More than Just the Final Straw : Stolen Elections as Revolutionary Triggers Author(s): Philipp Kuntz and Mark R. Thompson Source: Comparative Politics, Vol. 41, No. 3 (April 2009), pp. 253-272 Published by: Ph.D. Program in Political Science of the City University of New York Stable URL: http://www.jstor.org/stable/40599224 . Accessed: 03/09/2013 14:52

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. 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].

.

Ph.D. Program in Political Science of the City University of New York is collaborating with JSTOR to digitize, preserve and extend access to Comparative Politics.

http://www.jstor.org

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions More thanJust the Final Straw

StolenElections as RevolutionaryTriggers PhilippKuntz and MarkR. Thompson

Disillusionmentwith the failings of post-coldwar democratizationhas grownsteadily over the last decade. Many countriesare no longergoing through a rockyphase of democraticconsolidation but have insteadbecome "electoral authoritarian."1 While the reasonsfor such democratic decay are manifold,one aspectis noteworthyfrom a cross- nationalperspective: widespread political apathy in theface of a drifttowards electoral .This is allthe more remarkable given that recent waves of democratization in easternEurope, Asia, andAfrica have been characterized by transitions "from below," withpopular protest movements playing a decisiverole.2 Many examples - amongthem Milosevic'sSerbia, Shevardnadze's Georgia, and Ukraineunder Kuchma - revealthat people can remainpassive even in the face of grosshuman rights violations and/or a thoroughcorruption of government. Yet onlya shorttime after such pessimisticdiagnoses, tens or even hundredsof thousandsof citizens poured onto the streets of Belgrade (2000), Tbilisi(2003), andKiev (2004), fullof determination to get rid of their electoral authoritarian rulers. Vitali Silitski has nicelysummarized why so manypeople were astonishedby theserevolutionary "eruptions"in ,Georgia, and Ukraine:"The greateststory in all threerevolutions is theemergence of a truedemocratic spirit in societieswritten off by manyobservers (oftenincluding insiders) as passive,indifferent, submissive, and inherentlyunfit for democracy."3Not only oppositionactivists, but also powerholderswere caughtby surprise.It was onlya matterof weeksor even daysbefore these regimes surrendered power.How can sucha suddenand powerful mass mobilization be explained? The crucial turningpoint occurredonly a shorttime beforepeople started demonstrating.Massive antiregimeprotests broke out after"soft" authoritarian rulers engagedin blatantelectoral fraud. More precisely,interference in theelectoral process occurredat the last moment: conventional, that is, more subtle means of keeping elections undercontrol in advancehad failed.It becamea commonperception that the opposition had actuallydefeated the regime. The last resortfor cornered autocrats was to falsify crudelythe final results or to annulthe electoral contest altogether. Stolenelections can be definedas electionsin whichthe regime hinders an actualor perceivedopposition victory at theballot box throughblatant manipulation of thevote countor by annullingthe electoral result itself. While in a normativesense everyact

253

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April 2009

of manipulationamounts to an act of stealing,it is usefulto reservethe word "stolen" forelections in whichthe regime is believedto have lost thevoting despite attempts at manipulation.Ordinary people can reachthis conclusion in severaldifferent ways. Sometimesrulers are so surprisedby electoral defeat that voting officials start publishing actualresults before they receive orders to stop counting (or they go throughwith the count, onlyto havethe election annulled altogether). In othercases, exit polls andparallel vote countsprovide solid evidencefor the actual outcome. International and pro-opposition mediaas well as movementactivists can easilyspread such information in a heightened postelectoralatmosphere. There is thepossibility of hype in this process (what ultimately mattersis theperception of an election'sbeing stolen), but usually the opposition's claims arebased on somekind of credible evidence.4 Understoodthis way, a stolenelection is not just the final straw that breaks the autocrat's back.Rather, it constitutes a powerful transformatory event which fundamentally reshapes politicalcontestation and thus deserves to be distinguishedfrom other forms of electoral fraud.This findingis notonly of interestto studentsof contemporarydemocratization. Throughan analysisof stolenelections as triggeringevents, a usefulcontribution can also be made to the studyof revolutions.As the nextsection will show,attempts to emphasizerevolutionary précipitants have encounteredpersistent academic resistance. By systematicallydemonstrating the impact of a short-termevent like stolenelections, thismainstream view can be challenged. The significanceof stolenelections will be shownin threeways. First, a modelwill be introducedthat reveals the potential for theorizing this triggering event among different groupsof actors:citizens, activists, and regimemembers. Second, stolen elections will be situatedin thestructural context of electoralauthoritarianism. Blatant cheating on the scale ofannulling an oppositionvictory is crucialin overcomingstrong collective action barriersthat are peculiarto thistype of regime.Third, stolen elections were more than justthe final straw in thedemocratic uprisings that swept away the autocratic regimes of SlobodanMilosevic, Eduard Shevardnadze, and Leonid Kuchma.Had stolenelections been merelythe spark ignitingan alreadyexplosive situation, then other,equally reprehensiblepolitical acts shouldhave yielded a similareffect. Finally, examination of othercases beyondSerbia, Georgia, and Ukrainecan determinehow farthe argument aboutstolen elections can be carried.

TriggeringEvents and Theoriesof Revolution

Disagreementover long-termversus short-term causes has plaguedthe sociologyof revolutionssince it beginnings,but differences of opinionbecame more pronounced in thetwentieth century. Scholars in thenatural history of revolutiontradition reserved a place fortriggering events, as did latertheories or syntheseswritten in less metaphorical language.While the linkbetween triggers and revolutionaryconsequences remained

254

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson undertheorized,at least précipitants were recognized as beingof potentialimportance.5 Not surprisingly,proponents of the structuralistschool, who assumedthe impactof long-termdevelopments to be overriding,by and largeignored the events immediately precedinga revolution. Well-known works by Barrington Moore, Jeffery Paige, and Theda Skocpolare tellingin thisrespect.6 Skocpol, the mostinfluential structuralist theorist of revolutions,is bestknown for her resolute rejection of voluntarism,but her account leavesequally little room for the impact of short-termevents. Althoughpurely structuralist accounts no longerdominate the study of revolutions - Charles Kurzman speaks of a poststructuralistconsensus - skepticismabout the importanceof short-termevents persists.7 Only recently have some scholarsidentified short-termtemporality as a neglectedtopic and startedto builda new researchagenda aroundit.8 But so farexisting studies are dominatedby approachesstressing general "eventfulness."Sometimes this involves a culturalistemphasis, in whichthe cultural transformationbrought about by keyevents is stressedover resource distributional and politicalconsequences.9 In othercases wheresuch factorsare emphasized,a seriesof crucialevents ("accelerators") are consideredas steppingstones to revolution.10But this viewneglects singular events that lead to rapidpolitical change (metaphorically forging a riverdirectly to revolution). As nosystematic attempt has yet been undertaken todemonstrate the key role played by a trigger,strong objections persist against conceding such events much causal importance. Michael Kimmel,for example, in his standardaccount of revolutions,acknowledges triggersas an indispensableprerequisite but stresses the need to searchfor the "structural roots"of revolutions. He likensprerevolutionary situations to a housethat is "vulnerable toa strayspark from a match"because it consists of highly flammable materials. Once the househas burnt down there is a dangerof mistaking the immediate reason for the long-run causes."A straymatch does notcause a fire;it createsthe missing - oftenaccidental - ingredientin the fire-prone situation."11 Precipitating events in this picture are replaceable becauseit does not really matter what exactly starts the fire. Edgar Kiser and Margart Levi offera morescathing critique. They applaud the structuralist perspective for being aware "thatthe events most immediately preceding revolutions are 'triggerevents' rather than causes."12In thisperspective, short-term précipitants have lost any relevance. How may this dismissalbe foundedon empiricalgrounds? Timothy Wickham- Crowley,another outspoken critic, who ascribesonly a "severelylimited causal role" to triggeringevents, argues that explanations focusing on précipitantsresemble the (nowwidely despised) volcano model of revolutions.13A common objection against this approachhas been comparative,and Wickham-Crowleymounts a similarchallenge to theanalysis of triggeringevents. Are therenot numerous similar incidents which were notfollowed by a "volcanicsocial response?"14Why then should triggers be takenat all seriouslyin explainingrevolutions? Thisobjection remains valid only if the properties of sufficient(or quasi-sufficient) conditionsare ascribedto triggersand theirpotential is judged independentlyof the contextin whichthey are operating.The factthat the same eventdoes not lead to a

