sm huntington 1992

8/3/2019 sm huntington 1992

1/39

How Countries DemocratizeAuthor(s): Samuel P. HuntingtonSource: Political Science Quarterly, Vol. 106, No. 4 (Winter, 1991-1992), pp. 579-616Published by: The Academy of Political ScienceStable URL: http://www.jstor.org/stable/2151795

Accessed: 16/10/2010 06:37

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

http://www.jstor.org/action/showPublisher?publisherCode=aps.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

The Academy of Political Science is collaborating with JSTOR to digitize, preserve and extend access to

Political Science Quarterly.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=apshttp://www.jstor.org/stable/2151795?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=apshttp://www.jstor.org/action/showPublisher?publisherCode=apshttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/2151795?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=aps


2/39

How CountriesDemocratize

SAMUEL P. HUNTINGTONBetween 1974 and 1990 more than thirty countries n southernEurope, Latin America, East Asia, and EasternEurope shifted from authori-tarian o democraticystemsof government.This "globalemocratic evolution"is probably he mostimportantpolitical rend n the late twentiethcentury.It isthe third wave of democratizationn the modernera.A wave of democratizations a group of transitions romnondemocratic odemocratic egimes hat occurs within a specifiedperiodandthat significantly

outnumbers ransitions n the opposite direction n the sameperiod.The firstwavebeganin America n the early nineteenthcenturyand culminatedat theend of WorldWar I with about thirtycountrieshavingdemocraticregimes.Mussolini'smarchon Rome n 1922begana reversewave,andin 1942 herewereonly twelve democracies eft in the world. The Allied victory in WorldWarII and decolonizationstarted a second movement towarddemocracywhich,however, peteredout by the early 1960s when about thirty-sixcountrieshaddemocratic egimes.This wasthen followedby a second reversemovement o-wards authoritarianism,marked most dramaticallyby militarytake-overs nLatinAmericaand the seizureof powerby personaldespotssuch as FerdinandMarcos.The causesof the thirdwave,like those of its predecessors,werecomplexandpeculiar o thatwave.Thisarticle,however, s concernednotwiththewhyof thethirdwave but ratherwith the questionof how third wave democratizations

SAMUELP. HUNTINGTONs Eaton Professorof the Scienceof Governmentnddirector f theJohn M. Olin Institute or StrategicStudies at HarvardUniversity,and recent presidentof theAmericanPoliticalScienceAssociation.He has publishednumerous ooks and articleson the pro-cessesand problemsof democracies.This article s drawn romhis recentbook, The ThirdWave:Democratization in the Late Twentieth Century, published by the University of Oklahoma Press.Political Science Quarterly Volume 106 Number 4 1991-92 579


3/39

580 1POLITICAL SCIENCE QUARTERLYoccurred:he waysin whichpolitical eadersandpublics n the 1970sand 1980sendedauthoritarianystemsand createddemocratic nes. The routes of changewerediverse,aswere hepeopleprimarilyesponsibleor bringingaboutchange.Moreover,the startingand endingpoints of the processeswere asymmetric.Obviousdifferences xistamongdemocratic egimes: omearepresidential,omeare parliamentary,omeembodythe Gaullistmixtureof the two; so also somearetwo-party, omearemultiparty, ndmajordifferences xist nthenatureandstrengthof the parties.Thesedifferenceshavesignificanceorthe stabilityof thedemocratic ystems hatare created,butrelativelyittlefor theprocesseseadingto them.' Of greater mportances that in all democratic egimes he principalofficers of governmentare chosenthrough competitiveelectionsin whichthebulk of the populationcanparticipate.Democratic ystems hushavea commoninstitutionalcorethat establishes heiridentity.Authoritarian egimesas theterm s usedinthisstudy- aredefinedsimplyby the absenceof thisinstitutionalcore.Apartfromnot beingdemocratic heymayhavelittle else in common.Itwill, consequently,be necessaryo start he discussionof change n authoritarianregimesby identifying he differencesamongthoseregimesand the significanceof those differences or democratization rocesses.

AUTHORITARIAN REGIMESHistorically,nondemocratic egimeshave taken a wide varietyof forms. Theregimesdemocratizedn the firstwavewere generallyabsolutemonarchies, in-gering eudalaristocracies, ndthesuccessor tates o continental mpires.Thosedemocratizedn the secondwavehadbeenfasciststates,colonies,andpersonal-istic militarydictatorships ndoften had had some previousdemocratic xperi-ence.Theregimes hat movedto and towarddemocracyn the thirdwavegener-ally fell into threegroups:one-partysystems,militaryregimes,and personaldictatorships.

The one-partysystemswere createdby revolutionor Soviet impositionandincludedthe communistcountriesplus Taiwanand Mexico (withTurkeyalsofittingthismodelbeforeits secondwave democratizationn the 1940s).Inthesesystems,the partyeffectivelymonopolizedpower,accessto powerwasthroughthepartyorganization,andtheparty egitimatedts rule hrough deology.Thesesystemsoften achieveda relativelyhighlevelof politicalinstitutionalization.The militaryregimeswere created by coups d'etatreplacingdemocraticorciviliangovernments. n them, the militaryexercisedpoweron an institutionalbasis,withthemilitaryeaders ypicallyeithergoverning ollegiallyas a juntaorcirculatinghe top governmental ositionamong top generals.Militaryregimesexisted in large numbersin Latin America (where some approximated he

' See G. Bingham Powell, Jr., ContemporaryDemocracies: Participation, Stability, and Violence(Cambridge, MA: Harvard University Press, 1982), chaps. 5-9; Juan J. Linz, "Perils of Presiden-tialism," Journal of Democracy 1 (Winter 1990): 51-69.


4/39

HOW COUNTRIES DEMOCRATIZE | 581

bureaucratic-authoritarianmodel) and also in Greece, Turkey, Pakistan, Nigeria,and South Korea.Personal dictatorships were a third, more diverse group of nondemocraticsystems. The distinguishing characteristic of a personal dictatorship is that theindividual leader is the source of authority and that power depends on access to,closeness to, dependence on, and support from the leader. This category includedPortugal under Antonio Salazar and Marcello Caetano, Spain under FranciscoFranco, the Philippines under Ferdinand Marcos, India under Indira Ghandi,and Romania under Nicolae Ceausescu. Personal dictatorships had varied ori-gins. Those in the Philippines and India were the result of executive coups. Thosein Portugal and Spain began with military coups (which in the latter case led tocivil war)with the dictators subsequently establishing bases of power independentof the military. In Romania, a personal dictatorship evolved out of a one-partysystem. Chile under Augusto Pinochet originated as a military regime but ineffect became a personal dictatorship due to his prolonged tenure and his differ-ences with and dominance over the leaders of the military services. Some personaldictatorships, such as those of Marcos and Ceausescu, like those of AnastasioSomoza, Frangois Duvalier, Sese Seko Mobutu, and the shah, exemplified We-ber's model of sultanistic regimes characterized by patronage, nepotism, cro-nyism, and corruption.One-party systems, military regimes, and personal dictatorships suppressedboth competition and participation. The South African system differed fromthese in that it was basically a racial oligarchy with more than 70 percent of thepopulation excluded from politics but with fairly intense political competitionoccurring within the governing white community. Historical experience suggeststhat democratization proceeds more easily if competition expands before partici-pation.2 If this is the case, the prospects for successful democratization weregreater in South Africa than in countries with the other types of authoritariansystems. The process in South Africa would, in some measure, resemble thenineteenth-century democratizations in Europe in which the central feature wasthe expansion of the suffrage and the establishment of a more inclusive polity.In those cases exclusion had been based on economic, not racial, grounds. Hierar-chical communal systems, however, historically have been highly resistant topeaceful change.3 Competition within the oligarchythus favored successful SouthAfrican democratization; the racial definition of that oligarchy createdproblemsfor democratization.Particular regimes did not always fit neatly into particular categories. In the

2 Robert A. Dahl, Polyarchy:Participation nd OppositionNew Haven: Yale University Press,1971), 33-40.

3See Donald L. Horowitz, "Three Dimensions of Ethnic Politics," World Politics 23 (January1971): 232-36; Samuel P. Huntington and Jorge I. Dominguez, "Political Development" in Fred I.Greenstein and Nelson W. Polsby, eds., Handbook of Political Science, vol. 3 (Reading, MA:Addison-Wesley, 1975), 74-75.


