Explaining variation of executive instability in presidential regimes: Presidential interruptions in Latin America

Abstract

Through a conceptual and comparative analysis of 14 presidential interruptions in Latin America between 1980 and 2010, this article seeks to improve the current conceptualization of executive instability in presidential regimes and provide contingent answers to three debates: 1) ‘Do political institutions or pressure from below constitute the greater peril to presidential survival?’; 2) ‘Do presidential interruptions constitute a solution to an ongoing crisis, or further deepen the crisis?’; and 3) ‘Are presidential interruptions good or bad for democracy?’ The article argues that these questions have not been answered satisfactorily because the literature has assumed unit homogeneity, that is, that all interruptions are equal in terms of antecedents and aftermath, when in fact the cases of interruption demonstrate heterogeneity on these issues. This heterogeneity can be explained by two variables: the opposition’s primary motivation for challenging the president; and the degree of undemocratic behaviour demonstrated by the president and opposition during the crisis. Finally, based on these two variables, the article presents a typological map of crises and interruptions that helps define the scope conditions of the concept, captures the heterogeneity between the cases and seeks to provide a useful tool for future analysis of presidential interruptions.

Keywords

democracy executive instability Latin America presidentialism

Introduction

This article deals with executive instability in democratic presidential regimes, a phenomenon called presidential interruptions.¹ Despite criticism that presidential regimes cause regime instability (Linz, 1994), the fixed terms should provide government stability. Nevertheless, since the start of the third wave of democratization, 14 elected presidents in Latin America have been forced to leave office prematurely.² These important and critical events constitute the sample for this analysis. Through a conceptual and comparative analysis, I seek to provide answers to three unresolved debates in the literature on executive instability in Latin America: 1) ‘Do institutions or pressure from below (‘the streets’) constitute the greater peril to presidential survival?’; 2) ‘Do presidential interruptions constitute a solution to an ongoing crisis, or further deepen the crisis?’; and 3) ‘Are presidential interruptions good or bad for democracy?’

First, I argue that the literature has yet to come up with any agreement or satisfactory answers to these questions because most of the literature assumes that all interruptions are equal and ignores the systematic heterogeneity between the cases of interruption. Second, the literature has overlooked two variables that explain the heterogeneity linked to the three questions above: the opposition’s primary motivation for challenging the president; and the degree of undemocratic behaviour demonstrated by the president and opposition during the crisis. Third, I use these two variables to construct a typological map of presidential crises and interruptions. The typology sums up my argument, sorts the cases of interruption, delineates the scope of the concept and provides guidelines for case selection in the study of executive instability in presidential regimes. Consequently, this article not only seeks to resolve the ongoing debates on presidential interruptions, but also addresses and improves the conceptualization of the phenomenon itself.

I argue that the primary motivation for the opposition to remove the president can either be based on policy or on legal-constitutional issues. In the first case, street pressure tends to be more important for removing presidents than institutional pressure, and the interruptions do not solve the ongoing crisis, with the result that the social and political turmoil continues after the presidential ouster. In the second case, an opposition acting within democratic institutions tends to be the more important challenger, the presidential removal tends to solve the ongoing crisis, and the turmoil dies out rapidly. The second variable addresses the scope of the concept and the degree to which a presidential interruption is good or bad for democracy. The removal of an undemocratic president is normally good for democratic development, whereas the removal of a president who has not violated constitutional procedures by an undemocratic opposition, on the other hand, contributes to democratic erosion.

The article proceeds as follows. First, I introduce the topic of presidential interruption in Latin America and discuss the literature and the heterogeneity between the cases of presidential interruption. Second, I present and discuss the two omitted variables that I argue explain the observed heterogeneity between the cases. Third, I structure the analysis around the three questions above, and provide contingent answers to these questions. Fourth, I construct a typology of presidential crises and interruptions, sort the cases and discuss the potential merits of this typology.

Executive instability in presidential regimes

Quite separate from the debate regarding presidentialism’s merits related to regime stability (see, e.g., Cheibub, 2007; Linz, 1994; Shugart and Carey, 1992) is the fact that, barring democratic breakdowns, a presidential regime should provide executive stability. Nevertheless, in Latin America, many elected presidents have not been able to complete their terms, and during the current democratic period that began in 1978, executive instability has been decoupled from regime instability. Today, presidents may fall unexpectedly, but democracies tend to survive. In this article, I call this new form of executive instability ‘presidential interruption’, which is defined by a premature, extraordinary and forced exit of an elected president that does not lead to a democratic breakdown.³ Among students of presidential interruptions, there are some minor differences in the definition of the phenomenon and, consequently, the cases studied, but most authors would agree that 14 of the 15 cases listed in Table 1 qualify as interruptions, and that the removal of President Zelaya in Honduras in 2009 is a borderline case between an interruption and a breakdown. In the last section, I discuss briefly the Zelaya case in order to analyse the scope of presidential interruptions.

Table 1.

Presidential interruptions in Latin America since 1980.

President	Country	Year
Hernán Siles Zuazo	Bolivia	1985
Raúl Alfonsín	Argentina	1989
Fernando Collor de Melo	Brazil	1992
Carlos Andrés Pérez	Venezuela	1993
Jorge Serrano	Guatemala	1993
Joaquín Balaguer	Dominican Republic	1994/1996
Abdalá Bucaram	Ecuador	1997
Raúl Cubas	Paraguay	1999
Jamil Mahuad	Ecuador	2000
Alberto Fujimori	Peru	2000
Fernando de la Rúa	Argentina	2001
Gonzalo Sánchez de Lozada	Bolivia	2003
Lucio Gutiérrez	Ecuador	2005
Carlos Mesa	Bolivia	2005
Manuel Zelaya	Honduras	2009

Note: I include the case of Zelaya to analyse and define the scope of the concept of interruption (see the last section).

Democratic breakdowns and presidential interruptions constitute different phenomena, yet several scholars have applied the breakdown literature to the study of presidential interruptions. Valenzuela (2004) links presidential interruptions to the perils of presidentialism and institutional deadlocks, and argues that it is a combination of the regime type and the minority status of presidents that causes interruptions. Although their emphases vary, a host of other scholars agree with Valenzuela that institutions are key to explaining the fall of presidents (Mejía Acosta and Polga-Hecimovich, 2010; Morgenstern et al., 2008; Negretto, 2006; Pérez-Liñán, 2007).⁴ Against the institutionalist view, others argue that interruptions are triggered by popular mobilization from below, street protests and concentrated periods of social uproar (e.g. Hochstetler, 2006; Hochstetler and Edwards, 2009; Wolff, 2007; Zamosc, 2007). Hochstetler (2006) notes that all interruptions in South America were preceded by major protests on the streets, and that every presidential challenge not supported by popular mobilization failed. In fact, taking inspiration from Stepan (1971), she argues that civil society and the streets have taken over the role of political moderator (poder moderador) from the military in Latin America. Therefore, she concludes that it is the ‘streets’ and popular mobilization, not presidentialism or its institutions, that constitute the greatest peril for presidents. Although institutionalists also recognize the role played by the ‘streets’ (e.g. Pérez-Liñán, 2007, 2008), there is still no consensus emerging from the debate (see Llanos and Marsteintredet, 2010b).

