Democratization by Elections A New Mode ofTransition Edited by Staffan I. Lindberg The Johns Hopkins University Press Baltimore 6 All Elections Are Not the Same: Why Power-Sharing Elections Strengthen Democratization 148 Pippa Nonis © 2009 The Johns Hopkins University Press All rights reserved. Published 2009 Printed in the United States ofAmerica on acid-free paper 9 8 7 6 5 4 3 2 1 The Johns Hopkins University Press 2715 North Charles Street Baltimore, Maryland 21218-4363 www.press.jhu.edu Library of Congress Cataloging-in-Publication Data Democratization by elections: a new mode of transition I edited by Staffan I. Lindberg. p. cm. Includes bibliographical references and index. ISBN-I3: 978-0-8018'9318-6 (hardcover: alk. paper) 1SBN-1O: 0-8018-9318-6 (hardcover: alk. paper) ISBN-13: 978-0-8018-9319-3 (pbk. : alk. paper) ISBN-lO: 0-8018-9319-4(pbk.: alk. paper) 1. Elections. 2. Democratization. 3. Comparative government. r. Lindberg, 8taffan 1., 1969Jfl001.D465 2009 321.8-dc22 2008046463 A catalog record for this book is available from the British Library. SpeCial discounts are available for bulk purchases ofthis book. For more infonnation. please contact Special Sales at 410-516-6936 or spedalsales@press.jltu.edu. The Johns Hopkins University Press uses environmentally friendly book materials, including recycled text paper that is composed ofat least 30 percent post-consumer waste, whenever possible. All ofour book papers are acid-free, and our jackets and covers are printed on paper with recycled content. Contents List ofFigures List ofTables Foreword, by Larry Diamond Preface Introduction. Democratization by Elections: A New Mode ofTransition? Staffan 1. Lindberg PART I: The Democratizing Power of Elections: A Debate 1 The Power of Elections in Africa Revisited Staffan 1. Lindberg 2 The Relative Powerlessness of Elections in Latin America Jennifer 1. McCoy and Jonathan Hartlyn 3 Elections as Levers of Democratization: A Global Inquiry Jan Teorell and Axel Hadenius 4 Post-Cold War Political Regimes: When Do Elections Matter? Philip G. Roessler and Marc M. Howard 5 Harbinger of Democracy: Competitive Elections before the End ofAuthoritarianism Jason Brownlee vii ix xiii xxi 1 25 47 77 101 128 100 The Democratizing Power ofElections CHAPTER FOUR of elections, and fairly democratic elections in particular, have positive democratic side effects. Instead of being just a token of some kind of democratic achievement, as "pessimists" such as Carothers have claimed, the holding of elections has a cumulative although not very substantial consequence for democracy's future. And there is even, according to o:ur findings, evidence of a short-term effect. Current elections have a democratizing potential, but this potential applies only briefly in the wake of an election or, perhaps, in nonelectoral arenas such as civil liberties. Joining the optimist party, we would maintain that supporting the holding of elections, and preferably truly pluralistic and democratic elections, is indeed a desirable activity on the part of both domestic and international actors. Elections appear to set the stage for a process of democratic learning. a finding also reported by Brownlee in this book. Therefore, it is advantageous to introduce elections as soon as possible, and to make them ongoing and event~allymore pluralistic. The project may derail, but nevertheless it will payoff, both in the short and the long run. Having said this, the effects that We register are not very large in substantial terms. No democratizing miracles should thus be expected from the electoral experience. Post-Cold War Political Regimes When Do Elections Matter? Philip G. Roessler and Marc M. Howard The third wave of democratization and the end of the cold war signifi~ cantly altered the map of political regimes around the world. The col~ lapse of the Soviet Union was the death knell for most Communist governments, which were forced to open their political systems and introduce economic reforms. Other non~Communistdictators lost foreign patronage and found themselves vulnerable to domestic protest and international pressure. Multiparty elections were held in many African countries for the first time since the decade after independence. At first, political science scholarship treated the end ofthe cold war as initiating a political process in which these authoritarian regimes were in tran~ sition to democracies. Much research focused on explaining the sequence by which this democratization process would take place. But as many "transitions" stalled and the resulting regimes proved surprisir:tgly durable, other scholars recognized the need to understand the regimes as they existed, rather than understanding the extent to which they fell short of a set ofstan~ dards and criteria that were probably unrealistic