255

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April 2009 revolutioneverywhere (or not even in most of the cases) doesnot disprove the role it might assumein a spécifiecontext (for instance, a stolenelection in an electoralauthoritarian system).Instead it might be appropriateto reveal first a trigger'ssignificance for a smaller set of cases by employingbefore-after research designs. In a secondstep it can then be determinedif generalizationsacross time and space are possible.Critics furthermore overlookthat triggers may be relatedto revolutionaryoutcomes as well:their nature can decidewhether a movementtriumphs over the state's security forces and mayinfluence whoprevails if the old ordercollapses. On thesurface, disagreement over the place triggersdeserve in revolutionarytheory stemsfrom disagreement about their actual causal importance.There may well be a methodologicalfear involved as well.Would not bringing triggers back in meanopening thedoor to ever more contingency inthe explanation of revolutions, increasing the number of idiosyncraticcase studies?15Without doubt, any inclusionof short-termevents will resultin less parsimoniousexplanations. But thisdoes notmean that triggering events can notbe understoodtheoretically. Précipitants such as stolenelections can be linkedto well-establishedtheories, above all therich literature on socialmovements. In addition,it is possibleto contextualizetriggering events, that is, to showtheir relationship to longer- termstructures.

StolenElections as TriggeringEvents: A Model

Thisarticle is notthe place fora comprehensivetypology of triggering events or a review of theirrole in a varietyof revolutions.A singleexample may be enoughto cast severe doubton the view that there is nopurpose in studying triggers, as thisposition is expressed in a verygeneral fashion. The pointwill be exemplifiedby offeringa modelof stolen elections- an authoritarianregime trying to reversethe outcome of an electionthat has alreadybeen lost (de factoor at leastin the perception of most people concerned). Stolen electionsgreatly enhance the chances for staging a democraticrevolution in an electoral authoritariansystem. Democratic revolutions are largelynonviolent popular uprisings thattake place withthe goal of bringingabout a democratictransition. Analytically, thisvariable can be splitup intoa revolutionarysituation (the best indicator being mass mobilization)and a revolutionaryoutcome (the accomplishment of regimechange).16 In orderto explainwhy this exemplary trigger is so powerful,it is also necessaryto distinguishbetween different groups of actors: ordinary citizens, opposition activists, and regimeelites. By showingthat in an ideal-typesituation the influence of stolenelections occursat severallevels, the significanceof triggeringevents in overcomingcollective actionproblems can be underlined.

StolenElections and OrdinaryCitizens A stolenelection increases ordinary people's willingnessto participate in antiregimeprotests for two main reasons. On theone hand, it

256

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson conveysa strongmessage that a politicalopportunity for staging an uprisinghas emerged. Thisis importantfor individual calculations of the risks and chances of protesting before decidingto join in.17On theother hand, stolen elections produce deeply felt and widely sharedgrievances. Theperception of a goldenopportunity for rebellion results partly from the impression thatlarge numbers of people are preparedto join antiregimeprotests. The factthat the regimehas lostthe election, often by an overwhelmingmargin, documents exactly how widespreaddissatisfaction has become.Under less thandemocratic circumstances anger maybe widespread,but it mayalso be hiddenfrom view. When the opposition wins an election,private preferences become public, and individualsin societylearn they are not alonein their dislike of the regime. Collective action seems likely to receivebroad-based supportwithin society, reducing the chancesof sanctionsand approachingthe critical massnecessary for toppling the regime.18 A sense thatthe moment of decisionhas arrivedcan also be inferredfrom more directsigns of regime weakness. For reasons considered below, a stolenelection is likely to undermineregime cohesion and lead to open splitswithin the rulingelite. Such a developmentcan spurpopular protest because it indicatesthat the repressive capacities of theauthorities have diminished.A stolenelection also changespeople's impression ofthe top leadership, especially those hardliners who bear responsibility for stealing the vote.The loss ofan electiondeals a seriousblow to therulers by shatteringtheir nimbus ofinvincibility. Electoral authoritarians are oftenskilful tacticians who gain a reputation even amongtheir adversaries for knowing how to play thegame. For them,suffering electoraldefeat is especiallydevastating because it destroystheir reputation as political masterminds.The imageof vulnerabilitymay be fosteredif the regimeis so wrong- footedby the opposition victory that its postelectoral maneuvers look visibly helpless (for example,when delaying the announcement of finalresults). At thesame time,people's senseof political efficacy has increasedbecause it was theywho inflicted the defeat upon theregime.19 Thesubjective emergence of a politicalopportunity is by no meansa purelycognitive affair.It also possessesa strongemotional dimension.20 Feelings of optimismand hope, enthusiasmand joy, empowermentand pridecharacterize the postelectoral mood. The psychologicalimpact will be particularlystrong in societiesin whichthe regime had previouslyappeared unshakable. Having lost the electoral contest, rulers can be zestfully ridiculedlike the emperor without clothes. The secondway in whichstolen elections contribute to societalmobilization is by creatinga setof powerful grievances. Depriving people of their victory at theballot box leads to the disappointmentof expectationsbuilt up duringthe opposition'selection campaign.However naive it mayappear to outsiders,it is strikinghow enthusiastically votersparticipate in suchelections although there appears little chance that the regime would acceptdefeat. Casting one's ballotis nonethelessseen as a promisingway of bringingabout political change. Such hopesappear particularly well foundedunder the conditionsof electoralauthoritarianism, where manipulation is less comprehensiveor at