Less is known about the aftermath of presidential interruptions. The cited literature has discussed, however, whether interruptions constitute a solution to the crisis, and their implications for democracy. Valenzuela (2004) sees presidential interruptions as a problem for democratic development, and suggests parliamentarianism as a solution. Others argue that presidential interruptions mitigate the perils of presidentialism, since the regimes in question find a peaceful way out of serious political conflicts while avoiding full breakdown (Marsteintredet and Berntzen, 2008; Schamis, 2002). Hochstetler and Samuels (2011) adopt a position between the first two views; they contend that the predominant tendency is to re-equilibrate the political situation. Finally, other scholars suggest that the outcome strengthens congress vis-a-vis the presidency (Pérez-Liñán, 2005) and increases levels of horizontal and vertical accountability (Marsteintredet, 2008). At least implicitly, by invoking the concept of poder moderador, Hochstetler (2006) also hints that social movements may hold presidents accountable. The questions remain, however, whether presidential interruptions serve to resolve an ongoing political crisis or generate more conflict and instability, and whether they strengthen or weaken democracy.

The main reason for the lack of agreement regarding these questions is that the cited literature treats all interruptions as equal,⁵ when in fact the cases display considerable heterogeneity on key variables. The assumption of homogeneity leads researchers to expect similar antecedents, aftermaths and consequences in all cases of interruption. This problem is aggravated by the relatively few cases of interruptions, which makes analysing average causal effects and consequences more difficult. This article, on the other hand, seeks to identify and explain complexity, differences and similarities across the cases of presidential interruptions in Latin America.

Presidential interruptions in Latin America: A heterogeneous phenomenon

A cursory review demonstrates the great variation between the cases of presidential interruptions in terms of the principal challengers, aftermath of interruptions and implications for the regimes. Analysing the challengers, street protests are critical to the explanation of the fall of Presidents de la Rúa in Argentina (Wolff, 2007), Bucaram in Ecuador (Luna, 1997) and Sánchez de Lozada in Bolivia (Lucero, 2008: 123–127). On the other hand, the fall of Presidents Balaguer in the Dominican Republic and Serrano in Guatemala saw little or no street action (Hartlyn, 1998; -Villagrán de León, 1993). Finally, in the case of Fujimori, the greatest street demonstration against the president occurred in late July (La marcha de los cuatro suyos) (Carrión, 2006a: 315), while Fujimori fled to Japan as late as November 2000 after having lost the majority in congress.

To investigate whether the interruption is the culmination of an ongoing crisis or only deepens the crisis, I compare the degree to which social and political turmoil ends following the presidential ouster. Again, there is extensive variation in this respect. In Brazil and Paraguay, the street protests and inter-institutional conflict died out with the successful impeachment of Presidents Collor de Melo and Cubas, respectively. In Bolivia and Argentina, however, after the presidential interruptions, social and political turmoil continued and haunted the successors to the ousted presidents (Arce and Rice, 2009; Corrales, 1997; Malloy and Gamarra, 1988; Wolff, 2007).

In terms of the implications for the regimes in question, the impeachment of President Cubas was interpreted as important for the further democratization of Paraguay (Abente-Brun, 1999), and Schamis (2002) argues that the military’s absence during the crisis in Argentina in 2001/2002 was an important step in that country’s consolidation of democracy. Moreover, few would argue that the early exits of Presidents Fujimori and Balaguer harmed democratic regime development in Peru and the Dominican Republic. On the other hand, the civil–military operation that removed President Jamil Mahuad in Ecuador suggests that the implications of presidential interruptions vary from case to case. Indirectly, presidential interruptions in the Andes region have also been linked to concepts such as the crisis of democratic representation (Mainwaring et al., 2006) and democratic erosion (Mainwaring and Pérez-Liñán, 2005). I conclude that there is variation across cases on this aspect as well. The heterogeneity related to all three questions, however, remains unexplained.

Explaining variation: The opposition’s motivation and degree of undemocratic behaviour

Given the variation on key variables, it is no surprise that studies assuming unit homogeneity fail to provide satisfactory answers to the debates raised in the literature, and that researchers differ on these questions. The main problem is the lack of attention to the heterogeneity and complexity across cases of interruption. Therefore, we need different tools to improve the understanding of presidential interruptions in Latin America. I suggest that the variation across the cases of interruption can be explained by analysing two variables: the opposition’s motivation for challenging the president; and the president’s and opposition’s degree of undemocratic behaviour during the presidential crises.

The opposition’s motivation for challenging the president (interruptions as a reactive sequence)

The battle to remove a president is a struggle for power: the opposition seeks to discredit and remove the executive and sometimes also to increase its chances at seizing the presidency. Nevertheless, the challenge that topples a president is also a reaction to the chief executive’s prior behaviour, and can be understood as a reactive sequence (Mahoney, 2000). Consequently, removing a president is an extreme form of sanctioning and holding the chief executive accountable for actions or inactions, regardless of whether this form of accountability can be labelled as successful or not (Schmitter, 2004).

The opposition drives the process of presidential interruptions, and a common factor preceding all presidential interruptions is a challenge to the president (Hochstetler, 2006). The opposition’s motivation for mounting that challenge varies, however, and is linked to the president’s behaviour in office. In presidential regimes, the president can usually only be removed by impeachment. The impeachment procedure is normally restricted to a president’s illegal or unconstitutional behaviour. In contrast to parliamentary regimes, a president cannot, at least constitutionally, be removed for political reasons, or on the whim of the current majority in congress. Therefore, a natural distinction is whether the challengers are motivated by a president’s illegal or unconstitutional behaviour or by the president’s policies or political behaviour. If a president’s illegal or unconstitutional behaviour triggers a presidential challenge, congress, rather than street protesters, is likely to play an important role in deposing the president because the impeachment procedure is the adequate institutional response. Since, in these interruptions, the opposition’s demands are confined to removing the president, political and societal turmoil will decrease almost immediately after the president is removed.

The reverse is the case when the opposition seeks policy changes first, and demands for presidential removal appear only after some time. A chief executive’s (failed) policies do not warrant impeaching the president, and therefore congress is likely to play a minor role during interruptions motivated by policy demands. For the same reason, popular and street pressure are likely to play a much larger role in these cases. Widespread havoc in the streets is the only factor that could force a president to resign, or force congress to act against the incumbent. In these cases, however, unseating the president does not satisfy all the demands of the opposition. Therefore, the interruption is not the final solution to the ongoing crisis, and political and societal turmoil is likely to continue after the president’s ouster.

There are severable implications that follow from this distinction: if a president is challenged by an opposition motivated by the president’s illegal or unconstitutional behaviour, the most important challengers are likely to be institutional since such behaviour qualifies for impeachment, and the crisis is likely to die out almost immediately after the presidential interruption because all the demands of the challengers are met by the presidential removal. If opposition to a president is motivated by a president’s policies, the most important challengers are likely to be street-based since a president cannot be impeached because of political disagreements, and the crisis is likely to continue after the presidential interruption since only one of the opposition’s demands is met by presidential removal.

