The Third Wave 080612346X, 9780806123462

Citation preview

THE THIRD WAVE Democratization in

the Late Twentieth Century

By SAMUEL P. HUNTINGTON

UNIVERSITY OF OKLAHOMA PRESS

:

NORMAN AND LONDON

OTHER BOOXS BY SAMUEL P. HUNTINGTON The Soldier and the State: The Theory and Politics of

Civil-Military Relations (Cambridge, MA, 1957) The Cummon Defense: Strategic Programs in National Politics (New York, 1961)

Political Order in Changing Societies (New Haven, 1968) American Politics: The Promise of Disharmony (Cambridge, MA, 1981)

Library of Congress Cataloging-in-Publication Data Huntingion, Semucl P, The third wave : demoeratizaliun bn Use late twenieth cemury ¢ Samuel I Huntinglon. P cm. — (The Julion ). Rothluum distinguished keelure series ; v. 41 Includes Inbhepraphical references and ind 1SBN 0-60b1-2346-X (alk. puet) . v 2 " P h century. 3. World ok 4. World politcs— 3 . 1. Tide. 5. Serwes, JCanHEb 1991 mSagn—do 050600 ar The Third Wave: Democratization in the Late Tewentieth Cemtury is Volume 4 in the Julian J. Rothbaum Distinguished Lecture Series.

This book was wrilten under the auspices of the Harvard Universily Cenler for hiternational Affairs and lIts John M.

Stralegic Studies.

Olin Institute for

The paper in this book meets the guidelines for permanence and dnmg:.ily of the Commitice on Production Guidelines for Book

Longevity of the Council on Library Resources, Inc.&

Copyright 1998 © by Samuel P. Hunlinglon. Published by the Universily of Oklahoma Press, Norman, Publishing Division of the

University. All rights reserved. Manufactured in the U.S.A. First

edition.

To the memory of Warren and Anila

CONTENTS

Foreword, by Carl B. Alberl Preface 1.

page xi it

What? The Start of the Third Wave

The Meaning of Democracy

3

3

The Waves of Democratization The Issucs of Demacratization

3

13 26

Why? Explaining Waves 31 Explaining Democratization Waves 34 Explaining The Third Wave 40 Declining Legilimacy and the Performance

Dllemma 46 Economic Development and Economic Crises Religious Changes 72 New Palicies of Exlernal Actors

Transformations

124

vii

59

85

Demonstration Effects or Snowballing 100 From Causes to Causers 106 How? Processes of Democralization

Authoritarian Regimes 110 Transition Processes 121

3

109

viii

THE THIRD WAVE Guidglines for Democratizers «:

Reforming Authoritarian Systems

Replacements 142 Guidelines for Democralizers 2:

141

Overthrowing Authoritarian Regimes

Transplacements 151 Guidelines for Dentocriizers 3: 4.

Negotiating Regime Changes

149

162

How? Characteristics of Demoxcratizalion The Third Wave Democrahzauon Syndmme

promise dewfl

and the Partici 165

Elections: Stunning and Otherwise

5.

Low Levels of Violence

192

How Long?

164 164

174

208

Consolidation and Iis Problems 208 The Torturer Problem: Prosecute and Punish vs.

Forgive and Forget

211

Guidelines for Democratizers 4: Dealing with Autkorfiarian Crimes 231 The Praetorian Problem: Rebellious and Powerful Militaries 231

Guidelines for Democralizers 5: Curbing Military Power, Promoting Military Professionatism 251

Contextual Problems, Disillusionment, and

Authorilarian Nostalgia

253

Developing a Demosratie Polilical Culture

258

Institutionalizing Democratic Political Behavior Conditions Favoring Consolidation of 6.

New Democracies

Whither? “Third Wave Causcs:

270

260

Continuing, Weakening, Changing? 281

A Third Reverse Wave?

265

290

CONTENTS Furlher Demacratization: Obstacles and Opportunilies

ix 294

Economic Development and Political Leadership

Notes Index

315 317 343

FIGURES

1.1 2, 1 3. 1 5. 1

Democratization Waves and Economic Development as a Democratization Political Groups Invelved in Problems Confronting Third

Reverse Waves Factor in Democratization Wave Countries

~=

TABLES

3. 2 5. 1 ~

5.

w

5.

i

5.

-

6.

Democratizalion in the Modern World Economic Development and Third Wave Democratization Authoritarian Regimes and Liberalization/ Democratization Processes, 1974-90 Leadership Change and Reformn, 1973-90 Post-World War 1l Democratic Experience of Third Wave Countries Levels of Ecoromic Development of Third Wave Counlries External Environment and Democratic Consolidalion in Third Wave Countries Inauguration of Democracy in Third Wave Countries GNP per Capita 1988: Upper- and Middle-Income Nondemocratic Countlries x

62 113 132 271 272 273 275 314

FOREWORD

&w107, is devoted lo the themes of

govi

v and

ed

and

cmzen participalion in pubhr affzms values to which Julian J. hb has been g| his life. His lifelong ication to the Uni y of Ok the state, and his country is a tribute to the ideals 10 which the Rothbaum Leclure Senes is dcdncaled The books in the series make an end: ion to an und ing of American democracy. CasL B. ALBERT Forty-sixth Speaker of the United States House of Representatives

PREFACE

THIS BOOK IS ABOUT an imporlant—perhaps the most important-—global polilical devel of the late ieth century: the transition of some thirly countries from nondemocratic to democratic political systems. It is an effort to explain why, how, and with what immediale consequences this wave

of deinacratizalion occurred between 1974 and 1990. This book partakes of both theory and history, but it is neither a work of theory nor a work of history. It (alls somewhere in between; it is primarily explanalory. A good theory is precise, austere, elegant, and highlights the relalions among a few conceptual variables. Inevitably, no theory can explain fully a single event or group of events. An explanalion, in contrasl, is inevitably complex, dense, messy, and intellectually unsatisfying. It succeeds not by being austere but by being comprehensive. A good history describes chronologically and analyzes convincingly a sequence of evenls and shows why one event led to another. This sludy dees not do that ¢ither. Tt does not spell out the general course of democratization in the 19705 and 19805 nor does it describe the democ[t instead altempls to explain and lo analyz.e a particular group of regime transilions that occurred in a limiled period of time. In social science jargon, this study is neither nomothetic nor idiographic. Both theorists and historians are hence likely Lo find it unsatisfyxid

xiv

THE THIRD WAVE

ing. It does nol provide the generalizalions the lormer (reasure or the depth the latter prefer.

In its approach this study thus differs significantly (rom

several of my other books. In these other works 1 tried to develop generalizations or theorjes about the relations between key variables, such as polilical power and military professionalism, political participation and polilical institutionalizalion, and political ideals and political behavior. The proposilions about these relationships were generally set forth as timeless truths. In this book, however, the generalizations are limited to a discrete class of events of the 19705 and 1980s. One key point in the book, indeed, is that the democralizations of the third wave differed from those of carlier waves. In writing this book I was at times tempted to propound timeless Lruths, such as “Replacements are more violent than transformations.” [ would then have 10 remind mysell thal my evidence came from the limiled historical

cases | had studied and that I was writing an explanatory,

not a theoretical, work. So 1 had (o abjure Lhe timeless present tense and inslead wrile in the pasl tense: “Replacements were more violent than transformations.” Wilh a very few exceplions, | have done this. In some cases, Lhe universalily of lhe proposition seemed 56 clear I could nol resisl Lhe temp{ation (0 state il in more timeless terms. In addition, however,

almost no proposition even applied to all third wave cases. Hence

the reader will find words

like “lended

1o be,”

“gen-

erally,” “almost always,” and other such qualifiers scaltered throughout the text. In its final form, the above-cited proposition should read, “Replacements usually were more violent than lransflormalions.” This book was written in 1989 and 1990, as the class of evenls with which it was concerned was still unfolding. The book thus suffers all the problemc of contemporaneily and must be viewed as a prelimii and of these regime transitions. The book draws on the works of historians, political scientists, and other scholars who have

PREFACE

xv

written detailed monographs on particular events. It also refies extensively on journalistic reports of these events. When the third wave

more

possible.

sati

of democratization comes

y

explanation

of thisph

to an end, a fuller,

will be

My previous siudy of political change, Political Order in

Changing Socielies, focused on the problem of political staDility. 1 wrole that book because [ thought polilical order was a good thing. My purpose was to develop a general social science lheory of why, how, and under what circumstances order could and could not be achieved. The current book focuses on democratization. I have written il because 1 believe that democracy is good in itself and that, as argue in chapter 1, it has positive consequences

for individual

freedom,

do-

mestic stability, international peace, and the United States of

America. As in Political Order, 1 have attempted to keep my

analysis as detached as possible from my values; at least that is Lhe case in g5 percent of this book. It did seem to me, however, that it might be useful occasionally to spell out explicitly the implications of my analysis for people wishing to democratize their socielies. Consequently, al five places in the book I have abandoned the role of social scientis(, assumed that of polilical consultant, and set forlh some “Guidelines for Democralizers.” If thal makes me seem like an aspiring democratic Machiavelli, so be il. The

immediate stimulus to wriling this buck

was the invi-

tation to deliver the Julian |. Rothbaum Lectures al the Uni-

versity of Oklahoma in Navember 1989, In those lectures [ set forth the principal themes of the book without, of course, all the empirical evidence to butiress those themes. The bulk of the manuscript was written in late 1989 and in 1990, and [ have made no attempl to inchude in the analysis any events that occurred after 1990. I am greally indebled to the Carl Albert Congressional Research and Studies Center al the University of Oklahoma and to its director, Dr. Ronaid M. Peters, Jr., for inviting me to deliver these lectures. My wife, Nancy,

xvi

THE THIRD WAVE

and I also want lo record how much we appreviale the unfailing courlesy and hospilality we received al the Universily of Oklahoma from Dr. Peters, Julian and Irene Rothbaum, Joel Jankowsky, and Speaker and Mrs. Carl Albert. While the lecture invilation precipilated the writing of this book, the materal in it has been germinating in my mind for some time. In a few places in the manuscripl, 1 have drawn on two previous articles: “Will More Countries Become Democratic?” (Political Science Quarterly 99, Summer 1984, pp. 191-218) and “The Modest Meaning of Democracy,” in Democracy in the Americas: Slopping the Pendulunt, edited by Rubert A. Pastor (New York: Holines and Meier, 1989, PP- 11-28). Between 1987 and 1990 a John M. Olin Fellowship in Democracy and Development made il possible for me lo devote much more lime and effort than would otherwise have been the case to research on the subject of this book. Many people have also contributed both consciously and unwiltingly Lo this manuscript. Since 1983 I have laught a course on modern democracy in the core curriculum at Harvard Lhat has focused on Lhe problems of democralic transitions. Both sludenls and leaching assistants alike wifl recognize that much of the malerial in this book has come oul of that course; my thinking on this subject is much beller than it would have been as a result of their comunents and criticisms. Mary Kiraly, Young Jo Lee, Kevin Marchiora, and Adam Posen provided indispensable help in researching the material in this book and in making order oul of my files on this subject. Jeffrey Cimbalo not only performed these lasks but in addition carefully reviewed the accuracy of text and foolnotes in the final stages of manuscripl preparation. Juliel fllackell and Amy Englehard! applied lheu' very considerable ing skills to this iently, rapidly, and a«uralely producmg many dralls and seemingly endless revisions of drafts. The manuscripl was read entirely or in part by several colleagues. Houcl\ang Chehabi, Edwin Corr, Jorge Doming; Frances Hagopian, Fric linger, and

PREFACE

xvii

Tony Smith provided thoughtful, quite crilical, and very constructive writlen commenls. The members of the Harvard comparative politics discussion group helped with a lively discussion of the first half of lhe manuscript. 1 am very grateful 1o all these people for lheir interest in my work and for the major ibutions lhey made to imp. I the quality of my effort. In the end, however, I remain responsible for the argument, the evidence, and the errors of this study.

Cambridge, Massachuseils February 1991

SamueL P. HUNTINGTON

THE THIRD WAVE

CHAPTER

1

WHAT?

THE START OF THE THIRD WAVE

THE THIRD WAVE OF DEMOCRATIZATION m the modern world began, imp ibly and twenty-five minules after midnight, Thursday, April zs, 1974, in Lisbon, Portugal, when a radio station played the song “Grandola Vila Morena.” Thal broadcasl was the go-ahead signal for the military units in and around Lisbon lo carry oul the plans for a coup d’elat that had been carefully drawn up by lhe young officers leading the Movimento das Forcas Armadas (MFA). The coup was arried aut efficiently and successfully, with only minor resislance from the security police. Mililary unils occupied key ministries, broadcasting stalions, lhe post office, air-

ports, and telephone exchanges. By late moming, crowds

were flooding the streets, checring the soldiers, and placing carnations in the barrels of their rifles. By late afternoon the deposed

diclator,

Marcello Caetano,

had surrendered

to the

new military leaders of Portugal. The next day he flew into exile. So died the dictatorship that had been born in a similar mililary coup in 1926 and led for over thirty-five years by an auslere civilian, Anlénio Salazar, working in close collaboration with Portugal’s soldiers. The April 25 coup was an implausible beginning of a world3

4

TIIE THIRD WAVE

wide movemenl tv democracy because coups d'etat more [requenlly overlhruw democmlu: regimes than |ntroducc them Il was an because the i of democracy, much less the triggering of a global democratic movement, was far from the minds of leaders of the coup. The death of the dictatorship did not ensure the birth of democracy. [t did, however, unleash a huge array of popular, social, and political forces that had been effextively suppressed during the dictatorship. For eighteen months after the April coup, Portugal was in turmoil. The MFA officers splil inlo compeling conservalive, mod and Marxist factions. The polilical parties covered an equally wide spectrum, from the hard-line Communisl parly on the left lo fascist groups on the right. Six provisional governments succeeded each other, each exercising less authorily than its predecessor. Coups and countercoups were aempled. Workers and peasanls struck, demonstraled, and seized factories, farms, and media. Maderale parties won lhe national elections on the anniversary of Lhe coup in 1975, but by the (all of that ycar civil war appeared possible between the conservalive north and Lhe radical south. The revolutionary upheaval in Portugal secemed, in many respecls, o be a replay of 1917 Russia, with Caelano as Nicholas I1, lhe April coup as the February Revolulion, the dominant groups in the MFA as the Bolsheviks, similar wide-

spread economic turmoil and popular upheavel, and even the equivalent of the Kornilov conspiracy in General Spinola’s un-

successful right-wing coup attempl in March 1975. The resemblance was not losl on acule observers. In September 1974 Mario Soares, loreign minister of the provisional government and leader of the Portuguese Socialisi party, met with Secre-

tary of State Henry Kissinger in Washington. Kissinger be-

rated Soares and other moderates for nol acting more decisively to head off a Marxisl-Leninist dictatorship. “You are a Kerensky. . . . I believe your sincerity, but you are naive,” Kissinger lold Soares.

WIAT?

5

“l cerlainly don’t want Lo be a Kerensky,” replied Soares. “Neither did Kerensky,” shot back Kissinger.? Purtugal however, turned out lo be different from Russia. kys won; d t Soares went on lo become prime minisler and laler pmsldenl And the Lenin of the Porluguese revolution, lhe person who at the crucial momenl deployed disciplined force to produce the political result he desired, was a taciturn prodemocracy colonel named Anténio Ramalho Canes who on November 25, 1975, crushed the radical leftist elements in the armed lorces and ensured the [uture of democracy in Portugal. The movement loward democracy in Porlugal in 1974 and 1975 was dramalic bul not unique. Less obvious demacratic stirrings were occurring elsewhere. In 1973 in Brazil leaders of the outgoing guvernment of Gen, Emilio Médici developed

plans for political distensio or “decompression” and in 1974

Gen, Ernesto Geisel commilled his new government Lo start-

ing the process of political opening. In Spain Prime Minister

Carlos Arias cautiously moved the Franco dictatorship in a liberalizing direction while Lhe country awaited Lhe death of the diclalor. In Greece lensions were building up in the colonels’ regime thal led to its downfall in mid-1974 and, laler thal year, to the firstd ically electedg in the new wave of transitions, During (he fnllowing lifteen years this democratic wave became global in scope; about thirty countries shilled from authoritarianism to democracy,

and at least a score of other countries were affected by the democratic wave.

THE MEANING

OF DEMOCRACY

The transitions to democracy between 1974 and 1990 are the subject of this book. The first step in dealing with this subject is la clarify the meaning of democracy and democratization

as they are used in this book.

The concept of democracy as a form of government goes back to the Greek philosophers. lts modern usage, however,

6

THE THIRD WAVE

dates from the revolutionary upheavals in Western society at

the end of the eighteenth century. In the mid-twentieth century three general approaches emerged in the debates over the meaning of As a form of g democracy has been defined in terms of sources of authorily lor governmenl, purposes served by government, and procedures for constituling government. Setious problems of ambiguily and imprecisiun arisc when democracy is defined in terms of either source of authority or purposes, and a procedural definition is used in this study.? In olher governmental syslems people become leaders by reason of birth, lot, wealth, violence, cooptation, learning, appointment, or examinalion. The central procedure of democracy is the selection of Ieaders (hrough competitive clections by the people they govern. The most important modern formulation of this concept of democracy was by Joseph Schumpeler in 1942. In his pathbreaking study, Capitalism, Socialism, and Democracy, Schumpeter spelled out the deficiencies of what he lermed the “classical theory of democracy,” which defined democracy in terms of “the will of the people” (source) and “the common good” (purpase). Effeclively de molishing lhese approaches to the subjecl, Schumpeter advanced whal he labeled “another theory of demucracy.” The “democratic method,” he said, “is that institutional arrangemenl for arriving at political decisions in which individuals

acquire the power lo decide by means of a competitive slruggle for the people’s vole.”*

For some while after World War 11 a debate went on between those determined, in the classical vein, to define demacracy by source or purpose, and the growing number of theorists adhering to a procedural concept of democracy in the Schumpelerian mode. By the 19705 the debale was over, and Schumpeler had won. Theorists increasing Iy drew disi between rationalislic, ulopian, id: of democracy, on the one hand and ernpmca] descriplive, ional, and on the other, and

WHAT?

7

concluded Lhat only the latter type of definition provided the analylical precision and cmpirical referents that make the concept a useful one. Sweeping discussions of democracy in terms of normative theory sharply declined, at least in American scholarly discussions, and were replaced by efforts to understand the nature of democralic instilutions, how they function, and the reasons why they develop and coflapse.

The prevailing efforl was o make democracy less of a “hur

rah” word and more of a commonsense word.® Following in the Schumpelerian tradition, Lhis study delines a twenlieth-century polilical system as democralic to the extent that its most powerful collective decision makers are selected through fair, honest, and periodic eleclions in which candidales (reely compele for votes and in which virtually all the adult population is eligible lo vote. So defined, democracy involves the lwo dimensions—conleslalion and pamu pation—thal Robert Dahl saw as critical to mocracy or polyarchy. It also implies Lhe exislence of those civil and political freedoms to speak, publish, assemble, and organize Lhat are necessary lo political debate and the conducl of elev_lor.il campmgns ol y provides a number of bench marks—grouped largely along Dahl’s two dimensions—that make it possible o judge to what extent polilical syslems are democalic, lo compare systems, and to analyze whether

systems

are becoming

more

or less democratic.

To

Lhe extent, for inslance, thal a polilical system denies voting

participation to parl of ils society-—as the Soulh African system did (o the 70 percent of its population (hat was black, as Switzerland did lo the 50 percenl of its population that was {emale or as the United States did to Lhe 10 percent of ns population that were hern blacks—it is und i y. a system is ic to the extent thal no opposilion is permitted in elections, or thal he opposition is curbed or harassed in what it can do, or that opposition newspapers are censored or closed down, or thal votes are ma-

8

THIRD WAVE

nipulated or miscounted. In any society, the sustained (ailure of the major opposition polilical party lo win office necessarily raises questions concerning the degree of competition permitied Ly the syslem. In the late 1980s, lhe free-and-fairelections criterion of democracy became more use(ul by the |ncreas|ng observalion of elections Ly international groups. By 1990 the point had been reached where the first clection in a democratizing country would only be generally accepled as legitimate if il was observed by one or more reasonably competenl and detached leams of international observers, and if lhe observers certified Lthe election as meeting minimal

slandards of honesty and (aimess.

The p to y accords with the commonsense uses of the term. We all know that military coups, censorship, rigged elections, coercion and harassment of the opposition, jailing of political opponents, and prohibi tion of polilical meelings are incompatible with democracy. We all know also that informed polilical observers can apply the procedural conditions of democracy to exisling world politia! systems and rather easily come up with a list of those countries thal are clearly democratic, those lhat are clearly not, and those that fall somewhere in between, and that with minor exceptions different observers will compose identical lists. We all know also thal we can make and do make judg: ments as to how governments change over time and that no

one would dispute the proposition thal Argentina, Brazil, and Uruguay were more democratic in 1986 than they were

in 1976. Political regimes will never [il per(ectly into inlelleclually defined boxes, and any system |of classlficahon has to accept the exi of line, and mixed cases.

Historically,

the Kuumlnlang

(K.MT ) system

on lm-

wan, for instance, of 2 ianism, democracy, and lotalilarianism. In additien, governments that had democratic origins may end democracy by abolishing or severely limiting democratic procedures, as in Korea and Turkey in the late 1950s and in the Philippines in

wliaT?

9

1972. Yel with all its problems, the classification of regimes in lerms of their degree of pracedural democracy remains a relatively simple task. 1f popular election of lhe top decision makers is the essence of demmraq', then tho critical point in Lhe process of democ-

is the

rep

of 2

that was not

chosen this way by one thal is selected in a {ree, open, and fair election. The overall process of demacratization before and after that election, however, is usually complex and prolonged. [t involves bringing about the end of the nondemocralic regime, the inauguration of the democratic regime, and then the consolidation of the democratic system. Li alization, in contrasl, is the partial opening of an authorilarian syslem shon of :hoosmg govemmenlal leaders lhmugh [reely P Tegimes may release political pnsonen, open up some issues (or public debate, loosen censorship, sponsor elections for offices that have little power, permit some renewal of civil sociely, and take other steps in a democratic direction, withoul submitting top decision makers to the electoral test. Liberalization may or may not lead to full-scale democratizalion. Several additional points need lo be made in defining democracy. First, the ilion of d v in terms of elections is a minimal definilion. Ta some people democracy has or should

have much more sweeping and idealistic connolalions. To them, “true democracy” means liberté, egalité, fraternilé, effec-

tive cilizen contyol over policy, responsible government, honesty and openness in politics, informed and rational deliberation, equal participation and power, and various other civic virlues. These are, for lhe mosl part, good (hings and people can, if they wish, define democracy in these lerms. Doing so. however, raises all the problems that come up with the defi-

nitions of democracy by source or by purpust. Fuzzy norms

do not yield uscful analysis. Flections, open, (ree, and fair, are the cssence of democracy, lhe inescapable sine qua non.

10

TIIE THIRD WAVE

Gi produced by elections may be i i corrupt, shorlsighled, irresponsible, domi by special interests, and incapable of adopting policies d d by the public good. These qualities may make such governments undesirable but they do nol make them undemocratic. Democracy is one public virtue, not the only one, and the relation of democracy to other public virtues and vices can only be undersiood if democracy is clearly distinguished Irom other characteristics of polilical sysiems. Second, conceivably a society could choose ils political leaders through democratic means, but these political leaders might not exercise real power. They may be simply the fronts or puppets of some olher group. To the extent that the mosl powerful collective decision makers are not chosen through elections, the political system is not democratic. lmplicit in the concep! of democracy, however, are limitations on power. In demacracies elected decision makers do not exercise tolal

power. They share power with other groups in sociely. If

those democralically elected decision makers become, however, simply a (acade [or the exercise of much greater power by a nondemocratically chosen group, then clearly that political system is not democralic. Legitimate queslions may be raised, for inslance, as to whether Ihe elected governments in Japan in the lale 19208 and in Guatemala in the lale 1980s were sufficicnlly dominated by their military as not to be

Iruly democratic. It is also, however, easy for critics of a gov-

errunent, whether (rom the left or the right, to allege (hal the elected officials are simply lhe “tools” of some other group or that they exercise Lheir authorily only on the suf ferance of and within severe constrainls sel by some other group. Such allegations are often made, and they may be true. Bul they should not be judged Lo be true until they have been demonstraled to be true. Thal may be difficult, but it is not impossible.

A third issue concerns the fragility or stability of a demo-

cratic political system. One could incorporate into a definilion

WHAT?

n

of democracy a concepl of stability or instilulionalization. This typically refers to the degree 10 which the polilical syslem may be expecled lo semain in existence. Stabilily is a central dimension in the analysis of any political system. A political system may, however, be more or less demecralic and more or less slable. Systems that may be approprialely classified as equally democralic may differ greatly in lheir stability. Thus, in its survey of freedom in the world published at the beginning of 1984, Freedom House quite reasonably classified both New Zealand and Nigeria as “free.” When thal judgment was made, freedom may well have been no less in the latter than il was in the (ormer. [t was, however, much less stable: a military coup on New Year's Day 1084 effectively ended Nigerian democracy. Democralic and nondemocralic systems may be created but they may or may not endure. The siability of a system differs from the nature of the system.® Fourth, there is the issue of whether to treat democracy and { racy as a dich: or conli variable. Many analysts have preferred the laller approach and have developed measures of democracy combining indicators of fairness of elections, restrictions on political parties, freedom of the press, and other criteria. This approach is use{ul for some purposes, such as identifying variations in lhe degree of democracy among countries (United Slates, Sweden,

France, Japan) that would normally he considered (o be democralic or variations in the degree of authorilarianism in

nondemocralic countrics. It does, howevcr, pose many problems, such as the weighling of i dic approach better serves the purpose of this study because our concern is wilh the transition from a nondemocralic regime toad ic one. De has, , been defined in this study by a single, relatively clear and widely accepted criterion. Even when analysts use somewhat different measures, their judgments as to which political systems are democralic and which are nol correlale lo an extremely high

12

THE THIRD WAVE

degme This sludy will, consequently, treat democracy as a variable, gnizing that there will be some betwixt-and-between cases (e.g., Greece, 1915-36; Thailand, 1980-; Senegal, 1974-) that may be appropriately classified “semidemnocracies.” Fiflh, nondemocratic regimes do not have electoral compelition and widespread voting participation. Apari from these shared negative characteristics they have little else in common. The category includes absolute monarchies, bureaucratic empires, oligarchies, aristocracies, constilutional regimes wilh limited sulfrage, personal despotisms, fascis! and comununist regimes, mililary diclalorships, and other types of governance. Some of these forms were more prevalenl in previous eras; some are relatively modern. In particular, lotalitarian regimes emerged in the twentieth century after the beginning of demnocratization and atlempt the mass mobilization of their cilizenry to serve the purpuses of the regime. Social scientists have drawn an appropriate and important dislinction between these regimes and traditional nondemo cralic authoritarian systems. The former are characterized by: a single party, usually led by one man; a pervasive and power(ul secret police; a highly developed ideology selting forth the ideal sociely, which the tolalilarian movement is commilted to realizing; and government penetration and control of mass communications and all or most social and economic organizations. A tradilional authoritarian system, on the

other hand, is characterized by a single leader or small group

of leaders, no parly or a weak party, no mass mobilization, possibly a “mentality”” but no ideology , limiled government, “limiled, not responsible, political pluralism,” and no effort lo remake society and human nature * This dislinction between and auth ism is crucial to understanding twenlieth-century politics. In order (o avoid the semantic awkwardness in repeated use of the term “nondemocratic,” however, this study uses the lerm “authori-

WHAT?

13

larian” (o refer to all nondemocratic systems. Specilic forms of nondemocratic or authorilarian regimes are referred to as one-party systems, totalitarian systems, personal dictatorships, military regimes, and (he like. THE WAVES OF DEMOCRATIZATION

Political systems with democratic characteristics are not limited to modern times. In many areas of the world tribal chiels were elected for centuries and in some places democratic political instilulions long existed at the village level. In addition,

the concept of democracy was, of course, familiar to the ancient world. The democracy of the Greeks and the Romans, however, excluded women, slaves, and often other categori of people, such as resident aliens, from parlicipalion in political life. The extenl to which ruling bodies were, in practice, responsible to even

thiy restricted public was also often

limited. Modern democracy is not simply democracy of the village,

the tribe, or the clly stale; il is democracy of the nation-state

and its (1 i with the d P of the nalion-slate. The mmal push toward democracy in the West occurred in the first hall of the seventeenth century. Democratic ideas and democratic movements were an imporiant, although not a central, feature of the English Revolution. The Orders af Ci icul, adopled by the citizens of Hartford and neighboring towns on January 14, 1638, were the “first wrilten constitution of modern democracy.”? By and large, however, the Purilan upheavals did not leave a legacy of democtatic institulions in either England or America. For over a century after 1660 government in both Places tended to become even more closed and less broadly representative of the people than it had been earlier. In a variely ol ways, an aristocratic and oligarchic resurgence occurred. In 1750 no democratic institutions al the national level existed in the Western world. In 1goa such institutions existed

Numberof |

CHlegory| i L 2 K

u

]

¢ T

Fimt

i

| Wove | Revene | “wave | Revewe | Weve

H 1

0

G

9

3

3

3

4

o

5


There is truth in these arguments. Iis form of government is nol the only impartant thing about a country, nor even

probably the most important thing. The distinction between

order and anarchy is more fundamental than the dislinction between democracy and dictatorship. Yel thal distinction is also crucial for several reasons.

First, political democracy is closely associated with (reedom

of lhe individual. Democracies can and have abused indi-

vidual rights and liberties, and a well-regulated authoritarian

stale may provide a high degree of security and order for ils cilizens. Overall, however, the correlalion between the existence of democracy and the existence of individual liberty is extremely high. Indeed, some measure of the latter is an essenlial component of the former. Conversely, the long-lerm effect of the operation of democralic politics is probably lo broaden and deepen individual liberty, Liberty iy, in a sense, the peculiar virtue of democracy. If one is concerned with liberly as an ultimaie social value, one should also be concerned wilh lhe fale of democracy.

Second, political sidblhly and form of government are, as

was pointed out, two different variables. Yet they are also in~ terrelated. Democracies are oflen unruly, but they are nol often politically violenl. In Lhe modern world democratic systems lend to be less subjecl Lo civil violence than are nondemocralic sysiems. Democralic governments use far less violence against their citizens than do authorilarian ones. Democracies also provide accepted channels for the expression of dissent and opposition within the system. Both government and opposition thus liave fewer incentives to use violence against each other. Democracy also contributes o

WIHAT?

29

stability by providing regular opp ities for changing political leaders and changing public policies. In democracics, change rarely occurs dramalically overnight; it is almost always moderate and incremenlal. Democratic systems are much more immune to major revolulionary upheaval lhan authoritarian

anes.

Revolulion,

as Che

Guevara

once

said,

cannot succeed against a government that “has come into power through some form of popular vole, fraudulent or nol, and maintains at least an app of legality.”* Third, the spread of democracy has implicalions for international relations. Historically, demnrraues have fought wars as ofien as authorilarian A itarian countries have fought democratic countries and have fought each other. From the carly nineteenth century down to 1990, however, democracies did nol, with only Lrivial or formal exceptions, fight other democracies.” So long as this phenomenon conlinues, the spread of democracy in the world means the expansion u( a zone of peace in the world On the basis of pasl an owver ic world is likely to be a world relalively free of internalional violence. If, in particular, the Soviet Union and China become democracies like the other major powers, the probabilily of major inLerslate violence will be greatly reduced. A permanently divided world, on Lhe other hand, is likely

to be a violent world, Devclop in and economics arc intensifying Lhe interactions among countries. Tn 1858 Abraham Lincoln argued thal “a house divided against itsell cannot stand. This government cannot endure permanently hall-slave and hall-free.” The world at the end

of the (wentieth cenlury is not a single house, but it is becom-

ing more and more closely integrated. Interdependence is the trend of the times. How long can an increasingly mlerdependent world survive part-d ratic and parl TFinally, and

more p

hially,

the future of

d

y in

the warld is of special imporlance to Americans. The United

30

THE THIRD WAVE

States is the premier democratic country of the modern world,

and its identity as a nation is inseparable from its commitmen! to liberal and democralic values. Other nations may fundamentally change their political systems and continue their existence as nations. The United States does nol have thal oplion. Hence Amen:ans have a special interest in the development of a global ial lo d The futures of liberly, stability, peace, and the United Stales thus depend, in some measure, on the fulure of democracy. This sludy does not attempt to predict that future.

It does attempt to shed light on it by analyzing the wave of democralization that began in 1974. It allempls to explore the causes of his scries of Lransitions (chapter 2), the pracesses by which the Iransitions occurred and (he strategies of the supponers nnd opponenls of democracy (chaplers 3 and 4), and

g the

new

d

cracies

(chap-

ler ,) lt ends with some specu.lahons on the prospects for the further expansion of democratic regimes in lhe world (chapter 6). In dealing with these topies, use is made of existing social science Lheories and gencralizalions in an effort to see which ones may help explain the recent transilions. This book, however, is not an efiort to develop a general (hcory of lhe precanditions of d or the p ILis not an attempl ta explain why some counlnes have been democracies for over a century while others have been enduring dictatorships. Its purpose is the more modest one of at(empling lo explain why, how, and with what consequences a group of roughly contemporaneous transitions to democracy occurred in the 1970s and 1980s and to understand what these transitions may suggest about the future of democracy in the warld.

CHAPTER 2

WHY?

EXPLAINING WAVES DEMOCRATIZATION WAVES AND REVERSE WAVES are manifestations of a more general phenomenon in polilics. Al times in history, similar events happen more or less simultaneously wilhin different countries or political systems. In 1848 revolutions occurred in several European countries. In 1968 studenl protests erupled in many countries on several continents. In Latin America and Africa military coups in different countries often have been bunched logelher in time. Electi in d produce a swing to the left in one decade and a swing 1o the right in the next. The long wave of democralization in the nineleenth century was spread over sufficienl time to distinguish it significantly (rom later democratization and reverse waves. Each of the latter,

however, occurred during a relatively briel period of time. The problem is to identify the possible causes of waves such as these in polilics. Let

us

assume

a universe

of six countries,

numbered

1

through 6. Let us also assume thal within a relatively short period of time a similar event, democratization, or x, happens in each country. Whal could have caused this outbreak of x's?

Several explanalions are pussible.

Single cause. Conceivably all six x's could have a single cause, A, which occurred aparl (rom events in any of Lhe six countries. This might, for instance, be the rise of a new su3

32

THE THIRD WAVE

perpower or some olher major change in the international distribution of power. It might be a major war or other imporiant event that would have an impacl on many other societies. Several Lakin American nations, for instance, either introduced democralic regimes or held new national eleclions in 1945 and 1946. Evidence suggests thal these x developments were in considerable measure a result of a single cause, A, that is, Allied victory in World War II:

Parallel development. The x’s could be caused by simifar developments

in the same

independent

variables (a,, a., etc)

manifesting themselves more or less simultaneously in alt six countries. Theorists have argued, for instance, that a country is likely 10 develop democracy when it passes certain eco-

nomic development thresholds, achieving a particular level

of per capila gross national producl (GNP) or a particular literacy rale. In this case, democralic progress within each

country is caused by something within and particular to (hat counh'y bul similar causes may also be at work more or less in other P ing similar results: a4 —x, f———— 2, —ex, a. ——x, G ——

—

ox,

WHY?

k3]

Snowballing. An important cause for x in one country may be the occurrence of x in another country. If the x's occurred absolulely simultancously, this would be impossible. Perfect simultaneity, however, is very rare, and the possibility of isolated simul y is b ing rarer. Knowledge of significanl political events is in:rzasmgly transmilted almost instantaneously around the world. Hence, event x in one country is increasingly capable of lriggering a comparable evenl almost simullaneously in a different couniry. Demonstration effects are increasingly possible. Some unique and even idiosyncratic cause, a,, in one counlry may cause X, in thal coun-

try, but X, may cause evenls in other countries that, in tum, may then havea snowballing effect on still other countries.

Prevailing nostrum. 1L is possible that the immediale causes of event x in different counlries may differ significantly. These different causes, however, could prompt a common response if the elites in the different countries share a common belief in the efficacy of that response, the zcitgeist's prevailing remedy or nostrum. Just as six individuals may more or less su-nullaneously take aspirin lo cure six very different physical S0 six may ly engage in similar regime transitions to cope with very different sets of problems: inflation in one, breakdown of law and order in another, deepening economic recession in a third, military defeat in a fourth, and so on. in this case, the specilic individual causes (a,, b;, £y, etc.) of polilical change act on a com-

mon set of political beliefs, z, to produce similar x responses:

b2

THE TIIAD WAVE

4 ——wg——, fo———z——=1, These four possible explanations of political waves are not exhaustive; nor are they mutually exclusive; nor are Lhey necessarily contradictory. All four factors may be at waork in any

one situation. They arc models of explanation to be kept in mind in attempling to explain political waves.

EXPLAINING DEMOCRATIZATION WAVES The dependent variable of this study is not democracy but democratization. The purpose is to explain why some counLries that were authorilarian became democralic in a particular period of time. The focus is on regime change, not regime existence, This study lhus differs from studies that deal primarily with the characterislics of socielies with democratic and nondemocratic governments.

A number of studies, for inslance,

have shown high correlalions between various social and economic lactors and the existence of democratic inslilutions. As Dank Rustow has emphasi however, a ge-

netic explanalion differs from a functional one.! Almost all

wealthy countries are democratic and almost all democracies are wealthy. Thal correlalion alone says nothing aboul causation, and if the democracies were wealthy for a considerable lenglh of time before Lhey became democralic (as, relalively speaking, mos! northern European countrics were), then wealth, by ilsel(, is probably not a sufficient explanation of their transition from nondemocratic to democratic politics. Slmnlarly, Pusloncally 2 high correlation has exmcd between and d yel many were Prol-

wHY?

