Latin American Political Regimes in Comparative Perspective Adam Przeworski May 22, 2011

1

Introduction

I need to begin with a personal disclaimer: I am neither a Latin Americanist nor a Latin American. My interest in Latin America originates from the simple fact that is where the action was during most of the two past centuries. As of 1917, 19 out of 49 independent countries were Latin American; still by 1945, 19 out of 67 were. Between 1789 and 1957, when Ghana became independent, 166 out of 341 constitutions adopted in the entire world were promulgated in Latin America, at least 533 out of 1,821 national legislative elections transpired in Latin America, 260 out of 371 coups, and 117 out of 226 civil wars. Anyone who studies modern political history must consider this vast experience, just because this is where most of it occurred. My interest, however, is narrower than political history tout court. I want to understand the dynamics of political regimes and, even more narrowly, of those we call these days "democracies." And here the standard historiography is replete with stereotypes. The exotic image of a land gripped by dictatorships is about as accurate as that picturing the United States as the …nal realization of the Greek ideal of democracy, "the new Athens." Latin American countries tried democracy earlier than the rest of the world, except the United States.1 "Dictatorships," in the current meaning of the term, were rare. Several highly institutionalized regimes, rigorously observing term limits and tolerating at least some opposition, lasted for decades. Moreover, and I know that I am now locking horns with some Latin Americans as well, today several democracies in Latin I appreciate comments by José Antonio Aguilar Rivera, Robert Barros, Joanne FoxPrzeworski, Roberto Gargarella, Fernando Limongi, Gerardo Munck, Pasquale Pasquino, Julio Saguir, Juan Carlos Torre, and the editors of this volume. 1 Annino (1996: 10) observed that "el caso latinoamericano presenta una extraordinaria precocidad en el contexto internacional.... Si miramos al espacio euroatlántico en su conjunto es evidente que América Latina se encuentra en una situación de vanguardia." Hartlyn and Valenzuela (1994: 99-100) observe that Latin American regimes “were largely comparable to the restricted representative regimes in Europe of the same period.” Drake (2009: 2) stresses that “contrary to conventional wisdom, the countries leading the region in making democratic advances did not lag very far behind the United States and Europe.”

1

America –Argentina, Brazil, Chile, Costa Rica, Mexico, Uruguay –have nothing to envy those in the United States, Italy, or France. My point is that the history of democracy cannot be reduced to the experience of Great Britain, France, and the United States. Since stereotypes are hard to dispel, I take license to be polemical. There are two ethnocentrisms to combat: that of our period in history and that of the "center." Both warp our understanding. To demystify, one has to be careful about labels, and I try to reconstruct some of the language used at di¤erent times to characterize political regimes. But I also o¤er hard evidence, counting the incidence of various institutional forms, political regimes, and political events.2 What follows is …rst a conceptual history of political regimes. Then I introduce some aspects of political history that serves to construct really existing political regimes. Finally, I inquire into their dynamics.

2

Political Regimes: A Conceptual History

Distinctions among political regimes do not travel easily in history. Both the actual institutions and our labels for them are products of their times. Moreover, reality and language have histories of their own. The point of departure for any conceptual history of political regimes must be a warning against an anachronistic use of the label of "democracy." Due to the work of Dunn (2005), Hansen (2005), Manin (1997), Rosanvallon (1995) and several others, we now know that those who established …rst modern representative institutions in Great Britain, United States, and France did not think of them as "democracies." Indeed, "democracy" was a negative term, something to be avoided because it presented a danger to the security of property, often coded as "anarchy." This much is by now so well known that I refer the reader to the works cited above, as well as to their summary in Przeworski (2010: 3-5). I just want to emphasize that the universal use of the label of "democracy" to characterize desirable political regimes is very recent: it originated from a deliberate attempt in 1918 by Woodrow Wilson to neutralize the impact of Lenin’s "self-determination of nations" with "self-determination of the people" (Mandela 2007: 39¤, Graubard 2003: 665) and it became an unquestioned norm perhaps only in the 1980s, when Ronald Reagan launched the program of "democracy promotion" (Munck 2009: 2). It also bears emphasis that the language of "democracy" emanated from the United States, so it is an import, if not simply an export. I could not …nd a similar history of "democracy" in Latin America.3 It is clear that the founders of Latin American representative institutions shared the negative view of this system.4 But it seems that after several Latin American po2 All

the data used here are from the PIPE data set. could Posada-Carbó (2008: 16), and he knows better. 4 See McEvoy (2008) on José Ignacio Moreno in Peru, Posada-Carbó (2008) on Eloy Valenzuela in Colombia. Sarmiento (quoted in Zimmerman 2008: 12). referred to "la democracia consegrada por la Republica de 1810" but only in 1845 3 Neither

2

litical thinkers spent some years in the Philadelphia in the 1820s, some adopted a positive connotation for this term. The …rst to use the term "representative democracy" in a positive sense may have been the Peruvian constitutionalist Manuel Lorenzo de Vidaurre in 1827 (see Aguilar 2011, Chapter 3), but this history remains to be written. Note that one should not confuse the use of the language of "democracy" as a label for a political regime with the use of "democratic" or, in English of the eighteenth century, "democratical" element of mixed constitutions (Pasquino 2011).5 With an eye on Great Britain, several constitutional theorists proposed systems of representative government in which a "democratical" or "popular" element, embodied in the lower house of legislatures, would be counterpoised by an "aristocratic" one in the form of a Senate and at times also by a monarchy. These were not "democracies" as we now understand the term. The history of the concept of "dictatorship" is no less convoluted. Around 1810, its meaning was precise and clear because the common reference was to the design of this institution in Rome, where dictatorship was a power delegated (normally by the consuls upon a declaration of emergency by the senate) to someone else than those authorizing it, for a strictly de…ned period (normally six months), not to be used against the delegating body or its members (Nicolet 2004, Pasquino 2010). The duty of the Roman dictatorship, characterized by Schmitt as "commissarial" (McCormick 2004), was to return the polity to the constitutional status quo ante. This understanding of dictatorship was prevalent throughout the history of Latin America. Francisco Miranda may have been the …rst person to bear the title of ”dictator” in the modern era but this denomination was still based on the Roman concept of dictatorship.6 While periods of autocratic rule were not infrequent, they were almost invariably justi…ed by a need to respond to emergencies, crises, or exceptional circumstances. Dictators were "saviours" whose intervention was to be restricted to restoring the Roman salus publica. As Rippy (1965: 93) observed, ”Whether sincere or deliberatively deceptive, the documents of the period always employed expressions suggesting a crisis: liberator, restorer, regenerator, vindicator, deliverer, savior of the country, and so on. Somebody was constantly having to ’save’ these countries....”The crucial di¤erence from the Roman institution was that, although dictators almost always insisted that they are performing a task authorized by a constitution, claiming the mantle of "gobierno constitucional," the mission to save the country was unilaterally undertaken, by force and against the existing institutions. Nevertheless, with some exceptions discussed below, dictatorships were seen as something abnormal, something necessitated by exceptional circumstances, and something to self-dissolve when the situation is 5 According to Saguir’s (2011) account of the Argentine Constitutional Convention of 181619, democracy was seen as a danger because it portended anarchy but a democratic element, in the form of the lower house representing the lower classes and checked by the Senate, would be needed to absorb them into the constitutional system. 6 In 1808-9, Miranda wrote an Esquise de Gouvernement fédéral, a blueprint where he justi…ed justi…ed an exceptional dictatorship by invoking the experience of Rome (Aguilar 2000: 169).