in the first place. New studies began to focus on the emergence of these "hybrid regimes" (Karl 1995), which defied simple classification as democratic or authoritarian because they contained elements ofboth regime types (Diamond 2002). The work of Levitsky and Way (2002b) and Schedler (2002a) made important advancements in this burgeoning field by introducing the concepts of competitive and electoral authoritarianism, respectively. Building on these conceptualizations, a new and growing research program has emerged. Empirical studies have focused on the effect of external factors on changes in and the persistence of competitive authoritarian regimes (Levitsky and Way 2005), the determinants of liberalizing electoral outcomes within competitive authoritarian regimes (Howard and Roessler 2006), the conditions leading to mass political protest before and after elections in electoral authoritarian regimes (Schedler 2006b), and the democratizing power of elections (Lindberg 2006a). While this research program has produced a variety of rich and valuable conceptualizations of nondemocratic regimes, there has been less progress on how to operationalize them by measuring and scoring a global set of political regimes consistent with the systematized concepts of competitive and electoral authoritarianism (Munck and Snyder 2004). Operationalization is important because it helps uS to refine and check the validity of our systematized concepts (Adcock and Collier 2001). Moreover, it allows us to compare regime types systematically across a global sample, bridging artificial regional divisions that scholars tend to impose (Bunce 2003). This chapter has two key objectives: to contribute a clear and precise operationalization of post-cold war political regimes and to analyze empirically which regimes are more susceptible to elections as a mode of democratic transition. The first objective should provide a useful basis for examining and comparing the relative impact of elections on democratic processes within different regime types. Based on the degree of contestation and participation for the selection of the executive, we distinguish between five different regime types-closed authoritarianism, hegemonic authoritarianism, com~ petitive authoritarianism, electoral democracy, and liberal democracy-and develop a measurement scheme to score countries as belonging to one of these five regime types in any given year. Applying these criteria to all countries (with populations greater than 500,000) between 1987 and 2006 reveals several important empirical trends. First, we find that as democratic regimes around the world have surged, back~ sliding into authoritarianism has been surprisingly rare, occurring only 24 times out of a possible 1,454 country-years during the time period of study. Second, hegemonic authoritarian regimes have more than doubled since 1987 and emerged as the modal authoritarian regime type in 2005 (representing 38% of the world's authoritarian regimes). This trend may reflect a calculation by some authoritarian incumbents that significant electoral competition is too risky; others may have allowed minimal contestation as a nod to external donors demanding some liberalization. Interestingly, incumbents in hegemonic regimes rarely abandon multicandidate elections once they adopt the procedure-though contestation is so circumscribed in these elections that these incumbents almost never lose. Competitive authoritar~ ian regimes tend to be the most volatile regime type; more than half of the elections in these regimes either lead to a crackdown in contestation and opposition boycotts or, more frequently, a relatively free and fair election and opposition victory. Looking at geographic trends, we find that immediately after the end of the cold war, democratic transitions occurred disproportionately in Eastern Europe, but since then they have been quite diffuse. With the exception of North America and Europe, each region of the world has experienced at least one democratic transition since 1995.1 During this time period, the greatest number of democratic transitions (10) have occurred in Africa, though the region has also experienced the most incidences (16) ofcompetitive authori~ tarian regimes backsliding to other types of authoritarian regimes or collapsing altogether. Regarding the paper's second objective, our analysis suggests an important temporal distinction between the period at the end ofthe cold war (1987-94) and the last 12 years. For democratic transitions triggered by the end of the cold war, the type of authoritarian regime appears not to matter. In contrast, since 1995 democratic transitions have been Significantly more likely to occur in competitive authoritarian regimes-usually as a result of electionsthan in hegemonic or closed authoritarian