Explaining presidential instability in Latin America: evidence from Brazil, Argentina and Ecuador

Abstract

Purpose

This study attempts to figure out the factors that contributed to deposing certain elected presidents before the end of their constitutional terms, alongside tracing the new political context that prevailed in Latin America since 1978 and its impact on direct political participation and military behavior during presidential crises.

Design/methodology/approach

The paper uses the comparative method to investigate the causes of presidential instability in three case studies.

Findings

The likelihood of presidential instability increases when a president enacts austerity economic policies that marginalize large sectors of the citizenry, becomes implicated in acts of corruption and develops a hostile relationship with members of the ruling coalition.

Originality/value

This study integrates the social movement theory with analytical perspectives from parliamentary behavior to explain presidential instability. It attempts to investigate the dynamics of interaction between the acts of furious citizens and disloyal legislators through the in-depth analysis of three case studies.

Citation

Abofarha, E.A. and Nasreldein, R.I. (2022), "Explaining presidential instability in Latin America: evidence from Brazil, Argentina and Ecuador", Review of Economics and Political Science, Vol. 7 No. 1, pp. 56-70. https://doi.org/10.1108/REPS-04-2021-0041

1. Introduction

By the mid-1940s, William Stokes (1945) highlighted the phenomenon of “Democratic Caesarism”, arguing that Latin American presidents enjoyed almost total hegemony over the political process and were unaccountable to other institutions. Half a century later, Linz (1990) wrote his article about “the perils of presidentialism”, noting that the presidential system is more prone to gridlock and lack of institutional cooperation, thus inviting military intervention. However, from the 1990s onwards, certain developments cast doubts on the validity of both claims. In several Latin American countries, the parliament toppled the president, often amid massive popular mobilization. Despite the succession of several presidential crises, democracy proved more resilient than was initially imagined.

Many scholars argue that since the 1990s, Latin America has undergone a new pattern of political instability, by which they mean several elected presidents were toppled before the end of their constitutional terms. The main players of this political drama are angry social movements and/or parliamentary opposition (Perez-Liñán, 2001; Hochstetler and Edwards, 2009; Marsteintredet, 2009). Gone are the days when the military used to intervene directly in presidential crises, either to set up a military dictatorship (e.g. Brazil 1964, Argentina 1975, Chile 1973) or to support one competing faction over another (e.g. Brazil 1955, Ecuador 1961) (Perez-Liñán, 2007). As the army prefers to stay quartered and neutral during most episodes of institutional conflict, presidential instability can be safely described as a government crisis that does not necessarily undermine the democratic nature of the political system (Negretto, 2006; Mello, 2015).

Presidential instability is closely related to the new political context that prevailed in Latin America since the onset of the third wave of democratization. Before 1978, military coups were considered a normal part of the political process. By 1977, only three countries (i.e. Columbia, Costa Rica and Venezuela) were democratic. But by 1990, the number of military dictatorships declined to one (Dix, 1994). The consequences of this major political shift were immense. It meant that the army is no longer the final arbiter in the political system when it comes to determining the outcome of a conflict between the executive authority and the legislative authority. This demilitarization of presidential crises was coupled with lowering the costs of direct political participation and increasing the protesting capacity of civil society (Lehoucq and Perez-Liñán, 2014).

In fact, different scholars use different terms as synonyms for presidential instability, such as “presidential fall,” “presidential interruption” and “executive interruption.” All these terms refer to a form of political instability that takes place owing to the forced exit of a democratically elected president before the end of his constitutional term. In the new age of democracy, the president is toppled by the parliamentary opposition or/and the angry social movements, signaling a government crisis that takes place within the limits of the democratic system (Marsteintredet, 2009). In other words, democracy does not break down with the deposing of a certain president.