3

restored to normal.7 Dictators assumed power unilaterally but they were also to abdicate unilaterally. To cite Paz (1963: 3-4), ”It is signi…cant that the frequency of military coups has never faded (esmaecido) democratic legitimacy from the conscience of our people. For this reason, dictators assuming power almost invariably declared that their government is provisional and that they are ready to restore democratic institutions as soon as circumstances permit.”8 . The exceptions to this general pattern matter because they are a source of an anachronistic use of the term "dictatorship" in our times. This linguistic transformation is largely due to European di¢ culties in understanding regimes such as that those of the two Napoleons, Mussolini, Hitler, Stalin, or Franco. "Bonapartism," "Ceasarism," "imperialism," "fascism," "totalitarianism" were all terms groping to identify these regimes. According to Baehr and Richter (2004: 25), the term "dictatorship" in its modern sense was used in Europe only during two periods of the nineteenth century: referring to France between 1789 and 1815 and brie‡y after 1852 to the Second Empire. The …rst person to establish a "perpetual dictatorship" in Latin America was Dr.José Gaspár Rodriguez de Francia who in 1816 proclaimed himself El Dictador Perpetuo of Paraguay and ruled it until 1840 as El Supremo.9 The idea that dictatorship is necessary when religion fails to sustain order was developed by a Spanish thinker Juan Donoso-Cortés in a speech in 1849. It found echoes in Mexico (Aguillar 2011, Chapter 7) but an explicit argument that dictatorship should be permanent in Latin America was o¤ered only in 1919 by Vallenilla Lanz. The Soviets were the …rst to use it as a positive self-designation in the "dictatorship of the proletariat" but the term acquired widespread usage only when liberal opponents of the Fascist and Nazi regimes adopted it as the label designating what they were …ghting against. As a result, we came to include under the concept of dictatorship regimes that were "foundational," in Schmitt’s language "sovereign," based on the rejection of representative institutions, designed to be permanent. Moreover, the ambiguity is not only conceptual: several military governments in Latin America after 1930 were themselves split or confused as to whether their mission was to be only to "eradicate the foreign virus of subversion from the body of the nation" and to abdicate once this body was sanitized or to establish a new permanent authoritarian order.10 7 When Bolivar wanted to resign from his …rst of three dictatorships, for example, he was asked to keep the o¢ ce in the following terms: ”Remain, your Excellency, as a Dictator, improve your e¤orts at saving the Fatherland, and once you have done it, then restore full exercise of sovereignty by proposing a Democratic Government.” On Bolivar and dictatorship, see Aguilar (2000: Chapter V). 8 Already Bolivar, in the speech accepting the position of the Dictador Jefe Supremo de la República, announced that ”ya respiro devolviéndos esta autoridad.” (Discurso de Angostura, in Bolivar 1969: 93). 9 Francia is the protagonist of a richly documented historical novel by Augusto Roa Bastos, Yo el supremo but I could not …nd there any surprise at the notion of a perpetual dictator, an oxymoron in the language of the time. 1 0 The 1930 coup in Argentina, for example, was led by General Uriburu, who intended to replace individual by functional representation, while General Justo, who became president in 1931, participated in the coup "solamente como un soldado mas de la revolución," wanting only to depose President Irigoyen. See Ibarguen (1955), Pinedo (1946).

4

If not "democracy" and "dictatorship," as regimes are typically dichotomized in the post World War II era, then what? Following Machiavelli, the most important distinction during the nineteenth century was between monarchies and republics, with the third type distinguished by Montesquieu, "despotism," relegated to the exotic and less known Asia. While monarchical sentiments were frequent in the years following independence,11 all Latin American countries ended up establishing "republics." This is not to say that "republics" were all the same: as Gargarella (2000, 2011) amply documents, the weight of the "democratic" element varied importantly across di¤erent forms of republican governments. But the Western Hemisphere became the reserve of republics, as the United States would insist they be, while several European countries continued as monarchies, with a gradual transfer of control over governments from the crown to parliaments (Przeworski, Asadurian, Bohlken 2010). None of these terms –"democracy," "dictatorship" in its expanded meaning, "republic" or "monarchy," however, identi…es the type of political regimes that were frequent in Latin America and until quite recently in the rest of the world. The traditional Latin American reference was "oligarchic republics" but this term emphasizes political and economic exclusion and this characterization is insu¢ cient. They are recently identi…ed by a plethora of labels: "electoral authoritarianism" (Schedler 2006), "competitive authoritarianism" (Levitsky and Way 2010), "hybrid regimes" (Karl 1995, Diamond 2002), "semi-democracies," "authoritarian democracies," and what not. Those who coined these terms tend to think that the phenomenon is new12 but while "authoritarian" is a neologism, the phenomenon itself is as old as the institution of elections, beginning with 1797 in France (Crook 2002). Moreover, to label them "authoritarian" is to miss their central feature: this term suggests that someone is seen as naturally or historically endowed with the infallible authority to govern, while in these regimes the right to rule is granted only by "other-authorization" (Dunn 2010) in the form of elections. Louis XIV was authoritarian because he believed that his right to rule was due to the grace of God; eighteenth-century aristocracy was authoritarian because it maintained that its political authority was given by nature. The right to make laws belongs to the most intelligent, to the aristocracy of knowledge, created by nature, the Peruvian constitutionalist Bartolomé Herrera, declared in 1846 (Sobrevilla 2002: 196); the Peruvian José 1 1 In the National Assembly of Tucuman, Argentina (1816), General Belgrano put forward una monarquia temperada, a monarchical project having a king native to the Americas – a monarch of Inca descent rather than that of European lineage. General San Martin also favored a monarchical solution (López-Alves 2000: 179). Sentiments for monarchy under an Italian or British prince were present in Uruguay. Yet in the end only Brazil adopted this solution until it became a republic in 1889. In Mexico the …rst emperor, Agustin de Iturbide lasted two years, with a brief return of monarchy between 1862 and 1867. The reasons monarchical projects failed, according to Rippy (1965: 89) were that ”the royalties of Europe and the monarchists of America had di¢ culty in reaching an agreement, the United States was opposed to American kings, the princess were di¢ cult to …nd, and the people were not disposed to tolerate them.” 1 2 Levitsky and Way (2010: 14): "We contend that competitive authoritarianism is a new phenomenon...."

5

María Pando maintained that a “perpetual aristocracy ... is an imperative necessity”; the Venezuelan Andrés Bello wanted rulers to constitute un cuerpo de sabios (Gargarella 2005: 120): they were authoritarians. But even the repressive Soviet regime, whether or not they it was "totalitarian," still felt the need after 1936 to justify itself every few years by gleefully announcing that 99 percent of the people authorize the Communists to rule. That these elections were non-competitive, indeed some were just sham, makes no theoretical di¤erence: to legitimize their rule, the rulers had to rely on the ceremony of elections. Such regimes were recognized at the time simply as "representative government" and for the moment I leave it at that.

3 3.1

Facts and Distinctions "Turmoil" and Rule without Elections

Following O’Donnell, Schmitter, and Whitehead (1986), we are accustomed today to think about political history in terms of political regimes and transitions between them. But there are periods when countries do not have "regimes" of any kind. Regimes, after all, are some stable patterns of relations between the state and the society13 and there are times when everything is in turmoil. An election is held, someone wins, someone else contests, several people try to storm the presidential palace, someone drapes himself in a presidential mantle, only to be deposed by someone else, who then gets himself elected and in Latin America often proclaims a new constitution, only to be deposed again, etc.14 Consider a few schematized sequences of events following elections: (1) In the Dominican Republic a person designated by someone who held o¢ ce by force lost an election in 1878, the winner assumed o¢ ce but was overthrown by the military, which held another election, this time won by the incumbent military ruler, who assumed o¢ ce but was again overthrown by a di¤erent military faction, which in turn was deposed one year later. (2) In Panama the incumbent party lost an election in 1948 but the loser remained in o¢ ce, was replaced by another member of the party that lost, then by yet another one, who was in turn replaced by the original winner, who resigned in favor of his constitutional successor, who was forced out of o¢ ce three years later. (3) In Bolivia, the …rst round of presidential elections of 1979 did not generate a clear result, the winner of election by the Congress was overthrown by the military, a non-partisan 1 3 According to O’Donnell and Schmitter (1986: 73) a regime is "the ensemble of patterns, explicit or not, that determines the forms and channels of access to principal government positions, the characteristics of the actors who are admitted and excluded from such access, and the resources and strategies that they can use to gain access.... This necessarily involves institutionalization, .i.e., to be relevant the pattern de…ning a given regime must be habitually known, practiced, and accepted at least by those whom these same patterns de…ne as participants in the process." 1 4 Obviously my favorite nighttime reading is the impressive collection of some Latin American political chronologies, available from http://libraries.ucsd.edu/locations/sshl/resources/featured-collections/latin-americanelections-statistics/