35

estanl and nondemocratic for two or more centuries before they became democratic. To explain change in a dependent variable normally requires some (orm of change in the independent variable. The problem is complicated, however, by the fact that change in the independent variable may take the form of the persistence of the independent variable. Three years ol economic stagnation under an authoritarian regime may not engender its downlall but five years of stagnalion may do so. The cumulative effect of the independenl variable over lime eventually produces change In the dependent variable. Or, as Gabriel Almond has observed, “social and international change may continue for a long period and only begin Lo Lrigger change in the political system when a short-term kink or set of kinks occurs in the curve or curves.”? Change in this sense is obviously much more likely to have political effects when il involves independent variables such as economic and social trends rather than others. The dependenl variable is not only dynamic; it is also complex. People sometimes assume thal doing away with a dicip leads to the j ion of a y. In facl, however, nondemocralic regimes are more likely (o be replaced by other nondemocratic regimes than by demuocratic ones. [n addition, the factors responsible for the end of a nondemocralic regime may differ significantly from those Lhat lead to the creation of a democratic one. The economic (ailure

of an authorilarian regime may undermine that regime but Lhe economic success of an authoritarian regime may be more likely to create the basis for a democralic regime. Circumstances that

ik

to the initial

bli

of a demo-

cratic regime also may nolt contribute to its consolidation and long-lerm stabilily. At the simplesl level, democratizalion involves: (1) Lhe end of an authoritarian regime; (2) the installa-

tion of a democralic regime; and (3) the consolidation of the democratic regime. Different and contradictory causes may be responsible for each of these three developments.

36

THE THIRD WAVE

Analyzing |Lhe independent variable, the possible causes of democ also poses probl At one extreme is Lhe danger of lautology. Polilical ‘elites alter or overthrow authoritarian regimes and inslall and consolidale democratic ones. Why do political elites do this? They are, presumably, acling in terms of their interesls, values, and goals as they sce them. Il they want democracy, they will produce and gel democracy. Or as Rustow pul il, the creation of democracy requires {hat clites arrive al a “procedural consensus on the rules of the game.”* This (ocuses on what is probably the mosl immediate and significanl explanalory variable: the belic(s and actions of political elites. It is a powerful explanalory variable but not a salisfying one. Democracy can be created even if pevple do not want it. So it is not perhaps tautological to say that democracy will be created if people wanl democracy, but il is close to that. An explanation, someone has observed, is the place at which the mind comes to rest. Why do Lhe relevant

political

eliles

want

democracy?

Lnevitably,

the

mind

wants to move further along the causal chain. The di ion between independent and d denl variables is clearest if lhcy are of different orders, 1f as is oflen the cse, an economic variable is used to explain a political variable. The entire Marxist intellectual tradition tilts analysis in this direction. This is reinforced by the key-under-thelamppost factor. Economic data, including a variety of sla-

tistics on many different subjects, are available for a large number of societies, especially since World War It but for

Western societies back to the nincleenth century. Inevitably analysts are drawn 1o use Lhis dala and to see what correlational and causal conneclions may exist between economic factors and democratization. Sometimes such efforts are informed by theory; other limes they are not. Social scientists at limes speak of Lhe problem of overdeterminalion. By this, Lhey normally mean having a multiplicity of plausible theories lo explain an event and the consequent problem ol establishing Lhe relative validily of those

wiy?

37

theories. This is a problem, however, only for those conecrned with evaluating theories. The same problem does nol exist for those concerned wilh explaining evenls. In politics, almost everylling has many causes. Why did this candidate ralher than lhal candidate win an election? Clearly a variely of variables and theories behind those variables are necessary to explain something as simple as an eleclion outcome. To oceur historically, an event almost has lo be overdetermined theoretically. Such is clearly the case wilh democralization. Numerous theories have been advanced and numerous denl variables identi 10 explain di Amung the variables that have been said to contribute to democracy and democratization are the following: a high overall level of econornic wealth;

relatively equal distribution of income and/or wealth;

a market econom

P and social a feudal arislocracy al some point in the hlslory of sociely; the absence of feudalism in the saciety; a strong bourgeoisie (“no bourgeois. no democracy,” in Barringlon Moore’s succinct formulation); a strong middle class; high levels of literacy and education; an

rather than

y culture;

of political

fore the

Protestantism; socxal pluralism and strong intermediate

of po].\lual participation;

groups;

exp

democratic authority structures wilhin social groups, particu~

larly Lhose closely connected to palitics; Tow levels of civil violence; low levels of political polarization and extremisim; political leaders commilled to democracy: experience as a British colony;

itions

of

ion

an

F

occupalion by a prodemocratic foreign power;

influence by a prodemocratic foreign power; elite desire lo emulate democratic nations;

38

THE THIRD WAVE traditions of respect for law and individual rights; communal (ethnic, racial, religious) homogeneity;

communal (ethnic, racial, religious) heterogeneity; consensus on political and social values;

absence of consensus on political and social values. The lheories relating these factors to democracy and demoxcralization are almosl always plausible. Each variable and theory, however, is likely to have relevance 1o only a few cases. In the half cenlury after 1990, democratization occurred in India and Costa Rica, Venezuela and Turkey, Brazil and Botswana, Greece and Japan, The search for a common, umversa].ly present mdependcnl variable that might play a role in explai political d P in such different countries is almost cerl:nn ¢ to be unsuccessful if it is nat laulological. The causes of democratization differ substantially from one place to another and from one lime o

another. The multiplicily of theories and the diversity of experience suggesl propositions:

the probable validily of the following

(1) No single factor is sulficient lo explain the development of democracy in all countries or in a single counl (2) No single factor is necessary lo the developmemo( democracy in all countries. (3) Democratization in each counlry is the result of a combiunation of causes.

(1) The combination of causes producing democracy varies

from country lo country.

(5) The combination of causes generally responsible for ane

wave of democratization differs from that responsible for other waves. (6) The causes responsible for the initial regime changes in a democralization wave are likely to differ (rom those responsible for later regime changes in that wave. Reflecting on the diversity of societies that have democratic governments, Myron Weiner concluded that to explain democratizalion one should look at the “siralegies available 1o

why?

39

those who seek a democratic revolution.” * This advice appropriately highlights the crucial role of political leadership and political skill in bringing aboul democracy. It should not, however, lead to a total rejection of broader, contextual, social, economic, and cullural faclors in explaining democratic development. A chain or l'unnzl (choosc your melaphor) of ion exists; and i ocial, ic, cultural,

and, most immediately, political factors al operate, often in conflicting ways, either to (acilitate Lhe creation of democracy or to suslain authoritarianism. The causes of democratizalion are thus varied and (heir significance over lime is likely to vary considerably. This is not the place for any detailed historical analysis of what produced democratization before 1974. A brief summary of whal appear to be the principal causes of the firsl and second waves is, however,

in order if only to set the context for a more exlen-

sive discussion of the causes of the third wave. Economicd t jon, the emergence of the bourgeoisie and of a middle class, lhe development of a working class and its early organization, and the gradual decrease in economic inequality all seem to have played some role in the toward democrali in northern European countries in the nineteenth century. These were also, generally, countries where the intellectual ethos was shaped, in some measure, by locke, Dentham,

Mill, Montesquieu, Rousseau, and the impact of the ideals of

the French Revolulion. In the British selller countries—the Unlied States, Canada, Auslralia, New Zealand—-many of these same factors were at work, enhanced by much greater the k of existing slalus systems, and the more equal income distribution that was possible in frontier socit 5. Concenvablv, Prolestantism also democrati; quarters of the countries lhat developed democratic institutions before 1900 were overin religious The v.m-ry of the Western Allies "World War | and the

40

THE TIIRD WAVE

dismantling of empires thnl occurred after the war significanlly affecled d ion. The ipheral countries—Finland, Iceland, Treland—were relatively “successful in suslaining democratic systems; the more cenlrally located successor states of the Romanov, Hapsburg, and Hohenzollern empires were not. In brief, the primary factors responsible for the first wave of democratizalion appear to be ic and social the ic and social environment of the Brilish selller countries, and the vic-

tory of the Western Allies in World War I and the resulting breakup of the principal continental empires. Political and military faclors were clearly predominant in the second wave of democratization. Most of the counlries thal transiled to demacracy in this wave fall inlo one of three

categories. First, the victorious Western Allies imposed de-

mocracy on a number of countries: West Germany, ltaly, Japan, a major part of Austria, and South Korea. Second, many other ies moved in a d direction because the Weslern Allies had won the war. This category includes Greece, Turkey, Brazil, Argenlina, Peru, Ecuador, Venezuela,

Colombia.> Third, (he weakening of the Western slates by the war and the rising nationalism in their overseas colonies led them to iniliate the process of decolonization. A significant number of the new stales starled off as democracies and a somewhat lesser number sustained democralic institulions

for a respectable period of time. The viclory of the established

Weslern democracies in World War 11 and decolonization by those democracies afler the war were thus largely responsible for the second wave. These were historically discrete events. The third wave had to result from a differenl mix of causcs. EXPLAINING THE THIRD WAVE

g third wave democratizali requires answers to two quzsnons Tirst, why did some thirly countries with au-

wHY?

q1

thoritarian systems but not about one hundred other authorilarian countries shifl to democratic political systems? Second, why did regime changes in these countries occur in the 19705 and 19808 and nol al some other lime? With respect to the first queslion, whether or nol countries transited [rom authorilarianism could depend on the nature of their authoritarian regimes. In fact, however, the regimes thal moved toward democracy in the third wave were a diverse lot. They included one-party systems, military regimes, personal diclalorships, and the racial oligarchy in South Africa. Within each calegory of regime type, some countries did not democratize during the (iftecn years after 1974; China and Vietnam among one-parly systems; Burma ard fndonesia among mililary regimes; lraq and Cuba among personal dic: tatorships. The nalure of the authoritarian regime, consequendy, cannol explain why some regimes transited to democtacy and others did not. A different approach Lo answering this question could focus on the hislories of regime changes in those countries that did democralize. In the cyclical paitern, countries alternated back and forth belween democratic and authoritarian systems. Thn paltern was parlicularly prevalent in Latin America, such as Argenlina, Brazil, Peru, Bolivia, and F:uador, but il characterized other countries such as Turkey and Nigeria. These countries tended to osclllate beIween more pof; ic governme d more conservative mlhtary regimes. Under a demoxratic regime radicalism, corruption, and disorder reach unacceptable levels and the military overthrow it, 1o considerable popular reliel and acclaim. In due course, however, the coalilion supporting the military regime wravels, the military regime fails to deal cHeclively with the country’s economic problems, professionally inclined military officers become alarmed at the politicization of the armed forces, and, again to greal popular relief and acclaim, the military willidraw from or are pushed out

42

TIIE THIRD WAVE

of office. In these countries, the change of regime thus performs the same funclion as the change of parties in a slable democratic system. The counlry dees nal alternate between democratic and authoritarian political sysiems; the allernation of democracy and aulhorilarianism is the country’s political system. A second paltern of regime change was the second-try pattern. A counlry with an authoritarian system shills to a democralic one. Eilher the democralic system (ails because the country lacks the social bases (or democracy, or the leaders of the new democratic syslem pursue extremist policies that produce a draslic reaction, or some cataclysm (depression, war) undermines the regime. An authoritarian government then comes Lo power lor a grealer or shorter period of time. Eventually, however, a second and more successful effort is made to introduce democracy, with success increasing al least in parl because democratic leaders learned from the previous unsuccessful experience with democracy. In varying ways, a number of countnes—(ormany, ltaly, Austria, Japan, Vi la, Colombi stable democratic systems in the second wave after suffering reversals in their carlier efforts. Spain, Portugal, Greece, Korea, Czechoslovakia, and Poland will fit inla Lhis second-try pallern il their third wave democralic regimes stabilize. A third pattern was inlerruyled democracy. This involves ies that developd ic regimes that exist for u relallvely sustained period of time. At some poinl, however, izalion, or other conditions develop and lead lo the suspens:on of democralic processes. In the 1970s, demaocracy was suspended in India and the Philippines by democratically elected chief executives, in Uruguay by elected leaders in cooperalion with the military, and in Chile by military leaders overthrowing an elecled regime. These countries’ lengthy experience with democracy, however, made itimpossible for the political leaders who suspended democracy to do

wiy?

43

away entirely with democralic practices. In all four cases, they eventually feit compelled to submit to some form of popular vote, which they lost. A fourth pallern of change involved Lhe direct transition from a slable authorilarian sysiem 10 a stable democratic system, either through pradual evolulion over lime or the abrupt replacement of the former by the lalter. This pattern lypified the transitions of the first wave. If their democracy is consalidated third wave efforts in Romania, Bulgaria, Taiwan, Mexico, G la, Ll Salvador, Hond and Ni would approximate his pattern. Finally, there is the decolonization patlern. A democratic country imposes democratic inslitutions on its colonies. The colony becomes independent and, unlike most former colonies, successfully mainlains ils democratic institutions. Papua New Guinea was one third wave case. As Myron Weiner has pointed out, this pattern perlains primarily to former British colonies, most of which became independenl in the second wave.* Those lelt to become independent and democralic

in the third wave were mosily small and mostly insular. They included

Antigua

and

Barbuda,

Belize,

Dominica,

Kiribali,

Saint Christopher-Nevis, Saint Lucia, Saint Vincent and the Grenadines, Solomon Islands, Tuvalu, and Vanuatu. With Lhe possible exception of a very few remaining colonies (e.g., Hong Kong, Gibraltar, the Falklands), these countries are the

last legacy of the British Empire to democratization. Because

of their small size they are, unless stated to the contrary, ex-

cluded from analyses of third wave countries in this study. If A and D are used to represent refatively stable and longTasting authorilarian and democralic regimes and a and d are used to represent less stable and shorter-lived regimes, these five patterns of regime development may be portrayed as follows:

(1) cydlical: a—d-- a—d—a—d (2) second-try: A—d

--a—D

44

THE THIRD WAVE (3) interrupted democracy: A—D- -a—D (4) direct transition: A—] a—D. (5) decolonizatiol

The countries of the third wave encompass all five patterns of regime change. Twenty-three of iwenty-nine countries thal

democratized between 1974 and 1990, however, had had pre-

vious experience with democracy. In some cases, (his experience was distant in lime; in some cases il was brief; in some cases it was both distanl and briel. At some point. however, it occurred. Mosl of the countries with authoritarian systems in 1974 that did not democralize by 1590 had no previous experience wilh democracy. Thus an excellent predictor in 1974 of whether a counlry will an authoritarian government would become democralic was whether it had been democratic.

By

1989,

however,

the

third

wave

entered

a second

phase and began o affecl countries withaut previous sig-

nificant democratic experience, including Romania, Bulgaria, the Soviet Union, Taiwan, and Mexico. This posed a crucial question. To whal exlent would the third wave go beyond the first and second waves? Would counlries Lhal had not experienced demoxcracy in the past become stable democracies in the [ulure? Plausible answers to the quesiion why some countries but not others transiled o democracy are not necessarily answers to the second question of why these tansilions tovk place when they did and not at some other time. It does secm unlikely that the clustering of transilions within a decade and a hal( could be purely coincidental. It appears reasonable Lo assume that Lhese transitions were produced in parl by common causes affecling many countries, by parallel develupments within several countries, and by the impact of early transilions on later ones. Previous democratic experience, however, does nol explain why the shilt to democracy in lhese countries came m the 1970s and 1980s. Similarly, some the d ilions of the 19805 to a decply

winy?

45

felt and widespread “yearning for freedom” by people oppressed by authoritarian rulers. The presence of Lhis yearning may distinguish countries that did democratize {rom those thal did not, but it cannol explain why democratizalion occurred when it did. As the evenls of 1953, 1956, 1968, and 1980-81 show, Fasl Europeans had been yearning lor freedom for decades; they only gol il. however, in 1989, Why then and nol earlier? In olher countries, people may nol have had a yearning for freedom earlier in their history but developed il in Lhe 19705 and 1980s. The problem is to explain why this desirc emerged then. Analysis is forced to look for other d P that may have produced it. The question lo be answered is: What changes in plausible mdependenl variables in, most probably, the 1960s and 19708 prod the dependenl variable, d regime changes in the 1970s and 1980s? Five such changes seem 10 have played significant roles in bringing about the third wave transilions in the countries where they occurred and when they occurred: (1)the‘

ing

legil

of

ilarfan

sys-

tems in a world whéte democratic values were widely accepted, the dependence of those regimes on performance legilimacy, and

the undermining of that legitimacy by military defeats, economic

failures, and the oil shocks of 197374 and 1978-79; (2) the unprecedented plobal economic growth of the 19605, which raised living standards, increased educalion, and greally expandud the urban middle class i many countries; (3) the striking changes in the doctrine and activities of the Catholic Church

maniflesled

in the Second

Vatican Council

in

1463 -65 and the transformation of national churches [rom defenders of the stalus quo to opponents of nuthoritarianism and proponenis of social, economic, and political reform; (4) changces in the

policies of external actors, including in the

late 1960s the new attitude expanding its membership, ginning in 1974 toward Lhe mocracy in other countries,

of the European Community loward the major shift in U.S. policies bepromotion of human rights and deand Gorbachev’s dramatic change 1n

46

TIIE THIRD WAVE

the late 19808 in Sovie! policy toward maintaining the Soviel empire; and

) “snowballing”

or d

on

effects,

d by new

means of i jonal ication, of the first ilions to democracy in the third wave in stimulating and providing mod-

els for subsequent efforls at regime change in other countries. DECLINING LEGITIMACY AND THE PERFORMANCE DILEMMA

Legjitimacy is a mushy concept thal polilical analysts do well to avoid. Yel it is essential to underslanding the problems confronting authoritarian regimes in the late (wenlieth century. “The strongesl is never strong enough to be always the master,” observed Rousseau, “unless he transforms strength into right and obedience into duty.” Whence springs the “right” of authoritarian leaders to rule and the “duly” of their people 1o obey? In the past, tradition, rehglon lhr dwme nghl of lungs and social deference provi for ratic rule. In an age of Inerate, mobilized populations, these traditional rationales for authoritarianism lose Lheir efficacy. In modern limes authoritarianism has been justified by nation alism and by ideology. The cfficacy of the lormer as a basis for nondemocratic rule, however, depends in large part on the existence of a credible encmy to the national aspiralions

of a people. Nationalism is also a popular force and can

equally well fegiti as ian rule. The principal i ] ion for authoritarianism in modern times has been Mandsm-Leninism. It provides a rationale for one-party dictatorship and rule by a small self-perpeluating bureaucratic elite. Most authorilarian regimes in the lale twenlieth century, however, were not communist regimes. Together with the communist regimes they faced major problems in eslablishing and maintaining their legitimacy.

‘The victory of the Western Allies in World War Il produced

the second wave of democratization in practice. It also pro-

wWHY?

47

duced an even more pervasive and lasting change in ihe i tellectual environment of politics. People in most countries came to accept—if not Lo implement— the cheloric and ideas of democracy. A world democralic ethos came into being. Even those whose actions were clearly antidemocratic often juslified their actions by democratic values. Explicil argument againsl democracy as a concept almost disappeared from public debate in most countries of the world. “For the first time in the history of the world,” a UNESCO report noled in 1951,

“no doctrines are advanced

as antidemocratic.

The ac-

cusalion of anlidemocralic aclion or attitude is frequently directed against others, bul praclical politicians and political theorists agree in slressing the democratic element in the inslitutions lhey defend and the theories Lhey advocate.”” The pervasiveness of democratic norms rested in large part on the commitment to those norms of the most powerlul country in the world. The principal alternalive source of legitimacy, Marxism-Leninism, was espoused by the second stronges| power. Communisls, however, regularly paid tribute 1o the strength of d iv values by emphasizing the democratic elements in their ideology. by employlng democratic phraseology, and by playing down the role of the vanguard Leninist party and the diclatorship of the prolelariat. Many authoritarian regimes in the 1970s also confronted legitimacy problems because of the previous experience of their country with democracy. In a sense, the body politic of

their society had been infected with the democratic virus, and even if the previous demacratic regime had nol been terribly

successful, the belief remained that a truly legitimale governmenl had to be based on democratic practices. Authoritarian rulers were thus impelled to justify their own regimes by democratic rhetoric and claim that their regimes were truly democratic or would become so in the future once (hey had deall with the immediale problems confronting their sociely. The legilimacy problems of the authoritarian regimes varied with the nalure of the regime. One-parly systems that

48

THE THIRD WAVE

were a product of indigenous political development, such as in revolutionary communist slates, Mexico, and the Republic of China, had a more secure basis of legitimacy. Ideology and nalionalism could be joinlly harnessed lo bullress the regimes. In countries, as in Eastern Europe, where communism and Hie one-parly system were imposed by external forces, regimes could benefit from the ideology but nol from nalionalism, which was always a polential source of inslability. Over time, however, communist governments found it more difficult to invoke communist ideology to support their legilimacy. The appeal of the ideology declined as the state bureaucracy stagnated and socioeconomic inequalities ossified. Communist ideology also became the major obsiacle to economic growth and frusirated the abilily of the regime lo legitimate itself on the basis of economic performance. In communist slates, Marxism-Leninism thus inilially provided idevlogical legilimacy bul, when this weakened, ManxismLeninism made il impossible lo develop legitimacy bascd on economic per(ormance. Other things being equal, the legitimacy of mosl regimes declines over time, as choices are made, promises are unrealized, and frusirations develop. In mosl cases, lhe coalition supporting the regime also disintegrates over lime. Democratic systems, however, renew themselves through elections, which make it possible for a new coalition to come lo power wilh new

policies and new

promise for the future. In

conlrast, sell-renewal is a major problem for authoritarian regimes, and Lhe absence of mechanisms for self-renewal contributes significantly Lo the erosion of the legitimacy of those regimes. This problem is worst, of course, in personal dictalorships, where

sell-

1 (short of re

i

is impos-

sible due to the nature of the regime. In a few cases authoritarian regimes did develop mechanisms for regular replacement ol their lop leaders and thus for at least [imited renewal. In Mexico and in Brazil the principle thal no president could succeed himsell was well-insti-

WHY?

49

lutionalized. The benefits of such a system of routinized succession were at least twofold. First, it encouraged key ligures within the authorilarian establishment to hope Ihat next lime

around they mighl have a chance at the top office or a high

office and hence reduced their incentives lo become obstrepcrous or lo overthrow Lhe exisling leadership. Second, regular succession in the Lop leadership posilion made possible and even probable changes in policy. In Mexico, for decades presidents more of the Righl allernaled wilh presidents more of the Left. Somewhat similarly in Brazil, two loosely structured coaliti the Sorb group and the nationali compeled for power within the militaty. The Sorbonne group came to power after the 1964 coup; il was displaced when Gen. Artur Costa e Silva assumed Lhe presidency in 1967; it relurned to power with Geisel in 1974. Mechanisms and processes such as lhese made il possible for these regimes to avoid some of lhc dysfunclional aspects of authoritarianism and also provided means for g their 1 in at least a limited way. Each new president offered new promise il only because he was different from the previous president. From 1929 to 1989 Mexico had a uniquely stable authorilarian reglme requiring only a medest amount of mpressmn because

its

was

enhanced both by its

Y ide-

ology and by the regular change in ils politica! leaders. Indigenous communist regimes had the former bul not the lalter; Brazil had the latter but not the former.

The legitimacy problem of the military regimes and person-

alistic dictatorships of the 1970s, particularly those created during the second reverse wave, typically evolved through three phases. The displacement of the democratic regime by an authorilarian regime was almost always greeted with a sense of great relief and overwhelming approval by the public. In this initial phase, the new regime benelilted rom “negative legitimacy™ deriving (rom the failures of the demo-

cralic regime and its apparent differences from the democratic

regime. The new regimes typically justified themselves on the

50

THE TIIRD WAVE

grounds that lhey were combatling communism and internal subversion, reducing social turmoil, reestablishing law and order,

climinating

corruption

and

venal civilian politicians,

and enhancing national values. purpose, and coherence. The Greek colonels in 1967, (or instance, legitimaled themselves through appeals ta “the ideology of the anticommunist state”; and in ils initial years the Brazilian military regime similarly soughi legilimacy on N the basis of “the “anti’ appeals: . K st oy Inevitably, negative legitimacy declined with time. The authorilarian regimes of the 19605 and 1970s were almos! inescapably driven to look to performance as a principal if not Lhe principal source of legilimacy. In some cases, as in Peru and the Philippines, Lhe leaders of the authuritarian regimes promised economic and social reform. In most other cases they promised economic growlh and developmenl. The effort to base legitimacy on performance, however, gave rise to what can be lermed the performance dilemma. In democracies the legitimacy of rulers usually depends on the exlenl lo which (hey meet the expectations of key groups of volers, thal is, on their performance. The legilimacy of the system, however, depends on procedures, on Lhe abilily of the volers to choose their rulers through elections.

In office, the rulers,

eventually, fail lo perform, they lose legitimacy, Lhey are defeated in elections, and a different sel of rulers lakes over. The loss of performance legitimacy by the nuders leads to the reaffirmation of the procedural legitimacy of the system. In authorilarian systems other than one-parly syslems, however, no dlslmchon was possnblc belween rulcr leglbmacy and regime d the legitimacy of both the rulers and the Ieglumacy of the syslem. In their analysis of eight major crises of political development ranging (rom Lhe British Reform Acl of 1832 to Cardenas’s economic reforms in the 1930s, Gabriel Almond, Scolt C. Flanagan, and Rnberl J ‘Mundl found that every case was pe compo

WHY?

51

nenls (dep , food sh and famine).”"? Unsaus(a:lory economic pexfom\anm similarly played a key role in producing the crisis of authoritarian regimes in the 1970s. The drive for social and economic n:fom\ smgnaled quickly in the Philippines and was 1} d in Peru. The abilily of many authoritarian governmenls lo derive legiti from ic growth was und d bY the increases in oil prices in the 1970s and by the economic policies Lhe authoritarian governments followed. The oil price hike of 1973-74 lnggered a global economlc ion. It raised the y of democracy in the wilateral world of [urupe North America, and Japan, and il significanlly undermined the efforts of Third World authoritarian regimes to use economic performance to bolsler Lheir legitimacy. Counlries such as the Philippines, Spain, Portugal, Greece, Brazil, and Uruguay were parlicularly hard hit because of their overwhelming dependence on imporied oil. The second ol price increase in 1979 aggravaled the situation. In West Germany, Britain, France, Canada, and the United Stales, incumbenl parties were turned oul uf affice. In the Third World, the ability of the remaining authoritarian regimes tv bolster their legitimacy through ic pert was further k d. The oil price hikes and their economic consequences deserve a significan! portion of the credit (or the weakening of authoritarianism in the 1970s and early 1980s.

With rare exceptions, the policies adopled by authoritarian

governmenls to deal with the oif and debt crises often made the ion worse, p depression, inflation, low or negalive growth rates, expanded debt, or some combinalion of these conditions. and hence further undermining the regime’s legilimacy. The Philippines provides an example: The New Society’s economic foundation began (o crack wilh the oil crisis of 1973-1974 and the global recession that fol-

lowed. The Philippines, which imported go percent of its oil,

52

THE THIRD WAVE saw ils encrgy costs quadruple, while the prices exports fell. With the secord oil price squeeze of nomic slide accelerated. Marcos responded with ing and spending, doubling Manila’s foreign dcbl

of commodity 1979, the ecomore borrowbetween 1979

and 1983. Almost hall the debt was short-term, and . . - international lenders became nervous about the Philippines: Dur-

ing the last (our months of 1982 alone some $700 million in credits were withdrawn.*

From 1980 on per capila income in the Philippines declined sleadily. InA tina the ic policies of Martinez. de Hoz behween 1976 and 1980 created an artificial boom thal could not lasl. Imports became so cheap that local industry was collapsing under the competition. Exporls became so expensive that agriculture was priced out of the world markel. . . . In 1981, the balloon burst. . . . The economy plunged into recession almosl overnight. In nine months, both inflalion and unemployment exploded upwards. The peso, under iremendous speculalive pressure,

fell in value more than 400 percent. Argenlines holding dollar debts suddenly found that it took five times more pesos to an them back. They could not meel their payments. . . . Panicked savers, meanwhile, slarted a long-lerm run on the banks. The country’s reserves fell precipilously."

In Uruguay Lhe military regime in the early 1980s ran up the second largesl per capita foreign debt in Latin America and produced a four-year recession that cul real salaries to half

what they had been a decade earlier. In Portugal, the huge

cosls of its colonial war combined with the oil price increases and economic mismanagement to produce similarly unhappy economic results. One

of

the

economy. By had reached

other

casuallies

[of the

war]

was

Portugal's

the lime of the coup, the country’s inflation rale

30 percent (the highest in Western

Europe),

its

trade deficit was the worst ever, and unemployment was ris-

wHY?

53

ing—despile a sleady exodus of emigranis seeking work or avoiding conscription. Wilh nearly half of government spending going to the mili tary, Portugal’s rate of “(ixed” invesimenl--the kind that cre-

ales jobs and exports—twas the lowest in Western Europe. .

Portugal, which imports Bo percent of its energy and more than hallils food, was lul espedially hard by (he global bout of

recession and inflation thal

followed the 1973 Organization of

Petroleum Exporting Countries (OPEC) oil embargo. As Portugal’s exporl markets went soft, the cost of everything from

codfish to bullfight tickets raced ahead of wages. Though both unions and strikes were illegal in the New State, Communistled workers starled some 40 ma|or walkouls in 1973. Plants

owned by

and

Telegraph,

British Leyland, and other firms clused down. M Greece

experienced

Grundig,

4

growth under the inilitary regime thal look over in 1g67. The

new military group that assumed power at the end of 1973, however, “showed itself quite incapable of dealing with Lhe pressing problems confronting the country. . . . Inflation continued unchecked, and Greece, with few indigenous sources of energy, was particularly severely affected by the il crisis thal (ollowed the Yom Kippur War.”** In Peru, the “military regime’s elforts to spur the counlry’s economic growth . . .

misfired badly. Produclivity declined in both agriculture and

industry, real wages fell, open unemploymenl rose, inflalion climbed, and Peru’s public debl skyrocketed.”” Even Brazil

had problems with economic performance. As the regime began gradually to democratize in the later 1970, economic

problems increased, and “there was a marked delerioration in (he ability of government to deliver economic growth and to promise a better fulure for all Brazilians.” By 1982 large proporlions of the Brazilian public were atiributing these deficiencies Lo the policy failures of the government. Communis regimes were relatively insulated from oil price hikes and other developmenls in the world economy, al-

54

THE THIRD WAVE

though Poland and Hungary did runup substantial debts. Their poor ily a result of the command econamies that the Sovnel Union imposed on them after World War Il. During the 19505 these economies grew al very rapid rates. Growth slowed in the 1960s and slagnated in the 1970s and 19805. Economic failure engendered tremendous disaffection from and some opposition to the existing polilical-economic system. By itsell, however, this economic slagnation was not enough (o produce movement in a democratic direction. Tt did nol become a factor promoting democralizalion in those countries unlil the Soviet Union allowed that to happen. In Eastern Europe polilics was in command, and Soviet support inured the communist regimes (o the q of their poor i performance. Military failure conlribuled to the downfall or weakening, of al leas! five authorilarian regimes between 1974 and 1989, The military forces of two personal dictalorships became disaffecled because they were fighting insurgencies that they saw little chance of defealing. The obvious inability of the Porluguese government and military forces Lo win their colonial wars was the underlying cause behind the formation of the MFA and the April coup. In the Philippines, a somewhat similar

Reforim

of the

Armed

Forces

Movemenl

(RAM)

tacked the cronyism, inefficiency, and corruption had caused in the Philippine military. The p

al-

Marcos of the

growing strength of Lhe communist guernllas of the New Peo-

ple’s Army (NPA) during the last years of the Marcos regime contributed generally (o the weakening of the regime. The large costs of the Afghan War and the inability of lhe Sovict military to bring it to a success(ul conclusion contributed to the political liberalization Gorbachev brought to the Soviel Union. In Greece and Argentina conflicts provoked by military regimes produced defeal and the downfall of those regimes, The legjitimacy of an aulhoritarian regime was also under-

wHY?

55

mined if il did deliver on ils promises. By achieving its purpose,

it losl its purpose.

This reduced

the reasons

why

the

public should support the regime, given the other costs (e.g.. lack of freedom) connected with the regime. It promoted uncertainty and conflict within the regime aboul what new purposes il should pursue. In Argentina in 1980 and 1981, for inslance. the economy was in a tailspin. Al lhe same time, the regime had eliminated the A illas and reslored order (if nol law) to Lhe country. This removed a major reason for support of the regime, and the military governmenl “showed signs of ennui, precisely because it had achieved one of ils main objeclives: the defeat of the armed guerri]la groups.” Somewhal simitarly, in Brazil in 1974, “given the unequivocal establishment of order, the regime was forced to begin either the long-term mshluhonallzahon of its power or, y, the beginning of the li process.”*s A similar snruauon existed in Uruguay after the military g limi d the Tup Because khexr legitimacy was based on p criteria, auth regimes lost legitimacy if lhe\' did not perform and also lost it il they did perform. Conlronted with this erosion of legilimacy, authoritarian leaders could and did respond in one or more of five ways. Eirsl, they could simply refuse to acknowledge their increasing weakness with the hope or the conviction they would somehow survive in power. The weakness of feedback mech-

anisms in most authoritarian regimes and the delusions of

many personal dictators reinforced these tendencivs. Neither the hope nor (he belivt, however, was likely lo be warranted, Second, lhe regime could altempt to survive by becoming inive, by, in effect, replacing ing duty wilh coerced obedience. This often mqulmd a shift in regime leadership, such as happened in Greece in 1973, Argenlina in 1981, and China in 1989. [I the leaders of Lhe regime substantially agree on such a course, they may be able to delay significantly the of their decli

56

TIE THIRD WAVY

A Lhird option was lo provoke foreign conflict and attempt lo reslore legilimacy by an appeal lo nationalism. In the spring of 1974 the loannidis regime organized a coup which overthrew Archbishop Makarios on Cyprus and installed a governmenl favorable to enosis (union) with Greece. The Turks invaded Cyprus and [oannidis altempted to rally the Greek army and people to fight the Turks. The Greeks were, however, militarily incapable of taking on (he Turks, and their mililary commanders refused to do s0. The colonels’ regime (hus fell in humiliating fashion as a result, in effect, of a strike by ils own officer corps. In Argentina, the legitimacy of the military regime reached a low poinl in 1982 as a resull of its economic (ailure, and Gen. Leopoldo Galtieri altempled to reestablish support for his government by his invasion of the TFalklands. If he had been mililarily successful, he would have

become a major hero in Argentine hislory. His failure and the British recovery of the island precipitated the transition Lo democracy in lhe following year. The eflorls by authorilarian leaders to bolster fading legitimacy by foreign war face an inherent obstacle. The military forces of a military regime are involved in politics, may lack

an effective command siructure (as was the case in Argen: tina), and tend to become more politicized the older lhe re-

gime gets (which is one reason why professionally inclined military leaders often wish to end mililary regimes). In a personal dictatorship, like that of the shah of lran or Marcos in the Philippines, on the other hand, the military may not be

in politics but polities pervades the military because the dic-

tator’s major fear is a coup d’etal. Hence incompetents and cronies gel preference in appointments. Both military regimes and personal diclalorships are thus likely to have

armed forces low in military professionalism and military effectiveness. As the Greek and Argentine cases suggest, provoking a war is, consequently, a high-risk strategy. A fourth option was to allempl to establish some sem-

blance of democratic legitimacy for iheir regime. Mosl au-

wiy?

57

g

part from the

ideol gical y

based one-

party systems-—in 1 exislence in the early 19709 claimed thal in due course Lthey would restore democracy. As their performance legitimacy declined, they came under increasing pressure to deliver on that promise and they had increasing incentives to attempt to relegitimize themselves through elections. In some cases the leaders apparenlly believed hey could fairly win elections. That, however, was almosl never the case, particularly if the opposition achieved a minimum degree of unity. The performance dilemma thus gave way to the eleclion dilemma. Should they sponsor an election? If they

did, should they rig the election? if they did that, would they

gain in legitimacy? If they did not rig the election, would they lose it? Fiflh, the authoritarian leaders could seize the bull by the horns and take the lead in ending authorilarian rule and introducing a democralic system. This happened often but it almost always lirst required a change of leadership within the authoritarian system. Declining legilimacy usually provoked doubt in the minds of authoritarian leaders and divisions within the leadership over which response o choose. The resulling hesitation, disagreement, and fluctuations in action then further decreased the legitimacy of the regimes and encouraged political groups to think about successors to lhem. The successor regime did not need to be democratic. In

1978 and 1979, Inn and Nicaragua shxfred {rom modernizing

personaldic ps to Islamic ism and Marxism-Leninism respectively. As the third democratic wave gol underway in the mid-1970s, a number of transitions also occurred to Mandst-Leninisi regimes in Alrica and elsewhere in the Third World. After a struggle, Partugal went democralic; after independence,

its former colonies went Marxist-

1eninist. Detween the mid- 19603 and the early 1980s Lhe total number of ostensibly Marxisi-l.eninist Third World regimes increased [rom six to seventeen. These regimes were, for the

58

THE THIRD WAVE

most part, relatively narrowly based, and they did nol meet with economic success or suslained political stability. In the carly 1g60s, bolh Latin American politicians and U.S. leaders had seen the political choices for Latin America as reform or revolulion, John X Kennedy or Fidel Castro. With a few exceplions, however, Latin America got neither reform nor revolution but instead repression in the form of Lhe military and bureaucratic-authoritarian regimes. The economic failures of these regimes then eliminaled this form of governmenl as an dllern.'mve for Lhe immediate imure Right-wingdi hips, as in the Philippines and L] Salvador, ofien stimulale the growth of left-wing revolulivnary movements. In South America, however, the ruthless and successful repression by the military regimes physically eliminated many revolutionary extremisls and also kindled new apprecialion among Marxist and socialisl groups of the virtues of democtacy. In Lhe 1980s, as Juan Linz and Alfred Stepan noted, the Lalin American left came lo view “procedural democracy” as “a valuable nonn in itsell and as a political arrangement that offers both protection against slale terrorism and some hope of electoral progress toward social and economic democracy.” In a comparable vein, cne of the falhers of liberation (heology, Father Gustavo Guliérrez of Peru, observed in 1988 that “Experience with dictatorship has made liberation theologians appreciative of political rights.”