regimes. This finding underscores the changing nature of democratization over the last 20 years and speaks to the central question of this volume: How, if at all, do elections matter for democratization? We find that in the late 1980s and early 1990s many democratic transitions resulted from the sudden and severe weakening ofauthoritarian regimes and the rewriting of the rules of the game to allow for greater participation and competition in the selection of the executive, culminating in the holding of 102 The Democratizing Power ofElections Post-Cold War Political Regimes 103 a "founding election." In the past decade, in contrast, the dominant mode of transition has been via electoral processes in competitive authoritarian governments-that is, a subset of authoritarian regimes that allow a considerable degree of electoral contestation (as measured by the preceding election). Electoral processes provide an opportunity for opposition parties to coordinate their antigovernment activities and unify behind a single candidate or form a single coalition, which increase the costs for the incumbent to use force and fraud to stay in power. Our previous research has found opposition coalitions and incumbent turnover to be the key factors that drive liberalizing electoral outcomes in competitive authoritarian states (Howard and Roessler 2006). Bunce and Wolchik (Chapter 10, below) point to how opposition groups, often in conjunction with civil society groups, have skillfully applied a variety of techniques, including protest, voter-mobilization campaigns, dissemination of public opinion polls, and election monitoringcollectively referred to as the "electoral modeY'-to galvanize the public to resist incumbents' efforts to steal the election and help to usher in liberalizing change. While the data indicate that elections in competitive authoritarianism provide an arena for possible post-cold war democratization, the data also suggest the limitations of elections as levers of liberalization. (See Teorell and Hadenius in Chapter 3 of this volume for a systematic treatment of this subject.) First, electoral processes in competitive authoritarian states can also lead to backsliding, as incumbents-perhaps learning from prior elections as well as from the fate of their less repressive colleagues (see Bunce and Wolchik in Chapter 10 below)-institute more restrictive rules and harsher practices against the opposition to avoid the risk of defeat. The effect is that contestation becomes severely circumscribed and the government tilts in a more hegemonic direction. Second, the data suggest that elections in hegemonic authoritarian re~ gimes and electoral democracies have little effect on greater liberalization. Though rulers in hegemonic authoritarian regimes commit to continuous elections, albeit with significantly longer interim periods, on average this does not lead to political liberalization. The mean Freedom House civilliberties score in these countries is slightly worse in the years after they have held more than one consecutive election than in the year they made the transition to hegemonic authoritarianism (usually from closed authoritarianism) after the end of the cold war.' Moreover, these hegemonic regimes tend to be some of the most stable of all authoritarian regimes, underscoring the central point of Lust-Okar (Chapter 9), who argues that elections can reinforce authoritarianism if incumbents use them as an instrument to manage dissent and deepen their societal control. On the democratic side, consistent with Hartlyn and McCoy's analysis of Latin America (Chapter 2, above), in a global sample we see little evidence to suggest that holding successive elections in electoral democracies results in a change to liberal democracy. In the sections that follow, we begin by conceptualizing post-third wave political regimes, disaggregating them, and developing a coding scheme to operationalize these types and score all countries between 1987 and 2006. We then document and analyze trends in regime type frequency, proportion, and change over this time period. We also examine regional variation in regime types by means of world maps created with geographic information systems (GIS) software. We then analyze various trends in democratic transi~ tions and regime volatility, highlighting important differences between the regime types. '°5Post-Cold War Political Regimes Conceptualizing and Operationalizing Political Regimes The starting point of our measurement typology is the background concept of political regimes-the rules and procedures that determine how national, executive leaders are chosen. This concept covers all political systems, whether democratic or authoritarian. To disaggregate the broader concept of regimes into more systematized types, we distinguish them based