6

person was chosen pro-tempore but was again overthrown by the military, two more military coups followed, until the original winner assumed and served until the next election. Table 1 is a count of people who occupied the o¢ ce of the chief executive after an election until someone held it for at least one full year or a new election occurred, whichever was …rst. Table 1: Number of occupants of o¢ ce after elections Occupants 1 2 3 4 5 6 7 8 9

Number of observations 2776 (0) 134 (11) 63 (2) 25 (0) 11 (0) 10 (2) 7 (0) 1 (0) 2 (0)

of which Latin America 660 (0) 57 (1) 27 (1) 13 (3) 7 (0) 3 (0) 3 (0) 0 (0) 0 (0)

Note: If two elections occurred in the same year, I count from the …rst one until someone holds o¢ ce for a year after the second one or until the next election, whichever comes …rst. Numbers in parentheses are for years in which there were two elections.

Most elections resulted in the winner assuming o¢ ce and serving until the next one. But what kind of a "regime" is it if three, four, ..., as many as nine, people try to get into o¢ ce and no one succeeds to hold it even one year? What regime is there if over a long period the average tenure of chief executives is 6.5 months, as in Honduras between 1820s and 1876? For a lack of a better term, I call such periods simply "turmoil."15 Now, turmoil should be distinguished from the periods when some ruler holds power for a long period of time without holding elections. Such periods were surprisingly rare after countries experienced at least one election: the quali…cation is necessary because I do not have information about periods preceding the establishment of …rst representative institutions in each country, which means 1 5 Sa¤ord’s (2008: 349-50) characterization of the post-independence period in Latin America is so clear analytically that it merits being quoted in extenso: "Formal constitutional systems were enacted, most of which provided for the transfer of power through elections and guaranteed individual liberties. But these formal constitutional provisions frequently proved a dead letter. No political group believed that its adversaries would abide by them. Those who held power bent constitutional principles and often harshly repressed those in opposition in order to retain the government. Those out of power believed, generally correctly, that they could not gain possession of the state by means formally prescribed by the constitution, because those who held the government controlled the elections. Opposition politicians, both military and civilian, therefore waited for, and took advantage of, moments of government weakness in order to overthrow the ruling group. Governments were unable to resist these rebellions, often because they were too weak …nancially to maintain dominant military force or to provide su¢ cient patronage to buy the allegiance of potential rebels."

7

.4 .6 .8 Proportion of countries

1

Periods of Instability or Rule without Elections

sum_countries 100 150

200

that Saudi Arabia and other Golf countries that never held national elections are not included here. But it is still startling that there were only ten instances in the entire world during the past 200 years in which someone ruled without elections and coups during at least sixteen years in a country that had previously experienced an election. Libya under Gadda… holds the palm for not having held elections during thirty-nine years, closely followed by Spain under Franco. Francia’s reign in Paraguay is the third longest at twenty-nine years but there are only two other Latin American cases on this list. Pinochet’s rule of sixteen years is a major aberration in Latin American history as well as are the …rst seventeen years following the Cuban revolution of 1959. Figure 1 shows by year the proportion of countries, in Latin America and in the rest of the world, in which either elections were not held regularly or they were held but the winners did not complete their term in o¢ ce.16

.2

50

Latin America

0

0

Elsewhere 1800

1850

1900 year

1950

2000

Shaded bars give numbers of annual observations. Lowess smooth

Figure 1 Periods without regular elections were signi…cantly shorter in Latin America than in the rest of the world. If one considers the instances in which no elections were held or they were held but no one completed a full term in o¢ ce, there were 226 such periods in Latin America and they lasted on the average 3.83 years, while in the rest of the world there were 315 such instances with average duration of 5.62 years. The probability that such periods were shorter in Latin America is 1.0.17 Hence, these data con…rm the historians’observation that Latin American "dictatorships," while frequent, tended to be short-lived. 1 6 "Lowess

smooth," used in all the …gures, is just a slowly moving average. the statements about probabilities of di¤erences occurring by chance are based on t-tests with unpaired cases and unequal variances. 1 7 All

8

While power was usurped by the dictators, except for Francia’s rule in Paraguay, Pinochet’s in Chile, and the early years of communist rule in Cuba, they were rarely "foundational," as were Bonapartism, fascism, and communism in Europe. Periods without elections were shorter in Latin America because elections the results of which were not obeyed were more frequent on this continent. If we consider again all years in which regular elections were not held or they were held but the winners did not complete a term in o¢ ce, elections occurred in Latin America with the frequency of 0.20 per year, while in the rest of the world they transpired with the frequency of 0.12, a di¤erence which would occur by chance with probability 0.0000. Moreover, this is not an e¤ect of Africa or other countries that gained independence late: the di¤erence between Latin America and Europe is equally signi…cant (p=0.0013). As I argued elsewhere (Przeworski 2009a), the fact that Latin America experienced frequent breakdowns of constitutional order should not be taken as an indication of anti-democratic tendencies on this continent. Indeed, the contrary is true: Latin Americans tended to experiment with elections earlier and, most importantly, at lower levels of per capita income. And because in poorer countries attempts to select rulers by elections are more likely to end in coups, they often did end in this way, only to be repeated again.

3.2

Uncontested Elections

Events called "elections" in which no one is selected because no opposition is permitted are a puzzling phenomenon. Voting is not the same as electing: it can obviously have other functions (Przeworski 2008). It is controversial whether such elections reveal a need for "other-authorization" by people who in fact rule by force or they are intended to intimidate the potential opposition by demonstrating that the rulers can force large masses of people to perform this inconsequential ritual. (For the latter interpretation, see Gandhi and Przeworski 2006). It may well be that rulers who entered power by force did not tolerate opposition but still felt that they should be legitimized by having been elected, while stable one-party regimes used elections as an instrument of intimidation. It is striking that uncontested elections were frequent in Latin America during the nineteenth century but became rare exactly when Lenin’s invention of oneparty systems spread around the world, indicating that uncontested elections in Latin America were not "foundational" in the sense of establishing a permanent one-party system.

9

.5

Uncontested Elections

Proportion of elections .2 .3 .4

Latin America

0

.1

Elsewhere

1800

1850

1900 year

1950

2000

Lowess smooth

Figure 2

3.3

"Representative Government"

The conceptual di¢ culties in identifying what these days we call "democracies" arise from the fact that many countries over long periods of time regularly held contested elections which incumbents (persons, parties, or designated successors) always won. Labels aside, the observable features of these regimes were the following: (1) They held regular elections for the president or for the legislature. (2) They were politically pluralistic, in the sense that voters were o¤ered a choice in elections. More precisely elections were pluralistic if, with minor exception of instances where political parties agreed to a unique candidate (as in Chile in 1891), there was more than one candidate for president and voters in at least some districts faced a choice between candidates or lists in legislative elections.18 (3) The incumbent person or party or the candidate selected by the outgoing government never lost elections. The mechanism that maintained this non-competitive pluralism was simple. The incumbent government o¤ered candidates, the Ministry of Interior administered the elections, and the newly elected legislature validated the results. This system was universal in the world until 1920 when Austria, Czechoslovakia, and Canada, followed by Chile in 1925, Greece in 1927, as well as Brazil and Uruguay in 1932, transferred the administration or the validation of elections to an inde1 8 I do not consider as pluralistic those situations in which there was one organized party and independent candidates (as in Portugal in 1954), only those in which there was more than one party, faction, or "sentiment" or in which everyone ran as independent.