This paper attempts to analyze the causes of presidential instability in Latin America through the systematic comparison of three cases: Brazil, Argentina and Ecuador. It applies Mill's method of agreement in order to establish the existence of similar causal patterns among these three cases that agree on the outcome to be explained (i.e. presidential failure). The paper is divided into five major sections. The first section will introduce the theory of presidential instability by combining perspectives from parliamentary behavior and the social movement theory. From the second to the fourth section, the paper delves into the causes of presidential instability by analyzing the impact of neoliberal economic policies, political scandals and the collapse of the governing coalition, respectively, as the three main factors responsible for the forced exit of certain presidents. The fifth section explains the causes of military behavior during presidential crises in Latin America.

2. Presidential instability in Latin America: theoretical considerations

In their endeavor to account for episodes of presidential instability, scholars of Latin America resorted to two prominent explanations. The first explanation focused on the role of furious social movements and the protest capacity of civil society (Hochstetler, 2006; Zamosc, 2007), while the second one highlighted the absence of a legislative majority capable of protecting the president from hostile measures by the opposition (Linz, 1990; Valenzuela, 1993).

After the onset of the third wave of democratization, some scholars held a pessimistic view concerning the prospects for democratic stability in the continent. They feared that the inherent characteristics of presidential regimes (i.e. dual independent sources of legitimacy for the executive and the legislature and the president's fixed term in office) can again cause gridlock and lead the army to directly intervene in politics. This strand of thought has mainly concentrated on the role of institutional arrangements in explaining presidential instability and democratic breakdown in the region. According to Linz (1990) and Valenzuela (2004), the nature of the presidential system per se is responsible for the high incidence of presidential instability and democratic breakdown in Latin America. Since most Latin American countries adopt proportional representation for electing the parliament, the majority of presidents there lack majority legislative support. Gridlock and intense conflict will certainly ensue due to the lack of peaceful mechanisms (i.e. vote of no-confidence or the call for early elections) to resolve the conflict between a minority president and an adamant parliament. This strand of thought was criticized for its depiction of a deterministic relationship between presidentialism and political instability, its belief in the inherent superiority of the parliamentary system and its failure to forecast the resilience of democracy in Latin America despite the recurrence of presidential crises. However, its emphasis on the importance of building a legislative majority as a means to ensure the survival of the president proved to be of high validity (Mainwaring and Shugart, 1997).

Witnessing the prominent role of civil society during presidential crises in Latin America, a new strand of thought developed that uses the rich perspectives of social movement theory. The main emphasis here is on the conditions that lead to a surge in the protesting capacity of social movements and their adoption of one of the most radical goals. Nearly one-third of Latin American presidents who adopted neoliberal policies were toppled by furious citizens who capitalized on existing civil society mobilizing structures (Hochstetler, 2006). In addition, the emergence of certain social cleavages and their inadequate representation by the political system seem to increase the likelihood of societal polarization and convince excluded groups to seek radical goals, including the toppling of the president (Marsteintredet, 2009). Presidents who face massive nation-wide demonstrations calling for their removal face a unique dilemma and are left with a limited range of alternatives, as ignoring these demonstrations may lead to an expansion of their scope, while suppressing them would certainly deepen the legitimacy crisis a president faces (Hochstetler and Samuels, 2011). Despite the rich perspectives it provides, the sociological school fails to explain presidential instability even in the absence of massive anti-government demonstrations (Perez-Linan, 2014).

As recent developments have proved, presidential instability often takes place when the president lacks a legislative shield that protects him from hostile measures by the parliamentary opposition, and in the midst of massive social mobilization that attempts to topple him. Accordingly, street politics and parliamentary behavior should not be treated as competing explanations but shall be considered complementary to each other. A useful theory of presidential instability, therefore, must study the mechanisms of interaction between social protest and the behavior of parliamentarians, to figure out the conditions that make the toppling of certain presidents more likely. Social protest is likely to pose a real threat to the survival of a president when it succeeds in forming a multiclass alliance against him and manages to find allies from both the political elite and the media. On the other hand, parliamentarians succeed in protecting the president when the presidential party/ruling coalition enjoys internal cohesion and is large enough to abort or withstand calls for his impeachment (Perez-Liñán, 2014).

