mentioned above have all offered their classification of the political systems of developing countries. Up to the present, the most famous of these are the classifications proposed by Shils (1968) and Apter (1970). Shils identified and distinguished the following: "political democracies"; "tutelary" democracies; modernizing oligarchies; totalitarian oligarchies; and traditional oligarchies. Apter first considered that there were four types of political system in the independent developing countries, which he labeled mobilizing systems, modernizing autocracies, military oligarchies, and neomercantile systems. Subsequently, Apter extended this list to include bureaucratic, theocratic, and reconciliatory systems. All these classifications suffer from a fundamental deficiency: They all deflect attention from the most important aspects of such a system—its social character and the sociopolitical orientation in a given country's development. In my view, a fundamental classification should correspond to the basic historical division of societies into socioeconomic formations. We can then distinguish the following types: There are a handful of developing countries wherein there is still a feudal type of political system, which is gradually being transformed into a bourgeois type (e.g., Saudi Arabia, Bhutan); there are bourgeois types of political system at various stages of development (e.g., in Tunisia, Brazil); there are also transitional types of political system: in countries of socialist orientation (e.g., the People's Republic of Yemen, Ethiopia, and Nicaragua) moving toward the socialist model, and in the underdeveloped countries of capitalist orientation in tropical Africa and Oceania (e.g., Niger, Tonga) moving toward the bourgeois model. It is possible to make further distinctions within the framework of this typology according to social indicators. Thus, among the bourgeois political systems, we may distinguish, in particular, between capitalist-landowner systems in which capitalists are dominant (e.g., in Pakistan) and in which the landowner element has the upper hand (e.g., in Malaysia). Among countries of socialist orientation, we can distinguish between the national-democratic (e.g., in Algeria, Madagascar) and the people's democratic systems that have not yet

Veniamin

Chirkin

167

taken shape but are maturing (e.g., in Angola, People's Republic of Yemen). Further classification of political systems into "socially defined subcategories" depends on an assessment of the role played by different components, their methods of operation, and their interaction and correlation. In a generalized form, this involves three aspects: (1) the degree to which the system has been democratized (more specifically, the level of participation by its citizens in the administration of society and the state); (2) the structure of the political system, which must be considered; and (3) the way in which the political system functions. Let us take, for example, a national-democratic system of socialist orientation. It can be based either on a civilian regime, relying on civil agencies, norms, and ideology (as in Algeria since 1977), or on a military regime (as in Algeria between 1965 and 1977). These two varieties can also be distinguished in countries of capitalist orientation (e.g., the civilian regime in Kenya and Niger's pre-1989 military regime). It is important, however, to underline their social denomination: They are systems of a specifically capitalist orientation. All political systems of an exploiting type among developing countries (semifeudal, bourgeois, or of capitalist orientation) can be classified according to three types. In the liberal type, the basic structural principle is functional: There is pluralism and role autonomy. It is a multiparty system that recognizes the separation of powers, and several decisionmaking centers are functioning. The principle of formal equality among its citizens is proclaimed, and there are basic elements of democracy. As a whole, the system works in the interests of the exploiting classes and strata of the population. Nevertheless, the working people can obtain certain concessions from the ruling class by exerting pressure on the state (e.g., India). There is no denial of participation, pluralism, and role autonomy in the autocratic system; working people are, however, restricted or in certain cases excluded. These principles are extended only to a section of the political institutions. The activities of particular political parties and public organizations are permitted, though there may be several of them. In this system, also, formal participation takes place. In the Philippines, for example, referenda by commune were frequently held, before 1986, in which all citizens aged fifteen and over took part. Yet, in practice, the forces controlling the state are based on the principle of political guardianship. It is very difficult for the working class or the peasantry to gain any concessions from the exploiting state in this system because the leading organizations of working people (i.e., the Communist and Workers' parties) are either banned or subject to special restrictions. The almost closed nature of the system also adds to these difficulties. Such pressure on the state is not excluded, nevertheless, though the ruling class responds to it with harsh reprisals (e.g., in Indonesia). Finally, in the totalitarian system, even limited participation is abolished.

168

The Comparative

Study of

Third World

Political

Systems

Pluralism in the levels of the system is eliminated, though there remains a formal recognition of the difference between party and state institutions, and between the law and political norms outside the law. There is no role autonomy. There may be a single bourgeois or probourgeois party that is usually a party-movement and sometimes (e.g., in Zaire) may include all the adult population. Or, as in Bahrain, all parties may be forbidden. In either case, the political system forms a unitary centralized system controlled either by a solitary leader (e.g., in Ruanda) or by a ruling family (e.g., in Saudi Arabia). Political parties of the working people are either forbidden or considered to be outside the system, while other social organizations are integrated into the party-state structure. This is a closed system. Mass political activities or speeches by workers are forbidden; therefore, they cannot practically obtain concessions from the ruling class by putting pressure on the state. It is only by overthrowing or deeply transforming the totalitarian system that the working people in the population can improve their position. It is not the modernization, but the elimination of such a system, that meets the needs of the people.

References Apter, David E. 1965. The politics of modernization. Chicago: University of Chicago Press. -. 1970. "Political Systems and Developmental Change." In The Methodology of Comparative Research, eds. Robert T. Holt and John E. Turner. New York: Free Press. Bottomore, Tom 1979. Political sociology. London: Hutchinson. Calvert, P. 1983. Politics, power and revolution: An introduction to comparative politics. Brighton: Wheatsheaf Books. Coulter, E. M. 1981. Principles of politics and government. Boston: Allyn & Bacon. Dahl, Robert A. 1963. Modern political analysis. Englewood Cliffs, NJ: Prentice Hall. Dcbbasch, Charles, and I. Daudet. 1974. Lexique des termes politiques. Paris: Presses Universitaires de France. Dcbbasch, Charles, et al. 1983. Droit contitutionnel et institutions politiques. Paris: Economica. D e L e s p i n a s s e , P. 1981. Thinking about politics: American government in associationalperspective. New York: Free Press. Duverger, Maurice 1970. Institutions politiques et droit constitutionnel. Paris: Presses Universitaire de France. Entin, L. M., and M. L. Entin. 1986. Political science of development and the developing countries: A critique of non-Marxist conceptions (in Russian). Moscow: Science. Houriou, A. 1974. Droit constitutionnel et institutions politiques. Paris: Editions Monte hretien. Huntington, Samuel P. 1968. Political order in changing societies. New Haven, CT: Yale University Press. Kourevartis, G. A., and B. A. Dobrats. 1980. Society and politics: An overview and reappraisal of political sociology. Dubuque, IA: Kendall Hunt. Lavroff, D. G. 1975. Le système politique français. Paris: La Vie Republique.

Veniamln Chirkin

169

Leftwich, Adrian. 1983. Redefining politics: People, resources and power. London: Methuen. Lenin, Vladimir I. 1960-1970. 45 vols. Collected works. Vol. 5. Moscow: Foreign Languages Publishing House. . Collected works. Vol. 19. Moscow: Foreign Languages Publishing House. . Complete works. Vol. 33 (in Russian). Moscow: Politizdat. Marx, Karl 1974. Grundrisse der Kritik der politischen Ökonomie (Rohentwurf): 1857-1858. Berlin: Dietz. Politics in transitional societies: The challenge of change in Asia, Africa and Latin America. 1968. New York: Appleton-Century-Crofts. Ponton, G., and P. Gill. 1982. Introduction to politics. Oxford: Robertson. Roberts, G. K. 1971. A dictionary ofpolitical analysis. London: Longman. Shils, Edward. 1968. Political development in the new states. Paris: Mouton. Varma, S. P. 1985. History, tradition and culture in political development. Paper presented at the World Congress of the International Political Science Association, Paris, July 15-20. Von Beyme, K. 1980. Die politischen Theorien der Yegenwart. Munich: R. Piper.

19 Commentary Iqbal Narain Chandrakala Padia

We find lot of sense in some of Chirkin's critical emphases. He rightly insists that Western political institutions have so far served as a mere "external shell" to the political life of the Third World countries; that, in some of these countries, socialist ideas are being utilized more for influencing the masses than for solving their problems; and that, by and large, political scientists from the West have ignored the distinctive social character of political institutions of the countries in question. He has farther noted: 1. Political systems of developing countries are distinguished by great variety and, therefore, one has to be careful in devising "a well-thought-out methodology" for their study. 2. In this respect, bourgeois thinking has failed for two important reasons: First, it misses the class-dominated character of political relations of the Third World countries; second, it commits the mistake of assuming that their political relations are determined essentially by the individual's own nature and his cultural heritage rather than by the class structure of society. 3. Positively the only methodology that can be of help to us here is that which bases itself on dialectical materialism and the science of MarxismLeninism. All may not agree with Chirkin's formulations. There could also be several arguments on the other side, and some of these may be mentioned here. First, it does not seem proper to seek to evolve a common methodology for the study of all the developing societies. Each of them has its own culture, social history, and traditions and should, therefore, be expected and encouraged to work out its own individual methodology to capture its own sociocultural character. It would be dogmatic to start with a prefixed notion of a viable methodology and to apply it to every developing country. Such notions do not yield the desired results; and they soon come to appear irrelevant because of the cultural uniqueness of the country to which they are applied. We must realize that a methodology is not like a ready-made

170

Commentaryllqbal

Narain

and Chandrakala

Padia

171

outsized hat that can be worn on anyone's head. It is, in principle, a living plan that must never cease to respond sensitively to the specific character of the peoples for whom it is meant. We are not against universalism in knowledge. We are not constrained, however, to begin and end with a general methodology; we may move from the specific to the general. Second, Chirkin complains that bourgeois political scientists ignore the class-dominated character of political relations and the social nature of political institutions in Third World countries. There have been several studies with this Marxist approach. We could also argue that present-day "social nature" is not the only determinant of a political system. The wider forces of history, culture, and tradition must also be taken into account. It is common knowledge that such forces often influence the outlook of the educated class; and that, through the inevitable dominance of this class, they shape, not only the established customs, but the growing aspirations of a society. Scholars of Marxist and Leninist persuasion do not always appreciate the nuances of this interaction. They have always treated culture as merely the product of particular forms of society rather than as, also, a determinant. Third, the Marxist theory of dialectical materialism looks upon the mode of production alone as the chief determinant of radical changes in a civilization. It holds that history moves through the struggles of divergent social relations that arise because of the forces of production. Thus, Marx overemphasizes the role of objective historical forces in social change and minimizes the value of the part played by the individual's possible selfaffirmation against the forces of history. In what precise manner and to what extent the other factors determine our being and behavior depends, at least in part, on how and how far we let them influence us. Further, the individual not only assimilates influences discriminatingly but can radiate them for the social good. It is indeed the evidence of history that men of exceptional courage and wisdom have often transformed the face of society. Social progress arises from a clash between the lofty vision of some individuals and the actual state of things. For example: [WJho, before the socialist, had thought of class as a unit of repression? How many, before Freud, had sensed that children needed to be protected against their own parents? How many believed before Gandhi's rebirth, after the environmental crisis in the world, that modern technology, the supposed liberator of man, had become his most powerful oppressor [Nandy, 1987, 22]?

Fourth, class character (which is emphasized by dialectical materialism) is not by itself able to explain the divisions of caste and fragmentary village systems; nor can it explain why, in spite of the prophecy of Marx, neither a world bourgeois class nor a world proletariat class has emerged. Instead, what strikes the eye today is the emergence of a new and powerful class of educators, bureaucrats, and working people in countries of the Third World.

172

The Comparative

Study of

Third World

Political

Systems

Fifth, as far as the methodology that finds favor with Chirkin is concerned, we think it is suitable only for the study of industrialized societies, and not for countries with a predominantly agricultural economy. In the case of the agricultural societies, the problem is to balance indigenous social organizations, based on the norms of hierarchy, with the new legal and political institutions emerging in the wake of industrialization or, better still, in the context of efforts at social justice and equality. Marx never attempted a detailed analysis of the agricultural society of India (or of China). He merely hoped that British rule would sweep away "divisions of caste and village sentiments"; and that the undivided masses would then be able to generate their own political effectiveness for social change on the lines indicated by the history of industrialized societies. But, as is well known, the growth of the Indian political society has belied this hope. Caste and village system remained intact as British rule tottered; and yet, adequate political pressure could be generated for the attainment of independence. In fact, caste is still alive. Even in Kerala, caste and religion cannot be ignored altogether in the various class analyses that have so far been attempted. China's rural orientation to Marxism is too well known to need explication here. Finally, the present phase of economic liberalism and export-oriented economy, in general, and of glasnost and perestroika, in particular, may have implications for the Marxist-Leninist framework both in substantive and methodological terms. We are, however, too near the phenomenon to spell over the details. We may, however, recall in this context the observation made by Mikhail Gorbachev (1987, 21-22): In the social sciences scholastic theorization was encouraged and developed, but c r e a t i v e thinking was driven out from the s o c i a l s c i e n c e s and superfluous and voluntarist assessments and judgments were declared indisputable truths. Scientific, theoretical and other discussions, which are indispensable for the development of thought and for creative endeavor, were emasculated. Similar negative tendencies also affected culture, the arts and journalism, as well as the teaching process and medicine, where mediocrity, formalism and loud eulogizing surfaced too.

References Gorbachev, Mikhail. 1987. Perestroika: New thinking for our country and the world. London: Collins. Nandy, Ashis. 1987. Traditions, tyrannies and Utopias. Delhi, India: Oxford University Press.

9

Typologies of Third World Political Systems Dirk Berg-Schlosser Despite the flood of literature on a wide variety of Third World topics that has appeared in recent years, relatively few systematic, comprehensive, and longitudinal analyses of the performance of Third World political systems have been undertaken thus fair. Of these analyses, most have either tended to focus on a single type of rule (e.g., by the military) (Nordlinger 1970; Jackman 1976) or to employ rather crude distinctions between different types of political systems such as civil versus military systems (McKinlay & Cohan 1975) or democratic versus authoritarian systems (Dick 1974). In addition, the variables considered by these studies often have been based entirely upon narrow indicators of economic development, such as the GNP per capita growth rate, or have demonstrated other serious limitations with regard to scope and time. This lack of adequate conceptualization and measurement is one reason why rash generalizations as to the incompatibility of rapid development and democratic forms of government still abound (Heilbroner 1963, 132ff; Bhagwati 1966, 203f; Emerson 1971, 239). Lowenthal (1963, 187) once formulated this alleged dilemma quite unequivocally: "Every degree of freedom must be paid for by a retardation of development and every degree of acceleration with a loss of freedom. This is inevitable in the nature of the process." In contrast to such approaches, this chapter begins by examining some of the problems involved in adequately categorizing contemporary political systems in the Third World. After proposing a more differentiated, historically and regionally oriented approach on a medium level of abstraction, it proceeds to discuss various measures of the performance of distinct political system types. In this context, some of the shortcomings exhibited by presently employed measures are critically examined, and at least some supplementary criteria are proposed. This leads to a more comprehensive analysis of Third World political systems, and, in my view, to a more balanced assessment of their performance. The quantitative findings in this chapter are based on data taken from the World Handbook of Political and Social Indicators (Taylor & Jodice 1983) and from similar 173

174

Typologies

of Third

World

sources that recently have become available and that now cover a period of up to two decades for all Third World regions.