10

pendent body.19 The instruments at the disposal of incumbents included the use of the state apparatus, manipulation of the electoral rules, and often fraud.20 The idea of an o¢ cial government list submitted to voters for a plebiscitary approval was present already in France under the Directorate (Crook 2002), used under Restoration, and perfected under Napoleon III (Zeldin 1958). Promoting government candidates was not a transgression but a duty of public o¢ cials: the French Prime Minister, de Vilèlle, issued in 1822 a circular instructing ”All those who are members of my ministry must, to keep their jobs, contribute within the limits of their right to the election of M.P.s sincerely attached to the government”(quoted in Zeldin 1958: 79). The same was true in Latin America. As Domingo Santa Maria, the President of Chile between 1881 and 1886, unabashedly admitted, "Giving away votes to unworthy people, to the irrational passions of the parties, and even with universal su¤rage, is a suicide for a ruler and I will not commit suicide before a chimera."21 The stability of such systems – Brazil between 1833 and 1889, Mexico between 1946 and 1999, Chile between 1841 and 1890, Argentina between 1874 and 1915, Paraguay between 1963 and 1988, Uruguay between 1938 and 1966 –was remarkable. Here is the list of regimes that continuously held pluralistic elections in which the incumbents never experienced a defeat. Table 2: Long lasting regimes with regular pluralistic elections which incumbents always won 1 9 Based on Lehoucq (2002). According to IDEA’s 2006 survey of 214 countries and territories, the system in which the government administers and validates still prevails in 26 percent of countries covered, in 15 percent elections are administered by the executive and an independent judicial body certi…es, while electoral management bodies are nominally independent in 55 percent of countries . In remaining 4 percent, elections are not held. 2 0 On the di¢ culties of de…ning fraud, see Annino (1995: 15-18). On corrupt electoral practices in Latin America, see Posada-Carbó (2000). 2 1 Collier and Sater (1996: 58) report that “Delivering the vote was a vital aspect of the Intendant’s [equivalent of French prefet] work.... Yet Intendants could at times go too far .... When the young Intendant of Colchagua, Domingo Santa Maria [future president], interpreted the president’s instructions to win the elections ’at all costs’ a tri‡e too enthusiastically, this was seized by his enemies as the pretext for his dismissal.”

11

Country Luxembourg Norway Romania Netherlands Brazil Mexico Denmark Germany Italy Chilea Tunisia Japan Spainb Taiwan Austrian Empire Botswana Argentina Nepal Portugal

Year ended 1974 1890 1939 1873 1889 1999 1900 1918 1995 1890 2008 1992 1922 1999 1910 2008 1915 2004 1973

Years lasted 126 74 73 59 55 54 52 51 50 50 49 47 47 46 44 44 42 42 40

Reason ended Alternation Alternation War Alternation Monarchy abolished Alternation Alternation Monarchy abolished Alternation Civil war Continued as of 2008 Alternation Coup Alternation Alternation Continued as of 2008 Alternation Autocoup. Revolution

a The legislature was temporarily closed in 1841, otherwise the continuity would be dated back to 1831, with the duration of 60 years. b Control over government alternated regularly between two parties during this period but alternations always occurred before elections and governments newly chosen by the King never lost.

Pluralistic elections continuously won by incumbents were less frequent in Latin America during the nineteenth century because, as we have seen, for a long time winners of elections were often deposed or there was no opposition.

12

.6

Contested Elections Continously Won by Incumbents

Proportion of elections 0 .2 .4

Elsewhere

-.2

Latin America

1800

1850

1900 year

1950

2000

Lowess smooth

Figure 3 Now, were these "democracies" as we tend to de…ne such regimes today? Here is the crux of the di¢ culty. When one looks at this list, some of the regimes look like democracies and most observers classify them as such. Almost everyone thinks that post-war Japan and Italy were democracies even during the long periods during which elections did not produce a partisan alternation in o¢ ce. Most observers think the same about Botswana, where following independence in 1966 until today the same party won all elections. Przeworski et al. (2000) coped with this di¢ culty by invoking a retrospective criterion: a regime is a democracy if the period of successive victories of incumbents was followed by an instance of alternation under the same rules. But the results of applying this criterion obviously depend on the date when a country is observed. Hence, Botswana was not a democracy because alternation did not occur until today, even if it may occur some day, while Mexico was not a democracy before 2000 because the transfer of the administration of elections to an independent body (IFE) in 1990 constituted a change of rules. In any case, the alternation criterion will not get us very far with regard to the nineteenth-century Latin America for a simple reason that electoral defeats of incumbents were extremely rare and peaceful alternation in o¢ ce even less so. As Halperin-Donghi (1973: 116) wryly observed, “Among the many ways of overthrowing the government practiced in postrevolutionary Spanish America, defeat at the polls was conspicuously absent.” One may thus conclude that the alternation criterion is too exacting: as long as elections are contested and their results are obeyed, the regime is a democracy. But then all forms of "representative government" constitute democracy even if no opposition ever wins. Hence, we are in a quandary: either there were almost

13

no democracies in nineteenth-century Latin America or all the regimes in which there was a modicum of electoral opposition were democracies. Now, incumbents may win elections because they are genuinely popular but also because they exploit their control over the state apparatus, manipulate the rules, or commit fraud. These reasons cannot be isolated empirically so that incumbents always can, and do, claim that their victory is an authentic verdict of the people.22 It may be that incumbents exert so much political control that the chances of the opposition to win are practically zero but it may also be that the opposition had a fair chance and lost. This is what we would ideally want to know: whether incumbents could have been defeated. To put it di¤erently, instead of characterizing regimes ex post by the results of elections, we would want to characterize them ex ante by the probability that the opposition could win elections and assume o¢ ce as their result. If we knew which elections were "free and fair," to use the current jargon, we could separate those which incumbents won because they were unfair or unfree from those which they won because they were popular. Yet even if many researchers, including Smith (2005), are willing to make judgements of his sort, I fear that these assessments are often biased by the knowledge of the actual outcomes. Consider a "procedural" approach to this problem. Dahl (1971) lists several conditions, pertaining to rights and freedoms, under which elections should be considered democratic. The procedural approach is to observe if these conditions are satis…ed and to classify regimes on these bases, with the hypothesis that if they are satis…ed, elections are competitive.23 As argued by Munck (2009), this approach may be e¤ective today, given the widespread use of election monitoring. But given the paucity of equivalent information, the discriminating power of this approach is weak when applied to earlier elections. We do know whether elections were direct and whether balloting was secret, we know the franchise quali…cations, and for a smaller subset we have information about the actual proportion of the population eligible to vote, but we do not know whether elections were "clean" by modern standards. Statistical analyses show that incumbents were more likely to lose if elections were direct and if voting was secret or if electoral eligibility was higher.24 But the predictive power of these procedural features is too weak to provide reliable predictions of the instances in which incumbents lost. Moreover, per capita income alone has the same predictive power and in the presence of per capita income none of the institutional matter.25 Hence, the procedural approach does not take us far with 2 2 A Russian proponent of "sovereign democracy," Mikhail Leontiev (in an interview with a Polish newspaper, Dziennik, of 19 January 2008), exploits this ambiguity: "I do not understand what is undemocratic in that some force enjoying overwhelming social support wins elections." 2 3 This is, for example, the approach used by the Freedom House, with a well known ideological bias. 2 4 The "or" is due to colinearity: when secret, direct, and eligibility are introduced in the same speci…cation, only eligibility has a signi…cant.coe¢ cient. 2 5 These results are based on probit regressions with country-clustered standard errors, conditioned on the presence of opposition. The predictive power is measured by the area under receiver operating curves, which ranges from 0 to 1: this is perhaps the best measure of the