Degree of adherence to democratic rules during interruptions (the grey zone of executive instability)

A presidential interruption may also vary according to the degree to which the key actors play by democratic rules. This variable has two effects: first, it helps define the scope of the concept; and, second, it explains how an interruption affects democracy. An interruption is defined as an extraordinary and premature event; therefore, outside the scope of regular democratic elections. The scope boundaries, however, are not completely clear between a presidential interruption and a democratic breakdown. Both an interruption and a democratic breakdown may result from a coup that topples the president, but in the former case, democracy survives or is immediately restored. A Latin American example of a coup and interruption is the case of President Mahuad in Ecuador in 2000. Should democracy break down, however, the aftermath of the event should be analysed by other concepts and theories. A breakdown may be clearly identified by the closure of democratic institutions (e.g. Chile 1973), but, in some cases, democratic institutions survive and defining the new regime may be difficult (e.g. Honduras 2009). In the latter case, if key international actors such as the United Nations (UN), Organisation of American States (OAS), USA and neighbouring states argue that democracy has broken down, the case is considered a democratic breakdown and not an interruption. The reason is that, regardless of the objective truth of the matter, international condemnation will seriously affect the aftermath of the crisis, which will be driven by international pressure for re-democratization.

All changes of government involve risks for the regime in question (no matter how small), and between the two extremes of elections and coups, there is a continuum defined by the actors’ varying degree of adherence to democratic rules: both the opposition and the president may, to a different extent, adhere to democratic rules during the crisis. A president may attempt a self-coup when the opposition has done little more than obstruct legislation in congress; the opposition may ally with the military to overthrow the president even though a president has only failed to combat inflation. The effect of an interruption on a regime’s democratic development depends on whether it is the president or the opposition who represents the less democratic actor during the crisis. Removing an undemocratic president is a necessary step for democratic development, whereas allowing a disloyal opposition to remove a president is sufficient for the regime’s democratic deterioration.

Differentiating between interruptions based on the actors’ adherence to democratic rules has the following implication. If the president is the more undemocratic actor, removing him or her is likely to affect democracy positively. If the opposition is the more undemocratic actor and succeeds in removing the president, the interruption is likely to strengthen a semi-loyal or disloyal opposition, and thus harm democratic development.

Sources and coding of the variables

There are two variables to code: 1) the opposition’s primary and principal motivation for challenging the president; and 2) the actors’ degree of adherence to democratic rules during the process of interruption.

I argue that presidential interruptions and their outcomes should be analysed as reactive sequences in which the sequencing of events matters. Following the logic of path dependency (Mahoney, 2000), I establish a simplifying rule that codes the opposition’s motivation according to the first registered challenge. I do this for two reasons. First, the path-dependency literature holds that the order of events matters and that early stages of events are more important than later stages (Abbott, 1983; Pierson, 2004); the first event in the chain is the most important because it sets in motion a sequence of events. Second, given that some cases are fuzzy, a clear rule focusing on the first challenge makes discriminating between cases easier. I focus on the first registered challenge against presidents that can be connected to the removal of the president, as reported in the Latin American Weekly Report (LAWR, 1980–2009). This makes coding clear, replicable and verifiable. I corroborate the coding of LAWR with two additional sources of academic articles and books. LAWR reports weekly on political and economic ‘events’ and is therefore suited to register events such as congressional and/or ‘street’ challenges to presidents and, in particular, their timing (see also Hochstetler, 2006; Negretto, 2006; Pérez-Liñán, 2007).

I use the same sources to code the second variable: the actors’ degree of adherence to democratic rules. Here, I actually measure two variables: the president’s and the opposition’s behaviour, which in sum taps into the degree of threat to democracy during the interruption (see also Arceneaux and Pion-Berlin, 2007).⁶ I thus distinguish between different levels of undemocratic behaviour by each actor during the interruption. I create an ordinal variable of six values that measures the actors’ degree of adherence to democratic rules: 0 represents the least undemocratic behaviour (the opposition impeaching, the president does not retaliate against the challenge), 5 the maximum undemocratic behaviour (coup by the opposition, self-coup by the president). For a detailed list of sources, coding rules and placement of cases, see the Appendix.

Explaining variation of presidential interruptions in Latin America

I argue that the variation concerning the three issues related to presidential interruptions in Latin America can be explained by variation on the two variables discussed above. I now discuss the three questions and why the cases of interruption demonstrate such heterogeneity.

Do presidents fall by pressure from below or by institutions?

Table 2 displays whether the challenge was motivated primarily by policy issues or legal-constitutional issues. At the bottom of the table are presidents interrupted after legal-constitutional challenges: Presidents Collor de Melo in Brazil,⁷ Serrano in Guatemala, Cubas in Paraguay, Balaguer in Dominican Republic and Fujimori in Peru.⁸ In these cases, the level of street challenges was moderate to low. And, if there were street challenges, they were concentrated in time, appeared after congress or another institution had challenged the president, or functioned to pressure and convince congress to impeach the presidents, such as in the cases of Brazil and Paraguay.⁹

Table 2.

Explaining variation of presidential interruptions.

Presidents	Type of challenge	Street challenge	Institutional challenge	Continued turmoil after interruption
Hernán Siles Zuazo	Policy	High	Intermediate	Yes
Raúl Alfonsín	Policy	High	Low	Yes
Carlos Andrés Pérez	Policy	High	High	Yes
Abdalá Bucaram	Policy	High	Intermediate	Yes
Jamil Mahuad	Policy	High	Low	Yes
Fernando de la Rúa	Policy	High	Intermediate	Yes
Gonzalo Sánchez de Lozada	Policy	High	Low	Yes
Lucio Gutiérrez	Policy	High	Intermediate/high	Intermediate
Carlos Mesa	Policy	High	Intermediate	Yes
Fernando Collor de Melo	Legal-constitutional	Intermediate	High	No
Jorge Serrano	Legal-constitutional	Low	Intermediate/high	No
Joaquín Balaguer	Legal-constitutional	Low	Intermediate	No
Raúl Cubas	Legal-constitutional	Intermediate	High	Yes
Alberto Fujimori	Legal-constitutional	Low	Intermediate	No
Manuel Zelaya	Legal-constitutional	Low	High	Intermediate

Notes: Type of challenge given by primary challenge. For detailed sources and coding, see the Appendix to this article.

The reason for the relatively minor role of the streets in these latter cases is that the president had committed acts that were illegal or unconstitutional and thus was liable for institutional sanctions, such as impeachment. Collor de Melo and Cubas were impeached on charges of theft and murder, respectively, and in Guatemala, President Serrano’s self-coup was declared unconstitutional by the Constitutional Court, a declaration that unravelled Serrano’s scheme.¹⁰ In Peru, after having lost control of Congress, Fujimori feared impeachment over the Vladivideo scandal and fraudulent elections in April 2000; he fled the country before Congress could take action against him (Cameron, 2006; McClintock, 2006).