Classification of Political System Types There are basically two kinds of typologies of political systems: the constructed and the e n u m e r a t i v e ( H e m p e l 1952; von B e y m e 1980). Constructed typologies define a number of variables on a theoretical basis and then combine their distinct expressions into a comprehensive scheme. A well-known example of this is Aristotle's classification based on the number of rulers and their orientation toward their own self-interest or the common good. More recent typologies of this type, explicitly applied to Third World countries, include those by Apter (1965, 22ff) and Huntington (1968, 80f) as well as similar two-by-two, two-by-three (and so forth) tables. The basic problem with this approach lies in the fact that if too few variables and their discrete expressions are considered, the number of types arrived at is too limited and their d e f i n i t i o n too artificial to cover e m p i r i c a l reality satisfactorily. Thus, they tend to become intellectual straitjackets rather than heuristic tools of analysis. If, on the other hand, a larger number of variables is considered, the respective classifications soon tend to become innumerable and confusing in a multidimensional space. 1 Eckstein and Gurr (1974) also discuss a comprehensive scheme of dimensions for political indicators. They, however, refrain from a more comprehensive attempt at classifying political system types. Such attempts, more often than not, result in excessive neologisms of little practical value (Riggs 1970). The other extreme consists of listing types of political systems in a purely enumerative manner without a common and consistently applied t h e o r e t i c a l f r a m e of r e f e r e n c e . F r i e d r i c h ' s (1970) thirteen types of g o v e r n m e n t and, in a Third World c o n t e x t , S h i l s ' ( 1 9 5 8 ) f i v e f o l d classification are examples of this kind of typology. Because the elements of comparison and the potentially indefinite number of system types remain insufficiently specified, their application proves both theoretically and empirically unsatisfactory. The functional solution to this taxonomic puzzle (i.e., to consider universal processes of political systems, rather than concrete structures) does not help any further. Both systems theory, as most prominently developed by Easton (1965) and Almond and Powell (1966), and Marxist derivations of state theory (Altvater 1972; Mandel 1972) 2 remain at too high a level of abstraction to satisfy more concrete empirical needs. They cannot, in short, provide "fact-finding categories that own sufficient discriminating power" (Sartori 1970,1039). An even less satisfactory alternative is the selection of a single dimension for the puipose of cross-national political comparison. Examples

Dirk Berg-Schlosser

175

of this type of approach are the index of "political development" (Cutright 1963) or a general "index of political rights," such as employed in Freedom House publications (e.g., Gastil 1978-). Apart from the difficulty of operationalizing such indices satisfactorily, as well as certain apparent coding biases, distinctive qualitative differences between political systems become far too blurred to enable meaningful comparisons between different kinds of political systems and evaluations of their performance over time. There is no easy way out of this conceptual dilemma. Like Sartori (1970, 1053), I prefer to develop the discipline along a medium level of abstraction with better intermediate categories, and to maneuver both upwards and downwards, along a ladder of abstraction in such a way as to bring together assimilation and differentiation, a relatively precise descriptive content, macrotheory and empirical testing.

More concretely, I propose to look for certain characteristic clusters of properties within a multidimensional systems framework of variables. I have selected for this purpose a number of characteristics that compose some of the most important dimensions of political systems. 3 Among these characteristics are the party system, the basis of legitimacy, the pattern of recruitment of the head of the executive, the actual power structure, the scope of political control, the ideological orientation of the system, and the formal vertical and horizontal separations of power. Each variable has then been subdivided by what appeared to be the most relevant discrete expressions for the purpose at hand. Thus, party systems were distinguished according to their degree of competitiveness, the scale ranging from one-party states without any competitive elements through semicompetitive single party structures and restricted multiparty systems to fully competitive systems. 4 Legitimacy, in the tradition of Weber (1922) and Easton (1965), was considered to be based predominantly upon strong emotional attachment to a political leader, upon formal elective procedures (i.e., "legal-rational" acts in the Weberian sense), or upon a customary or similarly broad informal acceptance of political incumbents, or of a certain regime. This last category occurs in monarchies, f o r example, as well as in a n u m b e r of civilauthoritarian states, 5 and possibly in even those more broadly accepted socialist systems that are devoid of formalized competitive procedures. The recruitment of the head of the executive can be achieved by means of inheritance, direct election by the population at large, a majority vote' by parliament or a similar body, or through forceful usurpation. The actual power structure within the executive can be strongly personalized or clientclistic and have a small collective group as its base (e.g., a military junta or a politburo of a centralized party), or it can be more broadly composed to incorporate the most important plural elements of the society in

cW

M

E

a(A

o Ph o w o

£

u 3« H

¡vet» «J*

Ye*5'

1 Sod»AW'

I

I4 S3 f U X g

X g

X

iVet» X

tet» a*

jiet» .«i*

X

Co*'

X

i V,

e £

fr X

g

X X

X

iVc° s®**1

X

g g

X

X

g

Head of executive

ö MH o c .0 « c IS E a

X X

X

X

X

X

tcM

«1*A***

X

S

President of Prime Minister

Monarch

Military ruler

elective

personaKstic

customary or other broad aceptance

weak or no

multi-party

one-party with com pet. elements or restricted multi-party

of Third

Dominant basis of legitimacy

1« » Typologies

one-paity without compet elements

no-party

176 World

X

oW X

,.. o W ..tie'»'

S» S X

X g X

g g X

Coi X X

X X

X

X

X

X

X

X

X

X

X

Vertical separation of power

Horizontal separation of power

X

g X

Ideological orientation towards

X

Scope of Power

Power Structure

g

X

X

g

X X

X

X

X

X X

X X

X

X

X

X X

X X g

X X X

X X X X

S X X X

X

X X

X X X

X X X

federal system

centralized system

full separation

limited sep. with independent judie.

no separation

totalitarian

partial

middle classes or broader social basis

lower classes

"new" upper classes

"old" upper classes

communalistic groups

broadly represent.

collective group

Personalistic

Dirk Berg-Schlosser

m

£

g

X

g

178

Typologies

of Third

World

question. The scope of control can range from limited to, at the extreme end, totalitarian. Ideological orientation was considered in relation to its respective social basis. This can consist of a "communalistic" (e.g., ethnic or religious) group (Melson & Wolpe 1970) or of the lower, middle, or upper classes of society, corresponding to the more common political continuum of "left," "center," and "right." Finally, in a more formal constitutional sense, the vertical separation of power can be viewed as complete, in accordance with Montesquieu's model; as limited but with an independent judiciary, such as in parliamentary democracies where the prime minister can generally count upon a reliable majority in parliament; or as wholly nonexistent. Similarly, the horizontal separation of power differs according to whether a system is of a federal or centralized nature. This list does not purport to be exhaustive, nor do I claim all variables to be of equal importance or that all their possible expressions have been covered. Moreover, a number of attributes listed as dichotomies often represent, in fact, mere variations in degree. Nonetheless, for reasons of simplicity and economy, I have selected only those attributes considered to be the most useful and sufficiently discriminating for a comparative analysis of contemporary Third World states. After I had established the above categories, two independent coders proceeded to score present-day political systems on the basis of available information. In a number of instances (e.g., with respect to the ideological orientation of certain regimes), proper placement was once again a matter of judgment. For other variables that often occur in mixed forms, such as the personalistic and customary types of legitimacy, only the apparently dominant mode was chosen. Eventual errors in fact or points of contention can, in my opinion, be rectified with little difficulty. 6 Instead of proceeding to list all possible or actual combinations of the variables chosen for this study, 7 I continued by identifying those systems that exhibited the highest number of relevant common traits. In this way, I arrived at ten basic clusters of system characteristics, each of which received a summary label. The actual combination of characteristics selected to designate a particular type is reproduced in Table 9.1. The most traditional type of political system in the Third World today, whose roots lie in noncolonial or precolonial periods, is the monarchy. Its basis of legitimacy is custom and is often linked to mythical or spiritual assumptions regarding the foundation of the respective dynasty. Examples of this type of system are Ethiopia (until 1974), Morocco, and Saudi Arabia. Accession to the throne is generally regulated by heredity, and there is little, if any, open political competition. Formal separation of power is virtually nonexistent, and the power structures are highly personalistic and are reinforced by the prevalence of an aristocratic upper-class ideology. Most

Dirk Berg-Schlosser

179

regimes of this kind were built upon feudal agrarian structures of one variant or another. Today, however, they tend to have highly centralized and absolutist characteristics. Under present-day conditions, these systems are strongly prone to revolution (Huntington 1968, 171ff), a situation that was dramatically illustrated by the downfall of the shah of Iran in 1979. Another type of system, which is prevalent in the Latin American context and has evolved more permanent traditions, is what we call, for reasons of simplicity, the "old" oligarchy or, to use Wiarda's (1984) term, the traditional-authoritarian regime. This type of regime receives its support from feudalistic or "neofeudalistic" rural structures and, in recent decades, from segments of the urban upper classes, together with their foreign counterparts. In the old oligarchy's more extreme forms (e.g., in certain Central American states such as Guatemala and Nicaragua until 1979), support is derived from a small number of leading families that exercises almost exclusive control over political and economic life (Sotelo 1973; Guzmin Boeckler & Herbert 1972). Within such a system, one leader may be replaced relatively easily by another, sometimes through holding sham or manipulated elections, that do not alter the actual power structure. Political leadership in these states is also closely connected with the more traditional elements of the Catholic Church and the upper ranks of the military. Political activities by other groups and by the media are usually tightly controlled by the regime's highly developed repressive apparatus. Finally, other elements that could constitute a certain separation of power, such as an independent judiciary or a federal political structure, are equally lacking. The "new" oligarchy, as distinguished from the old, receives its support from the dominant urban groups in a contemporary "peripheral-capitalist" context. In this authoritarian or "hegemonic" (Dahl 1971) type of regime, open and institutionalized competition for public office does not occur. Usually, there is a single-party structure; in most instances, however, it is predominantly a formal and not very effective organization. Instead, great emphasis is placed upon a relatively large bureaucracy and upon other output aspects of the regime. The expression of public opinion and especially the media tend to be tightly controlled, while leadership is strongly centralized and often highly personalistic. In this respect, some potentially charismatic elements of legitimacy may exist. While widespread forms of legitimacy are lacking, compliance is established either in terms of passive acceptance of the regime or by outright repressive measures. Both the social base and the "inclusiveness" of this system vary in ethnic and class terms. In a majority of cases, a relatively wide ethnic base is combined with more restricted class interests that tend to favor the prosperous groups in society. Countries such as Cameroon, the Ivory Coast, Tunisia, and the Philippines (under Marcos) are cases in point. "Socialist" regimes also reveal a distinct pattern. They are characterized

180

Typologies

of Third

World

by an effective single-party organization, a centralized system of government, and an ideology directed toward a more egalitarian social order and a "noncapitalist" and "self-reliant" road to development. As a rule, freedom of expression and pluralistic forms of organization are widely curtailed in socialist regimes. There is, however, a great deal of variation within this category (Rosberg & Friedland 1964; Rosberg & Callaghy 1979). One group of countries (e.g., Tanzania and Guinea in Africa, Burma in Asia) attempted to found their specific type of socialism upon what they perceived to be (or what actually were) their societies' egalitarian traditions and culture. Another group of states (e.g., Algeria, Mozambique) advocated a more Marxistoriented and "scientific" brand of socialism. In both groups, there may exist some semicompetitive elements in the intraparty and parliamentary spheres. Whatever the differences among socialist regimes, they must be distinguished from their totalitarian "communist" counterparts, if only because of their generally "underdeveloped" condition and the relative lack of effective social control they can exercise. Communist states attempt to encompass all relevant social groups within their own organizational structure (e.g., through state-regulated trade unions, and youth, women's, and sports organizations). Social life is, thus, tightly controlled by an allpervasive bureaucratic apparatus and by the "democratic-centralist" single party. Elections, if held at all, offer no choice among candidates. Thus, they generally produce the acclamatory result of nearly 100 percent participation (Domínguez 1978; Saich 1981). In addition, most communist states in the Third World (e.g., North Korea, Vietnam, Cuba) have very close links with the Soviet Union. A further type of system can be termed "semicompetitive" (Hermet et al. 1978). In Latin America, for example, this system is also based upon the traditionally dominant classes of the region. In such a regime there is a greater measure of balance and competition between the rural and urban elements of these classes (Tomasek 1966; Anderson 1967). This type also comes close to what Wiarda (1984) calls an "open corporatist" system. The contending elements are often institutionalized in "conservative" and "liberal" parties such as those found in nineteenth-century Chile, Colombia, Uruguay, and in some Central American states. The "dominant one-party" system in postrevolutionary Mexico is also a somewhat special case in point. Active political participation in these regimes generally remains restricted to the middle and upper segments of the population, while mass mobilization is largely prevented or restricted to more symbolic functions. In a somewhat different social context, one-party systems embodying competitive elements (e.g., that of Kenya) and states in which "regulated competition" prevails (e.g., Singapore, Lebanon until 1975) can also be subsumed under this category. Semicompetitive regimes tend to follow established constitutional rules within a presidential or parliamentary system of government. Regular transfers of power occur within the set framework.

Dirk

Berg-Schlosser

181

and governmental structures are generally output-oriented, though they are generally less repressive than in the case of "oligarchic" systems. The media often enjoy a greater measure of freedom as well. A final type of civilian political system that we found useful to identify is the "polyarchy." This, according to Dahl's (1971, 202) definition, is "a regime in which opportunities for public contestation are available to the great bulk of the population." This term appears more suited to our present purposes than does the more common designation "democracy," because the latter term implies a variety of generally more ambitious Western normative connotations. Polyarchy designates a system in which political leaders are responsive to the largest part of the citizenry by means of regular and fair e l e c t i o n s , and w h e r e t h e r e e x i s t f r e e d o m of p o l i t i c a l t h o u g h t and organization, possibilities of legal opposition, a certain separation of power with an independent judiciary, and the institutionalized guarantee of the rule of law. It also presupposes certain social-structural and political-structural conditions that, thus far, have been only broadly defined (Dahl 1971, 203) and that include a relatively egalitarian social structure and a m o r e participatory political culture. Polyarchic systems can be found in the Third World in Botswana, Costa Rica, and Sri Lanka, among other places. In addition to the above-mentioned regimes dominated by civilian institutions, there are other systems, which are more immediately controlled by "men on horseback" and "the barrels of their guns" (Janowitz 1964; Buttner 1976). Inevitably, these military rulers have come to power by means of a more or less violent coup d'état after the previous civilian institutions had been seen to have failed. In the absence of significant countervailing powers (e.g., strong unions) (Huntington 1968), the military's monopoly over the physical means of coercion lets the less efficient civilian governments become fairly easy prey for small armed groups. The social bases of such regimes are also usually rather narrow. Some military rulers act as temporary caretakers and make genuine attempts to return their countries to civilian rule (e.g., to an elected polyarchic form of government). Others, on the contrary, seek to establish their power permanently. Among the military rulers who seek permanent power, I found it useful to distinguish among at least three distinct subtypes. The first has a strongly personalistic nature and includes some of the most despotic rulers of the Third World. Regimes of this kind center around a "strong man" and his most immediate following. Because there are few, if any, formalized input structures, these regimes rely quite heavily upon their highly centralized output apparatus. In many Latin American countries, this type of rule was exercised with particular forcefulness by the characteristic caudillo after the gaining of independence during the first half of the nineteenth century. Although the caudillo's ascent to power was largely due to his personal qualities (i.e., his military prowess and perhaps also a kind of charismatic appeal), his rule cannot be fully understood without reference to his links

182

Typologies

of Third

World

with the established landed oligarchy. The large haciendas remained one of the more stable elements in turbulent times, and caudillos often were (or soon became) hacendados as well. In cases where a caudillo managed to establish a more durable regime in a larger geographical area, this was o f t e n accomplished through a system of regional and local subpatrons, the caciques. However, since this type of rule was exceptionally personalistic, it remained inherently unstable. Caudillos were often overthrown by more successful rivals, or their system of caciquismo disintegrated following their death. This kind of political system, then, had to be rebuilt by a new caudillo, or it gave way, in the course of time, to other more durable political structures (see e.g., Anderson 1967; Gilmore 1966; Wolf 1966). During the past two decades, some Latin American states, such as Brazil and Argentina, have witnessed the emergence of a more "modern" and institutionalized type of military regime (Schmitter 1973; Collier 1979). In these countries, p o w e r has been assumed by the military on a more p e r m a n e n t , corporate basis, and not solely at the behest of a single personality. Within the leading ranks of the various segments of the aimed forces, a certain institutionalized transfer of power has been established. Although the political orientation of these regimes is definitely "national" and favorable to "modernization," they nonetheless have left older social structures essentially intact. Formal interest groups and parties are strongly regulated, if they exist at all, and the repressive nature of the regime is often particularly blatant. A third type of military regime attempts to establish its authority on a more permanent basis through the creation of a broadly organized singleparty system. In contrast to the other forms of military rule, which primarily cater either to the personal enrichment of the rulers or to the corporate i n v e s t s of the military together with their largely upper-class allies, these regimes exhibit a predominantly socialist and lower-class orientation (Tibi 1973; Stepan 1971). In such cases (e.g., Nasser's Egypt, Iraq and Peru after 1968), the power of the military was applied to the realization of significant and often overdue social reforms. The long-term success of such reforms depends, however, upon the government's ability to secure a broad measure of political participation from below. If these efforts fail, a shift to a semicompctitive type of regime (e.g., that of Sadat's Egypt), a polyarchic system (e.g., in Peru after 1980), or a return to a more unstable form of pcrsonalistic military rule may take place. In some instances, the boundaries among the various types of systems described here may become fluid. This is particularly the case with respect to long-term, institutionalized military rule. It is also true, however, of the gradual distinctions among more inclusive new oligarchies, semicompctitive, and polyarchic systems. For other purposes, further subtypes can be distinguished (e.g., certain constitutional variations of polyarchic forms of government). In view of the limited number of cases available, however, I

Dirk

Berg-Schlosser

183

was satisfied with the broader categorization presented so far. These system types must be analyzed not only at a given point in time, but also with regard to their dynamic aspects. For this purpose, I have selected 1960—a major watershed in the process of decolonialization, particularly in Africa—as the point of departure. The length of the period to be analyzed was determined by the availability of sufficient comparative data. Since I had to draw most heavily upon the data collected in available handbooks and by organizations such as the World Bank, 1980 was chosen as the termination point. Nevertheless, I feel the period under consideration is long enough and sufficiently independent of purely conjunctural changes to permit an evaluation of the effects that the political systems considered here have had on various major dependent variables. Of the 117 cases originally considered, fourteen had to be discarded because the countries in question had become independent only during the 1970s. Among the stable systems of this period, 8 twelve were monarchic, five old oligarchic, ten new oligarchic, eight socialist, five communist, seven semicompetitive, fifteen polyarchic, five personalistic-military, one corporate-military,9 and four socialist-military. The remaining three countries experienced one or more "irregular transfers of power" as a result of military interventions during this period and were classified here as "praetorian." 10 Information on these regimes is based on the "political events" data collected by Taylor and Jodice as well as on several other historical sources. For the period under consideration, it was not possible, in all instances, to subsume the empirical data exactly under the categories of our system typology. This problem arose, for example, with respect to Egypt during the Sadat era, when the system changed subtly from a basically military socialist to a more semicompetitive type. In such cases, the classification was made with respect to the dominant aspect of the system during the period 1960-1980.