14

the available data. Faced with these di¢ culties, I adopt a di¤erent tack. Instead of classifying political regimes, I investigate speci…c events: terminations of non-competitive pluralistic systems and the occurrence of the …rst partisan alternation in o¢ ce as a result of elections.26

4

Regime Dynamics

4.1

Terminations of "Representative Governments"

"Representative governments," as de…ned above, can end their life for a number of di¤erent reasons: (1) Rule without elections or elections after which the winners do not complete their terms; (2) Repression of all opposition while continuing to hold non-pluralistic elections; (3) Partisan alternation in o¢ ce as a result of an electoral defeat of the incumbent. The frequencies of these reasons for termination are listed in Table 3. Table 3: Causes of termination of pluralistic non-competitive regimes Cause turmoil, no elections repression of opposition alternation continuing total

World 89 (0.33) 42 (0.16) 93 (0.35) 42 (0.16) 266 (1.0)

LA 39 (0.51) 22 (0.29) 14 (0.18) 2 (0.03) 77 (1.0)

Statistical analysis27 shows that the determinants of these terminations are unsurprising, at least in so far as that higher per capita income makes termination by coups less likely. Turmoil is more likely to ensue if the rate of electoral participation –proportion of voters to the population –is higher: it seems that non-competitive but pluralistic regimes cannot withstand high participation. Yet higher participation does not increase the probability of suppressing the opposition or of partisan alternation. Older pluralistic regimes are less likely to repress opposition while presidential systems are more likely to do so. Representative government in countries which experienced in the past a breakdown of pluralistic regimes –whether or not they were competitive –is more likely to …t of probit predictions. The respective values are 0.6 for secret and direct, 0.59 for eligibility, and 0.59 for per capita income alone.. 2 6 Throughout text I consider only alternations in the partisan control over the o¢ ce of the chief executive that result from elections, but to avoid repetitions below I refer to them simply as alternations. 2 7 These results are based on multinomial logit of these reasons for termination, conditioned on the current state being "representative government," with country-clustered standard errors.

15

end in partisan alternation.28 Finally, presidential systems as well as systems with bicameral legislature are less likely to experience alternations. Latin American representative governments were more often terminated by breakdowns of constitutional order than those outside Latin America, while they were equally likely to end in repression of the opposition or in alternation. Note that while almost all Latin American political institutions were presidential, only about one-half presidential systems were located in Latin America. Hence, the e¤ect of presidentialism – to make alternations less likely and uncontested elections somewhat more likely –is independent of the location in Latin America.

4.2

The Origins of Political Competition

When and how do elections become competitive, with their results obeyed by the losers? Considered here are the instances in which either the incumbent rulers did not present themselves or their political allies in an election or did present themselves and lost, in both cases as long as the winner peacefully assumed o¢ ce and held it at least for one year or until the next election. There are quite a few instances in which the incumbent lost an election but the winner was prevented from assuming o¢ ce: such instances are not considered as alternations. Hence, the question concerns the origins of electoral competition, as evidenced by constitutionally regulated partisan alternation in o¢ ce. Table 4 lists …rst partisan alternations that occurred anywhere in the world as a result of elections. Only seven alternations occurred in Latin America during the nineteenth century, out of 328 presidential elections for which we have this information. Moreover, only in Colombia after 1837 and after 1848, Dominican Republic after 1849, and Argentina after 1878, did the victorious winner survive a full term in o¢ ce. If one were to apply to the nineteenth century the criteria used to classify regimes after 1945, these would have been the only democracies in Latin America. Table 4: Partisan alternations in o¢ ce as a result of elections 2 8 On

the e¤ect of past history on regime dynamics, see Przeworski et al. Przeworski (2009a).

16

(2000) and

Country US UK Colombia Spain Belgium Dom Rep Honduras Portugal Italy Argentina Liberia Netherlands France Sweden Costa Rica

First 1800 1835 1837 1837 1847 1849 1852 1860 1867 1868 1869 1874 1877 1884 1889

Second 1828 1841 1848 1857 1856 1853 1928 1864 1876 1916 1871 1877 1881 1905 1909

Third 1840 1852 1930 1865 1870 1978 1932 1865 1892 1989 1877 1888 1885 1911 1923

Note: Italic fonts indicate that coups or civil wars occurred between the previous alternation and the date appearing in italics.

Examining over time the proportion of elections that resulted in alternation shows that (1) alternations have been rare everywhere until quite recently, (2) for a long time they were less frequent in Latin America than elsewhere (which was bascially Europe), (3) they have become exceptionally frequent in Latin America recently. My hunch is that alternations were easier to tolerate by the losers in monarchies, where less is at stake in elections, than in presidential republics.29 The frequency of alternations in Latin America during the past three decades is thus even more striking. 2 9 It is interesting that this was the conclusion of Mexican conservatives in 1846. After only two governments completed their terms in twenty-two years, they concluded that solution to the instability is monarchy. See Aguilar (2011)

17

.2 .4 Proportion of elections

sum_elections 20 30

40

.6

Alternations by year

Latin America

0

0

10

Elsewhere

1800

1850

1900 year

1950

2000

Lowess smooth. Sum of annual number of elections in the entire world on the left scale.

Figure 4 To anticipate the conclusions, note that alternations have also become much more frequent in Europe in the past thirty years, so that this di¤erence between Latin America and the rest of the world is due largely to non-European countries. Even so, the rate of alternations in Latin America after 1980 is still slightly higher than in Europe. The circumstances under which …rst alternations tend to occur depend on the antecedent status quo.30 A direct transition from rule by force to competitive elections is more likely at higher levels of per capita income. Transitions from uncontested elections are more likely when the rulers fail to make people participate in the ceremony of "elections."31 Finally, we already know that alternations are more likely to happen under representative government in countries which experienced in the past a breakdown of pluralistic regimes and less likely in presidential systems with bicameral legislatures. Latin America di¤ers from the rest of the world only in that direct transitions from military rule to competitive elections have been more frequent than elsewhere. These patterns, however, do not add up to much: "transitions to democracy" are notoriously di¢ cult to predict because they occur for a large variety of reasons (Przeworski et. al 2000). 3 0 The statistical results summarized here are based on the coe¢ cients of multinomial logits of transition to competitive elections from turmoil, uncontested elections, and "representative government." 3 1 Przeworski (2011) argues that a fall of participation in one-party systems is a signal to the potential opposition that the regime is losing control and that, as a consequence, makes the fall of such regimes more likely. This argument has been con…rmed in a statistical analysis by Corvalan (2011).

18

It merits noting that neither the extent of franchise nor the proportion of the population participating in elections a¤ect the probability of the …rst alternation occurring under representative government. Because this is a topic that looms large in conceptual debates about democracy, it deserves more detailed attention. Note …rst that many people, including Dahl (1971) and Vanhanen (1997) but not only scholars, insist that a regime is not democratic unless the right to participate in elections is widespread. Dahl goes as far as to require that at least one half of the adult population have the right to vote, so for him the United States became a democracy ("polyarchy" in his language) only in 1950. These two dimensions – contestation and participation – are clearly distinct conceptually. Moreover, while political con‡icts before World War I focused mainly on su¤rage, after 1918 universal su¤rage became widespread and the main political issue became the right to contest elections. But the question is whether these conceptually distinct dimensions are historically related: whether contestation is a consequence of extended su¤rage. Here is the relevant evidence. Table 5: Proportion of Elections Lost by Incumbents, by Su¤rage Quali…cations Quali…cation Property Income and Literacy Income alone Income or Literacy "Independent" Universal Total

Males 0.13 (70) 0.07 (91) 0.20 (259) 0.14 (111) 0.11 (146) 0.14 (400) 0.14 (1077)

and Femalesa — (0) 0.23 (13) 0.00 (11) 0.15 (47) 0.17 (6) 0.26 (1520) 0.25 (1597)

Total 0.13 (70) 0.09 (104) 0.19 (270) 0.15 (158) 0.11 (152) 0.23 (1920) 0.21 (2674)

Note: The numbers of elections occurring under each type of quali…cations are in parentheses. a in 50 cases women faced additional restrictions (higher age, only widows of the military, etc.).