The collnpse of the communisl regimes in Eastern Europe further weakened the possibility of Marxism-Leninism as an allernative to other authoritarian regimes. Thus, while authorilarian regimes came in many forms-—military government, one-party system, personal lymn.nv, absolute monarchy, ra-

cial

ol

Islamic

di

by the 1980s. they were

not, by and large, perceived as alternatives lo each other. Outside of Africa and a few counlries elsewhere, democracy had come to be seen as the only legilimate and viable alter-

nalive to an authorilarian regime of any type.

wWHY? ECONOMIC

59 DEVELOPMENT

AND ECONOMIC CRISLS

The relation belween economic development, on the one hand, and democracy and democralizalion, on the other, is complex and probably varies in time and space. Economic factors have signifi impact on d; ization bul they are nol d inative. An overall ¢ lation exisls between the level of i and d y yel ne level or paltern ol economic development is in itself cither necessary or sufficient lo bring about democratization. Economic lactors affecled third wave democralizations in three ways. First, as was pointed out above, the oil price hikes in some countries and Marxisl-Leninist constraints in others created ic d that the authorilarian regimes. Second, Ly the early 1970s many countries had achieved overall levels of economic development thal provided an economic basis for democracy and that [acilitaled lransition to democracy. Third, in a few countries ly rapid growth deslabilized authoritarian regimes, lorcing lhem either lo liberalize or to intensily re pression. E in short, provided the basis for democracy; crises pmduccd by either rapid growth or economic recession weakened authoritarianism. All three factors did not appear in every country, but virtually no third wave country escaped all of them. They provided the economi¢ impetus and context for democratization in Lhe 19708

and 1980s.

Economic developrent. Cighleenth-century political theorists argued that wcnlll\y countries were Ilkely m be monarchies, while peor would be rep or d ies. This was a plausible hypothesis for agrarian societies. Industrialization, however, reversed lhe relation between level of wealth and form of government, and a posilive correlation belween wealth and democracy emerged in the nineteenth century. Tt has remained strong. Most wealthy countries

60

THE THIRD WAVE

are democralic and most democratic countries --India is the most dramalic exceplion- are wealthy. This relalionship was

highlighted by Seymour Martin Lipsel in 1959 and has been

strongly reinforced by a large number of subsequent studies.” In 1985, for inslance, Kenneth A. Dollen and Robert W. Jackman found Lhal in the 1960s “level of economic development has a pronounced effect on political democracy, even when other noneconomic faciors are considered. . . . GNP is the dominant explanalory variable.” ** In 1989 the Warld Bank classified as “high income” twenly-four countries with per capita incomes ranging from $6,010 (Spain) to $21,330 (SwitLerland) Three of these (Saudi Arabia, Kuwail, and United Arab were oil exp and d ic. Of the remaining twenty-one high-income countries, all except Singapore were democratic. Al the other extreme, the World Bank categorized as “poor” forty-two countries with per capita incomes ranging (rom S13o (Elhiopia) to S4so (Liberia). Only wo of these countries {India, Sri Lanka) had had any i with d Among the fifty-three “middle-income” countries, ranging from Senegal (per capita GNP of $520) to Oman (per capila GNP of $5,810), there were twenty-three democracies, twenly-five nondemocracies, and five countries thal could in 1989, be plausnbly classified as in from to d The correlation between wealth and dequaLy implies

that transitions lo democracy should occur primarily in countries at the middle levels of economic development. In poor

countries democratization is unlikely; in rich countries il has already occurred. In between there is a polilical transilion zone; ies in that p ic stratum are mosl likely 1o lransil tv democracy and most countries that transit to democracy will be in thal siralum. As countries develop economically and move into this zone, they become prospectls for democratization. During the first wave of democralizalion in the nineteenlh and early twentieth centuries democracy g 1l ged in northern Europ when

wHY?

61

Lheir per capila GNPs, in 1960 dollars, were between $300 and $500. In the 19205 and 19305, a variety of faclors, including economic crises, produced the first reverse wave to authoritarianism. Overall, however, economic development continued and hence the income level of the transition zone separating democracies and nondemocracies moved upward." The 19505 and 196os were years of impressive global economic growth, particularly among less developed countries. Between 1950 and 1975 per capita GNP of the developing counlries grew al an average rate of 3.4 percenl per year, a rate that “exceeded both official goals and private expectations,” % This rate was hislorically unprecedented for both the devl, ies and for the clop ies. In the

1960s, khe ‘decade

of developmenl,”

lhe

annual

GNP

growth rates of lhe developing countries averaged well over 35 percent, generally more than lwice the rales of European ies during their p phases of ic development.

The

rates

for

individual

countrics,

of

course,

varied considerably—higher in southern Lurope, Fasi Asia, the Middle East, and Latin America; lower in Soulh Asia and Alrica. Overall, however, the post-World War Il surge of economic growth that lasted unlil the oil shocks of 1973-74 moved many countries inlo Lhe transition zone, crealing within them the economic conditions (avorable to the development of democracy. In considerable measure, the wave of democratizations

that began

in 1974 was

the product

economic growth of the previous two decades.

of the

By the 1970s the center of the economic transition zone had moved upward from the prewar level of S300-$500 (1960

dollars)

to the $500-$1,000

range.

Nine,

or almosl

hall,

of

iwenty-one third wave democratizations occurred in countries within these limits; four occurred in countries in the $300-$500 range; two (Greece, Spain) in countries with per capita GNPs slightly over 51 000 (1960 dollars); and six (India, Pakistan, El S Bolivia, Philippi in counlries with 1960 per (zp)la GNPs of less than Sjm The

62

THE THIRD WAVE

range in the level of economic development, from India ($87) to Greece

($1,291)

was

substantial,

but aboul

two-thirds

the transitions were in countries between roughly $1,300 in per capita GNP (1960 dollars) al the time Lion. Transitions clearly were most likely to occur in al lhe middle and upper-middle levels of economic

of

$300 and of iransjcountries develop-

ment and, as expecled, were concenirated in the income zone

just above thal where Sunshine had found them concentraled before World War (L. The third wave transition zone also appears in the data presented in Table 2.1. Countries are classified according to their per capita GNDPs in 1976, as reported by the World Bank, and as Lo whether they had democralic political systems in 1974, whether they democratized or liberalized between 1974 and LE2.1 Lconomic Development and Third Wave Democratization (O] @ ] o @ )

Percentage

Demox-

1976 per

Demo.

ralized! Liberal

(in dollars}

1974

1974-89

Capita GNP 3,000

Total

cratic

ized

1

catic 31

2

3

1 1%

27

3t

that Democratized/

Non-

demo-

»

3 5

8

of Cauntries.

Toral

27 5 66

£

Liberal-

1zed+ 6

1 b3

» 76

24

n

px]

1

Source: Economic data are from World Bank, World Development Report

1978 (Washington: The World Bark, 1978), pp. 76-77. “During the period from

1974 o 1989, and Exdudmg countries that

were already democratic in 1974.

+Includes India, which became nondemocratic in 1975 and then democratized in 1977.

“Inchudes Nigeria, which transited 16 democracy in 1980 and back lo

military rule in 1984, and the Sudan, which traversed a similar course in

1986 and 198y,

WHY?

63

1989, or whether they had nondemocratic regimes throughout these years.™ These figures again indicate thal third wave countrics varied greatly in Lheir level of economic development, India and Pakistan having 1976 per capila GNPs less than $250 and Czechoslovakia and I:asl Germany havmg ones over $3.000. Twenty out of thirt; that liberalized or democralized, however, were in the middleincome range, neither poor nor weallhy, and half of the third wave countries had 1976 per capila GNPs belween $1,000 and $3,000. Three-quarters of the countries that were at this level of economic developmenl in 1976 and that had nondemocratic goveriunents in 1974 democralized or liberalized significantly by 1989. A social scientist in the mid-1g70s who wished ta predict future democralizalions, in short, would have done reasonably well if he had simply fingered the nondemocratic countrics in the $1,000-$3,000 transition zone. This is not lo argue thal democTatization is determined simply by economic development. Clearly il is not. In 1976 Czechoslovakia and East Germany were well up in the wealthy economic zone where they “should” have been democratic alrcady, and the Soviet Union, Bulgaria, Poland, and Hungary were high in the transition zone with per capita GNP over $2,000. Politics and external forces, however, held back their movement toward democracy until the end of the 1980s. Interesting to note, in an early 1960s study, Phillips Cutright established a strong correlation

belween

level of communi-

cations development and democracy. and used Lhis 1o highJight anomalous cases off his regression line. The principal European countries that were much less democratic then (han they “should” have been were ‘ipmn, Portugal, Poland, and Czech ia.Z In the less Iberian environment, political development caught up with economic development in the mid-1g70s; in Eastern Furope that did not happen until Soviet controls were removed lilteen years later. Five countries with 1976 per capila GNPs between $1,000 and $3,000 did not democratize by 1990. Iraq and Iran were

64

THE THIRD WAVE

large-population oil producers. l.ebanon had had a limited form of consociational democracy but disintegrated inlo civil war after the mid-1970s. Yugoslavia, which in some respects had been more liberal than other Casi European communist counlries, was overtaken by the democraltization surge of its ight in 1 Ithough ils hicst slates, Slovenia and Croatia, did begm to move in a dumn(‘rallc direction. The city-state of Singap the ing Third World country, remained throughout the n;sas under the generally benign but ruthlessly (irm rule of its philosopher-king. There, as in (he Soviet bloc, politics dominated economics. An economic transition zone can also be seen in a parallel analysls by Mitchel! Seligson argumg (hal in Latin America the holds thal made cray possible although not necessary were per capita GNI’ (1957 dollars) of $250 and so percent literacy. Of eleven Talin American countries. only three, Argentina, Chile, and Cosla Rica, were above these thresholds in 1957. By the 19805, however, seven more natlions —Brazil, Peru, Ecuador, El Salvador, Nicaragua, Guatemala, and, marginally, Honduras—had reached or passed them. Only Bollvna among the eleven nations covered in this study, d Iy below he thresholds. The eco nomic basis for democracy was thus emerging in Latin America. That did not, of course, guarantee (he emergence of democracy, yet by 1990 lransilions to democracy had occurred or were occurring, in all these countries. In a similar vein, Enrique Baloyra poinled oul that in Latin America old-style personalislic dictatorships (Paraguay) lended to survive longer than

new-slyle

bureaucratic authoritarian

regimes (Brazil).>

The viability of an authorilarian regime appears to be a {unction more of the narre of its society than of the nature of the Tegime. Why did economic development and the movement of countries inlo the upper-middle income levels promote democralization? The evidence suggesls thal sheer wealth itsel(

wiiy?

65

may not have been a crucial factor. [ran and Iraq were in the transition zone but did nol democralize. Three small-popuJation oil producers (Saudi Arabia, Libya, Kuwail) were undemocratic, although they had 1976 per capita GNPs over $4,000, ranking well up among the wealthy countries. The implicalion is Ihal broad- based economic developmeni volving sij ion may contri to democratizalion but wealth resulting from the sale of oil (and, probably, other natural resources) does nol. Oil revenues

accrue

to the slale: they therefore increase the power of the slate bureaucracy and, because lhey reduce or climinate the need for taxation, they also reduce the need for the government to solicit the acquicscence of ils subjects Lo taxalion. The lower the level of taxation, the less reason for publics to demand representation. “No taxation wilhout represenlation” was a political demand; “no representation willoul taxalion” is a political realily. In conlrast to patlerns in lhc oll stales, processes of economic Tead to a new, much

more dlvcrsc. complex, and interrelated

economy, which becomes increasingly difficull lor authorilarian regimes lo control. Economic development created new sources of wealth and power outside the siate and a functional need 1o devolve decision making. More directly, economic developmenl appears to have promoled changes in social structure and values thal. in turn, encouraged democ-

ratization. First, Lhe leve! of economic well-being wilhin a so-

ciety itself, it has been argued, shapes “the values and attiludes of its citizens,” foslering the development of {eclings of interpersonal trust, life satisfaclion, and competence, which, m lurn, com-:lale strongly wnlh the exislence of democratic » Gecond, increases the levels of education in sociely. Between 1960 and 1981 the proporuon of lhe relevant age group am-ndmg secondary school in rloping countries i ically.* More highly educated people tend to develop the characteristics of trust,

66

THE THIRD WAVE

satisfaction, and compelence thal go with democracy. Third, economic development makes greater resources available {or dlsmbunun among, soclal g'roups and hence Ianlllales accomand in the 1960s and 19705 bolh requu'ed and promoted the. opening of societies o forcign trade, investment, technology, tourism, and communications. Involvement of a country in the world economy created nongovernmental sources of wealth and influence and opened the sociely Lo Lhe impacl of the demaocratic ideas prevailing in the industrialized world. Governments, such as that in China, that wished lo open lheir economies to the world in order to promoie economic development and yel also {o maintain a dosed political system d an app confhcl Aularky and P were an impassibl T and lik izing foreign i an id, one. Tinally, ic devel the ion of the mlddle class:a larger and Iarper pmpoxlwn of snclery consists of busil teach. ers,

civil servanls,

managers,

technicians,

clencal and

sales

workers. Democracy is premised, in some measure, on majorily rule, and democracy is dilficult in a situalion of concentrated inequalities in which a large, impoverished majorily confronts a small, wealthy ofigarchy. Democracy may be possible in a relalively poor agricultural society, such as early nineteenth century United Stales or twentieth-century Costa Rica, where land p is ly equal. A middle class is normally, howeve(, the product of industrialization and econamic growth. In its early phases, the middle class is nol necessarily a force for democracy. At limes in Latin America and elscwhere, middle-class groups acquiesced in or actively supporied military coups designed to overthrow

radical governments and to reduce the political influence of labor and peasanl organizations. As the pracess of modemn-

ization

continued,

however,

rural

radical

movements

had

decreasing leverage on the polilical process, and the urban

wWRY?

67

middle class i in size p to the industrial working class. The polential threals democracy posed to middle-class groups lhus declined, and those groups became increasingly confident of their ability to advance their interesls through elecloral politics. Third wave movements for democratization were not led by fandlords, peasants, or (apart from Poland) industrial workers. In virlually every country the most active supporters of democratization came (rom Lhe urban middle class. In Argentina, for instance, the choice in the 19605 and 1970s had been between an elecled Peronista government based on the work-

ing class or a coup-originated military regime with middle-

class support. By the 19805, however, the middle class had become sufficiently numerous so as to provide the core element in the victory of the Radical party under Radil Alfonsin and to induce the Peronista candidates lo be sensilive Lo its interests. In Brazil the middle class overwhelmingly supported the 1964 coup. By the mid-1970s, however, “it was precisely those sectors which benefitied most [rom the years of the “economic miracle’ which were the mosl vocal in demanding a return to democratic rule: the populalion of the large and developed cities, and the middle class.”= in the Philippi middle-class p ionals and by people filled the ranks of the demonslrations agamst Marcos in 1984. The following year the core groups in the Aquino

campaign were “middle class, non-politician doctors and

lawyers, who had volunteered their services to the opposiLion candidates or to the citizen watchdog groups NAMFREL [National

Movement

for Free Clections],

rather than

{o any

party.”* In Spain economic developmenl had created “a nation of the modem middle classes,” which made possible the rapid and peaccful process of bringing the political system into line with sociely.? On Taiwan the “main actors for polit cal change” were “lhe newly emerged middle class-inteliectuals who came to age during the period of rapid economic

growth.” ® In Korea the movement for democracy in the 1980s

68

THE THIRD WAVE

only became a serious threat to the authoritarian regime afler the emergence by the 1980s of “a flourishing urban middle class " and middle-class professionals joined the smdenls in ding the end of author The of the "managenal and prolessional classes of Scoul . . . was perhaps of g-realer moment than any other” faclor in Iorcmg m 1987 porting the 1987 demagamsl the Chun reg|me the Economist asked, “What happens when tear gas meets the middle class in Seoul?”” The answer was soon clear: tear gas loses. In several countries, including Spain, Brazil, Peru, Lcuador, and the Philippines, the business comununity, which had previausly supported the creation of authorilarian regimes, played crucial roles in promoting the transitions lo democracy.” In contrast, where lhe urban middle class was smaller or weaker, as in China, Burma the Sudan, Bulgaria, and Romania, eilker d was or democ racy was unstable. The p of i which, if O'Donnell is correct, produced bureaucratic-authoritarianism in the 1960s, Lhus also provided the impetus for democratization in the 1980s. A plausible approximation of the causal connections that led (o this outcome is depicted in Figure 2.1, Rapid growth. The movement of counlries into the middieincome ranges of the economic transition zone thus led to changes

in social

structures,

beliefs,

and

culture

thal were

conducive to the emergence of democracy. Extremely high

rates of economic growth in some countries also gencrated dissatisfaction wilh the exisling authoritarian government. In the two decades before their iransilions in the mid-1g70s, Spain, Portugal, and Greece experienced explosive cconomic growth. Between 1913 and 1950 the average annual compound rate of growth of real outpul per capila had been nega tlve in Spain and less than 1 percent in Greece and Portugal. Belween 1950 and 1973, the rales were 5.2 percent in Spain, 5.3 percenl in Portugal, and 6.2 percent in Greece. GNP

more hight educated g.hn!r.:

—

ivictulture attitudes e action

n.g.er level of economic development

__

«competence

sup ort for democtatzation

lager middle dass

Figure 2.1. Economic Development as a Factor in Democratization

growlh rates in these (hree countries belween 1960 and 1973 were 6-8 percent compared to 4-5 percent in other Western European countries; and GNF per capita grew faster between 1960 and 1980 than in all other counkies of the Organization for E Co-operalion and Develop {OECD) except

Japan.®

Rapid economic growth creales rapidly the economic base for democracy that slower economic growlh creates more slowly. Tl also, however, raises expectations, exacerbales inequalities, and creates stresses and strains in the social fabric that st political ilization and d ds for polilical

participation. ln Greece, (or instance, in lhe 19508 and 1960s,

rapid, unequal gmwth d awareness, poli fi " leading to “social unrest and political rnohlxzalmn ¥ These pressures were a significant cause of the 1967 coup, one purpose ol which was o slifle them. Economic growth, however. continved under (he military mme unul 1973. The regime simul-

ly pursued two

policies. It “attempled o

halt and reverse the process of democratizalion. But at the same lime il was commilled lo rapid economic growth and

70

THE THIRD WAVE

dernization.”** Social jon and political discontent mounted. At the end of 1973, the oil price hike added an additional source of discontent and the regime either had to liberalize or lo intensify repression. Papadopoulos moved lentanvely toward liberalization; the sludems at the National 1 Uni y (Polytechnic) p d and d ded more. They were shot, and the hardllners under loannidis ousted Papadapol: only to fall th lves a hall year later when they provoked a military con(rontation in Cyprus. In Spain similar contradictions were produced by the “unprecedenled period of economic growth” in Lhe 1g60s. Leaders of the Franco regime hoped Ihal this growth would lead to a happy, contented populace uninterested in politics. “In practice, however, rapid economic change exacerbated or catalyzed major conflicts in Spanish society and promoled cullural, social, and political changes that placed the regime’s viability in doubl. Political arrangements, originally conslructed in a primarily agrarian soclety in Lhe aftermath of a bilil Civil War, app islic when confronted with the tensions of a rapidly changing industrialized society.”* The polilical demands generaled by rapid growth were superimposed on an economy that had fully established the economic and social base of democracy. In the 1960s, Laureano Lopez Rodo, Franco’s minister of planning, prophesied that Spain would become democratic when its per capita GNP reached

$2,000.

[1 did. The

transition was also furthered

by

the timely dealh of Franco in 1975. If he had not died then or if Juan Carlos had nol been committed lo creating a parliamentary democracy, polarization could have led to social violence and could have seriously diluted the prospects for democracy in Spain. As it was, however, the economic and social prerequisites for democracy existed in Spain in 1975 and l'\ent? skilled and committed leadership could bring democ ion about ively quickly and s In the late 1g60s and early 1970s, Brazil expenenced its. “economic miracle.” From 1968 to 1973, its GNP grew al an

whY?

71

average rate of close lo 10 percent a year. This intensilied its aiready highly unequal distribution of income and led some to portray Brazil as the epilome of capitalist development in which mullinational corporalions and their local partners benefiled and local workers and peasants suffered. It also led Emesto Geisel, when he became president of Brazil in 1974, to comment that “Brazil has done very well, bul Brazilians have done very poorly.” The pressures of rapid economic growth that had led to the demise of ihe military regime in Greece and to the Lransformation of the dictatorship in Spain were equally manifes! in Brazil. The military leaders of Brazil, however, were aware of these pressures and delermined (o accommodate them. In the last year of his regime President Medici began lo consider ways of bringing about disfensio (decompression). President Geisel and his top advisor, General Golbery do Couto e Silva, started this process and carried il forward to 1978. President Joao Figueiredo conlinued and broadened it inlo a process of nbertura, or opening. The actions of the two presidents averted intensified social confliel and paved the way for democracy. From the 1960s to the 1980s the growth rates achieved by South Korea and Taiwan were among the highest in the world. The Lwo socielies were transformed economically and socially. The p in these ies for di izati developed more slowly than in European and Latin American socielies for two reasons. First, Confucian cultural traditlons

emphasizing hierarchy, autherity, community, and loyally delayed the articulation by social groups of intense demands on the polity. Second, in conlrasl to other societies, rapid economic growth in Korea and Taiwan took place in the context of relalively equal patlerns of income distribution. The latter resulted from a variely of causes including land reform programs in lhe [ate 1940s and early 1950s and early atiainment of high levels of literacy and education. The inequalities associaled with rapid growth in Brazil were notably absent (rom these two East Asian countries. By Lhe rg8os, none-

72

TIIE THIRD WAVE

theless, ic devel P had ded o lhe point where p for polltu:nl particip compelled governments in both countries to begin processes of democratization. Very rapid economic growlh inevitably produced challenges for authoritarian leaders. It did nol necessarily lead them to introduce democracy. Between 1960 and 1975 Brazil's GNP grew al an average annual rale of 8 percenl. During lhese same years, Iran’s GNP grew al a rate of 10 percent. Between 1980 and 1987, China’s GNP also grew at an annual rate of 10 percent. These rales of growth generaled highly deslabllmng smsses and strams in these lluee authoritari. syslems, i and and s lated social groups lo make demands on their governments. The leaders of the three countries responded in three different ways. Geisel opened up; Deng cracked down; the shah shilly-shallied, Democracy, repression, and revolution were the respeclive results of their choices. sumnary. Over the long: i creales the basls for democralic teg\mcs In the short-tlerl'n, very rapid cconomic growth and economic crises may undermine authoritarian regimes. [f economic growth occurs without economic crisis democracy evolves slowly, as it did in nineteenth-cenlury Europe. If the destabilizing growlh or economic crisis occurs wilhout achievement of transition zone

wealth, the authoritarian regimes may fall but theic replace-

ment with long-lived democralic reglmes is highly problemalical. In Lhe third wave, the combination of substantial levels of economic developmenl and short-lerm economic erisis or failure was the economic formula most lavarable to the iransition from auth: to d B 7 RELIGIOUS CHANGES

Two devel

in religion p

democrati

in

the 19705 and 1980s. A strong correlation exists belween Western Chrislianity

wHY?

73

and democracy. Modern democracy developed first and most vigorously in Christian countries. As of 1988 Catholicism and/or Protestantism were the dominant religions in thirlynine of forly-six democralic countries. These thirly-nine democratic countries made up 57 perrenl ol a total of sixtyeight ies thal were p ly Western Christian. In contrast, only seven, or 12 percent of fifty-eight countries with other predominanl religions were democratic. Democracy was especiaily scarce among countries thal were predominantly Muslim, Buddhist, or Confucian.®

This correlation does nol prove causalion. Westen Christianity emphasizes, however, the dignity of the individual and the separale spheres of church and state. In many countries, Protestant and Catholic church leaders have been central in the struggles against repressive countries. It seems Jausibl ) ize that the expansion of Chri i encoumges demo-:mhc development.

Wihere did Christianity expand significantly in the 19605

and 1970s? The answer is, in very few places. The most prominent case was South Korea. Wilh only a brief break between them, Korea had first a semidemocratic civilian regime under Syngman Rhee in the 19505, 2 semidemocralic military regime under Park Chung Hee in the 1960, and military dictalorships under Park and Gen. Chun Doo Hwan in the 19708 and 1980s, with the transition lo democracy coming in 1987. At the end of World War N Korea was primarily a Buddhist

country with a Confucian overlay. Perhaps 1 percent of the

populalion was Chrislian. By the mid- |9&ys mughly 25 percent of the population was Christi largely Presbylerian, and one-fifth Catholic. ‘l'he Christian converls were primarily young, urban, and middle class. Their reasons for espousing Christianity stemmed from the profound social and economic change taking place in Korea, “For the millions who poured into the cities,” as one account described it. “and for many who stayed behind in the altered countryside, the quiescent Buddhism of Korea's agrarian age

74

THE THIRD WAVE

lost its appeal. Christianily with its message ol personal salvation and individual destiny offered a surer comfort in a time of confusion and change.” ¥ Christianity also offered a surer doctrinal and institutional basis for opposing political ion. Cl ianily, as one South Korean pul it, “made a difference, because it promoles the idea of equality and respect for some authority independent of the state.” Confucian authoritarianism and Buddhist passivity were replaced by Christian militancy. In 1974 (ve bishops led 5,000 Roman Catholics in the first major demonstration against President Park’s martial law regime. Many of the principal leaders of the opposition movements, such as Kim Dae Jung and Kim Young Sam, were Christian, and Proleslant and Catholic clergy, such as the Rev. Moon Ik llwan and Cardinal Kim Sou Hwan, were leaders in atlacking the repression of the military government. By Lhe early 1980s the churches had become “the principal forum for opposition to the regime.” In 1986 and 1987 Cardinal Kim, other Catholic leaders, and Lhe principal Prolestant organization, the National Council of Churches, all strongly supported the opposition campaign for direct election of the president. In the conflict wnl.h lhe government, “'Churches and cathedrals provided an | basis (or Lhe activities of Jiuman rights and justice, and a public space to share dissenting opinions and faiths. Catholic priests, Roman Catholic Association of Young

(‘alhohc Workers,

and

Urban

Industrial Mission and

were politicized and began to represent an imp part of the 'D . Myongdong Cathedml in Scoul had been a symbolic site for political dissidents.” Kore.l thus, in a sense, revelsed the Weber connection: d the expansion of Christianily, and the Christian churches, their leaders and communicants, were a major force bringing about the transition to democracy

in 1987 and

1988.

The second and more important religious developmenl encouraging democratization involved the far-reaching changes

wRY?

79

that occurred in the docirine, leadership, popular involvemenl, and political alignment n! the Roman Catholic Cl hurch globally and in many and democracy were linked with each other. The first democralic |mpulse in the Weslern world came with the Purilan revolution in the century. The overwhelmi majorily of the countries that beume democranc in the first wave of democralization in the nineteenth century were Prot estant. The second wave countries after World War II were diverse. N hel in the 1g60s a sj Telation existed between these (wo variables. For ninety-nine couniries, one siudy showed, “the greater the proportion of the population thal is Prolestant, the higher the level of democracy.” In contrast, Catholicism was associated with the absence of democracy or with limiled or lale democratic development. Catholicism, Lipset had noted, “appeared anlithetical to democracy in pre World War II Europe and in Latin America.” < Three plausible reasons were advanced to explain these relationships. Doclrinally, Protestantism siressed the individual conscience, the access of the individual to the Holy Writ in the Bible, and the direct relation of the individual to God. Catholicism emphasized the intermediary role of the priesthood epitomized in the Latin mass. Second, Proteslant chux:l\es were themselves more democraucally organized,

of the gation and with no or only a lmuled blSthn( The Catholic Church, in confrast, was an aulhonlanan orgnnlzanon with its mnk: of priests,

bishops,

archt

in the Pope

and the doclrine of papal mfalllbnlny Catholic countries, as Pierre Elliol Trudeau noted, “are authorilarian in spiritual matters; and since the dividing linc between the spiritual and the temporal may be very fine or cven confused, they are often disinclined to seek solutions in temporal affairs through the mere counling of heads.”+* Finally, there is the Weber thesi ism encourages ic enler-

76

TIIE THIRD WAVE

pnse, the d

ofab

lism,

and eco-

nomic wealth, which facilitate the emergence of democratic institutions, Until the 1960s

these arguments

and

the association

they

explained between religion and democracy seemed unchallengeable. That is no longer the case. The third wave of the 19705 and 19805 was overwhelmingly a Catholic wave. Two (Portugal and Spain) of the first three third wave countrics lo democratize were Catholic. Democralization (hen swept through six South American and three Central American countries. It moved on to the Philippines, the first Casl Asian country lo democratize, doubled back to Chile and affected Mexico, and then burst through in Catholic Poland and Hungary, the [irst East European countries to democratize. Catholic countries were in the lead in every region of the world, and the mosi Calholic region, Lalin America, was the region that democratized most fully. Overall, roughly three-quarters of the

countries

that transiled

those

countries

as

lo democracy

belween

1974

and 1989 were Catholic countries. How can this be explained? One partial answer, of course, is that by the early 1970s, mosl of the major Protestant countries in the world had become democratic. The principal exceplions were LEast Germany and South Africa, and Protestant church leaders promoted democratizalion in well

as

in

Korea.

If,

however,

many

more countries were Lo become democralic lhey would have

to be Catholic, Orthodox, or non-Christian. The question remains: Why Catholic? One partial explanalion may lie in he reversal ol another previously existing unfavorable Catholic correlation. Historically Protestant countries develped economically more rapidly than Catholic countries and achieved higher levels of economic well-being, Catholic countries were poor countries. Beginning in the 1950s, however, Catholic countries began lo have higher rales of economic growth than Proleslanl countries. In large part, of course, this was because thcy were generally at lower levels of

WHY?

77

economic E growth, hel undoubtedly facililated tramsilions to democracy in several Calholic countries.* A more pervasive cause ol the surge lo democracy in Catholic countries was change in the Catholic Church. Historically in Iberia, Latin America, and elsewhere the Church had been associated with the local eslablishmenl, the landowning oligarchy, and authoritarian government. In Lhe 1960s, the Church changed. The changes within the Church brought ap ! social i into opposition Lo dictatorial regimes, deprived those regimes of whalever legitimacy they might claim frum rehfnon and pmwded pmlechun support, to

rat

move-

menis. Before the mid-1960s, , Lhe Calhollr Chun:h usually accommodated itsell o authoritarian regimes and frequenty legilimaled them. After the mid-196us, the Church almosl invariably opposed authorilarian regimes; and in some countries, such as Brazil, Chile, the Philippines, Poland, and Central American couniries, it played a central vole in the efforts to change such regimes. This reposilioning of the Catholic Church

from

a bulwark

of the status quo,

usually

authori-

tarian, lo a {orce (or change, usually democralic, was a major political ph The social scientists of the 19505 were right: Catholicism then was an obslacle to democracy. After 1970, however, Catholicism was a force for democracy because of changes within the Catholic Church.*

These changes occurred at two levels. Al the global level

“Why and how these momentous changes occurred in the Catholic Church'are beyond the scope of this shady. George Weige! identifies several

factors rearhing back to the late ninetecnih century as responsible for the

shift in the Churcly’s position concesning the literal democratic srate, “Par. mong them, lw argues, were the Uinited States and the

Their influence, he says, culminated in Vatican If and uts DLlllmllflll on Religious Freedom {Dignilatis Humanae Personae), which was

d of the American expericnce and upcnmcnl and rspeaally M me

I\mcncan 1heglogian John

Courtney Murray. See George

cism and Democracy: The Other Twentieth-Century Revolu Democracies: Globul Change and U.S. Policy, ed. Brad Roberts

78

THE THIRD WAVE

change was introduced by Pope John XXIII. The changes sternmed from his own slyle and commitment and [rom the doctrines he arliculated in his encyclicals. Mosl important, however, the changes flowed from (he Second Vatican Council, which he called and which mel from 1962 to 1965. Vatican 11 stressed the legitimacy and need for social change, the importance of collegial aclion by bishops, priests, and laity, dedication lo helping Lthe poor, the conlingent characler of social and polilical structures, and the rights of individuals. Church leaders, Vatican Il asserled, have responsibilily to “pass moral judgments, even on mailers of the political order whenever basic personal rights . . . make such judgment necessary.”** These views were reaffirmed and elaborated in the Latin American bishops conferences in Medellin in 1968 and in Puebla in 1979 and in gatherings of blshOPS elsewhere

Equally

signiticant

changes were

in popular involvement and priestly aclivity ‘al the base of the Church. In Spain in the 1960s, (or instance, as Juan Linz puinted out, New generations of priesls, some of them lale vocalions, per-

haps a less rural recruitment of the clergy, a greater awareness

of social injustice and contact with the de-Christianized working class, sociological studics of religivus practice, he identificalion of the clergy with the cullural, linguistic minorilies in

the Basque country and Catalonia, and above all, lhe

impact

of the Second Vatican Council, produced a ferment of criticism

ress, 1990), pp. 20-25. To the extent that Weigel's argument is valid. the United States played two roles in bringiug about the thitd wave: dircetly through ks new polkies In the 19703 and 1980s (see below, pp. g1-98) and at one remove through its impact over time on the Catholic Church. Wilh rarc exceptions. sacial scientists ignarc the significance for cratic development of Lhe changes i olic Churehs they also ignored the comparably imporiant, if totally different, developments In [stam in the 1960s and 1g70s. One prescient exception on the Church was George:

C. Ladge’s 1570 book Engines of Change: Uniled States Intcrets and Revolition i Lt Amerets (New York: Alfred A Knopl).

wHY?

79

and unrest among younger catholic intellectuals, laymen, and clergy and conflicls with authority.« In Brazil the 1960s and 1970s saw the rapid spread throughout the country of popular ecclesiastical base communities (CEBs), which by 1974 numbered 40,000 and gave an enlirely new characler lo the Brazilian church. Simultaneously in the Philippines a “Christian left” developed that included priests and grass-roots activists, some of whom were Marxist and others of whom supported a social democracy that was both anti-imperialist and anticommunist. In the late 1970s in Argentina, the Church shilted dramatically away from its previous conservative coloration, with priesis mobilizing a massive youlh movement leading to “an extraordinary evangelical revival,” Similarly, in both Poland and Chile, a “grass-reots politicization of the church” occurred: “the loundation of the churcly's position in each country can be traced to movements of aggressive, young priests who have strongly identified themselves with the aspiralions of local society and sought to organizc or protect authentically representative and nonviolent social movements.”*” These new streams of L bilizati bined with the new streamn o( doctrine commg {rom the Valican created a new church that almost i iably came inlo PP i to governmenls. Tn non-communist counlries, the relations between the Church and authoritarian governments lended (o move

through three phases: acceplance, ambivalence, and opposi-

tion. Initially, conservalwe elemcnle wnlhm lhe hierarchy were usually b g the historical posilion of the Church as establishment paflncx and defender of social peace. Church leaders usually welcomed the eslablishment of the authorilarian regime. In Spain the Church helped Franco 1o viclory and long bolstered his government. Brazilian bishops adopled an “enthusiastically pro-military governmenl slance” immedialely after the 1964 coup. in Argenlina,

8o

TIHE TIIIRD WAVE

Chile, and elsewhere the Church similarly legitimated military takeovers. As the authorilarian regime sustained itsell in power and usually intensified its repression, and as the converging trends [rom the grass-roots and the Vatican were felt in the national hierarchy, the position of the Church typically shifted. In Brazll the Philippines, Chile, Central American c and t two strands of opposition thought and aclivity tended to develop within the Church. A socialist or “red” strand preached social justice, the evils of capitalism, the overwhelming need lo help the poor, and frequently mcorporaled into its thinking substanlial Manist ¢lements of “lit logy.” The latter infl did not lead the Church in the jon of d but in ies other than Nlcaragua ildid l\elp mobilize Catholics in opposition to thc leung i . in most ies also, a mod “yellow” sirand (as il was known in the Philippines) of upposilinn also developed that lypically included the major segment of the episcopacy and emphasized human rights and demacracy. As a result of these developments, the overall posilion of churches usually shifted from accommodation to ambivalence. Al some time, in most countries, a breaking point came in church-state relations when the national bishops” conference or a top church leader explicilly placed the Church in opposition to the regime. In Chile the brutal abuses of human rights by the military regime shorlly after coming 10 power led to an early break and the formation of the Vicariale of Solidarily in January 1976. In Brazil, the break also came early in the life of the mililary government, when the National Conference of Brazilian Bishops issued a paper denouncing the government’s national securily doclrine as “fascisl” and paved the way for active Church opposition by reminding Brazilians of Nazi Germany, where Chrislians “accepted govemmenl doctrines wnl\oul mcognmng Ihat they contradicted the true d d: of C ity.” Shortly thereafier the car-

wHY?

81

dinal of Sao Paulo dramatically confirmed the break when, in a calculated insult, he refused Lo celebrale a birthday mass for the military president of Brazil.® In Spain, Lhe break came much later in the history of the regime wilh the convening of a unique Assembly of Dishops and Priesls in Madrid in Seplember 971 For the 'ipamsh Church, this was an extraordinarily d ing: its p ings were widely veported; it ”pmvnded a lesson in democracy” to Lhe 5pamsh people and passed resolulions endarsing “the right to [reedom of expression, free association and free union meelings, in elfect all the rights whose exercise has always been very limited during the Franco regime.” As a result of the assembly ”thc Church clearly sepa.raled itsell from the state and d the role of legilimizer of the regime.”* These changes were fully supported by Lhe Valican. In the Philippines the break came in 1979 when Cardinal Jaime Sin demanded the end of martial law and new elections in which Marcos would not run. In Argenlina it came in 1981 with the issuance by the Church of a document on “Church and National Communily.” In Guatemala, the Church ceased lo be a defender of the existing order and became an advocate of social justice, reform, and democracy when the bishops issued “a series of more than fifteen pastoral lelters and public denunciations between 1983 and 1986, all calling for respect of human rights and social, economic, and land reform.”*' In Fl Salvador the Church under Archbishop Romero similarly

broke with Lhe governmeni alter 1977.

In virtually all the cases where Church leaders denounced the regime, the regime retaliated with inlensified alacks on Catholic clergy, aclivists, publicalions, organizalions, and property. Priests and Church workers were often seized, lortured, and, on occasion, killed. Martyrs were created. The resull was often polilical, ideological, and economic warfare belween church and state, wilh the Church becoming, as in Brazil, “the most conspicuous opponent”

to the authoritarian

slate and, in Chile, “the center of moral oppesition to the

82

THE THIRD WAVE

regime.”*2 In lhe I’hlhppmes and other :ountrles the Churrh became lhe ipal inslitulion d defending, human rights, and pushing for the transition (o democracy. The one interesting exception in a deeply Catholic counlry was Poland, where beginning in 1980 Solidarity ized the role of p i and lhe Polish Churrh under the cautmus (_ardmal J67ef Glemp, played for several years a medialing role between government and opposition. The national churches brought many resources lo their war against Church org and church buildings provided refuge and support for regime opponents. Church radio slations, newspapers, and periodicals articulated the opposition cause. As a countrywide and popular institution, the Church had, as in Brazil, a “national network of members

who

could

be

mobilized.”*

It was,

in a sense,

a

latent national political machine wilh hundreds or thousands of priests, nuns, and lay activists who could provide people power for opposition protests. The Church oflen had leaders, such as Cardinals Arns in Brazil, Sin in the Philippines, Romero in Ll Salvador, and Kim in Korca, who were highly skilled politically. The Church created organizations such as the Vicariale of Solidarity in Chile to provide support Lo the opposilion and Lhe second National Movementl for Free Elections (NAMFREL, or “Nunfrel” as it was often called) in the Philippines to promate return lo the electoral process and lo ensure honest elections.™ (The first NAMFREL had been created [or similar purposes by the U.5. Central Intelligence Agency

in 1953.)