5/39

582 | POLITICAL SCIENCE QUARTERLY

TABLE1Authoritarian egimes and Liberalization/Democratizationrocesses, 1974-90

RegimesProcesses One-Party Personal Military Racial OligarchyTransformation (Taiwan)a Spain Turkey

Hungary India Brazil(Mexico) Chile Peru(USSR) EcuadorBulgaria Guatemala

Nigeria*PakistanSudan*

16 5 3 8Transplacement Poland (Nepal) Uruguay (South Africa)

Czechoslovakia BoliviaNicaragua HondurasMongolia El Salvador

Korea11 4 1 5 1

Replacement East Germany Portugal GreecePhilippines ArgentinaRomania

6 1 3 2Intervention Grenada (Panama)

2 1 1Totals

35 11 7 16 1Note: The principal criterion of democratization is selection of a government through an open, competitive,

fully participatory, fairly administered election.a Parentheses indicate a country that significantly liberalized but did not democratize by 1990.* Indicates a country that reverted to authoritarianism.

early 1980s, for instance, Poland combined elements of a decaying one-partysystem and of a military-based martial law system led by a military officer whowas also secretary general of the Communist party. The communist system inRomania (like its counterpart in North Korea) started out as a one-party systembut by the 1980shad evolved into a sultanistic personal dictatorship. The Chileanregime between 1973 and 1989 was in part a military regime but also, in contrastto other South American military regimes, during its entire existence had onlyone leader who developed other sources of power. Hence it had many of thecharacteristics of a personal dictatorship. The Noriega dictatorship in Panama,on the other hand, was highly personalized but dependent almost entirely onmilitary power. The categorizations in Table 1, consequently, should be viewedas rough approximations. Where a regime combined elements of two types it is


6/39

HOW COUNTRIESDEMOCRATIZE | 583

categorized in terms of what seemed to be its dominant type as the transition gotunderway.In the second wave, democratization occurred in large measure through foreignimposition and decolonization. In the third wave, as we have seen, those twoprocesses were less significant, limited before 1990 to Grenada, Panama, andseveral relatively small former British colonies also mostly in the Caribbeanarea.While external influences often were significant causes of third wave democratiza-tions, the processes themselves were overwhelmingly indigenous. These processescan be located along a continuum in terms of the relative importance of governingand opposition groups as the sources of democratization. For analytical purposesit is useful to group the cases into three broad types of processes. Transformation(or, in Juan J. Linz's phrase, reforma) occurred when the elites in power tookthe lead in bringing about democracy. Replacement (Linz's ruptura) occurredwhen opposition groups took the lead in bringing about democracy, and theauthoritarian regime collapsed or was overthrown. What might be termed trans-placement or "ruptforma"occurred when democratization resulted largely fromjoint action by government and opposition groups.4 In virtually all cases groupsboth in power and out of power played some roles, and these categories simplydistinguish the relative importance of government and opposition.

As with regime types, historical cases of regime change did not necessarily fitneatly into theoretical categories. Almost all transitions, not just transplace-ments, involved some negotiation - explicit or implicit, overt or covert - betweengovernment and opposition groups. At times transitions began as one type andthen became another. In the early 1980s, for instance, P. W. Botha appeared tobe initiating a process of transformation in the South African political system,but he stopped short of democratizing it. Confronting a different political envi-ronment, his successor, F. W. de Klerk, shifted to a transplacement process ofnegotiation with the principal opposition group. Similarly, scholars agree thatthe Braziliangovernment initiated and controlled the transition process for manyyears. Some argue that it lost control over that process as a result of popularmobilization and strikesin 1979-1980; others, however, point to the government's

4For reasons that are undoubtedly deeply rooted in human nature, scholars often have the sameideas but prefer to use different words for those ideas. My tripartite division of transition processescoincides with that of Donald Share and Scott Mainwaring, but we have our own names for thoseprocesses:

Huntington Linz Share/Mainwaring(1) transformation = reforma = transaction(2) replacement = ruptura = breakdown/collapse(3) transplacement = - - extrication

See Juan J. Linz, "Crisis, Breakdown, and Reequilibration"in Juan J. Linz and Alfred Stepan, eds.,The Breakdown of Democratic Regimes (Baltimore: Johns Hopkins University Press, 1978), 35;Donald Share and Scott Mainwaring, "Transitions Through Transaction: Democratization in Braziland Spain" in Wayne A. Selcher, ed., Political Liberalization in Brazil: Dynamics, Dilemmas, andFuture Prospects (Boulder, CO: Westview Press, 1986), 177-79.


7/39

584 1POLITICALSCIENCE QUARTERLY

success nresisting trongoppositiondemands ordirectelectionof the presidentin the mid-1980s.Everyhistorical ase combinedelementsof two or moretransi-tion processes.Virtuallyeveryhistoricalcase, however,more clearlyapproxi-mated one type of processthan others.How did the nature of the authoritarian egimerelateto the natureof thetransitionprocess?As Table 1 suggests, therewas no one-to-onerelation.Yetthe formerdid have consequences or the latter. With threeexceptions,all thetransitions rommilitary egimesnvolved ransformation r transplacement.nthe three exceptions Argentina,Greece,and Panama militaryregimessuf-fered militarydefeats and collapsedas a result. Elsewheremilitaryrulers ookthe lead, at timesin responseto oppositionand popularpressure, n bringingabout thechange n regime.Militaryrulerswerebetterplacedto terminate heirregimes hanwere eadersof other regimes.The military eadersvirtuallyneverdefinedthemselvesas the permanent ulersof their country.They lield out theexpectationhat oncetheyhad correctedheevils hathad ed themto seizepowerthey wouldexit frompowerand return o their normalmilitary unctions.Themilitaryhad a permanentnstitutional oleotherthan politicsand governorship.At somepoint,consequently,hemilitaryeaders other han thoseinArgentina,Greece,andPanama)decided hatthe timehadcometo initiateareturno civiliandemocraticruleor to negotiatetheir withdrawal rom powerwith oppositiongroups.Almost always hisoccurredafter therehad beenat least one change nthe top leadershipof the militaryregime.5Militaryeadersalmost nvariablypositedtwo conditionsor "exitguarantees"for theirwithdrawalrom power. First, therewouldbe no prosecution,punish-ment, or other retaliationagainst militaryofficers for any acts they may havecommittedwhen they werein power. Second, the institutionalroles and au-tonomyof the militaryestablishmentwould be respected, ncluding ts overallresponsibilityor nationalsecurity, ts leadershipof the governmentministriesconcernedwith security,andoften its controlof arms ndustriesand other eco-nomic enterprises raditionallyunder military aegis. The abilityof the with-drawingmilitary o secureagreement f civilianpolitical eaders o these condi-tions dependedon theirrelativepower.In Brazil, Peru, and other instancesoftransformation,hemilitaryeadersdominated heprocessandcivilianpoliticalleadershad little choicebut to acquiesce o the demandsof the military.Whererelativepowerwas moreequal,asinUruguay,negotiationsedto somemodifica-tionsin themilitarydemands.Greekand Argentineanmilitaryeadersaskedforthesameassurances ther eadersdid. Theirrequests,however,wererejectedoutof hand by civilianleaders,and they had to agreeto a virtual unconditionalsurrender f power.6

I See Martin C. Needler, "The MilitaryWithdrawal from Power in South America,"ArmedForcesand Society 6 (Summer 1980): 621-23.

6 For discussion of the terms under which military rulers arranged their exits from power, seeRobert H. Dix, "The Breakdown of Authoritarian Regimes," Western Political Quarterly 35 (De-


8/39


It was thus relativelyeasy for militaryrulers o withdraw rom power and toresumeprofessionalmilitaryroles. The otherside of the coin, however, s thatit could also be relativelyeasyfor them to return o powerwhenexigenciesandtheir own interestswarranted.Onesuccessfulmilitarycoupin a countrymakesit impossiblefor political and military eadersto overlook the possibilityof asecond. The thirdwavedemocracieshatsucceededmilitaryregimesstarted ifeunder this shadow.Transformationand transplacement lso characterized he transitions romone-party ystems o democracyhrough1989,except orthoseinEast Germanyand Grenada.One-party egimeshad aninstitutional ramework nd ideologicallegitimacy hat differentiatedhemfromboth democraticandmilitaryregimes.They also had an assumptionof permanencehat distinguishedhem from mili-tary regimes.The distinctivecharacteristic f one-partysystemswas the closeinterweaving f partyand state.Thiscreated wo sets of problems, nstitutionaland ideological,in the transition o democracy.The institutionalproblems were most severewith Leninistparty states. InTaiwanas in communistcountries he "separation f the party from the state"was "thebiggestchallengeof a Leninistparty"ntheprocessof democratization.7In Hungary,Czechoslovakia,Poland, andEast Germanyconstitutionalprovi-sions for "theleadingrole" of the communistpartyhad to be abrogated.InTaiwancomparable"temporary rovisions"added to the constitution n 1950weresimilarlychallenged.In all Leninistpartysystemsmajorissues arose con-cerningownershipof physicalandfinancialassets didtheybelongto thepartyor the state?Theproperdispositionof thoseassetswas alsoin question shouldtheybe retainedbythe party,nationalizedbythegovernment, oldbytheparty

cember 1982): 567-68, for "exitguarantees";Myron Weiner, "EmpiricalDemocratic Theory and theTransition from Authoritarianism to Democracy," PS 20 (Fall 1987): 864-65; Enrique A. Baloyra,"Conclusion: Toward a Framework for the Study of Democratic Consolidation" in Enrique A.Baloyra, d., ComparingNew Democracies:TransitionndConsolidationn Mediterranean uropeand the Southern Cone (Boulder, CO: Westview Press, 1987), 299-300; Alfred Stepan, RethinkingMilitary Politics: Brazil and the Southern Cone (Princeton, NJ: Princeton University Press, 1988),64-65; Philip Mauceri, "Nine Cases of Transitions and Consolidations" in Robert A. Pastor, ed.,Democracy in theAmericas: Stopping the Pendulum (New York: Holmes and Meier, 1989), 225, 229;Luis A. Abugattas, "Populismand After: The Peruvian Experience" n JamesM. Malloy and MitchellA. Seligson,eds., AuthoritariansndDemocrats:RegimeTransitionn LatinAmerica Pittsburgh:University of Pittsburgh Press, 1987), 137-39; Aldo C. Vacs, "AuthoritarianBreakdown and Rede-mocratization in Argentina" in Malloy and Seligson, eds., Authoritarians and Democrats, 30-31;P. Nikiforos Diamandouros, "Transition to, and Consolidation of, Democratic Politics in Greece,1974-83: A Tentative Assessment" in Geoffrey Pridham, ed., The New Mediterranean Democracies:Regime Transition in Spain, Greece, and Portugal (London: Frank Cass, 1984), 54; Harry J. Psomi-ades, "Greece:From the Colonels' Rule to Democracy" in John H. Herz, ed., From Dictatorship toDemocracy:Copingwith the Legaciesof Authoritarianismnd TotalitarianismWestport,CT:Greenwood Press, 1982), 253-54.

