Competitive Authoritarianism Hybrid Regimes After the Cold War STEVEN LEVITSKY Hareard University LUCAN A. WAY University of Toronto Bibliotean Dantel Costo Villega EL COLEGIO DE MEXICO, A.C. CAMBRIDGE UNIVERSITY PRESSContents Ackneledgmenss Aeronyns and Abbreviations Part I: Intraduetion and Theory 1. Inoduction 2 Explaining Competive Anthoriurian Regime Trajectories: Incemational Linkage andthe Organizational Power of Incombents Part I: High Linkage and Democratization: Eastern Europe ‘and the Americas 3 Linkage, Leverage, and Democratization in Eastern Europe 4 Linkage, Leverage, nd Democratization in the Americas Part Il: The Dynamics of Competitive Autboritarianism in Law-Linkage Regions: The Former Soviet Union, Africa, and Asia 5 The Bvolution of Pow-Sovit Competitive Authortaranism 6 Africa: Traosions without Democratiation 7 Diverging Outcomes in Asa 8 Conclusion Appendix I: Measuring Competitive Aubortarinion and Authoritarian Stability Appendix I: Meanaring Leverage Appendis i: Measuring Linkage Appendix IV: Measuring Orgenizavonal Proer References Index page xi 37 85 130 181 183 399) 339 365 372 374 376 38r 493Introduction, “The end of the Cold War posed a fundamental challenge to authoritarian regimes. Single-party and military dicatorships collapsed throughout Africa, post-communist Eurasia, and much of Asia and Latin Armerica in the late 1980s and early 1990s. At the same time, the formal architecture of democracy — par- ticularly multiparty eleetions ~ diffused across the globe. “Transitions did not always lead to democracy, however. In much of Africa and the former Soviet Union, and in parts of Eastern Europe, Asia, and the Americas, new regimes combined electoral competition with varying degrees of authoritarianism. Unlike single-party or military dictatorships, post-Cold War regimes in Cambodia, Kenya, Malaysia, Mexico, Nigeria, Peru, Russia, Serbia, ‘Taiwan, Ukniine, Zimbabwe, and elsewhere were competitive in that opposi- tion forces used democratic institutions to contest vigorously ~ and, on occa- sion, suceessfilly -for power. Nevertheless, they were not democratic. Electoral manipalation, unfair media access, abuse of state resources, and varying degrees ‘of harassmen: and violence skewed the playing field in favor of incumbents. In ‘other words, 2ompetition was real but unfair." We characterize such regimes a compettie auboritarian, Competitive authoritarian regimes proliferated after the Cold War. B/ our count, 33 regimes were competitive authoritarian in t995 ~ 4 figure that exceeded the number of full democracies in the developing and post-communist world.” “The study of post-Cold War hybrid regimes was initially marked by a pro- nounced democratizing bias. Viewed through the lens of democratization, hybrid regimes were frequently categorized as flawed, incomplete, or “transi- tional” demozracies.' For example, Russia was treated as a case of “protracted” "On post-Cold War tybrid regimes, see * See, for example, the suring of Dison Carothers (032), Onaway (2003), Schedler (290s 3o-1)and Scholar (2o02b: 4) (Goo6s) and cluer oF aris inthe Apel? Faw a cgus, ee Caothers (002) oo. ura * See Caller and Lente (992)4 Competitive Aucboritarianion democratic transition during the 1990s,5 and its subsequent autocratic turn ‘was characterized as a “failure to consolidate” demoeray.* Likewise, Cam- bodia was described as a “nascent democracy” that was “ton the road to democratic consolidation””; Cameroon, Georgia, and Kazakhstan were labeled * democratizes";and the Central African Republic and Cango-Brazeaville were called “would-be democracies.”? Transitions that did not lead to democracy ‘were characterized as “stalled” of “lawed.” Thus, Zambia was said to be “stuck in transition"; Albania was labeled a case of “permanent transition”; and Haiti was suid ro be undergoing a “long,”* “ongoing,” and even “unending” transition, Such characterizations are misleading. The assumption that hybrid regimes, are (or should be) moving in a democratic direction lacks empirical foundation. Hybrid regimes followed diverse trajectories during the post-Cold War period. Although some of them demoeratized (e.g, Ghana, Mexico, and Slovakia), most did not. Many regimes cither remained stable (eg, Malaysia and Tanzania) or became increasingly authoritarian (e.g. Belarus and Russia) In other cases, 20t0- cratic governments fell hit were succeeded by new suthoritarians (eg, Georgia, “Madagascar, and Zambia). Indeed, some regimes experienced two or more tran sitions without democratizing.'* As of 2010, more than a dozen competitive authoritarian regimes had persisted for 15 years or more.® Rather than “par~ tial,” “incomplete,” or “unconsolidated” democracies, these cases should be con ‘eptualzed for what they are: a distinct, nondemocratic regime type. Instead fof eunumning the auch regimes are in transition to democracy, itis more use- fal to ask why some democratized and others did not. This is the goal of our study. "Tintok amine the netris ofl 35 rine hat were or ese ‘competiive authoritarian between 1990 and 1995."