In addition, of course, the Church was a transnational organization. On occasion Vatican influence could be brought lo bear, as well as Lhal of other nalional churches and Catholies in other countries. The Brazilian Church, for inslance, mobilized support abroad “through the Vatican, sympalhelic clergy and laymen in Europe and the U.S., and other human righls aclivists outside Brazil, thereby generating protests in

wiy?

83

the U.S. and European press. Criticism from those quariers made the Brazilian mililary especially uneasy.”* With the accession of John Paul Il, the Pope and the Vatican moved to central slage in the Church’s struggle against authoritarianism.

In March

1979 in his first encyclical John

Paul [l denounced violations of human rights and explicitly idenlified the Church as “the guardian” of freedom “which is the condition and basis for the human person’s true dignity.” Papal visits came to play a key role. John Paul Il seemed to have a way of showing up in [ull pontifical majesty at crilical points in democratizalion processes: Poland, June 1979, June

1983, and June 1987; Brazil, June-July 1980; the Philippines,

Tebruary 1981; Argenlina, June 1982; Gualemala, Nicaragua, El Salvador, Haiti, March 1983; Korea, May 1984; Chile, April 1987; Paraguay, May 1988, The purpose of hese visils, like Lhal of his many visils clsewhere, was always said to be pastoral. Their eflects were al-

mosl invariably political. In a few cases, as in Korea and the

Thilippines, local supporters of democratization expressed regrel that the Pope had not been more outspoken in backing their cause. More generally, however, he was quite explicit in supporting local churches in their siruggles with authoritarian governments, and in Poland, Guatemala, Nicaragua, Chile, Paraguay, and clsewhere he clearly identified himsel( wilh the opposition to the regime.* His grealest impact, of

course, was in Poland, where his dramatic 1979 visit, as one Polish bishop said, altered “the mentalily of fear, the fear of

police and tanks, of losing your job, of not gelting promoted, of being thrown out of school, of failing to get a passport. People learned that if Lhey ceased (o fear the system, lhe system was helpless.” This “first great pilgrimage,” Timothy Garlon Ash observed, was the “beginning of the end” of communism in Eastern Europe. ”Here, for the first time, we saw Lhat larg I yet ly peaceful and self-disciplined manifestation of social 1unity, lhe gentle crowd against the Party-stale, which was both the hallmark

84

THE THIRD WAVE

and Lhe essential domestic catalyst of change in 1989, in every country except Romania (and even in Romania, the crowds did nol slart the violence).”

Confronting

Pinochet in Chile

in 1987, the Pope spelled oul the relation of democracy to his mission: “1 am nol the evangelizer of democracy; | am the evangelizer of the Gospel. To the Gospel message, of course, belong all the problems of human rights; and, if democracy means human rights, it also belongs to the message of the Church.”® Finally, Church leaders and orgamzalnuns at times intervened politi al critical in the d

process. In 1978 in the Dominican Republic the Church de

nounced Ihe effort to stop the vole counting and prolong the tenure of Presidenl Belaguer. In 1989 in Panama Church leaders similarly denounced Noriega's theft of the eleclion and invited Panamaman troops Lo disobey orders that they acl against opp In Nicaragua Cardinal Obando y Bravo organized opposition to the Sandinisla government. In Chile Cardinal Juan Francisco Fresno, like his predecessor, Cardinal Radl Silva Enriquez, was in the fore(ronl of the struggle against the Pinochet regime and in August 1985 played an active role in bringing together the leaders of cleven political parlies 1o sign the National Accord demanding constitutional reform and elections. At a crucial moment in 1986 in the campaign for demacracy in Korea, Car-

dinal Kim in an open political move explicitly endorsed the need (or “constitutional revision” and stated “We have to bring democracy to Korea urgently.”® The most extreme political involvemenl of a Church leader was undoubtedly in Lhe Philippi Cardinal Sin ialed the arrangemcnls between Aquino and Salvador Laurel to join together on a united opposition ticket. A month before the election, the cardinal senl a lctter Lo all 2,000 parishes in the Philippines instructing Catholics to vole for “persons who embody the Gospel values of humilily, truth, honesty, respect for human rights and life.” This was unlikely 10 leave

WHY?

85

any doubt in people’s minds as to whom

the cardinal was

supporling, but he followed il up anyway with a virtually ex-

plicit endorsemeni of Aquino. Alter Marcos attempled to sleal lhe election and the military revolt occurred at Camp Craeme, he used the Church organization and the Church radio stalion to mobilize the populace on Lhe military’s behalf. “The religious overlone of the lhree-day revolt could not be mistaken as nuns and priests occupied the front lines of the human barricades, and rebel generals hoisted up the statue of the Virgin Mary before the crowd. After Marcos had finally gone io Hawaii, Cardinal Sin presided over a triumphal thanksgiving mass at the Luncta, chanling ‘Cory’ and flashing the Laban sign.”© Cardinal Sin may have played a more active and more powerful role in bringing about the end of 2 regime and a change in national polilical leadership than any Catholic prelale since the seventeenth centu Allin all, il it were not for the changes within the Catholic Church and the resulting actions of the Church againsl authorilarianism, fewer third wave (ransitions to democracy would have occurred and many that did occur would have occurred lalcr In country after country the choice between d and auth isn becamep ified in the conflict berween the cardinal and the dictator. Catholicism was second only to economic developmcent as a pervasive (orce making for democratization in the 1970s and 19803. The logo of the third wave could well be a crucifix superimposed

on a dollar sign,

NFEW POLICIES OF EXTERNAL ACTORS

Democralization in a couniry may be influenced, perhaps decisively, by the actions of governmenls and instilutions external to thal counlry. In fifteen of the bventy-nine democratic countries in 1970, as Robert Dahl pointed out, dernocralic regimes were instituted either during periods of foreign rule or as the country became independent from foreign rule.* Obviously foreign actors may also overlthrow demo-

66

THE THIRD WAVE

cratic regimes or prevenl countries that might otherwise become democratic {rom doing so. Foreign actors can perhaps best be thought of as hastening or relarding Lhe effects of economic and social development on democratization. As has been pointed out, when countries reach a certain social and economic level, Lhey enter a transition zone where the probability of their moving in a democratic direction incveases kedly. Foreign infl may lead lo d izalion elforts before countries reach that zone or they may retard or prevent democratization by countries which have reached that level of developmenl. Jonathan Sunshine, for instance, argues thal external influences in Europe before 1830 were fundamentally antidemocratic and hence held up democratizalion. Betwecen 1830 and 1930, however, lhe external environment was neulral wilh respect to demacralization; hence democraticization proceeded in different countries more or less at Lhe pace sel by ic and social
Juan Carlos exercised his power and prerogatives lo the full in moving Spain toward democracy, not leas! in the surprise selection of Sudrez as prime minister. Botha and Gorbachev, as we have seen,

created

powerful

new

presidential

offices

for

(hem-

selves. Salinas dramalically asserted his powers during his first years as Mexico's president. The [irsl requirement for reform leaders was to purge the governmental, military, and, where appropriate, party buies, replacing d in top offices with supporters of reform. This was lypu:ally done in selective fashion s0 as nol to provoke a strong reaction and so as to promote fissions within the slandpalter ranks. In addition to weakening standpatters, reform leaders also tried to reassure and convert them. In military regimes, the reformers argued that it was time to go back, after a necessary but limiled authorilarian i to thed iples Lhal were the basis of their cuunlrys political sysrem In this sense, they appealed for a “return to legitimacy.” In the nonmilitary authoritarian systems, reformers invoked “backward legilimacy” and stressed elemenls of continuily with the past.» In Spain, for instance, the monarchy was reeslablished and Suarez adhered to the provisions of lhe Franco constilution in abolishing Lhal conslilulion: no Francoist could claim that

there were procedural irregularities. In Mexico and South Af-

rica the reformers in (he PRI and National party casl themselves in the radilions of those parlies. On Taiwan the KMT reformers appealed Lo Sun Yal-Sen’s three principles. Backward legilimacy had two appeals and two eflects: it legitimated the new order because it was a producl of the old, and it retrospectively legitimated the old order because it had produced the new. Il elicited consensus from all except apposition extremists who had no use for either the old authoritarian regime or the new democratic one. Reformers also appealed lo slandpatters on the grounds that they were preempting the radical PP and hence mini 3 n-

HOW? PROCESSES OF DEMOCRATIZATION

139

slability and violence. Suarez, for inslance, asked the Spanish army to support him for these reasons and the dominani elemenls in Lhe army accepted the transition because there “was no illegitimacy, no disorder in the streels, no significant threat of breakdown and subversion.” Inevitably, the reformers also (ound that, as Geisel pul it, lhey could “not advance without some retreals” and that hence, on occasion, as in the 1977 “April package” in Brazil, they had to make concessions (0 the standpatters.”™ Coopting the opposition. Once in power the democralic reformers usually moved quickly lo begin the process of democralization. This normally involved consultations with leaders of the opposition, the pohmal parties, nnd major social groups and inst In some ively formal negotiations oceurred and quite explicit agreements or pacis were reached. In olher cases, the consullations and negotiations were more informal. In Ecuador and Nigeria the gove menl appointed commissions to develop plans and polici for the new

system.

In Spain,

Peru,

ngerla, and eventually

in Brazil elected blies drafted new In several instances referenda were held to approve the new constitutional arrangements. As the reformers alienated standpallers within the governing coalition, they had to reinforce themselves by developing supporl within the opposition and by expanding the palitical arena and appealing to the new groups lhal were becoming

politically active as a resull of the opening. Skillful reformers

used the increased pressure [rom lhese groups for democratizalion to weaken the standpatters, and used Lhe Lhreat of a slandpaller coup as well as the attraclions of a share in power to strenglhen moderate groups in the opposition. To these ends, reformers in government negotiated with the principal opposilion groups and arrived al explicit or tacit agreements with them. In Spain, for instance, the Communist parly recognized that it was too weak to follow a “radical rupturista policy” and instead went along with a *'rupiura pac-

140

THE THIRD WAVE

fada” even though the pacl was “purely lacil.” In Oclober 1977 Suarez won the agreement of the Communist and Socialisl parties Lo Lhe Pactos de la Moncloa comprising a mixture of fairly severe economic austerity measures and some social reforms. Secrel negotialions with Sanliago Carillo, the principal Communist leader, “played on the PCF. [Partido Comunista

de

Espana]

leader’s

anxiety

to be

near

the

levers

of

power and secured his backing for an austerity package.”? in Hungary explicit negotiations occurred in the fall of 1989 between the Communist party and the Opposilion Round Table representing the principal other parties and groups. In Brazil informal understandings developed betwcen the government and the opposilion parlies, the N D: ico Brasileiro (MDB) and the Partido Movimento

Democratico Brasi-

leiro (PMDB). On Taiwan in 1986 the government and the opposilion arrived at an und ing on the within which political change would take place and, in a week-long conference in July 1990, agreed on a full schedule of democralization. Moderation and cooperation by the democratic opposition—tiheir involvement in the process as junior partners— were essential su:cessful Lransformation. In almost all

the princip

rties—the

MDB PMDB

jsts and ¢Commurusls in Spain, the Democratic Progressive Parly (DPP) on Taiwan, the Civic Forum

in Hungary, the Alianza Popular Revolucionaria Americana

(APRA) in Peru, the Christian Demodrats in Chile—were led by moderates and followed moderate policies, at times in the face of iderable provocation by d groups in the government.

Skidmore’s summary of whal occurred in Brazil neally calches the central relationships involved in transformation processes: In

the

end,

liberalizalion

was

Lhe product

of an

inlense

dialectical relationship between the government and the op-

HOW? PROCESSES OF DEMDCRATIZATION

141

position. The military who (avored nbertura had lo proceed

cauliously, for fear of arousing the hardliners. Their overtures

to the opposition were designed to draw out the “responsible”

elements, thercby showing there were moderales ready 10 co-

operate with the governmenl. Al the same time, the opposition constantly pressed the government lo end ils arbilrary excesses, thereby reminding the mililary thal their rule lacked the had to remind the radicals that they would play into the hands of the hardliners if they pushed (oo hard. This intricate polilical relationship functioned successfully because there was a consensus among bolh military and civilians in (avor of a return to an (almost) open political system.” Guidelines for Democratizers 1: Reforming Authoritarian Systems The principal lessons of the Spanish, Brazilian, and other transformalions for democratic reformers in authoritarian governments include the following: (1) Secure your political base. As quickly as possible place supporlers of democratization in key power positions in the government, the party, and the military,

(2) Maintain_backward lcgmmncy, that is, make changes

through the

f the

regime

and reassure slandpaner groups wnh symbeolic concessions, following a course of twa steps forward, onc step backward. (3) Gradunlly shift your own cunsnluem-y s0 as lo reduce your ir

g change

and

to

brmden your cnnsllmency in the dlredmn of opposition groups supporting democracy.

{4) Be prepared for the standpatters Lo take some extreme aclion to slop change (c.g., a coup attempt)—possibly cven stimulate them to do so—and then crack down on them ruthlessly, isolating and discrediling the more extreme opponents of change. (5)

Scize and keep control of the initiative in the democrali-

zation process. Only lead from strength and never introduce demacratization measures in fesponse o obvious pressure from more extreme radical opposition groups.

142

THE THIRD WAVE

(6) Keep expectations low as (o how far change can go; talk in terms of maintaining an ongoing process rather than achieving some (ully elaborated democranc ulopm @) position party, which the key groups in society (including the mililary) will accepl as a plausible nonthreatening alternative governmenl.

(8) Creale a sense of inevitability aboul the process of demac-

1alization so that il becomes widely accepted as a necessary and natural course of development even if Lo some people it remains an undesirable one. REPLACLMENTS

Replacements involve a very dilferent process from transformations. Reformers within the regime are weak or nonexistent. The dominant elements in government are slandpallers staunchly opposed to regime change. Democratization consequently results from the opposition gaining strength and the government losing sirength until the governmeni collapses or is overthrown. The [ormer opposition groups come to power and Lhe conflict then often enters a new phasc as groups in the new governmenl struggle among Ihemselves

over the nature of the regime they should institute. Replace

ment, in short, involves three distinci phases: Ihe struggle to

produce the fall, The (all, and the struggle after the (all. Mosl third wave democratizations required some cooperation from Lhose in power. Only six replacements had oc-

curred by 1990. Replacements were rare in transitions from one-parly systems {(one oul of eleven) and military regimes

(wo out of sixleen) and more common in transitions from

personal dictatorships (three oul of seven). As we have pointed out, with some exceptions (Gandhi, Evren, Pinochet), leaders who created authorilarian regimes did not end those regimes. Changes of leadership within authorilarian systems were

much more likely in military regimes through “second phase"

coups of, in one-party systems, through regular succession or the aclion of constituled parly bodies. Personal diclators,

HOW? PROCESSES OF DEMOCRATIZATION

143

however, seldom relired voluntarily, and the nature of their power—personal rather than mililary or organizational— made it difficult for opponents within the regime lo oust them and, indeed, made it unlikely that such oppanents would exisl in any significant numbers or sirength. The personal dictator was Lhus likely to hang on until he died or until the regime itself came (0 an end. The life of the regime became the life of the dictator. Politically and at times lilerally (e.g., Franco, Ceausescu)

the deaths of the dictator and the regime

coincided. Democratic reformers were notably weak in or missing from the authorilarian regimes that disappeared in replacements. In Argentina and Greece, the liberalizing leaders Viola and Papadopoulos were forced oul of power and succeeded by military hard-liners. In Portugal Caetano initiated some liberalizing reforms and then backed away from them. In the Philippines, Romania, and East Germany, the entourages of Marcos, Ceausescu, and Honecker contained few if any democrats or even liberals. In all six cases standpaliers monopolized power, and the possibility of initiating reform from within was almosl totally absent. An aulhoritarian system exists because the government is politically stronger than the opposition. It is replaced when the government becomes weaker than the opposilion. Hence replacement requires the opposition lo wear down the goverament and shift the balance of power in ils favor. When they were initiated, the aulhoritarian regimes involved in the lhird wave were almost always popular and widely supported. They usually had the backing of a broad coalilion of groups. Over lime, however, as with any government, their strength deteriorated. The Greek and Argentine mililary regimes sulfered the humilialiun of military defeal. The Portuguese and Philippine regimes were unable to win counterinsurgency wars, and the Philippine regime created a martyr and stole an eleclion, The Romanian regime followed policies that deeply anlagonized its people and isolated itself from

144

THE THIRD WAVE

them; hence it was ] ble to the ¢ lati balling ol the antiauth i hrough Eastern Eu rope. The case of Cast Germany was more ambiguous. Although the regime was relalively successful in some respecls, the inevilable comparison wilh West Germany was an inherenl weakness, and the opening of the transit corridor through Hungary dramatically undermined the regime's authorily. The parly leadership resigned in early December 1989, and a caretaker governmenl took over. The regime’s authority, however, evaporaled, and with il the reasons for the Easl German slale. The crosion of support for the regime somelimes occurred openly, bul, given lhe repressive character ot authorilarian regimes, il was more likely lo occur covertly. Authorilarian leaders were often unaware of how unpopular they were. Covert disalfection then manifested itsel{ when some triggering evenl expused lhe weakness of the regime. In Greece and Argentina il was mililary defeat. In Portugal and East Germany il was the explicit turning against the regime of its ultimate source of power—the army in Portugal, the S0~ viel Union in Fast Germany. The actions of the Turks, the Brilish, the Porluguese military, and Gorbachev galvanized and brought into the open the disaffection (rom the regime of other groups in those societies. In all these cases. only a few weak groups rallied to the support of lhe regime. Many people had become disaffecled from lhe regime but, because it was an ian regime, a ing event was required to crystalize the dlsal'fectwn are the uni ition; they oppose whalever regime exisls in their SDCIer By themselves, however, students do not bring down regimes. Lacking subslanlial support (rom other groups in lhe populalion, lhey were gunned down by the military and police in Greece in November 1973, Burma in Seplember 1988, and China in June 1989. The military are the ultimale support of regimes. 1! they withdraw their support, if they carry oul a coup against the re-

HOW? PROCESSES OF DEMOCRATIZATION

145

gime, or if they refuse lo use force against those who threaten to overthrow the regime, the regime (alls. In between the perpetual opposition of the students and Ui necessary support of the military are other groups whose supporl for or opposition to Lhe regime depends on circumstances. In noncommunist authorilarian systems, such as the Philippines, these groups tended to disaffect in sequence. The disaffection of the students was fallowed by lhal of intellectuals in general and then by Lhe leaders of previously exisling political parties, many of whom may have supported or acquiesced in the authoritarian Lakeover. Typically the broader reaches of Lhe ‘middle cl hite-collar workers, p Is, small i became ali d. In a Catholic counlry, Church leaders also were early and effective opponentls of the regime. If labor unions existed and were not totally controlled by the government, al some point they joined the opposition. So also, and most important, did larger business groups and (he bourgeoisic. In due course, the Uniled Stales or other foreign sources of support became disaffected. Finally and conclusively, the military decided not o support the government or actively lo side with the opposition against the government. 1n five out of six replacements, consequenily, the exceplion being Argenlina, military disaffeclion was essenlial to bringing down the regime. In the personal dictatorships in Portu-

gal, the Philippines, and Romania, this military disaffection was promoted by lhe dictator's policies weakening military

and

cor upting the officer corps,

and creating cmnpetmg pammdllary and :ecunry forces. Opposilion to the governmenl normally (Portugal was (he only exception) had Lo be widespread before the military deserted the government. If disaffection was nol widespread, it was either because the most probable sources of opposilion—the middle class, isie, religious group: small and weak or because the regime had Lhe support of these groups, usually as a result of successful policies for economic devel-

146

THE THIRD WAVE

opment. In Burma and China the armed forces brulally suppressed prolests that were largely student-led. In socielies that were more highly developed economically, oppusition lo authoritarianism conunanded a wider range of support. When Lhis opposilion took lo the streels in the Philippines, Easl Germany,

and

Romania,

military unils did

not

fire on

broadly representative groups of their fellow cilizens. A popular image of jons is thal govemmenks are broughl down by “people power, “ the mass of ged citizensd and (orcing a change of regime. Some form of mass achon did take place in almosl every third wave regime change. Mass demonstrations, protesls, and sirikes played central roles, however, in only about six transitions completed or underway at the end of the 1980s. These included the replacements in the Philippines, Fast Germany, and Romania, and the transplacemenls in Korea, Poland, and Czechoslovakia. In Chile frequent nass actions attempled, withoul success, Lo alter Pino-

chel’s plan for Lransformation. In East Germany. uniquely, both “exit” and “voice,” in Hirschman’s terms, played major roles, with protest laking the form first of massive depariure of citizens from the country and then of massive sireet demonstralions in Leipzig and Berlin. In the Philippines, Portugal, Romania, and Greece, when the regime collapsed, it collapsed quickly. One day the authorilarian government was in power, the nexl day il was not. In Argentina and Fast Germany, the authoritarian regimes were quickly delegilimated bul clung lo power while atlempting lo negotiale terms for the change in regime. In Argentina, the military g of General Reynaldo Bignone, which look uvenn ]uly 1982 im: 'alely after the l'alklands defeat, was " 1§ some regime control over the lransll.\on Ior six months. [n December 1982, however mounting pubhc opposilion and the deled to mass protests, a general slnke Bngnones schedulmg of elections, and the

HOW? PROCESSES OF DEMOCRATIZATION

147

rejection by the uniled opposition parties of the terms proposed by the military for the Lransfer of power. The aulhorily of Lhe lame-duck military regime continued to deteriorate unLil it was replaced by the Alionsin government elected in October 1983. “The military government collapsed,” one author observed;

“it had

no influence over the choice of candidales

or the election itself, it excluded no one, and reserved neither powers nor veto prerogatives for itsell in the (uture. In addition, it was unable (o guaranlee either its autonomy in relation 1o the future constitutional government or the prcrni of a fulure military policy, and, even less- -given lhe winning the basis lor an ag on the ongoing struggle against the guerrillas.”™ In East Germany in early 1990 a somewhat similar situation exisled, with a weak and discredited comununist governmenl clinging to power, and its prime minister, Hans Modrow, playmg the role of Blgnune The in and backward legitimacy was absent from replacements. The institutions, procedures, ideas, and individuals connected with the previous regime were insiead considered tainted and the emphasis was on a sharp, clean break with the past. Those who succeeded the authoritarian rulers based their rule on “forward legitimacy,” whal they would bring aboul in the (uture and their lack of involvement in or connection with

the previous regime. In

and

the leaders of the

aulhoritarian regimes usually lefl politics and went back to the barracks or private life quietly and with some respect and dignity. Authoritarian leaders who lost power through replacements, in contrasl, sulfered unhappy fates. Marcos and Caelano were forced into exile. Ceausescu was summarily executed. The mililary officers who ran Greece and Argentina were tried and imprisuned. In East Germany punishments were threatened againsl Honecker and other former leaders in notable contrast to the absence of such action in Poland, 1lungary, arnd Czechoslovakia. The dictalors removed by for-

148

TILE THIRD WAVE

eign intervention in Grenada and Panama were similarly subjected to prosecution and punishmenl. The peaceful collapse of an authorilarian regime usually produced a glorious if brie( moment of public euphoria, of carnations and champagne, absent from transformations. The collapse also crealed a polential vacuum of authority absenl from transformations. In Greece and the Philippines, the vacuum was quickly filled by the accession to power of Karamanlis and Aquino, popular polilical lcaders who guided their countries lo democracy. In iran the authorily vacuum was filled by the ayatollah, who guided Iran elsewhere. In Argenlina and Fast Germany the Bignone and Modrow governments weakly filled the inlerim between the collapse of the authorilarian regimes and the clection of democralic govemnmenls. Before the fall, opposition groups are united by their desire to bring aboul the fall. Alter the fall, divisions appear among them and they struggle over the distribulion of power and the nature of the new regime that musl be eslablished. The fale of democracy was determined by Lhe relative power of democratic moderates and anlidemocratic radicals. In Argentina and Greece, the authoritarian regimes had nol been in power for long, polilical parlies quickly reappeared, and an overwhelming consensus existed among polilical leaders and groups on the need quickly to rees!ahhsh democratic institutions. In the Philippines overt opp lod apari from the NPA insurgency, also was minimal. In Nicaragua, Iran, Portugal and Romania the abrupt col-

lapsc of the dictatorships led lo siruggles among the former

opposition groups and parties as to who would exercise power and whal sort of regime would be created. In Nicaragua and Iran the democratic moderates lost out. In Portugal, as was noted on the opening pages of this book, a state of revolutionary ferment existed between April 1974 and November 1975. A consolidation of power by the antidemocratic Marxist-leninist coalilion of the Communist party and lefi-

HOW? PROCESSES OF DEMOCRATIZATION

149

wing military officers was entirely possible. In the end, after

inlense struggles between military factions, mass mobilizations, demonstrations, and strikes, the military action by Fanes settled Portugal on a democratic course. “What started as a coup,” as Robert Harvey observed, “became a revolution which was stopped by a reaction before il became an anarchy. Out of the lumult a democracy was bon.” ® The chaices in Portugal were belween bourgeois democracy and Marxist-Leninist diclatorship. The choices in Romania in 1990 were less clcar, but democracy also was not inevitable. The lack of effectively organized opposilion parties and groups, the absence of previous experience with democracy, the violence involved in the overthrow of Ceausescu, lhe

deep desire for revenge against people associaled with the

dic bined with the widesp of much of the population wilh the diclalorship, the many leaders of the new governmenl who had been parl of the old regime—all did not augur well for the emergence of democracy. Al the end of 1989 some Romanians enthusiastically compared what was happening in their country to whal had happened (wo hundred years earlier in France. They mighl also have noled thal the French Revolution ended in a military diclatorship. Guidelines for Democratizers 2: Overthrowing Authoritarian Regimes suggests (he following The historyof rep ing 1o d i d for opposition P an authorilarian regime:*

guideli

*Myron Weiner has formulated a similar and more concise sel of recom mendalions: “For those who seek demoxratization the lessons are these: mo-

Villze large-scale non-violent oppusilion to the regime, seek support from the

cenler and, if necessary, from the conservative right, cestain the lelt and keep them from dorminating the agenda of the movement, woo smionsof the military, seek sympathetic coverage (rom the western medin, and the United $tates for support ” “Empirical Democrauc Theory and the ey sition from Authoritarianism to Demacracy.” PS 20 (Fall 1987), p. B66.

150

THE THIRD WAVE

(1) Focus altention on Lhe illegitimacy or dubious legitimacy

of the authoritarian regime; thal is its most vulnerable point. Al-

tack the regime on gencral issues thal are of widespread concem,

such as corruplion and brutalily. If the regime is performing success{ully (parlicularly economically) these atlacks will not be elfective. Once its performance falters (as it must). highlighting its illegitimacy becomes the single most important lever for dislodging it from power. (2) Like democrahc rulers, authoritarian rulels aver lime alien-

ale ers these d groups ta support democmcy as the nece»ary aliernative to the cur-

rent system. Make particular efforls |0 enlist business leaders,

middle-class professionals, religious figures, and politieal party leaders, mosl of whom pmbabfi' supported creation of the au-

thorilarian system. The more “respectable” and “responsible” !he c

osition 2 ippears, the easier it is Lo win more supporters.

ultivate genemls In the last analysis, whether the regime colhpse:, or not depends on whether they support the regime, join you in opposition to it, or sland by on the sidelines. Support

from the military could be helpful when the crisis comes, but all

you really need is military unwillingness Lo defend the regime. (4) Practice and preach nonviolence. (See pp. 196ff. below.) Among other Lhings, Lhis will make it easier for you to win over the securily [orees: soldiers do nol tend to be sympathetic to people who have been hurling Mololov cocktails at them. (5) Seize every opportunily to express opposition 1o the regime, including participalion in elections il organizes. (See below, pp. 185¢(.) (6) Develop :ontdcls with the global media, foreign human rights such as churches, In parmular mobdlze supporlers in the United Slales. American congressmenbers are always looking for moral causes (o get publicity for themselves and (0 use against the American

administralion. Dramatize your cause to them and provide them

with material for TV photo opportunilies and headline-making

speeches.

(7) Promote unlly among opposition gmups Allempl to crebrella that will facilitale conpemmn among such groups. This will be difficull and, as the enamples of the Philippines, Chile, Korea, and South Africa show, authorilarian rulers are oflen expert in promoting opposi-

HOW? PROCESSES OF DEMOCRATIZATION

151

tion disunily. One test of your qualifications to become a democratic leader of your country is your abilily 10 pvercome these obstacles and secure some measure of opposition unity. Remeinber Gabriel Almond’s truth: “Greal leaders are greal coalition builders.”» (8) When the authorilarian regime falls, be prepared quickly to fill the vacuum of authority thal results. This can be done by: pushing to the fore a popular, charismatic, democratically inclined leader; promptly organizing elections lo provide popular lo a new g and building leglhma:y by getting support of foreign and lrnnsnalmnal actors (international

organizations,

the

United

States,

the

European

Communily, Lhe Catholic Church). Recognize Lhat some of your former cvalilion partners will want

to establish a new dictator-

ship of their own and quietly organize the supporters of democracy to counler this effort if it materializes. TRANSPLACEMENTS

In d by the combined actions of governmenl and upposmon Within the govemment the balance between standpatlers and reformers is such that the government is willing lo negotiate a change of regime-—unlike the situation of standpatter dominance that leads to replacement—bul il is unwilling (o initiate a change of regime. It has lo be pushed and/or pulled into formal or informal negotialions with the opposition. Within the opposilion democratic moderates are strong enough Lo prevail over antidemocratic radicals, bul they are nol sirong enough

to overthrow the govenment. Hence they too see virlues in

negoliation, Approximately eleven of thisty-(ive liberalizations and de macratizations that oceurred or began in the 1970s and 19805 i d the . model. The most notable ones were in Poland, Czechoslovakia, Uruguay, and Korea; the regime changes in Bolivia, Honduras, El Salvador, and Nicaragua also involved significant elements of transplacement. In El Salvador and Honduras the negotiatlons were in part with the Uniled Slates government, acling as a surrogate

152

THE THIRD WAVE

for democrallc moderates. In 198y dnd 1990, South Africa began a process, and N lia and Nepal appeared lo be moving in that direction. Some features of transplacement were also present in Chile. The Pinochet regime was slmng Pnough however, to resisl opposilion pressure to and stubbornly adhered 1o the schedule Iur reglme v.hzmge lhnl it laid down in 1980, In the d groups in both g and ppositi cognized that they were ble of unil d g the nature of the future

political system in theis socicty. Government and opposition

leaders often developed these views after testing each other’s strength and resolve in a political dialectic. Initially, the op-

position usually believed thal it would be able to bring about

the downfall of the governmenl at some point in lhe nol loo distant future. This belief was on occasion wildly unrealistic, bul so long as opposition leaders held to il, serious negotialions with the government were impossible. In contrast, the government usually initially believed that it could eflectively contain and supp the opp without i unacceptable cosls. Transplacements occurred when the belie(s of both changed. The opposilion realized that it was not strong enough to overthrow the government. The government realized thal the opposition was strong enough (o increase significantly the costs of nonnegotiation in terms of increased repression leading to further alienation of groups from the

government, intensified divisions within the ruling coalition,

increased possibility of a hard-line takeover of the government, and significant losses in inlernational legitimacy. ‘The transplacement dialectic often involved a distinct sequence of sleps. First, the government engaged in some liberalization and began 1o lose power and aulhorily. Second, the oppositi loited this I ing by and kening,of Lhe government (o expand ils support and intensify ils acli ties with the hope and expectation it would shorily be able lo bring down the g Third, the g reacted

HOW? PROCESSES OF DEMOCRATIZATION

153

y to contain and supp ilization of political power by the opposition. Fourlh, gl)vemmenl and opposition leaders perceived a slandolfl emergmg and began to explore the possibilities of a neg ition. This fourth step was nol, however, inevitable. Conceivably, the government, perhaps after a change of leadership, could brutally use its military and police forces lo restore ils power, at least lemporarily. Or the opposilion could conlinue to develop its slrength, further eroding the power of (he government and eventually bringing about its down/all. Transplacements thus required some rough equality of sirength between government and opposition as well as uncertainty on each side as to who would prevail in a major test of strength. In these circumstances, the risks of negotiation and compromise appeared less than the risks of confrontation and catastrophe. The political process leading to transplacement was thus often marked by a seesawing back and forth of strikes, protests, and demonstrations,

on the one hand, and repression,

jailings, police violence, stales of siege, and martial law, on the other. Cycles of protesl and repression in Poland, Czechoslovakia, Uruguay, Korea, and Chile eventually led to negoHaled agreemenls between government and opposition in all cases except that of Chile. In Uruguay, for instance, mounting prolests and demonstrations in Lhe (all of 1983 stimulated Lhe negotiations leading to the military wilhdrawal from power. In Bolivia in 1978

“a series of conflicts and protest movements” preceded the

military’s agreeing to a limetable for elections.™ Tn Korea as in Uruguay, the military regime had earlier forcefully suppressed protesls. In the spring of 1987, however, the demonstralions became more massive and broad-based and increasingly involved Lhe middle class. The government first reacted in its usual fashion bul then shifted, agreed Lo negotiate, and accepted the central demands of the opposition. Ln Poland the 1988 sirikes had a similar impact. As one commentator explained, “The strikes made the round tabie nol only pos-

154

THE THIRD WAVE

sible, but necessary—for both sides. Paradoxically, the strikes were strong enough to compel the communists to go to Lhe round table, yet too weak to allow Solidarity’s leaders lo refuse negotiations. That's why Lhe round table talks took place.” ¥ In transplacements, the eyeball-to eyeball confrontation in the central square of the capital between massed protesters and scrried ranks of police revealed each side’s strenglhs and weaknesses. The opposition could mobilize massive supporl; the government could contain and withstand opposition pressure. Polilics in South Africa in the 1980s also evolved along the lines of the (our-step model. In the late 1970s P. W. Botha began the process of liberalizing reform, arousing black expectations and then [rustrating them when the 1983 constilution denied blacks a national political role. This led lo uprisings in the black townships in 1984 and 1985, which stimulated black hopes Ihal the collapse of the Afrikaner-dominated regime was imminent. The government’s forceful and effective suppression of black and while dissent then compelled the opposition dmsllcally Lo revise lhm hopeb At the same time, the uf gs allracled i ion, stimulated d ion of both the id system and the governmenl's tactics, and led the United States and European governmentls Lo intensify economic sanctions against South Africa. As the hopes for revolution of the African NaHonal Congress (ANC) radicals declined, Lhe worries of the Na-

lional parly government about international legitimacy and

the economie fulure increased. In the mid-19705, Joe Slovo, bead of the South African Communist party and the ANC's military organization, argued that the ANC could overthrow the government and win power Lhrough sustained guerrilla warfare and revolution. In the lale 1g80s he remained committed to the use of violence, but saw negotialions as the more likely route lor achieving ANC goals. After becoming president of South Alrica in 1989, T. W. de Klerk also empha-

HOW? PROCRSSES OF DEMOCRATIZATION sized (he

imp

of

iations.

155 The lesson of

Rhod;

he said, was that “When the opporiunity was there for real, constructive negolialion, il was nol grasped. . .. Tt went wrong because in the reahry of lheu cm:umslanccs lhey wailed too long before g in and

dialogue.