7 Tun-jen Cheng, "Democratizingthe Quasi-Leninist Regime in Taiwan," World Politics 41 (July1989): 496.


9/39


to the highest bidder, or distributed in some equitable manner among social andpolitical groups? In Nicaragua, for instance, after losing the election in February1990, the Sandinista government apparently moved quickly "to transfer largeamounts of Government property to Sandinista hands." "They are selling housesto themselves, selling vehicles to themselves," alleged one anti-Sandinista busi-nessman.8 Similar allegations were made about the disposal of government prop-erty to the Communist party as Solidarity was about to take over the governmentin Poland. (In a parallel move in Chile, the Pinochet government as it went outof power transferredto the military establishment property and records that hadbelonged to other government agencies.)

In some countries party militias had to be disbanded or brought under govern-ment control, and in almost all one-party states the regular armed forces had tobe depoliticized. In Poland, as in most communist countries, for instance, allarmy officers had to be members of the Communist party; in 1989, however,Polish army officers lobbied parliament to prohibit officers from being membersof any political party.9In Nicaragua the Sandinista People's Army had been thearmy of the movement, became also the army of the state, and then had to beconverted into being only the latter. The question of whether party cells withineconomic enterprises should continue was also a highly controversial issue. Fi-nally, where the single party remained in power, there was the question of therelation between its leaders in government and the top party bodies such as thePolitburo and the central committee. In the Leninist state the latter dictatedpolicyto the former. Yet this relationship was hardly compatible with the supremacyofelected parliamentary bodies and responsible cabinets in a democratic state.The other distinctive set of problems was ideological. In one-party systems,the ideology of the party defined the identity of the state. Hence opposition tothe party amounted to treason to the state. To legitimize opposition to the partyit was necessary to establish some other identity for the state. This problemmanifested itself in three contexts. First, in Poland, Hungary, Czechoslovakia,Romania, and Bulgaria, communist ideology and rule had been imposed by theSoviet Union. The ideology was not essential to defining the identity of thecountry. In fact, in at least three of these countries nationalism opposed commu-nism. When the communist parties in these countries gave up their claim toundisputed rule based on that ideology, the countries redefined themselves from"people's republics" to "republics" and reestablished nationalism rather thanideology as the basis of the state. These changes hence occurred relatively easily.

Second, severalone-party systems where democratization became an issue hadbeen createdby national revolutions. In these cases - China, Mexico, Nicaragua,and Turkey-the nature and purpose of the state were defined by the ideologyof the party. In China the regime staunchly adheredto its ideology and identified

8 New York Times, 9 March 1990; 11 March 1990.9 Bronislaw Geremek, "Postcommuninism and Democracy in Poland," Washington Quarterly13

(Summer 1990): 129.


10/39


democratic opposition to communism with treason to the state. In Turkey, thegovernment followed an uncertain and ambivalent policy toward Islamic groupschallenging the secular basis of the Kemalist state. In Mexico the leadership ofthe Partido Revolucionario Institucional (PRI) held somewhat comparable viewsconcerning the liberal challenge of the opposition Partido Accion Nacional(PAN) to the revolutionary, socialist, corporatist character of the PRI state. InNicaragua Sandinista ideology was the basis of not just the program of a partybut also of the legitimacy of the state created by the Nicaraguan revolution.Third, in some instances the ideology of the single party defined both the natureof the state and its geographical scope. In Yugoslavia and the Soviet Unioncommunist ideology provided the ideological legitimacy for multinational states.If the ideology were rejected, the basis for the state would disappear and eachnationality could legitimately claim its own state. In East Germany communismprovided the ideological basis for a separate state; when the ideology was aban-doned, the rationale for an East German state disappeared. The ideology of theKuomintang (KMT) defined the government on Taiwan as the government ofChina, and the regime saw opposition elements supporting an independentTaiwan as subversive. The problem here was less serious than in the other threecases because the ideology legitimated an aspiration rather than a reality. TheKMT government functioned in fact as the highly successful government ofTaiwan even though in its own eyes its legitimacy depended on the myth that itwas the rightful government of all China.

When the military give up their control of government, they do not also giveup their control of the instruments of violence with which they could resumecontrol of government. Democratization of a one-party system, however, meansthat the monopolistic party places at risk its control of government and becomesone more party competing in a multiparty system. In this sense its separationfrom power is less complete than it is for the military when they withdraw. Theparty remains a political actor. Defeated in the 1990 election, the Sandinistascould hope "tofight again another day"and come back to power through electoralmeans.10In Bulgaria and Romania former communist parties won elections; inother East European countries they had less sanguine expectations of partici-pating in a coalition government sometime in the future.After democratization a former monopolistic partyis in no betterposition thanany other political group to reinstate an authoritariansystem. The party gives upits monopoly of power but not the opportunity to compete for power by demo-cratic means. When they return to the barracks, the military give up both, butthey also retain the capacity to reacquire power by nondemocratic means. Thetransition from a one-party system to democracy, consequently, is likely to bemore difficult than the transition from a military regime to democracy, but it isalso likely to be more permanent.1"The difficulties of transforming one-party

10 New York Times, 11 March 1990."For a similar conclusion, see I. William Zartman, "Transition to Democracy from Single-Party


11/39

588 | POLITICALSCIENCE QUARTERLY

systemsareperhapsreflected in the fact that as of 1990the leadersof such regimesin Taiwan, Mexico, and the Soviet Union had initiated the liberalization of theirregimes but were moving only slowly toward full democratization.

The leadersof personal dictatorships were less likely than those of militaryandone-party regimes to give up power voluntarily. Personal dictators in countriesthat transited to democracy as well as those that did not usually tried to remainin office as long as they could. This often created tensions between a narrowlybased political system and an increasingly complex and modern economy andsociety.12It also led on occasion to the violent overthrow of the dictator, ashappened in Cuba, Nicaragua, Haiti, and Iran, and the dictator's replacementby another authoritarianregime. In the third wave of democratization, uprisingssimilarly overthrew personal dictatorships in Portugal, the Philippines, and Ro-mania. In Spain, the dictator died and his successors led a classic case of demo-cratic transformation from above. In India and in Chile, the leaders in powersubmitted themselves to elections in the apparent but mistaken belief that thevoters would confirm them in office. When this did not happen, they, unlikeMarcos and Manuel Noriega, accepted the electoral verdict. In the cases of sul-tanistic regimes, the transitions to democracy were complicated by the weaknessof political partiesand other institutions. Transitionsto democracy from personaldictatorship thus occurred when the founding dictator died and his successorsdecided on democratization, when the dictator was overthrown, and when he orshe miscalculated the support that the dictator could win in an election.

TRANSITION PROCESSESThe third wave transitions were complex political processes involving a varietyof groups struggling for power and for and against democracy and other goals.In terms of their attitudes toward democratization, the crucial participantsin theprocesses were the standpatters, liberal reformers, and democratic reformers inthe governing coalition, and democratic moderates and revolutionary extremistsin the opposition. In noncommunist authoritarian systems, the standpatterswithin the government werenormally perceivedas right-wing, fascist, and nation-alist. The opponents of democratization in the opposition were normally left-wing, revolutionary, and Marxist-Leninist. Supporters of democracy in bothgovernment and opposition could be conceived as occupying middle positions onthe left-right continuum. In communist systems left and right were less clear.Standpatters were normally thought of as Stalinist or Brezhnevite. Within the

Regimes: Lessons from North Africa" (Paper presented to Annual Meeting, American PoliticalScience Association, Atlanta, 31 August-3 September 1989), 2-4.

12 See RichardK. Betts and Samuel P. Huntington, "Dead Dictators and Rioting Mobs: Does theDemise of Authoritarian Rulers Lead to Political Instability?" International Security 10 (Winter1985-86): 112-46.


12/39


FIGURE1PoliticalGroups Involved n Democratization

Attitudes Toward DemocracyAgainst For Against

ReformersGovernment Democratizers Liberals Standpatters

Radical DemocraticOpposition Extremists Moderates

opposition,the extremistopponentsof democracywerenot revolutionaryeft-wingersbut often nationalistgroups thoughtof as right-wing.Within hegoverning oalitionsome groupsoften cameto favor democratiza-tion, while othersopposed t, and otherssupported imitedreformor liberaliza-tion (see Figure 1). Opposition- ttitudestoward democracywere also usuallydivided.Supporters f theexistingdictatorship lwaysopposeddemocracy; p-ponentsof theexistingdictatorship ften opposeddemocracy.Almost nvariably,however, they used the rhetoric of democracy n their efforts to replacetheexistingauthoritarianegimewith one of their own. Thegroups nvolved n thepolitics of democratizationhus had both conflicting and common objectives.Reformersand standpatters ividedoverliberalization nddemocratization utpresumablyhad a common interestin constraining he power of oppositiongroups. Moderatesand radicalshad a commoninterest n bringingdown theexisting regimeand gettinginto powerbut disagreedabout what sort of newregimeshouldbe created.Reformersandmoderateshad a commoninterest ncreatingdemocracybut often dividedover howthecostsof creatingt shouldbeborneandhowpowerwithin t shouldbe apportioned.Standpatters ndradicalsweretotallyopposedon the issue of who shouldrule buthad a common nterestin weakening he democratic roups nthecenterandin polarizingpolitics n thesociety.The attitudesandgoalsof particularndividualsandgroupsat timeschangedin the democratization rocess.If democratization id not produce he dangersthey feared, peoplewho had been liberalreformersor evenstandpattersmightcome to acceptdemocracy.Similarly,participationntheprocessesof democrati-zation could lead membersof extremistopposition groupsto moderatetheirrevolutionarypropensitiesandacceptthe constraintsandopportunitiesdemoc-racyoffered.