on the degree to which the rules adopted to select authoritative national leaders allow for contestation and participation in selection of a government (Dahl 1971). These rules are (1) whether selection is through national elections or through lineage, party decree, or military orders; (2) whether there are national elections for an executive, whether the rules and procedures allow for contestation; (3) whether the elections are free and fair or fraudulent; and (4) whether the regime is based on the rule of law and "political and civic pluralism;' or whether the rights and liberties of some individual and groups are still violated (Diamond 1999, 8-13). Figure 4.1 presents a tree diagram that illustrates the key distinctions between the five different types of regimes in the world today. The four main The Democratizing Power ofElections '°4 106 The Democratizing Power ofElections Post-Cold War Political Regimes 107 Liberal democracy Paradigmatic cases: Sweden, United States Figure 4.1. Disaggregation of political regimes by various dimensions of democracy factors that distinguish regimes are listed on the left of the figure, and the regime types are listed on the right. EmplOying these rules, we identify five global regime types: closed authoritarianism, hegemonic authoritarianism, competitive authoritarianism, electoral democracy, and liberal democracy. Two important caveats are necessary regarding this figure. First, it is in~ tended to depict a typology, not a linear or teleolOgical progression from one regime to the next. The last decade has shown that, contrary to the "democratizing bias" (Levitsky and Way 2002b, 51) of much of the earlier democratization literature, these regime types can be stable and enduring, or can even revert to a more consolidated form of authoritarianism. Second, our conceptualization scheme revolves around the institution of national ele'ctions because no better objective and parsimonious metric of contestation and participation exists. But we are also conscious of the "fallacy of electoralism" (Karl 1995) and the pitfalls of focusing on the significance of elections at the expense of other important attributes of democracy. Indeed, democ· racy involves much more than just elections. Robust civil society, effective and independent legislatures and judiciaries, and a civilianized military are just three of the many factors that are necessary for a consolidated democRegimes racy (Linz and Stepan 1996). At the same time, however, democracy cannot be less than free and fair elections. Until a country's selection of national leaders occurs conSistently through a public, competitive, and free and fair process, the deepening of democracy will remain elusive. Regime Types Building on the work of Schumpeter (1942), Dahl (1971), Diamond (1999), and others, we distinguish regimes based on the degree of contestation and participation in the selection of national leaders. Closed authoritarian regimes are those in which the selection of a country's leaders is the responSibility of a small group of elites from the ruling family, the army, or a political party; the citizenry is constitutionally excluded from participating in the selection. Thus, there are no multicandidate national elections; there may be referendums or plebiscites, but no elections that allow for contestation between the incumbent and another candidate. To enforce their monopoly on executive recruitment, the elites ban opposition political parties, rely heavily on repression to maintain political control, and squash free media and civil society. Distinctfrom closed systems are electoral authoritarian regimes (see Schedler 2002a; 2006a)-those in which the executive recruitment process does allow for regular national elections, where there is a choice in candidates (if rival candidates choose to participate rather than boycott), and in which a substantial segment of the citizenry is able to participate, but in which the integrity of the process is fundamentally violated by the incumbent administration's application of rules, procedures, and practices that tilt the playing field in its favor to try to guarantee political survival. Important variation exists among electoral authoritarian regimes, however, depending upon the degree to which the playing field favors the incumbent and infringes upon the opposition's opportunity to contest the election. The cases of Egypt in 2005 and Ethiopia in 2005 are illustrative. In Egypt electoral participation was restricted to a limited number of opposition parties licensed by the Political Parties Committee, which is controlled by the ruling party, while independent candidates were required to collect signatures from 5% of the country's elected officials (again, almost all of whom belonged to the ruling political party) (Freedom House 