System Analysis and Evaluation In addition to the question of a regime's mere durability, its actual performance (in terms of the outcomes of the policies it proclaims and executes) must also be examined systematically (Eckstein 1971). An evaluation will be undertaken on two principal levels for the cases considered here. The first level is the goal that all developing countries, at least officially, seek to achieve today: namely, internal social and economic development in a material sense. The second level represents the fulfilment of certain normative criteria that are considered to be characteristic of a good political order. This is, of course, not a comprehensive list of potential criteria, which would necessarily have to include such subjective factors as the immaterial satisfaction provided by a regime (Campbell et al. 1976), and the

184

Typologies

of Third

World

"reinforcement of supports" (Easton 1965). Our criteria must also be operationalized within the limits of the comparative and longitudinal data available. "Development" is one of the most elusive notions in the social sciences. In its multidimensionality and with its manifold normative and political implications, it defies a simple and straightforward definition. 11 For our present purposes, however, I will restrict the use of the term to its purely material and economic aspects. But even in this regard, I am far from general agreement on its most appropriate indicators (e.g., UNRISD 1969; Ward 1980). Thus, we must once again make a choice of relevant criteria, both on theoretical and practical grounds. The most common single indicator of economic development remains the GNP per capita growth rate. In spite of the many shortcomings of this purely monetary indicator (e.g., incomparability of real purchasing power and its international incomparability), and despite its inadequate coverage of certain important factors such as, for example subsistence income, the satisfaction of real needs, social costs, aspects of income distribution, and external drains, we do not yet possess another readily available indicator that could replace it in a generally more satisfactory manner. I will continue to use the GNP per capita growth rate here, even though there is a certain amount of "growth without development" (Amin 1973).12 This indicator will be supplemented, however, by another measure covering the direct improvement in the material well-being of a larger number of persons: the physical quality of life index (PQLI) developed by the Overseas Development Council. This indicator seeks to provide a "composite measure of international social economic performance" that "sums the complex interrelationships among policies that affect infant mortality, life expectancy and literacy rates" (Morris 1979).13 Even though a number of problems still remain, I am satisfied inasmuch as this measure covers very basic aspects of development under Third World conditions. It measures results instead of inputs; it is sensitive to distribution effects; and it lends itself, in a relatively simple and straightforward manner, to international comparisons on the basis of available data. For purposes of longitudinal comparison, the computation of a disparity reduction rate (DRR) 14 for the PQLI at different points in time also avoids simplistic assumptions of unilinear forms of development (Grant 1978). The results of this twofold measurement of development perfoimance, shown in Table 9.2, reveal interesting differences among the types of political systems we have been considering. These findings are remarkable as regards both the general discussion of various indices of development and the respective capabilities of different types of political systems. On the one hand, they demonstrate that large discrepancies between the two measures employed in this study exist. (The overall product-moment correlation is < 0.466, p < 0.001). Even more important, however, are the clear distinctions

Dirk Berg-Schlosser

185

between these two measures with respect to the types of political system under consideration. T h e c o r p o r a t e - m i l i t a r y regimes, the stable m o n a r c h i e s , and the communist systems have, by far, the highest GNP per capita growth rates. They are followed by the socialist-military, semicompetitive, and new oligarchic systems. At first glance, therefore, Lowenthal's statement appears confirmed: Countries with stable authoritarian regimes have the highest rate of economic growth. Socialist systems, the unstable praetorian states, and the more static Latin American oligarchies rank, on the contrary, distinctly below the mean for all countries. A look at the regional distribution, however, shows that the high scores achieved by the monarchies are due mainly to the special situation enjoyed by the oil-exporting countries of the Middle East. The data indicating changes in the "physical quality of life" present a somewhat different picture. Here, the communist states achieve the best scores, while the praetorian, new oligarchic, monarchic, and personalisticmilitary systems all rank below the mean. The results for the polyarchic and, to a somewhat lesser extent, semicompetitive systems remain remarkable, h o w e v e r , as they achieve quite s a t i s f a c t o r y scores on both indices. Lowenthal's assertion might thus be modified as follows: Authoritarian systems produce higher level of "development"—right-wing regimes predominantly in terms of GNP growth (but with a fairly unequal measure of distribution); left-wing regimes mainly with respect to an improved physical quality of life for a larger portion of the population. As will be shown below, however, both achievements have high normative costs. Of particular importance is the discovery that the reversel of Lowenthal's statement cannot be c o n f i r m e d . T h e positive scores achieved by the more pluralistic semicompetitive and polyarchic systems on both accounts do not justify premature pessimism on grounds of economic development alone with regard to the chances of democratic government in the Third World. Some of the causal factors relating to this differentiated pattern of development come to light in an analysis of the allocation of public resources by the various system types, shown in Table 9.3. On the one hand, the share of government spending in the gross domestic product (GDP) is noticeably high for the socialist and communist states, though the available data for the latter are extremely scanty, as it is for the corporate- and socialist-military systems. On the other hand, the more traditional Latin American oligarchies and the semicompetitive states exhibit the lowest values in this respect. The same is true of the monarchies if the special situation of the wealthy oilproducing states in the Middle East is taken into account. The use of public funds differs widely, though, from system to system. Expenditure for (unproductive) military purposes is very high in both the communist and socialist military systems. The level of spending f o r educational and health purposes is lowest in the corporate military, the old oligarchic, the semicompetitive, and the personalistic-military systems.

186

Typologies

of Third

« O O H N O « © —' o ^ ö o o

World

osor-'oosoavr0 — 0 0 0 0 0 0

O O oo r- ci o — —< r4 —i —^ —Î ts

V) O O O Tj- O 0 0 0 0 0 0

O JS S s

©^oeooor-*©ro h o o m ^ O -

N > O O V N ^ 00 dcs^OTf^ri —

cn O ^ O (N O

(S^QO^QOONin

^Hr-oefSTj-r-w-ip

00 vj TT r-. 1». ^ ^ >0 ^ n

eu z

2 g

w

I S I I I

I I * s J H u v  H f1 . S i » g-.2 "

1 î S 4» M

Vi

k

1

s s S -S e Ä B I1) O U 8 Ü S U I V 8. V&

" 3 3 3

5

g

1

3 3 3

•O -O -D XI § 3 3 3 3 3 « u n u n i

1 £

Dirk

Berg-Schlosser

tn — -H n « — oo

od

~

»o

--o-

^onooinnie^oov dd-HO«Hdddd

Vi TT N

u 3 T3 c < ufl E

c

T C3 a 4>

O co es o en v> o vn Ö «•»; es o o o

ert u s

es 00 r» en o vo es >ri co vi d en v> ri en

t'en

, oo Os O eS es fes «S ^ en es es

o o V) o en TT orö es es d Os O

(S

o r- , O O Tf t-. O Nen Os o vo d d â d o

oo 00 TJ; en Osr* so iri Osr-^ (S T esjî es

*

vO

m ^ q es es oo en tt • es 6 ri^ ^ es -h h ^

a h^ 73 3 rU= 32 S a. §•1 1O Hg 5

h

r-

u 33 1-2 c C^ u QSo

«

H

«

World

c i a 1> Ö w CL o 00 jr m s-1£^ 4 a5 S CL. c 5Q O So

Tc3 (d ä C e

o COm e». en r- t- rd en o cs soes

o m o \o ^ ö ö o Öd d

1

en o

00 Cl es ^ 00 q , q es Ö

*

es es

r- r- O rt en - h o ên d vS O ^ vi en -«r »n vi

d

o O , o o wo n o o o o oo o

* **

en

TT O

m Tt ( o o tt -h wn 00 > O OO d o -- O —

o o , o o o »n on tT o d o O < ^ O —i

» uQ p >O 8S J3 1 M ä« 1 1 §at ['o 2 o Q>$ V)i E_e o 1 'Sa JJ O «u o u V o u •c — 3 3 3 3 3 3 3 « «h a a 3 a a a a a oo co co co co co coû-

S e

as? sE •3 \O U>

asto M C Q O) 00ts Ë» 9* S " g ac w2 g g 'Je pd c SS §ü s. : 8 g l u l l .. «> 4)« J) O g «3 33 3 3 3 3 u w a a 3 3 3 3 3 2 ,o CS CS

oo n » (s Tf n « o oo o « (S CS CS CS CS —

189

Berg-Schlosser

c e M i n o o e M h ^ o ^ i n es — ^ o — — — © « ~

oo o

s à s

s t1 s

s a s

M

S O.S

£ < 2

•9 k -e duri t/1 .¿S 00

.5 -e a a l a S i

S

g

%^

I

sS

I I e § 1 Ì ° J S ci iE rS • " ? 13 4 s» « •§ s u >» o u « . 1U js .a c; -S < 3 E c =a u a) O S Ï S gS'S.a ï5 o J¡ o « % fi V ft

z « s i £

•g o-.S . S s f g -a a n S3 -o ¡2 Ä P D S 5 »-H oi rn ^ io

fe ^

° S2 o « 0 e3 > 1 e i Ep § .3 E §•§ O §

Ì 1 C » E g

8 n •a

•S.Ï.SS

I ¡ 5 1

ts m o> u c ef

S •a § 'i

•i D

i-HiScn^vSvdt-^oo

1 1 I.S

Ê .5 u uA JZ C M cg 9M a a

.a

S. S. ob -a fi S •S « - -S î - -s 1 s 1 1 -a -S z E E e

G •S S cj ir ON

s s« I I I I

c s

S « «go -A « > ë § ! O 9 o e J

S S § 'g •S * § i "3 o g .3 MZmíS rt pi m

324

Arab Political

G

I

Systems

5

6

3 8

a

£

2 oo o •o

1 I ? Sx C2

"8 >

•a

•s

c

a

1

•S

s a •a .1 S * 8

u E

a

11 S -ä "

E 6 c •8 1 1 '1 î

S 1 •s o I«

i »

•ai l a £ 1

!

-C m -C M fla s äsS 3

I so S O *û T3 _T

.c O

ä ii ä

S

s s -a

» .2

•s * -a S .Sä

s 3

S

I

«

g 1

E .a 3 oa. &

s •I 3 S

% S

S£

S 3

•S

•s £

s .3

1

I •a

8M m

1

S.

S

I

•a I" •a 3 S,! a ° i 8 5

&

Carlos

Huneeus

345

T h e i n c l u s i o n a r y r e g i m e s are characterized by the f a c t that the government, from above, tries to mobilize the lower strata, particularly the peasants, which requires specific institutional and political mechanisms. The best example of this subtype was the Revolutionary Movement of the Armed Forces in Peru, which, after overthrowing the government headed by President Fernando Belaúnde Terry in 1968, built such mechanisms and followed mobilizing policies. This attempt at "South American reformism" (Mansilla 1977) failed because it was not possible to channel the mobilization of the popular, peasant, and worker sectors through the institutions formed by the government (SINAMOS for the peasants), and the economic development strategy did not achieve the results that it promised its adherents. These failures increased social conflict, which had to be, definitively, repressed by the military. Unlike the traditional military regimes, the new military regimes are characterized by the armed forces, as an institution, assuming responsibilities of political and administrative direction at the national and regional levels (Stepan 1971). This implies that there is no personalization of power, but that power tends to be institutionalized. It is even possible to witness a rotation in the directive positions of the government, which corresponds to changes in the correlation of forces within the military. Thus, for example, in the case of the authoritarian regime of Brazil that appeared in 1964, there have been five presidents, all elected by the military and later confirmed by Parliament (Castello Branco, Costa e Silva, Garrastazú Mèdici, Geisel, and Figueiredo). In the case of Argentina, we also find this rotation, although in this case the political differences among members of the military high command are not always handled consistently, as manifested by the conflicts in the designation of the successor to General González Videla. In fact, the designation of Galtieri was described by observers as a palace revolution within the armed forces. In the case of Uruguay, we also find this rotation, because of the failure of the military government to attain the support of the people in its project for a constitution in the November 1980 referendum. The defeat of the military in the referendum led to the substitution of the civilian president, Méndez, by General Alvarez. The new military regimes do not exclude the personalization of power, as happened with General Pinochet in Chile. In fact, in this case, General Pinochet succeeded in consolidating his political position through ties and loyalties of a personal kind. Over time, this broke the agreement among the members of the government junta that the presidency would be rotated among them, with the purpose of assuring a well-rounded direction. From the institutional viewpoint, we can find in this subtype a significant degree of institutionalization of power. This is due to the existence of mechanisms established in the pursuit of legitimization and its constant renovation. In the case of Brazil, we find a high degree of institutional differentiation, which, in turn, rests on preexisting institutional development

346

Latin American

Political

Systems

tendencies, such as bicameralism and federalism. In Argentina, Chile, and Uruguay, we find more clearly an abrupt rupture with the country's institutional traditions. The cut, however, is not complete, judging from the utilization of electoral mechanisms of the sort represented by referenda in Chile (twice) and Uruguay (once) to seek new sources of legitimization for the regime. In the case of Chile, General Pinochet could consolidate his position as president through the Consulta of 1978 and the 1980 referendum (Huneeus 1981b). Finally, there are authoritarian regimes in a strict sense of the word. In this case, there is a political system with institutions endowed with attributes and clear competitions, whose positions of power are distributed among the governing elite through established rules, and which succeed in co-opting important nuclei of the population. With it, the cores of social conflict are reduced and the employment of repression as an instrument to assure political stability diminishes. In L a t i n A m e r i c a , M e x i c o c o n s t i t u t e s the b e s t e x a m p l e of an authoritarian regime. In fact, after the bloody civil war, a political system was constituted that has achieved the integration of the principal social groups by means of a distribution of power following proportional criteria. The institutional differentiation is strongly integrated and endowed with a d e c i s i o n m a k i n g c a p a c i t y t h r o u g h the p r e s e n c e of the I n s t i t u t i o n a l Revolutionary party (PRI). Until now, all those reforms that have tended to improve the democratic input of the political system have not modified substantially the bases of the authoritarian regime (Mols 1981).