? The study spans fie regions, including sx countries in the Americas (dhe Dominican Republic, Guyana, Hi ‘Mexico, Peru, and Nicaragua): six in Eastern Europe (Altania, Croatia, Mace~ Reguably, democratization occured. in regimes held tegulr mulipary cections Mesco and Taivan before incumbents lort and some oil So for three (Zimbabwe), ection for (Senegal and even five Malaysia and "8 Cases of bie democrsition followed by Mexico) decals. without democraizing. fevenion to compestie athoriwianism The holding of lecions. thus cannot (cay Madogasae 1993-1997) awe sored a5 explsin why some competine auhoritar bnsablessthertarian fat regimes denocratzed wheresoxher did 10s Ciser in which incumbents remained in not Nether can sey explain why Guyan, power butte filer ha nt yee Deen Mexico, and Tainan demoxraied via clec= ompleed as of December 2038 (eg, tins ding ne roges bat noe daring pre: Cameroon tnd Tunsaia) ae scored vioos deouden Move generally, Brownlee sere (Geer) shiv dat holding of ean 5 Caothess (2002) and. Brownlee (22072) dae elactns fas no independent cas sake nse pines, inpacc on authoritarian aby. 24 See Lindberg (0962, 3008, 20054 2000. See alo Rigger (1999, 2000) All of our asraduction 3 autocratic incumbents broughe litte institutional change, and successor parties ‘did not govern democratieally. Such cases are too numerous to be ignored or treated as exceptions. [EXPLAINING DIVERGENT OUTCOMES: THE ARGUMENT IN BRIEF “Thisbookexplains the diverging trajectories of competitive anthoritarian regimes since 1990. Asa starting point, we assume that incumbents seck to retain power tnd that they are willing to use extralegal means to do so. We argue that incar~ ‘ents’ capacity to hold onto power ~ and the fate of competitive authoritarian ‘regimes tore generally ~ hinges primarily on two factors: (1) linkage ro the West, or the density of tics (economic, politcal, diplomatic, social, and organizational) and cross-border flows (of capital, goods and servies, people, and information) between particular countries and the United States and the BU; and (2) incu bent’ orgonizatinal power, or the scope and cohesion of tate and governing-party stroctares ‘We make a three-step argument. First, where linkage to the West was exten= sive, asin Eastern Europe and the Americas, competitive authoritarian regimes democratized during the post-Cold War period. By heightening the interna~ ‘onal salience of autocratic abuse, increasing the likelihood of Western response, ‘expanding the number of domestic actors with a stake in avoiding international isolation, and shifting the balance of resources and prestige in favor of opposi- tions, linkage raised the cost of building and sustaining authoritarian rule. High Finkage created powerful incentives for authoritarian rulers to abandon power, rather than erack dowa, in the face of opposition challenges. Tt also created incentives for successor governments to rule democratically. Among igh linkage cases, nota single authoritarian government remained in power through 2008 and neatly every transition resulted in democracy. This outcome occurred even where domestic conditions for democracy were unfavorable e ., Guyana, Macedonia, and Romania). ‘Where linkage was low, a8 in most of Africa and the former Soviet Union, external democratizing pressure was weaker. Consequently, regime outeomes ‘were driven primarily by domestic factors, particularly the organizational power of incumbents. Where state and/or governing parties were well organized and cohesive, as in Malaysia and Zimbabvse, incumbents were able to manage elite conflict and thware even serious opposition challenges (both inthe streets and at the balloc box), and comperitive authoritarian regimes survived. Indeed, in nearly all low-linkage cases in which incumbents had developed coercive and/or party cnganinations, atoerats or their chosen successors remained in power through 2068. ‘Where state and governing-party structures were underdevelopest and lacked cohesion, regimes were less stable. Because incumbents lacked the organiza tional and coercive tools to prevent clite defection, steal elections, oF crack ddawn on protest, they were vulnerable to even relatively weak opposition chal- lenges. Consequently, regimes were more open to contingency than in other cases.24 Comgecisive Aatboritarianione In this context, a third facror~ states’ vulnerability to Western democratizing pressure Qwhich we call Wstem Ieoerage) ~ was often decsive. Where countries’ strategic ot economic importance inhibited external pressure (e.g., Russia), oF Where assistance from counter-hegemonie powers blunted the impact of that pressure (¢.g,, Cameroon, Gabon, and post-1994 Belarus), even relatively weak: regimes survived. Where Western leverage was high, such governments were more likely to fall. In these eases, turnover created an opportunity for democ- ratization, Indeed, fragile democracies emerged in Benin, Mali, and Ukraine. However, inthe absence of close ties to the West or a strong domestic push for democracy, transitions frequently brought to power new authoritarian govern— iments (eg, Georgia, Malawi, and Zambia). In low-linkaze cases, therefore, low ‘organizational power was associated with unstable competitive authoritarianism. Like all theories of regime change, ours