257

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April2009 leastless visible to the voters. Outrage when the election is stolenis correspondinglygreat. Frustratedhopes set stolenelections apart from many other incidents of authoritarian injusticethat are not preceded by increasingexpectations.21 Moreover,it should already be clearthat postelectoral dissatisfaction is deeply moral in nature.In JamesJasper's phrase, one can speakof "moralshocks" caused by stolen elections.22Moreover, blatant fraud works like a magnifyingglass in that it focuses on the unjustcharacter of the regime. Even if the regime had littlepopular legitimacy before the election,the actual stealing of the result creates a clearmoral front between the people's willand an evilregime bent on thwartingit. Again,this focus is of particularimportance in an electoralauthoritarian setting. Such regimesare notcompletely devoid of formallegitimacy, or at leastpeople do not see themas outrightdictatorships. As theparticipation before stolen elections indicates, institutionalchannels are stillregarded as holdingthe potential for change. A flagrant stealingof the vote, however, rips away the last remnants of regime legitimacy and removes whateverdoubts there may have remained about the rightfulness ofrevolutionary action. Disrespectingthe choice of the electoratedeeply offends people's dignityas citizens, engenderinga strongsense of moralobligation to join proteststhat can outweighthe perceivedcosts of participation or overcomea possiblefree rider dilemma. It is not only the qualityof grievances(fundamentally heightened and morally grounded)that matters after a stolenelection, but also theextent to which these grievances areshared among the population. While authoritarian regimes constantly violate citizens' rights,few incidentspersonally affect most of the populaceat the same time.Stolen electionsdo. Stealingthe results creates an imaginedcommunity of millionsof robbed voters.

Advantagesfor OppositionActivists All of the mechanismselaborated so farcan also be appliedto thenarrower group of oppositionactivists. Yet it makessense to deal withthem separately because of theirspecific task of mobilizingthe population. Stolen electionsfacilitate this effort in tworespects. First, it becomeseasier to createresonant collectiveaction frames.23 Second, in terms of the resource mobilization paradigm, at the momentan electionis stolen,opposition activists already have an effectiveorganizational apparatusat theirdisposal. Revolutionaryactivists offer their own interpretationof the politicalsituation in orderto drawpeople into antiregime protests. The resonanceof socialmovement frames partlydepends upon their credibility. A well-definedand stronglyemotionalizing event likea stolenelection helps in constructingthe right frames. Diagnostic framing, with its attributionof blameto theregime, now coincideswith people's first-hand experience of electoralmanipulation. The same holdsfor the opposition's attempts to conveya sense of urgency.Quick reactionis requiredbefore the windowof opportunity,temporarily openedby theelection, closes again.The opposition'sshopworn talk about the need for urgentaction suddenly sounds very convincing. Electoralcampaigning improves the opposition'sorganizational capabilities. It is

258

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson rarelyenough just to have a convincingmessage against an adversarywho enjoysvast structuraladvantages. The oppositionalso needs to forma broad coalition,choose a commoncandidate, redouble its efforts to reach out to voters, and engage in poll-watching activities.Building a broadelectoral coalition can be extremelydifficult for opposition parties,and divide-and-rule strategies are common tools of electoral authoritarians. But it maybe evenharder to forgea unitedfront seeking revolutionary change. As longas the systemis notentirely closed, factions of the opposition may opt for more moderate means ofchallenging the regime (not to mentionthe possibility of their being coopted). Once an electionis called,pressure for opposition unity rises, and those who do notjoin a coalition maybe marginalizedduring the electoral campaign. But even if divisions persist, a broad- basedcoalition is likelyto emergeafter the elections as theregime becomes the object of popularprotest that has entereda revolutionaryphase. Oppositionunity often finds its expression in theselection of a singlecandidate who runsfor president or primeminister. Campaigning creates strong ties betweenhim or herand the voters, particularly through personal campaign appearances since opposition presencein themedia is oftenseverely limited. By thetime of the election a clearleader figurehas emerged.Door-to-door canvassing efforts by volunteersand a pushto getout thevote on electionday further contribute to theopposition's organizational capabilities. Perhapsmost important, the opposition often tries to createa networkof independent electionobservers to comeup witha votecount independent of thegovernment's. Not onlyis thistask crucial in showingthat the elections have in factbeen stolen, but it also generatesa networkof activiststhat can be mobilizedfor further collective action after theballoting. In short,efforts to achievevictory at theballot box enhancecapacities for postelectoralmobilization as well. A similarpoint can be made regardinginternational assistance. Foreign support is oftencrucial in overcomingthe tremendous hurdles set up by an electoral-authoritarian regime.By aiding the construction ofcampaign and poll-watching infrastructures, external actorssimultaneously prepare the ground for ensuing protest movements. Under normal circumstancesit is extremelydifficult for outsiders to promotepolitical participation amongan apatheticcitizenry (let alone to encouragethe emergence of a powerfulpopular movement).Yet when foreign assistance is targetedat electoralcontests, it can contribute to the outbreakof massiveprotest movements. It is no coincidencethat international actors,who want to stemthe current tide of electoralauthoritarianism, have increasingly beendrawn to electionsas focalpoints. Ultimately,however, it dependson domesticcircumstances whether a strategy of promotingelectoral revolutions from abroad succeeds.Moreover, the impression of massiveexternal interference in recentcolorful revolutions (often exaggerated to thepoint that foreign forces are depictedas pullingthe strings)may provokedrastic authoritariancountermeasures. As a consequence,electoral playing fields have become less competitive,and promotion of international democracy a moredifficult endeavor.24

The Impacton theRegime Elite Stolenelections are not only triggers in society.They

259

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April2009

can also lead to openrifts among regime elites. Elite defections are importantfor several reasons.Citizens perceive them as an indicatorof regimeweakness, which in turnwill spurtheir willingness to engagein acts of protest.Mobilization will also be facilitated oncethe state-controlled media machinery crumbles from within. Last but not least, elite divisionsundermine the ability to repressmass mobilization.Even whenits legitimacy is low,a regimecan holdonto power by retaining the loyalty of key civilian officials and militaryelites. But a stolenelection is likelyto undercut this loyal behavior sufficiently to weakenthe regime severely. As a consequence,not only the probability for a revolutionary situation,but also thelikelihood of a revolutionaryoutcome increases Whydo stolenelections put the loyalty of the regime staff to thetest? One reasonis thatthe unjust character of theregime has becomeobvious to itsown members.A loss ofelite legitimacy may be crucialfor a peacefuluprising, as examplesfrom Communist easternEurope demonstrate.25 Ideology is farless importantin electoralauthoritarian systems,but upholding certain democratic standards is ofgenuine concern for at least some membersof theelite. Having long cooperatedwithin a pseudo-democraticframework, theybecome disgusted by the blatant disregard of the people's will and sidewith regime opponentsdue to genuinefeelings of shame. More frequently,however, defections from the ruling circle occur out of strategic calculation.The electionshowed that the currentregime lacks legitimacyand thatan alternativecenter of power has emerged,backed by a largepart of thepopulation. Now thatthe regime has been openlyquestioned and is undergreat pressure, beneficiaries of thecurrent system have to takea stance:either to stickwith a discreditedleadership or abandonit. Many prefer to changesides before it is too late.On theother hand, parts of theelite may have been plottingways of oustingthe leadership for some time already. Theymay have neverhad an opportunityto translatetheir plans intoaction, however. The weakeningof the regimeinternationally and domesticallyafter a stolenelection signalsthat the time to takeaction has come.Simultaneously, itbecomes harder for these groupsor individuals to assumepower by themselves. It is difficultto replacethe current leadershipin orderto installanother dictatorship since a newopposition center of power, backedby an unambiguouspopular mandate, has surfaced.In sum,it may be said thata stolenelection has an enablingand constrainingeffect on regimedefections. The main argumentspresented here are summarizedin Figure1 .