Explanatory Factors of the Variety and Instability of Political Systems I have demonstrated that the theses of variety and instability are a form of approximation of an extraordinarily complex and changing reality. In no way are we seeking monocausal or reductionist answers. In fact, there is always someone to explain the failure of democracy in Latin America as being due to the absence or weakness of a democratic political culture (i.e., cultural interpretation). For others, the cause is located in the existence of solid economic structures that obstruct the introduction of democratic regimes and permit only the emergence of authoritarian regimes. I believe that we must explore all the national and international factors that can help to explain this phenomenon, not only those found in the conditions of the twentieth century, but, more precisely, the structural factors that existed at the moment of the upsurge of the national states in the nineteenth century. It is necessary to include in the analysis the factors that later influenced these countries' political development. A historical-systematic analysis, therefore, could show the factors that explain the variety and instability of Latin American political

Carlos

Huneeus

347

systems. Following Rokkan (1975), it would be wise to warn that the states of Latin America, unlike the European states, can be characterized as latecomers (i.e., states that must develop when there is already an international system on the road to consolidation and must occupy a peripheral position in the capitalist system). Moreover, one must consider that the Latin American states have had to confront, in a short period of time, the diverse crises and challenges that, in the case of Western democracies, were handled during a much longer time span and without having the pressures and limitations of the international system that the states of Latin America must face. As latecomers, they encounter existing models of political development in the East and West. These exert considerable economic and ideological influence by means of the modern communications media, which is today a very efficient means for inducing intensive social mobilization (Rokkan 1975, 574). All these factors tend to be superimposed on Latin America and constitute the parameters which affect Latin American political development. These parameters considerably impede the possibilities of a democratic development. Along with the analysis of these structural parameters, it is necessary to analyze the socioeconomic and political characteristics of each country in particular. These factors include, inter alia, the enormous inequality among the social strata with regard to access to political power, the huge social and economic differences among the distinct strata and social classes, and the lesser or greater differentiation of the social and productive structure. These internal and external circumstances condition the stability of not only the democratic, but also the authoritarian, regimes. Both regime types often must introduce institutional mechanisms in order to give an image of participation and competition, with the goal of generating levels of social consensus to provide political stability to the system. It is possible to visualize the power structure peculiar to Latin Amcrica at this level of generalization. In effect, the axis of the power structure is determined by the social groups that are located in the most modern, dynamic, and dependent sector of the economy, which is the most active sector. They are, in the first place, national capitalist groups that cannot always, or do not always, wish to elude conflicts of interest with foreign capital. On the other hand, we find the top functionaries of large state enterprises (Elsenhans 1981), who, in order to strengthen their position of power, tend to favor foreign capital interests to the detriment of national private capital interests. An example of this has been the policy of nationalizations in diverse countries, which has been linked to a process of growing income concentration and not to more egalitarian development tendencies. Development strategies based on growth with intensive use of capital

348

Latin American

Political

Systems

and, therefore, labor reduction, and containing high rates of income concentration, can hardly resolve the social problems facing Latin American societies (Prebisch 1981). In effect, the lack of social participation extracts consensus from the development model. It generates a "crisis of structural legitimacy" at the level of the political system (Graciarena & Franco 1978). This has been shown with complete clarity in the case of the monetarist model in Argentina and Chile, which has grounded its economic propositions in eliminating democratic mechanisms and participation of the people. These models cannot perceive that, in those countries where there has been an era of political participation and social mobilization, their elimination and the breakup of union organizations and political parties are only obtainable through a strong, consistent repressive policy that generates tensions and conflicts that are not always possible to measure and control. In that sense, the "disappearance" of people, torture, and even civil war tend to be an inevitable consequence of these economic policies and their political assumptions.

Re democratization? I have asserted that, until now, political development in Latin America has been characterized by the variety and instability of the political systems. This does not mean, however, that future development is going to be always dominated by these two tendencies. In fact, reccnt changes of political regimes tend to show that the number of countries initiating experiences of democratic regimes is increasing. These new experiences emerge, moreover, because of the failure of authoritarian regimes that have not been able to resolve the most basic needs of the population and are unable to generate a constant, significant loyalty in the population (Handelman & Sanders 1981). With the failure of these authoritarian experiences in countries that had, formerly, had democratic experiences (which were not always appreciated by social actors and some elites), the high costs of repression and the economic hardship of authoritarianism leads to a revaluation of democracy by these same social and elite groups (Moulian 1981). It even seems likely that the support bases of these new democratic orders are considerably broader and more solid than those formerly experienced. My interest in this last part of the chapter is to give some reflection to the effect of the "legacy" of the authoritarian regimes on the new democratic experiences. First, the authoritarian experiences tend to generate profound divisions in the social system, since the mechanisms of control and coercion have been oriented toward a social sector (workers) and not to all social classes and actors. In that sense, there is a division between groups favored by the economic and political model (e.g., entrepreneurs and the military) and broad social sectors prejudiced by it.

Carlos

Huneeus

349

The "structural asymmetry" is reflected, above all, by the system of postauthoritarian industrial relations. There, we find that the business organizations have greater organizational cohesion and increased resources. In contrast to this, union organizations are atomized by coercion and have a considerably weaker direction because of the repression that they have experienced. Also, this "structural asymmetry" exists in the relations between the military and political parties. In fact, the government's experience of the military normally has led it to increase considerably the size of the armed forces (i.e., arms and salaries), as well as to train ranks of officers in the management of political issues and state administration. Political parties, on the other hand, have been excluded f r o m political activity, and their l e a d e r s h i p h a s b e e n c o e r c i v e l y d e c i m a t e d . In that s e n s e , in the postauthoritarian phase, the military could have greater cohesion and capacity of response in the face of political crises than do the parties (e.g., in A r g e n t i n a ) . T h i s m a y n o u r i s h a v i c i o u s circle of d e m o c r a c y and authoritarianism and work to the detriment of democratic tendencies. In the second place, authoritarian regimes leave scars in the political culture of the population and the elites. As a matter of fact, the employment of mass media to diffuse the ideological message of authoritarian regimes l e a v e s i m p r i n t s on the y o u t h who a b s o r b e d the t r a n s m i s s i o n of antidemocratic values through the educational system; all of this contributes to the emergence of political tendencies among many of the youth in the direction of apathy and cynicism. They lack interest in public affairs and in the strengthening of individualist and particularist postures in the socicty. With this, the possibility of building a democratic order erodes because the capacity for active support of it has been impaired. Should a democracy be installed, the tendency towards cynicism and apathy weakens its capability successfully to confront crisies and impedes the consolidation of democracy. In the t h i r d p l a c e , the s t r o n g s o c i a l c o n t r o l that is e x e r t e d by authoritarian regimes leads to the tendency of affected social sectors to put off the realization of their interests until a time when the political system opens up, at which time, a difficult situation arises. Democracy confronts an enormous accumulation of social demands that were repressed in the a u t h o r i t a r i a n p h a s e . Without the s a t i s f a c t i o n of these d e m a n d s , the democratic order cannot obtain a moderately legitimate social consensus. It should not seem odd, therefore, that the first stages of a new democracy must endure strong upheavals produced by this explosion of demands, and that the government tends to find itself overcome by the expectations of the population. In order to avoid a situation in which the government reconstitutes social order by means of coercive policies, the principal social and political groups need to seek political formulae of consensus in order to stabilize a democratic order. The consensus need not necessarily be on the basis of policies of "grand coalition" or of "broad fronts," but it should include a differentiated

350

Latin American

Political

Systems

strategy for the distribution of costs and benefits from political experience. Only a democracy with a wide and solid support base can avoid failure and a return to an authoritarian regime.

Note 1. Cuba is not included in this typology because its political system is exceptional for Latin America. In fact, it would be more appropriate to include it in a comparative analysis of communist regimes than among the Latin American political systems (Von Beyme 1975; Brunner & Meissner 1979). This decision is supported by the type of socioeconomic and political development that Cuba has pursued since the end of the 1960s. Abandoning the intent to pursue a road of its own from the ideological, political, and developmental viewpoint (Fabian 1981), Cuba has initiated a dynamic of a progressive ideological, political, and cultural movement toward the Soviet model. This development tendency responds to a variety of factors whose analysis escapes the purposes of this chapter and cannot be condensed. My typology also excludes the Nicaragua of the Frente Sandinista, because it is still not clear what type of socialism will be adopted, and what institutional forms it will have.

References Alleman, F. R. 1974. Macht und Ohnmacht der Guerilla. Munich: Piper. Bau, M. 1982. Entwicklung und Partizipation. Brunner, G., and B. Meissner (eds.). 1979. Verfassungen der kommunistischen Staaten. Paderborn: Schoningh. Collier, David (ed.). 1979. The new authoritarianism in Latin America. Princeton, NJ: Princeton University Press. Di Telia, Torcuato S. 1965. "Populism and Reform in Latin America." In Obstacles to change in Latin America, ed. C. Veliz. London: Oxford University Press. Dix, R. H. 1973. "Oppositions and Development." In Regimes and oppositions, ed. Robert A. Dahl. New Haven & London: Yale University Press. Elsenhans, H. 1981. Abhängiger Kapitalismus oder bürokratische Entwicklungsgesellschaft: Versuch über den Staat in der Dritten Welt. Frankfurt: Campus Verlago Fabian, H. 1981. Das kubanische Entwicklungsmodell. Opladen: Westdutcher Verlag. Foxley, A. 1979. Políticas de estabilización y sus efectos sobre el empleo y la distribución del ingreso: una perspectiva latinoamericana. Estudios Cieplan (Santiago.) 2; original in English published in 1971. Goldenberg, B. 1971. Kommumsmus in Lateinamerika. Graciarena, J., and R. Franco. 1978. Social formations and power structures in Latin America. London: Sage. Handelman, H., and T. G. Sanders (eds.). 1981. Military government and the movement toward democracy in South America. Bloomington: Indiana University Press. Huneeus, Carlos 1981a. Der Zusammenbruch der Demokratie in Chile: Eine vergleichende Analyse. Heidelberg: Exprint Verlag. . 1981b. Elecciones no-competitivas en las dictaduras burocrático-authoritarias

Carlos Huneeus

351

en América Latina. Revista Española de Investigaciones Sociológicas 13. Krumwiede, H. W. 1980. Politik und katolische Kirche in gesellschaftlichen Moder misierungsprozess: Tradition und Entwicklung in Kolumbien. Hamburg: Hoffman und Campe. Levine, Daniel H. 1981. Religion and politics in Latin America. Princeton: NJ: Princeton University Press. Mansilla, H. C. F. 1977 .Der sudamerikanische Reformismus. Rheinstetten: Schindele. M e d i n a Echevarría, J. 1964. Consideraciones sociológicas sobre el desarrollo económico en América Latina. Buenos Aires: Solar/Hachette. Mols, M. 1981. Mexico im 20 Jahrhundert: Politisches System, Regierungsprozess und politische Partizipation. Paderborn: Schöning. Moulian, T. 1981. "Crítica a la crítica marxista de las democracias burguesas." In DESCO: América Latina 80: Democracia y movimiento popular. Lima: DESCO. Nohlen, D. 1973. Chile: Das sozialistische Experiment. Hamburg: H o f f m a n und Campe. . 1 9 8 2 . " R e g i m e w e c h s e l in L a t e i n a m e r i k a : Ü b e r l e g u n g e n z u r Demokratisierung autoritärer Systeme." In Lateinamerika Herrschaft, ed. K. Lindenberg. , a n d F. N u s c h e i e r ( e d s . ) . 1982. Handbuch der Dritten Welt. Vol. 2: Sudamerika-, Vol. 3: Mittelamerika und Karibik. Hamburg: Hoffman und Campe. O'Donnell, Guillermo. 1973. Modernization and Bureaucratic authoritarianism: Studies in South American politics. Berkeley: Institute of International Studies, University of California. Prebisch, R. 1981. Capitalismo periférico: Crisis y transformación. Mexico: Fondo de Cultura Económica. R o k k a n , Stein 1975. " D i m e n s i o n s of State Formation and National Building: A Possible Paradigm for Research on Variations within Europe." In The formation of states in Western Europe, ed. C h a r l e s Tilly. P r i n c e t o n , N J : P r i n c e t o n University Press. Sotelo, I., K. Eber, and B. Moltmann. 1975. Die bewaffneten Technokraten, Militar und Politik in Lateinamerika. Hannover: Fackeitrager. Stepan, A. 1971. The military in politics: Changing patterns in Brazil. New Haven, CT: Yale University Press. . 1978. The state and society: Peru in comparative perspective. Princeton, NJ: Princeton University Press. Von B e y m e , K. 1975. Ökonomie und Politik in Sozialismus: Ein Vergleich der Entwicklung in den sozialistischen Landern. Munich: Waldmann, P. 1974. Der Peronismus 1943 bis 1955. Hamburg: Hoffman und Campe. . 1 9 8 2 . " G e w e r k s c h a f t e n in L a t e i n a m e r i k a . " In Internationale Gewerkschaftshandbuch, ed. S. Mielke. Opladen: Laske und Budrich.

IB Commentary George Philip

There is no doubt that Huneeus is right to identify variety and instability as the key features of Latin American politics. In late 1987, there was a fairly orthodox Marxist-Leninist system in Cuba, a less defined but similar system in Nicaragua, a capitalist single-party system in Mexico, orthodox democracy in Venezuela and Colombia, and military-backed dictatorship in Chile. Currently, a number of other countries have recently moved from military dictatorship to some form of elected civilian regime. As to instability, only Mexico has maintained the same form of rule consistently since 1945, and things have been changing even there. How is one to gain understanding of such a varied region, which students of comparative politics have generally found so difficult to grasp? Huneeus attempts to do so by developing yet another set of classifications, but it is not at all clear that this is in fact the best way forward. Given his classificatory system, it is not clear why he has chosen to leave Cuba and Nicaragua out of account (Mexico, after all, has a dominant party system as well), but the more serious problem is that politics in many Latin American countries changes so quickly as to endow any system of classification with a built-in obsolescence. A further difficulty is that classifiers can rarely resist the temptation to put a variety of different situations into a single case, which they do not readily fit; a notorious e x a m p l e of such an enterprise is O ' D o n n e l l ' s notion of bureaucratic authoritarianism. Nor is Huneeus wholly free of this tendency. For example, he states that exclusionary authoritarianism is "based on monetarism and strict policies of stabilization," but this is only intermittently true of the Argentine regime between 1976 and 1983 and not true at all of Brazilian policy after around 1967. Also, he notes the "reduction of tasks and functions of state enterprises," which is not true of Brazil, and that political parties are prohibited, which is not true of Brazil or even, as a completely general statement, of Pinochet's Chile. As an alternative approach, one might begin by observing that one reason for the observed variety and instability is that most Latin American countries have neither strong political institutions nor a strongly 352

Commentary/George

Philip

353

institutionalized political system (Huntington 1968; Anderson 1967). An exception to this general rule can be found in Mexico and, probably, Cuba where the single- or, at any rate, dominant-party systems are relatively institutionalized, albeit, like most Latin American political arrangements, on a far more individualized basis than are their European counterparts. The role of dominant-party systems could indeed be made the subject of comparative analysis. In most other countries, however, the form that politics takes (i.e., the nature of the formal political system) is less important than is the nature of the power elites and social forces that interact in politics. Making due allowances for regional variations, politics involves a corporate military (and to some extent a corporate church), a wealthy or plutocratic class whose political loyalties are essentially instrumental, and forms of mass politics whose present or past interaction with conservative forces is the motivating force behind most political conflict. The combination of weak political (though not social) institutions on the one hand and developing social "modernization" on the other has generally led, not as Huntington (1968) believed it would, to political decay, but rather to a series of ad hoc arrangements that have been designed, above all, to reflect the power of civilian elites and an institutionalized military (Philip 1985). Conversely, in Central America outside Mexico, the absence of a corporate and professional military and the weakness of the civilian elites have enabled revolutionary threats to arise. These threats have brought Washington directly and heavily into the area. If this interpretation is valid, then (except in Cuba and Mexico) it is more useful to look at power elites (or, if one is a Gramscian, hegemonic blocs) than at political institutions as such. To put the same point a little differently, democracies and dictatorships may come and go, but the military, the church, and the plutocracy (often described, in the Latin American context, as the oligarchy) tend to go on forever. The differences among the political systems of the various countries reflect mainly the different ways in which these forces have arranged power among themselves and the different forms that mass politics have taken. On the first point, for example, the behavior of military governments owes a great deal to the position of the military in society before the coup. On the second, moderate popular movements have at times been accepted within some kind of democratic framework (e.g., in Venezuela), while radical popular movements (e.g., Unidad Popular in Chile and Acción Democrática in Venezuela between 1945 and 1948) have been repressed. Particularly characteristic of South America have been the attempts made by previously conventional military or promilitary figures (e.g., Vargas and Perón) to use the power of the state to build a coalition of urban supporters and to seek to generate mass appeal. "Populism," in its many forms, appeared for some time to be a way of resolving some of the tensions inherent in rapid social m o d e r n i z a t i o n , but eventually it b e c a m e discredited as the internal contradictions of the various movements became increasingly apparent.