cannot explain all cases, Regime out- comes are influenced by a variety of factors ~ including cconomic performance, the strength and strategies of opposition movements, leadership, and historical contingetcy ~ that ie outside of our theoretical framework. [cis not surprising, therefore, that some ofthe regimes analyzed in this study follow trajectories not predicted by our theory (e.., democratization in Benin, Ghana, and Ukraine). Nevertheless, linkage, leverage, and organizational powerexplain striking nur ber of eases. ‘THEORETICAL IMPLICATIONS ‘Our research has a range of implications for the study of contemporary political regimes and regime change. For exaraple, it contributes to the emerging liter ature on the international dimension of regime change. The massive wave of democratization that swept across the developing world inthe r980s and ro90s defied nearly all established theories of democratization. Framed in terms of Dahl's cost of toleration versus cost of suppression, many leading theories ‘expect stable democracy to emerge when either (1) increased societal wealth or ‘equality reduces the cost of toleration"; or (2)a strengthening of civil society or ‘opposition forces ~ often a product of socioeconomic mcdernization~ increases the cost of repression.” Neither of these phenomena oecurred on a large scale prior to the transitions in Latin Americ, Africa, or communist Burasia. What Dahl (978 15 167 See Dati (1931) and Rucchemeyer, Ste- 6 Lipee(engo/tg6tiDubl(.g72}Praewooki pens and Stephens (1992) ‘nal Long (1997 Bois 2003) Acemogia We thane Dail Waldner for deoring out sed Robinson (008) steno mothe pine Insroduction 25 sransnatjonal civil society, and new information technologies. We organize these ‘arious mechanisms into ovo dimensions: Western Jeerage and linkage to the ‘Wes. This framework enables us to capture cross-national variation in the nature and degree of external democratizing pressure. We find that the impact of the {international environment varied considerably across cases and regions, and that this variation was rooted, to large degree, inthe extent of counties’ tiesto the ‘West. Where linkage was high (e.g, Fastern Europe and the Americas), regimes often democratized ~even in the absence of favorable domestic conditions; where irwas low (eg, Africa and the former Soviet Union), domestic factors predomi- nated, Moreover, we find that although political conditionalty and other forms of direct (or leverage-hased) pressure may be effective, the democratiing impact of conditionality is far greater in countries with extensive linkage to the West Second, this book highlights the role of incumbent organizational power in shaping regime outcomes. Recent studies of democratization have given consid- rable attention to the role of societal or opposition-centered factors, including civil society,? organized fabor,"* mass protest," and opposition cohesion, inundermining authoritarianism and/or installing democracy. However, in much of post-Cold War Africa, Asia, and post-communist Eurasia civil societies and ‘opposition parties were weak and fragmented; as a result the societal push for democratization was meager." In many of these cases, regime outcomes were rooted less in the character or behavior of opposition movements than in incum- bent’ capacity to thwart thes. Where incumbents possessed a powerful coercive uyppuiatus and/or party organization, oven well-organized and cohesive opposi- tion challenges often failed. By contrast, where incumbents lacked the orga- rizational tools needed to steal clections, co-opt opponents, or crack down on protest, transitions occurred even when oppositions were weak. Indeed, this hook Shows thae suecessul opposition movements were often rooted in state and party weakness, Much of the financial and organizational muscle behind successful ‘opposition challenges in Zambia (r990-1991), Kenya (2002), Georgia (2003), and Ukraine (2004) was provided by ex-government officials who had defected ‘only weeks or months before the transition. “Two implications are worth noting. First, although strong parties and states are widely and correctly ~ viewed! as critical to democratic stability, they also are critical ro stable authoritarianism.’ Where incumbents lacked strong state and party organizations, they rarely survived during the post-Cold War period. In a competitive authoritarian context, therefore, succesful state- or party-building (€g, Zimbabwe in the 1980s, Armenia and Cambodia in the rogos, and Russia in the 2000s) may contribute not to democratization but rather to authoritarian See Fish (1993) Diamond (i999 and 9? Howard end Roeser (2006. Hloward (2009. 