The relativepowerof the groupsshapedthe natureof the democratizationprocessand often changedduringthat process.If standpattersdominated hegovernmentand extremistshe opposition,democratizationwasimpossible,as,for example,where a right-wingpersonaldictator determined o hang on topowerconfrontedan oppositiondominatedby Marxist-Leninists. ransition odemocracywas,of course,facilitatedf prodemocraticroupsweredominant n


13/39

590 | POLITICALSCIENCE QUARTERLY

both thegovernment ndopposition.Thedifferencesnpowerbetween eformersandmoderates,however, hapedhowthe processoccurred. n1976,forinstance,the Spanishoppositionurgeda complete "democraticbreak"or rupturawiththe Franco legacyand creationof a provisionalgovernmentand a constituentassembly o formulatea newconstitutionalorder.Adolfo Suarezwas powerfulenough,however, o fend this off and producedemocratizationworking hroughthe Francoconstitutionalmechanism."3If democratic roupswere strong n theoppositionbut not in the government,democratization ependedon eventsun-dermining he governmentand bringing he opposition o power.If democraticgroupsweredominant n the governing oalition, butnot in the opposition, theeffort at democratization ould be threatenedby insurgentviolence and by abacklashincrease n power of standpattergroupspossibly leading to a coupd'etat.The threecrucialnteractionsn democratization rocesseswere hose betweengovernment ndopposition,between eformers ndstandpattersnthe governingcoalition,andbetweenmoderatesandextremistsn the opposition.In all transi-tions these threecentral nteractionsplayedsome role. The relative mportanceandthe conflictualorcooperative haracter f these nteractions, owever,variedwith theoverallnatureof the transitionprocess.Intransformations,he interac-tion betweenreformersand standpatterswithin the governingcoalition was ofcentral importance;and the transformationonly occurred f reformerswerestronger hanstandpatters,f the governmentwas stronger han the opposition,and if the moderateswerestronger hanthe extremists.As the transformationwenton, oppositionmoderateswereoften coopted into the governing oalitionwhilestandpatter roupsopposingdemocratization efected rom it. Inreplace-ments, theinteractions etweengovernment ndoppositionand betweenmoder-atesandextremistswere mportant;heoppositioneventuallyhad to be strongerthan thegovernment,andthemoderateshad to be stronger hanthe extremists.A successivedefectionof groupsoften led to the downfallof the regimeandinauguration f the democratic ystem.In transplacements,he central nterac-tion was betweenreformersand moderatesnot widelyunequal n power,witheach being able to dominatethe antidemocratic roupson its side of the linebetween he governmentand the opposition.In sometransplacements, overn-ment and formeropposition groups agreedon at least a temporary haringofpower.

TRANSFORMATIONSIn transformationshosein power n the authoritarianegime ake the leadandplay the decisive role in endingthat regimeand changing t into a democratic

13 See Raymond Carr, "Introduction: The Spanish Transition to Democracyin Historical Perspec-tive" in Robert P. Clark and Michael H. Haltzel, eds., Spain in the 1980s: TheDemocratic Transitionand a New International Role (Cambridge, MA: Ballinger, 1987), 3-4.


14/39


system. The line between transformations and transplacements is fuzzy, andsome cases might be legitimately classified in either category. Overall, however,transformations accounted for approximately sixteenout of thirty-five thirdwavetransitions that had occurred or that appeared to be underway by the end of the1980s. These sixteen cases of liberalization or democratization included changesfrom five one-party systems, three personal dictatorships, and eight militaryregimes. Transformation requires the government to be stronger than the opposi-tion. Consequently, it occurredin well-establishedmilitary regimes wheregovern-ments clearly controlled the ultimate means of coercion vis-a-vis the oppositionand/or vis-a-vis authoritarian systems that had been successful economically,such as Spain, Brazil, Taiwan, Mexico, and, compared to other communist states,Hungary. The leaders of these countries had the power to move their countriestoward democracy if they wanted to. In every case the opposition was, at leastat the beginning of the process, markedly weakerthan the government. In Brazil,for example, as Alfred Stepan points out, when "liberalizationbegan, there wasno significant political opposition, no economic crisis, and no collapse of thecoercive apparatus due to defeat in war."'14n Brazil and elsewhere the peoplebest situated to end the authoritarian regime were the leaders of the regime- andthey did.The prototypical cases of transformation were Spain, Brazil, and, among com-munist regimes, Hungary. The most important case, if it materializes, will be theSoviet Union. The Brazilian transition was "liberation from above" or "regime-initiated liberalization." In Spain "it was a question of reformist elements associ-ated with the incumbent dictatorship, initiating processes of political changefrom within the established regime.""5The two transitions differed significantly,however, in their duration. In Spain in less than three and a half years after thedeath of Franco, a democratizing prime minister had replaced a liberalizing one,the Franco legislature had voted the end of the regime, political reform had beenendorsed in a referendum, political parties (includingthe Communist party) werelegalized, a new assembly was elected, a democratic constitution was draftedand approved in a referendum, the major political actors reached agreement oneconomic policy, and parliamentaryelections were held under the new constitu-tion. Adolfo Suairez eportedly told his cabinet that "his strategy would be basedon speed. He would keep ahead of the game by introducing specific measuresfaster than the continuistas of the Francoist establishment could respond tothem." While the reforms were compressed within a short period of time, how-ever, they were also undertaken sequentially. Hence, it has also been arguedthat"By staggeringthe reforms, Suarezavoided antagonizing too many sectors of the

14 Alfred Stepan, "Introduction," in Stepan, ed., Democratizing Brazil: Problems of Transitionand Consolidation (New York: Oxford University Press, 1989), ix.

15 Ibid.; Scott Mainwaring, "The Transition to Democracy in Brazil," Journal of InteramericanStudiesand WorldAffairs 28 (Spring 1986): 149; Kenneth Medhurst, "Spain'sEvolutionary Pathwayfrom Dictatorship to Democracy in Pridham, ed., New Mediterranean Democracies, 30.


15/39


franquist regime simultaneously. The last set of democratic reforms provokedopen hostility from the military and other franquist hardliners, but the President[SuairezJhad greatly gained considerable momentum and support." In effect,then, Suairezfollowed a highly compressed version of the Kemalist "Fabianstrategy, blitzkrieg tactics" pattern of reform.16In Brazil, in contrast, President Ernesto Geisel determined that political changewas to be "gradual, slow, and sure." The process began at the end of the Mediciadministration in 1973, continued through the Geisel and Figueiredo administra-tions, jumped forward with the installation of a civilian president in 1985, andculminated in the adoption of a new constitution in 1988and the popular electionof a president in 1989. The regime-decreed movements toward democratizationwere interspersed with actions taken to reassure hardliners in the military andelsewhere. In effect, Presidents Geisel and Figueiredo followed a two-steps for-ward, one-step backward policy. The result was a creeping democratization inwhich the control of the government over the process was never seriously chal-lenged. In 1973 Brazil had a repressive military dictatorship; in 1989 it was afull-scale democracy. It is customary to date the arrival of democracy in Brazilin January 1985, when the electoral college chose a civilian president. In fact,however, there was no clear break; the genius of the Brazilian transformation isthat it is virtually impossible to say at what point Brazil stopped being a dictator-ship and became a democracy.Spain and Brazil were the prototypical cases of change from above, and theSpanish case in particularbecame the model for subsequent democratizations inLatin America and Eastern Europe. In 1988 and 1989, for instance, Hungarianleaders consulted extensively with Spanish leaders on how to introduce democ-racy, and in April 1989a Spanish delegation went to Budapest to offer advice. Sixmonths later one commentator pointed to the similarities in the two transitions:

The astyearsof theKadar radidbear omeresemblanceothebenignauthoritarianismof Franco'sdecayingdictatorship. mrePozsgayplaysthepartof PrinceJuan Carlosin this comparison.He is a reassuring ymbolof continuity n the midstof radicalchange.Liberal-mindedconomicexpertswith links to the old establishment nd thenewentrepreneuriallassprovidea technocratic lite for the transition,much as thenewbourgeois litesassociatedwithOpusDei didin Spain.Theoppositionpartiesalsofigure n thisanalogy,emergingromundergroundn muchthesamewaytheSpanishexilesdid once twas safe to come out. And asinSpain, heHungarian ppositionistsmoderate n style, radicallydemocraticn substanceare playinga vital role in thereinventionof democracy.1716 Paul Preston, The Triumph of Democracy in Spain (London:Methuen, 1986), 93; Donald Shareand Scott Mainwaring, "Transitions Through Transaction: Democratization in Brazil and Spain" in

Wayne A. Selcher, ed., PoliticalLiberalization in Brazil: Dynamics,Dilemmas, and Future Prospects(Boulder, CO: Westview Press, 1986), 179; Samuel P. Huntington, Political Order in ChangingSocieties (New Haven: Yale University Press, 1968), 344-57.