We

must

not make

that mistake,

and we are

determined not to repeal thal mistake.”® The lwo palitical leaders were learning from their own experience and that of others. In Chile, in contrasl, lhe government was willing and able Lo avoid negotiation. Major strikes erupted in the spring of 1983, bul a national general strike was suppressed by the government. Beginning in May 1983 Lhe opposilion organized massive monthly demonstrations on “Days of Nalional Pro-

test.” These were broken up by the police, usually with sev-

eral people being killed. Economic problems and the opposition prolests forced the Pinochet government lo iniliate a dialoguc with the opposition. The economy Lhen began to recover, however, and the middle cla; became alarmed at the breakdown of law and order. A national strike in October 1984 was pul down with consMerable bloodshed. Shortly the g pused the slate of siege that had been cancellcd in 1979. The opposilion efforts thus failed lo cvexlhmw the _government or to induce il to engage |n i The iti ad

its strength and deres et government's.”* It had also underestimated PinocheUs tenacity and political skill and

the willingness of Chilean security forces te shoot unarmed civilian demonstrators. Transplacements required leaders on both sides willing to risk negoliations. Divisions of opinion over negotiations usually exisled within governing eliles. At times, the top leaders had lobe p by their colieagues and by cire to negoliate wilh the opposition. In 1989, for instance, Adam Michnik argued that Poland, like Hungary, was following

“the Spanish way 1o democracy.” At one level, he was right

156

THE THIRD WAVE.

in that both the Spanish and Polish Lransilions were basically peaceful. At a more particular level, however, the Spanish analogy

did

nol hold

for Poland because Jaruzelski was

not

a Juan Carlos or Suarez (whereas Imre Pozsgay in Hungary in considerable measure was). Jaruzelski was a reluctant democral who had to be pushed by the deterioration of his counlry and his regime into negotiations with Solidarity.* In Uruguay the president, General Gregorio Alvarez, wanted lo prolong his power and postpone democratization and had lo be forced by the olher members of the mililary junta to move ahead with (he regime change. In Chile, General Pinochet was somewhal similarly under pressure from other junla liy the air force General Fernando Vlaltlu:n, lo be more forthcoming in dealing with the opposition, bul Pinochel successfully resisted this pressure. In other countries changes occurred in the top leadership belore serious negoliations with the opposition began. In Korea the government of General Chun Doo Hwan followed a slaunch dp policy of i pposilion demands and suppressing opposition nclwny in 1987, however, lhe governing party designated Roh Tae Woo as its candidale lo succeed Chun. Roh dramalically reversed Chun’s policies, announced a political opening, and enlemd mlo negolialions with the opp leader.® In Czect ia the I C ist parly general secremy, Gustav Ilusak, was succeeded by the mildly reformist

Milos Jakes in December 1987. Once the opposition became

mobilized in the fall of 198y, however, Jakes was replaced by

the reformer Karel Urbanek. Urbanek and the reformist prime minister, Ladislav Adamec, then negotiated arrangements for the Lransilion to democracy with Vaclav Havel and the other leaders of the opposition Civic Forum. In South Africa, de Klerk moved beyond his pred ‘s aborled tion process (rom above to L1ansplacement-lype negotialions with black opposition leaders. Uncerlainty, ambiguity, and

HOW? PROCESSES OF DEMOCRATIZATION

157

division of opinion aver democratization thus tended o characterize lhe ruling circles in transplacemen situations. These regimes were not overwhelmingly committed either to holding on lo power ruthlessly or lo moving decisively loward democracy. Disagreement and unceriainty existed not only on the govemment side in transplacements. In (acl, the one group more likety lo be divided against ilself than the leaders of a decaying ian g are the opp leaders who aspire to replace them In replacement situations the government supp the opposilion and the opposilion has an overriding common interest in bringing down the government. As the Philippine and Nicarag; les indicate, even under these condilions securing unily among opposition leaders and parties may be extremely difficult, and the unity achieved is often tenuous and [ragile. In transplacements, where it is a question nol of overthrowing the government bul of negotiating with il, opposition unity is even more dilficult (o achieve. Il was not achicved in Korea, and hence the governmenlal candidale, Roh Tae Woo, was clected president with a minority of the vote, as the two opposition candidales split the antigovernmenl majority by opposing each other. In Uruguay, because its leader was slill imprisoned, one opposilion party—lhe National party—rejected the agreement reached belween the lwo other parties and the military. In South Afric.-. a major obslacle to democratlc reform was the many divisions within the opposition belween parli and nonparliamenlary groups, Afrikaner and Enyhsh bla:k and white, and among black idcological and (ribal groups. At no time belore the 1990s did the South Alrican government confront anything bul a bewildering multiplicity of opposition groups whose differences among themselves were oficn as great as their differences with the goverument. In Chile the opposilion was seriously divided into a large number of parlics, factions, and coalitions. In 1983, the mod-

158

TIE TIIRD WAVE

crate centrist opposition parties were able to join together in the Democralic Alliance. In August 1985 a broader group of 2 dozen parties joined in the National Accord calling for a transition to democtacy. Yet conflicts over leadership and tactics conlinued. [n 1986 ' the Chilean opposilion mobilized massive prolests, hoping lo duplicate in Sanliago what had just happened in Manila. The oppaosition, however, was divided and its militancy frighlened conservatives. The problem, as one observer pul it al lhe lime, was lhal “the general is not being that has got itself lugelher under the leadershlp of a respected figure. There is no Chilean

Cory.Ӵ

In Poland,

on

the other hand,

things were different. Lech Walesa was a Polish Cory, and .:olu:lanly dnmmalv:d the npposmcn for mosl of a decade. In occurred so quickly that

dlf[crcuccs among opposition political groups did not have

lune to materialize. d. have lo be sufficlenlly stmng within the opposition to be credible negotiating partners wilh the governmenl. Almost always some groups within lhe opposilion reject negolialions with the govemnmenl, They fear lhal negotialions will produce undesirable compromises and they hope that continued opposition pressure will bring about the collapse or the overthrow of the

government. In Poland in 1988-8g, right-wing opposition

groups urged a boycott of the Round Table talks. In Chile lefi-

wing opposition groups carried out lerrorist attacks Lhat undermined the efforts of the moderale opposilion lo negoliate with the governmenl. Similarly, in Korea radicals rejecled the agreement on eleclions reached by the government and the leading opposition groups. In Uruguay, the opposition was dominated by leaders of moderate political parties and extremists were less of a problem. For negotiations to occur each party had to concede some degree of legitimacy to the other. The opposition had to rec-

HOW? PROCESSES OF DEMOCRATIZATION

159

ognize the government as a worthy partner in change and implicitly if not explicitly acquiesce in its current right to rule. The government, in turn, had lo accept the opposition groups as | of of soci ety. The ),ovemmenl g could do this more ensnly if the opposition groups had not engaged in violence. Negoliations were also easier if the opposilion groups, such as polilical parties under a military regime, had previously been legitimate parlicipants in the political process. It was easier for the opposilion to negotiale if the government had used only limited violence against il and if there were some democratic reformers in the government whom il had reason 1o believe shared its goals. n I unlike i and replacemenls,

govemmenl

viously

had

leaders often negotialed the basic lerms

of the regime change with opposition leaders (hey had preunder

arrest:

Lech Walesa,

Vaclav

[Havel, Jurge

Ballle Ibanez, Kim Dae Jung and Kim Young Sam, Waller Sisulu and Nelson Mandela. There were good reasons for this. Opposition leaders who have been in prison have not been fighting Lhe gov violently or folently; Lhey have been living with it. They have also experienced the realily of government power. Governmental leaders who released their captives were usually interesled in reform, and those released were usually moderate enough (o be willing Lo negotiate with their former captors. Imprisonment alsv enhanced the moral

authority of the former prisoners. This helped them to unile

the opposilion groups, al least lemporarily, the prospect {o the govemmen( lhat they of Lheir fol rearhed Al one poinl in lhe Brazilian transition, reportedly told an opposilion leader, “You under control and we will contrul ours.”® under control ofien requires the cooperalivn

and lo hold out could secure the was General Golbery get your radicals Getting radicals of the other side.

160

THE THIRD WAVE

In transplacement negotialions, each party has an inlerest in strengthening the other party so thal he can deal more effectively with the exiremists on his side. In June 1990, for in-

stance, Nelson Mandela commented

on the problems F. W. de

Klerk was having with white hard-liners and said thal the ANC had appealed “(o whites Lo assist de Klerk. We are also trying to address Lhe problems of white opposition to him. Discussions have already been started with influential sectors in Lhe right wing.” At the same time, Mandcla said that his own desire to meel with Chief Mengosuthu Buthelezi had been vetoed by militants within the ANC and that he had to accepl that decision because he was “a loyal and disciplined member of the A.N.C.”™ De Klerk obviously had an interest in strengthening Mandela and helping him deal with his militant left-wing opposilinn Neguhahons [or regime change were at times preceded by ions” about the diti for entering into negohallons In South Atfrica, the government precondition was that the ANC renounce violence. ANC preconditions were that the government urban opposition groups and release political prisoners. In some cases prencgotiations concerned which opposition individuals and groups would be involved in the negotiations. were imes lengthy and imes brief. They uhen were interrupted as one parly or ihe other broke themn off. As the negoliations conlinued, however, the politi-

cal {ulure of each of the parties became more engaged with

their success. If the negotiations failed, standpallers within the governing coalition and radicals in the opposilion stood ready lo capilalize on that failure and to bring down lhe leaders who had engaged in negotiations. A common inferest emerged and the sense of a common fate. “/{T}n a way,” Nelson Mandela observed in August 1990, “there is an alliance now” between the ANC and the National party. “We are on one boal, one ship,” agreed National Party leader R. F. Bolha, “and the sharks to the lefl and the sharks to the right are not

HOW? FROCESSES OF DEMOCRATIZATION

161

going lo dlslmbmsh between us when we fall overboard.” ly. a Lhe parties became more w:llmg lo compronuse in order Lo reach an agreement. The agreements Lhey reached often generated attacks (rom others in governmenl and opposilion who thouglht the negotialors had conceded oo much. The specific agreements reflecled, of course, issues peculiar to their countries. Of central |mpon:mce in almosl all negotiations, hawever, was Lhe of In former officials of the authoritarian regime were almost never punished; in replacements Lhey almost always were. In transplacements this was often an issue lo be negotialed; the military leaders in Uruguay and Korea, for instance, demanded guaranlces agamsl prosecution and punishmenl for any human rights In other situali invulved arrangements for the sharmg ol power or for changes in power Lhrough elections. In Poland each side was guaranteed an explicil share of the seals in the legislature. In Czechoslovakia positions in the cabinet were divided between the lwo parties. In both these countries coalition governments dc isl and the opposilion thal their interests would be protected during the transition. In Korea lhe governing parly agreed lo a direct, open election for the presidency on the assumption, and possibly Lhe understanding, that at leasl two major opposition candidates would run, thereby making highly probable victory for Lhe government party’s candidale. The risks of con(rontalion and of losing thus impel governmenl and oppusilion lo negoliate with each other; and guarantees that neither will lose everything become the basis for agrecment. Bolh gel the opportunily lo share in power or to compele [or power. Opposition leaders know they will not be senl back lo prison; govemment leaders krow they will not have 10 flee into exile. Mulual reduction in risk prompts reformers and moderales to cooperate in establishing democracy.

162

THE

THIRD

WAVE

Guidelines for Democratizers 3: Negotiating Regime Changes For democratic refunm»rs in gwemmem:

0]

g th

or

syslems (see above, pp. 191-42), first isolate and weaken your and your hold on the government and polmr.al machinery. (2) Also

lollawmg lhosc gu|delmes,

surprise bolh

opp

Ip

seize

the initialive and

wilh the

you are willing (o make, bul never make concessions under ob-

vious opposilion pressure.

(3) Secure endorsement of the concepl of negotiations from

leading, generals or other top officials in lhe securily establishment.

(4) Do what you can 10 enhance lhe stature, authorily, and

moderation of your principal apposition negoliating partner. Lstablish confidential and reliable back-channels (or negotiating key cenlral questions with opposition leaders. (6) i the negotiation succeeds, you very probably will be in the opposition. Your prime concern, consequently, should be se-

curing guarantees and safeguards for the rights of the oppusition and of groups that have been associated wilh your government

(e-g., the mllnary) Everything else is negotiable. For d

in the

(1) Be prepared 1o mobilize your supporters for demonsirations when these will weaken the slan atters in the government. Too many marches and protests, however, are likely to

strengihen them, weaken your negotiating partner, and arouse middle-class concern aboul law and order. (2) Be moderate; appear stalesmanlike,

(3) Be prepared to negotiale and, il necessary, make concessions on all issues excepl the holding of free and fair elections.

(4) Recognize the high probability that you will win those

elections and do not take aclions that will seriously complicale your governing your counlry.

For both g

(1) The political

and

favorable loa

11OW? PROCESSES OF DEMOCRATIZATION

163

will not lasi indefinitely. Seize the opportunity they present and move quickly to resolve the central issues. ) Recogmze thal your political future and thal of your partner depend on your success in reaching agreement on the {ransilion lo democracy. (3) Resisl the demands of leaders and groups on your side that cilher delay the negolialing process or threaten the core interest of your negotialing partner.

(4) Recognize that that agrecment you reach will be the only

radicals and d) il, but Lhey will nol be able 10 produce an allemanve that commands broad sup) (5) When in doubl, compromise.

CHAPTER 4 HOW? CHARACTERISTICS OF DEMOCRATIZATION

THE THIRD WAVE DEMOCRATIZATION

SYNDROME

COUNTRIES TRANSITED TO DEMOCRACY in different ways. For all their differences, however, third wave Lransformalions, replacements, and transplacements had important characteristics in common. Of the more than twenty-flive democratizations that had occurred by or appeared to be underway in 1990, only lwo, Panama and Grenada, were the result of foreign invasion and imposition. Mosl of the other transitions were alike in whal they lacked. With the paxlldl and debatable of > no regime was brought down by a prolonged guerrilla insurgency or civil war. What mighl be termed revolulionary upheavals occurred in two

cases, Portugal and Romania, bul the Porfugucse revolution

involved very little violence and the Romanian revolution was an urban uprising assisted by the armed forces and was very briel. Significant fighting between armed military units occurred only in Romania, the Philippines, Bolivia, and Nicaragua. lixcept in lhe Philippines, Romania, and Fast Germany, outraged mobs of citizens did nol slorm into presidential palaces. How were democracies made? They were made by the methods ol democracy; there was no other way. They were made through neg: i and 164

HOW? CHARACTERISTICS OF DEMOCRATIZATION

165

They were made through demonstralions, campaigns, and eleclions,

and

through

the

nonviolenl

resolution

of differ-

ences. They were made by polilical leaders in governments and oppositions who had the courage both (o challenge (he status quo and to subordinate Lhe immediale interests of their foll to the longer-lerm needs of d . They were made by leaders in both government and uppm\lmn who withslood the provocalions lo viclence of opposition radicals and government standpatlers. They were made by leaders in governmen! and opposition who had the wisdom to recognize that iin pohllcs no one hasa monopoly on Lruth or viriue. C ! , ani 1 were the third wave democratization syndromc In varying degrees Lhey characterized most of the transformalions,

placements of that wave. COMPROMISE

replacemenits, and trans-

AND THE PARTICIPATION/

MODERATION TRADE-OFF

Negotiations and compromise among political elites were at the heart of lhe democralization processes. The leaders of the key political forces and social groups in society bargained with each olher, explicilly or implicitly, and worked out acceptable if not sahsrymg arrangemenls for the transilion to democracy. Neg and ag) between reformors and moderates werc, of course, the central elements of transplacements. In transformations the pro-

cess was oflen implicil, as the reformers in the government

opened up Lhe political process and opposition groups modified their demands and moderated their laclics to take part in thal process. At times also explicil agreements were reached between reformers Icading the transition and opposition inoderates, whom Ihey wished to coopt. [n both replacements and P democratic opposilion groups negotinled agreements among, hemselves. Om:c in power through former opposi normally steered middle courses making concessions as needed Lo refarmers,

166

THE TILRD WAVE

slandpatlers, and radicals. Whether the initiative for democralization came (rom lhe government, from the opposition, or from both, al some poini the key players reached agreements on We crucial aspects of the democratization process and the new system Lhat was to be created. The agreements on introducing democracy took many forms.

The

transilions in Brazil, Peru,

Ecuador,

and

Bohvla

were generally ct erized by “tentalive between Lhe opposition and an ofl"nal carelaker coalition lrying to manage the Lransition from the authoritarian regime.” These understandings usually only involved “lacil agreement on some procedural—primarily clecloral— ground rules for the transition.” In other inslances, regime changes resembled the second wave Lransilions in Colombia and Venezuela in 1957 and 1958 in which very explicit pacts were negohated among the inleresled parties.! In Spain, the government under Juan

Carlos

and

Suarez

dominated

the

transition,

but

government and opposition engaged in “Lhe politics of compromise” in machmg & in the on the I for the new d and in the so-called Pact of Moncloa in October 1977. In this pact all the relevani political parties, including the Socialists and Communists, agreed on a comprehensive economic program including limitalions on wages, currency devaluation, monetary policy, increased public invesiment, restrictions on social security, lax reform, trade union aclivities. control of nationalized induslries, and other matlers.” In Poland Solidarity and the Communist party negotiated the Round Table agreements

in March and April 1989. In Hungary government and oppo-

sition leaders during the summer of 1989, Laler thal (all governmenl and opposition leaders arrived at arrangements for lhe lransilion in Czechoslovakia. In Uruguay the military and party leaders reached agreemenl in the Club Naval Pact of August 1984. In almost all cases the principal parucnpams wete Lhe leaders of government and opposition political parties. In many in-

HOW? CIIARACTERISTICS OF DEMOCRATIZATION

167

stances, implicit or explicil agreements were also reached wilh the leaders of key sucial and inslitutional forces in Lhe society, including, besides the military, the business community, labor unions, and, where appropriale, the church. Whelher or not formal pacls were negotiated, agreement—

implicit or explicil—was easier (0 reach when no vas! discrep-

ancy in power resources existed among the participants and when group leaders were able lo exercise subslantial control over their followers. Agreement also came more easily when the negoliations were carried out in secrel among a relalively small number of leaders. In Spain, even in the constitulional bly, key a were jated. as the Spanish said, “behind thc curtain.”? In Poland the Round Table negotialions were relatively public, but the mosl important issues were

discussed

in

“secrel,

parallel

talks . . . conducted

in

isolation from (he mass media™ at a villa at Magdalenka outside Warsaw. “Many fewer people ook part in the secret talks than in the open ones.” The leaders of the negotiations, General Czeslaw Kiszczak and Lech Walesa, appeared occasionally al the open lalks ”only to depart soon thereafter for Magdalenka, where

lin private.” C

the private talks, only “Vague and highly diplomatic communiques were issued, as il by the ambassadors of two countries thal had been al war unlil recently.”? It was at these

gs that the basic ag were reached between the Communist party and Solidarity. The compromises reached in negotiations al times crealed

Pproblems for polilical leaders when, as in Spain, they “would have been unacceptable to their [ollowers.” In Spain, Suarez emphasized, il was necessary to break with the past patlen of “imposition by some Spaniards on the rest” and instead develop a broad-gauged consensus. Agreement among political parties, however, another government leader pointed out, led to “a process of disaffection of Spanish sociely [rom polilical parties, because political parties in Spain at thal lime were not adequately performing Lheir function of represen-

168

THE THIRD WAVE

lation of interests.”* A Communist party journal commented on Lhe “disenchantment wilhin and outside our ranks™ alllicting the parly because “all parties are saying the same Lhing” and “there is no clear Communist identity in the PCE.” [n Bo-

livia in 1980 rank-and-file union members “widely criticized” the pact their leaders signed with the mililary and political leaders g ions; and in Ni in ti-Sandinista labor union bers similarly fell “betrayed by President-Elect Violeta Chamorro's agreeing Lo app()lnl Humberlo Ortega chief of the armed forces.* In l’oland ti pp s of both5i ily and the C par(y were alienated by the compromises made by leaders of those two organizations. General Jaruzelski was atlacked by Communist party comewallves “for havmg yielded loo mud\ power to the opposition” and for aband “fund; Communist tenets.” Lech Walesa, on the other hand, was attacked al union rallies for slarling negotiations with the government and denounced by radicals who had “grown to abhor his striving for compromise with the Communist aulhonlles " Sohdanry supporters pamcularly 0b|ev.ted to (he union k ip's supporl for as | of Poland. One leading Solidarity journalist commented that as a result of the 1989 negotialions over forming a government "both sides will be afraid of their respeclive electorates . am:l the logic of evenls will push the communis!| party and idarity inlo back hidden from the pub-

lic.”” In Hungary major elements of the agreement reached in

1989 by the C ist party and the opposition Round Tablewere hall d by di groups and defeated in a referendum two months laler. The collapse of aulhorilarian regimes was almost always exhilaraling; the creation of democratic regimes was often disillusioning. Few political leaders who pul logether the compromises creating those regimes escaped the charge of having “sold out” the interests of their constituents. The extent of this disalfection was, in a sense, a measure of their success.

110W? CHARACTERISTICS OF DEMOCRATIZATION

169

“How is the slate to be made democralic,” asked the Polish scholar Wojtek Lamentlowicz, “by two artificially uniled and oligarchically ruled constellations of political forces?” That, however, may be the most effeclive way. In Lhe third wave, democracies were oflen made by leaders willing to betray the interests of their followers in order lo achieve that goal * A central compromise in most cases of democratization was what mightbe termed “the democratic bargain,” the trade-off between participalion and moderation. Implicilly or explicilly g processes leading lo democralization, the scope of participation was broadened and more political figures and groups gained the opportunily lo compele for power

and to win power on the implicit or explicil understanding

thal they would be moderate in their tactics and policies. In this sense the third wave duplicated experiences of the first wave in Europe, when suffrage was extended to the working class and socialist parties abandoned their commitment to violent revolution and tempered their policies. In late nineleenlh-cenmry llaly Clovamu Giolilti pursued an expllcll pclicy of “deradicali by incorporation.” In lale century Spain, Porlugal, and Greece, moderalion was “the price of power,” and the socialist partics of those countries “won their victories and achieved the lolerance of their once mlranexgenl foes by dint of demdwal.\zanon e ian regimes, by definiti restrict political participation. The ruling groups in authoritarian sys-

tems often found particular opposition leaders and political

parties especially anathema. Democratization required accepling these groups as legilimate participants in politics. For decades the military of Peru and Argentina used force to prevent the Apristas and the Peronistas from achiev-

ing or exercising power. In the democratizalion processes in

those countries in the 1980s, the military accepted nol the participation in politics of their old opponents but eventually their acquisition of power. A few months after ing office in the Greck transition, Karamanlis legalized

only also takthe

170

THE THIRD WAVE

Greek Communist party. Facing a much more difficult situation and the severe threat of a right-wing coup, Suarez also the Spanish C ist party in April 1977; Juan Caclos played a key role in securing the “grudging acceptance” of this by the armed forces. In Uruguay the Club Naval agreement legalized Lhe leflist Broad Front. In the Brazilian Lransition Lhe ban on the old pre-1964 political leaders was lifted in 1979 and in 1985 the legislalure legalized previously outlawed Mandist parties and gave illiterales the vote. In 1689 Chilean voters amended their conslilution to legalize the Communist party." In 1987 in Turkey first the parliament and then the electorate revoked the mililary-decreed bans that prevented one hundred politicians from engaging in polilics. In South Africa in 1990 the African National Congress was unbanned, imprisoned political leaders were freed, and exiles were allowed to return lo their country. The other side of lhe democralic bargain was moderation in tactics and palicies by the included leaders and groups. This often involved lheir agreeing to abandon violence and any commitment to revolution, 10 accept existing basic social, economic, and political inslilutions (e.g., private properly and the market system, aulonomy of the military, the privileges of the Catholic Church), and to work (hrough elections and parliamentary procedures in order Lo achieve power and pul through their policies." In Lhe Spanish transilion, lhe

army accepted the sacialists and communists as parlicipants in Spanish politics, the socialists accepted capitalism, and the communists abandoned their republicanism and accepled the manarchy as well as special arrangements for the Catholic Church. By persuading his sacialist followers in 1979 to abandon their commitment to Marxism, Felipe Gonz4lez paved the way for their electoral viclory tliree years later. In Portugal, Mario Soares similarly led the socialists to the cenier, Coining back into power in 1983 as the head of a coalition including conservative parties, he accepled the need “to abandon the Marxist, collectivist and—to a lesser extent—

HOW? CHARACTERISTICS OF DEMOCRATIZATION

171

etaiist Jeanings, to which an imporiant section of Uheir party adhered,” and imposed a rather severe auslerity program. In Greece Papandreou distanced himself [rom the “more exIreme and polemical posilions” he had espoused in the past and acted with restraint in office.” In Peru APRA moved lo the center; in Argentina Lhe Peronists moved lo the right; in Poland Solidarity moved first lo the center and then lo the

right, In Brazil, Lhe opposition party, the MDB, “cooperaled

with the government’s political game. . . . the opposition was dinarily

d

Even with

ive cassacoes, the

MDB effeclively urged moderation on its more radical members.” In the referendum campaign on Pinochel in 1988 in Chile the opposition coalition similarly pursued a consciously and explicitly moderale course.™ Transitions were thus helped by the deradicalization of new participants and [ormer leftists. They were also helped if those who first came to power in the new regime were not far distant politically from the previous authoritarian rulers. Karamanlis was a reassuring moderale conservative (o lhe anticommunist Greek military, and Soares was a reassuring moderate socialist to at least some elements of the radicalized Porluguese military. Juan Carlos and Suérez had indisputable conservalive Francoist credentials. Aylwin, a Chrislian Democral, was conservative enought to satis(y the Chilean military. By and large, 1he inilial acquisition of power by conservative and centrisl leaders eased the transition lo democracy (rom hori regimes.' In hal similar fashion, the subsequent coming lo power of socialist leaders often eased the introduclion of economic reforms and austerity programs. The willingness and ability of leaders lo reach compromises were alfected by the prevailing attiludes toward compromise

in their society. Some cultures appear to be more favorable Lo compromise than are olhers, and lhe legilimacy of and value placed on compromise may vary over time in a society. Historically, Spaniards, Poles, and Koreans, for inslance, were

172

THE THIRD WAVE

noted more (or the value they placed on principle and honor than on compromise. Il can at least be hypolhesized that the tragic consequences of those emphases logether with socioeconomic development produced changes in Lheir national values by the 1g970s and 1980s. In all three countries political leaders manifested an appreciation of the need for compromise in order to move their countries toward democracy. When, for inslance, the governing and opposition parlies in Korea reached agreemenl on a new constitution, it was reported [hal lhe “committee of negotialors [rom the iwo parties did something today that Koreans themselves acknowledge is a rarity in their politics: They compromised.” s Democratization occurs in a society when thal rarity becomes a reality, and democracy is stabilized when thal realily becomes commonplace. Vistorically Lhe (irst efforts to establish democracy in countries frequently fail; second efforts vften succeed. One plausible reason for this pattern is that learning occurred, and that does appear lo be the case in several instances. Venezuela was a dramalic example in the second wave. The first meaningful effort Lo establish demucracy in Venezuelan history occurred during the so-called Irienio of 1945-48. In 1945 a military coup overlhrew the diclator and introduced democratic

politics thal for the next three years was dominated by the reformist Accién Democrélica (AD). The AD government pur-

sued radical policies that alicnated many groups and led to extreme polarization. “Democracy in lhe trienio was flawed by the absence of a sense of Lrusl and mutual guarantees amang major social and polma] roups. - . Secure in their AD’s leaders d the need o compmnuse with intense minarities, no matter how small in size.” This first attempl al democracy ended in a coup in 1948. Ten ycars laler, when Lhe mililary dictatorship of Gen. Marcos Pérez Jiménez was dissolving, lhe leaders of AD and other

democratic groups explicilly acted “to reduce inter-party tension and violence, to accentuate inlerests and procedures,

HOW? CHARACTFRISTICS OF DEMOCRATIZATION

173

and remave, insofar as possible, issues of survival and legitimacy from the political scene.” The leaders of the successful democralization in 1958 were in large part those who had led the unsuccessiul democralization of 1945. They had benefitted from (heir expericnce; their 1958 aclions reflected their “recognition of the lessons leamed from the breakdown of democracy during the trienio (1945-1948).” % Similar learning experiences on lhe need for compromise and moderation occurred in third wave transitions. In Spain, for instance, Juan Carlos was reported by one political leader to be concerned with “why the monarchy collapsed in 1932. Te wanls 1o avoid making the same mistakes as his grandiather.” The leaders of the Communist parly were also convinced that the mistakes of the 1930s had to be avoided. *The memory of the past obliges us,” one of them said, “to take these circumstances into account, that is, to fallow a policy of moderation. . . . We cannot allow ourselves Lhe luxury of expressing opinions (hat might be misunderstood, thal could be, or appear lo be, extremist.”” The peaceful transition in Argenlina in 1983 similarly “suggesls that there may have been a tenuous leamning process al work in Argentina in comparison o the transition of 1971- 1973, when many sides of the political spectrum resorted to violence.”* In Peru the military and APRA wenl through a similar learning process. In Poland in 1981 Solidarity moved in a radical direclion, threatening, (he direct overthrow of (he Marxist-Leninisl sys-

tem; the government reacted with a martial law crackdown, oullawing the union and imprisoning ils leaders. Seven years

later, both sides had lcamed their lessons: radicalism Jeads to repression, and repression does nol work. They pursued policies of moderation and compromise in leading Poland loward democracy in 1988 and 1989. Later democratizers not only received a snowballing impetus Lo regime change from those who had done it earlier, they also learned lessons (rom the previous experience of others. The Latin Americans and East Europeans look lo heart the

174

THE THIRD WAVE

lessons in moderation of the Spanish model. In transplacements lhv groups domlndnt in both the government and Lhe have to be lo ise and moderatlon lru\spldcemenls tended 1o occur more frequently in the

later Lhan in the earlier years of the third wave, which sug

gests that the groups involved may have leamned the desir-

ability of compromise from the earlier experiences of others.

In Korea the opposition learned the lessons of peaceful people power from the Philippines, and the government learned the advantages of compromise from the lale of Marcos. The transplacement in Czechoslovakia was in some measure, as Timothy Garton Ash put it, the beneficiary, and what happened there the culmination, of I Central P learning procc ith Poland bemg

the first, but paying the heaviest price. A student

occupalion strike? of course, as in Poland! Nonv‘lolence7 The

First C

of all Central

Europ

Pup-

pet parties coming alive? As in East Germany. "A “tound table” to negoliale the lransition? As in Poland and Ilungary. And so on. Politically, Czechoslovakia had what call Lhe “advantages of backwardness.” They could learn from the others” examples; and [rom their mislakes.* ELECTIONS: STUNNING

AND OTHERWISE

Llections are lhe way democracy operales. In the third wave they were also a way of weakening and ending authoritarian

regimes. They were a vehicle of democralization as well as

lhe goal of democratization. Democratizalion was brought

about by authoritarian rulers who, for one reason or another, ventured 1o hold elections, and by opposition groups who pushed for clections and participated in them. The lesson of the third wave is that elections are not only the life of democracy; they are also Lhe death of dictatorship. When their performance legitimacy declined, authoritarian rulers often came under increasing pressure and had increasing incentives to allempt to renew their legitimacy through

HOW? CIIARACTERISTICS OF DEMOCRATIZATION

175

elections. Rulers sponsored elections believing they would either prolong their regime or their rule or that of associates. The rulers were almost always disappointed. With very few exceptions, the parties or candidates associated with authorilarian regimes lost or did very poorly in the regime-sponsored clections. The results of these elections oflen surprised bolh the leaders of the opposition and the leaders of the government. In the first fifteen years of the third wave this “stunning election” pattern was a pervasive one. |l occurred in alf three types of transilion processes. Consider the following;

(1) As part of his decompression policy General Geisel allowed fairly competilive congressional elections 10 be held in November 1574 in Brazil. The government party, the Alianga Nacional Renovadora (ARENA), expected an easy win against the oprosi(ion party, MDB, and as late as October “few informed political observers would have bel against them.” The election results, however, “stunned everyone, including the most opti-

mislic MDB sirategists.”* The MDB doubled its representation

in the lower house of Congress, almost tripled ils representalion in the Senate, and increased ils control of state legislatures from one o six.

(2) In January 1577 in India, Indira Gandhi, who had been ex-

ercising emergency powers, abruplly called parliamentary elec-

tions (or March. Gandhi was the commanding ligure in Indian

polilics, bul the Janata apposition coalition won an gverwhelming victory. For the first time in history the Congress party lost control of the pational government, winning only 34 perceat of

the popular volc, the firsi time also that it had ever gotten less than 40 percenl.

(3) In Peru’s transition election in May 1980, the military governmenl suppurled the APRA party and passed an electoral law d to enhance its The election, however, pmduced “surprising, results.” “APRA sulfend an electoral collapse, oblaining 27 percent of the vote.” Accién Popular, the opposition party most removed from the military, won “an asionishing victory,” getling 45.5 percent of the vole and winning lhe presidency, a majority in the Assembly, and a pluralily in the Senate.®

(4) In November 1980 the military government of Uruguay

176

THE THIRD WAVE

held a

dum

on a proposed

new

itution

that would

have given the military an instilutionalized permanent veto on Fovemmenl policies. “To the amazement of the army,” the pubic rejected il by a vote of 57 pereent Lo 43 percent. The result “stunned Lhe military as much as the gpposition.”* Two years later the military horized election of delegates to

conventions of the major partics. Opponents of the mililary

swept (he field, with one close associate of the military, former President Jorge Pacheco Areco, getting only 27.8 percent of the vole in his party. (5) The ition military g in Argenlina sponsored national clections in October 1983. The Radical party, led by Rac) Allonsin, long a critic of the military, won a “stunning” victory with an unprecedenled 52 percent of the popular vote.

The candidate of the olher major party, the Perondstas, had “the

open or tacit support of various military sectors” and gol 40 percenl of the vote.” For the first lime in Argenline history the PeTonislas los| a free election. (6) In November 1983 the military government in Turkey sponsored eleclions for a return to civilian rule, and Whe leaders of the governmenl organized and explicitly backed the Nationalist Demuocratic parly, headed by a retired general. “"To the surprise of Turkey’s military rulers,” however, “the wrong man won.” The Nationalist Democrats ran a poor Lhird, getting only 23 percent of the vote, while the opposition Motherland parly swept

into power with 45 percent of the vote. (7) In the February

1985 elections {or the National Assembly

in Korca, the newly formed opposilion party, the New Korea

De

ic

party,

di

p

well,” winning 102 of the

276 scats in the legislature. This followed a campaign that “was tightly lled by the g and opposilion parties

charged that a fair vote was impossible.”

(8) In 1085 the militacy ruler of Pakistan, Genera) Zia-ul-Hugq,

organized parliamentary elections bul initially prohibited pol cal parties from nominaling candidates. The pariics formally boycotled the election. Despite these circumslances, a “large number of candidates who held high positions in the martial

law regime or who were identified as supporicrs of Zia were defealed.” {9) In Chile in October 1988 General Pinocher submilted him-

sel( 10 a yes or no vale on his continued rule. A year before the

1OW? CHARACTERISTICS OF DEMOCRATIZATION

177

referendun, informed opinion held thal he would “win by a landslide.”7 Buoyed by an economic surge, Lhe general himself was confident of victory. As the campaign progressed, however, the opposition mobilized public opinion against him. The electorale lurned down General Pinochet’s bid for eight more years in power by a vote of 55 percent to 43 percent. (10) In March

1989, for the first time in over sevenly

years,

Soviel valers had a chance to vote freely for members of their national legislature. The resull was most surprising, “a mortifying rebuke 1o the high and mighty,” including the defeal of lhe Leningrad parly chief and Politburo members, the party chiefs in Moscow, Kjev, Lvov, and Minsk, several regional party bosses, al least two top military commanders, as well as other leading figures of the communist establishment. = (11) In the June 198g Polish elections Solidarity scored a totally

unanticipaled sweeping victory, winning g9 of 100 seats in the Senate and 160 of 161 seals for which it could compete in the

lower house. Thisty-three of thinly-five high-level government

candidales running unopposed did not get the necessary 50 percent of the votes rcquircg for election. The resulls were described stunmng and reporhedly greeled with “disbeliel” by bolh who were “unprepared”

for them.®

o

(12) The February 1990 victory in Nicaragua of the National

Opposition Union (UNO) under Violeta Chamorro was the most

pervasively stunning election anywhere up to thal date. It was variously described as a “stunning clectoral defeal,” a “stunning upset,” a “stlunning election defeal,” and a ““stunning expression

of Nicaragua’s popular will,” which “stunned many political ana-

lysts” and produced “stunned Sandinistas.”® Despite anticipations thal the Sandinistas benefitting from their control of the government and access to its resources would win easily, Chamorro carried eight oul of ninc administrative regions and got

55.2 percent of the vote to 40.8 percent for Daniel Orlega. (13) Tn May 1990 in Myanmar (formerly Burma) the military

State Law and Order Restoralion Council sponsored the first

multiparty elections in thirly years. The resull was “astounding™

and “a surpl

Tlu: opposition National League for Democ-

racy (NLD) won a “stunning landslide victory” and elected 392

of the 485 contested seats in the National Assembly; (he militarybacked National Unily Party won

10 seats. During Lhe campaign

178

THE THIRD WAVE

the NLD’s leader and tour hundred of its activists were under detention and Lhe party was subjected to a variety of consiraints and harassments.*

(24) In June 1990 in the fiest multi-party elections in Algeria’s

twenly-cight years of

indep

lhe

Islamic Sal-

valion Fron. “scored a stunning success,” which was grested by officials in North Alrica and Europe with “stunned silence. i The

lslarmc Front won

cnnlrol of lhnrry -two provinces and 853

ipal councils. Thep single party, the National Liberation l‘mnl won fourteen provinces and 487 municipal conncils. In all these cases authoritarian rulers sponsored eleclions and losl or did much worse than lhey and others anlicipaled. Were there any exceptions to this pattern, where aulhoritarian rulers sponsored (air eleclions and they or their friends won fairly? One ambiguous case is the Seplember 1980 referendum in Chile in which 68 percent of the voters approved a conslitulion proposed by General Pinochet. The opposilion, however, was severely reslricted, there was no electoral registration, and no meaningful way of monitoring fraud.® In Korea in 1987, the candidate backed by the military government, Roh Tae Woo, was elected president with a pluralily ol 36 percent againsl Lhree olher candidales. The lwo opposilion

candidates who had long campaigned against military rule gol a combined Lotal of 54 percent of the vole. If Lhey had united lhelr rom-s they pmbablv would have won the election, More

i

from the

ing

election pat-

tern occurred in the elections in Romania, Bulgaria, and Mongolia in May, June,

and July 1990. In Romania,

the National

Salvation Fronl took over Lhe government after the downfall of Ceausescu and five months laler won a substanlial viclory

in the May 1990 elections, In Bulgaria the Communist parly, which had governed the country for decades, renamed itself the Socialist party and won control of the Grand National Assembly. In Mongolia the general secrelary and other top leaders of the Communist party were replaced, opposition parties formed, and competitive clections held in which the

HOW? CHARACTERISTICS OF DEMOCRATIZATION

179

communisls won 6u-70 percenl of the seals in parliament. In all cases the leading figures on the winning side had been officials in the communist regimes. How can these app from the g election pattern be explained? Three factors may have been relevanl. Firsl, the new leaders distanced themselves (rom the former authoritarian rulers. Clearly neither Ceausescu nor Todoer Zhivkov could have won fair eleclions in Lheir countries in 1990. lon Iliescu, the leader of the National Salvation Fronl, had been an official iin the Ceausescu regime, but was expelled [rom the Cq parly’s Central C Petar Mladenov and his associates in Bulgaria had themselves thrown Zhivkov, Lhe long-term dictator, out of office; they had been the reformers ousting the standpatters in the Bulgarian transformalion process. A less dramatic change in leadership occurred in Mongolia. Second, cocrcion and fraud in the conduct of the campaigns and Lhe elections may have played a parl. In Romania and Bulgeria inlernational observers were divided as to the extenl lo which the governing groups unfairly allempled to influence the elections: such eflforls were apparently more prevalenl in Rom.uua than in Bulgaria. In both cases, (oreign d certain el of coercion and unfair pm:mes, but their prevalent opinion was that Lthese did not

decisively affect Lhe outcome of the elections.