The StructuralContext of StolenElections as Triggers

The discussionthus far has shownthat a multitudeof causal mechanisms can be specified whenmodeling the triggering effect of a stolenelection. In addition,theory building is possiblethrough locating triggering events in theirstructural environment. Examples of theevent- structure relationship between stolen elections and electoralauthoritarianism havealready been mentioned, but in the following the focus shifts directly to the question whythis specific kind of election fraud helps to ripensuch regimes for revolution.

260

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson

Figure 1 StolenElections as RevolutionaryTriggers

StolenElection

Oppositionactivists Citizens Regime electioncampaign outrageand perception of splitdue to shame, promotesopposition opportunityenhance opportunistic unityand strengthens willingnessto participate calculations;palace coup organizationalcapacities; inanti-regime protests «4 inface of well-organized stolenelection facilitates anti-regimemobilization anti-regimeframing unlikelyoutcome

RevolutionarySituation / I / RevolutionaryOutcome

Fromone pointof view,to stagean uprisingin an electoralauthoritarian regime appearsto be a comparativelyeasy task. Unlike in "harder"authoritarian regimes, there are no intricatecommunication problems to be overcome,and the fearof suffering fromrepression is muchlower for would-be protesters. Such relativelyopen political environmentsprobably formed a necessarybackground condition for the postelectoral uprisingsconsidered in thisarticle. On theother hand, many recent studies of revolution claim thatrevolutions occur only in highlyexclusionary political contexts and/or regimesthat resort to violent,arbitrary repression. People are forcedto rebelbecause rulersshow no willingnessto reform,offer no channelsfor political participation, and/ or indiscriminatelyrepress political opponents. Sultanistic dictatorships (like Nicaragua underSomoza) or posttotalitarian Communist rulers defying reform (such as Honeckerin East Germany)had to be broughtdown through popular uprisings, while in neighboring countriesinsurrections were not encompassingenough to achieve a revolutionary breakthroughor regimeswere transformed through negotiation between softliners who dominatedthe government and a moderateopposition.26 More open or institutionalized systemsare likelyto divertrevolutionary pressure or avoid its emergencein the first place. ParaphrasingLeon Trotsky,people resortto revolutionaryaction only if thereis "no otherway out."27 Jeff Goodwin claims that "[e]ven imperfect and poorly consolidated democraciestend to diffuserevolutionary pressures."28 It followsfrom this argument that electoral authoritarian regimes are rather resistant to revolutionssince they allow relativelyextensive political pluralism and outrightacts of repressionare notcommon. Most important,elections, while not providing an even playingfield, are held on a regularbasis and often develop into fierce contests. Opposition

261

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April2009 partiesmay find it hard to forma broadrevolutionary coalition under such circumstances. A plethoraof them typically exists, and somefactions may simply prefer to vie forpower throughinstitutionalized channels. With elections still an option,it is also moredifficult to drawbroader support from society. Citizens may preferto wait forthe opposition to offera crediblealternative at the ballotbox ratherthan engage in relativelycostly protestactions with questionable prospects of success.Finally, as was said before,even manipulatedelections bestow some degreeof legitimacyon theregime, which makes attemptsto stagean overthrowlook like an insurrection. By stealingan electionoutright, however, electoral autocrats change the character of theirrule abruptly. The uncertainty produced by the political competition still going on in thesekinds of regimesis replacedby thecertainty that change through institutionalized channelsalone will be impossible.Political closure through a stolenelection leads to the realizationthat there is "no otherway out" thanrevolutionary action. The regimehas strippeditself of anysemblance of democraticlegitimacy, and an uprisingmay now be judgeda "revoltin defenseof the constitution."29

The Importanceof Stolen Elections as Triggersin Serbia,Georgia, and Ukraine

The theoreticaldiscussion has shownhow stolenelections can triggersudden and significantbehavioral changes in threegroups of actors:citizens, activists, and members ofthe regime. It was also arguedthat relatively open regime environments tend to inhibit oppositionmobilization, but circumstances alter dramatically when incumbents deprive theopposition of victory by stealing elections. Yet how can itbe demonstratedthat stolen electionswere in factof overridingimportance for ousting electoral autocrats in Serbia (September-October2000), Georgia (November 2003), and Ukraine (November-December 2004)? One possibilitywould be to offera thickanalytic narrative by corroboratingthe causal mechanismsoutlined above withempirical evidence. As such detailedprocess- tracingwould go beyondthe scope of thisarticle, an attemptwill be made insteadto establishthe centrality of stolen elections by comparing stolen elections to otherpotential triggers.If thejust-the-final-straw perspective on stolenelections were correct,these previousevents should have likewiseled to revolutionaryexplosions. Indeed, several suchwould-be triggers occurred in these three countries: attacks on theantiregime media, theassassination of regime critics, and, in thecase of Serbia,a lostwar. Yet they did not lead to popularuprisings against the regime. Itmay be objectedthat such a juxtapositionof stolen elections with triggers that failed is misleading.Instead of viewingthese events in isolation,it mightbe arguedthat each was a crucialstep toward revolution. Stolen elections appear more powerful only because theybenefited from a cumulativebuild-up of revolutionarypotential. More generally, longitudinalcomparisons that highlight the importance of a particularfactor run the risk of overlookingthe fact that other factors have changedover time as well. In answerto