0s Oa the weakness of cl secety im post 1 See Rasschemeje,Stphans and Stephens communist counties, soe Hoard (09). (2), Caller (199), and Belin (20). (Gn oppastion eskness in Acs, 86 "9 See" Bratton and tun de Walle (1997), Roker and van de Walle (2009) Beisinger (2000, 3997), Thompson and" See Hlmingsin (ig 1970) and, more Kune 2004, 2005} Bunce and Wolehieeecendy, Way (2cosa) Brownlee (20078), (Goby and Tucker (2007) and lace (areheoming26 Competitive Authoritarianions ‘consolidation. Second, many post-Cold War transitions were rooted more in the weakness of incumbent governments than in the strength, strategies, oF mobilization of opposition forces. Such transitions were marked by a paradox: “The weakness of state and governing-party onganizationsmade it more likely hat fan antoerat would be forced from power bet fe likely that the transition would result in democracy. Transitions by collapse generally cecurred in a context of ‘weak states, parties, and civil societies ~ conditions that were hardly propitious for democratization. Because both institutional and societal checks on successor ‘governments tended to be weak, transitions often gave rse to new authoritarian incumbents. “This bookalso peaks to the emerging literature on politcal partiesand author~ itarian stability, Scholars such as Barbara Geddes, Jason Brownlee, and Beatriz ‘Magaloni have highlighted the rote that partes play in ma ntaining elite cohesion, which is widely viewed as central to authoritarian stability.!25 For these scholars, parties manage elite conflict mainly through the organization and distribution of patronage. By providing institutional mechanisms for rulers to reward loy~ alists and by lengthening actors’ time horizons through the provision of future ‘opportunities for career advancement, partes encourage elite cooperation over defection." Not all ruling parties arc alike, however. As our study demonstrates, author= itarian parties vary considerably in their organizational strength and cohesion. ‘This variation has important implications for regime stability, Indeed, our case analyses show thar stricey patronaye-based parties ~eveninsecucionaized ies — are often vulnerable to collapse during periods of erisis. During the post-Cold War period, established raling parties in Kenya, Malawi, Senegal, and Zambia were decimated by defection in the face of economic and/or succession crises. By contrast, cohesion is greater in parties that are bound by salient ethnic or ideological ties or a shared history of violent struggle, sich as revolutionary or liberation movements (eg Frelimo in Mozambigue, tke FSLN in Nicaragua, and ZANU-PF in Zimbabwe). Such nonmaterial bonds often help hold patties together even in the face of declining patronage resoure:s. Frelimo, the FSLN, and ZANU remained intact despite severe economic crises and serious threats to their hold on power. Thus, partes that combined patronage with nonmaterial ties ~ such as those rooted in violent conflict or struggle ~ provided the most robust bases for authoritarian rule during the post-Cold War era. ‘THE DISTINCTIVE LOGIC OF COMPETITIVE. AUTHORITARIAN POLITICS “Thisbookalso highlights the importance of taking seriously the dynamics of con- temporary authoritarian regimes.” Until recently, the ssumption that hybrid regimes were “in transition” to democracy biased analyses in important ways "5 Gailes ogg; Brownlee (2coysh Magaloni_ Her, weecho he cals of Ling (202 32-8), (Go08). See also Sith 2003) ‘Brown 2905), thee (2oo6t, and Snyder 16 Gees (999) and Browne (22072. G08) = Introduton ” Scholars gave disproportionate attention to factors that shaped the performance and stability of democracy, such as constitutional design, executive-egislative relations, electoral and party systems, and voting behavior. As a result, the fac- (ors that contre to building and sustaining contemporary nondemocracies, fs wel a the internal dynamics of these regimes, were left underexplored.** In ‘reating competitive authoritarian regimesas “transitional” democracies, scholars often assumed that political processes (e., candidate selection, electoral cam _paigns,and legislative polities) worked more or less as they do under democracies. JYetsuch assumptions ae often misguided, The coexistence of meaningful demo- cratic institutions and authoritarian incumbents creates distinctive opportunities and constraints for actors, which — in important areas of political life— generate distinct patterns of political behavior. We examine some of these areas in the following sections. Informal Institutions (One characteristic of competitive authoritarianism is the centrality of informal instivutions." Informal insticutions exist in all regimes but, given the disjunc- ‘ure between formal (ie., democratic) rules and actual behavior that is inber- ent to competitive authoritarianism, their role in such regimes may be partic arly important, Recent work suggests that actors frequently employ informal institutions as a “second-best” strategy when they cannot achieve their goals AInough formal institions but Bnd the cost of changing those institutions to be probibitive.'