The third, and mosl important, factor was the nature of

the societies. As has been pointed oul, urban middle-class

groups were the principal forces pushing for democralizalion in third wave countries. In 1980 only 17 percent of the Romanian population lived in cities of over 500,000 people, compared, for instance, 10 37 percent in Hungary. Both Bulgaria

and Romania were still in considerable measurc peasant socielies and both were at lower Jevels of economic development than other Easlern and southern Curopean third wave countries, as well as some st Asian and Lalin American third wave lia was still in iderable mea-

160

THE THIRD WAVE

sure a pastoral society, with three-quarlers of ils population living outside ils only major city and with less than 600 miles of surfaced roads. In all three countries, the opposilion parlies were strong in the cities; the successor parties to the communist regimes won in Lhe countryside, which provided enough voles lo return them to office. Whelher acounlry pm— duced a ing resuit in an authori -Sp election was perhaps ‘one tesl of whelher it had reachied a level of socloeconomic development thal would be supportive of a demacratic regime. The (requency of the stunning election pattern in the lransitions (rom authorilaranism to democracy raises three major queslions. First, why did authoritarian rulers or groups closely associaled with them so consistenlly lose these eleclions? The most obvious and probably most valid answer is that they lost the elections for the same reasons that leaders and parties that have been in power for substantial periods of time lose elections in democratic states. All leaders eventually lose their inilial support and legitimacy. Publics look for an alternalive. In most stunning elections, the vulers clearly were casling protest votes against lhe existing authorilarian rulers. They were probably voling against the existing authoritarian syslem. They may or may not have been casling aflirmative votes in favor of democracy. They could nol, however, vole against the incumbents without also voling against the sys. tem. The defeats of incumbents in a large number of industrialized democracies in the 1g70s and early 1480s did nol destroy democracy; they renewed il. The clecloral defeal of authorilarian rulers, in contrasl, usually meant the effective end of the authoritarian regime. The protest character of the votes in (he stunning clections is reflected in the tenuous nature of opposition unity. [ndivid uals and groups represenling very different political ideologies and wilh very dillerenl grievances againel the regime came together lo vole against the regime. The opposition was

HOW? CHARACTERISTICS OF DEMOCRATIZATION

181

often a motley coalition of a large number of parties with little in common except their opposition to the incumbent rulers. In both Nicaragua and Chile, for inslance, the opposition coalitions were composed of fourleen parlies ranging in views [rom the extreme right to lhe extreme left. In Bulgaria, where the governurient party won, the opposition coalition included sixteen parties and movements. In many eleclions, the leading OPPDGIlIUn parry was 2 new and recen(ly formed party lhat, ils view, could serve as a new and undiscredited vehicle for whalever resentments and (rustrations voters had with the regime. It seems highly unlikely, for instance, thal a majorily of Algerians in 1950 were slaunchly ited to Islamic{1 Voling for the [slamic Salvation Front, however, was Lhe most cffective way of expressing their opposition to the party thal had governed Algeria for three drrudes In addmun, there was the widows and oppusition groups coalesced around the survwmg Temale relatives of martyred national heroes: Corazon Aquino, Benazir Bhutto, Violela Chamorro, Aung San Suu Kyi. These leaders dramatized Lhe issue of good versus Lhe evil of the incumbent regime and provided a magnetic symbol and personality around which all manner of dissidenl groups could rally. Overall. publics rarely missed the opportunily to cast protest voles against long-term authorilarian incumbents. Second, given this pattern of stunning defeals, why did authorilarian rulers sponsor elections they were so likely to

lose? They appeared to be driven by a varicty of faclors, in-

cluding perception of the need lo revive their declining legilimacy at home, the prevalence of democratic norms globally and in their sodiety, and the desire for inlernalional respecl and legitimacy (symbolized by a formal welcome by the president on the While

House

lawn).

In addition,

in most cases,

undoubtedly the risks of having an clection seemed small. Authoritarian regimes normally provide few feedback mechanisms, and dictators nalurally tended to believe (hat they

182

THE THIRD WAVE

had sufficient rapport with their publics to win Lheir endorsement. The aulhoritarian leaders also, of course, controlled the

political organizations were allowed to cmr and substanlial financial resources; hence it was reasonable to assume that lhey could overwhelm what usually appeared lo be a weak, narrowly based, disorganized. and fragmented opposition. Aulhoritarian rulers naturally concluded, “How can [ lose?” In considerable measure, third wave democratizalions moved forward on the false confidence of dictators. The confidence of authoritarian rulers that they could win the elections they sponsored was undoubtedly further bolslered by the extent to which they felt they could manipulale electoral procedures. Three devices were frequently employed. Some leaders attempled (o affect the resulis of elections by aflecting the liming of elections. Both they and Lheir opponents usually thought that earlier elections would benelil the governmenl because it was organized and commanded public atention, whereas later elections would bene[it the opposition by giving it time lo get organized, appeal to the public, and mobilize its supporters. Marcos called a snap elechun with the hupe that the oppasition would be d and in Brazil, the opposition supporled the delay of the munlupnl eleclions scheduled for May 1980 because Lhey feared they would not be ready for them.* In the Polish Round Table negoliations the government pushed

for early elections, and Solidarity was viewed as making a

major concession when it agreed to them. {n Hungary the government wanled an early popular election for president because it assumed its probable candidale, Lmre Pozsgay, had high visibility and a very good chance of winning. The opposition feared that would happen and pushed through a proposal for a referendum on the issue, in which the public Jecting lhe [irst presid In Czechoslovakia simitar concerns were expressed aboul the advantage

HOW? CHARACTERISTICS OF DEMOCRATIZATION

183

that early elections would give the communists; and in Romania in February 1990, opposilion leaders said (hey wanted 1o delay the elections scheduled for May because they had inadequale resources and lime o prepare for them. The logic of Lhese government and opposilion posilions on election timing seems obvious: upposilion groups would benefil from more time to prepare for the elcction. Liltle empirical evidence exisls, however, to support this logic. Back in Turkey’s second wave democratization, for example, the government moved up eleclions to July 1946 in order “to catch Lhe new [opposition] party before it could (ully organize,”% but that parly did extremely well in those elections. In the Korean elections in February 1985, the opposilion New Korea Democratic Parly, formed a (ew weeks before the election, won 2g percent of the vote and 67 of 184 elecled seats in the National Assembly.* DBy all impartial accounts, Ferdinand Marcos lost his snap election, and Solidarity scored a sweeping victory in the eatly elections lo which il had agreed very reluctantly. The evidence is not conclusive, bul it clearly does nol support the proposition that governments always benefitted from calling early elections and that oppositions suffered by parlicipating in them. Second, authoritarian rulers often rigged elections by establishing electoral systems hlghly favorable to the governmenl, by ing and intm Lhe opposition, and by in the paign. If carried to an exlrcme of course, these tactics ensured the government’s victory but made a mockery of the election. In mosl of the slunming elections listed above, however, groups in power made efforts, at times extensive ones, (o tilt the clection in their favor, and yet did not succeed. Over the course of a decade, from 1974 to 1984, lhe Brazilian government regularly revised its laws on elections, parties, and campaigning in hopes of stopping lhe steady growth of opposilion power. It did not succeed. Again the evidence is fragmentary,

184

THE THIRD WAVE

bul what there is does suggesl thal, unless lhey were carried to an extreme, rigging tactics were unlikely lo ensure governmenl viclory.

It mampulalmn of elechon hmmg and procndures did not suffice, the al itarian rulers was oulrighl (raud and lheft Aukhomanan rulers can steal elections if they want to. Often in the past they were able to steal eleclions quielly, in nonobvious ways, so thal although everyone knew Lhe elections were stolen, no one could prove it. In the July 1978 elections in Bolivia, for inslance, General

Banzer engaged in “massive [raud”

in arranging for his can-

didate, Gen. Pereda Asbun, {v receive exactly the required 50 percent of the vote.” As the third wave wenl on, however, democratization became a recognized phenomenon of global polilics, the media paid more auenlnon lo n and elections were subject 10 i By lhe late lgfim foreign observers had become afamlha.r and p in almost all i In some instances such missions were sponsored by lhe United Nations, the Organizalion of American Stales, or olher intergovernmental bodies. In olher cases private organizalions provided (his servlce By 1990 the National Democralic Institute for Affairs had urganized international observer missions for Lhird wave eleclions in some thirteen counlries. Delegations from the U.S. Congress and other legislatures were also present in some cases. Former

President Jimmy Carter played an active role in and added

his authority to several of these missions. Foreign observers made it difficult if not impossible for govemments quielly and secretly to steal an election. Blatant theft, as in the Philippines and Panama, however, defeated lhe purpose of havmg the elechon which was w enlance the and i y of the rulers. If, on the other hand, the government refused to allow “impartial” external observers to witness Lhe vole, that in itsell now became proof that it was rigging the election. The emergence and

HOW? CHARACTERISTICS OF DEMOCRATIZATION

185

prevalence of the foreign observer phenomenon was a major development of the 1980s and significantly enhanced the importance of elections in the democratization process. Authoritarian rulers who sponsored elections to bolster their declining legitimacy were in a no-win position. If they played the game fairly, they suffered a “stunning” defeat. If they i timing and d in other than extreme fashion, they probably also were defeaied. If they slole the election, Lhey lost legitimacy rather than gaining it. Thy reasons Lhat led them 1o hold elections—decreasing legitimacy and opposition pressure—were also the reasons lhey Jost Lhose elections. The insoluble dilermma they faced was neally summed up by Gen. Fernando Malthei, commander of the Chilean air force, shorlly before the 1988 election: “If Lhe government’s candidate wins everyone will say it was a fraud. I( he loses everyone will say il was a fair election. So it is more in our interests (han anyone else’s to be able to show it was an absolutely (air election.”* In 1990 (he Sandinistas, similarly confident of victory, felt lhe need to have a fair election and invited masses of foreign observers who would be able

1o testify 1o both its faimess and their victory. The resulls

of both clections underscored Malthei’s point. Aulhoritarian rulers could only legitimate their regime through elections by ending lheir regime through elections.

Third, eleclions sponsored by authoritarian regimes also posed problems for opposition groups. Should they participate in the election or should they boycoll it? Given the pat-

tern of stunning defeats of authoritarian rulers, what. if any. was the rationale for the opposition not taking ad ge of the opportunity offered by Lhe aulhoritarian-sponsored election? These issues did not normally arise when a regime transition was clearly und y: |! d were in power and moving deci forward in a process, if the leaders of a mililary governmenl explicitly said that they are returning to the barracks, if leaders of government and opposilion were agreed on a transplacemenl pro-

186

THE THIRD WAVE

cess, In Lheseci the principal opposition groups normally had no reason nol to participate. At the other exlreme opposition democrats had LitUe Lo gain by accepting appoi in auth ian governments and thus lending legulm\a:y lo those governments. I{ they did, they would divorce themselves [rom their constituencies and make themselves dependent on the authoritarian rulers. The Jaruzelski government in Poland and the Bolha government in South Africa, (or msh‘mce, both attempted lo lure opposition leaders to positi consullative councils. From the wewpoml of pmmoung democracy, Polish and Soulh African opposilion leaders correclly refused to pamupatc Oppusmon demorrals also generally did not in for thal were powerless and 5|mply govemmznt tools, In 1973 lor instance, Pap pted (o bulster his[ re gime by pmrmsmg parliamentary elections. The leaders of the Greek political parties refused to participate. George Mavors, head of the oullawed Center Union party, said il well: the promised “eleclions have a single purpose: lo legilimize the dictatorship covering il by a castrated Parliament which will not have the power to debale, let alone decide, any of the ration’s vital malters.”> In between

those extremes,

lhe boycoll

issue mosl

often

arose when elections werc called either by a standpatier re-

gime or by a liberalizing regime whose ullimale intentions with respect to democratization were unclear. Philippine op-

posilion leaders, for instance, hotly debated and disagreed on the boycotl issue with respect to lhe Marcos-sponsored 1978 and 1984 National Assembly elections and the 1986 presidenlial election. Mosl but not all black South Alrican polilical leaders urged boycolts of the 1983 and 1988 municipal clections; coloured and Asian political leaders were divided about parlicipalion in the 1984 and 198y parliamentary elections. Three out of four opposition parties urged boycott of (he 1974 Dominican Republic presidential elections when it appeared

HOW? CHARACTBRISTICS OF DEMOCRATIZATION

187

that the Belaguer g had little or no i ion of being willing lo surrender power. At U.S. government urging, opposilion groups boycotted the 1984 Nicaraguan elections. Pakistani party leaders urged a boycott of the 1985 National Assembly eleclions orgamzed by lhe Zia regime as il began to li The two princir ion parties (the Movemenl for Democracy iin Algeria and the Socialist Forces Front) but nol the Islamic Salvalion Fronl urged abslention {rom the 1990 local and provincial elections in Algeria. or liberali d elec lions to enhance their reglmc Icglhmacy and/or to extend their personal tenure in power. Parlicipation in the eleclions by al least some opposition groups lhus was essential lo Lhe governments, Marcos, for inslance, welcomed Benigno Aquino's decision to run for the Assembly in 1978 while in prison under a death sentence “because it legilimized the exercise.”

Hence governments often tried to defleat boycolt efforts. In

the Pakislan assembly election of 1985, “propaganda against and instigation of the boycolt of the polls was declared a cognizable offense” and “newspapers were ordered in February 1985 not to publish any stalement in favor of Lhe boycoll campaign.” In the 1988 municipal elections, the South African governmen! similarly “clamped down on pro-boycott opposilion groups and made it unlaw(ul for individuals lo urge a boycotl.”* Boycoll campaigns changed the vlection issue from whom

to vole for Lo whelther to vote. The success of these efforts

varied, depending on lhe unity of the opposition groups in support of the boycott, public perception of the intentions of the government, and the public’s previous experience with voting. Most South African blacks had never voted previously, so it is nol surprising thal only 20 percent of potential black voters voted in Ihe municipal elections in 1983 and about 30 percenl in 1988. Thirty percent of coloured voters and zo percent of Indian volers parlicipaled in the 1984 parliamentary election in South Africa and voling in 1989 was

188

THE THIRD WAVE

comparably low. In the Dominican Republic in 1974 lhe abstention rale was again about 70 perceni. Some boycoll efforts were less successful. The principal oppusilion groups urged Spanish voters to boycoll Lhe referendum on political reform in December 1976, but 77 percent of the voters wenl lo the polls. This election, howcvcr was d by a g cearly dto reform. In the May 1984 malional assembly elections in Lhe Philippines, 8o percent of the electorate voled despite the appeals of leftisl opposilion groups that they not do so. In the 1985 Pakistani National Assembly elections boycotted by the political parties, many opposition candidales were clected, and the parlies subsequenlly admitted they had made a mistake in urging abstention. About 40 percent of the polential voters abslained from the 1990 Algerian local and provincial elections. " Whal then were Lhe efficacy and wnsdom of boycolls as a strategy for the d PP boycott did nol end Lhe auLhoritarian regime or remove the gavernment from power. It did reduce ils leg.lm\ary. which was one reason whyg $0 ly against boycott efforts. Unsuccessful boyu)lks. on the other hand, were evidence of

opp

. More

imp-

, boy-

cotls olten meant opportunities foregone, choosing nonproduclive exit instead of meaninglul voice. l’arlncxpahon in an election paign itsell often ished an , depending on government restrictions, Lo rilicize the govem— menl, mobilize and organize opposilion supporters, and appeal lo the public. [t was an incentive lo polilical activity, and authoritarian regimes are only changed or brought down by political activity. If the eleclion was conducted with minimal fairness, (he opposition often did dinarily well. In the best of ci il won a ing” victory and toppled the govemmenl. Short of this, opposition candidates made elfective in autf elecPaig!i P

HOW? CHARACTERISTICS OF DEMOCRATIZATION

189

tions in Brazil, Taiwan, Mexico, the Philippines, Pakislan, and the Soviet Union. Lven modest opposition electoral successes could be capilalized upon to weaken the government. Peping Cojuangco argued hat the Philippine opposition should contest the 1984 National Assembly elections, although they clearly would not win a majority, because "If you get thirty [National Assembly seals] Lhis time, Lhen people will believe it can be done. and in Lhe next election that thirly will multiply.” [n Brazil in the 19705 the government restricted opposilion electoral campaigning and changed electoral rules lo hamper the opposi tion. The opposilion party, ihe MDB, however, contesled every cleclion as best it could, gradually expanded its strength in and control of legislalive bodies, used these posilions to pressure the government to move forward in democratization, and, as a result, increasingly came lo be viewed as a ponsible alternalive gov: - Al the same lime its activity strengthened the hand of the democralic reformers in the government in dealing with the strong standpatter resistance wilhin the military.

In South Africa boycolt proponents greatly reduced partici-

pation in the eleclions to the coloured and Indian houses of

Parliament in 1994. At least some of those who were elected,

however, made effeclive use of their positions in the campaign against apartheid. The 1985 session of Parliament re-

pealed the laws prohibiling inlerracial marriage and sex and

prevenling (ormation of mulliracial political parties. It also I d ictions on black resid and emp in urban areas. Coloured and Indian members of Parliament gave “impetus” to these changes. Allan Hendrickse, leader of the coloured Labour party, “forced Mr. Botha’s hand against the law which prevents the formalion of multiracial parties, by pulting up Coloured candidates for the Indian chambers.

The Labour party may also have helped Lo relax the ‘pass

laws,’” which control the movement of non-whites, and en-

190

THE TILIRD WAVE

couraged one black ity lo resist the plans to displace il from its ancestral home.”#* Subsequently, Hendrickse used his control of lhe coloured house of Parliament lo demand that President Botha repeal the Group Arcas Act in return for [endrickse agrecing to amend the conslitulion so as to postpone parliamentary elections from 1989 to 1992. Bolha refused to agree to the deal, and the elections were held in 1989. In South Africa and elsewhere, elected representatives found ways o lobby the government and lo bargain with it in the inleresls of democralic reforms. ‘Those most prone lo boycolt elections were radical oppositlion groups opposed to democracy. Marxisl-Leninisl insurgenls in the Philippines and El Salvador spurned elections.

Opposition groups, such as the Palestine Liberation Organization (PLO) and Lhe ANC in the 1980s, which refused lo

participale in elections, however limited or unfair, were likely 1o be dominaled by leaders not adverse to replacing one undemocratic regime with another. In the 1984 Philippine elections (he communist-dominaled National Democratic Front led the boycott campaign against Corazon Aquino and Lhe other democratic candidates who participated in the election against lhe Marcos regime. They also intensified their use of violence during the campaign. “The opposilion running in these elections,” one Comununist parly leader said, “are only PP isls. The real opposition are the boycotters.”< e

was right. His “real opposition” opposed both the Marcos regime and democracy. His political opporlunisls were using ballots rather than bullets to revive Philippine democracy. The lesson of the third wave seems clear: authoritarian leaders who wanled to stay in power should not have called

elections; opposition groups who wanted democracy should

not have boycolled the elections aulhoritarian leaders did call.

In the course of the transitions to democracy, eleclions also tended to promote political moderation. They provided incentives to move loward the center for both opposition par-

HOW? CHARACTERISTICS OF DEMOCRATIZATION

191

lies (hal wished lo gain power and government partics thal wished 10 retain power. In their first election in April 1975 Portuguese volers decisively rcjecled radical Marxist alternatives and gave their support to moderate centrist parties. Two years later Spanish volers acted similarly in their first election, which was appropriately described as “a triumph for moderalion and for the desire for change.”* Volers in Greece, El Salvador, Peru, the Philippines, and elsewhere similarly gave only minuscule elecioral support to leftist revolutionaries, and, of course, Nicaraguan voters ousted the leftists [rom power. With only a few exceplions, publics consistently rejecled the old authoritarian regimes and those who had been associated with them and also rejected extremist alternatives to lhose regimes. “No dictators, no revolutionaries,” was, in effecl, the maxim of voters in third wave transition elections. Elections were one road away from authoritarianism. Revolution was the other. Revolulionaries rejected elections, “We shall nol allow puppet organizalions to put up candidales,” the head of the ANC military wing said in connection with the 1988 municipal elections in Soulth Africa. “We shall use revolutionary violence to prevent blacks [rom collaborating.” The contrast between the two paths was well summed up by Alvaro Cunhal, head of the Portuguese Communist party, in 1976: In lhe Portug lution lwo p have intervened, lwo d! i p difierent ch. with tics. On one side, the revolulionary dynamic, created by lhe inlervention of material force—popular and military—direcily and g liberties,

defealing and throwing out the fascists, opposing the counter-

revolulionary attempts, bringing aboul profound social and § o create a stale in service of (he Revolulion, and the crealion of organs ol power

(including military organs) which will guaraniee the demo-

cralic process and correspond 1o the revolulionary transformalion.

192

THE THIRD WAVE

On the other side, the elecloral process, underslood as the Lhulce by universal suffrage of the organs of power, tending to any social (o a previous constilutional legality, and nol recognizing the intervention of the military in political life, or the creative, predominant intervention of lhe masses in the revolutionary process.® In lhe third wave the "elecloral dynamic” led from authorito d the * i y dynamic” led Irom one form of authorilarianism to another. LOW LEVELS OF VIOLENCE

Major political changes alimost always involve violence. The third wave was no P Almost every democrati between 1974 and 1990 involved some violence, yel the over all level of violence was not high. Taking place as they did through compromise and elections, most third wave democwere rel ly peaceful pared to other regime transilions. Polilical violence involves people doing physical injury to persons or properly in order to affect the composilion or behavior of government. An imperfect but widely used measure of political violence is the number of deaths (hat occur for political reasons during a particular time or in connection with a particular event. Estimating with even very rough accuracy the number of polmcsl deaths in the third wave is ex-

d

ization

difficult. p violence Lhat is part of has to be distinguished from violence that

may occur during democratization, such as the routine murder by gt of d (which is an inherent characteristic of many authoritarian regimes), and from ethnic condlict, which may be a product of liberalizalion and demecratizalion. In a minority of countries demncratization efforts were accompanied by major violence. The mosl extensive violence occurred where Lhere was suslained armed conflict between and op guerrilla over an B

HOW? CHARACTERISTICS OF DEMOCRATIZATION

193

extended period of time. In Guatemala, El Salvador, the Philippines, and Peru, Marxist-Leninists waged insurgent wars against aulhonlamn governments; these regimes were e placed by ically elected go e cies, however, continued. In at least Guatemala and El Salvador the political deaths resulling {rom the insurgencies against the authoritarian regimes were considerable. Estimates of those killed in Guatemala hetween 1978 and the clec tion of Vinicio Cereza in 1985 range from 40,000 0 100,000. Political killings in El Salvador between the 1978 reform coup and the coming, into office of Duarle in 1984 have been estimated at 30,000 to 45,000. These killings were the result of the often brutal and indiscriminate application of violence by the securily lorces defending authoritarian governments in wars against insurgenl movements attempting to overlhrow those governments and establish Mands|-Leninist regimes. The killings were a product not of demacralizalion efforts by either government or opposilion but of a war between two nondemacratic groups. In Nicaragua 23,000 people are estimated to have been killed in the civil war between 1981 and 1990. Whether or not a contra military victory would have produced a democratic governmenl in Nicaragua is uncertain. The confra insurgency, however, was one of the many [actors that induced the Sandinista regime to call elections. After a democralic regime came into office, the contras, unlike the MarxistLeninists in Gualemala and E) Salvador, ended their insur-

gency and disbanded. The casuallles in lhe N|c:uaguan civil

war, unlike those in El Sal and Peru, thus may be appropnnmly vnewad as P.lrl of the costs of democratization. Nicaragua stands alone among lhird wave countries in the number of people killed in the siruggle 1o democratize. For the period 1974-90, South Africa probably ranked second. In South Africa 575 people were killed in the Soweto massacre in 1976; some 207 were killed by government forces, the ANC, and others belween 1977 and 1984; aboul 3,500 are es-

194

TIIE THIRD WAVE

limaled to have been killed in the uprisings in the black townships from 1964 0 1988, and from 3,500 to 5,000 more in the fighting belween black groups from 1985 to 1990. Overall perhaps 9,500 to 10,000 South Africans died in political violence between 1976 and 1990, Individual incidents or actions also produced significant numbers of deaths in some countries. L.S. invasions resulied in the killing of some 120 lo 150 people in Grenada. and al least 550 and possibly as many as Boo in Panama. The Korean army killed at least 200 and possibly up to 1,000 people in the Kwangju incident of May 1980. Probably 1,000 and perhaps

as many as 3,000 people were killed by the Burmese mili-

lary when they supp d the d in August and September 1960. Somewhere belween 400 and 1,000 people were killed in Deijing in the Chinese government’s crackdown in June 1989. At leasl 746 people were killed in

political violence in Bucharest in December 1589 and probably

several hundred others were killed in Timisoara and elsewhere. Apparently more than 200 people were killed in Lhe coups d’etal in Bolivia in 1979 and 1980.% In the great majority of third wave countries, however, the averall levels of violence were quile low. Such was clearly the case in Lhe first transitions in southern Curope. In the coup that began Lhe third wave in Portugal, for instance, 5 people were killed and 15 wounded. In the year thal followed less than a dozen people died in political violence. Several more

people were apparently killed in peasant anticommunisl up-

heavals in northern Porlugal in the summer of 1975. One soldier was killed in the aborlive righl-wing coup of March 11, 1975. and 3 more in the November 25, 1975, coup and countercoup.®® In a year and a half of revolutionary upheaval in Portugal, however, total political deaths probably numbered less than 100. The Spanish transilion also was relatively nonviolent. During Lhe four years {rom 1975 through 1978, 205 political deaths were reported: 13 by extreme right-wing groups, 23 by Marxist-Lenirist groups, 62 by Lhe police

HOW? CHARACTERISTICS OF DEMOCRATIZATION

195

and Civil Guard, and 107 by the ETA Dasque leftist separa-

tists.** Apart from Lhe army killing 34 people in the Poly

technic incident, the Greek transition was relatively free of violence. The transitions {rom military regimes in the South Ameri can countries, with the partial exception of Chile, were also

generally peaceful. Virhaally no blood was shed during the

regime changes in Poland, Easl Germany, Hungary, and Czechoslovakia. These transitions were, as Timothy Garton Ash put it, “remarkable far iheir almosl complete lack of violence. . . . No baslilles were stormed, no guillotines erecled. Lamp paosts were used only for streel lighting.”* In Taiwan the most dramatic evenl in the struggle over democratization was the so-called “Kaohsiung incident,” in which no one was killed but 183 unarmed policemen were injured. In the Philippines, a country with what is generally considered a violence-prone culture, both the Marcos regime and the Marxist insurgents killed people, bul the numbers were limited and the principal opposition groups did not employ violence. In Korea, after the Kwangju incident, violence was restrained and very few political deaths occurred. The return lo democracy in India and Turkey, after their brie( experience with authorilarian rule, also involved minimal violence, as did the transilion in Nigeria. Violence obviously occurred in the third wave, bul overall

it was quite limited. The tolal political deaths in more than thirty democratization efforts apart from Nicaragua belween

1974 and 1990 probably numbered about 20,000 and were heavily concentrated in South Africa and mainland Asia. IL was, obviously, tragic that Lhese people were killed. Campared, however, to the hundreds of thousands of people killed in individual communal conflicts, civil wars, and international wars, and considering the positive results achieved in termg of political change, the cost in human Iwe; of the third wave was inarily low. Democrati accounted for an infinitesimal poriion of the political deaths in

196

THE THIRD WAVE

the world beteen 1974 and 1990. From the “revolution of the carnations” in Lisbon lo the “velvet revolution” in Prague,

the third wave was overwhelmingly a peaceful wave.

What explains the low levels of violence in these regime changes? First, the experience of some countries wilh substantial civil violence before democratization began or early in the process of democratizalion encouraged both government and opposition to abjure violence. Spain and Greece suffered bloody and divisive civil wars before and alter World War II. Military governments in Bmzil, Uruguay, and Argenlina fought ruthless “dirty wars” against terrorist groups in the 19605 and 1970s. One effect of these wars was to reduce drasticallyor to eli radical opposition of the Marighella type, commitled exclusively to violence.” A secand eflect was 1o produce in all seclors of society lhe reaction of “munca mds.” In December 1982, for inslance, Argenlina's transition was marred when some participants in an authorized protest charged police barriers in front of Ihe presidential palace; the police responded with tear gas, and one person was shot by a governmenl security agent. President Bignone and opposilion party leaders reacted promptly to prevent (urther violence and the Church declared a “‘Day of Natiunal Reconciliation.” Afier this, the Argenlme transilion inued to be ly peaceful comp to lhose in

many other countries.”** Somewhat similarly, in the Korean prolests of 1086 and 1987, the police were careful (o avoid

using firearins so as nol (o replicale (he Kwangju massacre. In Taiwan both opposition and government lactics were simi-

larly heavily affected by memories of the Kavhsiung incident.

In December 1986, for mstame, the principal leaders of the Taiwan opp d those of a crowd who had stoned police cars at the airport, announced that “security should be firsl and freedom second.” and cancelled plans for twenly rallies. In Leipzig in Fast Germany in Oclo-

HOW? CHARACTERISTICS OF DEMOCRATIZATION

197

ber 1989, both ¢ i ities and opp recognized the need Lo avoid “another Beiji Second, different levels of vialence were, in some measure, associated with different processes of transition. About half the transilions belween 1974 and 1990 were lransformations in which d in the g were strong, enough lo initiale and in considerable measure (o control (he process of regime change. The government thus had little incentive Lo resorl to violence, and the opposition had little opportunily to do so. One nolable exceplion was Chile, where the government adhered toa rigid schedule [or regime transand the opp used massive d to attempt to accelerate change and to I‘mce the government to In dominant in the govemmcnt and democratic moderales dominant in the opposition had a common interest in minimizing violence as they struggled to agree on the terms for transilion. Among replacements, the record was more varied, and the two military interventions caused significant bloodshed in two small countries. The prevnlcnce of lransformahons and, toa lesser degree, however, violence in the third wave. Third, the willi of dp to order the use of violence againsl opposition gruups varied considerably, as did the willingness of the security forces lo carry out such orders. [n China, Hurma, South Africa, and Chile,

tough-minded leaders endorsed the use of force, and police

and mililary unils ruthlessly employed vm]ence to suppress peaceful and not-so-pe: In other cases, however, government leaders did not act decisively and seemed reluctant 1o direct the use of force against their own cilizens. Like Lthe shah of Iran, Marcas walfled back and forth in his instructions to his mililary as opposilion protests mounted after the February 1986 election. In Poland, East Germany, and Czechuslovakia, the communist govern-

198

THE THIRD WAVE

ments for many years did not hesitate to use force lo suppress

opposition. At the crucial moments of regime transition in

1988 and 1yBy, however, lhey refrained (rom doing so. In Leipzig on October g, 1989, the situalion was apparently one of touch and go: a massive opposition demonstralion was planned and “riot police, state security forces, and members of the paramililary faclory ‘combat groups’ stood ready lo clear East Germany's Tienanmen Square with truncheons and, it was subsequently reporied, live ammunition.” The order (o use force againsi lhe 70,000 demonstrators, however, never came. This was apparently the result of action by the local civic and party leaders with the belated endorsement of the national Communist leader, Egon Krenz. Overall, in Easl Curope, aparl [rom Romania, there was, as Garlon Ash noled, an amazing absence of “major connierrevolutionary violence.”* In both Lhe Philippines and East Europe, probably the principal reason behind the reluctance of government leaders 1o use force at the crisis moments in the democratizalion process was the expressed opposition la such use by theg of the relevant sup . [n contrast, the influence of restraining superpnwers was lotally absent in China, Burma, Romania, and South Alfrica, and was very weak in Chile. When governmenls do anthorize the use of violence, the violence only becomes real when those orders are obeyed. Ultima ratio regum is not guns bul the wi]ling-nes: of those

who have guns 1o use them on behall of the regime. That

also varied iderably. Armies g y do nolL like to train their weapons on the citizens whom they have a duty Lo defend. Police and inlernal security troops were normally more willing than regular military units Lo use violence to quell disorder and protest. Authoritarian rulers often created special securily forces -the Securilate in Romania, “dignity battalions” in Panama, interior ministry lroops in many countries—especially recruiled and trained to support the regime.

HOW? CIIARACTERISTICS OF DEMOCRATIZATION

199

Soldiers and police were less likely to obey orders 10 use violence if they could identify with the people they were ordered to shoot. Authoritarian regimes, ¢ q y, attempled to ensure lhat there were social, ethnic, or racial differences between the users and largets of regime violence. The South African government regularly assigned black policemen to tribal areas differen! from their own. The Soviel government attempled to pursue a similar policy with respect to Lhat country’s nationalities. The Chinese government used Ppeasant armies {rom distant provinces against the students in Tiananmen Square. The more homogeneous a sociely was, the more difficult it was for the regime to use violence to suppress npposmon The same prmc:ple also pnol.ubly explains hi duced by U.S. forces In a similar vein, the more broad-gauged and representa-

tive of mainstream citizenry a mass demonsiration was, the

more reluctant military and police units were lo use violence against il. In September 1984 the Manila police used “guns, clubs, and tear gas” to break up an anligovernment demonstration of 3,000 people, composed largely of students and leftists, with 34 people (12 with sholgun wounds) being hospitalized.* The following month the police did not interfere with a demonsLration of 30,000 Feople organized by business groups and Cardinal Sin. In the confronlation oulside Camp Cracme in February 1986, Filipino army units clearly did not wish to fire on the massed religious workers, professionals, and housewives. Similarly, Korean security forces were much more willing to use force against radical student demonstralors Ihan against mndd]uclass clerical workers, technicians, and b In Czechoslovakia the defense minister assured leaders o[ the Civic Forum that the Czech army would not shool Czech cilizens. In Romania, army unils refused to fire on demonsiralors in Timisoara; and the army then lurned against (he regime and played a crucial role in

the

jtate,

which

ined loyal to the

200

THE THIRD WAVE

Ceausescus. Even in China, some army units apparenlly refused to fire on civilians, and their officers were subsequently investigaled and court-martialed.*” The use of force against the opposition was thus likely to be more effective if (1) the society was socially or communally helerogeneuus, or (2) il was at a relalively low level of socioian regimes in socielies in

which economic development had produced a substantial

middle class sympathelic lo democratizalion were more reluctant 1o order the use of force to suppress dissent, and lhe securily troops of those regimes were less likely lo carry out such orders. Fourlh, oppesilion groups also varied widely in the cxtent to which they used, tolerated, or rejected the use of violence. This issue posed in a much more crilical and dramatic way some of Lhe same questions that came up in the debates over whether or nol to boycott regime-sponsored elections. In many cases, the police and security forces of the regime imprisoned hundreds or thousands of people and tortured or murdered them. The activities of the official forces were often supplemenied by those of the semiofficial, paramilitary “death squads.” Under these circumstances the templalion was strong to dramalize and further the opposilion cause by blowing up governmenl installations, throwing Molotov cockrails at govemmenl vehxcles, shooting at soldiers and police, kidnapp iously brutal officials. To

whal extent should a democratic opposition resort 1o these

tactics against a repressive regime that uses violence? The answers of opposition groups varied across the spectrum from never to sumetimes to always. The answers corrclaled highly with the overall commilment of the groups lo democra Democratic moderates rejected violence; radical groups advocated violence. In most third wave ies the principal opposition groups pursued democracy lhrough nonviolent

HOW? CHARACTERISTICS OF DEMOCRATIZATION

201

means. The Catholic Churd\ as we have seen, was a major force pushing for democ in many and the Pope, local bishops, and most clergy also vigorously espoused nonviolence.* Middle-class urban businesspeople, prolessionals, and clerical workers, who often dominated the democralic opposition, usually rejected violence and atlemplcd Lo minimize violence. Political party leaders had a stake in employing lhe melhods in which they were presumably skilled ise, elections—and in avoiding the tactics of lerronsm and insurgency in which others would excel. The social sources of the moderate opposition movements thus shaped not only Lheir support for democracy bul also their support for nonviolent means of realizing democracy. The extent to which Lhe opposition was commitled to nonviolence varied among countries.

“In a revolution,” Benigno

Aquino said in Lhe speech he drafted for delivery at Manila Airpori, “there can really be no victors, only victims. We do not have to destroy in order (o build.”* In he years following his murder, Corazon Aquino firmly maintained this commilment to nonviolence which, of course, culminated in the massive demonstration of people power by busmesspeuple, students,

and

nuns “that brought

down

the Marcos

regime

in

February 1986. In Eastem Europe, Sohdamy J’mm the slart opposed y tactics and supp lence. As one Solidarity leader put it during the organizalion’s underground phase, when the temptalion to use violence was presumably strongest, Solidarity was “againsl any acts of violence, street batlles, hit squads, acts of terror, armed organizalio nd we do not accept any responsibility for violent acts.”