262

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson thisobjection, two alternativeexplanations derived from analyses of otherauthors will be considered:the possibility, first, that in successfuluprisings regime opponents could capitalizeon previousmobilization efforts and, second, that the ruling clique had started to fragmentin the meantime. On thefirst potential trigger, assaults on mediacritical of theregime, an important featureof most electoral authoritarian systems is the existence of independent oropposition- leaningmedia witha considerableaudience. In orderto keep theirinfluence under control,rulers usually employ "more subtle means of repressionthan their counterparts inauthoritarian regimes."30 But occasionally they may feel compelled to resort to outright repression.Because open attacksmean a departurefrom the conventional pattern, they maybe interpretedas majorinfringements on political liberties and thusprovoke public outrage.One highprofile incident occurred in Serbiajust monthsbefore the ouster of Milosevic.In May 2000 police raidedan importantopposition broadcasting center in Belgrade.Yet protests against this blatant crackdown were short-lived and did not reverse thetrend of decliningparticipation in theanti-Milosevic movement. In Georgiaa raidof the independentRustavi-2 television company in late 2001 attractedseveral thousand people to antiregimedemonstrations. This numberis not negligibleif it is takeninto accountthat the revolutionary movement in Tbilisiin 2003 was muchsmaller than the movementsin Belgrade and Kiev. Nevertheless, the protests did not develop the dynamic necessaryto toppleShevardnadze. In the followingyears, Georgia's president sided openlywith the hardliner faction of his regime. Attackson themedia overlap with another possible catalyst of antiregime action, the murderof regime critics. Critical journalists are often targeted for assassination. Among the cases consideredhere, the murder of Ukrainian investigative reporter Georgiy Gongadze in September2000 standsout. Leaked audiotapes that implicated President Kuchma and otherhigh-ranking officials in the killing sparked off mass protests in early 200 1 . Although theregime emerged badly weakened from the crisis, the protest eventually petered out. Themovement resurfaced in the following years, but again demonstrators did not succeed in gettingrid of Kuchma.In thetwo other countries considered, assassinations fell short of havinga similarmobilizational effect. The slayingsof Slavko Curuvija,a Serbian newspaperowner and editor,in April 1999 and of Georgiantelevision anchor Giorgi Sanaia in July2001 wereboth believed to be politicallymotivated. Several thousand people attendedthe funerals,but no protestserupted. In a similarvein, assassination attemptsagainst Serbian opposition leader Vuk Draskovic(in October1999 and June 2000) failedto boostthe ailing anti-Milosevic movement. Serbia's strongman,however, had to survivea crisisthat leaders in Georgiaand Ukrainedid not face: a lostwar. When Milosevic agreed to NATO's peace termsin June 1999,it meantBelgrade's de factoloss of Kosovo. Whathad remainedof Milosevic's nationalsavior image was irrevocablyshattered. The economicoutlook was bleak as well. Yet forover a yearthe oppositionwaged an increasinglydesperate struggle to unseatthe regime. Daily marchesand ralliesstarted in September,but the number of demonstratorssoon declined, and the movement ultimately fizzled out in midDecember

263

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April2009

1999.The patternrepeated itself when the opposition started another round of protests in spring2000. Mostcitizens responded with resignation rather than political activism to the catastrophicresults of Milosevic's policies. Allthese examples indicate stolen elections were a muchmore powerful revolutionary trigger.What accounts for this difference? First, in thecase of politicalassassinations it is oftendifficult to establisha clearchain of responsibility (even as ofthis writing, courts have failedto uncoverwho commissionedsome of the crimesmentioned above). By contrast,the attribution of blamein case of a stolenelection is a straightforwardmatter. Second,even whereit is obviousthat responsibility for outrageous acts lies withthe regime,not all partsof societymay feelpersonally affected. For instance,media that becomethe target of government attacks frequently do notreach beyond a certainregion (forexample, the May 2000 raidagainst oppositional media in Belgradewas a localized event,whereas a coupleof months later hundreds of thousands from all overthe country flockedto thecapital to defendtheir vote). Third, among all theevents considered, only stolenelections allowed citizens and oppositionists to drawclear inferences about regime vulnerability.These inferencesin turnaffected the calculationsof potentialdefectors fromthe regime. It was obviouslynot the right moment to changesides when large-scale protestshad failedto takeplace. Fourth,none of the alternative triggers was precededby risingexpectations that could then have been disappointed. By contrast,voters in Serbia andUkraine looked upon the impending presidential elections as watershedevents. Finally, incidentsnot related to large-scaleelectoral fraud, however despicable, did not constitute a descentinto full authoritarianism. Sometimes rulers could even make concessions to the public(for example, by settingup "criminalinvestigations" or by dismissinga cabinet memberas a scapegoat).Such attempts to deflectpopular anger were not an optionafter stolenelections, however, because the vote had raisedfundamental questions about the distributionof power.Depriving the opposition of itsvictory meant that semiautocratic rulersno longercould reap the benefits of retaining a relatively open political order. One exampleof theimportance of suchelectoral authoritarian payoffs can be foundin Paul D'Anieri'sobservation that "protesters living in tentsin centralKyiv were able to force theresignation of the Ukrainian prime minister in 1990,when the country was stillruled by theSoviets, but the same tactics... failed in 2002, in largepart because Kuchma... has a strongelectoral basis forhis legitimacy."31In other words, before stolen elections in 2004 theelectoral authoritarian system established in Ukraine was openenough to absorb popularpressure even as itfaced the severe crisis prompted by the Gongadze murder. Thevarying impact of events may be rootedelsewhere, though. Arguably, itcould be tracedbacked not so muchto their intrinsic features as toa changein external circumstances. In all threecases, learningprocesses had occurredprior to the massivepostelection mobilizations,and (especially in theform of youthmovements) had been graduallygaining strength. In Ukrainethis tendency was visiblesince the "Kuchmagate" scandalin late2000. In thewords of TarasKuzio, "[t]heexperience of popularprotests duringthose four years prepared and equipped Ukraine's opposition members to lead the OrangeRevolution."32 Yet because suchprotests had repeatedlyfailed, one of themost

264

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson

importantlessons taken to heartby oppositionactors was thatobstinate rulers had to be challengedthrough elections. Moreover, many of theskills, resources, and networks thatproved so valuablein organizingthe eventual uprisings were only developed in the electioncampaigns.33 Finally, it wouldbe problematicto subscribeto theview thatan experienced,organized, and well-fundedcivil societycould somehow"make" these revolutions.In orderto bringdown unyielding rulers, activists needed to mobilizemany morepeople than they themselves could muster, and much of the revolutionary rank and fileseem to have made their decision to join ina spontaneousmanner. For example, when Milosevicdenied the opposition its victory in theSeptember 2000 presidentialelection, workersat theKolubara coal minedecided to takethings in theirown handsand went on strikewithout having received any clear instructions.34 To be sure,Serbia's organized oppositionhad calledfor a nationwidework stoppage, but similar appeals had fallenon deafears in themonths before. In hindsight,itis temptingto viewsuccessful democratic revolutions as havingbeen inevitable.It is easilyforgotten how pessimisticallythe prospects for protest activities werejudged not long before stolen elections took place. Opinion polls carried out in Serbia justmonths before indicated that most citizens did not view street protests as a usefultool forbringing about political change (this finding helps explain the muted, if not apathetic, responseto oppositionleaders who organizeda seriesof protestsbefore elections had beencalled).35 Sociologists who carried out preelection opinion polls in Ukrainein 2004 remainedskeptical about whether electoral fraud would lead to popularoutrage.36 Even oppositionleaders, who reckoned with the necessity of taking to thestreets, were stunned bythe massive turnout.37 In Georgia two respected and otherwisenot overly pessimistic observersstated only half a yearbefore the Rose Revolutionthat "Georgia today is a differentcountry from what it was in the1980s, when it was possibleto rallythe masses in supportof nationalindependence and democracy:the mass enthusiasmof thosedays will be difficultto recover."38There is no evidenceearlier protests had increasedthe willingnessto takepart in furtherantiregime activities (beyond perhaps a hardcore of regimeopponents). One can even go a stepfurther and arguethat previous mass mobilization,despite enhancingorganizational skills and capacities,actually discouraged further protest. Serbia,with its decade-longhistory of protests,is a case in point.By the late 1990s Milosevichad outlivednumerous protest movements, and citizenshad grownextremely skepticalof ralliesand demonstrationsas a meansof affectingpolitical change. Even afterthe Kosovo war, when a certainmomentum was regained,it did not take long until a similarpattern emerged: opposition protest rallies became smaller and smaller as citizens grewdisillusioned by previousfailures. Longer-term mobilization periods can lead to growingpessimism and protestfatigue instead of strengtheningpeople's determination to bringa regimedown. It is thusdifficult to arguethat stolen elections evoked a morepowerful response than priorincidents because they tapped into more fertile ground. But perhaps the chances for a revolutionhad improved because the regime had grown more fragile. It has indeedbeen