2006). These and other restrictions, such as a ban on religiOUS parties and the barring of international monitors, ensured that most opposition parties were excluded and that the incumbent, Hosni Closed authoritarian Paradigmatic cases: Chilla, Saudi Arabia Hegemonic authoritarian Paradigmatic cases: Burkina Faso, Uzbekistan Competitiveauthoritarian Paradigmatic cases: Armenia, Gambia No Electoral democracy Paradigmatic cases: Brazil, Philippines / \Yes No No / \ No Elections / \ / \ Yes Contested elections Yes Free and fair elections Yes Freedom, pluralism, rule of law Mubarak, easily won reelection. In contrast, in Ethiopia the main opposition parties were allowed to participate in the electoral process, hold political rallies, and have access to the media. Despite a more open electoral process, the opposition's ability to compete fairly was hindered by irregularities, fraud, and lack of transparency in the counting of votes, all of which contributed to the ruling party's electoral victory (European Union 2005). To capture the differences in contestation highlighted by the Egyptian and Ethiopian cases, we distinguish between two types of electoral authoritarian regimes-hegemonic and competitive. In hegemonic authoritarian regimes the restrictions on opposition parties and their political activities, bias in state~ownedmedia coverage, and other forms of repression so severely circumscribe contestation that the incumbent candidate or party does not face the possibility of losing (Munck 2006,33), often leading to a de facto one-party state.3 Thus, hegemonic authoritarian regimes absolutely violate Bunce's maxim on the central elements of democracy (20m, 45): "freedom, uncertain results, and certain procedures:' In hegemonic authoritarian regimes, the dominance of the political system by the incumbent and the ruling party ensures that there is never any uncertainty in the outcome of national elections; the incumbent nearly always prevails. Though elections are rendered meaningless in the selection of the executive as the outcome is a foregone conclusion, they are not irrelevant; elections, particularly legislative ones, often serve as a key instrument employed by rulers to manage society nonviolently and consolidate political control (Gandhi and Przeworski 2006; Lust-Okar, this volume). Competitive authoritarian systems, on the other hand, permit a substan~ tively higher degree of contestation, leading to greater uncertainty in the outcome of the elections between the ruling party and a legal and legitimate opposition, which usually chooses to participate, rather than to boycott the election. But the incumbent government still uses fraud, repression, and other illiberal means "to create an uneven playing field between government and opposition" (Levitsky and Way 2002b, 53) to try to ensure that it ultimately prevails in the electoral contest-even though it sometimes loses (Howard and Roessler 2006; see the appendix to this chapter). Democracies can be distinguished from the broader set ofelectoral author~ itarian regimes by "the freedom, fairness, inclusiveness, and meaningfulness of elections" (Diamond 2002, 28). At one end of the democratic spectrum are electoral democracies, which permit a competitive process for the selec- Operationalization To disaggregate political regimes into the five mutually exclusive types conceptualized in the previous section, we employ the criteria illustrated in Figure 4.1. Our primary sources for the coding distinctions are the two most commonly used indices of regimes, Freedom House (various years) and Polity (various years).' By using a combination of these two indices-both of which are imperfect, of course-we are able to have a firmer, more reliable basis on which to make our regime type determinations. We code c6untries as closed authoritarian when there are no multicandidate national elections for the direct or indirect selection of the executive.6 Also included in this classification are regimes with referenda for the president or unopposed "elections" in which rival candidates or parties are for~ '°9Post-Cold War Political Regimes tion of the executive that is held under genuinely free and fair conditions. Even if the outcomes of elections in electoral democracies are occasionally one~sided, and even if there are sporadic violations of civil liberties, there is a much more level playing field between the incumbent and the opposition. Finally, liberal democracies go a step beyond: they are strictly bound by the state's constitution and the rule of law, with horizontal accountability among officeholders, protection of pluralism and freedoms, and the lack of "reserved domains of power for the military or other actors not accountable to the electorate" (Diamond 1999, 10). Having defined and explained our relatively