What about the impeachment of Carlos Andrés Pérez? The first challenge to Pérez was a policy challenge, and the case follows the pattern predicted by theory: street pressure and challenges were at a high level since the great riot in Caracas (the caracazo) against Pérez’s neo liberal turnaround in 1989 (López Maya, 2003). President Pérez, however, fell as a result of impeachment in 1993, after taking personal advantage of the exchange system RECADI (Oficina del Régimen de Cambios Diferenciales) just before his new policies abolished it in 1989. The scandal burst in November of 1992, but before that, Congress, without a formal cause, unsuccessfully attempted to remove Pérez several times as a reaction to the social pressure against him. The RECADI scheme, however, gave Congress the constitutional pretext to impeach and remove the president. In this case, then, the levels of both street and institutional pressure were high and contributed to Pérez’s downfall.

In contrast to the cases in which the opposition is motivated by legal issues, popular pressure is relatively more important when the issue at stake is policy.¹¹ The reason is that failed, or unpopular, economic (or other) policies do not constitute constitutional grounds to remove a president. This fact also explains why in Ecuador, for instance, congressional elites and former Presidents Borja, Hurtado and Febrés Cordero urged people to take to the streets to help them oust Presidents Bucaram, Mahuad and Gutiérrez (Hernández et al., 2000; Ospina Peralta, 2005; Pachano, 1997, 2005). Lacking a constitutional rationale for presidential removal, popular pressure creates a generalized sentiment of an ungovernable situation that helps put pressure and strains on the administration,¹² and convinces other institutions, such as congress or a supreme court, to act against the president, often after the fact.¹³

Do presidential interruptions constitute a solution to the ongoing crisis, or further deepen the crisis?

When challengers are motivated by a president’s violations of the law or the constitution, the removal of the president satisfies the opposition’s demands. Therefore, social or political turmoil ends after the president’s removal. In these cases, the opposition’s battles have been won, the president has been held accountable and there are no longer any reasons to continue the fight. Therefore, the regime re-equilibrates rapidly (Hochstetler and Samuels, 2011), and the presidential interruption is indeed a solution to the ongoing crisis. In the aftermath of the cases of Collor de Melo, Serrano, Balaguer and Fujimori, there were few major street protests, institutional deadlocks or challenges to their successors. The exception to this pattern is the aftermath of the impeachment of President Cubas in Paraguay. Here, the turmoil continued, since the government of Cubas’s successor, President Gonzalez Macchi, was involved in several other cases of corruption (Pérez-Liñán, 2007: 130–131). None of this turmoil, however, was related to the issues that toppled Cubas, namely, the unconstitutional liberation of General Lino Oviedo and the killing of Vice-President Argaña. These were resolved with Cubas’s impeachment.

In the remaining cases, the popular pressure that toppled the presidents did not end with their removal. The opposition initially enunciated policy demands, and only after some time engaged in anti-establishment protests and voiced demands for presidential removals. The removal of the president in these cases only satisfied one of the demands of the opposition, and did not end the political crisis.¹⁴ The policy demands were not immediately satisfied, or easy to satisfy at all. Therefore, following presidential interruptions, the opposition remained mobilized on policy issues and the level of conflicts continued, leading to more challenges and presidential interruptions in some cases. In contrast to an interruption motivated by a president’s illegal action, interruptions motivated by policy may be considered part of the processes of social change. The successors to the interrupted presidents immediately meet demands for policy changes, and seek to resolve the problems that generated the protest. Consequently, these cases of interruption also tend to be followed not only by continued turmoil, but also by policy changes, which by themselves might cause a continuation of social and institutional pressure against the presidents (see Marsteintredet, 2010: 256–261).

In Argentina after Alfonsín’s resignation in June 1989, political street protests, strikes and inter-institutional conflicts continued at the same level throughout 1990 (Corrales, 1997). Not until 1991, when hyperinflation, one of the principal causes of the fall of Alfonsín, was brought under control, did the level of conflicts on the streets and in congress diminish (Weyland, 2002: 126–127). In Bolivia after Siles Zuazo’s early exit in 1985, street conflicts and strikes continued at a high level, especially in reaction to the neoliberal reform package launched by President Paz Estenssorro. They ended only after heavy repression (Conaghan et al., 1990: 24–25). In Venezuela, the high level of protest and pressure continued during the presidencies of both successors to President Pérez (caretaker Ramón Velásquez and Rafael Caldera) (López Maya and Lander, 2005).

In the interruptions in Ecuador (1997, 2000, 2005), Bolivia (2003, 2005) and Argentina (2001), not only did the level of conflict remain at a high level after the presidential interruptions, but also more interruptions and challenges followed.¹⁵ In Ecuador, the conflicts surrounding the economic policies that toppled Bucaram in 1997 continued after his fall, and the recently mobilized indigenous groups maintained the pressure against Mahuad (Zamosc, 2007). In 2000, Lucio Gutiérrez allied with the indigenous movement and removed Mahuad, and in the 2002 elections, Gutiérrez went to the polls in alliance with the indigenous party Pachakutik, promising new economic policies. Gutiérrez’s continuation of his toppled predecessor’s policies, several scandals and his attempts to take control of the Supreme Court assured him strong opposition in Congress and the streets, and led to his downfall in 2005 (Polga-Hecimovich, 2010). In Bolivia, the water wars and the thorny issue of gas exports through Chilean ports (Crabtree, 2005) generated enough opposition against President Sánchez de Lozada to oust him from the presidency. Congressional and street pressure against his successor, President Mesa, was no lower than against Sánchez de Lozada. In 2005, Mesa, unable to satisfy the opposition’s political demands, fell as well (Buitrago, 2010). In Argentina, the economic crisis led to de la Rúa’s early exit, but the piquetero movement continued its pressure from the streets against Duhalde, de la Rúa’s successor (Wolff, 2007: 6), and ended the presidency of Duhalde early, in April 2003. In sum, the aftermath of these interruptions contrasts starkly with the aftermath of interruptions motivated by legal-constitutional issues.

Are presidential interruptions good or bad for democracy?

The overarching question regarding presidential interruptions in Latin America is whether the events strengthen or weaken democratic development.¹⁶ Due to the heterogeneity across cases, it is difficult to come up with a single answer that covers all cases. In my view, the answer hinges upon whether it is the president or the opposition that has demonstrated the most undemocratic behaviour prior to the interruption, but also, partly, whether or not there is a valid legal or constitutional ground for removing the president. Also helpful for this analysis is the counterfactual: given the president’s and opposition’s behaviour, would presidential survival have been better for democracy? Table 3 portrays the degree of undemocratic behaviour demonstrated by the opposition and the president during the interruptions.

Table 3.

Interruptions and degrees of undemocratic behaviour.

Presidents	Opposition’s motivation	President’s degree of undemocratic behaviour	Opposition’s degree of undemocratic behaviour
Hernán Siles Zuazo	Policy	0	2
Raúl Alfonsín	Policy	0	1
Carlos Andrés Pérez	Policy	2	2
Abdalá Bucaram	Policy	1	1
Jamil Mahuad	Policy	0	5
Fernando de la Rúa	Policy	2	2
Gonzalo Sánchez de Lozada	Policy	2	1
Lucio Gutiérrez	Policy	3	1
Carlos Mesa	Policy	0	1
Fernando Collor de Melo	Legal-constitutional	0	0
Jorge Serrano	Legal-constitutional	5	0
Joaquín Balaguer	Legal-constitutional	4	0
Raúl Cubas	Legal-constitutional	3	0
Alberto Fujimori	Legal-constitutional	4	1
Manuel Zelaya	Legal-constitutional	3	5

Notes: Based on a scale from 0 (no undemocratic behaviour) to 5 (maximum undemocratic behaviour). For detailed coding and sources, see the Appendix to this article.