17 Jacques Rupnik, "Hungary's Quiet Revolution," New Republic, 20 November 1989, 20; NewYork Times, 16 April 1989.


16/39


Third wave transformations usually evolved through five major phases, fourof which occurred within the authoritarian system.Emergence of reformers. The first step was the emergence of a group of leadersor potential leaders within the authoritarian regime who believed that movementin the direction of democracy was desirable or necessary. Why did they concludethis? The reasons why people became democratic reformers varied greatly fromcountry to country and seldom were clear. They can, however, be grouped intofive categories. First, reformers often concluded that the costs of staying inpower -such as politicizing their armed forces, dividing the coalition that hadsupported them, grappling with seemingly unsolvable problems (usually eco-nomic), and increasing repression-had reached the point where a graceful exitfrom power was desirable. The leaders of military regimes were particularlysensitive to the corrosive effects of political involvement on the integrity, profes-sionalism, coherence, and command structureof the military. "We all directly orindirectly," General Morales Bermudez observed as he led Peru toward democ-racy, "had been witnesses to what was happening to this institution fundamentalto our fatherland, and in the same vein, to the other institutions. And we don'twant that." In a similar vein, General Fernando Matthei, head of the Chilean airforce, warned, "If the transition toward democracy is not initiated promptly, weshall ruin the armed forces in a way no Marxist infiltration could."'8Second, in some cases reformers wished to reduce the risks they faced if theyheld on to power and then eventually lost it. If the opposition seemed to begaining strength, arrangingfor a democratic transition was one way of achievingthis. It is, after all, preferable to risk losing office than to risk losing life.Third, in some cases, including India, Chile, and Turkey, authoritarianleadersbelieved that they or their associates would not lose office. Having made commit-ments to restore democratic institutions and being faced with declining legitimacyand support these rulers could see the desirability of attempting to renew theirlegitimacy by organizing elections in anticipation that the voters would continuethem in power. This anticipation was usually wrong.Fourth, reformers often believed that democratizing would produce benefitsfor their country: increase its international legitimacy, reduce U.S. or othersanctions against their regime, and open the door to economic and militaryassistance, International MonetaryFund (IMF) loans, invitations to Washington,and inclusion in international gatheringsdominated by the leadersof the Westernalliance.

Finally, in many cases, including Spain, Brazil, Hungary, and Turkey andsome other military regimes, reformers believed that democracy was the "right"form of government and that their country had evolved to the point where, likeother developed and respected countries, it too should have a democratic politicalsystem.

18 Quotedby Abugattasn Malloyand Seligson,eds., Authoritariansnd Democrats,129, andby SylviaT. Borzutzky,"ThePinochetRegime:Crisisand Consolidation"n ibid., 85.


17/39


Liberal reformers tended to see liberalization as a way of defusing oppositionto their regime without fully democratizing the regime. They would ease upon repression, restore some civil liberties, reduce censorship, permit broaderdiscussion of public issues, and allow civil society-associations, churches,unions, business organizations -greater scope to conduct their affairs. Liberal-izers did not, however, wish to introduce fully participatory competitive electionsthat could cause current leaders to lose power. They wanted to create a kinder,gentler, more secure and stable authoritarianismwithout altering fundamentallythe nature of their system. Some reformers were undoubtedly unsure themselveshow far they wished to go in opening up the politics of their country. They alsoat times undoubtedly felt the need to veil their intentions: democratizers tendedto reassurestandpattersby giving the impression that they were only liberalizing;liberalizersattempted to win broader popular support by creating the impressionthey were democratizing. Debates consequently ragedover how far Geisel, Botha,Gorbachev, and others "really"wanted to go.The emergenceof liberalizersand democratizerswithin an authoritariansystemcreates a first-order force for political change. It also, however, can have asecond-order effect. In military regimes in particular it divides the ruling group,further politicizes the military, and hence leads more officers to believe that "themilitary as government" must be ended in order to preserve "the military asinstitution." The debate over whether or not to withdraw from government initself becomes an argument to withdraw from government.Acquiring power. Democratic reformers not only had to exist within the au-thoritarian regime, they also had to be in power in that regime. How did thiscome about? In three cases leaders who created the authoritarian regime presidedover its transition to democracy. In India and Turkey, authoritarianregimesweredefined from the start as interruptions in the formal pattern of democracy. Theregimes were short-lived, ending with elections organized by the authoritarianleaders in the false anticipation that they or the candidates they supported wouldwin those elections. In Chile General Pinochet created the regime, remained inpower for seventeen years, established a lengthy schedule for the transition todemocracy, implemented that schedule in anticipation that the voters wouldextend him in office for eight more years, and exited grudgingly from powerwhen they did not. Otherwise those who created authoritarian regimes or wholed such regimes for prolonged periods of time did not take the lead in endingthose regimes. In all these cases, transformation occurred because reformersreplaced standpatters in power.Reformers came to power in authoritarian regimes in three ways. First, inSpain and Taiwan, the founding and long-ruling authoritarian leaders, Francoand Chiang Kai-shek died. Their designated successors, Juan Carlos and ChiangChing-kuo, succeeded to the mantle, responded to the momentous social andeconomic changes that had occurred in their countries, and began the process ofdemocratization. In the Soviet Union, the deaths in the course of three years ofLeonid Brezhnev, Yuri Andropov, and Konstantine Chernenko allowed Gorba-


18/39


chev to come to power. In a sense, Franco, Chiang, and Brezhnev died in time;Deng Xiao-ping did not.In Brazil and in Mexico, the authoritarian system itself provided for regularchange in leadership. This made the acquisition of power by reformers possiblebut not necessary. In Brazil two factions existed in the military. Repressionreached its peak between 1969 and 1972 during the presidency of GeneralEmilioMedici, a hard-liner. In a major struggle within the military establishment at theend of his term, the soft-line Sorbonne group was able to secure the nominationof General Ernesto Geisel for president, in part because his brother was ministerof war. Guided by his chief associate, General Golbery do Couto e Silva, Geiselbegan the process of democratization and acted decisively to ensure that hewould, in turn, be succeeded in 1978 by another member of the Sorbonne group,General Joao Batista Figueiredo. In Mexico, outgoing President Jose LopezPortillo in 1981 followed standard practice in selecting his minister of planningandbudgets, Miguel de la Madrid, as his successor. De la Madrid was an economicand political liberalizer and, rejectingmore traditional and old-guard candidates,chose a young reforming technocrat, Carlos Salinas, to continue the opening upprocess.Where authoritarian leadersdid not die and were not regularlychanged, demo-cratic reformers had to oust the ruler and install prodemocratic leadership. Inmilitary governments, other than Brazil, this meant the replacement by coupd'etatof one militaryleaderby another: Morales BermudezreplacedJuan Velascoin Peru; Alfredo Poveda replaced Guillermo Rodriguez Lara in Ecuador; OscarMejia replaced Jose Rios Montt in Guatemala; Murtala Muhammed replacedYacubu Gowon in Nigeria.19In the one-party system in Hungary, reformersmobilized their strength and deposed the long-ruling Janos Kadar at a specialparty conference in May 1988, replacing him as secretary general with KarolyGrosz. Grosz, however, was only a semireformer, and a year later the CentralCommittee replaced him with a four-person presidium dominated by reformers.In October 1989 one of them, Rezso Nyers, became party president. In Bulgariainthe fall of 1989, reform-minded Communist party leaders ousted Todor Zhivkovfrom the dominant position he had occupied for thirty-five years. The leadershipchanges associated with some liberalizingand democratizing reforms are summa-rized in Table 2.Thefailure of liberalization. A critical issue in the third wave concerned therole of liberal reformers and the stability of a liberalized authoritarian polity.Liberal reformers who succeeded standpatter leaders usually turned out to betransition figures with brief stays in power. In Taiwan, Hungary, and Mexico,liberalizers were quickly succeeded by more democratically oriented reformers.In Brazil, although some analysts are dubious, it seems reasonably clear that

19See Needler, "The Military Withdrawal,"621-23 on "second phase"coups and the observationthat "the military government that returns power to civilian hands is not the same one that seizedpower from the constitutional government in the first place."


19/39

596 j POLITICAL SCIENCE QUARTERLY

TABLELeadershipChangeand Reform, 1973-90

FirstStandpat Reform Reform Democratic

Country Leader Change Leader I Change Leader 11 ElectionNigeria Gowon July1975coup Murtala February976death Obasanjo August 979

MohammedEcuador Rodriguez January976coup Poveda - - April 979

LaraPeru Velasco August 975coup Morales - - May 980

BermudezBrazil Medici March 974succession Geisel March 979 uccession Figueiredo January 985Guatemala Rios August 983coup Mejia - - December985

MonttSpain Franco November975death Juan - Juan March 979

Carlos CarlosCarrero December 973death Arias July1976ouster SuArez

BlancoTaiwan Chiang April975death Chiang January 988death Lee

Kai-shek Ching-kuo Teng-huiHungary Kadar May 988ouster Grosz May-October989ouster Nyers-PozsgayMarch 990Mexico Portillo December 982succession De la December 988succession Salinas

MadridSouthAfrica Vorster September978ouster Botha September989ouster de KlerkUSSR Chernenko March 985death GorbachevBulgaria Zhivkov November989ouster Mladenov - - June1990

Geisel and Golbery do Couto e Silva were committed to meaningful democratiza-tion from the start.20Even if they did just intend to liberalize the authoritariansystem rather than replace it, Joao Figueiredo extended the process to democrati-zation. "I have to make this country into a democracy," he said in 1978 beforetaking office, and he did.21In Spain the hard-line prime minister, Admiral Luis Carrero Blanco, wasassassinated in December 1973, and Franco appointed Carlos Arias Navarro tosucceed him. Arias was the classic liberal reformer. He wished to modify theFranco regimein order to preserveit. In a famous speech on 12February 1974, heproposed an opening (apertura)and recommended a number of modest reforms

including, for instance, permitting political associations to function but not polit-ical parties. He "was too much of a conservative and Francoist at heart to carryout a true democratization of the regime." His reform proposals were torpedoed

20 Stepan,RethinkingMilitaryPolitics, 32-40; andThomasE. Skidmore,"Brazil's lowRoadtoDemocratization:974-1985"nStepan,ed., Democratizing razil,33. This nterpretationoincideswith my own impression f Golbery's ntentions hat I formed n 1974workingwith him on plansfor Brazil'sdemocratization. or a contraryargument, ee Silvio R. Duncan Barettaand JohnMarkoff, "Brazil'sAbertura:A Transition rom Whatto What?" n Malloy and Seligson, eds.,Authoritarians nd Democrats,45-46.21 Quoted n FranciscoWeffort, "WhyDemocracy?"n Stepan,ed., Democratizing razil, 332.