265

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April2009

frequentlyremarked that the three rulers of the countries discussed here were very badly weakenedby the time they faced popular revolts because of increasing divisions within the regimeelite. Thus, the postelection mobilization against Shevardnadze was spearheaded by formermembers of his Citizens'Union of Georgia.Their withdrawal from the ruling partyin 2001 and2002 was decisivefor the poor showing of progovernment forces in the 2003 elections.At the same time, it is highlyimprobable that these politicians would have beencapable of coordinating sustained street protests prior to the elections - afterall, they did noteven manage to forman electoralalliance. The impactof regimefragmentation is mostfrequently invoked in the case of the OrangeRevolution. With the loyaltyof Ukraine'soligarchs to theregime in doubt,several former regime officials now leading oppositionists,and membersof thestate apparatus favoring the opposition, the regime seemedunstable even before the vote in 2004.39 Elite support was a preconditionnot only forwinning elections against a well-oiledpolitical machine, but also forthe success of theOrange Revolution itself. On theother hand, a closerlook revealsthat many crucial riftsin theregime appeared only after the votes had been cast. In thewords of Henry E. Hale, Ukraine's"elite stand-off gradually began to resolveitself after the first and secondrounds of the election took place, giving elites who were hedging their bets more informationabout who was likelyto win."40 Defection is nota decisionto be takenlightly, evenwhere loyalty remains fragile. In Serbiathere were few defections at all beforethe September2000 vote,and whilethere were certainly tensions within the inner circle of power,most observers did notforecast its breakup in theimmediate future. Only when Milosevic'sdefeat at thepolls seemedto spell disasterdid theheads of his extensive securityapparatus decide to jump ship.Finally, regime fragmentation does notsay very muchabout how rulers are goingto lose power- a massivepopular uprising, an internal coup,or somethingelse. But themode of transitionmay, of course,strongly influence futurepolitical developments.

OtherStolen Elections That TriggeredRevolutions

The triggerof stolenelections greatly facilitates the emergenceand breakthroughof revolutionarymovements in electoralauthoritarian settings. In Serbia,Georgia, and Ukraine,only a shorttime before the stolen elections, other events had failedto triggera similarpopular response. While this combination of factors certainly does not amount to a sufficientcondition, the underlying theoretical reasoning points to a broaderset of cases, and it wouldbe instructiveto see if otherstolen elections, in similarregime contexts, werealso followedby a regimechange from below. Additional positive cases wouldlend furthersupport to theargument, whereas stolen elections without a subsequentuprising wouldrequire that the scope conditions of this trigger be specific. Thereare at least two more electoral authoritarian regimes that experienced an uprising aftera stolenelection. In thePhilippines under Ferdinand Marcos, mass mobilization had

266

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson been underwayfor several years already before he was finallyousted in 1986. But the impressiveprotest movement set off by the assassination of opposition leader Benigno S. Aquino,Jr. in 1983failed to achieve its objective of getting rid of Marcos. Not long before thecrucial election Marcos seemed as firmlyin poweras ever.41With hope in a peaceful uprisingdecreasing, the initiative shifted to moreradical organizers. If Marcoshad not announcedelections at all, theremight have neverbeen a democraticrevolution, and a Communistinsurrection might have succeededin removinghim frompower instead. Anotherconsequence of Marcos' downfall in thewake of a stolenelection was thatarmy officerswho started to rebelwere unable to establisha militaryregime as theyhad hoped. Theywere too dependenton thepowerful position that opposition candidate Corazon C. Aquinohad acquiredby de factowinning the election. Thus, this case notonly reaffirms theview of stolen elections as beingmore effective than other triggers. It also showsmore clearlythan previous examples that a seeminglysmall event may have a decisiveimpact on whois goingto replacethe ancien régime. Almostcompletely overlooked by studiesof electoralrevolutions is Madagascar, wherein 2002 a month-longuprising brought about the removalof presidentDidier Ratsiraka,another unrepentant election loser. It was not the firsttime that Ratsiraka was forcedout of officeby a massiveprotest movement; a month-longgeneral strike effectivelyended his regimein 1991. However,Madagascar's young democracy never consolidated,and onlya coupleof yearslater Ratsiraka returned to power.Despite the authoritariansetback, few people were willing to taketo thestreets once again.Thus, all themore astonishing was theoutpouring of public anger in 2002 afterRatsiraka tried to forcehis challenger,Marc Ravalomanana, into a secondround of presidentialelections, despiteevidence that his opponent had alreadygarnered the majority of votes. If the Philippinesand Madagascarare added to the threeexamples from ,it becomes obvious that stolen elections turned weak or extremelyrevolutionary potentialsinto mighty popular insurrections in quite diverse environments. Furthermore, powerfulmovements in defenseof election victories seem to haveemerged even without theinternational diffusion effects that certainly helped inspire the opposition in Georgia andUkraine.42

The Dog That Didn't Bark: StolenElections That Did Not TriggerRevolutions

Are therecases of electoralauthoritarianism in which a rulerbrazenly snatched away an electoralvictory from the handsof the oppositionand managedto stayin power nonetheless?In general,the universe of stolenelections is muchsmaller than it might appearat firstglance. It is importantto recallthat the argument advanced in thisarticle is foundedon a narrowdefinition of stolenelections. Lesser forms of electionfraud are disregardedbecause some of the most crucial causal mechanismsoutlined above pertain onlyto a situationin whichthere is widespreadperception that a regimeactually lost at