The classification of cases coincides with a similar coding of crises and democratic dilemmas in Latin America (see Arceneaux and Pion-Berlin, 2007), a fact that increases the validity of the coding. Table 3 shows that there are cases in which one of the main actors has violated core democratic principles: the opposition in a civil–military coup in Ecuador in 2000 (coded 5); Presidents Fujimori and Balaguer through organizing fraudulent elections (coded 4); and President Serrano for organizing a self-coup (coded 5). The combined scores of both actors in the case of Zelaya demonstrates the highest level of undemocratic behaviour.¹⁷ At the other end of the scale, we find the impeachment of Collor de Melo, which involved a thieving president and a moderate opposition that impeached the president. Between these cases, we find a host of cases that score moderately (1–3) on the scale of undemocratic behaviour. This is not surprising: presidential interruptions occur somewhere in the grey zone between regular democratic elections and fully fledged democratic breakdowns.

In the extreme cases of presidents violating core democratic principles, we find presidents who illegally attempt to extend their terms as president. In these cases, either the president leaves power (the case of Serrano) or democracy ends (the case of Fujimori’s self-coup in 1992). The removals of Fujimori and Balaguer in Peru and the Dominican Republic were therefore regarded as a second transition to democracy (Cameron, 2006; Hartlyn, 1998; McClintock, 2006), while the removal of President Serrano and Vice-President Espina saved that democracy from a breakdown. Although not sufficient, the removal of these presidents was clearly necessary for positive democratic development. Given the fact that these presidents were removed due to their undemocratic behaviour, new leaders were given a window of opportunity to make necessary democratic reforms, which could affect democratic development positively.¹⁸ To conclude, in the cases in which the president demonstrates apparently undemocratic behaviour, the removal of the president is clearly better for democracy than would be presidential survival. I hold the same conclusion to be valid in the two remaining cases, where the president proved to be more undemocratic than the opposition.¹⁹

Looking at the cases in which the opposition was the more undemocratic actor, it becomes clear why it is so important to distinguish between the president’s and the opposition’s behaviour when evaluating the effect a presidential interruption has on democracy. The most extreme case is the civil–military coup against President Jamil Mahuad in Ecuador in early 2000. The coup was orchestrated by the indigenous movement Confederación de Nacionalidades Indígenas de Ecuador (CONAIE) and a group of mid-level officers in the army. After Mahuad was removed, Lieutenant Colonel Lucio Gutiérrez and CONAIE leader Antonio Vargas declared a government of national salvation. Only strong international pressure, which made the top military brass condemn the coup, saved electoral democracy. An interruption due to undemocratic behaviour on the part of the opposition demonstrates that a disloyal or semi-loyal opposition is able to disrupt the democratic game, act as a veto-player over policies and contend for power through coups. In this case, the removal of the president is worse for democracy than the counterfactual, namely, the survival of a democratically elected president and the failure of a disloyal opposition. Thus, taking into account the degree of undemocratic behaviour by the key actors helps sort cases according to the impact that interruptions have on democracy.

The remaining cases score on the lower end of the scales of undemocratic behaviour, and in these cases, conclusions regarding the effect of interruptions on democracy are less obvious. Therefore, for these cases, it is useful to consider the opposition’s motivation for removing a president. In presidential democracies, removing a president for policy reasons is more harmful for democracy than removing a president for impeachment-qualifying behaviour (e.g. corruption). The impeachment of a corrupt president (the case of Collor) should not have any effect on democracy. On the other hand, seeking to remove a president between regular elections for policy reasons demonstrates that the opposition is semi-loyal towards the democratic rules of the game, and is inclined to play outside the rules. This sort of semi-loyalty can clearly be observed in the cases of interruption in Argentina, Bolivia and even in the case of Carlos Andrés Pérez before he was eventually impeached. When the challenges commenced in these cases, there were no formal grounds for removing the president. Except for the case of Pérez in Venezuela, and barring police violence, the presidents in question during these crises did not behave in a manner that unequivocally was undemocratic or qualified for impeachment. Even though the opposition does not seek to eliminate the democratic system, by using their strength in the streets rather than waiting for elections to demonstrate strength in the voting booth, the opposition is clearly semi-loyal. The success and demonstration of strength of a semi-loyal opposition should, as Linz (1978) argues, be a concern for any democracy. Therefore, for these cases, Pérez-Liñán’s (2007: 211) conclusion that ‘there is no clear democratic principle to support the argument that protests should trump votes’ is valid.

Towards a typology of presidential interruptions?

O’Donnell (1988: 24) wrote that: ‘The term crisis is too general. Since the social and/or political dimensions of a crisis can attain very different levels of intensity, it is useful to distinguish among several kinds of crises.’ Presidential interruptions are the result of different types of crises, which generate different types of interruption.

However, given the variation between the cases, one might ask whether presidential interruption is an analytically useful concept. I believe the answer is ‘yes’, for two reasons. First, all the cases fit the definition of the concept and do not fit the definition of neighbouring phenomena, such as breakdown or regular governmental turnover via elections, and, thus, cannot be analysed by theories explaining these other phenomena. There is therefore a need for theories and concepts that analyse the cases in between. Second, instead of discarding the concept, which is now widely used, I suggest that it is preferable to systematize the variation between cases by constructing a typology of presidential crises and interruptions.

The overarching concept of interest here is the phenomenon of executive instability in presidential regimes, that is, presidential interruption. By generating different types of interruptions, the typology sums up, systematizes, maps and explains the analytically interesting variation described above. The typology has two additional advantages: it helps researchers select comparable negative cases for further comparison and case studies; and it situates the concept of interruptions, distinguishing it from presidential removals through regular elections on the one extreme and from democratic breakdowns on the other.

With some adjustments for within-type variation and fuzzy cases, Figure 1 maps presidential interruptions in Latin America according to the two variables presented above. The vertical axis in Figure 1 measures the opposition’s motivation for removing presidents, which ranges from exclusively legal or constitutional issues (bottom) to exclusively policy issues (top). It is converted from a dichotomous measure (see Table 2) to a fuzzy scale.²⁰ Towards the middle, we find the more fuzzy cases in which the opposition mobilized on both legal and policy issues. The horizontal axis measures the extent to which the interruption threatened democratic survival, and also defines the scope of the concept. The variable is coded as the maximum value of the scores of the president’s and opposition’s degree of undemocratic behaviour, as reported in Table 3.²¹ Allowing for fuzzy-scale variables, the figure pictures the perceived within-type variation and degree of fuzzy membership, instead of just generating a 2x2 table, which is the common strategy for multidimensional typologies (Collier et al., 2008). The dotted, vertical line on the right denotes the often blurry scope boundaries of the concept towards coups and democratic breakdowns. Marked by the dashed horizontal and vertical lines, the figure generates four types of crises and interruptions based on the cases’ score on these variables.