354

Latin

American

Political

Systems

Huneeus does not simply offer a classification but seeks to explain Latin American politics in terms of the "latecomer" status of the area and the development policies adopted. Much of what he says here is valid. However, an economist might ask why many Latin American countries have adopted development strategies that are more inegalitarian, as well as being less successful, than those adopted in the East Asian newly industrialized countries (e.g., South Korea). A more market-oriented policy is not necessarily more regressive socially than is a more state-led path. Mexico, for example, with its revolutionary tradition and its energetic use of the state to promote growth, has an unusually regressive pattern of income distribution, even for Latin America. It is possible to make only a few brief remarks here on this very general theme. European social democratic thought has generally considered it axiomatic that market-led development generates extreme inequalities, and that a democratic state both should and can use central state power to limit these inequalities and to ensure that the broad majority of the population benefits from economic progress. Latin American dependencia theory has, by and large, accepted this assumption. Market-oriented development is associated with inequality, popular discontent and dictatorship, while a more mixed economy is associated with democracy and greater social harmony. Huneeus seems, broadly speaking, also to have accepted this assumption. Even without the neoconservative critique of the role of the state in liberal democracies, it has been clear for some time that this equation does not work particularly well in Latin America, where there has been an extended role of the state, very great inequality, and varying degrees of political discontent and repression. Studies of public policy in Latin America (outside Cuba) have found many examples of policies favoring the middle class and rich (e.g., artificially low gasoline prices, a tendency to overvalue currencies, public housing projects that house mainly the state bureaucracy, and the expansion of higher education before the provision of universal primary education). One reason for this may be that clientelism, which is pervasive in Latin America as a basis for bureaucratic appointment, tends to influence the provision of state services away from the poor and toward those with palanca (pull). Another is simply that the middle class, considered either as individuals or as a collective, is more powerful than the poor and better able to influence the allocation of state services in its favor. It is not clear, either, that the role of the state in most Latin American countries has been particularly conducive to developmental success. The tendency in the last fifteen years has, notoriously, been for Latin American governments to borrow heavily, invest in unviable projects, and find themselves unable to repay or even service the loans. Whatever the final outcome of the debt crisis, there is no doubt that the region as a whole has lost several years of economic growth as a consequence. This can hardly make life easier for the emergent democratic governments.

Commentary/George

Philip

355

Clearly, there is very much more that could be written on this topic, but the apparent success of some of the Asian economies and the clear failure of at least some of the development strategies pursued in Latin America should, if nothing more, give pause for thought (Duran 1985). One of the conditions making for political authoritarianism in the 1970s, at any rate in the Southern Cone, was a growing frustration among elites about relative (in the case of Uruguay, absolute) economic decline. Similarly, one of the reasons for the relatively smooth transition to democracy in Venezuela after 1958 was the availability of oil wealth (Levine 1985). For the future, as in the past, the prospects for democracy in Latin America may be crucially dependent upon the ability of democratic governments to pursue successful development strategies.

References Anderson, C. W. 1967. Politics and economic change in Latin America. New York: Van Nostrand. Duran, E. (ed.) 1985. Latin America in the world recession. London: Cambridge University Press. Huntington, Samuel P. 1968. Political order in changing societies. New Haven, CT: Yale University Press. Levine, Daniel H. 1985. The transition to democracy: Are there lessons from Venezuela? Bulletin of Latin American Research, 4(2). Philip, George 1985. The military in South American politics. London: Croom Helm.

IB Commentary Karen L. Remmer

The diversity of political systems found in L a t i n A m e r i c a p r e s e n t s comparativists with a considerable theoretical challenge. Liberal democracies coexist with personalistic dictatorships and a full-blown Leninist regime, as well as with several other variations and combinations of democratic and authoritarian forms of rule. Moreover, the range of political experience encompasses both highly stable and unstable political systems; robust and fragile state structures; socialist and market-oriented development ideologies; corporatist and pluralist patterns of interest representation; and a full spectrum of international alignments. Making sense of this political diversity is not simply a matter of inventing a typology or classificatory scheme that can be shown to correspond in some way to empirical reality. The litmus test of conceptual distinctions is theoretical utility. Do they resolve outstanding theoretical issues, facilitate analysis of political similarities and differences, augment our understanding of important phenomena, or provide us with new analytical insights? From this perspective, the problem for students of Latin America is one of developing concepts that not only describe the region's political diversity but also enhance our ability to understand causes and consequences. In order to do so, typologies or other classificatory efforts must meet two basic criteria. First, they must be capable of subsuming the full range of the region's polities (in the Latin American case, they include both Cuba and Nicaragua). To exclude any polity on the grounds of exceptionalism is highly problematic. After eight years in power, the institutions and orientation of the Sandinista regime are no more provisional or undefined than those of Latin America's many new liberal democracies. Similarly, while Cuba can be u s e f u l l y compared with other Leninist regimes, the availability of a non-Latin American comparative framework is anything but unique to the Cuban case. Venezuela and Costa Rica, for example, can be appropriately compared with other liberal democracies, Mexico with other one-party systems, or Paraguay with other dictatorships. Second, if concepts are to prove useful in analyzing political diversity, 356

Commentary!Karen

L. Remmer

357

their breadth and complexity must be strictly delimited. Otherwise, important questions are defined away in advance, closing the door to causal analysis and explanation. Concepts with a large number of defining characteristics also produce uncertainty regarding empirical referents and a corresponding tendency toward the proliferation of classificatory categories. These problems are well-illustrated by the theoretical debate surrounding the concept of bureaucratic authoritarianism. By defining bureaucratic authoritarianism in terms of a complex constellation of traits, O'Donnell's seminal work (1973; 1978) drew attention to the linkages among socioeconomic development, regimes, ruling coalitions, and public policy in Latin America. At the same time, however, his multidimensional concept generated enormous conceptual confusion and much unproductive debate as to whether particular cases, such as M e x i c o or P e r u , c o u l d be c o n s i d e r e d i n s t a n c e s of b u r e a u c r a t i c authoritarianism against which one might test significant propositions. To add to these difficulties, so many different variables were built into the definition that the c o n c e p t i m p e d e d c o m p a r a t i v e a n a l y s i s of the c a u s e s and consequences of political change and variation. Questions concerning the interrelationship of variables became definitional rather than empirical issues (Remmer & Merkx 1982). These problems are not peculiar to O'Donnell's work. Bossert's (1983) elaborate typology of Latin American regimes, for example, similarly closes off major avenues of research, including the possibility of analyzing the relationship between political participation and stability, because it utilizes four different variables to classify systems: the strength of the state; stability; ideological orientation of governing elites; and degree of democratic participation in politics. Bossert's four defining characteristics also leave us with sixteen different types of political systems—nearly as many as the number of nations in Latin America. The point of replacing proper names with abstractions from spatiotemporal parameters is thus effectively lost. Huneeus' contribution to this book offers a far more useful set of concepts for comparative analysis. He explicitly links his classification of political systems to an analysis of political change and presents concepts that do not define away key questions about political similarities and differences in Latin America. Nevertheless, Huneeus utilizes an unnecessarily broad and complicated set of criteria to classify systems into nine categories (populism, four varieties of democracy, and four of authoritarianism). Participation levels, o p p o s i t i o n f r e e d o m s , party system, political l e a d e r s h i p , the institutional role of the military, ideology, public policy orientation, and the social composition of the governing coalition all figure in his delineation of Latin American system types. Most of these definitional criteria appear superfluous and even counterproductive inasmuch as they create categories that are not mutually exclusive. The multiplicity of definitional criteria also detracts from the typology's core, which consists of variations in regime or what might be called the rules of the political game. If all criteria of

358

Latin American

Political

Systems

classification except those of participation and competition are eliminated, this core remains. At the same time, it becomes possible to reduce the overlap among categories, such as that between populism and inclusionary military authoritarianism; expand the comprehensiveness of the typology to cover the entire range of Latin American political systems; cut back the number of categories to a more manageable four; and open the door to conceptualizing political system similarities and differences in terms of two intersecting continua. The resulting two-dimensional scheme, as indicated by Figure 15.1, takes us back to the simple but rather elegant mode of dealing with the problem of regime classification that was presented by Dahl (1971) more than a decade ago. Conceptualizing similarities and differences among Latin American nations in terms of two intersecting continua offers several advantages over the more usual typology or classificatory scheme. First, the simplicity of the conceptual distinctions does not define away important theoretical questions,

Figure 15.1 Two-Dimensional Classification of Regimes

Exclusionary democracy

Inclusionary democracy

v

au >

•3

•3 8. E o

U

Exclusionary authoritarianism

i '

1 1

Inclusivcncss

Inclusionary authoritarianism

Commentary!Karen

L. Remmer

359

such as the relationship between regime and policy. Second, unlike complex concepts that differentiate among political systems on the basis of a set of interrelated traits or syndromes, the conceptual distinctions presented in Figure 15.1 facilitate comparisons not only among Latin American nations but also between the Latin American region and other parts of the world. These distinctions also explicitly address the problem that most actual regimes do not fully conform to any of the polar types represented in the extreme corners of the figure. The classificatory terminology that is conventionally applied to the discussion of political systems obscures this problem, leading to endless labeling disputes of the type that plagued the discussion of bureaucratic authoritarianism. Latin Americanists have traditionally distinguished among regimes on the basis of their competitiveness, as represented by the vertical axis in Figure 15.1. The contrasts that have been drawn between military and civilian r e g i m e s , d e m o c r a c i e s and dictatorships, as well as c o m p e t i t i v e and noncompetitive regimes, all reflect an underlying concern with the right of opposition elements to compete for popular support and exercise the classic liberal freedoms of expression, organization, and suffrage. Scaling regimes along the dimension of inclusiveness or the breadth of the right to participate in political affairs is less conventional, partially because it is often assumed that participation and competition covary. Yet, intensely competitive regimes may limit participation to a small sector of the population, as in Chile before 1958, whereas regimes that restrict opposition freedoms may nevertheless encourage participation, as in Cuba or Nicaragua. The latter point was explicitly recognized by the literature on bureaucratic authoritarianism, which distinguished between exclusionary (or bureaucratic) authoritarianism and inclusionary (or populist) authoritarianism. Figure 15.1 builds on this theoretical insight by extending the distinction to competitive regimes. Separating regimes that fall at the exclusionary and inclusionary poles of the political spectrum is particularly important for developing an adequate understanding of political change in Latin America. As Huneeus suggests, inclusionary and exclusionary forms of rule emerge under very different c o n d i t i o n s and h a v e dissimilar impacts. T h e f a i l u r e to o b s e r v e this fundamental point is responsible for much of the vacuous theorizing about politics in Latin America. Searching for a common denominator behind such disparate phenomena as Velasco Alvarado's reformist military government in Peru and Pinochet's repressive regime in Chile, for example, is unlikely to take one m u c h beyond the tautological proposition that the military intervenes to defend its institutional self interest. The effort to uncover generic policy contrasts between competitive and authoritarian regimes has proved equally self-defeating. Again, tautological propositions, such as that competitive regimes are more likely to tolerate political opposition, have been the chief result. Generating interesting hypotheses about Latin American democracy has

360

Latin American

Political

Systems

also proved difficult. The main conclusion of the recent literature on democratization has been that generalization is difficult, if not impossible, because virtu and fortuna shape political events. According to Malloy (1987), for example, "there is obviously a key voluntary dimension to the process [of democratization] that not only makes it difficult to formulate a general theory but also precludes neat deterministic theories based on general laws." However gratifying it may be to place human talent and choice at the center of transitions to democracy, such an analysis does not take us very far. The literature on democratization, like the literature on the military, has simply attempted to generalize about too much, lumping together very disparate political phenomena. I n c l u s i o n a r y and e x c l u s i o n a r y a u t h o r i t a r i a n regimes have very different points of vulnerability and break down for d i f f e r e n t r e a s o n s . W h e r e a s the f o r m e r s u c c u m b to m o u n t i n g class polarization, the latter characteristically disintegrate in response to the erosion of the class tensions associated with their emergence. Likewise, the conditions favoring the emergence of competitive rule are not necessarily identical for inclusionary and exclusionary regimes. The distinction between exclusionary and inclusionary forms of rule is consequently fundamental to understanding not only the political diversity of Latin Amcrica but also patterns of change and instability through time. Theories emphasizing political culture, the structural constraints of dependent capitalism, or other commonalities among the countries of the region c a n n o t e x p l a i n L a t i n A m e r i c a ' s p o l i t i c a l diversity, n o r can modernization theory in either its original or inverted form. Whether one looks at longitudinal or cross-sectional data, no unilinear relationship can be drawn between socioeconomic development and regime type in Latin America. Does this mean, as scholars such as Therborn (1979) have s u g g e s t e d , that the p a t t e r n of r e g i m e c h a n g e in L a t i n A m e r i c a is predominantly cyclical in character and is driven by accident, human choice, or c o n j u n c t u r a l c o n f i g u r a t i o n s rather than underlying sociopolitical structures? Because such a theoretical perspective ignores evolutionary trends as well as cross-national similarities and differences, the answer is no. Looking first at longitudinal trends, important empirical relationships exist between socioeconomic structure and politics in Latin America; however, these relationships have been overlooked because of the failure to distinguish between inclusionary and exclusionary forms of competitive rule. For example, the sharp upsurge in authoritarianism in Latin Amcrica during the 1960s was not merely a conjunctural outcome. It reflected the declining viability of exclusionary competitive rule in the face of increased educational opportunities, urbanization, and the transformation of rural areas. Once such a relationship is recognized, democratization in Latin America obviously cannot be interpreted simply as a cyclical or rccurring phenomenon entirely divorced from underlying socioeconomic change. Over time, exclusionary forms of competitive rule have given way to inclusionary ones. Cycles of

Commentary/Karen

L. Remmer

361

democratization consequently have often involved more than a simple restoration of the status quo ante. The return to competitive rule in Peru during the late 1970s, for example, involved a dramatic expansion of political participation and a corresponding shift in the alignment and correlation of political forces. Hence, while competitive and authoritarian regimes have rotated in and out of power in Peru, as in other parts of Latin America, they have done so within the framework of a broader evolutionary or sequential pattern of regime change. In 1968, an exclusionary competitive regime gave way to the inclusionary military regime of Velasco Alvarado. The mobilizing policies of that regime in turn paved the way for the emergence of the most inclusionary regime in Peruvian history. Similar sequences of regime transformation may be observed elsewhere as well. Inclusionary authoritarian regimes, such as Velasco Alvarado's in Peru or Juan Perdn's in Argentina, have paved the way for inclusionary competition or exclusionary authoritarianism, but they have also closed the door to the possibility of political exclusion without coercion. The historical record also suggests that inclusionary democracy, which encourages the development of mass allegiances to political parties and trade unions, limits opportunities for the establishment of inclusionary authoritarianism. Hcnce, inclusionary d e m o c r a c i e s may alternate in p o w e r with exclusionary authoritarianism, as in Argentina since the 1950s; but inclusionary democracy u n d e r m i n e s the viability of either i n c l u s i o n a r y a u t h o r i t a r i a n i s m or exclusionary democracy. Cross-national variations in patterns of political instability emphasize the problems of understanding Latin American politics simply in terms of human volition, conjunctural configurations, or cyclical instability. The waves of competitive and authoritarian rule that have swept through the region have not affected all countries equally. Cuba, Paraguay, Costa Rica, Colombia, Venezuela, Mexico, and Chile stand out as notable exceptions. Space limitations preclude a case-by-case analysis of the conditions that have contributed to long periods of stability in these countries, but, once again, the distinction between inclusionary and exclusionary rule helps elucidate relationships between socioeconomic structures and political patterns in Latin America. High levels of socioeconomic inequality, for example, have been linked with comparatively stable competitive regimes of an exclusionary character (e.g., Colombia and pre-1964 Chile). Inclusionary competition, however, has provided a basis for political stability only in Costa Rica and Venezuela: settings where economic and institutional resources have m o d e r a t e d social cleavages. The reasons are relatively self-evident. C o n d i t i o n s of s e v e r e social inequality create strong i n c e n t i v e s and opportunities for political exclusion. The implications of these arguments for the process of democratization in Latin America are twofold. First, if competitive and authoritarian regimes have alternated in power within a broader evolutionary framework, the

362

Latin

American

Political

Systems

inevitability of a new cycle of democratic breakdown cannot be taken for granted. The recent authoritarian experience of a country such as Argentina may have altered the political game as profoundly as the inclusionary regime of Juan Perón did before it. Inter alia, the sheer brutality and incompetence of recent military regimes have left authoritarian rule deeply discredited and generated a new commitment to democratic institutions on the part of groups across the political spectrum. The second implication of this analysis, however, is that the viability of democratic institutions in the region is likely to vary with regime inclusiveness and underlying sociopolitical conditions. In Central America, where highly inegalitarian social structures create strong pressures for political exclusion, as well as in South American countries that have yet to experience a traumatic shift from inclusionary democracy to exclusionary authoritarianism, future democratic breakdowns appear not only possible but probable.