20/39

HOW COUNTRIES DEMOCRATIZE j 597

by the standpatters of the "bunker," including Franco; at the same time theproposals stimulated the opposition to demand a more extensive opening. In theend, Arias "discredited aperturismo just as Luis Carrero Blanco had discreditedimmobilism."22In November 1975 Franco died and Juan Carlos succeeded himas chief of state. Juan Carlos was committed to transforming Spain into a true,European-style parliamentary democracy, Arias resisted this change, and in July1976 Juan Carlos replaced him with Adolfo Suarez, who moved quickly to intro-duce democracy.The transition from liberalized authoritarianism, however, could move back-ward as well as forward. A limited opening could raise expectations of furtherchange that could lead to instability, upheaval, and even violence; these, in turn,could provoke an antidemocratic reaction and replacement of the liberalizingleadership with standpatter leaders. In Greece, George Papadopoulos attemptedto shift from a standpatter to a liberalizing stance; this led to the Polytechnicstudent demonstration and its bloody suppression; a reaction followed and theliberalizing Papadopoulos was replaced by the hard-line Dimitrios Ioannidis. InArgentina General Roberto Viola succeeded the hard-line General Jorge Videlaas president and began to liberalize. This produced a reaction in the military,Viola's ouster, and his replacement by hard-line General Leopoldo Galtieri. InChina ultimate power presumably rested with Deng Xiao-ping. In 1987, howeverZhao Ziyang became general secretary of the Communist party and began toopen up the political system. This led to the massive student demonstrations inTiananmen Square in the spring of 1989, which, in turn, provoked a hard-linereaction, the crushing of the student movement, the ouster of Zhao, and hisreplacement by Li Peng. In Burma, General Ne Win, who had ruled Burma fortwenty-six years, ostensibly retired from office in July 1988 and was replaced byGeneral Sein Lwin, another hard-liner. Mounting protests and violence forcedSein Lwin out within three weeks. He was succeeded by a civilian and presumedmoderate, Maung Maung, who proposed elections and attempted to negotiatewith opposition groups. Protests continued, however, and in Septemberthe armydeposed Maung Maung, took control of the government, bloodily suppressedthe demonstrations, and ended the movement toward liberalization.The dilemmas of the liberalizerwere reflected in the experiencesof P. W. Bothaand Mikhail Gorbachev. Both leaders introduced major liberalizing reforms intheir societies. Botha came to power in 1978 with the slogan "Adapt or die"and legalized black trade unions, repealed the marriage laws, established mixedtrading zones, granted citizenship to urban blacks, permitted blacks to acquirefreehold title, substantially reduced petty apartheid, increased significantly in-vestment in black education, abolished the pass laws, provided for elected blacktownship councils, and created houses of parliament representingcoloureds and

22Raymond Carr and Juan Pablo Fusi Aizpurua, Spain: Dictatorship to Democracy, 2d ed.(London: Allen & Unwin, 1981), 198-206.


21/39


Asians, althoughnot blacks. Gorbachevopened up public discussion, greatlyreducedcensorship,dramatically hallenged he powerof the Communistpartyapparat,and introducedat leastmodest forms of government esponsibilityoan elected egislature.Both leadersgavetheirsocietiesnew constitutions ncorpo-ratingmanyreformsand also creatingnew and verypowerfulposts of president,whichthey then assumed.It seemsprobable hat neitherBotha nor Gorbachev,however, wanted fundamental hangein their political systems. Their reformsweredesigned o improveand to moderate,but also to bolster he existing ystemand makeit moreacceptable o theirsocieties. Theythemselvessaid as muchrepeatedly.Botha did not intendto end white power;Gorbachevdid not intendto end communistpower.As liberalreformers hey wantedto changebut alsoto preservehe systems hattheyled andin whosebureaucraciesheyhadspenttheir careers.Botha's iberalizingbut not democratizing eformsstimulated ntensifiedde-mands from South African blacks for their full incorporation nto the politicalsystem. In September1984 black townships eruptedwith proteststhat led toviolence, repression,and the deploymentof military orces into the townships.The efforts at reformsimultaneously nded, and Botha the reformerwas widelyviewed as having become Botha the repressor.The reform process only gotunderway gain n 1989whenBothawasreplacedbyF. W. deKlerk,whosemoreextensive reforms led to criticismsfrom Botha and his resignationfrom theNationalparty.In 1989and 1990Gorbachev'siberalizing ut not democratizingreformsappeared o be stimulating omparable pheaval,protests,and violenceinthe SovietUnion. As in SouthAfrica, communal roups oughteachotherandthe centralauthorities.The dilemma or Gorbachevwas clear.Movingforwardtoward full-scaledemocratizationwouldmeannot only the end of communistpower nthe SovietUnionbutveryprobably heend of the SovietUnion.Leadinga hard-linereaction o the upheavalswould mean the end of his efforts at eco-nomicreform,his greatly mproved elationswiththeWest,and hisglobal mageas a creativeand humane eader. Andrei Sakharovput the choicessquarely oGorbachevn 1989:"Amiddlecourse nsituationsikethese s almost mpossible.The countryand you personallyare at a crossroads either ncrease he processof changemaximally rtryto retain headministrativeommand ystemwith allits qualities."23Where t wastried,liberalizationtimulated he desirefor democratizationnsomegroupsand the desirefor repressionn others.Theexperience f thethirdwavestronglysuggests hat liberalizedauthoritarianisms not a stableequilib-rium;the halfwayhouse does not stand.

Backward egitimacy: ubduingthestandpatters.The achievementof powerenabled he reformers o startdemocratizing ut it did not eliminate he ability23 Quoted in David Remnick, "The Struggle for Light," New York Review of Books, 16 August

1990, 6.


22/39


of the standpatterso challenge he reformers.The standpatter lementsof whathad been the governingcoalition-the Francoist"bunker"n Spain, the militaryhard-linersn Braziland other Latin Americancountires, he Stalinists n Hun-gary, the mainlander ld guard n the KMT, the party bosses and bureaucracyin the PRI, the Verkrampte ing of the Nationalparty-did not give up easily.In the government,military,andpartybureaucraciestandpattersworked o stopor slow down the processesof change. In the non-one-party ystemsBrazil,Ecuador,Peru, Guatemala,Nigeria,and Spain standpatter roups nthemili-tary attemptedcoups d'etatand made other efforts to dislodge the reformersfrom power. In South Africa and in Hungary,standpatter actionsbroke awayfrom the dominantparties,charging hem with betraying he basic principles nwhich the partieswere based.Reformgovernments ttempted o neutralize tandpatter ppositionby weak-ening, reassuring, ndconverting he standpatters.Countering tandpatteresis-tance often requireda concentrationof powerin the reformchief executive.Geiselassertedhimselfas"dictator f theabertura"norder o forcethe Brazilianmilitaryout of politics.24 uanCarlosexercisedhis powerand prerogativeso thefull in moving Spain toward democracy,not least in the surpriseselection ofSuarezas primeminister.Botha and Gorbachev,as we have seen, createdpow-erful new presidentialoffices for themselves.Salinas dramatically ssertedhispowers duringhis first yearsas Mexico'spresident.The first requirementor reform eaderswas to purge he governmental,mili-tary, and,whereappropriate, artybureaucracies, eplacing tandpattersn topofficeswith supporters f reform.This wastypicallydoneinselective ashionsoas not to provokea strong reactionand so as to promotefissions withinthestandpatter anks.In addition to weakeningstandpatters, eform leadersalsotried o reassure ndconvert hem.Inmilitary egimes, he reformers rgued hatit was time to go back, after a necessarybut limitedauthoritariannterlude, othe democraticprincipleshat werethe basisof theircountry'spoliticalsystem.In this sense, they appealedfor a "return o legitimacy." n the nonmilitaryauthoritarian ystems,reformers nvoked "backwardegitimacy" nd stressedelementsof continuitywiththe past.25n Spain,for instance, hemonarchywasreestablished ndSuarezadhered o the provisionsof the Francoconstitutionnabolishing hat constitution:no Francoist ouldclaim hattherewereproceduralirregularities.nMexicoand SouthAfricathereformersnthePRIandNationalpartycast themselves n the traditionsof those parties.On Taiwan the KMTreformersappealed o Sun Yat-Sen's hreeprinciples.Backwardegitimacyhadtwo appealsand two effects: it legitimatedhe new

24 See Stepan, Rethinking Military Politics, 42-43.25 Giuseppe Di Palma highlightedthe significance of backward legitimacy in "Founding Coalitions

in Southern Europe: Legitimacy and Hegemony," Government and Opposition 15 (Spring 1980):170. See also Nancy Bermeo, "Redemocratizationand Transition Elections: A Comparison of Spainand Portugal," Comparative Politics 19 (January 1987): 218.