Figure 1.

A typological map of crises and presidential interruptions in Latin America.

The types of interruptions in the figure are here labelled as crises rather than interruptions to underscore the point that the typology can be used for negative case selection, that is, crises or challenges that did not end in presidential interruptions. In an interruption resulting from a legal crisis, the threat to democracy is quite low and the president is removed due to a violation of the law or the constitution. This is the kind of interruption that one could expect in a presidential regime. The remaining types of interruption are the ones that constitute the South or Latin American anomaly (Hochstetler and Edwards, 2009), and, I believe, are only possible in less institutionalized presidential democracies. Interruptions stemming from a constitutional crisis are cases in which there are legal-constitutional grounds for challenging the presidents, and the threat to democracy is considered quite high. In other words, democratic survival is at stake. In Latin America, the president has constituted the most undemocratic actor in interruptions resulting from this type of crisis; the opposite pattern is unobserved, but not impossible.²² Interruptions resulting from a policy crisis are the most numerous type in Latin America, and describe cases that indeed show parliamentary features, not because of the procedures by which the president is removed (Marsteintredet and Berntzen, 2008), but because the president was removed on account of failed or unpopular policies. Finally, an interruption may result from a combined constitutional and policy crisis, which involves the government and its policies, and endangers democratic survival. Due to the policy nature of the latter two types of crisis and interruption, these might be seen as part of more general political development and social change. The suggested typology thus captures the heterogeneity of the cases of interruption in Latin America, re-establishes unit homogeneity within each type, helps organize and analyse the data, and enhances our understanding of both the similarities and differences between cases of executive instability in presidential regimes.

The typology also addresses the scope of the concept. The case of the civil–military coup against President Zelaya in Honduras in 2009 merits further discussion. Is it a case of interruption or democratic breakdown? As in Ecuador in 2000, the democratic institutions were not closed down; in addition, all relevant elites and institutions inside Honduras argued that Zelaya had been removed in a constitutional manner. The case, however, falls outside the scope of presidential interruptions because the OAS, USA and most Latin American nations argued that what went down in Honduras was a coup. The international interpretation of the case as a coup and breakdown considerably affected the aftermath, and the case can no longer be explained by theories of interruption.

Finally, the suggested typology helps address a problematic topic in the study of presidential interruptions: the selection of negative cases (Mahoney and Goertz, 2004). A presidential interruption is a rare event of medium-N size. Identifying negative cases is difficult, and strategies have varied from large-N approaches with panel data using country-years of presidential survival as negative cases, or a somewhat more qualitative approach using challenges as negative cases (Hochstetler, 2006; Pérez-Liñán, 2007: 200). The suggested typology, however, addresses this issue and guides researchers in selecting negative cases (e.g. crises, challenges) that may fall within the same type as an actual interruption (most similar cases), or in the opposite corner on the diagonal (most different cases). One could also focus on only one of the dimensions and compare positive and negative cases of interruptions that constituted high threats to the regime. The typology can also be used together with other techniques for case selection (Gerring, 2007: 86–150), and, as such, could be helpful in framing arguments and research questions in future studies of executive instability in presidential regimes.

Concluding remarks

This conceptual and comparative analysis has sought to improve the conceptualization of executive instability in presidential regimes and explain the variation and heterogeneity among the cases of presidential interruptions in Latin America. The literature has treated all interruptions as being equal, and differences between cases of presidential interruptions have been neglected. The result has been a lack of satisfactory answers to the debates concerning whether institutions or the streets are the greater peril to executive stability; why some presidential interruptions constitute the end of a political crisis, while other interruptions only constitute a chapter of a lengthy process of political and social unrest; and what potential effect an interruption may have on democracy. I answered these questions through an analysis centred around two variables: the opposition’s motivation for challenging the president; and the actors’ degree of undemocratic behaviour during these crises. These variables explain the observed heterogeneity and provide contingent answers to these questions. Therefore, this analysis contributes to the debates regarding executive instability in Latin America and helps shed light on questions that find few satisfactory answers in analyses that assume that all interruptions are identical. The two variables also formed the basis for a typology of crises and interruptions that captures and maps the heterogeneity of the cases of interruption, helps delimit the scope of the concept, guides researchers in the selection of negative and positive cases of interruption, and provides a useful tool for comparing cases of executive instability in presidential regimes. The typology points to four types of crises and interruptions, which generate unit homogeneity within each type. Therefore, this article not only provides new answers to the current debates in the literature, it also improves the conceptualization of presidential interruptions. Finally, my analysis should, hopefully, serve as a warning against oversimplifying these highly complex events.

Footnotes

Appendix: Coding and sources for types of presidential interruptions

Table A2.

Coding of the level of threat to democracy.

President	President’s behaviour	Opposition’s behaviour	MAX value ‘Threat to democracy’
Siles Zuazo	0 (Peaceful resignation)	2 (Kidnapping of president)	2
Raúl Alfonsín	0 (Peaceful resignation)	1 (Street protests)	1
Collor de Melo	0 (Corruption)	0 (Impeachment)	0
Carlos Andrés Pérez	1 (Corruption, violent handling of protests)	2 (Street protests, first dubious congressional manoeuvre, then impeachment)	2
Jorge Serrano	5 (Self-coup)	0 (Impeachment)	5
Joaquín Balaguer	4 (Electoral fraud)	0 (Diplomacy, negotiations)	4
Abdalá Bucaram	1 (Refusing to be removed)	1 (Street protests)	1
Raúl Cubas	3 (Killing of Vice-President)	0 (Impeachment)	3
Jamil Mahuad	0 (Peaceful resignation)	5 (Civil–military coup)	5
Alberto Fujimori	4 (Electoral fraud)	1 (Street protests)	4
Fernando de la Rúa	1 (Police violence)	1 (Street protests)	1
Gonzalo Sánchez de Lozada	1 (Police violence)	1 (Street protests)	1
Lucio Gutiérrez	3 (Closing of Supreme Court)	1 (Street protests)	3
Carlos Mesa	0 (Peaceful resignation)	1 (Street protests)	1
Manuel Zelaya	3 (Illegal referendum)	5 (Civil–military coup)	5

Sources: See Table A1.

Notes: Based on a scale from 0 (no undemocratic behaviour) to 5 (maximum undemocratic behaviour). Coding rules: 0: Opposition – Demonstrations, Impeachment attempt; President – No retaliation, peaceful resignation; 1: Opposition – Continued street demonstrations, dubious congressional manoeuvre; President – Non-violent sabotaging of protests, for example, jailing opposition; 2: Opposition – Illegal behaviour/violent protests/threats of use of violence against president; President – Illegal behaviour/violent handling of protests/threats of use of violence against opposition leaders. 3: Opposition – Violence against government; President – Violent, targeted attacks, or unconstitutional attack on opposition/key actors/institutions; 4: Opposition – Boycotts democratic game, for example, elections, aims to delegitimize president sufficiently to have him ousted; President – Electoral fraud; 5: Opposition – Civil–military coup: President – Self-coup.