abstract typology of regime types, we still need to show how they can best be measured empirically in the real world. There are two ways whereby one can identify how countries should be classified. One is to select cases on the basis ofthe "I know it when 1 see it" formula, namely, by analyzing countries independently and determining which ones fit the overall definition. The other is to establish criteria derived from the coding of other data sources and "let the chips fall where they may:' Both are plausible and defensible strategies. We have chosen the latter, thereby avoiding the temptation to select cases based on our subjective judgments, and instead applying a common, precise, and systematic set of criteria based on existing indices. Any classification system is, of course, arbitrary, but by applying these criteria consistently, we aim to contribute a more objective measurement of these regime types-while recognizing that no such measurement is perfect and that there may still be some disagreement about the inclusion or exclusion of individual cases.4 The Democratizing Power ofElections108 mally banned. since the citizenry is given no choice in the selection of the executive and its participation is inconsequential? AB long as a regime does not allow multicandidate national elections for the selection of the executive, it maintains a closed authoritarian score for each country-year. If there is a change in the rules and procedures by which the executive is selected between 1987 and 2006 (Le., if direct multicandidate presidential elections are introduced), then the closed authoritarian regime is reclassified based on whether the new rules allow for contestation, a free and fair electoral process, and the protection of the rule of law and other freedoms.' The other four regime types do hold national executive elections, of course, albeit in quite different ways. In our operationalization. we first distinguish between countries on either side of what we consider the "democratic threshold," or the minimal requirements to be considered an electoral democracy. Countries with either a Freedom House political rights score of 2 or better or a Polity score of 6 or higher are coded as minimally democratic? Conversely, countries that have both Freedom House scores of 3 or worse and Polity scores of 5 or lower are considered electoral authoritarian.to Within the category of electoral authoritarianism, we distinguish between hegemonic and competitive authoritarianism based on the degree of contes~ tation-Le., the degree to which rules and practices allow for the possibility of incumbent defeat (Munck 2006). Comparing and distinguishing between regimes on the basis of contestation is tricky, however. The rules and prac~ tices incumbents employ to manipulate elections, constrain the opposition, and try to guarantee reelection are rarely transparent and vary widely across countries (Schedler 2006a, 7-10). Thus, there exist few objective and analogous indicators that allow us to capture precisely the integrity ofthe electoral process or the degree to which the rules and practices allow each participant an equal possibility of electoral victory. To differentiate between competitive and hegemonic regimes, we use the outcome of the previous election as the distinguishing criterion.B If the winning party or candidate received more than 70% of the popular vote or 70% of the seats in parliament in the previous election, we code the regime as hegemonic. 12 A country keeps its categorization as a hegemoniC government until the next election unless there is a significant change in the rules and procedures for selecting the executive prior to the next election.13 If the win~ Global Trends in Regime Types, 1987-2006 111Post-Cold War Political Regimes ning party or candidate received less than 70% of the popular vote or of tbe parliamentary seats, the regime is coded as competitive.14 While an electoral percentage threshold has been criticized for conBating contestation with competitiveness (Munck 2006, 34), we believe that it captures the degree of contestation fairly well, since many authoritarian in~umbents who gain 70% or better benefit from a boycott by one or more of the major opposition parties." A boycott by the opposition suggests that the regime's electoral rules were so restrictive, and its practices so repressive, that the opposition forces calculated that they had little or no chance at all of winning. By contrast, in elections in which the incumbent received less than 70%, there was often broader participation by the opposition party members, who calculated that though the playing field was tilted against them, they still had a chance of electoral victo_ry.16 In other words, this rule partially captures the opposition's own calculations about the integrity of the