267

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April2009

thepolls. The factthat recent electoral shenanigans by authoritarian rulers in Azerbaijan, Armenia,and Belaruswere not followed by democraticuprisings does notchallenge the role of large-scalefraud in themaking of theRose or Orangerevolutions because the definitionof stolen elections used here only applies to thelatter cases. Ratherthan seeing thefailure of postelectoralprotests in Georgia'sand Ukraine'sneighbors as proofthat stolenelections were not of overridingimportance, it is moreplausible to arguethat the differencebetween stolen elections and electoralfraud that falls short of this category is a keyexplanation of thesevarying outcomes.43 Other cases (suchas Kenyain 1997 and Zimbabwein 2002) thatmay possibly be invokedto questionthe significance of stolen electionsalso fallshort in thesense of the term used here.44 Thereis one clear-cutexample of a stolenelection that did not lead to theoverthrow of an electoralauthoritarian regime, and interestinglyit occurred in one of the cases consideredinitially - Serbia.In November1996 Milosevicdeprived the opposition of a numberof victories in communalelections. The massiveprotest that occurred as a result stronglysupports the argument about the dynamic unleashed by suchan act,especially sincein the months before other potentially explosive events had failedto incitea similar response.45But Milosevicmanaged to stayin powerfor almost four more years. It is ratherobvious why there was no revolutionaryoutcome then: the demand that unified the protesters- the recognition of theopposition victories in severaltowns and cities- did notfocus on thecountry's most important, that is, national,political institutions. When Milosevicfinally gave in, the movementquickly disintegrated. It is thusnecessary to add a qualificationto thetheory. If an electionis limitedin scope (forinstance, for local officesonly), an autocraticruler may well reverse his decision of stealing, concede defeat, and stayin power nonetheless. To be sure,there are also instancesof key electionsbeing stolenthat have not set offdemocratic revolutions. But theyhave happenednot in electoral,but in "hard" authoritariansettings. In Panama(1989), Burma(1990), Algeria(1991-92), andNigeria (1993), themost prominent examples since the late 1980s,stolen elections left a lasting impact.They led to thereshuffling of regimeelites in Algeria and Nigeria, hastened the U.S. invasionof Panama,and forcedthe Burmesemilitary to keep tightcontrol over electionwinner Aung San Suu Kyi. But itwould be farfetchedto speakof revolutionary consequences.Peaceful mass protestsstaged in Panamaand especiallyNigeria turned out to be eithershort-lived or of limitedreach. The brutalcivil war unleashedby the refusalto allow an Islamistvictory in Algeriaalso can not be likenedto the sortof popularmobilization considered in thisarticle. In Burmano majorpostelectoral protests tookplace. Electionsin thesecountries that fall well shortof theelectoral authoritarian regimetype were preceded by onlya briefphase of politicalliberalization. They lacked (at leastfor a decade) an electoraltradition that would have producedexpectations of minimumstandards for elections to whichincumbents must adhere. Outrage in electoral authoritarianregimes, by contrast,is arguablygreater after stolen elections because such standardsexisted there. Elections (despite repeated manipulation) were seen as a vested rightthat simply could not be takenaway.

268

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson

A moreimportant point, though, is thatintimidation was muchmore prevalent in these"harder" authoritarian regimes (all are,or were, military dictatorships). In Panama, Noriega's DignityBattalions moved with bruteforce against protesting opposition politicians.Repression in Algeria was evengreater. But sometimes mere fear of reprisal, based on theexperience of past repression,is enoughto preemptpopular protest. The Burmeseexample is themost notable in this respect: in 1990people remained intimidated bythe horrendous massacre carried out by the armed forces against protesters two years earlier. Thesenegative examples thus underline the importance of an electoralauthoritarian - andthereby civilian - contextfor the operation of stolen elections as triggers.On theother hand,the fact that a legacyof political violence may easily ward off postelectoral protests shouldnot lead to theconclusion that it was actuallythe relative openness of electoral authoritarianismthat produced impressive uprisings from Ukraine to Madagascar.This moresubtle form of authoritarian rule tends to generate its own collective action dilemmas, makinga successfulpopular overthrow by no meansa foregoneconclusion.

Conclusion

Paul Piersonhas recentlyissued a plea thatmore attention be paid to the"slow-moving dimensionsof social life"instead of focusingon "triggeringor precipitatingfactors."46 Whenit comesto thestudy of revolutions,however, short-term causes in theform of eventsmay not have been studied enough. In thisarticle, the example of a specifickind of electoralfraud disputes the still widely held view that short-term catalysts are negligible incomparison with the long-term developments that precede them. A rangeof theoretical approachesfrom social movementstudies can be fruitfullyemployed to showwhy this particulartrigger should be so powerful.In the contextof electoralauthoritarianism, stolenelections assume a particularlysignificant role. As neoinstitutionalisttheories of revolutionsuggest, the relative openness of electoralauthoritarian regimes inhibits mass protest.But when such leaders steal elections, their rule undergoes closure, increasing the probabilityof successful mass protest. Stolen elections are triggering events that suddenly transformthe structural setting in favorof revolutionary efforts. For Serbia,Georgia, and Ukraine, the relevance of stolenelections was underscored whenearlier potential events failed to bringabout a revolutionaryendgame. While these otherevents were not insignificant and weakenedthe regimes, they did notnecessarily make the emergenceof a revolutionarysituation more likely.Even when additional variables(like previousmobilization efforts and increasingregime fragmentation) are takeninto account, it is difficultto discernhow they should have fundamentally altered the chancesof revolution (this conclusion becomes especially apparent when the widespread apathyprior to stolenelections is recalled).The inclusionof othercases furthersupports thenexus between stolen elections and an electoralauthoritarian context.

269

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April2009

Undoubtedly,the analysis of a singletype of triggeringevent, unfolding in a rather specificpolitical environment, does notallow inferencesto be drawnabout other kinds of triggers,contexts, revolutions, or, more generally, protest movements. There might well be instancesof popularupheavals where the nature of thetriggering event was of lesserimportance. But this question needs to be answeredempirically. If theanalysis of theexemplary catalyst of stolenelections is correct,it revealsthat triggers can notbe dismissedout of handby merelycontending that structural factors carry overwhelming explanatoryweight.

NOTES

This articlewas originallypresented to the"Conference on AuthoritarianRegimes - Conditionsfor Stability and Change"at theSwedish Institute in Istanbulin May 2005. In particular,the authors would like to thank JasonBrownlee, Axel Hadenius, Andreas Schedler, Dan Slater,Richard Snyder, Lucan Way, and the anonymous reviewersfor Comparative Politics. 1. An overviewof this research agenda can be foundin Andreas Schedler, ed., Electoral Authoritarianism: The DynamicsofUnfree Competition (Boulder: Lynne Rienner, 2006). 2. MarkR. Thompson,Democratic Revolutions: Asia and EasternEurope (: Routledge, 2004). 3. "Has theAge ofRevolutions Ended?," Transitions Online, Jan. 13, 2005. 4. The potentialfor stirring up popularanger by merely alleging an electionwas stolenbecame apparent after Mexico's2006 presidentialelection. In theabsence of compelling evidence, however, neither streets protests nor a self-proclamationas the legitimate president could prevent the official winner from being installed. 5. See, forinstance, Lyford P. Edwards,The Natural History of Revolution (Chicago: University of Chicago Press,1970); ChalmersJohnson, Revolutionary Change (Boston: Little Brown, 1966). 6. BarringtonMoore, Social Originsof Dictatorship and Democracy:Lord and Peasantin theMaking of the ModernWorld (Boston: Beacon Press,1966); JefferyM. Paige,Agrarian Revolution: Social Movementsand Export-Agriculturein the UnderdevelopedWorld (New York:Free Press,1975); Theda Skocpol,States and Social Revolutions:A ComparativeAnalysis of France, Russia, and China(Cambridge: Cambridge University Press,1979). 7. CharlesKurzman, "The Post-StructuralistConsensus in Social MovementTheory," in JeffGoodwin and JamesM. Jasper,eds., Rethinking Social Movements:Structure, Meaning, and Emotion(Lanham: Rowman and Littlefield,2004), pp. 111-20. 8. WilliamH. Sewell,Jr., "Historical Events as Transformationsof Structures: Inventing Revolution at the Bastille,"Theory and Society,25 (December1996), 841-81; Doug MeAdam, Sidney Tarrow, and Charles Tilly,Dynamics of Contention(Cambridge: Cambridge University Press, 2001); JamesM. Jasper,The Art of MoralProtest: Culture, Biography, and Creativityin Social Movements(Chicago: University of Chicago Press, 1997). 9. Sewell,p. 862. 10. Jean-PierreReed, "Emotions in Context:Revolutionary Accelerators, Hope, Moral Outrage,and Other Emotionsin theMaking of Nicaragua's Revolution," Theory and Society,33 (December2004), 653-703. 11.Michael S. Kimmel,Revolution: A SociologicalInterpretation (Cambridge: Polity Press, 1990), pp. 9-10. 12. EdgarKiser and MargaretLevi, "UsingCounterfactuals in HistoricalAnalysis: Theories of Revolution," in PhilipE. Tetlockand AaronBelkin, eds., CounterfactualThought Experiments in WorldPolitics: Logical, Methodologicaland PsychologicalPerspectives (Princeton: Princeton University Press, 1996), p. 204. 13.Timothy Wickham-Crowley, "Structural Theories of Revolution," in John Foran, ed., Theorizing Revolutions (London:Routledge, 1997), p. 68 n27.