® By raising the cost of formal (e.g, single-party) authoritarian rule, the post-Cold War international environment created incentives for incumbents tw employ informal mechanisms of coercion and control while maintaining the formal architecture of democracy. Because informal means of coercion are more difficult for international observers to identify than formal mechanisms ofrepres- sion (eg., press censorship or bans on opposition), they were often critical to the survival of post-Cold War antocracies, “This book highlights a range of informal rules, practices, and orgunizations tsed by incombents in competitive authoritarian regimes. Inthe electoral arena, for example, incumbents who eannot cancel elections or ban oppesition candi dates frequently turn to illicit strategies such as vote buying, bellot-box staffing, and manipulation of the vote count. Although they ate frequently ad hoe, SF This lacuna began go be filed in he soos, outside of ofcally sanesoned channels See Browalee (2072, 20072, 20070), Sher Fede and Levitsky 3009). On informal (2003, 2010), Way (2003, 2004, 20058), _insruions. snd_poiteal regimes, soe Balin (2009), Smith (200s, 2007), Wald O'Donnall gy), Laut (2000), Collins ner (Goos), Scheller (aco), Magsloai (2002, 3003), and Hime and Levicky (Gone), Greene (2007), Lus-Okat (007), {200g 200) arden (Gea), Pepinsky (20096), and" Meeshon(ago4! 50-7, Helmke nd Levitsky lays (fordcoming G209) 2 Informal institions ay be defined as“ See” Morais and Sehedler (2003), socially shared rates usully unwritten that Scheer (sab), and Hard and McCoy dre rested, eaumusicsted, and enforced 2006).28 Compeitive Autboritarianisn practices such as ballor-box stuffing (e.g., Mexico) and vote-buying (ex, “Taiwan) may be instirutionaized ‘Another informal instiution found in many competitive authoritarian regimes is organized corruption. Bribery, blackmail, proxy owrership, and other illicit exchanges are often critical to sustaining authoritarian governing coalitions." For eample, in Cambodia, Peru, Russia, Ukraine, and elsewhere, corruption networks played central roe in ensuring the compliance of state actors during the r9g0s."¥| In Malaysia, Mexico, Pera, Senegal, Russa, and Taiwan, institu- ‘ionalized corruption and patronage and proxy-ownership networks bound key ceonomic, medi, and eivil-society actors to governing partes. ‘Competitive authoritarian governments also employ informal mechanisms of repression. For example, many of them use “legal” repression, or the discre- ‘onary use of legal instraments~ suchas tax authorities and libel laws ~ to target opposition and the media. Although such repression is formal in the sense that it entails the (often technically correct) application of tre law, iis an informal institution in that enforcement is widely known to be selective. The value of this form of repression ists legal veneer: Prosecution for tax fraud or corruption can be presented tothe worldas enforcement ofthe rule of aw rather than repression, Finally, authoritarian incumbents employ informal or “privatized” violence to suppress opposition.'** When the cost of imposing martial law or banning opposition activity is prohibitively high, incambents may opt for violence that is “orchestrated by the state... but carried out by nonstate accor, such as vigi- lates, parsiitarie, and milaias*°* Examples include organized war veterans in Armenia and Zimbabwe, “ethnic warriors” in Kenya, miners in Romania, party youth wings” in Kenya and Malawi, “kick-down-the-door gangs” in Guyana, cbiméresin Haiti and “divine mobs” in Nicaragua. Because such thug groups are ‘not formally linked to state security forces, they providea “certain invisibility as far as international opinion is concemed.”"" They therefore help incumbents achieve the goal of “containing the broad popular challerge to their government, while ateempring to distance themselves from human-rights abuses."7 ‘Succession Polities ‘Competitive authoritarianism also generates distinct challenges in the reali of ‘executive succession, Snocession posesa serious challenge to most autocracies." Unlike most democracies, authoritarian succession is often a high-stakes game. ‘Outgoing incumbents often face serious risks, including possible seizure of ‘wealth and prosecution for corruption or human-righ’ violations." Indeed, ‘many former rulers in competitive authoritarian regimes have been exiled or 19° See Darden G098), 1 Holmquist and Ford 994: 1) "8 On Cambodia, = Gots (2003). On 40 Roeser (o0e 211) Uline, see Darden (2008). On Pere, 8 See Brownlee (2007) see Ropigios’ (1000), Durand (203), and" For this eso, immunity offen a cena ern (2006) Sue for depacng autocrat. This was the Soe Kirsch (2000) and Rowsler 2009) case for example, in Georgia, Rossin Set- "46 Rarsler (2005309). 1, Ulrsne, Zimbabwe, and elsewhere © jutroducton 2» Jmprisoued after leaving ofice.'® For this reason, incumbents often seek a suc- gesor who they can trust to protect them.'S* At the same time, however, they fee e challenge that does not exisin other authoritarian regimes: the need to-win competitive elections. A loyal suecessor is of no value ifhe or she loses elections. ‘Trastworthiness and electability are often in tension with one another. On the ‘one hand, the most electrally viable candidates are often figures with indepen resources and/or support bases, which make them more difficult to control ‘On the other hand, regime insiders ~ particularly those who lack independent ature or resources ~ are more likely to remain loyal, particulary if their close onnection to the regime makes them vulnerable to blackmail. However, such _ politicians often lack the voter appeal to win elections." PY Finding a successor who i both electable and trustworthy is often difficult In Malawi (2999), Kenya (2002), and Ukraine (2004), outgoing rulers erred on the side of safety, choosing loyal but weak candidates who lost elections, In Ukraine, for example, President Kuchma chose Viktor Yanukowych—s corrupt official with 4 criminal past~ apparently because he could be controlled via blackmail, bt the ‘mpopular Yanukorych lost the 2004 election. By contrast, in Malawi (2004) and Zambia (2001), successors won elections bat subsequently tuned on thei __ patrons. In Peru, the inability to find a viable successor contributed to Fujimori’s decision co seck an illegal (and, uldimarely, ill-fated) third term in 2000. Party Behavior Finally, parey behavior is distinee under competitive authoritarian regimes. As Scott Mainwaring has noted, standard assumptions about party behavior ~ for ‘example, that partis are vote-maximizing ~ hold only where elections are the only game in town.”'5} In such a context, parties take the political regime as given and work within it: They participate in elections, seeking to maximize ‘votes; if they lose, they turn to parliamentary opposition. In unconsolidated © democracies and hybrid regimes, however, parties often play a “dual game” that ~ encompasses both electoral and regime objectives. In other words, conven- tional vote-maximizing strategie are complemented ~ and sometimes trumped ~ by strategies aimed at shoring up or undermining the existing regime. Parties cleavly play a dual game in competitive authoritarian regimes. On the one hand, unlike most authoritarian regimes, parties must take seriously elec tions and other democratic institutions; their ability to guin or maintain power Former presidents who were proscured ° Iegperhaps for his reason hatunikehege- ser leaving ofce ininde Fat Neno in mone roginesinAzerbajan, Jordar, Nor Alanis, Levon TerPetosian in Arine- Kota, and ya no compesie sutra ti, Kamuzu Bends and Babli Muloai in ianregime exept Gabo tadarventadjnas ‘Malawi Alber Fujimoriin Peru, Slobodon te rberemin berween 1990 nd 2010, Milofei in Serbia, and Feedek Chiluba 5S Mainwaring (200) in Zambia, Nene, Fujimori and Niloseié 58 Meiovating (2008 8-17). See abo Scheer sree imine Googe) ‘0. Tepoftenforthis ison dit many amoerats For an insight dsusion of this dul opetora dynastic saloon, tapping 2 son or game, ste Schaler s0001, 2090) tthe dese ela (role 1207).x Competitive Authoritarianism hinges ~ a least, in part ~ on their ability to win votes and control legislanres ‘On the other hand, however, competing on a skewed playing feld often requires scrategies that have litte 1o do with vote-maximization, “hus, in all but afew of ‘our cases,5® opposition parties combined conventional (ie, electoral or par- Tiamentary) strategies with extra-institurional ones. Fo- example, oppositions | ‘may boycott elections in an effort to undermine their demesic or international legisimacy.'5° Major opposition parties boycotted at lest one round of presi- dental or parliamentary elections in Ghana (1992), Cameroon (1992 and 1997) Hici (1995 and 2000), Zambia (1996), Zimbabwe (1996 ind 2008), Mali (1997), Serbia (1997), Pers (2000), Benin (2001), and Senega! (2207). ‘When opposition parties participate in elections, conditions may induce them to adopt strategies that difler markedly from those seen in democratic regimes. One is thug mobilization. In a context of widespread violence or lawlessness, candidates" ability to win vores may be just as important as their ability to phys- ically protect or deliver chem. ‘Thus, although recruiting and deploying armed thugs rarely enhances partes’ electoral appeal, it can be critica to their ability t0 ‘campaign and protect the vote. As Zoran Bind, the main architect of Serbia's “bulldozer revolution” in 2000, stated, oppositions must “leary show they are ready to use violence to fight back in case of repression... Security forces must realize they cannot resort t0 violence without risk”? Indeed, successful oppo- ‘tions mobilized both votes and thugs in Benin (1991). Malawi (1994), Serbia (2000), and Kenya (2002). ‘Opposition strategies also differ between elections. Rather than confine its activities co parliament, oppositions in competitive authoritarian regimes may engage in mass protest aimed at toppling the government (or forcing it to under- take democtatiing reform) before the end of its mandate. Such tactis were adopted in Cameroon (199%), Madagascar (1991 and 2009), Albania (iggr and 1997), Ukraine (1993), Venezuela (2002), Haiti (2003), and Georgia (2003 and 2007). ‘Alternatively, opposition parties may adopr a coalitional strategy, joining the government in porsuit of state resources, media access protection, and other Benefits." Although often characterized as “naked oppcrtunism,”™" coaitional In. Boewans, tke Dominicn Repti, Romani, Slralia and posmrgy: Ghana, ‘opposition eaten wasted to con sontional Ge, eeetr} abd parent) chinnels. Conventnsl opposition state ‘Ber geneclly predominate where comet ‘Se utontanin egies are relive soe aint eppositon partes pes the resources needed to suave (ually due tthe exis tenve of aust private sector and soci. 15) Seu expeily Lindberg (cote). In Atfica, ‘opposition pander boyeoed moe chan ¢ ‘ied of presidential eens becween 108% tn 3903 Lindberg 20066 150-1. 