Acknowledgements

A previous version of this article was presented at the annual meeting of the American Political Science Association, Toronto, September 2009. I appreciate helpful comments from Michael E. Alvarez, Einar Berntzen, Javier Corrales, Jonathan Hartlyn, Kathryn Hochstetler, Scott Mainwaring and Gerardo Munck. I also thank four anonymous reviewers for valuable criticism and advice.

Funding

I appreciate economic support from the Christian Michelsen Institute in Bergen through ‘Nationalgaven’ to finalize this article.

Notes

Author biography

Leiv Marsteintredet is associate professor at the Department of Literature, Area Studies and European Languages, University of Oslo, and holds a PhD in Comparative Politics. His research interests include political crises in presidential regimes, presidentialism, political institutions and regime development in Latin America: Marsteintredet has co-edited (with Mariana Llanos) Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies (New York, NY: Palgrave, 2010) and has published articles in Comparative Politics, Latin American Politics and Society and América Latina Hoy.

References

Abbott

(1983) Sequences of social events: Concepts and methods for the analysis of order in social processes. Historical Methods 16(4): 129–147.

Abente-Brun

(1999) ‘People power’ in Paraguay. Journal of Democracy 10(3): 93–100.

Alfonsín

(2004) Memoria política. Transición a la democracia y derechos humanos. Buenos Aires: Fondo de Cultura Económica.

Arce

Rice

(2009) Societal protest in post-stabilization Bolivia. Latin American Research Review 44(1): 88–101.

Arceneaux

Pion-Berlin

(2007) Issues, threats, and institutions: Explaining OAS responses to democratic dilemmas in Latin America. Latin American Politics and Society 49(2): 1–31.

Buitrago

(2010) Civil society, social protest and presidential breakdowns in Bolivia. In: Llanos

Marsteintredet

(eds) Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies. New York, NY, and London: Palgrave Macmillan, 91–107.

Cameron

(2006) Endogenous regime breakdown: The Vladivideo and the fall of Peru’s Fujimori. In: Carrión

(eds) The Fujimori Legacy: The Rise of Electoral Authoritarianism in Peru. University Park, PA: Pennsylvania State University Press, 268–293.

Carey

(2005) Presidential vs. parliamentary government. In: Ménard

Shirley

(eds) Handbook of New Institutional Economics. Dordrecht: Springer, 91–123.

Carrión

(2006a) Conclusion: The rise and fall of electoral authoritarianism in Peru. In: Carrión

(eds) The Fujimori Legacy: The Rise of Electoral Authoritarianism in Peru. University Park, PA: Pennsylvania State University Press, 294–318.

10.

Carrión

(ed.) (2006b) The Fujimori Legacy: The Rise of Electoral Authoritarianism in Peru. University Park, PA: Pennsylvania State University Press.

11.

Cheibub

(2007) Presidentialism, Parliamentarism, and Democracy, Cambridge Studies in Comparative Politics. Cambridge: Cambridge University Press.

12.

Cheibub Figuereido

(2010) The Collor impeachment and presidential government in Brazil. In: Llanos

Marsteintredet

(eds) Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies. New York, NY, and London: Palgrave Macmillan.

13.

Collier

LaPorte

Seawright

(2008) Typologies: Forming concepts and creating categorical variables. In: Box-Steffensmeier

Brady

Collier

(eds) The Oxford Handbook of Political Methodology. Oxford: Oxford University Press, 152–173.

14.

Conaghan

Malloy

Abugattas

(1990) Business and the boys – The politics of neoliberalism in the Central Andes. Latin American Research Review 25(2): 3–30.

15.

Cooper

Legler

(2001) The OAS in Peru: A model for the future? Journal of Democracy 12(4): 123–136.

16.

Coppedge

(1994) Prospects of democratic governability in Venezuela. Journal of Interamerican Studies and World Affairs 36(2): 39–64.

17.

Corrales

(1997) Do economic crises contribute to economic reform? Argentina and Venezuela in the 1990s. Political Science Quarterly 112(4): 617–644.

18.

Crabtree

(2005) Patterns of Protest: Politics and Social Movements in Bolivia. London: Latin America Bureau.

19.

Díaz Santana

(1996) Trauma Electoral. Santo Domingo: Mograf S.A.

20.

Fukuyama

Dressel

Chang

B-S

(2005) Facing the perils of presidentialism? Journal of Democracy 16(2): 102–116.

21.

Gerring

(2007) Case Study Research: Principles and Practices. Cambridge: Cambridge University Press.

22.

Hartlyn

(1998) The Struggle for Democratic Politics in the Dominican Republic, H. Eugene and Lillian Youngs Lehman Series. Chapel Hill, NC: University of North Carolina Press.

23.

Hernández

Aráuz

Rodríguez

. (2000) 21 de enero. La vorágine que acabó con Mahuad. Quito: C.A. El comercio.

24.

Hochstetler

(2006) Rethinking presidentialism: Challenges and presidential falls in South America. Comparative Politics 38(4): 401–418.

25.

Hochstetler

Edwards

(2009) Failed presidencies: Identifying and explaining a South American anomaly. Journal of Politics in Latin America 1(2): 31–57.

26.

Hochstetler

Samuels

(2011) Crisis and rapid reequilibration: The consequences of presidential challenge and failure in Latin America. Comparative Politics 43: 127–145.

27.

INCEP (Instituto Centroamericano de Estudios Políticos) (1993) La crisis político-constitucional de Guatemala. Del golpe de Estado de Jorge Serrano a la Presidencia Constitucional de Ramiro de León Carpio. In: Panorama Centroamericano. Ciudad de Guatemala: Instituto Centroamericano de Estudios Políticos.

28.

Kim

Bahry

(2008) Interrupted presidencies in third wave democracies. The Journal of Politics 70(3): 807–822.

29.

Lalander

(2010) The impeachment of Carlos Andrés Pérez and the collapse of the Venezuelan Partyarchy. In: Llanos

Marsteintredet

(eds) Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies. New York, NY, and London: Palgrave Macmillan.

30.

LAWR (Latin American Weekly Report) (1980–2009). London.

31.

Linz

(1978) The Breakdown of Democratic Regimes: Crisis, Breakdown & Reequilibration. Baltimore, MD: Johns Hopkins University Press.

32.

Linz

(1994) Presidential or parliamentary democracy: Does it make a difference? In: Linz

Valenzuela

(eds) The Failure of Presidential Democracy. Baltimore, MD: Johns Hopkins University Press, 3–90.

33.

Llanos

(2010) Presidential breakdowns in Argentina. In: Llanos

Marsteintredet

(eds) Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies. New York, NY, and London: Palgrave Macmillan.

34.

Llanos

Marsteintredet

(2010a) Introduction: Presidentialism and presidential breakdowns in Latin America. In: Llanos

Marsteintredet

(eds) Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies. New York, NY, and London: Palgrave Macmillan.

35.

Llanos

Marsteintredet

(eds) (2010b) Presidential Breakdowns in Latin America. Causes and Outcomes of Executive Instability in Developing Democracies. New York, NY, and London: Palgrave Macmillan.

36.