“We

know

lots of revolulions,

great revolutlons,

and magnificent people,” said Walesa, “who after taking over power, produced systems Lhal were much worse than Lhe ones they destroyed.” Those who start by storming bastilles, Adam Michnik similarly warned, end up building their own.®

202

THE THIRD WAVE

Solid ided the model for iol ilion movements lhat brought about the transitions in East Germany and Czechoslovakia. In South Alrica, the African Nalional Congress adhered lo a policy of nonviolence for almost a hal( century until the Sharpeville massacre in 1960, after which the ANC changed ils policy, espoused the use of violence, and created its own military organization, Umkhonto we Sizwe. Other black leaders, such as Archbishop Desmond Tulu and Chief Mengosuthu Buthelezi, continued o adhere to nonviolence. “Bloody revolutions fought against terrible oppression,” Buthelezi warned, “do not automatically bring aboul great improvements.”* In Korea, also, the principal opposition groups rejected the use of violence, although their demonstrations iin the mid-1980s were often accompanied by violence p by radical sludents. In many countries some opposition groups were, of course, firmly committed to the use of violence against the nondemocratic regimes they confronted. These included MarxistLeninist and Maoist opposilion groups in Cl Salvador, the Philippines, Peru, and Gualemala thai loughi against authoritarian regimes and their democratic successors. The Communist party and jated leflist i L tions in Chile engaged in sustained violence agamsl the Pinochet regime, and {rom 1960 to 1990 the Alrican National

Congress employed violence against Lhe South African regime. Opposition groups used violence against three Lypes of targets: (1) government officials (political leaders, police officers,

soldiers) and installations (police stalions, electrical pylons,

depots, ilities); (2) “colat ” hat is, individuals who i PP the opposition or belonged (o social or radical groups supporting the opposition but who also allegedly acted as informers, agents, or officials for the nondemocralic

regime;

(3) random

civilian

fa-

cilities, such as slores, shopping cenlers, and thealers, which were altacked presumably with a view to demonstrating the

HOW? CHARACTERISTICS OF DEMOCRATIZATION

203

strength of Lhe opposilion and the inability of the government to provide security. Much debate went on within opposition groups as lo the relative merils of these different lypes of targels and particularly over the moralily and eiectiveness of random terrorist atlacks againsl ci ms. In addition, opposition groups commilted to violence debated the relative merits of rural and urban guerrilla warfare and the desirability and tuning of major offensives and popular uprisings. In the 19705 and 1980s, leaders of the African National Congress repeatedly affirmed the role of violence as a tactic in their struggle against apartheid. “Violence,” as ANC leader Thabo Mbeki said in 1987, “is a very important element lo achieve change.”® Inilially the ANC focused on government installations: police stations, electrical pylons, power plants, and olher facililies. Between October 1976 and December 1984, the ANC reporledly carried out 262 armed altacks largelv agamsl su(h targels. In the three years following the starl 4 of ined unrest in black hips, the number of allacks reportedly quadrupled, and blacks associated wilh Lhe regime increasingly became largets. After police fired on demonstrators in Sharpeville in Seplember 1964, mobs killed six black councilors, including the deputy mayor

of Sharpeville. In the following years blacks killed hundreds

of other blacks suspected to be collaborators. During the nine months afler September 1984, black radieals attacked 120 black councilors, killed five, and bumed down the homes of seventy-five. In July 1985 only two of thirty-cight councils were reported to be still operaling. Finally in late 1985 and 1986 a major increase occurred in the third type of attack, with more assaulis against “solt” civilian targets occurring in the first six months of 1986 than in the previous three years combined. Statements by ANC leaders stupgested that serious differences exisled among them as lo the wisdom of these bombings.% In Chile the radical opposition groups concentrated primarily on government installations and officials. During the

204

THE THIRD WAVE

first three months of 1984, for instance, reportedly lhere were eighty bombings of rail lines, utilities, and radio siations, wilh Lhe central area of the country being blacked out three limes as a result. On October 29, 1984, lwelve bombings in five cities damaged governmen! offices, banks, and telephone centers. All in all, apparently over 400 lerrorist attacks occurred in Chile during 1984, and some 1,000 during a twelve-

month period in 1985-86.% Chilean opposilion violence culminaled in the September 1986 effort by the Manuel Rodriguez Patriotic Front (MRPT)

to assassinate Pinochel, in which

the general escaped serious harm though five of his bodyguards were killed. In virtually all countries 2 central tactic of the opposition was the mass rally, march, or demonstration against the regime. Such demonstrations mobilize and focus discontent, enable the oppusition to test the breadth of its support and the effectiveness of its organization, gencrale publicity that is often international in scope, enhance divisions wilhin (he regime on lhe appropriate response, and, i( the regime responds violenlly, can creale martyrs and new causes for outrage. Opposition mass prolests were lypically called on four types of occasions: (1) In sume instances, opposition groups organized demonstrations on a recurring basis. In Chile in 198384, for instance, the opposition held monthly protest demonstrations thal involved significant vialence by both police and proteslers. In Leipzig in 1989 there were weekly Monday evening peaceful demonstralions against the regime. (2) Opposil

oups

on the an-

niversaries of notable events, such as the Sharpeville and Soweta massacres in South Africa, the Kwangju massacre in Korea, the

coup against Salvador Allende in Chile, and the murder of Be-

nigno Aquino in the Philippines.

(3) Demonstrations were oxgamud as parl or a campaign 1o induce or compelthe demands. In both Brazil and Korea, formslance a s«.nl.-s< of massive

1HOW? CHARACTERISTICS OF DEMOCRATIZATION

205

s d

ds for direct

popular clection of lhe presldcnl

(4) Finally, opp

groups

n re-

sponse lo government outrages, such as the kll]mg ol’peacel’ul or polilical pri or other ly infamous acts of police brulahly In some Anstancefi, most natably in South Alrica, an acl of brulality would lead lo a demnonstration, often in the form of the funeral for the victim of the brutality, which would then provoke new acts of brutalily, giving rise for the need for further funeral demonstrations. This sequence of events led the South African government 1o ban funeral proces-

sions in August 1985

Whatever occasioned them, mass demonstralions almost invariably provided a setting [or violence. Even when moderale mai of d ions were (irmly commitled (o non\'lolence, at least some parlicipants in the demenstrations were susceptible io violence. Radicals would lake advantage of the cover and protective support provided by the demonstrations to throw stones or gasoline bombs at police and government vehicles. Often violence-prane groups would splil off from the main demonstralion to launch attacks on government largets. On the other side, even a peaceful ion could and [req ly did furnish an excuse for the police to resorl to violence. Mass prolests, in shorl, at times (1) inadvertently generated violence, (2) provided an opporlunity for radicals lo use violence. (3) provided an opportunity for regime standpallers to use violence, and {4) provided an opportunity for slandpatier agents to allack the police and thus creale a justification for massive govern-

ment violence against the opposition.

The use of violence moderate opposition usually were younger were those supporting cized the lo

was a major issue between radical and groups. Those commiited to violence and more likely to be studenls ll\an nonv:olenr larnrs They typi al.ly criti; 1 as “opp istic” and as

de facto collaborators with the regnme

In Koren, for instance,

206

THE THIRD WAVE

a maijor gulf existed belween Kim Dae Jung, Kim Young Sam, and other leaders of the mainstream opposition parties, on the one side, and, on the other, the young radical students and associaled toughs who filled the ranks of the protest marches and used (hose occasions for attacks on the police. At (imes student d d opposition leaders as much as they denounced the leaders of the government. These denunciations, according to Kim Dae Jung, “astonished” the moderate leaders.® They placed Lhose leaders in (he dilficult position of wanting to dissociate (hemselves from (he taclics the studenls were using while at the same time wishing to mobilize Lhe students for the massive peaceful demonstrations by which they hoped to bring down the regime. When the government agreed Lo a popular (ree election in 1987, the sludents remained unsatisfied, alienaled, and commilted to using prolests and violence lo promote socialist reforms and the end of American influence in Korea. In Chile, the mainstream opposilion p()lmml leaders similarly d to distance t Ives (rom the

party, the MRPF, and other groups using violence against the

regime. In the Philippines the Aquino forces rejected both the use of violence and cooperation wilh those who were using violence. In South Alrica, in conlrast, opposilion groups and leaders pursuing nonviolence, such as Tutu and other religious leaders, Buthelezi, and the liberal coalition —the

United Democratic Front, had littie choice but Lo cooperate

wilh the ANC. Inevilably, opposition radicals and angry crowds were termpled Lo resort to violence, and mainstream leaders were often hard pul to restrain them. Adam Michnik and Desmond Tulu shared the experience of having been imprisoned by the nondemocralic regimes in their countries. They also shared the experience of risking their own lives o prevent the lynching of g agents by ition mobs. A variety of factors thus mduced the Tevels of violence in

HOW? CHARACTERISTICS OF DEMOCRATIZATION

207

third wave democratizations. In considerable measire these factors also contribuled to the success of the democratization efforts. Violenl external intervention produced demactacy in several cases in the second wave and in Grenada and Panama in the third wave. Violence by groups within a society did not have the same result, The leaders of authoritarian regimes can successfully use violence to sustain Lheir rule; their radical opponents may successlully use violence to overthrow those regimes. The former action prevents democracy from coming inlo being; the latier kills it at birth. Throughout hislory armed revolls have almosl never produced democratic regimes. In nine oul of eleven unsuccessful efforts at democratizalion between 1860 and 1960, substantial civil violence had occurred during the twenly years preceding the democratization altempt. Only two ol eight successful efforts at democralization during the same period had been preceded by subslantial civil violence.* Similarly, between 1974 and 1990 violent upheavals cnded authoritarian regimes in Nicaragua, Yemen, Ethiopia, Iran, lHaiti, Romania, and elsewhere. In no case, with the possible but highly problematic exception of Romania, did democracy resuil. The resort to violence increased the power of the specialists in violence in both government and the opposition. Governinents created by moderation and compromise ruled by moderation and compromise. Governments produced by violence ruled by violence.

CHAPTER 5 HOW LONG?

CONSOLIDATION AND ITS PROBLEMS THE DEMOCRATIC ReFORMERS in Counlry A have achieved power and begun the transformation of Lheir country’s polilical system. The standpatter dictator in Country B has flown into exile on a U.S. Air Force jet, generating wild euphoria among his people, and the moderale democrats of the (ormer opposition now face the challenges of governing. Government and opposition democralizers in Country C have sacrificed the immediate interests of their constituents and agreed on the essentials of a new democralic system. In all three counlrics for the fi:sl time in years [ree and (air elections have

duced

larly chosen

" what lhen7 \Mm problems do the new democratic syslems

confront? Does democracy endure? Do the new systems con-

solidate or collapse? In both the firstand second reverse waves lwenly countries with democratic political systems shilted to itarian forms of g How many ol the thirty countries that lxaxmled 1o democracy in the 1970s and 1980s are likely to shift back to some form of authorilarianism? Two reversals occurred in Africa in the 19Ras: Nigeria in 1984 and the Sudan in 198y, Were these idiosyncratic or Lhe first manifestations of what would be a much more extensive collapse of new democratic governments? 208

oW LONG?

209

Speculations about the future are rarely illuminating; predictions about lhe future are often embarrassing. Previous chapters have examined the what, why, and how of third wave transilions to democracy. This chapter attempis lo inue this empirical app by analyzing: (1) two major transition problems confronting the new democracies; (2) the sleps involved in lhe development of democratic political institutions and a democralic political culture; and (3) the fac-

tors that may plausibly affect the probability of democratic

consolidation. Countries confronted three types of problems in developing and c Lheir new d: ic politicnl syslems. Transition problems siemmed directly from the phenomenon of regime change from aulhomanamsm to democracy. They included the [ of ing new constiluti and eleclural systems, weeding out proaulhonlanan officials and p g them with d ones, repealing or modifying laws thal were itable for d g or drasY tically changing authoritarian agencies such as the secret police, and, in former one-party syslems, separating parly and government property, functions, and personnel. Two key transition problems in many countries concerned (1) how to treal authoritarian officials who had blatantly violaled human rights, “’the torturer problem™ and (2) how to reduce military involvement in politics and establish a professional pattern of civil-military relations, “the praetorian problem.”

A second category of problems might be lermed comtextual

problems. These stemmed [rom the nature of the society, ils economy, culture, and history, and were in some degree endemic lo the country, whalever its form ol governmenl. The authorilarian rulers did nol resolve these problems and, in all probability, neither would the democratic rulers. As these problems were specific to individual countries and not to the of iti the i differed (rom :ountry to country. Among those that were prevalenl among lhird wave democracies, however, were in-

210

THE TIIRD WAVE

surgencies, communal

erly, socioeconomic

conflicl, regional anlagonisms,

inequalily,

inflation,

external debl,

pov~ low

rales of economic growth. Commenlators often siressed the threats these problems posed to the consolidalion of new democracies. In facl, however, apart from a low level of economic development, the number and sevcnty of a country’s to be only modestly relaled v its success or failure in cnnsolndalmg, democracy Finally, as the new democracies became consolidated and achieved a certain stability, they would confront systemic probfems stemming from lhe workings of a democralic system. Authoritarian polilical syslems sufier from problems thal derive from their particular nature, such as overly concenlraled decision making, deficient feedback, dependence on pecformance legitimacy. Other problems tend to be peculiarly characteristic of democralic systems: stalemate, the lnablhty to reach decisions, P w0 by vested economic interests. These pmhlems have afllicted long-standing democracies, and new third wave democracies presumably would not be immune lo (hem. The timing of these three types of prablems is presented schemalically in Figure 5.1, The following sections of this chapter deal with the problems posed by torhxrefi and soldner: dnd then discuss the role af in the ¢ id:

will be made to analyze the systemic problems arising from

the workings of a democracy as they are, in a sense, largely a postconsolidalion phenomenon.

Type of

pmblem: Transnllonal Systemic

Political Cvolution System L]

Phase e

System —

Figure 5.1. Problems Confronting Third Wave Countries

HOW LONG?

211 THE TONTURER

PROBLEM.

PROSECUTE AND PUNISH VS. FORGIVE AND FORGET

Among other things, new demoxratic regimes had to decide

what Lo do with the symbols, doclrines, organizations, laws, civil servants, and leaders of the authorilarian system. Beneath these issues often lay fundamenlal queslions of national identity and political legilimacy. One common issue concerned crimes committed by officials of the predecessor regime.? Democratic governmenls succeeding each olher seize opportunilies 10 expose and prosecute incompelence, corruption, and [mud by officials of the predt.c..ssor government. [ autharitarian governments faced a much more serious, emotion-charged, and politically sensitive issue. How should the democralic government respond to charges of gross violations of human rights—murder, kidnapping, lorture, rape, imprisonment withoul trial—commilled by (he officials of the authoritarian regimes? Was Lhe appropriate course to prosecule and punish or to [orgive and forget? The autheritarian regimes of the 1970s and 1980s provided ample reason [or lhis issue lo emerge. During lhe years of military rule, almost 9,000 Argentines disappeared, presumably murdered by the sccurity forces. and many others

were kidnapped and tortured. During its years of military

rule Uruguay had, according to one estimale, the highest pro-

portion of political prisoners of any country in the world.

One oul of every 50 Uruguayans was detained al one point or another, and many were lortured. Aboul 200 people disappeared or were killed in detention. In Greece those tor-

tured or olherwise abused apparently numbered in the hun-

dreds. In Chile, about 8oo civilians were killed during or immediately alter the coup in 1973 and another 1,200 were killed in later years. When an amnesly was declared in 1979, some 7,000 political prisoners were freed {rom jail. The Ceausescu regime violated basic human righls of Lhousands of Romani

212

THE THIRD WAVE

ans. Central American diclalorships of both Lhe Righl and the Left treated their populations, particularly racial minorilies, in similarly brutal fashion. Even in Brazil, some 81 civilians were killed and 45 disappeared in the war against the urban guerrillas between 1966 and 1975.% Acts againsl paclicular individuals were at limes supplemented by mass violence against demonstralors, as in the Kwangju and Polytechnic killings in Korea and Greece. These actions by the authoritarian guvernments of the late hwenlieth century did not differ greatly (rom those of earlier authoritarian regimes. Their behavior became a central issue

in Their societies largely because of the development of global

concern

about

human

rights

in the

1970s.

This

manifested

itself in the human rights legislation of the U.S. Congress, the emergence of human rights organizalions such as Amnesty Intemational, Freedom House, Americas Watch, and CSCE-related organizations, the more active role of intergovernmenlal human rights organizalions, and the Carler ad-

ministration’s effcctively pulting human rights on the world’s

agenda. As a result, once democratic governmenls came lo power they could not avoid conlronting violations of human rights by their predecessor regimes, even if, as was usually the case, efforts at puni: would P in their societies. The central importance of human rights is reflected in the nalure of the charges Lhal were brought against authorilarian officials. ‘The leaders of the Greek mililary regime were charged with carrying out a coup d’etat and were convicted of high treason. In virlually every other country, however, the charges and indictments concerned Lhe murder, kidnapping, and torture of individuals. In many countrics Where had

been broad popular support for the creation of lhe autlri-

tarian syslem and hence it would have been politically embarrassing and difficult to prosecute people for creating (hal system. The global concern with human rights focused oul-

HOW LONG?

213

rage not on the illegality of the regime bul on (he illegal aclions of ils agenls. Authorilarian officials were prosecuted not because they killed constitutional democracy bul because they killed individual people. In those countries where there had been egregious violations of human righls, major debates took place over (he acLion the democralic regime should take. Il was argued Lhal the perpetrators of such crimes should be proseculed and punished because: (1) Truth and justice require it; the successor regime has the

moral duly to punish vicious crimes against humanity. (2) Proseculion is a moral vbligation owed

to the victims and

their families. (3) Democracy is based on law, and the point must be made that neither high officials nor military or police officers are above

the law. As one Uruguayan judge put it, crilicizing the demo-

cralic government’s amnesly proposal, “Democracy isn’t just freedom of opinion, the right to hold elections, and so forth. Jt's the rule of law. Without equal application of the law, democracy is dead. The government is acting like a husband whose wife is cheating on him. He knows il, everybody knows it, bul hc goes on insisling thal everything is fine and praying every day that he isn’t guing to be forced to confront the truth, because then he’d have lo do something about it.” ¢ (4) Prosecution is necessary

lo deter future violations of hu-

man rights by securily officials. (5) Prosecution is essenlial to establish (he viability of Lhe

democratic system. 1( the military and police establishimenls can

prevent proseculion through political influence or the threat of a coup, democracy does not really exist in the country, and the slrugglc Lo establish democracy must go on. (6)

is

sary 10 assert the sup

of demo-

cralic values and norms and to encourage the pubhc lo believe in them. “Unless major crimes arc investigaled and punished, there can be no real growth of trust, no ‘implanting” of democratic norms in the society at large; and therelore no genuine ‘consolidation’ of democracy.” (7) Bven il most ian crimes are not d, ata

214

THE THIRD WAVE

very minimum it is necessary to bring into Ihe open the extent of the crimes and the identity of those responsible and thus establish a full and unchallengeable public record. The principle of accountabilily is essential to democracy, and accountability re quires “expusing the truth” and insisting “that people not be sacrificed for the greater good; thal lheir sulfcring should be disclosed and the responsibilily of the state and its agents for causing that sulfering be made clear.”* Those oppused to prosecution made countering arguments: (1) Democracy has Lo be based on reconciliation, on the major

groups in society selling aside the divisions of Lhe pasl.

(2) The process of democratization involves the explicit or im-

plicit understanding among groups that there will be no retribution for past outrages. (3) In many cases, both opposilion groups and government

forces grossly violaled human rights. A general amnesty for all provides a far stronger base for democracy than efforts ko prosecute one side or the other or both. (4) The crimes of the authorilarian

the time by the overriding

olficials were justified at

need to suppress terrorism, defeat

Marxist-Leninist guerrillas, and restore law and order 1o the so-

ciely, and their aclions were widely supported by the public al thal lime.

(5) Many people andgroups in the sociely shared in Lhe guilt

for the crimes committed

by (he authorilarian regime. **We have

all become used to the totalitarian system,” argued Vaclav Havel,

“and accepled il as an immutable fact, thus helping 1o perpetuate

it. In other words, we are aII—lhuuFH naturally lo various degrees—responsible for the crealion of the totalilarian machinery. None of us is just its victim; we are all responsible for it "

(6) Amnesly is necessary to establish the new democracy on a

solid basis. Even il a legal and moral argument could be made for prosecution, that would fall before the moral imperative of

crealing a stabie democracy. The consolidation of democracy should take p over the of i

Or as President Sanguinetti of Urusuay put it: “What is more just—to consolidate the peace of a country where human rights

are guaranteed today or to seek retroaciive justice that could compromise thal peace?”*

HOW LONG?

215

Such, in summary, were the arguments for and against proseculion of authoritarian crimes in third wave countries. In aclual praclice what happened was little affected by moral

and legal considerations. Il was shaped almost exclusively by

politics, by the nature of the democratizalion process, and by

Lhe distribution of political power during and after the tran-

sition. In the end, the working of polilics in the third wave counlries undermined Lhe elforts to prasecule and punish authorilarian criminals. In a few countries, summary justice was meled out to a few individuals; in almost all counlries, no effective prosecution and punishmenl occurred. In the countries that democratized before 1990, only in Greece were therea number of itarian olficials subjected o meaningful trial and punishmenl. Given the heated political debale and the emotional intensity of this issue, how can this result be explained? First, about half of the pre-1990 democralizations were transformations initiated and guided by the leaders of the existing aulhoritarian regimes. These Jeaders were usually democratic reformers who had, in most cases, displaced previous standpatter leaders. The latter obviously did nol wish to be prosecuted for crimes they may have committed. The reformers wanled to induce the standpatlers 10 acquiesce in the dcmocmlualmn process, and assuring them that lhey would not be punished by Lhe regime was essential lo achieving that. In addition, authorilarian regimes in which reformers were able Lo displace slandpatters in power were likely to have been responsible for fewer and less heinous human rights violations than those regimes where standpalters slayed in power to the bitler end. Hence virlually every authorilarian regime that initiated its Lransformation to democracy also decrecd an amnesty as a part of

that process. These amncsties typically applied (o any crimes

commilted during a specified period of lime by either the agents of the regime or members of the opposition. The Brazilian and Chilean regimes enacted such amnesties in 1979.

216

THE THIRD WAVE

Guatemala did so in 1986. The Turkish generals guaranteed Lheir immunity against proseculion before allowing a democralic government to be clected in 1983. In these and olher cases of transformation, the authori tarian regimes not only acted in their own interests in legislating amnesty, they also had the power to make lhe amnesties stick. The argument can be and has been made that as a result Lhe successor democralic regimes were not real democracies because they lacked the power to hnng Lo justice those in the aut ian regimes who ¢ crimes. In Guatemala, (or instance, the mililary declared an amnesty (or themselves four days before turning over the government to the democratically elected civilian president, Vinicio Cerezo,

in January 1986. Cerezo accepted Lhe amnesty, extended it lusther, and readily admitted that he would not long continue as presi ifhe pted Lo any Gi lan soldiers for the apparently nu erous human rights violations thal occurred during the military regime. Hence, il was argued, the Cerezo government was nol “a democralic governmenl or even a governmenl lhal is making a transition to

democracy.”® The same charge could be and was made againsl the postPinochet regime in Chile. The multi-parly coalition backing

Palricio Aylwin for president in the 1989 election adopted a program urging legislation lo overlurn the 1979 amnesly law and to set aside for one year the len-year statute of limitations for prosecutions for murder and other violent crimes. Chilean mililary leaders warned of the consequences if efforts were made W acl along these lines. In October 1989, while still president, General Pinochet declared: “The day lhey louch any of my men will be the end of the state law.” The commander of the Chilean air force, Gen. Fernando Matthei, for ten years had struggled with Pinochet and urged the desirability of speedier democratization. Efforts to repeal the amnesty law, however, he warned in 1989, would Lhreaten that process. The armed forces will “not accept” prosecutions: “If

HOW LONG?

217

they are going lo Lry to put us in the pillory, as in Argentina, that is going to have the most grave consequences.” Leftist oppusmon Ieaders continued to insisL on the possibility of d ic leaders, however, emphasized only lhe desirability of investigations. A few days before becoming president, Aylwin went oul of his way to reassure the mililary. “The idea of a trial is nol in my mind,” he said. “It is nol my intention (o promote trials. . . . [l am nol in the mood] to persecule or lo antagonize General Pinochet or anyone else.”" In keeping with his theme of national reconciliation, immediatcly after taking office Aylwin freed political prisoners imprisoned by the Pinochet regime who had not committed acls of violence. The crealion of a democralic system always involves comise among the politi 1 groups as to what that government can "and cannot do. The establishment of Venczuelan democracy in the lale 19508 required commitments 10 honor the privileges of the Church, to respect private property, and to underiake fand reform. Was the resulling political system undemecratic because of these constraints? [s a systemn undemocratic because the governmenl lacks either the power or the will to prosecute criminals in the predecessor authoritarian regime? If it is, then no democralic system established (hrough ion is democratic because clearly no authorilarian leaders will make their system democralic if they expect themselves or their associates lo

be prosecutcd and punished as a result. Governments that

are strong enough lo bring aboul transformations are strong enough 10 exact thal price. If they had not been, possibly half of the pre-1990 third wave transilions would not have occurred. To reject ammsty in these cases is to exclude the mos! form of d "The situation was quite different with authoritarian regimes thal did nol exil with sirength bul were replaced (rom weakness. Normally such regimes did not anticipate lheir end and hence did not atlempt to protect their members by enact-

218

THE THIRD WAVE

ing amnesties. The most nolable exception was Argentina, where General Bignone led an interim military regime for sixteen months from the downfall of the junta to the election of a democralic government. His government made successive cfforts, all of which (ailed. lo pwlecl the mllllnry and police {rom prosecution. It first an agreement on nonproseculion wilth cmhan leaders, this effort “was widely ridiculed.” IL then issued a lelevised reporl on ils war against leftist terrorists in the hopes of justifying its actions; this “backfired badly.” It next atiempled lo negoliale a secrel agreement with conscrvative union leaders in the Peronist movemenl, bul this was exposed by other opposilion leaders and had to be abandoned." Finally, a [ew wecks before the clections, Bignone's government decreed a “Law of Nalional Pacification” which granted immunity from proseculion and investigalion to all military and police officers for virlually any actions, including *‘common crimes and related military crimes,” taken during the war against lerrorism. The amnesty also applied to a small number of terrorists, perhaps a dozen, who had nol been arrested and convicled and who were not in exile. This law was immediately denounced by all opposition political leaders. The democralic government repealed Lhe amnesly law two weeks after laking office in December 1983. Officials of authoritarian governmenlts (hat cullapsed or

were overthrown were largets for punishment. After (heir re-

moval [rom power by U.S. military forces, Bernard Coard and thirteen other leaders of the Grenadan communist regime were convicled of murder and other crimes and received long jail sentences. If Noriega had not been taken lo the Uniled Slates for proseculion on drug charges, he would have faced a variely of charges in his own country. The Ceausescus got

rather summary jusice. Before her election, Corazon Aquino

threatened Marcos with prosecution, which he escaped by flying into exile. Honecker and his associates initially avoided serious punishment by being old and sick, bul al the end of

HOW LONG?

219

1990 Honecker was charged wilh ordering the killing of East Germans who had atlemptex to escape over the Berlin Wall. The most extensive and serious elforts Lo prosecule and punish authorilarian criminals were made in Greece and Argenlina. The situalion in these two countries seems on the surface lo have been ralher similar. Both military governments were clearly guilty of substantial violations of human rights. Both governments collapsed alter suffering mililary humiliation in external conflicts. [n both cases, the successor democratic g were fully lled by one party and headed by a highly respecled and popular leader. In both cases also, when the democratic governmenl was elected, public opinion jon and punishment of Lhose guilly of human nghls violations. In Greece, as one analyst pul i, ihis was “the most sensitive and c.xploslve of popular demands.” ' The same could be said of

th new g

d to respond to

popular demand and moral imperative hy developing a program to deal with human rights violations. Both got also similar problems in developing their policies. They had 1o determine who should be prosecuted lor what crimes, in what manner, when, and before which tribunals. When personal dictalorships were overthrown in Colombia, Venezuela, the Philippines, and Roma-

nia, prosecution and punishment were largely limiled to the

dictator,

his (amily, and close associates. The replacement of

B

a military regime provided a more dilficult challenge. In both Argentina and Greece it was clearly necessary to prosecute Lhe top leaders of the military government. But how far down the military and police hierarchies should the prosecution go? The Alfonsin government attempled to deal with this issue by dividing those potentially guilty into three categories: (1) those who gave orders to violale human rights; (2)

lhose who carried out the orders;

(3) those who engaged in human rights violations beyond the

actions they werc ordered 1o take.

220

THE TIIIRD WAVE

Under Lhe government's policies mililary and police officers who [ell inlo calegories (1) and (3) would be indicted and tried; Lhose in calegory (2) would be tried only il they knew that the orders Lhey were given were clearly illegilimate. The conditions that confronted the Alfons(n and Karamanlis governments were thus quite similar. The results the two governments achieved were very different. By August 1975, nine months after Karamanlis had been elected prime minister, Uie eighteen lop officials of the military governmenl had been indicted, lried, and convicied of high treason. In the first torture trial that imunediately followed, thirty-iwo military police (fourieen officers and eighteen enlisted men) were indicted and sixteen were convicled. Three other military, naval, and police trials in 1975, lwo more in lale 1976, and a trial of the top military government leaders for the Polylechnic massacre produced addilional convictions and prison sentences. All in all, between 100 and 400 trials for torture were apparently held throughout Greece, and a large number of people were convicted of human rights violations in those trials.”* By the end of 1976, two years after the democratic government came to power, substantial justice had been done and the prosecution and punishment issue laid to rest in Greek politics.

Human rights viclati

greater ahsalutely and Greece. The i

Angentina

proporhonally

were

than

I

they were in by Alfonsin

concluded that the security forces had ”forcnbly disappeared”

at least 8,960 people, that they had maintained a network of 340 secrel delention and torture centers, that about 200 officers, whom it named, had been identified as directly involved in these operalions, and that many more had also participaled in the repression.'* For seven years after Alfonsin came to power, the issue of how lo handle human rights vielations

aitated and at Limes c

Argenline

politics, stimulat:

ing at least three mililary coup attempts. When all was said and done, a total of sixieen officers (including members of Lhe

HOW LONG?

221

ruling military juntas) were broughl to trial and len were convicted of human rights violations. in contrast lo Greece, the efforts lo prosecute and punish in Argentina served neither justice nor democracy and instead produced a moral and political shambles. In 19g0 the issue was still a highly divisive factor in Argentine politics. Argentina was lefl with memories of Lhe civic trauma caused by lhe efforls to pmsecule authoritarian criminals that emories of the civic and personal trauma caused by Lhe horrendous crimes they committed. How can the difference in the Argenline and Greek outcomes be explained? In part it was due to the fact that a serious inlernal security threat had existed in Argenlina, that lhe Peronist governmenl preceding the mililary junta had direcled the military to “eliminate” the lerrorists, and that major seclors of the Argentine public had quietly accepted and in some measure approved lhe mlhless tactics the military d ing that mission. The less subslanna] human rights violations of the Greek military also had less juslification. Tn addition. the Greek military regime had been a colonels’ regime, opposed by some more senior officers. Hence elements of the Greek rmlllary were supportive of the government’s prosecution efforts, while the Argentine mililary, whatever their other dillerences, unanimously opposed such efforts. These factors, however, expiain only part of the difference belween Greek success and Argen-

line failure in dealing with this issue. Far morc important

were the policies and slralegies of the lwo governments.

Karamanlis did two things. First, he moved quickly when

popular support was greatest. 11is was a “policy of swifl, decisive, credible, but also contained and circumscribed retribution.”** Confirmed in office by elections in 1974, K: lis prepared plans for prosecution of human righls violators. Supporl for such action was greatly increased by a serious milifary coup attempt against him in February 1975. Capitalizing on the popular reaclion against this, he moved quickly

222

THE THIRD WAVE

to Taunch indictments and trials lhrough the normal civilian courts, with the resull that the process was substantially concluded within eighleen months. Second, Karamanlis acted to reassure the officer corps that he was not challenging it institutionally. Ilis “policy of chleldmg the ulficer corps from criticism and his ds and requirements [orestalled a possd)le backlash by officers on active duty. The policy of a limiled purge of junta principals conducted under regular legal procedures and after public passion had subsided also relieved much of the anxiety of the officer corps.” The torturer issue then virtually disappeared from Greek politics for fourteen years only to be revived briefly in December 1990 when the conservative government announced that it would pardon seven of the eight junta members in prison. The government reversed itsell quickly, however, following public outcry and the reported refusal of

President Karamanlis lo sign the decree.!*

The failure of the elflorts at prosecution in Argenlina and the serious crises they generated for Argentine democracy largely resulted {rom Lhe policies of the Alfonsin government. These produced an extended delay in the trial and punishment of the human rights violalors and encouraged the mililary lo resist this process. As time went on, public outrage

and support (o |

gave way loindifference, and the

military regamed influence and stalus from their humilialion

of 1982 1983. Immediately on taking office, Alfonsin repealed the Big-

none g amnesty, app a civilian led by Ernesto Sabalo, a novelist, to investigate military crimes, initiated the prosecution of nine top military officers, and sccured passage of legislalion providing the legal basis for prosecuting human righls violalors throughout the military and police hierarchies. These aclions provoked fear, concern, and opposition within the military establishment. At the same time, however, Alfonstn also secured legislation

that military officers charged with human rights

violations

HOW LONG?

223

would be initially tried in military courts. He thus provided the mililary first with the incentive and then with the means 10 obstrucl prosecutions. The case of (he nine generals and admirals went belore the Armed Forces Supreme Council in early 1584. Eight months laler, in September, the Council reported Lhat il had found “nothing objectionable” about the aclions of the junla members. The case was then transferred to the civilian Federal Appeals Court. The Lrial there lasted another year with a judgment in December 1985 convicting five and acquitting (our of the junta members. During 1986 seven more high-ranking of-

ficers were indicted and five were cventually convicted of hu-

man rights vi i were proceeding and indictments being drawn up against many other officers. Public opinion polls in 1984 and 1985 showed widespread supporl for prosecutions, particularly of the lop military commanders. Yel concern with this issue also began to fade. A year afler Alfonsin took office, it was reporled that “ /\Imady, many Argenlines may be losing interest in desaparecidos.” The ion were smaller. One aclivist complamed 'we are .\Jwavs the same bunch.” 7 Public opinion was changing, mnlnlary resistance was growing, and in December 1986 Alfonsin proposed a punto final or “full stop” bill that would prohibit initiation of addilional cases. Human rights groups denounced this proposal, but most Argentines were indifferent, and an eflori to mobilize a 24-hour prolest against the bill failed. Congress passed the bill and deadline charges were filed against some 200 officers, including many on active duly. The military made clear lhat they strongly opposed trying the latter. At the same lime, the mos! prominent prosecution group, the Mothers of the Plaza de Mayo, split, and its weekly marches came to resemble “a reunion of aging cousins or old friends, many accompanied by husbands, children and grandchildren.” ' In April 1987 an efforl to compel the appearance of an offi-

224

THE TIIRD WAVE

cer in court produced the so-called Caster Rebellion, in which army units on fwo bases mutinied and made several demands on the government. Alfonsin successfully induced the insurgents Lo surrender but acquiesced Lo their principal demands. lle Gircd the army chief of staff and, over the vigorous objections of human righls groups, secured passage of a “due obedience” law, which effectively barred prosecution of everyone except a hand[ul of retired olficers. In January and December 1988 additional military uprisings were supp , but lhese put more pressure on the governmeni to forego prosecuhons. In 1989 the Peronist candidate, Carlos Menem, was elecled president. The Peronist party by then had developed close ties to Lhe military, and Menem appoinied a supporter of amnesty as mnister of defense and a favorite of the military mulineers as army chief of stall. In October 1989 Menem pardoned sixty guerrillas and all the military and police officers charged or potentially chargeable with human rights violations, except for the five junta members still in prison. On December 29, 1990, Menem pardoned the five junta members, another general extradited from the United Slates in 1988 and awailing trial on thirty-cight charges of murder, and a Montenero guerrilla leader. 1lis action stimulated intense bitterness, antagonism, and outrage. Almost 50,000 people protested in Buenos Aires. “This is,” said (ormer President Alfonsin, “the saddest day in Argenline history.”” In Grecce the confrontalion belween the demogratic gov-

ernment and the military came to a climax in an atiempted

coup three months after the government was elected. In Argentina the climax came in an attempled coup three and a half years after the govcrnmem was elected. The coup allempt in Greece p and d suppori for prosecution. The coup atlempt in Argentina induced the governmenl to give up prosecution. Alfonsfn’s failure lo move quickly and decisively in 1984. when public opinion supported action, made human rights prosecutions the victim of changes in power relationships and public allitudes. The re-

HOW 1.

225

sult, as Ernesto Sabalo put it, was lthal “A man who steals a pocketbook is thrown inlo jail, and the man who has Lortured

goes [ree,” [n contrast 1o what happened in transformations and replacements, in transplacements the terms of amnesty were negoliated explicitly or implicitly between government and opposition. In Nicaragua, for instance. the Sandinistas advanced onc amnesty proposal but then modified it to meet the objections of the democralic opposilion. In Korea President Chun Doo Hwan undoubledly backed his colleague Roh Tae Woo for president on the assumption and with an implicil understanding that he and his associales would not be prosecuted for any actions they took during their seven years of authorilarian rule. Once Roh was elecled, however, arose for an g for crimes iled by Chun government officials. These included, most notably, the Kwangju massacre, but in addition Lhere were “many welldocumented allegations of torture, viclimizalion and unexplained deaths while in custody.””" Opposition assemblymen demanded punishment for Chun and [ive others they held responsible for the massacre. During 1988 and 1989 the issue was intensely debated in Korea. In November 1988 Chun made a public apology and then retired to an isolated Buddhist monastery. Meanwhile, equally intense negotialions went on between the Roh governmen! and the opposilion,

focused on hearings, prosecutions, and possible reparations to the Kwangju victims. In the end, a “backroom deal” was

reached under which Chun testified in a televised hearing before the Assembly. This somewhal raucous affair did not salis{y lhe opposilion bul it did help Roh to distinguish his government from that of his predecessor. In Nicaragua and Korea negotiations occurred between government and opposition alter a govemmpnl had hfl'n elected d ically. In other i an was reached between authoritarian and opposition leaders before a new government was elected. In Uruguay, for in-

226

TITE THIRD WAVE

stance, the military and some polilical leaders negotiated arrangements for the transilion to democracy at the Club Naval in the sunmer of 1984. Controversy exists as lo what extent the mililary received guarantees against proseculion, Subsequently both General Medina and President Sang netti said that the issue was never discussed. Some civilians said that the mililary were assured Lhal the government would not prosecute them, although it would not prevent privale citizens (rom bringing suits. Others said the mililary were assured that these aclions too would be blocked, and Wilson Ferreira, leader of ll\e upposmon Blanco party, subly justified abandoning his opposition Lo amnesty on the grounds that the military had recelved guarantees against prosecution in the Club Naval negotiations {in which he had

not taken part because he was in prison).® The Uruguayan generals were under considerable domeslic and inlernational pressure (from the Uniled States and the Argentine

and

Brazilian

democratizations)

(o end

their

rule. They were nol, however, driven from office. They negoliated their way out, and, as one Uruguayan political columnist remarked, “The generals believe they are leaving office with heads high and flags flying.”? It was certainly widely assumed at the time that some understanding concerning prosccution had been reached al the Club Naval. Given the scope and seriousness of the human rights viola-

tions, it seems likely that the generals surrendercd power ei iher because they received some assurance or becauge both

sides assumed serious pruseculions were oul of the queslion. In the year and a half after the elected government took office in March 1985, Uruguayan cilizens initiated thirty-eight cases against 150 officials alleging murder, torture, kidnapping, rape, and other crimes. The military declared they would not allow their members to appear in court. Wishing to avoid a major confronlation, which could have meant the end of his regime, Sanguineili proposed a blanket amnesty

for the military, justificd in part by the fact that he had al-

HOW LONG?