19 Se chino (207) ‘30 Tumi 21) "0 Such contr ae dinince om those in democratic raines i at last 40. Was Eee they are ual not neesiy for ned arcoften onrlted othe formation of p= Bamentry morse. In most se neu ‘ents lead enjoy ach majors, Second, they generllyhavenoprogratamaticorie= ‘gia bases. "bons (203: 47-9. Alo Chege (1096 359 Introduction a sguategies may be critical to party survival, In countries characterized by extreme underdevelopment (e.g, Cambodia and Malawi) or extensive state control of the economy (e.g., Belarus and! Gabon), civil society and the private sector are generally small and impoverished, leaving the opposition with limited access to resources. Unless parties have a generous external patron (e.g., Nicaragua and Slovakia) or established organizations, identities, and core constituencies (e., Albania, Guyana, and Malaysia), joining the government may be the only viable means of securing the resources and media access necessary to remain a viable political force. ‘From a vote-maximizing standpoint, coaltional strategies are often subop- ‘imal. Joining an unpopular (and, in many cases, corrupt and repressive) gow= ernment may erorle opposition parties’ electoral and activist bases." Flowever, sphere access to resources isso limited that four or five years in opposition can be tantamount to politieal suicide (eg., much of Aftica and the former Soviet Union), politicians may conelude that jaining the governments isthe best means of preserving their organizations in order to “play another day.” CCoaliional strategies a times have been successful. In Ukraine, afer oligarch ‘Yolia Tymoshenko's bank accounts were frozen in the late 19908, she abandoned ‘the opposition and created the progovernment Fatherland Party. An alliance ‘withthe government allowed Tymoshenko to regain her assets and build a pow= erful organization before moving back into opposition, where she would become a major player in the Orange Revolution." Similarly, the entry of Abdonlaye Wale’> Senegalese Denne Party (PDS) entry into governs walitions ia 1991 and 1995 brought the party access to patronage resources that it used for ‘organization building, while other opposition parties languished.™ Wade won ‘the presidency in 2000. In Kenya, opposition leader Raila Odinga led his National Development Party (NDP) into “partnership” with the Moi government dur- ing the late 19908 in exchange for police protection and access to patronage resources.’ In 2003, the NDP joined the cabinet, which “permitted Odinga to organize dissent from within.”"® A year later, Odinga fed a massive defection ‘thathelped ensure the 2002 electoral defeat ofthe Kenya Aftican National Union _ (KANU). 6 Forcrample aor Cambodia's langet oppor sition pay, FUNCINPEC, Jena he Bun Sen goverment in 1908, it cae wo be viewed asa government lp dog” (Marston pare bacaust he had never eoopecited ith ocr 98) and sured electoral dectine. In Milofevi. Cameroon, the opposition UNDP joined "5 By contras. poitians ~ such as Olek- the goverment afer the 1997 ection nd sande Moros ~ who eemined in opposition sre" ut wiped au in the 2008 lege throughout the Kachina pornd remained tye elon Ans Conds, Angus 3, atginal and enjoyed es electoral sucess. 200% pp. 1) In Serbs, several opposi- 4 Beck 1999-(205-208). tian partes align with the Mists gor-°% Kanpinga (aooy' 12-13 Ndegwa (2003¢ fenuient ding the gg to gem secon 50. to patronage, but Mleleié teed these §. Brown (Geog: 36 ste also Odlambo- aMbiances to eicredi these pares (and thas Mbai 220: 78-80) spline the opposition) a key womens. Indeed, Vj Kofunes emerged the strongest opposition callenger it 2000 in3 Competitive Aurboritarianism Unser competv authoritarian, therefore operon ares pay a a sme, trying to win by the existing rules while siraltaneously seeking to change sre Thitncae th sltoughoppsison pores take esol eco competition and vote-maximization, they may also pursue strategies (eg.,elec- toral boycotts, mass protest) aimed at undermining the regime. Moreover, they ray adopt strategies (eg, thug mobilization, alliances with unpopular gover iments) that ~ although suboptimal from a vote-seeking standpoint ~ allow them to compere and suevive on a skewed playing field. CASE SELECTION AND METHODS (Our study examines all 35 regimes in the world that were or became competitive tuchortaran between 1990 and r995, We exlade fom the analysis other sypes oflaybrid (or “party free”) regimes, including variety cf regimes in which polit- ical competition exists but nonelected officials retain considerable power, such as {@) those in which the most important executive office is not elected (e., Iran, Jordan, Kuwait, and Morocco)"; (2) regimes in which top executive positions are filled via elections but the authority of elected governments is seriously con- strained by the military or other nonelected bodies (eg., Guatemala, Pakistan, ‘Thailand, and Turkey in the early 1990s)"; and (3) competitive regimes under foreign ozeupation (eg, Lebanon in the early 19908). In all oF these regimes, the power of actors outside the electoral process generates 1 distinct set of dynamics [nd challenges not