Presidential approval rating, as influenced by the state of the economy and the president's commitment to clean politics, is key to understanding some actors' (un)willingness to seek deposing a president. Deteriorating economic conditions and/or leaks about his personal involvement in allegations of corruption may act as the grievances that help trigger social mobilization against him. Furthermore, when conventional channels of interest aggregation fail to represent the demands of important sectors of the population adequately, political protest might arise as a compensatory mechanism, as social movements tend to fill the void left by political parties and trade unions. Countries characterized by poor trust in political parties, due to their corrupt practices or elitist nature, and that have only a limited capacity, at best, to suppress, are likely to witness rising levels of protest by sectors which feel excluded and marginalized (Pugh, 2008). Starting from the 1980s, Latin America witnessed a rise in the mobilization capacity of civil society as nationwide massive demonstrations have swept many countries for several weeks or months and managed to impose major changes in public policy or even in the composition of the political elite. The fabric and tactics of these social movements varied from one case to another, but they were all united in defying the state power and the prevailing policies deemed to further marginalize and impoverish them (Frajman, 2006).

Politicians everywhere are driven by political ambition. Their first and basic objective is to get re-elected (Morgenstern, 2002). They are also rational actors who seek to maximize their interests, and their decisions are made in conformity with calculations about expected losses and gains (Hinojosa and Perez-Liñán, 2006). Since the impeachment process is usually carried out by the lower chamber of the parliament, the decision to declare whether the president is guilty or innocent is a political decision par excellence (Broderick, 1974; Havens and McNeil, 1978). This decision is influenced by several factors, such as the president's approval rating, the ruling coalition's internal cohesion, and the strategy the president employs in dealing with parliamentarians' needs and requests (Pérez-Liñán, 2014). In Columbia, for example, President Samper was able to withstand calls to impeach him and he managed to complete his term despite alleged ties to organized crime. The key to presidential survival in this case is his high favorability rates among the electorate and the role of the Liberal majority in the parliament that blocked any impeachment vote and encouraged splits within the ranks of Conservative opposition regarding the charges (Hinojosa and Pérez-Liñán, 2006).

In addition, the more fragmented the party system, the more difficult the president's mission to maintain the ruling coalition cohesion, as he/she has to deal with an increasing number of small parties. The distribution of policy concessions or patronage may seem unfair to some parties, encouraging them to leave the president's side. The president's negotiation skills and ability to impose discipline over the coalition members are important factors in maintaining the ruling party cohesion (Silva, 2016; Altman, 2000).

Parties also matter for the survival of presidents. In countries with moderately low institutionalized parties, such as Ecuador, it is not uncommon for outsiders who lack negotiation skills and tend to pursue confrontation with other constitutional bodies to reach the office of the presidency. Furthermore, political interactions are unpredictable and short-term goals of parties make cooperation less common. On the other hand, strong parties are able to incentivize cooperation among different factions and coalitions as they possess long-term goals and serve as forums to reveal preferences and broker agreements. Inter-branch crises are less likely as ruling parties with strong structures tend to curb the authority of the president. For example, in Chile, a country with a tradition of high institutionalization of political parties, President Sebastian Pinera was able to complete his term despite being the target of an unprecedented wave of massive demonstrations due to the efforts of ruling and opposition parties to protect him (Martínez, 2017). The legacy of democratic rule serves as another contextual factor that impacts the likelihood of the incidence of inter-branch crisis. As democracy deepens in a country, political actors become more experienced and tend to avoid behavior that might trigger presidential instability, such as betraying electoral promises, the formation of fragile alliances that quickly disintegrate and committing stark abuses of power (Martínez, 2020).

To sum up, some political forces, whether they are opposition political parties in the parliament and/or certain social movements, pose a challenge to the president when they threaten his survival in office. This threat usually takes the form of calls for his impeachment and/or massive demonstrations that paralyze the country and signal the legitimacy crisis the president faces. The impact of these political forces is not direct. Rather, its impact on the fate of the president is influenced by its relative strength compared to that of the countervailing “shield” created by the president's supporters (Pérez-Liñán, 2014).