270

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions PhilippKuntz and MarkR. Thompson

14. Ibid.,pp. 47^8. 15.An earlierstatement by Harry Eckstein on thestudy of "internal wars" reflects this position: preconditions are seen as "amenableto systematicinquiry," précipitants "as defy[ing]systematic inquiry." Harry Eckstein, "On theEtiology of InternalWars," History and Theory,4 (February1965), 140. 16. Applyingthe label "regime change" to theevents discussed here will not go uncontested.The very features of electoralauthoritarianism place thiskind of regimeclose to the categoryof proceduraldemocracy, thus obscuringthat a realtransition has occurred.And althougha popularuprising does notguarantee democratic consolidation,itopens the possibility of democratization in thefirst place. 17. On theconcept of political opportunity structures, see Goodwinand Jasper, eds. 18. Preferencefalsification poses a particularobstacle to collectiveaction in highlyoppressive contexts, but also in electoralauthoritarian regimes the fullextent of politicaldissatisfaction may not be known.On the conceptof preferencefalsification, see TimurKuran, Private Truths, Public Lies: TheSocial Consequencesof PreferenceFalsification (Cambridge, Mass.: HarvardUniversity Press, 1995). 19. The frameworkso farcould easilybe restatedin thelanguage of rationalchoice theory. Cf. JoshuaA. Tucker,"Enough! Electoral Fraud, Collective Action Problems, and Post-CommunistColored Revolutions," Perspectiveson Politics,5 (September2007), 535-51,for such an approach. 20. On therole of emotions in popular protest, for example, see JeffGoodwin, James M. Jasper,and Francesca Polletta,eds., PassionatePolitics: Emotions and Social Movements(Chicago: Universityof Chicago Press, 2001). 2 1. The logicbehind this argument is derivedfrom relative deprivation theory. il. Jasper,p. ιυο. 23. On trameanalysis, see HankJohnston and John A. Noakes,eds., r ramesoj Protest:Social Movementsand theFraming Perspective (Lanham: Rowman and Littlefield,2005). 24. ThomasCarothers, "The Backlashagainst Democracy Promotion," Foreign Affairs, 85 (March-April 2006),55-^9. 25. See, forexample, on thecase of theGerman Democratic Republic, Daniel V. Friedheim,"Democratic Transitionthrough Regime Collapse" (Ph.D. diss.,Yale University,1997). 26. Thebest account of this argument is JeffGoodwin, No OtherWay Out: States and Revolutionary Movements, 1945-1991(Cambridge: Cambridge University Press, 2001). 27. Ibid.,p. 26. 28. Ibid.,p. 303. 29. GhiaNodia, "Breaking the Mold of Powerlessness: The Meaning of Georgia's Latest Revolution," in Zurab Karumidzeand JamesV. Wertsch,eds., Enough! The Rose Revolutionin theRepublic of Georgia(New York: Nova SciencePublishers, 2005), p. 99. 30. StevenLevitsky and LucanA. Way,"The Rise of CompetitiveAuthoritarianism," Journal of Democracy, 13 (April2002), 58. 31. Paul D'Anieri,"Leonid Kuchma and the Personalization of the Ukrainian Presidency," Problems of Post- Communism,50 (September-October2003), 59-60. 32. TarasKuzio, "The Opposition'sRoad to Success,"Journal of Democracy, 16 (April2005), 129. 33. Cf.Joerg Forbrig and Pavol Demes, Reclaiming Democracy : Civil Society and Electoral Change in Central and EasternEurope (Washington. D.C.: The GermanMarshall Fund of the . 20071 34. DarkoMarinkovic, "Strike at Kolubara- a Case Study,"South East EuropeReview, 6 (2003), 41-71. 35. BetweenDisappointment and Hope- Public Opinionin Serbia,September 1999 (Belgrade:Centar za ProucavanieAlternativa. 1999). 36. IhorZhdanov, Oleksandr Lytvynenko, and Yuriy Yakymenko, "Ukraine before Elections - WhatWill It Be Like?,"National Security & Defence,6 (June2004), 2-14. 37. Kuzio,p. 125. 38. David Usupashviliand GhiaNodia, Electoral Processes in Georgia:Attempts to EstablishDemocracy in Georgia(Stockholm: IDEA, 2003),p. 14. 39. See, forinstance, Lucan A. Way,"Kuchma's Failed Authoritarianism," Journal of Democracy, 16 (April 2005), 131-45. 40. HenryE. Hale, "RegimeCycles: Democracy, Autocracy, and Revolutionin Post-SovietEurasia," World Politics,58 (October2005), 154; Way,pp. 143-44;Kuzio, p. 128. 41. See MarkR. Thompson,The Anti- Marcos Struggle: Personalistic Rule and DemocraticTransition in the Philippines(New Haven:Yale UniversityPress, 1995), ch. 7. 42. On diffusiondynamics in thepostcommunist world, see MarkR. Beissinger,"Structure and Examplein ModularPolitical Phenomena: The Diffusionof Bulldozer/Rose/Orange/TulipRevolutions," Perspectives on

271

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions ComparitivePolitics April 2009

Politics,5 (June2007), 259-76. 43. David R. Marples,"Color Revolutions: The BelarusCase," Communistand Post-CommunistStudies, 39 (September2006), 363, makesthis point in an analysisof revolutionaryfailure after the 2006 presidential elections in Reianis 44. The mostrecent Kenyan election, in December2007, does appearto have been stolen,but its violent consequenceshave been shaped by ethnic divisions, a variablenot discussed in theargument developed here. 45. Cf.Florian Bieber, "The SerbianOpposition and Civil Society: Roots of the Delayed Transition in Serbia," Internationaljournalof Politics. Culture and Society.17 ( September2003). 84. 46. Paul Pierson,Politics in Time:History, Institutions, and Social Analysis(Princeton: Princeton University Press,2004), p. 14.

272

This content downloaded from 141.213.236.110 on Tue, 3 Sep 2013 14:52:50 PM All use subject to JSTOR Terms and Conditions