Two aspects of this table merit attention. When su¤rage was limited to males, the incidence of incumbents losing was not di¤erent when it was additionally restricted by property, income, or literacy than when it was universal.32 Yet incumbents lost more frequently when su¤rage was quali…ed speci…cally by income (the di¤erence from universal male su¤rage is signi…cant, p=0.0151). The reason is that income was the most restrictive criterion in its e¤ect on the proportion of the population quali…ed to vote, even more restrictive than the property requirements which were very lax in Canada and New Zealand, and more restrictive than the literacy quali…cation, which continued in place well after most males were literate. The proportion of elections lost by incumbents became clearly higher when women gained the right to vote. Yet statistical analysis shows that the e¤ect of 3 2 The

"independent" category is sui generis. See Przeworski (2009b).

19

female su¤rage vanishes when controlled by per capita income, as does the e¤ect of universal su¤rage. In turn, the e¤ect of income quali…cations does survive in the presence of this control (p= 0.0320). These …nding add up to the conclusion that su¤rage and competition constitute historically independent dimensions. If anything, the "oligarchical republics," at least those in which the access to the oligarchy was regulated by income and gender, were more competitive than "mass democracies." Hence, there are grounds to believe that elites were more willing to compete when competition was restricted to elites. Statistical analyses indicate the conditions under which it happened but not how it happened. The circumstances under which …rst competitive elections occur are highly varied. "Jump starting" –competitive elections following a civil war – occurred after wars destroyed most of what the …ghting was about, so that little was to gain by continuing (Wantchekon 2004). Transitions from uncontested elections often entailed overthrowing the dictator by force, while those from institutionalized one-party systems were highly conditioned by geopolitical factors. Elections became competitive in this context, I believe, when none among the forces opposed to the extant regime could impose itself by force over its allies in overthrowing it.33 The question I …nd fascinating is why at some de…nite moments the incumbent rulers, whether they entered o¢ ce by force or by non-competitive elections, become willing to risk their hold on power at the polls. What is crucial in my view is whether the incumbents expect that if they were to hold a competitive election, giving someone else a chance to win, the eventual winner would reciprocate. As the experience of the U.S. elections of 1800 demonstrates (see Weisberger 2000, Dunn 2004), the decision to release the reigns of power is extremely di¢ cult: Je¤erson may have prevailed only as a consequence of Madison’s threat to mobilize the Virginia militia. And it took twenty-eight years, seven electoral periods, before the next alternation occurred in the United States. In New Granada in 1837, General Santander, who believed that the country was not ready for a civilian chief executive, supported General Obando to be his successor. Yet a civilian, Dr.José Ignacio de Márquez won the plurality of electoral votes and the Congress con…rmed his victory (Posada Carbó 1999). According to Bushnell (1993: 90), “Santander then delivered his o¢ ce to someone he had opposed –taking pains to point out, in a proclamation, that he had thus respected the will of the people and the law of the land.” Obando did rise against Márquez two years later, but was defeated, and Márquez completed his term.34 Party lines were ‡uid until around 1850, but regular elections followed and terms of o¢ ce were completed until the coup of 1854. 3 3 The prospect of joining Europe was obviously also important in the Eastern European part of the former Soviet block. 3 4 Posada-Carbó (2008: 31) points out the contrast with the events in Venezuela at two years earlier: "Although in Venezuela President Páez’s favourite, General Soublette, also lost at the polls and Páez handed in power to the victor, Dr José M. Vargas -a civilian-, the latter was ousted seven months later, when Páez returned to the presidency and then practically ruled Venezuela for the next two decades."

20

One might expect that partisan alternations would be easier in monarchies because the monarchs can provide the guarantee to the losers that they could return to o¢ ce. This was generally true, but not without a signi…cant and sometimes long resistance from the kings, who had partisan preferences of their own. In England, the king appointed a Tory prime minister in spite of the Tory electoral defeat in 1834 and only the repeated victory of the opposition forced him to accept Melbourne government. In Belgium, Liberals had to win twice before assuming o¢ ce in 1847, in Denmark minority right-wing governments stayed in o¢ ce in spite of repeated defeats between 1872 and 1901, in the Netherlands the same was true between 1856 and 1871. The expectation that the losers can return to o¢ ce is crucial because it means that that holding onto power is not a matter of survival, so the stakes are not very high. When incumbents believe that an electoral defeat may mean a loss of life or at least of their fortunes, as in contemporary Russia, the risk is just too high (see Makarenko 2011 on "tolerable uncertainty"). To put it di¤erently, what matters is not whether the incumbents would lose but what they would lose.35 Whether this belief is a matter of trust, of relations of physical force, or of the degree to which the interests between the rulers and the opposition happen to converge, is unfortunately not a question that can be resolved by a recourse to observation. The fact that alternations are more likely to transpire at higher income levels indicates that the size of the stakes matter: when incomes are higher, there is less to gain by holding onto power and less to lose by releasing the reigns of power. But it is striking that elections that resulted in a peaceful partisan alternation in o¢ ce tend to be followed by periods during which competitive elections become the norm, coups are less frequent, and elected governments complete their terms.36 Hence, there is strong evidence of path dependence, speci…cally that once political leaders see that having lost o¢ ce is not a disaster, they are willing to put their hold onto power at risk again.

5

Latin America in a Comparative Context

One cannot attend a Latin American meeting without hearing two canonical phrases: "In my country, as in the rest of Latin America, ..." and "In contrast to advanced countries, in Latin America ...." If we are to believe what we hear, Latin America is internally homogenous and di¤erent from the rest of the world or at least from "advanced countries," "established democracies," or something of the sort. The evidence presented here indicates that this claim may have been true in a distant past. But it is true no longer. 3 5 This

is literally what a communist reformer said to me on the streets of Warsaw in 1987. the …rst alternation occurs in a country, about one subsequent election in three (0.35 of 1355 elections) results in peaceful alternations. Before the …rst alternation, coups occur with annual frequency of 0.0682, after the …rst alternation with the frequency of 0.0310. Before the …rst alternation, years of completed constitutional terms constitute 0.63 of 6210 years, after the …rst alternation they make 0.82 of 4753 years. The e¤ect of …rst alternation on coups and completed terms survive in probit regressions controlled for per capita income. 3 6 Once

21

To summarize the patterns presented above, consider the annual incidence of "rule by force" in Latin America compared to Europe. A note of caution with regard to this regionalization is, however, in order …rst. Regionalizations also have a history, often resulting from geopolitical interests. My native country, Poland, was a part of Christian Europe, Slavic Countries, Middle Europe (Mitteleuropa), "Eastern Europe" during the Cold War, is now a part of "PostCommunist Countries," and perhaps will be simply in "Europe" within a near future. The idea of the "Americas" and later "Western Hemisphere" served to separate it from monarchical Europe: as Rojas (2009: 15) observes, for many in‡uential Latin American intellectuals at the time of independence, "lo americano ... no estuviera adjetivado por lo ’latino’ o lo ’hispano’." The invention of a "Latin race" by a Frenchman, Michel Chevalier, was intended to juxtapose it to Anglo-Saxon domination.37 "Latin America" is a product of the 1850s. "Europe," in turn, changed meanings several times, extending and contracting to the East and the South (Davies 1998). How arbitrary are these regionalizations is evidenced by a football game between Aktobe Lento, Kazakhstan, and FC Tbilisi, Georgia (Lento won 2:0), which took place within the European Champions League. With this caveat, I compare the Spanish and Portuguese speaking countries south of the Rio Grande and the Caribbean Islands with countries bordered by the Ural in the East and the Mediterranean in the South. Included under "rule of force" are instances in which no elections were held regularly or they were held but their results were not obeyed or they were held without opposition. The complement to these cases are thus instances in which pluralistic elections were held regularly, regardless of their results. 3 7 In

the words of José María Torres Caicedo (Las dos Américas, 1857), "La raza de la América latina Al frente tiene la sajona raza Enemiga mortal que ya amenaza Su libertad destruir y su pendón.".