3. Austerity economic measures and the political price of neoliberal policies

Presidential instability in Latin America partially mirrors the widespread anger among civil society groups owing to the retreat of the state from intervening in the economy. Since the 1990s, several Latin American countries have witnessed the ascendancy of neoliberalism and the free market economy model. Privatization, currency devaluation and reduction of government subsidies became the main ingredients in the economic policy there. These policies were adopted in response to the debt crisis that hit Latin America in the 1980s and the conditionality imposed by the International Monetary Fund (IMF) and other international financial institutions (Gwynne and Kay, 2000).

Neoliberalism as an economic ideology had the actual effect of increasing the plight of large sectors of the population, as it led to rising levels of poverty and unemployment (Petras, 1997). The repercussions of economic liberalization led to massive waves of social protest in several countries, such as Paraguay (1994), Venezuela (1999), Peru (2002) and Bolivia (2003). In all these cases, the free market economy model aggravated deprivation of basic needs among the poor and the middle class, whereas low levels of trust in political parties, in particular, convinced furious citizens to resort to unconventional means of participation (Bejar and Mores, 2016). Since the chief executive in Latin America embodies national authority and state power, he is considered the main actor responsible for the population's welfare. In many cases, the president became the direct target of criticism as high rates of unemployment and inflation were judged as the outcomes of his mismanagement of the economy (Valenzuela, 2004). On the other hand, the high institutionalization of the party system in Uruguay, Chile and Costa Rica encouraged angry citizens to resort to elections and political parties as a means to express their interests and demands, leading to a decrease in protest activities (Bejar and Mores, 2016).

In Brazil, Color de Mello (1990–1992) assumed the presidency amid rising expectations of economic recovery due to the return of the democratic rule. He promised to have a “new Brazil” capable of achieving economic modernization and social justice, and his main constituency was mainly the poor and the less-educated (Weyland, 1993). However, he could not achieve any of his promises, as he failed to fight inflation and unemployment or create a stable economy. Measures such as privatizing state-owned companies, raising the prices of oil and public services, and firing public sector employees proved detrimental to his popularity. In March 1990, a poll found out that 63% of Brazilians thought the president is doing a good job, but in October 1991 this percentage declined to only 10%. As De Mello's popularity deteriorated, the Congress became more oppositional to his policies and the media more bold in exposing allegations of government corruption (Perez-Liñán, 2007).

In contrast to the background of Color de Mello, Dilma Rousseff (2010–2016) was not a political outsider but an experienced political cadre. Although Brazil achieved one of the highest levels of economic growth under her predecessor Lula da Silva, the national economy began to suffer since 2012, as real incomes declined and unemployment rates soared. Mainstream media, owned by right-wing political actors, seized the opportunity and began to criticize Rousseff's lack of competence. Brazil's lavish spending on the preparations for hosting World Cup 2014 and the massive demonstrations that erupted in May 2013 by the “Let Free Fare Movement” against the increase of public transportation fares compounded the impasse of Rousseff. They led to further deterioration in her popularity, as they highlighted many aspects of Rousseff's alleged mismanagement of Brazil's national priorities (Onabsi, 2019).

While the fate of the two Brazilian presidents was eventually sealed by impeachment, Argentinean president Fernando de la Rua (1999–2001) resigned amid a massive wave of looting and violence, with the poor and unemployed banging empty pots and chanting “Let's throw everyone out”. These demonstrations were unprecedented in scale and signaled a massive decline in the president's legitimacy and popularity. In fact, Argentina has embarked on a bold platform of economic liberalization since 1990, but the repercussions of neoliberalism were felt starting from the mid-1990s as large sectors of the population suffered unemployment, poverty and stagnation (Schamis, 2002). These desperate conditions were coupled with the retreat of the state from providing social protection to large sectors of the population. In fact, unemployment rates rose from 6% in the 1980s to 17% in the 1990s; and the percentage of Argentinians who lived below the poverty line reached 36% in 2000. Large sectors suffered the harsh consequences of neoliberalism as their real incomes declined, their safety net vanished and their jobs were lost (Villalon, 2007).