References Bossert, Thomas John. 1983. Can we return to the regime for comparative policy analysis? or the state and health policy in Central America. Comparative Politics 15 (July):419-441. Dahl, Robert A. 1971. Polyarchy: Participation and opposition. New Haven, CT: Yale University Press. Malloy, James M. 1987. "The Politics of Transition in Latin America." In Authoritarians and democrats: Regime transition in latin America, ed. James M. Malloy and Mitchell A. Seligson. Pittsburgh, PA: Pittsburgh University Press. O'Donnell, Guillermo. 1973. Modernization and bureaucratic authoritarianism: Studies in South American politics. Politics of Modernization Series No. 9. Berkeley, CA: Institute of International Studies, University of California. . 1978. Reflections on the patterns of change in the bureaucratic-authoritarian state. Latin American Research Review 13:3-38. Remmer, Karen L., and Gilbert W. Merkx. 1982. Bureaucratic authoritarianism revisited. Latin American Research Review 17:3^40. Therborn, Goran. 1979. The travail of Latin American democracy. New Left Review (January-April): 71-109.

16 Conclusions and Summary: A Long Journey Begins with a Single Step Jim Seroka Scholars from antiquity through current times have wrestled with the problem of how to compare and categorize political systems. This book has taken a giant step toward that worthy goal. While it is apparent that a quick answer will not soon be forthcoming, it is also apparent that many of the prior attempts at classification may not be relevant and will probably not advance us appreciably. The contributors to this book have conclusively demonstrated that future efforts at typology building must be team-centered and group-oriented. We have also experienced the necessity for prior agreement about the purpose for a classification scheme, and that we must select only those research questions that are most appropriate to our research agenda. Before we can provide answers, we must agree upon the appropriate questions. Before we argue about the nuances of competing research methods, we must agree upon the conceptual image of the political regime or state. A major achievement of the contributors has been the willingness to ask those questions and to propose various starting points. Cumulatively, the chapters and their accompanying commentaries have done much to advance our understanding of comparative political systems. Nevertheless, considerably more synthesis remains to be done. On the one hand, much of the diversity and variation that we have seen in modern political life no longer appears to be random, and considerable hope exists for true comparative inquiry. On the other hand, our understanding about the nature of political life is far from complete, and no global political typology readily suggests itself. E a c h of the c o n t r i b u t o r s and c o m m e n t a t o r s shared an i d e n t i c a l frustration: Prior attempts to classify and categorize polities are inadequate for today's needs and modern political research. Marradi's well-reasoned essay on classification sets a standard that offers hope to the global comparative endeavor and a mechanism through which we can measure our progress. Smith underscores conclusively that a new approach to comparative inquiry must be taken and that the paths of the ancients, while heuristically 363

364

Conclusions

and

Summary

interesting, cannot rescue us from our current morass. In brief, we cannot avoid the hard work that is necessary, nor should we spend time expecting the discovery of some magic wand to relieve us of this effort. Each of the partial attempts at categorization, ranging from Lijphart's discussion about democratic political systems through Huneeus's articulation of variation within Latin American systems, argues convincingly that prior typologies and categorization schemes that were applied to that group of nations are inadequate and often fundamentally flawed. None have argued that his or her approach has global applications or that she or he has accomplished more than the first tentative steps toward a comprehensive and dynamic typology for a global subsystem. Each, however, has offered an approach that is promising and deserves fuller scholarly attention. Each would agree with Remmer's comments that our classification schemes must be developed further in order to test hypotheses and to advance beyond trivial conclusions or tautological statements. Unlike the case of many of the natural sciences, political science does not naturally suggest a f o r m u l a from which to begin the process of differentiating among political systems. There is no natural hierarchy, no natural starting point for categorization. For example, should ideology be the initial differentiating factor, as suggested by this book's separate treatment for democratic, authoritarian, and socialist political systems? Should level of development, number of political parties, or military rule be the major differentiating factor, as demonstrated by the chapters on Third World political systems, political parties, and military regimes? Should geography and culture be that initial point of departure, as suggested by the chapters on s u b - S a h a r a n A f r i c a , A r a b p o l i t i c a l s y s t e m s , and L a t i n A m e r i c a n nation-states? As Marradi points out, the starting point is critical. Where one begins determines where one concludes the study and how one makes that journey. The order of the additional differentiating factors is also critical. If one a c c e p t s i d e o l o g y as the initial criterion f o r d i f f e r e n t i a t i o n , s h o u l d geography/culture, or economic development, or institutional differences become the secondary or tertiary differentiating variable? Currently, there is no clear path to decide these matters. In its entirety, this book does not argue for a single method for differentiating among political systems. Rather, it presents a rich diversity in approaches with the hope that one of them or a particular combination, may p r o v i d e the s p r i n g b o a r d f o r m o r e g l o b a l a p p l i c a t i o n s of p o l i t i c a l classification. The contributors diverge considerably on the issue of purpose (i.c, the intent of classification), also. Some asked whether one type of political regime is better than another and under what circumstances it is most appropriate. Others examined the impact of regime type upon regime stability, national wealth, or regime power. Still others correlated the type of

Jim Seroka

365

regime to cultural and historical concerns. Again, this apparent incongruity was intentional in order to give each contributor the maximum freedom and to permit each to exhibit the fullest potential of the classification scheme. One commonality among all the contributors was the tension between descriptions of the legal institutional arrangements and political reality. Each recognized that institutions such as parties, legislatures, and executives were not constants even within a self-selected subset of political systems, let alone on a global basis or across time. Each felt a frustration about the difficulties inherent in making predictions, but each expressed a strong belief that prediction should be the eventual goal of a typology. When we turn our attention to the individual chapters, we can identify a host of achievements, albeit on more modest levels. In the first part, Marradi's discussion of classification laid the groundwork for the succeeding classification studies and attempts to form subglobal typologies. He appropriately set the standard for what must be done. Smith's review of premodern political typologies eradicated any remnants of a general illusion that classical scholars possessed a unanimity of judgment about these issues and had solved the research problems of political classification for their time. It is simultaneously refreshing and frustrating to note that even such giants of antiquity as Socrates, Aristotle, and Polybius did not claim closure on this subject, and that their efforts and struggles are reflected in our present debates. This book's discussion of ideologically based classification schemes is an honest acknowledgment that we need to find ways to classify a particular subset of the global community before we attempt to present a classification scheme that includes all nation-states. Our fundamental criteria revolved around how groups of nation-states perceive themselves; Western democratic n a t i o n s and s o c i a l i s t n a t i o n - s t a t e s p e r c e i v e t h e m s e l v e s as distinctive—different from each other and from other nation-state groupings as well. The major problem with reliance on self-identification to determine grouping, however, is that there is a considerable number of nation-states who claim membership in a particular grouping, but who may not be accepted as fully legitimate members by the central core of that group. Is South Yemen a socialist state? Precisely when did Spain become a Western democratic state? Answers to these critical questions were left to each scholar to handle in the most appropriate way. Lijphart's comparative analysis of democratic political systems of Western type supplements his early work on the subject. Not only did he provide an analysis of the critical behavioral and legal structural differences among democratic nations, he also discussed the implications of the typology to political behavior, regime stability, and the pluralist basis of the state. Koch's commentary applauds Lijphart's efforts but also indicates that there can be considerable consolidation in viewpoints and convergence of ideas across the chasms of geopolitical boundaries, specifically between Eastern

366

Conclusions

and

Summary

and Western Europe. Authoritarianism as a political grouping proves to be a more intractable conceptual problem than that presented by the democratic subset of nations. As Morlino demonstrates, there is considerable variation among the authoritarian nation-states and there is considerable inherent instability within this genre. The analyses and commentaries on socialist national groupings indicate that there is considerable disagreement regarding the essential characteristics of socialism and variation within this group. Topomin makes a strong case for claiming the socialist state as a distinctive global national subset, but Huang appropriately points out that variation within the socialist world must also be addressed and examined in more detail. My essay on a socialist typology does deal with the problems of intrasocialist variation and change, but it is correctly criticized by Jovicic for its incompleteness. Again, we are left with the problems, addressed by Marradi, of closure and the ranking of variables in the classification scheme. Chirkin's, Riggs', and Berg-Schlosser's analyses of Third World political systems each approach the task of typology-building for Third World countries very differently. Each, however, was forced to deal with the inherent problem that the category, Third World, is a catchall for nation-states that belong to neither the "first" nor "second" world power blocs and that they have little in common beyond this. Chirkin, using a well-reasoned Marxist approach, pinpointed four types of Third World systems, but, as Narain and Padia stipulate, important cultural differences among the Third World nations were not given sufficient weight. Riggs sidesteps the land mine of cultural variation and deals with significant issues such as regime stability. Berg-Schlosser's extensive data analysis provides strong evidence that certain types of Third World regimes perform better than others do. For both Riggs and Berg-Schlosser, however, political regimes make a difference in political and national performance. One significant and time-honored mechanism for differentiating among political systems is to examine the number of political parties and their impact on political behavior. Wiatr argues persuasively that political parties can be the first differentiating variable in the attempt to categorize and group political systems. However, as Bacot and Journes note in their commentary, counting parties is not enough. An analysis based on the number of parties must be linked to a general theory of change and the expected impact of party groupings on fundamental political behavior. Bebler points out in his chapter on military rule that there is a need to examine variation among military regimes. Not all military regimes are oppressive authoritarian dictatorships; not all military governments are antidemocratic; not all share a common set of political behaviors. Bebler also demonstrates that it is incorrect to classify all military regimes as transitory or to argue that changes in the military leadership are inconsequential to the

Jim Seroka

367

regime and populace. In the commentary on Bebler's work, Bienen notes these positive achievements, but he also recommends that the militarycivilian relationships should become a part of the central core of the typology. The chapters on geographic area classifications belie any tendency to minimize variation within these clusters. As highlighted earlier, variation within sub-Saharan Africa, the Arab world, and Latin America is substantial and may be as significant as variation among these and other geocultural groupings. Welch's review of African political systems conclusively d e m o n s t r a t e s the need to include g e o g r a p h i c v a r i a b l e s in t y p o l o g y construction as well as the difficulty in predicting and explaining change within a geocultural cluster. Nevertheless, at this time, African polities are conclusively in a group by themselves. K o r a n y ' s presentation is the purest form of a historically based geopolitical analysis that we provide. He makes a convincing argument that Arab political regimes march to their own drummer and fulfil distinctive purposes that can be traced to early forms of agriculture and tribal patterns of governance. The clear distinctiveness of the Arab world offers a major challenge to any transcultural analytical classification of political regimes. Huneeus's comparative study of Latin American political systems gives perspective to the bewildering diversity and rapidity of regime change that we witness in that area. This geocultural analysis, unlike its two predecessors, presents a classification and approach that may have wider applicability in the Third World. Indeed, as both Philips and Remmer are quick to point out, Huneeus's typology of Latin American systems should be expanded to i n c l u d e the M a r x i s t - o r i e n t e d r e g i m e s in Cuba and N i c a r a g u a , the English-speaking Caribbean nation-states, and relatively stable and relatively powerful Mexico. With these additions, a significant step will have been made toward a truly region-based political classification. In s u m m a r y , the e f f o r t s of the c o n t r i b u t o r s and c o m m e n t a t o r s demonstrate that the problem of categorization of political systems is still vibrant, far from exhausted and sterile. As this book demonstrates, scholars from around the globe and from every major political regime type are concerned about these issues, are eager to share ideas, and are open to new concepts, methods, and approaches. The worldwide community of political scientists concerned with classificatory issues seems prepared to cooperate in this endeavor. On this basis, if for no other reason, there is cause for optimism that the global problem of political classification may one day witness substantial success.

About the Contributors

Paul Bacot is maître de conférences in political science at the Université Lumière Lyon II in France. He is author of Les dirigeants du Parti Socialiste; he also coedited Les nouvelles idéologies, Atlas électoral de Lyon et de Rhône and has contributed to Géopolitiques des régions françaises. Anton Bebler is professor of political science at Edvard Kardely University in Ljubljana. He is author of numerous articles and books, including Military Rule in Africa, Marxism and Militaria, Defense and Security Aspects of Nonalignment, and Disarmament. Dirk Berg-Schlosser is presently professor of political science at Philipps University, Marburg. His major publications include Politische Kultur and Tradition and Change in Kenya. He is editor of Die politischen Probleme der Dritten Welt, coeditor of Vergleichende Politikwissenschaft and Politische Kultur in Deutschland, and co-author of Politische Stabilität und Entwicklung Einführung in die Politikwissenschaft. Henry Bienen is James S. McDonnell Distinguished Professor of Politics and International Affairs and director of the Center of International Studies at Princeton University. He is editor of the journal World Politics and author of many books on the politics of economic development, conflict and change, and the military in politics. Friedemann Büttner is professor of politics and contemporary history of the Middle East at the Free University of Berlin. He is author of The Crisis of the Islamic Order and coeditor of Reform and Revolution in the Islamic World; he is co-author of The Near East PLOETZ: A Reference Book on the History of the Arabo-Islamic World and Reform in Uniform? Military Rule and Development in the Third World (all titles translated from German). Veniamin Chirkin is head of the Department of State and Law in Developing 369

370

About

the

Contributors

Countries at the Institute for State and Law of the USSR Academy of Sciences. He is author of thirteen books, or political science, the theory of state and law, and comparative constitutional law, three of which have been published in English: Constitutional Law and Political Institutions, and the co-authored A Socialist Oriented State and Fundamentals of Socialist Theory of the State and Law. Richard Dale is associate professor of political science at Southern Illinois University at Carbondale. He has written and published extensively on Southern Africa, primarily Botswana, Namibia, and South Africa. Huang Zhong-Liang is associate professor of political science at Beijing University. He is a director of the Chinese USSR and Eastern Europe Institute and is a member of the council of the National Communist Movement History Institute. He is a specialist in socialism and politics, particularly the politics and political systems of socialist countries, and co-author of The Evolution of the USSR Political System and The International Communist Movement: A Concise Edition. Carlos Huneeus is professor of political science at the Catholic University of Santiago de Chile and senior researcher at the Center for the Study of Contemporary Reality. He is the author of La Union de Centro Democratico y La Transición a La Democracia en Espana and has written extensively on comparative politics and public policy. Zachary Irwin is associate professor of political science at Pennsylvania State University in Erie. He has published extensively on church-state relations and human rights in Eastern Europe and the Soviet Union. Claude Journès is maitre de conférences in political science at the Università Lumière Lyon II in France. Among his published works are L'extremegauche en Grande-Bretagne and L'État britannique, Les nouvelles idéologies which he coedited, and La coutume et la loi, and Police et politique, both edited volumes. Miodrag Jovicic was professor of comparative political systems at the Faculty of Law, Novi Sad, until 1985 and is now a member of the Serbian Academy of Sciences and Arts. He is the author of several books in comparative constitutional law and political systems and numerous professional journal articles. Burkhard Koch is professor at the East German Institute of International Politics and Economics in Berlin. His publications concentrate on correlations between ideology and politics, and on East-West relations, with

About

the Contributors

371

particular emphasis on humanitarian aspects. His books include Neokonservative Ideologic und USA-Konfrontationsstrategie and the edited volume Ideologische Gegner-Partner im Friedenskampf. Bahgat Korany is professor in the Department of Political Science and director of the University Program of Arab Studies at the Université de Montreal. He has published four books and many articles, some of which have been translated into Arabic, Italian, Spanish, and Chinese. His first book, Social Change, Charisma and International Behavior, was awarded the 1976 Hauchman Prize. Arend Lijphart is professor of political science at the University of California, San Diego. His field of specialization is comparative politics, and his current research entails the comparative study of democratic regimes and of electoral systems. His most recent books are Democracies: Patterns of Major itarian and Consensus Government in Twenty-One Countries, Power-Sharing in South Africa, and, coedited, Choosing an Electoral System: Issues and Alternatives and Electoral Laws and Their Political Consequences. Alberto Marradi is professor of methodology of social research at the University of Florence. He has published widely on méthodologie issues and is currently the vice-president of the International Political Science Association/International Studies Association Committee on Conceptual and Terminological Analysis. Leonardo Morlino is professor of political science, Faculty of Political Science, University of Florence, and coeditor of Revista Italiana de Scienza Politico. His recent publications include Come cambiano i regimi politici: Strumenti di analisi, and Dalla democrazia al autoritarismo: II caso spagnolo in prospettiva comparata. Iqbal Narain is professor of political science at the University of Rajasthan, Jaipur. He has written and edited a number of books, including From Dyarchy to Self-Government, Pattern of Urban-Rural Relationships, State Politics in India, Panchayati Raj Planning and Democracy, Fourth General Elections in India, vols. 1 and 2, Literature, Social Consciousness and Polity, and Development, Politics and Social Theory. He is at present member and secretary of the Indian Council of Social Science Research, New Delhi. Recently, he has been writing on India's educational system in the light of his experience as a vice-chancellor of Rajasthan and Benaras Hindu universities. Chandrakala Padia is associate professor of political science in Banaras Hindu University, Varanasi. Her field of specialization is political theory with