found under competitive authoritarian, We also exckede “illiberal” electoral regimes, in which mainstream partiescompete ons reasonably level playing field but widespread human- or civl-rights abuse ~ often targeting nonmainstream political partes or ethnic groups - persist (et, Colombia and Sri Lanka in the early 19908), Because violations do not directly affect mainstreara political competition, such hybrid regimes are not competitive authoritarian. ‘We also exclude cases in which competitive authoritarianism collapses before the completion of single presidental or parliamentary term," as well as cases in which state collapse makes idifcult to identify any kindof organized political regime." Finally, we limit ourstady to regimes that were competitive authoritar~ jan prior to 1995 in order to evaluate the impact of our variables over a significant period (at least 3 years). Thus, cases that became competitive authoritarian after 1995 (¢., Nigeria and Venezaela) are exchuded from the sample." ‘Our criteria for scoring eases (and the actual coding) are elaborated in [Appendix I. Our criteria for democracy are strict. Regimes “cross the line from kewl, Ug seiclled from the snn- petitive othortavian government fl prey ple beenwe there were no elections for the tomass pros in 1997. Trceweve and polite pares wee banaed "° Cases node Ange, Bonia-Hlerzegenina, between 1990 snd 1998 Litera, Siew Leone, and Zaire/Congo it °6 Oxher tuclary regis during the early the 1990s Space inch Bangladesh, Bl Salvador," Osher rogines thar might be charac- | tevied a0 campeiive suthontarian afer oss include Congo-Beezzile, Gambit, Kyrgystan, Niger, ard Uganda Honduras, and Nepal "6 Caves include Niger, where 2 compesive suthoriarian government was copped 2 ‘coup fo 1996 ar Bulga, where »eom= Introduction 3 democratic to competitive authoritarian if we find evidence of centrally coor- inated or tolerated electoral manipulation, systematic civil-liberties violations Ge, abuse is repeated rather than an exceptional event and is orchestrated or approved by the national government) or an uneven playing ld (., opposition parties are denied significant access to finance or mass media or state institutions are systematically deployed against the opposition). ‘Our method of coring may be illustrated with reference to few cases that fall near the border between competitive authoritarianism and democracy. During the initial period (1990-1095), we scored Botswana as competitive authoritarian dducto extreme inequalities in access to media and finanee; the Dominican Repub- licas competitive authoritarian due to the Balaguer government's packing of the electoral commission and large-scale manipulation of voter rolls; and Slovakia as competitive authoritarian due to Meciar’s abuse of media and harassment of parliamentary opposition. On the other side of the line, Brazil and che Phitip- pines suffered serious problems of democratic governance ~ including extensive lentelism, corruption, and/or 2 weak rule of law ~ in the early ro9es, but we found no evidence of systematic electoral abuse, civil-libertes violations against political opposition, or skewed access to media or finance. Hence, these cases were scored! as democratic and excluded from the analysis, ‘Turning to regime outcomes in 2008, we scored Senegal as competitive sthoritarian due to harassment and arrest of opposition politicians and jour- nalists, and we scored Georgia as competitive authoricarian due to harassment df umjur media in the 200g electious aud dosute uf television sativus dating, the 2007 state of emergency. On the other side of the line, we scored Benin as democratic because the 2006 election was widely characterized as clean and ‘we found no evidence of serious abuse under President Yayi Boni, Likewise, Guyana, Macedonia, Romania, Serbia, and Ukraine were scored as democratic because ~ notwithstanding repeated institutional crises and serious problems of ‘corruption ~ elections were clean, critics suffered no systematic harassment, and | opposition parties enjoyed access to media and finance. With respect to the line berween competitive and fill authoritarianism, our ‘main criterion is whether opposition partis can use democratic institutions to ‘compete seriously for power. If parties or candidatesare routinely exchided, either formally or effectively, from competing in elections forthe national executive," or if electoral fraud isso extensive that voting is essentially meaningless, then __ regimes were scored as noncompetitive and excluded from analysis." Based on these eriteria, Cambodia, Serbia, and Zimbabwe in the 19908 were scored 28 competitive becamse ~ notwithstanding widespread state violence - opposition parties were able to seriously contest national elections. Likewise, Malaysia ‘was scored as competitive because, despite highly institutionalized authoritarian controls, opposition partes operated legally and seriously contested nearly all DY Vablecandistermayhedfecsvelydetered "Examples inclade Burk Faso, Beh fiom runing via severe psa repesion pia, Kasasan, Usheartan in the eatly forthe imposidon of exvere legal admin 1ggon ‘erate, and financial obstacles to elocarl Parcipeion.