22

Proportion of countries .4 .6 .8

1

Rule by Force in Latin America and in Europe

.2

Latin America

0

Europe

1800

1850

1900 year

1950

2000

Lowess smooth

Figure 5 As is well known, civil wars and other forms of political violence continued in Latin America long after independence and when order …nally prevailed around 1870, it was imposed by force. Europe in the meantime began to hold elections, which were typically pluralistic, even if non-competitive. Pluralistic elections became more frequent in Latin America after 1870 and rule by force reached its lowest level by about 1920. The end of World War I gave birth to several new countries in Europe and some of them quickly succumbed to the rule of force. Most puzzling, and to my best knowledge never analyzed jointly, is the simultaneous eruption of political instability in mid-1920s in Europe and Latin America.38 The aftermath of World War II saw the rise of several one-party regimes in what became Eastern Europe, while pluralistic electoral regimes became more frequent again in Western Europe and well as in Latin America. The major contrast is that Greece was the only European country where pluralism subsequently collapsed, while not a single democracy that existed in Latin America as of 1946 survived. As I argued elsewhere (Przeworski 2009a, also Cheibub 2007), the di¤erent fates of democracy on the two continents must have been due at least in part to the outcome of the war, which resulted in defeat of authoritarian forces in Western Europe, while it left them intact in Latin America. In consequence, just when the last regimes based on force in 3 8 According to Rouquie (1994: 223), “Between February and December of 1930, the military were involved in the overthrow of governments in no fewer than six, widely di¤ering Latin American nations - Argentina, Brazil, the Dominican Republic, Bolivia, Peru, and Guatemala. The same year also saw four unsuccesfull attempts to seize power by force in other Latin American countries. Over the following years, Ecuador and el Salvador in 1931, and Chile in 1932, joined the list of countries in which military-provoked political shifts and unscheduled changes of the executive had taken place.”

23

Western Europe –Greece, Portugal, and Spain –collapsed, the Southern Cone was in the grip of exceptionally brutal military dictatorships. But all this is a past, now distant by at least a generation.39 Pluralistic electoral regimes are as entrenched now in most countries of Latin America as they are in most of Europe. The di¤erences do not run across regional lines: Argentina, Brazil, Chile, Costa Rica, Mexico, Uruguay have stable and reasonably well functioning democratic regimes that have nothing to envy the United States, Italy, or France. I know that this assertion will raise objections: Latin Americans will point out to corruption, unscrupulous tactics of politicians, in‡uence of money over the media, in some cases even to electoral irregularities. But it takes a second to …nd the same phenomena in several "old," "developed," "long-established" democracies. One can quibble whether democracy in Sweden is "better" than in Chile, in France than in Argentina, in the United States than in Brazil, but these will be quibbles. Lula could not have won an elections in the United States and even if the election of Obama was a miracle, the former was much more e¤ective in implementing his program of reforms than the latter. In turn, signi…cant intra-regional di¤erences remain on both continents: Argentina has more in common with France than with Honduras, Sweden more in common with Chile than with Belarus. This is a new world, and I do not hesitate to say, a "democratic" one. True, everyday life of really existing democracies is not an inspiring spectacle. But con‡icts, peace, and liberty do not coexist easily: this is why through most of history, civil peace could be maintained only by force. Given the history recounted here we should not lose the sight of how privileged we are to be free from oppression, free to process our con‡icts in peace. When Donoso– Cortés could say in the middle of the nineteenth century, "What is good is the correction that disobedient peoples receive from tyrants and tyrants receive from revolutions," we can now say that what is good is that people are free under laws and the lawmakers are chosen by the people.

6

References

Aguilar Rivera, José Antonio. 2000. En pos de la quimera. Re‡exiones sobre el experimento constitucional atlántico. México: CIDE. Aguilar Rivera, José Antonio . 2010.“Las ideologías políticas en América Latina: de la metamorfosis al ocaso del liberalismo, 1870-1930.” Ms. Aguilar Rivera, José Antonio . 2011. Ausentes del Universo: Re‡exiones sobre el Pensamineto Político Hispanoamericano en la Era de la Construcción Nacional: 1821-1850. Ms. Mexico: CIDE. Annino, Antonio. 1995. “Introducción.” In Annino (ed.). Historia de las elecciones en Iberoamérica, siglo XIX. México: Fondo de Cultura Económica. 3 9 Perhaps the most striking piece of evidence is that one can now participate in large academic conferences on Latin America in which the word "military" is not mentioned once.

24

Annino, Antonio 1998. ”Vote et décalage de la citoyenneté dans les pays andins et meso-americains.”In Ra¤aele Romanelli (ed.), How Did They Become Voters? The History of Franchise in Modern European Representation. The Hague: Kluwer. Pages 155-182. Baehr, Peter, and Melvin Richter. Eds. 2004. Dictatorship in History and Theory: Bonapartism, Ceasarism, and Totalitarianism. Cambridge: Cambridge University Press. Bolívar, Simon. 1969. Escritos politicos. Edited by Graciela Soriano, Madrid: Alianza Editorial. Bushnell, David. 1993. The Making of Modern Colombia. Berkeley: University of California Press Calderón, Francisco Garcia. 1912. Les démocraties latines de l’Amérique. Paris. Cheibub, José Antonio. 2007. Presidentialism, Parliamentarism, and Democracy. New York: Cambridge University Press. Collier, Simon and William F. Sater. 1996. A History of Chile, 1808-1994. Cambridge: Cambridge University Press. Corvalan, Alejandro. 2011. "Is Participation Harmful to Democracy?" Working Paper, Department of Economics, New York University. Crook, Malcom. 2002. Elections in the French Revolution. Cambridge: Cambridge University Press. Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New Haven: Yale University Press. Diamond, Larry. 2002. "Thinking about Hybrid Regimes." Journal of Democracy 13 (no.2): 21-35. Davies, Norman. 1998. Europe: A History. New York: Harper. Drake, Paul. 2009. Between Tyranny and Anarchy: A History of Democracy in Latin America, 1800-2006 Stanford, Cal.: Stanford University Press. Dunn, John. 2005. Democracy: A History. New York: Atlantic Monthly Press. Dunn, John. 2010. "Judging Democracy as Form of Government for Given Territories: Utopia or Apologetics?" In Adam Przeworski and others, Democracy in the Russian Mirror. Forthcoming. Dunn, Susan. 2004. Je¤ erson’s Second Revolution: The Election Crisis of 1800 and the Triumph of Republicanism. Boston: Houghton Mi- in. Gandhi, Jennifer and Adam Przeworski. 2006. "Cooperation, Cooptation, and Rebellion under Dictatorships." Economics&Politics 18: 1-26. Gargarella, Roberto. 2000. Los fundamentos legales de la desigualidad: El constitucionalismo en América (1776-1860). Madrid: Siglo XXI. Gargarella, Roberto. 2010. 200 años de constitucionalismo en América Latina (1810-2010). Ms. Graubard, Stephen R. 2003. "Democracy." In The Dictionary of the History of Ideas. University of Virginia Library: The Electronic Text Center. (http://etext.lib.virginia.edu/cgi-local/DHI/dhi.cgi?id=dv1-78) Halperin-Donghi, Tulio. 1973. The Aftermath of Revolution in Latin America. New York: Harper&Row. 25