372

About

the

Contributors

special reference to Russell. George Philip is reader in Latin American politics with a joint position at the London School of Economics and Political Science and the London Institute of Latin American Studies. His writings include The Rise and Fall of the Peruvian Military Radicals 1968-76, Oil and Politics in Latin America, The Mexican Economy. He is coeditor of Political Dilemmas of Military Regimes and editor of The Military in South American Politics. Karen L. Remmer is professor and chair of the Department of Political Science at the University of New Mexico and associate editor of the Latin American Research Review. She has a special interest in issues of regime change and public policy in Latin America. Her publications include Party Competition in Argentina and Chile and Military Rule in Latin America. Fred W. Riggs is professor emeritus at the University of Hawaii. He is a specialist in comparative politics and administration, with special reference to Third World countries and is best known for Administration in Developing Countries: The Theory of Prismatic Society and Thailand: The Modernization of a Bureaucratic Polity. He is the author or editor of approximately a dozen books and 200 journal articles. In recent years, he has directed much of his attention to problems of conceptualization and terminology. Jim Seroka is professor of political science at the Pennsylvania State University at Erie, Behrend College. His research interests include comparative communist systems and problems in public administration. He is editor of Rural Public Administration: Problems and Prospects, coeditor of Devoted Socialism in the Soviet Block, and co-author of Political Organizations in Socialist Yugoslavia. Bruce James Smith is associate professor of political science at Allegheny College. His interests include U.S. political thought, republican theory, and the psychology of public life. He is author of Politics and Remembrance: Republican Themes in Machiavelli, Burke, and Tocqueville and is currently working on a book titled The Anthropology of Justice: An Inquiry Into the Nature and Origin of the Political Sentiments. Boris N. Topornin is professor and director of the Institute for State and Law of the Academy of Sciences in the USSR. He is involved with comparative legal and constitutional issues and is the author of Politicheskaja Sistema Socializma (1972) and Konstitucija Stranji Sovetov (1982). Claude E. Welch, Jr., is distinguished Service Professor at the State

About

the Contributors

373

University of New York at Buffalo. His publications have focused on Africa, human rights, and the political roles of armed forces. Major books include No Farewell to Arms? Military Disengagement from Politics in Africa and Latin America, Human Rights and Development in Africa, Anatomy of Rebellion, Civilian Control of the Military, Military Role and Rule, Revolution and Political Change, Soldier and State in Africa, Political Modernization, and Dream of Unity: Pan-Africanism and Political Unification in West Africa. He is also editor of Armed Forces and Society. Jerzy J. Wiatr is professor of political sociology at the University of Warsaw. He has also been president of the Polish Political Science Association and vice-president of the International Political Science Association. He is author of numerous books on political and sociological theory, military sociology, and political sociology that have been published in Poland, the United States, the USSR, Japan, Yugoslavia, Hungary, Czechoslovakia, and the Netherlands. His most recent book is The Soldier and the Nation: The Role of the Military in Polish Politics, 1918-1985.

Index

Abboud, 267 Acción Democrática, 353 Accountability, 214 Adanson, 19 Administrative capability, 206 Administration, 210-211 Afghanistan, 144 Africa, 98,114,183, 267,289, 298 African political systems, 281-302 African socialism, 228 Agrarian Reform law, 313 Albania, 143,144, 147, 152 Algeria, 166,167,180,266, 269,271, 288, 293 Allende, Salvador, 231, 342 Ali, Mohamed, 315 Almond, G., 77, 84,174,206 Alvarado, V., 359 Amin, Idi, 263 Anglo-Dutch system, 268 Angola, 159,167, 271 Anthropology, 281,284, 285,288, 306308 Anticolonialism, 285 Antidemocratic youth, 349 Antifeudal stage, 123 Apathy, 349 Apter, D„ 15, 165, 166, 174, 286 Aquino, Corazon, 221 Arab League, 303 Arab political system, 303-335, 367 Arab nationalism, 303 Arab radicalism, 332 Arab socialism, 332 Aramco, 317 Arendt, H., 50 Argentina, 103, 113,182,214, 247,267, 277,337, 340, 342-343, 345-346, 348,352-355, 361-362, 365 Aristocracy, 7, 51,60, 66, 209,268 Aristotle, 6,15, 17,32,45,47,49, 52-62,

174,252-253, 365 Arithmetical criterion, 252, 257 Army, 97 Aron, R„ 256 Asad, Hafiz, 331 Asia, 97-98 Ataturk, Kemal, 263 Australia, 80, 267,270 Austria, 80-83, 255,268 Authoritarian crisis, 102, 104 Authoritarian rule, 202 Authoritarian systems, 173,179, 185, 289, 304,333, 337-339, 343-350, 355, 357,360-362, 364, 366 Authoritarianism, 89-115,264 Authority, 207,211 Autocracy, 255 Autocratic system, 167 Autonomization, 101 Autonomy, 92,225 Baath Party, 266,305,331, 334 Bacot, P., 366 Bahrain, 168, 303 Bangladesh, 202, 203, 266, 268, 278 Barongo, Y„ 289-290 Balandier, G., 285 Banks, A., 3 Bane, 267 Barton, 27-28 Base typology, 243 Bebler, A., 276-278, 366-367 Bedouin-based political system, 304, 307, 308, 313, 320, 330, 333 Behaviorism, 165 Belgium, 79, 82,255 Benin, 220, 269,29 Berg-Schlosser, 202-203,366 Berger, J., 22,27 Bertsch, G., 3 Bhutan, 166

375

376

Index

Bicameralism, 74-75, 80,137 Bicameral, 137 Bienen, H., 367 Bios politikos, 62 Bipartisan regime, 257 "Black September War" (1970), 331 Blalock, H„ 7 Blondel, J., 256-257 Bolivia, 220, 26 Bosert, T„ 357 Bossism, 214 Botswana, 181 Bottomore, T., 16 Boumedien, 267 Bourgeois model, 165,16 Bourgeoisie, 123,209,229,230-231, 263, 268 Bratton, M„ 113 Braudel, F„ 30 Brazil, 103,112-113,166,182,220,247, 267,277, 337,343, 345, 352-355 British West Indies, 253 Buddhist philosophy, 20 Bulgaria, 123, 129,130,136-138, 144-145 Bulgarian Agrarian People's Union, 129 Bunge, M., 16 Bureaucracy, 95, 97, 209-214, 217, 219, 222, 223, 227, 251 Bureaucratic authoritarianism, 352-355, 357, 359 Bureaucratic culture, 315 Bureaucratic models, 199 Bureaucratic policy, 214 Bureaucratic polity, 208 Bureaucratic power, 206,213 Burkina Faso, 269 Burma, 180, 203,227, 266,268, 277 Burundi, 269 Butler, D„ 76 Cabinet stability, 84, 90 Callisthenes, 60 Calogero, G., 16 Cameroon, 179 Camp David Agreements, 305 Canada, 80-82 Capecchi, V., 27, 28 Capitalism, 124,159,160,206, 209,212, 218,231,244,347 Capitalist-landowner systems, 166 Carnap, R., 35 Categorization, 364 Catholic Church, 179

CaudiUo, 181,182 Central African Republic, 269 Central America, 98, 343 Central authority, 311 Central Committee, 271 Centralized government, 74, 312 Chad, 269 Charismatic leader, 228, 309,310 Chariot, J., 250 Chazan, N„ 293 Chief executive, 214-216, 222-223, 227 Chile, 111, 113,180, 231,267, 337,340346, 348,352-355,361 China, 130,135-138,144-145,147,171, 221, 254,269, 271 Chinese People's Political Consultative Conference, 136 Christian Democrat, 342 Christian Democratic Italy, 253 Christian Democratic Union, 129 Circulation of elites, 292 Civil elites, 103 Civil liberty, 202-203,226, 299 Civil Liberty Index, 190, 194 Civil war, 337, 348 Civil-Military regimes, 103, 107 Civilian-Military relation, 276, 299 Civil society, 104, 257 Civil opposition, 106 Civil rights, 220, 225 Civil Rights Index, 194 Civilian-dominated political system, 261274 Civilian regime, 167 Civilocracy, 261-262 Class, 120-121,123,135, 170, 354 Class conflict, 164 Class structure, 156 Classification, 6,11,17, 53,174,249, 250, 255-256, 303-304,356, 363, 365,367 Clausewitz, Karl Von, 269 Clientelism, 354 Club of Paris, 293 Cluster analysis, 19 Coalition, 100-101,105, 111, 113 Coercive resources, 100, 349 Coleman, J., 285 Colonial experience, 202 Colonialism, 281-285,287, 291 Colombia, 180, 277,340, 352-355, 361 Comintern, 269 Communist party, 121,125-126, 128-

Index

130,137,144,148,150-151,156, 232, 245, 252, 257, 269, 271, 342 Communist Party of Cuba, 128 Communist Party of Czechoslovakia, 128 Communist Party of the Soviet Union, 118,120,125-126,128-130 Communist system, 185, 194-195 Comparative politics, 118, 206-207, 211 Comparative public administration, 210 Competence, 211 Competitiveness, 206, 244, 340, 342, 347 Congo, 159, 269 Consensual executive, 73 Consensus, 348 Consensus democracy, 71,76-77,79, 84, 86, 89 Conservative party, 254 Consociational democracy, 76-77,79, 89 Consolidation, 100-101 Constitution, 45-46, 66, 207 Constitutional cycle, 59 Constitutional form, 60 Constitutional Law, 118 Constitutional theory, 66 Constitutive system, 212, 214, 224, 234 Content analysis, 25 Control, 207, 211 Coombs, C., 35 Copt, 314 Corporate federalism, 80 Corporate-military regimes, 185 Corporatist, 356 Corruption, 60 Costa Rica, 181,221, 253,271, 337,356, 361 Council of Ministers, 131 Council of State, 127-128 Coup d'e'tat, 181,210, 214-215, 219-221, 278, 282, 288, 310, 331, 334 Crone, P., 309 Crossnational statistics 195 Cuba, 123,128-130,137,144,180,229, 269,352-355, 359, 361, 367 Cultural norm, 208 Culture, 218,250 Cumming, R. D., 61 Cybernetics, 161 Cynicism, 349 Cyprus, 221 Czechoslovakia, 127-128,129-131,136, 137-139,145

377

Dahl, R., 85,181,206 Dahomey, 293 Darius, 47-48,53 Death squad, 194 Debbasch, C., 257 Debt crisis, 354 Decentralization, 202 Decentralized government, 74 Decisionmaking, 283 Decolonialization, 183 de Gaulle, Charles, 266 Deinstitutionalization, 98 De Jourenel, B., 50,66 Delegation, 137 Demagogy, 7 Democracy, 7,47,49, 51-53, 56, 60, 84, 93,99,121,129,166-167,173, 202, 203,244,254, 257,310, 338-340, 343, 346, 350, 359 Democratic, 305,364 Democratic development, 338, 347,361 Democratic Peasants' party, 129 Democratic political system, 71-87, 88, 348, 365 Democratic quality, 90 Dendrogram, 18, 31 Dependencia theory, 354 Despotism, 47, 50 Deutsch, K„ 14, 143 Developing countries, 170, 184, 205 Development, 194,251, 319,347 Di Palma, G., 250 Dialectical materialism, 160,171 Diaresis, 15 Dichotomy, 210 Dictablandas, 97 Dictatorship, 226, 343, 356, 359 Dictatorship of the proletariat, 121-122, 123,156 Disparity reduction rate, 184 Dissident, 227 Diversity, 304 Divine sovereignty, 210 Divisionis, 22 Djibouti, 221 Dominant one-party system, 180,243, 255, 352-355 Dufrenoy, P., 16 Duverger, M„ 5, 249,251, 253,257, 286 Dyarchy, 292 East Germany, 127,130, 137,144,145 Eastern Europe, 97

378

Index

Easton, D„ 174-175 Eckstein, H., 3 Economic decline, 355 Economic development, 144,166, 184, 185,195, 202, 272, 338, 345 Economic growth, 185, 218, 354 Economic liberalization, 334 ECOWAS, 294 Ecuador, 220 Education, 354 Efficacy, 101 Efficiency, 205 Egypt, 182,183, 267,303, 305,310-314, 316-319, 331, 333-334 Elected assembly, 211-212, 216-217, 219,226-227 Election, 136,180-181,202 Electoral participation, 342 Electoral registration, 342 Electoral system, 95, 211-212 Elites, 104, 148, 288, 315, 348, 349, 357 El Salvador, 159 Emerson, R., 257 Engel, Friedrich, 117 Entrepreneur, 348 Environmental protection, 218 Ephorus, 60 Epistemology, 36 Equality, 56-57,167, 210 Equatorial Guinea, 264 Essentialist Fallacy, The, 31 Ethical standard, 203 Ethiopia, 144,166,178,228, 266 Ethos, 54 Eurocentrism, 264 Evénementialisme, 308 Everitt, B„ 17 Exclusionary authoritarianism, 362 Exclusionary democracy, 361 Executive, 89,175, 365 Executive dominance, 73 Executive-legislative balance, 73 Executive power-sharing, 73 Exhaustiveness, 14,23-26 Extensional classification, 12,17,19-20, 30-31

Federalism, 77 Feudal type, 166 Fifth Republic, 81, 216,245,266 Finer, S., 263 Finland, 82 First world, 4, 205 Fleron, F., 148 Foreign capital, 343, 347 Formalism, 207-208 Forms of property, 156 Fox, J., 22 France, 82,161,254, 256,286 Free societies, 64 Freedom, 57,225 Freedom House, 190,194 Freedom of press, 85,233 Freedom of speech, 242 Frei, Eduardo, 342 French Fourth Republic, 85 Friedrich, C„ 78 Fundamentum divisionis, 13-16,18-22

Faisal City, 319 Farmers' Union of Bulgaria, 136 Fascism, 257, 262 Federal government, 74 Federal-Unitary dimension, 80

Hayek, F., 35 Head of government, 212 Hegemonic party system, 243,245, 247, 252,253, 255,258 Heide, 15

Gabon, 271 Galtung, J., 18,19 Garrison state, 209,263 Geographic area classification, 367 Germani, G., 94 Germany, 80,82,83,268 Ghana, 160, 278,286, 293 Ghene, 15 Glasnost, 300 Gold Coast, 284 Gorbachev, M„ 172 Gordon, M., 7 Grand Coalition, 81 Great October Socialist Revolution, 122 Greece, 245,247,270 Green Revolution, 294 Grenada, 220 Gross domestic investment rate, 190 Gross National Product, 184,185,190, 195, 202, 203,276 Guatemala, 179 Guerrilla organization, 341 Guinea, 180,264, 269 Guyana, 227,271

Index

Helms, C., 311 Hempel, C., 12,15, 27, 33, 35 Herodotus, 48-50, 252 Hispanoamérica, 338 Historian of science, 117 Historicocultural context, 202 Historical-systematic analysis, 346 Hobbes, 58 Hodgkin, T„ 285, 286 Huang, 366 Hudson, M.C., 307 Human rights, 190,195 Huneeus, C., 352-355, 356-362, 367 Hungary, 129, 130, 138, 144, 145 Hungarian Socialist Workers' Party, 128 Huntington, S., 166,174, 299 Hussein, Saddam, 305 Huxley, J., 33 Hwan, Chun Doo, 267 Hydraulic society, 312-313, 320 Ideology, 97, 100,142,144,148, 167, 175,178,180, 206, 244, 251, 253, 254, 257, 258, 333, 357 Ikhwan, 313, 316 Imperial power, 233, 307 Inclusionary authoritarian, 361 Inclusionary democracy, 361-362 Inclusionary regime, 345 Inclusiveness, 206 Independence, 92, 206, 229, 281 Index of Political Rights, 175 India, 159,171, 202, 241-242, 253, 268, 271 Indirect rule, 283,287 Indonesia, 167, 266, 269, 277 Industrialization, 211, 218 Inequality, 347,362 Infant mortality rate, 144 Infitah, 319, 331, 334 Instability, 224, 278, 292, 346-348, 352355, 366 Institutionalization, 101,165, 207-208, 305, 337, 340 Intensional classification, 12-20, 23, 26, 32 Intensity of penetration, 309 International African Institute, 284 International Monetary Fund, 292 International Parliamentary Union, 147 Iraq, 182, 266, 268, 278, 312, 330,334 Iran, 111,235,271,309 Islam, 303,308, 311, 314-316, 318, 321,