23/39


order because it was a product of the old, and it retrospectively legitimated theold order because it had produced the new. It elicited consensus from all exceptopposition extremists who had no use for either the old authoritarian regime orthe new democratic one. Reformers also appealed to standpatters on the groundsthat they were preemptingthe radicalopposition and hence minimizing instabilityand violence. Suarez, for instance, asked the Spanish army to support him forthese reasons and the dominant elements in the army accepted the transitionbecause there "was no illegitimacy, no disorderin the streets, no significant threatof breakdown and subversion." Inevitably, the reformers also found that, asGeisel put it, they could "not advance without some retreats" and that hence, onoccasion, as in the 1977 "April package" in Brazil, they had to make concessionsto the standpatters.26Coopting the opposition. Once in power the democratic reformers usuallymoved quickly to begin the process of democratization. This normally involvedconsultations with leaders of the opposition, the political parties, and majorsocial groups and institutions. In some instances relatively formal negotiationsoccurred and quite explicit agreements or pacts were reached. In other cases, theconsultations and negotiations were more informal. In Ecuador and Nigeria thegovernment appointed commissions to develop plans and policies for the newsystem. In Spain, Peru, Nigeria, and eventually in Brazil elected assembliesdrafted new constitutions. In several instances referenda were held to approvethe new constitutional arrangements.As the reformers alienated standpatters within the governing coalition, theyhad to reinforce themselves by developing support within the opposition and byexpandingthe political arena and appealingto the new groups that werebecomingpolitically active as a result of the opening. Skillful reformers used the increasedpressurefrom these groups for democratization to weaken the standpatters, andused the threat of a standpatter coup as well as the attractions of a share in powerto strengthen moderate groups in the opposition.

To these ends, reformers in government negotiated with the principal opposi-tion groups and arrived at explicit or tacit agreements with them. In Spain, forinstance, the Communist party recognized that it was too weak to follow a"radical rupturista policy" and instead went along with a "rupturapactada"eventhough the pact was "purelytacit." In October 1977 Suarez won the agreementof the Communist and Socialist parties to the Pactos de la Moncloa comprisinga mixture of fairly severe economic austerity measures and some social reforms.Secret negotiations with Santiago Carrillo, the principal Communist leader,"playedon the PCE [Partido Comunista de Espafial leader's anxiety to be near thelevers of power and secured his backing for an austerity package."27 n Hungary

26 StanleyG. Payne, "TheRole of the Armed Forces n the SpanishTransition"n ClarkandHaltzel, eds., Spain in the 1980s, 86; Stepan, Rethinking Military Politics, 36.

27 Thesespresentedby the CentralCommittee,Ninth Congress,CommunistParty of Spain, 5-9 April 1978,quoted n JuanJ. Linz, "S'me Comparative houghtson the Transition o Democracy


24/39


explicit negotiations occurred in the fall of 1989 between the Communist partyand the Opposition Round Table representing the principal other parties andgroups. In Brazil informal understandings developed between the governmentand the opposition parties, the Movimento Democratico Brasileiro (MDB) andthe Partido Movimento Democratico Brasileiro (PMDB). On Taiwan in 1986 thegovernment and the opposition arrived at an understanding on the parameterswithin which political change would take place and, in a week-long conferencein July 1990, agreed on a full schedule of democratization.Moderation and cooperation by the democratic opposition- their involvementin the process as junior partners were essential to successful transformation. Inalmost all countries, the principalopposition parties - the MDB-PMDB in Brazil,the Socialists and Communists in Spain, the Democratic Progressive Party (DPP)on Taiwan, the Civic Forum in Hungary, the Alianza Popular RevolucionariaAmericana (APRA) in Peru, the Christian Democrats in Chile - were led bymoderates and followed moderate policies, at times in the face of considerableprovocation by standpatter groups in the government.Thomas E. Skidmore's summary of what occurred in Brazil neatly catches thecentral relationship involved in transformation processes:

Intheend, liberalizationwas theproductof anintensedialectical elationship etweenthegovernment nd theopposition.Themilitarywhofavoredaberturahad to proceedcautiously,for fear of arousing he hardliners.Theirovertureso theoppositionweredesignedo drawout the"responsible"lements, hereby howing hereweremoderatesreadyto cooperatewiththegovernment.At the sametime, the oppositionconstantlypressed he government o end its arbitrary xcesses,therebyremindinghe militarythattheirrulelacked egitimacy.Meanwhile,heoppositionmoderateshad to remindthe radicals hat theywouldplay into the hands of the hardlinersf they pushedtoohard.Thisintricatepoliticalrelationship unctionedsuccessfullybecausethere was aconsensusamongboth militaryand civilians n favor of a return o an (almost)openpoliticalsystem.28

Guidelines for Democratizers 1:Reforming Authoritarian SystemsThe principal lessons of the Spanish, Brazilian, and other transformations fordemocratic reformers in authoritarian governments include the following:(1) Secureyour political base. As quickly as possible place supportersof democ-ratization in key power positions in the government, the party, and the military.(2) Maintain backward legitimacy, that is, make changes through the estab-lished procedures of the nondemocratic regime and reassure standpatter groupsin Portugal and Spain" n Jorge Bragade Macedo and Simon Serfaty, eds., Portugal Since theRevolution: Economic and Political Perspectives (Boulder, CO: Westview Press, 1981), 44; Preston,Triumph of Democracy in Spain, 137.

28 Skidmore,"Brazil's low Road")inStepan,ed., Democratizing Brazil, 34.


25/39


with symbolic concessions, following a course of two steps forward, one stepbackward.(3) Gradually shift your own constituency so as to reduce your dependence ongovernment groups opposing change and to broaden your constituency in thedirection of opposition groups supporting democracy.

(4) Be prepared for the standpatters to take some extreme action to stop change(for example, a coup attempt) - possibly even stimulate them to do so - andthen crack down on them ruthlessly, isolating and discrediting the more extremeopponents of change.

(5) Seize and keep control of the initiative in the democratization process. Onlylead from strength and never introduce democratization measures in response toobvious pressure from more extreme radical opposition groups.

(6) Keep expectations low as to how far change can go; talk in terms of main-taining an ongoing process rather than achieving some fully elaborated demo-cratic utopia.

(7) Encourage development of a responsible, moderate opposition party, whichthe key groups in society (including the military) will accept as a plausible non-threatening alternative government.

(8) Create a sense of inevitability about the process of democratization so thatit becomes widely accepted as a necessary and natural course of developmenteven if to some people it remains an undesirable one.

REPLACEMENTSReplacements involve a very different process from transformations. Reformerswithin the regime are weak or nonexistent. The dominant elements in governmentare standpatters staunchly opposed to regime change. Democratization conse-quently results from the opposition gaining strength and the government losingstrengthuntil the government collapses or is overthrown. The former oppositiongroups come to power and the conflict then often enters a new phase as groupsin the new government struggle among themselves over the nature of the regimethey should institute. Replacement, in short, involves three distinct phases: thestruggle to produce the fall, the fall, and the struggle after the fall.Most third wave democratizations required some cooperation from those inpower. Only six replacements had occurred by 1990. Replacements were rare intransitions from one-party systems (one out of eleven) and military regimes (twoout of sixteen) and more common in transitions from personal dictatorships(three out of seven). As we have pointed out, with some exceptions (Gandhi,Kenan Evren, Pinochet), leaders who created authoritarian regimes did not endthose regimes. Changes of leadership within authoritarian systems were muchmore likely in military regimes through "second phase" coups or, in one-partysystems, through regular succession or the action of constituted party bodies.Personal dictators, however, seldom retired voluntarily, and the nature of their


26/39


power - personal rather than military or organizational - made it difficult foropponents within the regime to oust them and, indeed, made it unlikely that suchopponents would exist in any significant numbers or strength. The personaldictator was thus likely to hang on until he died or until the regime itself cameto an end. The life of the regime became the life of the dictator. Politically andat times literally (for example, Franco, Ceausescu) the deaths of the dictator andthe regime coincided.Democratic reformers were notably weak in or missing from the authoritarianregimes that disappeared in replacements. In Argentina and Greece, the liberal-izing leaders Viola and Papadopoulos were forced out of power and succeededby military hard-liners. In Portugal Caetano initiated some liberalizing reformsand then backed away from them. In the Philippines, Romania, and East Ger-many, the entourages of Marcos, Ceausescu, and Erich Honecker contained fewif any democrats or even liberals. In all six cases standpatters monopolized power,and the possibility of initiating reform from within was almost totally absent.An authoritarian system exists because the government is politically strongerthan the opposition. It is replaced when the government becomes weaker thanthe opposition. Hence replacement requires the opposition to wear down thegovernment and shift the balance of power in its favor. When they were initiated,the authoritarian regimes involved in the third wave were almost always popularand widely supported. They usually had the backing of a broad coalition ofgroups. Over time, however, as with any government, their strengthdeteriorated.The Greek and Argentine military regimes suffered the humiliation of militarydefeat. The Portuguese and Philippine regimes were unable to win counterinsur-gency wars, and the Philippine regime created a martyr and stole an election.The Romanian regime followed policies that deeply antagonized its people andisolated itself from them; hence it was vulnerable to the cumulative snowballingof the antiauthoritarian movement throughout Eastern Europe. The case of EastGermany was more ambiguous. Although the regime was relatively successful insome respects, the inevitable comparison with West Germany was an inherentweakness, and the opening of the transit corridor through Hungary dramaticallyundermined the regime's authority. The party leadership resigned in early De-cember 1989, and a caretaker government took over. The regime's authority,however, evaporated, and with it the reasons for the East German state.The erosion of support for the regime sometimes occurred openly, but, giventhe repressive character of authoritarian regimes, it was more likely to occurcovertly. Authoritarian leaders were often unaware of how unpopular they were.Covert disaffection then manifested itself when some triggering event exposedthe weakness fo the regime. In Greece and Argentina it was military defeat. InPortugal and East Germany it was the explicit turning against the regime ofits ultimate source of power-the army in Portugal, the Soviet Union in EastGermany. The actions of the Turks, the British, the Portuguese military, andGorbachevgalvanized and brought into the open the disaffection from the regimeof other groups in those societies. In all these cases, only a few weak groups