Halperin-Donghi, Tulio. 1987. "En el transfondo de la novela de dictadores: la dictadura hispanoamericana como problema histórico." In El espejo de la Historia: Problemas argentinos y perspectivas latinoaméricanas. Buenos Aires. Pages 17-39. Hansen, Mogens Herman. 2005. The Tradition of Ancient Democracy and Its Importance for Modern Democracy. Copenhagen: Royal Danish Academy of Arts and Letters. Hartlyn, Jonathan and Arturo Valenzuela. 1994. “Democracy in Latin America since 1930,” pp. 99-162, 610-622, in Leslie Bethell, (ed.), The Cambridge History of Latin America Vol. VI. Latin America since 1930 Part 2. Politics and Society. New York: Cambridge University Press. Ibarguen, Carlos. 1955. La historia que he vivido. Buenos Aires: Ediciones Peuser. Karl, Terry Lynn. 1995. "The Hybrid Regimes of Central America." Journal of Democracy 6 (no.3): 72-87. Lehoucq, Fabrice. 2002. “Can Parties Police Themselves? Electoral Governance and Democratization.” International Political Science Review 23 (1): 29-46. López-Alves, Fernando. 2000. State Formation and Democracy in Latin America, 1810-1900. Durham: Duke University Press. McEvoy, Carmen. 2008. "The Forking Path: Elections and Democracy in Peru, 1912-1872." Paper presented at the Seminar on the Origins of Democracy in the Americas, Notre Dame University, September 18-21. McCormick, John P. "From Constitutional Techniques to Ceasarist Ploy: Carl Schmitt on Dictatorship, Liberalism, and Emergency Powers." In Peter Baehr and Melvin Richter (eds.), Dictatorship in History and Theory: Bonapartism, Ceasarism, and Totalitarianism. Cambridge: Cambridge University Press. Pages 197-220. Makarenko, Boris. 2011. "The Role of Elections in Democracy." In In Adam Przeworski and others, Democracy in a Russian Mirror. Forthcoming. Manela, Erez. 2007. The Wilsonian Moment: Self-Determination and the International Origins of Anticolonial Nationalism. Oxford: Oxford University Press. Manin, Bernard. 1997. The Principles of Representative Government. Cambridge: Cambridge University Press. Munck, Gerardo L. 2009. Measuring Democracy: A Bridge between Scholarship and Politics. Baltimore: The John Hopkins University Press. Nicolet, Claude. 2004. "Dictatorship in Rome." In Peter Baehr and Melvin Richter (eds.), Dictatorship in History and Theory: Bonapartism, Ceasarism, and Totalitarianism. Cambridge: Cambridge University Press. Pages 263-278. O’Donnell, Guillermo, Philippe C. Schmitter, and Laurence Whitehead. 1986. Transitions from Authoritarian Rule: Tentative Conclusions about Uncertain Democracies. Baltimore: The Johns Hopkins University Press. Pasquino, Pasquale. 2010. "Machiavel: dictature et salus reipublicae." In Brigitte Krulic (ed.), Raison(s) d’Etat(s) en Europe, Traditions, usages, et recompositions. Pages 12-34. 26

Pasquino, Pasquale. 2011. "Democracy: ancient and modern, good and bad." In Adam Przeworski and others, Democracy in a Russian Mirror. Forthcoming. Paz, Octavio. 1965. ”A Democracia e a América Latina.” Caderno de Cultura de O Estado de São Paulo, ano II, numero 128. Pinedo, Federico. 1946. En Tiempos de la Republica. Buenos Aires: Editorial Mundo Forense. Posada-Carbó, Eduardo. 1999. “Alternancia y república: Elecciones en la Nueva Granada y Venezuela, 1835-37.” In Hilda Sábato (ed.), Ciudadanía política y formación de las naciones. Perspectivas históricas de América Latina. México: Colegio de México. Posada-Carbó, Eduardo. 2000. “Electoral Juggling: A Comparative History of the Corruption of Su¤rage in Latin America, 1830-1930.” Journal of Latin American Studies 32: 611-644. Posada-Carbó, Eduardo. 2008. "Origins of Democracy in Colombia, 18081886." Paper presented at the Seminar on the Origins of Democracy in the Americas, Notre Dame University, September 18-21. Przeworski, Adam. 2008. "Constraints and Choices: Electoral Participation in Historical Perspective." Comparative Political Studies 20: 1-27. Przeworski, Adam. 2009a. "The Mechanics of Regime Instability in Latin America." Journal of Politics in Latin America 1: 5-36. Przeworski, Adam. 2009b. "Conquered or Granted? A History of Su¤rage Extensions." British Journal of Political Science 39: 291-321. Przeworski, Adam. 2010. Democracy and the Limits of Self-Government. New York: Cambridge University Press. In Spanish: Que Esperar de la Democracia. Buenos Aires: Siglo XXI. Przeworski, Adam. 2011. "Political Institutions and Political Order." In Adam Przeworski and others, Democracy in the Russian Mirror. Forthcoming. Przeworski, Adam. 2011. PIPE: Political Institutions and Political Events Data Set. Department of Politics, New York University. Przeworski, Adam, Michael E. Alvarez, José Antonio Cheibub, and Fernando Limongi. 2000. Democracy and Development. New York: Cambridge University Press. Przeworski, Adam, Tamar Asadurian, and Anjali Bohlken. 2010. "The Origins of Parliamentary Responsibility." Revised paper presented at the Conference on Constitutional Design, University of Chicago Law School, October 16-17, 2009. Rippy, Fred J. 1965. ”Monarchy or Republic?”In Hugh M. Hamill, Jr. (ed.), Dictatorship in Spanish America. New York: Alfred. A. Knopf. Pages 86-94. Rojas, Rafael. 2009. Las repúblicas de aire. Utopía y desencanto en la revolución de Hispanoamérica. México: Taurus. Rosanvallon, Pierre. 1995. ”The History of the Word ’Democracy’ in France.” Journal of Democracy 5(4): 140-154. Rouquie Alain. 1994. “The military in Latin American Politics since 1930.” In Leslie Bethell (ed.), The Cambridge History of Latin America, Volume VI,

27

Part 2: 1930 to the Present. Cambridge: Cambridge University Press. Pages 233-306. Sa¤ord, Frank. 2008 (1985). "Politics, Ideology, and Society in PostIndependence Spanish America." In Cambridge History of Latin America, vol. 3, available from Cambridge Histories on Line, Cambridge University Press, pages 347-421. Saguir, Julio. 2011. "El Congreso de 1816-19 y la Constitución de 1819." Ms. Schedler, Andreas. 2006. Electoral Authoritarianism: The Dynamics of Unfree Competition. Boulder, CO: Lynn Rienner. Smith, Peter H. 2005. Democracy in Latin America. New York: Oxford University Press. Vanhanen, Tatu. 1997. Prospects of Democracy: A Study of 172 Countries. New York: Routledge. Vallenilla Lanz, Laureano. 1919. Cesarismo democratico: estudios sobre las bases sociólogicas de la constitución efectiva de Venezuela. Caracas. Wantchekon, Leonard. 2004. "The Paradox of ’Warlord Democracy’: A Theoretical Investigation." American Political Science Review 98: 17-33. Weisberger, Bernard A. 2000. America on Fire: Je¤ erson, Adams, and the First Contested Election. New York: HarperCollins. Zeldin, Theodore. 1958. The Political System of Napoleon III. New York: W.W. Norton. Zimmermann, Eduardo. 2008. "Elecciones y Representación politíca: Los Origines de una Traditición Democratica en la Argentina, 1810-1880." Paper presented at the Seminar on the Origins of Democracy in the Americas, Notre Dame University, September 18-21.

28