379

334 Islamic States, 272 Israel, 80, 82,270, 318 Italian Christian Democracy, 258 Italy, 80, 82, 97, 244, 256, 257, 268 Ivory Coast, 179, 271 Jackson, R„ 212 Japan, 80, 82, 83, 253, 270, 272 Japanese Miracle, 218 Jaruzelski, Wojciech, 266, 267 Johnson, C., 148 Jordan, 271, 306, 309 Joumés, P., 366 Jovicic, M., 366 Judicial review, 75 Judiciary, 178, 181, 202 Junta, 214 Justice, 55, 58 Kampuchea, 143-145 Kanet, R., 148 Kaplan, A., 35,207, 209,213, 227 Kenya, 159, 167,180, 268, 271, 288 Kerala, 171 Kerekou, 267 Khadduri, M„ 309 Khaldun, Ibn, 308, 311 Khan, Ayub, 266 Khomeini, Ayatollah, 111 Kingship, 60 Kinship, 114 Kiribati, 159 Komsomol, 125 Koraat, 330-335 Koran, 367 Korea, 138, 144 Kuomintang, 229 Kuwait, 271, 316, 321 Labour Party, 254 Laos, 144 LaPalombara, J., 254, 256-257 Lasswell, H„ 207, 209, 213, 227 Latin America, 102,112,180, 194, 221, 231, 267, 281, 337, 364, 367 Laubier, P., 252 Lavau, G„ 249 Lazarsfeld, P., 15,16, 27, 28,35 Leadership succession, 148 Lebanon, 180, 269, 304, 330 Legal-Institutional models, 147, 365 Legislature, 85, 147, 223, 251, 365

380

Index

Legitimacy, 101-102, 104, 112, 175,178, 179, 208, 214, 216, 234, 291, 307, 341-343, 348 Lehner, F., 76 Lenin, V. I., 117, 122, 126, 162-163, 269, 309 Leninism, 120, 356 Levine, D., 355 Liberal democracy, 356 Liberal individualism, 155 Liberals, 89,129 Liberia, 269 Libya, 113,331 Life expectancy, 144 Lijphart, A., 90, 244, 254,306, 364-365 Limited pluralism, 93 Linz, J., 91, 93,95, 111, 333 Literacy rate, 190 Literary clubs, 125 Local politics, 287 Locke, J., 45 Löwenthal, R„ 173,185, 194 Loyalty, 348, 353 Lundberg, G., 33 Luxembourg, 81-82 Lycurgus, 60-61 Machiavelli, N„ 46, 54, 59,61, 63 Macridis, R., 3 Madagascar, 166, 269 Majoritarian democracy, 84-86,90 Malawi, 271 Malaysia, 159, 166,271 Mali, 269, 293 Malvinas (Falklands), 214,338 Maoism, 114,271,309 Mariam, Mengistu Haile, 266 Market economy, 156 Markovitz, L„ 299 Marradi, A., 363, 364-366 Martial Law, 266 Marx, K., 91, 117, 163, 171-172 Marxism, 134, 170-171, 174,180, 208, 228, 232, 243, 255, 269, 334, 366 Marxism-Leninism, 114, 119, 121-122, 126,128, 144, 150-151, 160, 162, 172, 352-355 Mass liberal democracy, 92 Mass media, 349 Maximum consensus, 72 Maximum majoritarianism, 72 Measurement, 34 Media, 181

Megabyzus, 47-48 Metacephaly, 251 Methodology, 170 Mexico, 180,221, 231,265, 268, 271, 337, 346, 352-355, 356-357, 361, 367 Middle class, 59 Middle East, 113, 185, 210, 305 Middleton, D., 284 Military coup, 235, 241 Military regime, 95, 98-107, 113, 160, 167,173,175, 179,181-185, 195, 202-203, 210-211,214, 217-219, 222, 225, 227, 245, 255, 261-274, 287, 291,292, 304-305, 317-318, 331, 334, 337, 339, 343, 345, 348, 349, 352-359, 361, 364, 366 Mendeleev, 7, 8 Minimal winning coalition, 81 Mobilization, 95-98,105-106, 112, 114, 280 Modernization, 112, 148, 154, 182, 203, 285, 287, 307, 338, 340 Monarchy, 7, 178,183, 185,190, 194195, 209-212, 215, 224, 234, 245, 255, 262, 271, 305, 318, 330, 333 Mongolia, 129, 131, 144, 262 Morgenthau, R., 286 Morocco, 178, 271, 293, 330 Movimento Democràtica Brasileiro, 112 Mozambique, 180, 271 Mubarak, M„ 267 Mujib, 203 Multicameralism, 137 Multidimensional party system, 74 Multicollinearity, 195 Multiparty system, 74, 82, 84, 85, 89, 129,153,167, 243, 252-258, 310, 341 Mutual exclusiveness, 13-14, 22-26 Naked Rule, 225, 234 Napolean, 315, 317 Narain, I., 366 Nasser, Gamal A., 263, 305, 310, 315, 318, 331 National Assembly, 127 National cohesion, 309 National Democratic party, 129 National front, 130 Nationalism, 124, 282, 285, 332, 333, 347 Natural regime, 60

Index

Natural regime, 60 Naturalness, 306 Near East, 267 Neoinstitutional framework, 206-207 Neo-Marxist theory, 208, 306 Nepal, 271 Netherlands, 80, 82 New socialist man, 147 New Zealand, 79, 82, 270 Nicaragua, 159,166,179, 231, 271, 337, 343, 352-355, 356, 359, 367 Niger, 159,166-167, 269,293 Nigeria, 278, 287, 290 Nile River, 312, 320 Nominal classification, 306 Nominal scale, 34 Noncompetitive system, 252 Nonparty regime, 245 Nordlinger, E., 263 Normativism, 165, 183 North Korea, 128,145,147,180 Numerical taxonomy, 19 Oil embargo, 318 Oil wealth, 355 Oligarchy, 7, 51, 53, 55-56, 179, 181183,190,194-195, 206, 341 Oligarchic political system, 341 Oman, 271 One-party system, 129, 216, 219, 228, 243, 247, 252, 254, 256 OPEC, 289 Oppenheim, F., 15, 27 Opposition, 105, 217, 219, 225, 226, 230-231, 233, 337, 340, 357, 359 Order, 205 Orescu, S„ 143 Organization of African Unity, 267, 293, 303 Orwell, George, 142 Otanes, 47,48 Ottoman Empire, 209, 311 Outliers, 195 Overseas Development Council, 184 Owusu, M„ 287 Pakistan, 166, 202-203, 227, 268 Palestine Liberation Organization, 303 Pan-Arabism, 310 Panchayat Raj, 202 Papau New Guinea, 159 Paraguay, 159,267, 343, 356, 361 Pariah entrepreneur, 209

381

Parliamentary system, 102, 136, 175, 213, 215-217, 219-220, 222-223, 227, 232-233, 268, 334, 345 Parteinenstaat, 254 Participation, 167 Party competition, 340 Party flexibility, 257 Party membership, 232 Party of the proletariat, 254 Party organization, 224 Party system, 138, 211-212, 215, 357 Pasic, N„ 255 Patron-client networks, 299 Peaceful coexistence, 203 Peasants, 120-121, 304, 320, 330, 333, 341 People's democracy, 122-123, 166 Per capita income, 218 Perez, Jimenez, 103 Pericles, 45 Peronism, 112,263, 342, 353, 361-363 Personal rule, 208, 222 Peru, 103, 182,228, 267, 337, 345, 357, 359, 361 Physical Quality of Life Index, 184-185 Philippines, 167, 179, 221, 264, 268 Piao, Lin, 271 Pinochet, General, 111, 221, 231, 345, 346, 352-355, 359 Plato, 15, 32,49-52, 54-57, 59, 61-62, 252 Pluralism, 82, 83, 86,91-92, 96, 98,112, 185, 246, 250, 256, 356 Plutocracy, 209 Poland, 111, 128,129,136-137, 144, 145, 147,152, 258, 266, 271 Polarization, 254 Police power, 225 Polish People's Army, 262 Polish United Workers' Party, 129, 136, 138 Politburo, 271 Political action committees, 231 Political coalitions, 99 Political culture, 255, 304, 309, 313, 314-316,346 Political decay, 60 Political democracy, 144 Political development, 145, 175, 202, 347-348 Political economy, 104 Political elite, 304, 309,317 Political instability, 361

382

Index

Political liberalization, 334 Political mobilization, 243 Political modernization, 165, 251 Political participation, 287, 337,339, 340, 342, 347, 348, 361 Political party, 95,113,168,243, 305, 348, 349, 364, 366 Political repression, 194, 202 Political stability, 58, 271 Political system, 160, 304 Pollution, 218 Polyarchy, 181-183, 190, 194, 202, 206 Polybius, 46, 59,64, 365 Polymorphous party, 258 Popular Front, 342 Popular initiative, 76 Popular sovereignty, 210, 256 Population density, 218 Populism, 293,339, 353,357-358 Poverty, 54 Powell, B„ 174 Power, 50 Portugal, 113, 247, 264, 265, 266, 270 Praetorian system, 183, 190, 194, 263, 343 Predominant party system, 253 Presidentialist system, 216, 219-224, 230-232 Presidium of the General Assembly, 131 Presidium of the Supreme Soviet, 127, 131 Prime Minister, 178, 213 Promajoritarian bias, 85 Propaganda, 98 Proportional representation, 74,223 Protestantism, 218 Pseudoconstitutionalism, 207, 214,217, 235 Public opinion, 54, 179 Public policy, 211,218, 357 Puppet, 212 Pye, L„ 250 Qaddafi, 266 Qatar, 303 Radical party, 342 Ranney, A., 76 Ratio relationship, 25-26 Rationalism, 210 Rattray, R. 284 Recruitment, 175 Redemocratization, 113, 348-350

Referendum, 75-76,102 Regime, 45, 57, 65, 208 Religion, 202-203, 246 Remmer, K., 367 Representative government, 75,127,210 Repression, 105, 348 Republic, 212, 219, 330, 333 Residual category, 26 Responsibility, 211 Revolution, 123,179, 215, 219-220, 231, 263 Revolutionary Movement of the Armed Forces, 345 Riggs, F., 251, 256, 366 Rokkano, S., 347 Role autonomy, 167 Roman state, 63, 309 Romania, 128-130, 137-138, 144-145 Rosberg, C„ 212 Rosenblith, W., 35 Rouquié, A., 93 Ruanda, 159,168, 269 Ruling elite, 225 Ruralism, 288,293, 308 Sadat, 305, 309, 318 Salah, Ahamed Ben, 331 Sartori, G„ 35,175, 253-257, 304, 305 Saudi, Al, 313, 316 Saudi Arabia, 159, 166,178, 271, 310312, 314, 316-319, 333 SADCC, 294 Schall, J., 66 Scheffler, I., 12 Scientific socialism, 121 Second world, 205 Secretary general, 213 Secularism, 211 Segmental autonomy, 78-79,81 Segmental economy, 80 Seiler, D„ 253, 257-258 Semicompetitive, 180 Senate, 211 Senengal, 271 Separation of power, 181, 216 Seroka, J., 154 Shah of Iran, 179 Shils, E„ 166 Sierra Leone, 267 Singapore, 180,271 Single party system, 175,179, 217,226, 282, 352-355 Size of party, 257

Index

Size of party, 257 Skilling, G., 148 Skupstina, 128 Slavery, 50 Social class, 54, 55, 58, 233, 347 Social Democrat, 89 Social Democratic party, 232 Social Democratic Sweden, 253 Social Democratic thought, 354 Social ecology, 306, 309, 311, 330 Social participation, 348 Social reform, 182 Social welfare benefits, 233 Socialist democracy, 121 Socialist internationalism, 122, 150 Socialist property, 120 Socialist systems, 117-132, 141-157, 160,175, 179-180,194-195,206, 212, 231-232, 244, 253, 263, 315, 321, 364, 366 Socialist Unity party, 128-129 Sociocultural norm, 202 Socioeconomic development, 93, 357, 360 Socrates, 45,49-51, 54,55,64, 365 Somalia, 160, 266, 293, 303, 306 South Africa, 244, 270, 299 South America, 268 Southeast Asia, 267 South Korea, 203, 267, 269, 354 South Vietnam, 220 South Yemen, 144, 365 Soviet of Nationalities, 131 Soviet of the Union, 131 Soviet Union. See USSR Spain, 113, 247, 270, 365 Spoils system, 230 Sri Lanka, 159,181, 202-203, 242,268, 271 Stability, 83-84, 90, 92, 205,244, 357, 364 Stalinism, 114, 147, 277, 309 State, 125-126,135-136 State property, 120 Stateless socialism, 125,150 Stroessner, General, 267, 343 Strong man, 181 Steady-state economy, 218 Stevens, S., 35 Structural asymmetry, 349 Structural functionalism, 147, 161, 284 Sub-Saharan Africa, 113,367 Sudan, 277, 287,303-320

383

Suez War (1956), 331 Suffrage, 359 Suharto, 266 Sumaria, 309 Supressed-party regime, 255 Surinam, 227-228, 269 Survivability, 206, 211, 234 Sweden, 244 Swiss Federal Council, 73 Switzerland, 79-83,269 Syria, 266,269, 278, 314, 331,334 Taiwan, 219, 269 Tanzania, 180, 271 Taxonomy, 16,18, 24,29-35, 206, 219, 303 Technocracy, 209, 263 Territorial federalism, 80 Terry, Fernando Belaunde, 345 Thailand, 266 Theocratic autocracy, 245, 255, 271, 310, 330 Therbom, G., 360 Third world, 8, 157-242, 271-272, 299, 304, 307, 321, 367 Thrasymachus, 49 Thucydides, 45 Timocracy, 51, 55 Togo, 267 Tonga, 166 Topornin, B„ 133-137, 366 Torgerson, W., 35 Totalitarianism, 97, 141,148, 150, 154, 167, 244, 254, 257, 262 Toulemonde, B„ 253, 257 Tiryakian, E., 36 Town vs. countryside, 306 Trade unions, 95,125,128,161,163 Traditional regime, 97,203, 262,281, 285, 287-288, 291 Transitional regimes, 97,99-107,166 Tribalism, 114, 286, 313, 316 Tunisia, 166,179, 269,271, 331 Turkey, 113, 277 Turnbull, C., 285 Two-party system, 82, 85,243,245 Typology, 6, 14, 15, 18, 26-30, 35, 71, 88,97,142-143,149, 151,153-155, 157, 166, 174, 208, 213, 244, 246247, 253, 261-274, 281-302, 291292, 303-304, 306, 330, 338, 358, 363-366 Tyranny, 7, 52, 53, 208

384

Index

Uganda, 263, 276,278 Ulama, 313, 316,318 Unicameralism, 74,75, 80 Unidad Popular in Chile, 353 Unitary government, 74 Union organizations, 348-349 United Arab Emirates, 271, 303 United Arab Republic, 314 United Kingdom, 79, 89, 215, 252, 286, 338 United Nations, 190, 227, 261, 303 United Peasants' party, 129 United States, 80, 82-83, 216, 220, 224, 229-231, 253, 271, 276, 286, 298 Unwritten constitution, 75 Urbanization, 308 Uruguay, 103,113,180, 267, 337, 345, 346,355 USSR, 114, 118, 120-121, 123-124,126127,129-131,135, 137-138,141, 144, 180, 215, 229, 253, 257, 262, 267,271, 309-310,319 Vargas, 353 Varma, S.P., 165 Velasco, Alvarado, 361 Venezuala, 103, 220, 352-356, 361 Videla, Gonzalez, 341, 345 Vietnam, 128-129,144-145, 180,269 Virocrat, 209 Voegelin, C., 27 Vote of No Confidence, 233 Voter, 217 Wahhabism, 314 Wealth, 54, 55, 209 Weber, M„ 29, 175, 218 Weight of political party, 253 Weiner, M„ 256-257 Welch, C„ 263,298-302, 367 West Germany, 161 West Pakistan, 203 Western democracy, 250 Western Sahara, 294 Western Samoa, 159 Westminister Model, 79, 80, 83,88 Whitaker, C„ 290 Wiatr, J., 251,255, 258, 366 Wolin, S„ 48, 50, 62 Women's associations, 128 Work-collective, 121 Working class, 120

World Bank, 183 World War I, 300 World War II, 284, 333 Xenophon, 60 Yemen, 159,166,167,310, 321, 330 Yemeni civil war, 310 Young, C„ 295 Youth leagues, 128 Yugoslavia, 123,128-131, 135, 137-138, 144-145, 147, 269 Zaire, 159,168, 269,278 Zartman, L„ 305, 321 Zelditch, M„ 27 Zimbabwe, 271 Zoon politikon, 61