Llanos

Marsteintredet

(2010c) Ruptura y continuidad: La caída de ‘Mel’ Zelaya en perspectica comparada. América Latina Hoy 55: 173–197.

37.

López Maya

(2003) The Venezuelan Caracazo of 1989: Popular protest and institutional weakness. Journal of Latin American Studies 35: 117–137.

38.

López Maya

Lander

(2005) Popular protest in Venezuela – Novelties and continuities. Latin American Perspectives 32(2): 92–108.

39.

Lucero

(2008) Struggles of Voice: The Politics of Indigenous Representation in the Andes. Pittsburght, PA: University of Pittsburgh Press.

40.

Luna

(1997) Bucaram, ¡¡Fuera!!: La voz de los movimientos profundos. In: Báez

Ribadeneira

Acosta

(eds) ¿Y ahora qué …? Una contribución al análisis político-histórico del país. Quito: Eskeletra Editorial, 229–264.

41.

Mahoney

(2000) Path dependence in historical sociology. Theory and Society 29(4): 507–548.

42.

Mahoney

Goertz

(2004) The possibility principle: Choosing negative cases in comparative research. American Political Science Review 98(4): 653–669.

43.

Mainwaring

Pérez-Liñán

(2005) Latin American democratization since 1978: Democratic transitions, breakdowns, and erosions. In: Hagopian

Mainwaring

(eds) The Third Wave of Democratization in Latin America: Advances and Setbacks. Cambridge: Cambridge University Press, 14–63.

44.

Mainwaring

Brinks

Perez-Liñán

(2001) Classifying political regimes in Latin America, 1945–1999. Studies in Comparative International Development 36(1): 37–65.

45.

Mainwaring

Bejarano

Leongómez

(eds) (2006) The Crisis of Democratic Representation in the Andes. Stanford, CA: Stanford University Press.

46.

Malloy

Gamarra

(1988) Revolution and Reaction: Bolivia, 1964–1985. New Brunswick, NJ, and Oxford: Transaction Books.

47.

Marsteintredet

(2008) Las consecuencias sobre el régimen de las interrupciones presidenciales en América Latina. América Latina Hoy 49: 31–50.

48.

Marsteintredet

(2010) Presidential Interruptions in Latin America: Concepts, Causes, and Outcomes. Monograph, Department of Comparative Politics, University of Bergen, Bergen.

49.

Marsteintredet

Berntzen

(2008) Reducing the perils of presidentialism in Latin America through presidential interruptions. Comparative Politics 41(1): 83–101.

50.

Mayorga

(2010) Enlargement of democracy and changes in the Bolivian party system. In: Lawson

Lanzaro

(eds) Political Parties and Democracy, Volume I. The Americas. Santa Barbara, CA: Praeger.

51.

Mayorga

(1994) Gobernabilidad y reforma política. La experiencia de Bolivia. América Latina Hoy 8: 35–60.

52.

McClintock

(2006) Electoral authoritarian versus partially democratic regimes: The case of the Fujimori government and the 2000 elections. In: Carrión

(ed.) The Fujimori Legacy: The Rise of Electoral Authoritarianism. University Park, PA: Pennsylvania State University Press, 242–267.

53.

Mejía Acosta

Polga-Hecimovich

(2010) Parliamentary solutions to presidential crises in Ecuador. In: Llanos

Marsteintredet

(eds) Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies. New York, NY, and London: Palgrave Macmillan, 73–90.

54.

Morgenstern

Negri

Pérez-Liñán

(2008) Parliamentary opposition in non-parliamentary regimes: Latin America. The Journal of Legislative Studies 14(1/2): 160–189.

55.

Negretto

(2006) Minority presidents and types of government in Latin America. Latin American Politics and Society 48(3): 63–92.

56.

O’Donnell

(1988) Bureaucratic Authoritarianism: Argentina 1966–1973 in Comparative Perspective. McGuire

Flory

trans. Berkeley and Los Angeles, CA: University of California Press.

57.

Ospina Peralta

(2005) El abril que se llevó al coronel que no murió en el intento. Ecuador Debate 65.

58.

Pachano

(1997) Bucaram ¡fuera! Bucaram ¿fuera? In: Báez

Ribadeneira

Acosta

. (eds) ¿Y ahora qué ...? Una contribución al análisis político-histórico del país. Quito: Eskeletra Editorial, 229–264.

59.

Pachano

(2005) Ecuador: Cuando la inestabilidad se vuelve estable. Íconos. Revista de Ciencias Sociales 23: 37–44.

60.

Pérez-Liñán

(2005) Democratization and constitutional crises in presidential regimes – Toward congressional supremacy? Comparative Political Studies 38(1): 51–74.

61.

Pérez-Liñán

(2007) Presidential Impeachment and the New Political Instability in Latin America. Cambridge: Cambridge University Press.

62.

Pérez-Liñán

(2008) Instituciones, coaliciones callejeras e inestabilidad política: Perspectivas teóricas sobre las crisis presidenciales. América Latina Hoy 49: 105–126.

63.

Pierson

(2004) Politics in Time: History, Institutions, and Social Analysis. Princeton, NJ: Princeton University Press.

64.

Polga-Hecimovich

(2010) Políticas, militares y ciudadanos: Un análisis de las caídas presidenciales en Ecuador (1997–2005). Quito: Universidad Andina Simón Bolivar Ecuador, Ediciones Abya-Yala, Corporación Editora Nacional.

65.

Ruhl

(2010) Honduras unravels. Journal of Democracy 21(2): 93–107.

66.

Schamis

(2002) Argentina: Crisis and democratic consolidation. Journal of Democracy 13(2): 81–94.

67.

Schmitter

(2004) The ambiguous virtues of accountability. Journal of Democracy 15(4): 47–60.

68.

Shugart

Carey

(1992) Presidents and Assemblies: Constitutional Design and Electoral Dynamics. Cambridge: Cambridge University Press.

69.

Slatopolsky Cantis

(1995) La Caída de los tres Gobiernos Radicales. Vol. 27, Colección Estudios. Buenos Aires: Editorial Centro de Estudios Unión para la Nueva Mayoría.

70.

Stepan

(1971) The Military in Politics: Changing Patterns in Brazil. Princeton, NJ: Princeton University Press.

71.

Valenzuela

(2004) Latin American presidencies interrupted. Journal of Democracy 15(4): 5–19.

72.

Villagrán

León

(1993) Thwarting the Guatemalan coup. Journal of Democracy 4(4): 117–124.

73.

Weyland

(1993) The rise and fall of President Collor and its impact on Brazilian democracy. Journal of Interamerican Studies and World Affairs 35(1): 1–37.

74.

Weyland

(2002) The Politics of Market Reform in Fragile Democracies: Argentina, Brazil, Peru and Venezuela. Princeton, NJ: Princeton University Press.

75.

Wolff

(2007) (De-) Mobilising the marginalised: A comparison of the Argentine Piqueteros and Ecuador’s indigenous movement. Journal of Latin American studies 39(1): 1–29.

76.

Zamosc

(2007) The Indian movement and political democracy in Ecuador. Latin American Politics and Society 49(3): 1–34.