electoral process, and the opposition is probably the best judge ofwhether the electoral process offers the possibility of non-incumbent electoral victory. We broadly distinguish democratic regimes from authoritarian ones on the basis of countries' passing the democratic threshold described abovei.e., a Freedom House political rights score s 2 or a Polity score ~ 6. And we differentiate between the two types of democracies by coding countries that receive a score of both 10 on Polity and 1 on Freedom House political rights as liberal democracies, with the others being classified as electoral democra- cies. Table 4.1 summarizes the coding rules employed to operationalize political regimes. (See the appendix to this chapter for various tests to check the validity of the measurement scheme.) Our universe of cases includes all political regimes in countries with pop~ ulations over 500,000. It begins in the year 1987 (or the year of the country's independence) and extends until 2006. Our unit of analysis is the type of political regime in any given country-year. Applying the coding rules from Table 4.1, Figure 4.2 maps out the frequency and pattern ofregime types over this 20-year span. As the figure shows, there has been a striking increase in democracies as a proportion of total regimes (a nearly 25% gain from 1987 The Democratizing Power ofElections110 112 The Democratizing Power ofElections Post-Cold War Political Regimes 113 Table 4.1 Operationalizing political regimes 1 8 0 - , - - - - - - - - - - - - - - - - - - - - - Regime type Measurement criteria to 2006). This shift (in which the proportion of democratic regimes in the world reached 50% for the first time in history in 2000) is a consequence not only of the third wave of democratization and the end of the cold war, but also of a steady increase in democratic regimes between 1999 and 2006. Most of the increase in democratic regimes, however, has been due to a rise in electoral democracies, which have increased in frequency by almost 150% (from 25 to 62) between 1987 and 2006 and represent the modal regime type in the world from 1992 onward. In contrast, liberal democracies have not kept pace; the increase in these regimes since the end of the cold war has been less than 15%. As with democratic regimes, there have been important changes in the trajectories of different types of authoritarian regimes. As illustrated in Figures 4.2 and 4.3, the number of closed authoritarian regimes declined precipitously between 1987 and 2006. Most of the drop in closed authoritarian regimes was due to the end of the cold war and the collapse of the Soviet Union. But the downtrend continued between 1998 and 2005, with a slight uptick in 2006. In contrast to closed authoritarian regimes, hegemonic authoritarian regimes rebounded after the end of the cold war, more than doubling since 1992, and actually becoming the modal authoritarian regime type by 2005. Competitive authoritarian regimes have experienced greater volatility. This type surged after the end of the cold war as incumbents of closed or hegemonic authoritarian regimes, facing tremendous international and domestic pressures, were forced to open their political systems. Multiparty elections were held to appease international donors and domestic opposition, but the autocratic incumbents frequently employed force and fraud to try to guarantee their political survival (Joseph 1997; Levitsky and Way 2002b). Thus, by 1995 there were more competitive authoritarian regimes than liberal democracies. But over the next 10 years the number of competitive authoritarian regimes declined sharply, from 19% of all regimes in 1995 to less than 10% in 2004 though slightly rebounding in 2005 and 2006. The data presented in Figures 4.2 and 4.3 show the overall numbers and proportions of each regime type, but not their geographical location. The next three figures are world maps that illustrate the regime types of each country in the world with a population greater than 500,000 at three different time periods: 1987, 1996, and 2006. Figure 4.2. Frequency of regime types, 1987-2006 I!IlI Hegemonkauthoritarian D liberal democracy III Closed authoritarian f2l Electoral democracy III Collapsed/other. II Competitive authoritarian o 150 -1-----1 30 120 :D.~ ...~ 908 't5 0 z 60 No multicandidate national elections for selection of executive FH ;:::: 3 and Polity..:::; 5 and winner received;:::: 70% of the vote or seats in previous election FH ;:::: 3 and Polity 5. 5 and winner received < 70% of the vote or seats in previous election FH~ 2 or Polity:?: 6 FH := 1 and Polity := 10 Competitive authoritarianism Hegemonic authoritarianism Closed authoritarianism Electoral democracy Liberal democracy 60% _r_------------------------------: 40\\1 +------2lIt------------------------------;:-:::O=