227

ready pardoned the lerrorists and other political prisoners the military had imprisoned. The opposition parties in the legislalure defealed his proposal and submitted a plan for a partial amnesty, which also was defeated. The first trial at which military officers were supposed lo appear in courl was scheduled for December 23. Al Lhe last moment, the opposition polilical leadership shifted its position, the legislature worked inlo the night of December 22, a general amnesty bill was approved, the (irst trial was cancelled, and no further trials were held. The political balile in Uruguay, however, was only just beginning. When the amnesty bill was being debated, public apinion polls showed 72 percent of the public supporting punishment of those convicled of human rights violations.*

Two months after the bill was passed a broad coalition of op-

position politicians, human rights activists, victims and Lheir (amilics, clergymen, joumnalists, lawyers and others slarted a campaign to submit the amnesty law to a referendum, This required Lhe signatures on petilions of one quarler of the voters in the last cleclion. or 535,701 signatures from a total population of slightly over 3,000,000, The campaign to get those signatures dominated Uruguayan polilics for two years, The government, Electoral Courl, the mililary, and some opposilion political leaders made strenuous efforts, by fair means and foul, to obstruct the gathering and the verifi-

cation of the signatures. In the end, in December 1988, the Electoral Court held thal referendum supporiers had pro-

duced 187 more valid signatures than were required, and the Congress scheduled Lhe referendum for April 16, 1989. After a bilter campaign thal included not-s iled threats from the military that it would not acquiesce in repeal of the law, the Uruguayan public voled 53 percent lo 40 percent in favor of amnesty. This oulcome resolved the issue bul satisfied no one. Those opposed Lo amnesty lost; but on the other side, there was, as President Sanguinetli said, “no sense of tri-

umph.”= It was nine years after the beginning of the demoe-

228

THE THIRD WAVE

ratization process and almost five years after a democratically elected government had assumed power. In Eastern Europe, apart [rom Romania and East Germany, the initial overall tendency was to (orgive and forget. The issue of punishment never really arose in Hungary; and Havel in Czechoslovakia, Mazowiecki in Poland, and Yakolev in the Sovict Union all argued againsl criminal charges. In several countries,

however,

demands

arose for investigation

and prosecution of those responsible for the most notorious crimes. The former communist leader in Prague was tried and sentenced to four years in prison flor using violence against demonstrators. Two generals in Lhe Polich secret police were arrested and charged with “insligating and direcling”™ lhe murder of Father Jerzy Popieluszko in 1984. Todor Zhivkov was taken inlo custody in Bulgaria (or six months and then released in July 1990, with charges against him still being investigated. The record of democralic governments in bringing Lo justice authoritarian officials who had committed crimes yields some indisputable conclusions, Juslice was a function of politieal power. Officials of strong authoritarian regimes lhat voluntarily ended themselves were not prosecuted: officials of weak authoritarian regimes that collapsed were punished if they were promptly prosecuted by the new democratic government.

“Justice,” Emeslo Sabalo once remarked,

“works in

this way. It is slow. The only quick justice belongs to totali-

tarian and despotic countries.”* He was wrong. Democralic

justice cannot be summary justice of Lhe sort meted out to the Ceausescus, bul it also cannol be slow justice. The popular suppart and indignation necessary to make justice a political

realily fade; (he discredited groups associated with the au-

thoritarian regime reestablish (heir legilimacy and influence. In new demaocratic regimes, justice comes quickly or it does not come at all. With the end of authorilarian regimes, there were demands for truth as well as for juslice. In Argenlina this demand was

HOW LONG?

229

met by the report in Seplember 1984 of Lhe Sabato commission, appomled by President A]Ions[n the previous December. The eard (rom victims, Lheir families, officials, and others, exhaustively mined official records, and visited detention and torlure centers. Its 400 page summary reporl was backed by 50,000 pages of documentation. The revelations and evidence in the reporl increased the probability lhat the promise of its litle would be realized: “Nunca Mas” (Never Again). ln the Phnhppmes, President ‘Aquino appointed a Presi on Human Righls to investigale primarily abuses by the police rather than the

military, who had, after all, helped her come lo pawer. The

chairman of the commnittee died shortly aiter his appointment, and the committec disbanded in less than a year without having had much impact. In Drazil, the transformation process and the muth lower level of criminal violence precluded the ic go? from king an investigation like that in Argentina. The Archdiocese of Sio Paulo, however,

did produce a reporl based on official records that

was somewhal stmilar to the Sabato report and had the same title: “Nunca Mais.” In Chile, Aylwin opposed prosecution bul determined lhat Lhe truth should out. The government

appointed a Commission for Truth and Reconciliation which

was lo be the “moral conscience of the nation™ and investi-

gate and reporl (ully on political killings and disappearances

during the military rule. The assumption was thal those re-

sponsible for these crimes would nol be proseculed but that

viclims and their families would be compensaled.?” In Uruguay the desirability of “truth” as well as of “justice” was the subject of debate. The supporters of amnesty also argued against investigation and exposure of authoritarian crimes. Forgelling was essential, as well as (orgiving, “Amnesly does not mean that the crimes did not take place; it means forgetling them,” argued Sen. Jorge Batlle. “The botlom line,” Sanguinetti said, “is thal either we're going to look lo the future or to the pasl. . . . If the French were still

230

THE

TITRD WAVE

thinking aboul the Night of St Bartholomew, they’dbe slaughtering each other to this day.” Olkers in Uruguay and elsewhere argued that truth was, if anything, more important than justice; that it was essential to lay bare all the facts

of what happened in order lo provide some consolation to

the victims and their families, to expose and humiliate (he torlurers, and Lo develop a public conscience and determination that would ensure that such actions would indeed never happen again. Aryeh Neier concisely laid out this case: “By knowing what happened, a nation is able to debate honestly why and how dreadful crimes came (o be commitled. To identify those responsible, and to show what Lhey did, is to mark them with a public stigma that is a punishment in itsell, and toidentify the victims, and recall how they were tortured and killed, is a way ol acknowledging their worth and dignity.” The main constraint on both prosecution and disclosure in the former communisl counlries was the pervasiveness of lhe commwunist regime, the extent 1o which so many people accepted it and collaborated with it, and the fear of whal proseculions or inveslgalions mighl reveal. Zhivkov’s scheduled i before the ian parli was rep dl postponed because of the people he might name. The mosl acute problem was posed by the immense files of the secrel police. Should Lhese be generally opened up, made available only to p locked up, or yed? The East German files reportediy included the names of six million people, and in 1990 several parliamentarians and ministers in the successor regime were exposed as collabarators with the police. Many feared that an unselective opening of the files could poison public life in the new democracies, and a general opening of those in East Germany could also have elfecls on West Germany. In Romania the huge files of the Securitate were kept at a secret localion under military guard. “If we publish the files as some people hove suggested,” a government official observed, “there could literally be something

HOW LONG?

231

worse than a civil war with friend lurning agains! friend once they find out what is contained in them.”* In some respects, truth as well as justice was a threat lo democracy.

Guidelines for Democratizers §:

Dealing with Authoritarian Crines (1) If transformation or transplacement occurred, do not altempt to prosecule authorilarian officials for human rights violations. The polilical costs of such an effort will outweigh any moral gains. (2) I replacement occurred and you feel il is morally and politieally desirable, prosecute the leaders of the authorilarian regime promplly (within one year of your coming into power)

while making clear that you will not prosecute middle- and Iower-ranking officials.

(3) Devise a means to achieve a full and dispassionate public accounting, of how and why the crimes were commilled. (4) Recognize that on the issue of “prosecute and punish vs. forgive and forgel”, each allcrnative presents grave problems, and that the leasl unsatisfaclory course mi?' well be: do not rosecule, do not punish, do nol forgive, and, above all, do not

orgel.

‘THE PRAETORIAN PROBLEM: REBELLIOUS AND POWERFUL MILITARIES

The problem of dealing with the criminal actions of authori-

tarian officials overlapped with a broader, more lasting, and politically more serious problem confrenting many new democracies: the need to curb Lhe political power of the military eslablishment and (o make the armed forces inio a professional body committed to providing for {he external security of the country. The civil-military problems of Lhe new democracies took one of three (orms, depending on the type of autherilarian regime, the power of the military establishment,

232

THE THIRD WAVE

and the nalure of the Lransition process. The military forces of one-parly dictalorships, wilh the glaring exceplion of Nicaragua and the partial exceplion of Poland, had generally been under the firm control of the party. They did nol attempt coups and did not play a major role in the politics of the regime. In communist states, and to a lesser degree in the Republic of China, most military officers belonged to the ruling party, parly cells and organizalions permeated Lhe military hierarchy, and the top party bodies shaped military as well as other policies. The problem in the successor democralic states to these regimes was lo separate party (rom army and lo replace military subordination to one party with military subordination Lo a democratic muitiparty system. In Eastern Curopean countries the separalion of party and army went relatively smoothly. In the Soviet Union intense debates occurred over “departization” of the military; legislalion passed in 1990 changed the functions of the Main Political Administration bul le(l intact the structure of party cells despile arguments that “There are no organs in the armies of democralic countries (hat implant the ideology of a single party.”* In general, however, the democratic successors to ane-party dictatorships faced fewer difficulties in establishing civilian contro} than the democracies that followed mililary regimes and personal dictatorships. Different and more serious problems arose (rom military establishments that had been replaced (that is, overthrown)

in the transilion process or had become highly politicized in

personal dictatorships. The middle- and lower-ranking officers in these i often had well-developed political views or ideologies, grealy resented their loss of power and slalus, and felt threatened by the forces active in and dominant in the new demuxratic politics. Hence they ofien engaged in a variety of political aclivilies designed to bring down the new democralic regime or to force changes in its leadership or policies. The most dramatic political ac-

HOW LONG?

233

tions were, of course, military uprisings and attempted coups d’etat. Coups were atlempted or seriously planned in al least ten countries that democratized between Lhe mid-1970s and Lhe late-1980s. In Nigeria and the Sudan, the coups were success{ul, and military regimes were reestablished where they had ended a {ew vears earlier. A distinclion, however,

should be

made between coups that are reactions to the perceived failures of a democratic system and coups that are reactions to Lhe prospective success of democralization. Both Lhe Nigerian and Sudanese coups partake of more of the former than of the latter. The Nigerian coup came afler a disputed eleclion thal relurned to power a president widely regarded as corrupt and ineffective; the Sudanese coup followed three years of inept civilian government during which the economic and insurgency problems ol the country had greatly worsened. Coup atlemplts also occurred in Gualemala and Ecvador, where the military had volunlarily exited from power. These efforts appear to have been related to intramilitary struggles. In Gualemala the elected civilian government clearly held power al lhe sufferance of the armed forces, and the attempted coups of May 1988 and May 1989 were quickly suppressed by the governument and the military leadership, with many people wondering aboul Ihe molives of the coup makers. As one lan politician “The only explateading G nation is that lhis is a very isolated group. The army is the

power that designed this process [of democratization). They

are nol crazy—they are in control.”® Coup allempts were most {requent in countries at the opposite extreme from Guatemala, countries where the armed forces had been humilialed by defeat or politicized by personal dlictators. Seven coup allempts or conspiracies are repurted 1o have D:curred in Greece; Seven coups or military uprisings were pled in the Philipp Give in A three in Spain. Belween 1974 and 1990, however, except for

234

TUE THTRD WAVE

the ambiguous cases of Nigeria and the Sudan, no democratizing government was overthrown by a military coup d’etat. ven lhe apparent (ragilily of new democracies, why was this the case? Overthrowing a government by coup d"etal normally requires support from cither the mi high command, or important civilian groups, or an influential foreign actor, or some combinalion of these. The antidemocratization coup makers were moslly middle-ranking olficers. The same officers often led successive coup efforts: Lt. Col. Aldo Rico and Lt. Col. Mot d Ali Seineldin in Argentina, Col. Gregorio Honasan in the Philippines, Lt. Col. Antonio Tejero Molina and Lt. Gen. Jaime Milans del Bosch in Spain. In virtually all cases, the top military leadership supported the government or al leasl did not back the coup attempt. The coup makers also were generally unable to mobilize the support of significant civilian groups or of influential foreign governments. In these respects, the transilion coup attempts of the third democratization wave differed significantly from the successful coups of the second reverse wave. One of the most serious antidemocratizalion coup efforis, for instance, occurred in the Philippines in December 1989 and involved extensive fighting and casuallies. Some political leaders tacitly supporled the coup but they were nol aclive in Lheir support; key civilian groups opposed it. So also did Minister of Defense Gen. Fidel V. Ramos and the military high command. And at a critical momen, the United States decisively inter-

vened, with American fighter planes laking lo he air to deter

the procoup [orces. In general, the of elected g to overthrow by coups varied wilh the leve] of socio-economic development of the socicty. In rural societics and in poor societies, coup-prone military officers could ofien find aclive support and cooperalion among civilian eliles. As the power of landowners and resource exlraclors declined and that ol the bourgeoisie and middle class grew, the social basis

HOW LONG?

235

for mililary coups withered. In Peru the agrarian reforms of

the Velasco military regime wiped out the landowning elite

and hence greatly reduced the coup threat to the subsequenl democralic regime. In Spain, business leaders and other civilian elites opposed military coups and supported democracy. Alfonsin argued that coups in Argentina “have always been civilian-military in character” and (hal the key 1o preventing them was to break the link between ihe military and civilian groups.¥ The true key, however, was (he changing balance of forces among civilian groups in Argentine sociely and the emerging numerical dominance of the middle classes, who no longer would have to rely on the mililary to prolecl themselves against the organized power of Peronist labor unions Thus, when coups threalened in May 1985 and in the Easler Week crisis of April 1987 Alfonsfn was able to mobilize hundreds of Lh Is of supporters to d on his behalf inthe slreets of Buenos Aires. Massed people power countered mililary firepower. In Augusl 1474 in the first phase of the Greek transilion, military leaders initially delied Karamanlis’ demand thal certain tank units be removed (rom Athens Karamanlis replied, “Either you take the lanks oul of Athens, or the people will decide the issue al Constitulion Square.”* The lanks lelt. If the politically dominant (.wnllan group in Greece had slill been landlords and the cl group had still been peasants, the outcome very likely would have been dilflerent.

The efforls to overthrow new democratic governments failed

because coup makers were unable to win o Lheir side middleclass and other groups in the political coalition that had made democratization possible. The coup atlempts were, in effect, desperam rearguard actions by minority slandpatier elcments in the military. Like the [ 1944 Ard they were efforts Lo reverse the course of events after the war had been lost. A coupalttemptagainsl a new democralicregime is a sign that democratization is working. The failure of that

236

altempt is a sign that d

THE THIRD WAVE

ization

is working

While no government was overthrown by a transition coup belween 1974 and 1990, coup elforts and military uprisings

did on occasion affect government action. The coup allempts

in Guaternala in May 1988 and in Argenlina in April 1987 and December 1988 were aimed at, among other things, forcing changes in the top military leadership. In the latter two cases, lhey achieved lhen ob)ocuves The April 1987 Easler Rebellion in Arg also led the g ta enact its “due obedience” law, effectively precluding prosecution of

aclive duly officers for human rights violations. Coup altempts were thus one way by which dissident officers could pressure governments to change officials or policies. Governments also were generally lenient in their trealment of coup makers. In Spain, Tejero and Milans del Bosch were sentenced to thirty years in prison, but this was exceptional. When proseculed and convicted, leaders of coup cfforts normally received fairly minimal punishments, and rank-and-file parlicipants in such efforts were normally not punished at all. Coup efforts were the most extreme form of political aclion by disgruniled officers. At times, such officers also engaged in other, more normal forms of political activity, Midway between the February 1981 and October 1982 coup attempls in Spain, [or instance, disconlenl manilested itsell among middle- and lower-ranking officers, with 100 junior and noncommissianed officers signing a public manifesto protesting

press treatment of the military and arguing thal the military

"in order to better [ulfil] their mission, do nol need to be professionalized, democratized, or purged.” Reportedly only vigorous action by the top military officers prevented several hundred olher officers from signing (his declaralion.™ In Guatemala, for several weeks before the May 1988 coup effort, an anonymous military group issued communiqués from “the Officers of the Mountains.”

The Portuguese and Philippine armed forces became heavily

HOW LONG?

237

politicized during the period of authoritarian rule. In both cases middle-ranking officers formed associations to promote reform and dei ¢y, the Movi das Forcas Armad (MFA) in Portugal and the Reform of the Armed Forces Movement (RAM) in the Philippines. The MFA was the key group in ending the Caclano regime; the RAM would have been the key group in averthrowing Lhe Marcos regime, if il had not been preempled by Aquino’s elecloral victory. Once regime changes occurred in both countries, many of the officers that

had been in the

p osing

the dictatorships also

came to oppose the d g This ity of mililary oppositi ized in lhe Philippines by Col. Gregorio Honasan, aleadu\g RAM member, who led two coups against the Aquino government, and in Portugal by Col. Otelo Saraiva de Carvalho, the intellectual leader of the MFA, who after 1980 was associated with the April 25 Popular Forces - (PF-25), an underground group that waged

a lerrorist

paign

against T

y- In

1983 other officers formed a nonviolent parallel organization, Lhe Association of the 25th of April, to keep alive the original radical and revolutionary goals of the MFA. This organizalion reportedly had substantial support among both aclive and retired military officers. Over time the coup allempts against new democratic regimes became less (requent. One important measure of democratic consolidation (see below, pp. 256f(.) is the turnover of

governmental contro! [rom one party to another as the resuit

of elections. In Greece and Spain, the prospective victories of leftist parties, Lhe Panellinio Sosialistiky Kiniina (PASOK) and

the Socialists, slimulaled coup rumors and, apparently, coup conspiracies. The acquisition ol power in l’orlugal by the con-

Social D

party p

d similar specula-

tion, as the resull of the dominance of Marxist and leflist ideologies among the Porluguese military. By the mid-1y8os, however, all three countries had moved beyond the point

238

‘THE THIRD WAVE

where transition coup attempts were feasible. This did not necessarily mean that coups were (orever impossible in those countries. If the democralic syslem (ailed to produce minimally effective governmen, or il important groups deserted the prodemocracy coalition, coups might again be attempted. They would not, however. be transition coups. Their larget would be democracy, not democratizalion. Dernocratic regimes that [ollowed military regimes Lhat left office voluntarily usually faced a different set of problems. They were confronted nol by possible coups by alienated and discontented officers who opposed democratization, but rather by the continuing power and influence of the military leaders who had made democratizalion pussible. As has been noted, such military leaders in effect defined their terms for giving up office. The problem for the elected leaders of the new democracy was lo reduce the power and privileges of the military i m alevel comp with the funclioning of a constituti y. In ies al lower levels of ic and social devel such as Guatemala and El Salvador, this was difficult if niol impossible to accomplish. Whatever curbing of military influence occurred in El Salvador

was,

in considerable

measure,

a product

of

U.S. government power rather than the power of civilian Presidents Duarte and Alfredo Cristiani. In other countries

where the mililary had been the sponsors of democratizalion,

over lime new governments reduced military privileges, just

a3 in countries where the mililary had been the victims of democratization, over lime new governments reduced m tary coup allempts.

In Turkey, Brazil, Chile, Portugal, Nicaragua, and other countries, powerful military establishments allempled to continue into Lhe pusura.nsmon period powers and prcrogahves

that might be ” (or a | democracy. First, they mmled that special provisions be ll’l' cluded in to the mililary resp

1How LoNG?

239

Lo provide for law and order and nalional securily, “to guaranlee the institutional order of lhe republic” (Chile), or to “preserve Lhe conquests of the revolution” (Portugal). Implicil in Lhese provisions was the possibility thal on Uheir own initialive mlhlary lcaders cou]d intervene in polilics and fake action (incl ing an elecled government) to ensure that these xesponsnh iies were met. Second, actions of the mililary regime were in some cases made irreversible. In Turkey, for instance, the mililary decreed that 631 laws they had enacted during their rule between 1980 and 1983 could not be changed or criticized. The power lo revoke martial law was given 1o the head of the armed forces. In Chile the nuhlary government passed laws igning to the military esl power to buy and sell cquxpmenl and properly withoul securing the approval of the

governiment.

Third, rew governmental bodies dominated by the military were i created. The Portug, ion, for example, provided for a Council of the Revolution, the members of which came from the armed forces, to advise the governmenl and Lo judge Lhe conslilutionality of laws. In Turkey, the National Securily Council that had been the cenler of power in the mililary regime was reconstituled with a membership of retired military officers as an advisory council to the president.* Fourth, top military officers themselves al times assumed

key positions in new democratic governments. In countries

with American-style presidential systems, mililary officers might occupy cabinet positions under a civilian president; in Brazil, six out of twenty-two lo lwenly six :abmol members were military officers. In Y regimes, mililary officers like Fanes in I’orluga] and Evren in Turkey, both of whom had directed the transition Lo democracy, became presidenl and engaged in struggles with civilian

prime ministers over the powers of their respective offices. In

240

THE THIRD WAVE

Chile, the president in the authoritari; t. General Pinochel, slayed on as commander in chief of the army in the democrallc government. In Nicaragua, the delense minister of the Gen. I Ortega, stayed nn as army chlef in the democratic governmenl, and the S i their dom e in and control of the officer corps. Fifth, the military often attempted to guarantee the future autonomy of the armed forces, particularly Lhe independence of Lheir personnel and finances, from control by the elecled civilian government. [n Brazil the military ensured (hat (hey would have full control over promotions. The Chilean mililary decreed that the chiefs of the armed forces and the police could not be removed for seven years, that the newly clecled government could not reduce the size of the military, and that the armed forces could set their own hudgel The Chilean example imp the N guan military. ] ing the models thal might be followed in creating a posl -election balance between their military forces and the newly elected government, Sandinisla leaders and officials speak openly aboul such counlries as Chile—casting themselves in the role of {hat counlry’s mililary dictator, Augusto Pinochet.” The Sandinislas, however, ouldid Pinochet and promulgaled a law (dated before they surrendered power bul possibly writlen aflerward) thal perpetualed and enhanced the power of the Sandinista army. This law gave the commander in chiel of the armed lorces rather than the president the power to ap point{ the new commander; the commander in chief was also

given, among other things, the power to appoint all military

officers. to procure arms, cquipment, and other assets, lo or ganize and deploy the military. to acquire and conslruct lacililies, Lo establish businesses lo meel the needs of the armed [orces, and to prepare the military budget.” Arrangemenls such as these were significant infringements on lhe normal authority of elected governments. Overall, it

HOW LONG?

241

was probably easier for new democratic regimes 1o eslablish their control over rebellious militaries Lhal were weak Lhan over cooperative mililaries that were strong. Yet while militaries that cooperaled in democratization might attempt to prolong their power, lhe record suggesls thal in countries at middle levels of economic developmenl mililary power tended to diminish over Eme. In Portugal, for inslance, the military overthrew (he dictalorship on Lheir own initiative, dominated the government for two years thereafter, and had greal prestige with the Porluguese public. Yet the institutional manifestation of military power, the Council of the Revolution, was abolished in 1982, a Law of National Dofenca was passed subordinating the military to a cabinet responsible to parliament. the powers of the president were curbed, and Mario Soares, the greal opponent of Colonel Eanes, succeeded him as president. Ten years afler the revolution civilmilitary relations in Portugal were “approximating the de sired model of objective civilian control.” ® In Brazil the mililary gave up conlrol of governmenl wilh flags flymg their power and preslige inlacl. Yel Brazil’s first P elected president, Fernando Collor, arrested one leadmg general for making political statements, reprimanded another, Lthe brother of a former military presidenl, for defending thal general, and reduced mililary repre senlation in the cabinet from six to four. In addition, he

reduced Ihe size of the principal intelligence agency, the Servigio Nacional de Informagés {SN), which had always been

headed by a general, and put a divilian in charge. Civilians replaced military officers on the presidential staff and were placed in key positions dealing with nuclear power and the Amazon, hwo issues of great concern to lhe mililary. President Collor also drastically cul the military budget and refused lo grant cost-of-living increases to the mililary, generating considerable grumbling and protesls within the ranks. The disgruntled officers threatened the government, how-

242

THE THIRD WAVE

ever, not with a coup but with a lawsuitl.” In Peru the mililary were also historically a major political (orce; yel on assuming power in July 1990, President Alberlo Fujimori abruptly fired the heads of the navy and air force. In Chile, less than a year afler surrendering the presidency, the redoubtable General Pinochet suffered from allegations of plion by of his family and reportedly considered retiring as army commander. “Pinochet is a cal, not a liger,” observed one leading Chilean politician.”» In Turkey the military have been consisienlly viewed as a revered national inslitution and identilied with the values of the Kemalist secular republic. Yet in 1987, four years afler the military gave up polilical office, a referendum restored politi 16

Rio de Janeiro, Au-

43- International Herald Tritane, May 30, 1”0, p 5, New York Timies, March n, n9az,P Aq; Times (London), March 29,

Danapoulos, “From Balconies (o £y p 91-9:, Salvador Giner, “Sothem European Socialism in Transition,” in New Meditrrranean Democracies, ed. Pridham,p. 151; New York Times, March 31, 1982,p. A,

45. Quoted in New York Times, July 22, 198, p. 3, 46. "On the Edge of Europe: A Survey of Portugal.” Econamist, June 30, P-7

47. New York Timzs, October 28, 1983, p. As, January 24, 1984, p. Az Times (London), March 20, 5990, p. 13. 48. Danopoulos, “From Balconies to Tanks,” p. 93; Zagorski, “Civil-Milury I.elnlmlu, Kos Danopoul

“Erom Balconies o Tanks,” p. 89; Theodore A, Cou-

loumhs and Prodromos M. Yannls, ~The Stability Quotient of Greoce's Past974 Demacratic Institations,” Journal of Modern Greek Studies 1 {October :w;) P. 366; New York Times, Janary 15, 1980, p.6; Econamist, Jan. 21, 1989,

P- 40 50. nopoulos, “From Balconies to Tanks,” pp. 85, 93; New York Times, March 3t opa, 51. New YorkTms, necemhere 1990, p. Aus. 52 Waskinglon Psi. May 5. 1984, . Avy; Washingion Post Nl Weckly, November g, 1987, p. 17; Juan de Onis, “Brazil on the Tightrope Toward Democracy,” Foreign Affairs 63 (Fall 1969). p. 128; New York Times, June s, 1989, p. Ag; Tzvetan Todorov, Posi-Totalitarian Depression.” New Republic, June 25, 1990, pp. 23 -25: New York Times, November g, 1990, p. A1, A10. 53, Peter McDonough, Samuel H. Bames, Antorio Lopes Pina, “The Growth of Democratic Legitimacy ‘Amricun Political Scicnce Review 8o (September 1936), p. 743 New York Times, May 7, 1989, p. 2E; Thomas C Bruneny, “Popular Support for Democracy in Post-revolutionary Portu, Results from & Survey, - i n Soarch of Moders Portugat. The Revelasion and Is Consequences, ed. Lawrcnce 5. Graham and Douglas L. Wheeler (Madison: University of Wisconsin Press, lgfl;) PP- 35-36; McClintock, “Prospects for Democratic Consohdation,” p. 1 ntri

mour Martin Lipart, “Democracy

i

rating Jv Obstructing Factors,” in Freedom in

the World: Political Rnglm rnd Corl Ltics 1987~ 1988, vd. RaymondD. Gastil (New York. Frecdom Hou 986) p- 231 s5. Linz and Stepan, ical Cralting, ” in Demwcracy in the Americas, ed.

Pastor,pp. 46, 58-3. "ind ki, Zimmerman, “Economic and Political Reactions 10 World Economic Crises of the 19305 in Six European Countries™

{Paper presented for convenrion, Midwest Political Scicnce Associalion, Chicago, April 10-c2, 1986). p. 51, quuted in Linz and Stepan, “Pofitical Crafting,” p. 46; Rabert A, Dix, Book Review, Americen Political Science Review 83 (Seplember 1989). p. 1055

NOTES TO PAGES 260-285

»9

56. Linz and Stepan, “Political Cralting.” in Democracy in the Americas, ed. 7. Enrique A. baloyra, “Public Opinion and Support for Democratic Regimnes, Venezuela 1973 -198)” (Paper prepared for Annual Meeting, American Political Science Assaciation, New Orleans, La., August 25— September

Pastor, P- 49 McClintock, “Prospects for Democralic Consolidation,” p. 127.

1, |9Bs) Pp- 10-11. . Makram Haluani, “Waiting for the Revolution: The Relative Deprivationnl the J-Curve Lugic in the Case of Venezuela, 1968-1989 (Paper prepared {or Annual Mceling, American Palilical $cience Association, Allanta,

Georgia, August J1-September 3, \989) pp-9-10. 59. McDonough, Bammes, Lopez Pina, “Democalic Legitimacy in Spain,”

P-753 60. These figures on the economy and public opinion come from Linz and Stepan, “Tolitical Ceaftng,” in Demacracy in the Americas, ed. Pastor, - 43-45PP e 5% Guilermo Orone, “Challenges to Democralization in Drazil,” World Policy Joursal s (Spring 1088), 2P 381300, 2. McDonough, Bames, Lopez. Pina, “Democralic Legitimacy in Spein,” 75: 753 McClintock, *Pmspens fon Democuatic Consolidation.” p. 140. 63, For these and other daia on the in Cerman political culture, see Kendall L. Baker, Ruseel ). Dalon, Kai 1lidcbrande, Germany Trans: formed: Politicat Culture and the New Politics (Cambridge: Harvard University Press, 1981), passimm bute p . 273, 2b7; David P. Conradi, “Changing German Political Culture,” ivic Cullure Revisited, ed. Ga. brielA_ Almond and Sidrey Verba (Boston: Lt Brown, 1o, pp. 213-72, ard “Wes! Germany;: A Kemade Polifical Culture?” Comparaie Pofitcal Studsy(July19741, 238, Balion,Hiddbrandi, Germany Transformed,pp. 69-69, 185; Warten M Touncishi, Japancst Poltiet Style (New York: H-qm 1966), 17 PP b nkcrmational Herald Tribune, May 10, 1990, p. 1, May 16, 1990, p. &, May 21, 1990, p. 2 Times (London), May 11, 1990, p. 1 . See Juan], Linz, “The Perils of Presidentialism,” journal of Democracy 1 (Winler 1990), pp. 51~70, and the follow-up articles by Donald Hurowitz, Seymour Martin Lipsel, and Juan]. Linz, Journai of Democrecy 1 (Fall 1990),

PP-73-91-

Chapter 6: WHITHER?

1. For useful ceports, sce Economist, Seplember m, 1988, pp- 43-44, December 24, 1088, pp. 6166, and August 5, 1989, p. 7 2 Economist, May6, 1989, p. 34, November 11, mflq

Pp. 40-41: Times

(London), Apri) 12, 1990, p. 12; The Observer, May 29, 1990, p. 16. ; Tlms (l.undnn) april 24, 1990, p. 11 Secrelary James Baker, ’Demmmy and Amerienn Diplomacy,” (Addms World Affatrs Council, Dallas,Texas, March 30, 1990), and Ronald s remarks lo the Engluh«pe.]ung Grion, London, quoled in New Yok Times, June 14, 1989, . A

340

NOTES TO PAGES 285-310

5. Stan Sessel “A Rich Country Gone Wrong,” New Yorker, October 9, 1 R e vok Times, December 28, 1989, p. A3; Intermational Herald Tribune, May 12-1), 1990, P, 7. Times (London), May 27, 1990, p. Azt; Time, May a1. 1990, pp. J4~35: Datly Telegraph, M.nd- 29. 1990, p. 13; New York Times, February 27, 1999,

p- Ao, April g 10, p. A 8. Gearge F. Kennan, m Cloud of Danger (Boston: Litile, Brown, 1977). 41-43 PP 2SR William Wallace, The Tllns,’amwlwn of Western Europe (London: Royal [nstitute of International Affairs, Pinter, 1950), pp. 1619, and Michael Howard, “The remaking of Europe,” Suroical 32 (March-April 1990), PP 102-3. 10. Yu-sheng Lin, * The f Contucianism on China’s Search for Political Mnd:mlly in Confuconism and Modernization: A Symposium, ed. Joseph I Wi Nan Publishing Co.. 1987), p. 25. For a sonwwhat different inlerpretation of human rights and the rule of law in the Confucian tradition, see Stephen B. Young and Nguyen e Huy, The Tradition of Human Righls in China and Vietram (New Haven: 5: Cenler for Internalional and Area Studies, Council on Southeast Asia Studies, 1990). They argue that there was a duality of virlue and power in the Confucian tradition, bul they admil thal power became increasingly concentraled in modern timey. 11. See Danlel Kelliher, “The Political Consequencesof China's Reforms,” Comwmlnx Politics 18 (July 1986), pp. 488-90, and Andrew ). Nathan, Chi(New York: Alfred A. Krop!, 1595). e Eeononist April 23, 1938, p. 37, November s, 1982,p. Times. May 20, 1982, p. A3, July 10, 1988, p. Hz; lan Boruis, New York Times Mayezine, Jure n, 2989, p. 26 13. Lucian W. Pye with Mary W, Pye, Asion Potoer and Palitics: The Cultaral Dimensions of Authority (Cumbndge Harvard University Press, 1985), pPp- 232-36. Pt Now York Times, Decerber 5. 1907 p. A Gregory Henderson, Korea; The Dolitics of the Vortex (Cambridge: Harvard Uaversity Press. 1969), P- 365, p-1

15- Ecomomist, January 27, 1990,p. 31; New York Timss, January 23, 1990,

16. Goh Chok Tong, quoted in New Yark Times, August 14, 1995, p. A13.

17. Lucian W. Pye, “Asia 1986—A1| Exceptional Year,” Freadom al kssur 9q

(January - Februaty 1987), p. 1 18. Emest Geliner, “Up from Imperialism,” The New Republi, May 1989, pp. 35- 36: K. Sicphen Humphreys, “lslam and Politial Values n Saui Arabia, Egypi. and Syria,” Mrddlebfl!wmal 33 (Winber 1979), pp. 6-7. 19. New YorkTimes, July 1,

0. For briet

Mahraz lepahani, “The

ion of the many forms of lslamic politics, see

Varisties of Muslim Experlence.” Wilson Quarierly 13

{Autumn 1939), pp. 63-72.

NOTES TO PAGES 312313

41

21. World Bank World Development Report 1990 (New York: Odord University Press, 1900), pp. 8-11, 16, 160, and Sub-Saharan Africa: From Crisis to St G Washington: World Bank, 1990) T2 These ond subsequent figures on per capita GNP and growth rates of GNP and GDP come from Warld Bank, Werld Deoelopmen Report 1990,

pp- 178-81.

INDEX

Aberturn: 22,71, 137,1.

Abortion, Cathalic Chunh and: 282 Abugattas, Ly

Accibn Demucritica (AD): 72

Acxién Poputar: 175

Acrountability, democracy and: 214 AD (Accion Democrdtica): 172 Ad.u-nec ud.slav 100, 156

Aighzn or o}

elections in, 178, 181, 187, 308— 305; election boycolt in, 188; economic healh of, 312, 314; ser also lslamic Salvation Front Alianca Necional Renovadora (ARENA): 1 Allanza Popular Revolucionaria Americana: see APRA Allende, Sahvador: 204 Alliance for Alvarez, Gregurio:

Saharan, 295; economic growth in, 312, 316;

per capita GNP in,

372 abstacles to demucralization in, 315 Atrican Nalianal Congress: see ANC AlD (Agency lnr lnlemalmnal Develdemfl‘flmlmn

2B7; openu\g up” of, 256 Alforsin, Ravl: b7, 107, 147, 278,

Amnesty:

214,

in Chile,

211, 215, 216-17; in Uruguay, 213, 226-17. 129; in Brozil, L in Guatemala, 216, 2 216; nepenledmAlgpnlm 28, W

of,

nesty

International: 96, 212

ANC (African National Congress): 25, 134, 160, 206; unbanning of,

207: and Argentine mililary, 103,

243, 245, 247, 248, 250, 251; and

human rights abuses, 219, 220, 222~24, 229; coups resisted by, 135, 353 Algeria: liberalization in, 25, 287;

15

zonia, anllum ‘military cone for: American Revolution: 16 Americas Walch: 212

enduing authoritarianism of sub-

170; elections

boycolted by, 190,

291, and violence, 19, 202, 203;

and nonviolence, 202; we also Umkhonto we Sizwe

Andropov, Yuri: 130 343

THE THIRD WAVE

344

Angola: 313; U.S. support for rebels in, 94 Anlicommunism: 50; Franco ammy and 243; of Argenline military,

Anhgua and Barbuda: 4

iteracy in, 64; per capita GNP in, @4; middle dlass of, &7; political in, &7; Catholic Church

in, 79-80, 8x; John Paul Il in, 83; lmlmry establishment of, 1 244, 246, 24B. 249; liberalization

horrence of, 154; South African

in, 135; human rights abuses in, 147. 211, 219-21, 222-25, 228political compromise in, 173; in-

Ser wiso Group Areas Aci; pass

Chile, 24

officer corps commitment o, 244 Jaws

Apertura; 134

APRA (Alianza Popular Revolucionana Americana): 19, 140; Peruvian

military vs., 169; accommudationism of, 173, 173; collapse o, 175: Ilaul IA.II of, 266; sce also Alan

April z;5Popular Forces (PE-25): 237 Aquino, Berigno: 23, 187, 201; marder of, 104, 204; nonviolence of,

206 Aquino, Corazon C.: o7, - 85, 107, 2B1, 201, 25, 208;1S, support of, 94; accession of, 148, 237; National Demacratic Fron( vs., 190;

nowviolence of, 206; plans for

Marcos of, 216; and Marcos hu-

man rights abuscs, 229; and military, 243, 250; communist Insurgenls vs., 247 Arabs, liberalization of: vos, 287-88; see also Islam; Middle East Aq, Zehra F.: 321, n. 18 |anca Nacional Renava-

Argenting, democratization cf, B,

17, 40, 103, 104105, 278; authoritarianism in, ' 2 coupsatiempled coups in, 18. 19, 233, 275, 2)6, 291 (sex also Easter

Mion); democracy in, 18- 19, 22; political oscillation In, 41, 146 47 economic furmoil n, 53. 55;

196; Tepression genlina, human rights abuses in);

terim government of. 218; and (ser also

Greal

Falkland lalands and D&

pendencies, war over); chronlc inflation in, 254; elections in,

267; {omifin uence on, 274; rmilitacy delcat of (v Falkdand Islnmh and Dependencies, war uver); see also Alfonsin, Raul;

Bignone, Reynaldo; Falkland Islands and Dependencics, war

over; Menem, Carlos; Perunislas; Sabato, Erneslo

Arias, Carlos: 5. 134

Arigtocracy: 12; democracy and, 37 Ams, Cardinal Paulo Evaristo: 82

melhy C.'\Hen 83, 104, 174,

195, 198, 326

AU

influence in, 293;

in. 285; Chinese aulhy

ism in Enst, 295: coups in, 297 economic growth in, 3x2; obstacles 1o demacratization in Easl, ns

Asians, Souath Alnan 136, 167; in Pa riiament,

Associalion ol’ lhp a5th of April: 237

Atatiirk, Mustafa Komal: yo7; se also Kemalism

Aung San Suu Kyi: 181 Austalia: 39 Austria: Hitler and, 13; democracy in, 18, 40, 43 and greal depression, z