27/39

604 1POLITICAL SCIENCE QUARTERLY

rallied o the supportof the regime.Manypeople had becomedisaffected romthe regimebut, because t was an authoritarian egime, a triggering vent wasrequired o crystalize he disaffection.Studentsare the universalopposition;they oppose whateverregimeexists intheir society. By themselves,however,students do not bring down regimes.Lacking substantialsupport from other groups in the population, they weregunneddownby the militaryand police in Greece n November1973, Burma nSeptember1988, andChina n June 1989.The militaryare theultimatesupportof regimes.If theywithdraw heir support, f they carryout a coup against heregime,orif theyrefuse o use force against hose who threaten o overthrow heregime, he regime alls. In between he perpetual ppositionof thestudentsandthe necessarysupportof the military are other groups whose support for oroppositionto the regimedependson circumstances.n noncommunist uthori-tariansystems,such as the Philippines, hesegroupstendedto disaffectin se-quence.The disaffectionof the studentswasfollowedby thatof intellectualsngeneraland then by the leadersof previously xistingpoliticalparties,manyofwhommayhavesupported r acquiescednthe authoritarianakeover.Typicallythe broaderreaches of the middle class-white-collar workers,professionals,small businessproprietors-becamealienated.In a Catholiccountry,Churchleadersalso wereearly and effective opponentsof the regime.If labor unionsexisted and werenot totallycontrolledby the government,at some point theyjoinedthe opposition.So also, andmost important,did largerbusinessgroupsand thebourgeoisie. n duecourse,the United Statesorotherforeignsourcesofsupportbecamedisaffected.Finallyand conclusively, he militarydecidednotto supportthe governmentor activelyto side with the oppositionagainst thegovernment.In five out of sixreplacements, onsequently, he exceptionbeing Argentina,militarydisaffectionwas essential o bringingdown the regime.In the personaldictatorshipsn Portugal,the Philippines,andRomania,this militarydisaffec-tionwaspromotedbythe dictator'spoliciesweakeningmilitaryprofessionalism,politicizing ndcorruptingheofficercorps,andcreating ompetingparamilitaryand security orces.Opposition o the governmentnormally Portugalwas theonly exception)hadto bewidespread efore hemilitarydesertedhegovernment.If disaffectionwas not widespread, t was either becausethe most probablesourcesof opposition-the middleclass, bourgeoisie, religiousgroups weresmall and weakor because he regimehad the supportof thesegroups, usuallyas a resultof successfulpoliciesforeconomicdevelopment. n Burmaand Chinathe armed forcesbrutallysuppressedproteststhat werelargelystudent-led.Insocieties hatweremorehighlydeveloped conomically, pposition o authoritar-ianismcommandeda widerrangeof support.Whenthis opposition ook to thestreets n the Philippines,EastGermany,and Romania,militaryunits did notfire on broadlyrepresentative roupsof their fellow citizens.A popular mageof democratic ransitionss thatrepressive overnments rebroughtdown by "peoplepower,"the mass mobilizationof outragedcitizens


28/39


demanding and eventually forcing a change of regime. Some form of mass actiondid take place in almost every third wave regime change. Mass demonstrations,protests, and strikes played central roles, however, in only about six transitionscompleted or underway at the end of the 1980s. These included the replacementsin the Philippines, East Germany, and Romania, and the transplacements inKorea, Poland, and Czechoslovakia. In Chile frequent mass actions attempted,without success, to alter Pinochet's plan for transformation. In East Germany,uniquely, both "exit" and "voice," in Albert Hirschman's terms, played majorroles, with protest taking the form first of massive departure of citizens from thecountry and then of massive street demonstrations in Leipzig and Berlin.29

Inthe Philippines, Portugal, Romania, and Greece, when the regime collapsed,it collapsed quickly. One day the authoritarian government was in power, thenext day it was not. In Argentina and East Germany, the authoritarian regimeswere quickly delegitimated but clung to power while attempting to negotiateterms for the change in regime. In Argentina, the successor military governmentof General Reynaldo Bignone, which took over in July 1982 immediately afterthe Falklands defeat, was "relatively successful" in maintaining some regimecontrol over the transition for six months. In December 1982, however, mountingpublic opposition and the development of opposition organizations led to massprotests, a general strike, Bignone's scheduling of elections, and the rejection bythe united opposition partiesof the termsproposed by the military for the transferof power. The authority of the lame-duck militaryregime continued to deteriorateuntil it was replaced by the Alfonsin government elected in October 1983. "Themilitary government collapsed," one author observed; "it had no influence overthe choice of candidates or the election itself, it excluded no one, and reservedneither powers nor veto prerogatives for itself in the future. In addition, it wasunable to guarantee either its autonomy in relation to the future constitutionalgovernment or the promise of a future military policy, and, even less -given thewinning candidate -the basis for an agreement on the ongoing struggle againstthe guerrillas."30 n East Germany in early 1990 a somewhat similar situationexisted, with a weak and discredited communist government clinging to power,and its prime minister, Hans Modrow, playing the role of Bignone.The emphasis in transformations on proceduralcontinuity and backwardlegiti-macy was absent from replacements. The institutions, procedures, ideas, andindividuals connected with the previous regime were instead considered taintedand the emphasis was on a sharp, clean break with the past. Those who succeededthe authoritarianrulers based their rule on "forwardlegitimacy,"what they wouldbring about in the future, and their lack of involvement in or connection withthe previous regime.

29 Albert 0. Hirschman, Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations,and States (Cambridge, MA: Harvard University Press, 1970).

30 Virgilio R. Beltran, "PoliticalTransition in Argentina: 1982 to 1985,"Armed Forces and Society13 (Winter 1987): 217; Scott Mainwaringand Eduardo J. Viola, "Brazil and Argentina in the 1980s,"Journal of International Affairs 38 (Winter 1985): 206-9.


29/39


In transformations and transplacements the leaders of the authoritarian re-gimes usually left politics and went back to the barracks or private life quietlyand with some respect and dignity. Authoritarian leaderswho lost power throughreplacements, in contrast, suffered unhappy fates. Marcos and Caetano wereforced into exile. Ceausescu was summarily executed. The military officers whoran Greece and Argentina were tried and imprisoned. In East Germany punish-ments were threatened against Honecker and other former leaders in notablecontrast to the absence of such action in Poland, Hungary, and Czechoslovakia.The dictators removed by foreign intervention in Grenada and Panama weresimilarly subjected to prosecution and punishment.

The peaceful collapse of an authoritarian regime usually produced a gloriousif brief moment of public euphoria, of carnations and champagne, absent fromtransformations. The collapse also created a potential vacuum of authority absentfrom transformations. In Greece and the Philippines, the vacuum was quicklyfilled by the accession to power of Constantine Karamanlis and Corazon C.Aquino, popular political leaders who guided their countires to democracy. InIran the authority vacuum was filled by the ayatollah, who guided Iran elsewhere.In Argentina and East Germany the Bignone and Modrow governments weaklyfilled the interim between the collapse of the authoritarian regimes and the elec-tion of democratic governments.Before the fall, opposition groups are united by their desire to bring about thefall. After the fall, divisions appear among them and they struggle over thedistribution of power and the nature of the new regime that must be established.The fate of democracy was determined by the relative power of democraticmoderates and antidemocratic radicals. In Argentina and Greece, the authori-tarian regimes had not been in power for long, political parties quickly reap-peared, and an overwhelming consensus existed among political leaders andgroups on the need quickly to reestablish democratic institutions. In the Philip-pines overt opposition to democracy, apart from the NPA insurgency, also wasminimal.In Nicaragua, Iran, Portugal, and Romania the abrupt collapse of the dictator-ships led to struggles among the former opposition groups and parties as to whowould exercise power and what sort of regime would be created. In Nicaraguaand Iran the democratic moderates lost out. In Portugal, a state of revolutionaryferment existed between April 1974 and November 1975. A consolidation ofpower by the antidemocratic Marxist-Leninist coalition of the Communist partyand left-wing military officers was entirely possible. In the end, after intensestruggles between military factions, mass mobilizations, demonstrations, andstrikes, the military action by Antonio Ramalho Eanes settled Portugal on ademocratic course. "What started as a coup," as Robert Harvey observed, "be-came a revolution which was stopped by a reaction before it became an anarchy.Out of the tumult a democracy was born."'"

31 Robert Harvey, Portugal: Birth of a Democracy (London: Macmillan, 1978), 2.


30/39


The choices in Portugal were between bourgeois democracy and Marxist-Leninist dictatorship. The choices in Romania in 1990 were less clear, but democ-racy also was not inevitable. The lack of effectively organized opposition partiesand groups, the absence of previous experience with democracy, the violenceinvolved in the overthrow of Ceausescu, thedeep desire for revengeagainst peopleassociated with the dictatorship combined with the widespread involvement ofmuch of the population with the dictatorship, the many leaders of the new govern-ment who had been part of the old regime-all did not augur well for the emer-gence of democracy. At the end of 1989 some Romanians enthusiastically com-pared what was happening in their country to what had happened two hu

sm huntington 1992

Documents

Transcript of sm huntington 1992