Starting from the mid-1990s, the increase in the size of the informal sector and the unemployed was associated with high levels of political protest. The Piqueteros movement in Argentina, whose main activity was the blockade of main roads, constituted the core of this protest movement (Reyes, 2015). The crisis of legitimacy that conventional institutions faced (i.e. political parties and trade unions), owing to their elitist practices, their neglect of victims of economic liberalization, and their corrupt practices led many societal actors to seek new and unconventional means of political participation. In addition, the rich heritage of the Argentinian civil society during the military regime provided these actors with experienced cadres who have a history of political struggle (Villalon, 2007).

It can be argued that the crisis of December 2001 finds its origins in the economic policies and practices of Carlos Menem's administration. However, during the Peronist administration (1989–1999), the ruling party benefited from the success it achieved in fighting inflation and attracting foreign investment, and from its close ties with organized labor and small parties in the parliament. President De la Rúa, on the other hand, assumed rule while Argentina was on the brink of economic collapse. The Radical Party failed to fight stagnation or provide a remedy for the social costs of neoliberalism, while the government adopted austerity measures owing to the demands of the IMF and the liquidity crisis (Armony and Armony, 2005).

With the persistence of deteriorating economic conditions, public frustration reached high levels in October 2001. In the mid-term legislative elections, the percentage of those who cast null votes reached 22% of the electorate. Meanwhile, the Radical Party lost almost half of the votes it obtained in the 1999 elections. This loss was a manifestation of a severe legitimacy crisis in Argentina, as a large percentage declared their opposition to prevailing political practices and signaled their loss of trust in the whole political class. Amid a massive wave of capital flight, the Minister for Economy decided to impose harsh restrictions on Argentinians’ freedom to withdraw money from the banking system. This measure led to depriving millions of citizens from their savings. On December 18 and 19, 2001, Argentina has witnessed massive waves of protest and looting (Levitsky and Murillo, 2003). The demonstrators' clash with police forces resulted in the killing of twenty citizens. As the Peronist party declared that it would prepare for the president's impeachment and the army refused to quell the uprising, Fernando de la Rua found himself obliged to present resignation on December 20 and fled the presidential palace (Levitsky and Murillo, 2005).

The severity of economic conditions was felt in Ecuador as well. Ecuador has undergone three episodes of presidential instability: the ouster of Abdala Bucaram (1996–1997), Jamil Mahuad (1998–2000) and Lucio Gutierrez (2003–2005). The three presidents won the run-off elections with a very low margin and all of them adopted neoliberal economic policies. They adopted austerity measures that estranged opposition parties and popular organizations. In the case of Bucaram, major social movements organized a general strike that encouraged the parliament to declare Bucaram's “mental incapacity” to rule. In the case of Mahuad, a civil-military coup organized by “Confederación de Nacionalidades Indígenas del Ecuador (CONAIE)”, the largest organization representing indigenous people and a group of junior military officers ousted Mahuad after two years in office (Zamosc, 2013). Indigenous people constituted the core of the popular front against Bucaram and Mahuad. Their role in the two presidential crises mirrors their political strength. Indigenous people were excluded from political participation until 1978, when the literacy requirement was abolished as a prerequisite to voting (Jameson, 2011). In addition, they represent 40% of the population according to some estimates. Indigenous people participated in many episodes of popular mobilization in the 1990s, and they fiercely opposed neoliberal economic policies (Zamosc, 2007). This opposition to neoliberalism continued during Lucio Gutierrez's rule. Although the economy was recovering under Gutierrez, his orientation to the right and adoption of austerity measures, such as suspending subsidies offered on home gas, caused his popularity to decline and his governing coalition to disintegrate (Martínez, 2015).

4. The president's involvement in allegations of corruption

Presidential instability in the three countries was associated not only with unpopular economic policies but also with allegations of corruption and abuse of power that involved the head of state. In other words, the root causes of the wide mobilization against the president were not confined to opposing economic liberalization, but they included angry reactions to political scandals as well (Frajman, 2006). Corruption and abuse of power are not new phenomena in Latin America. However, the explosion of the number of political scandals reported by news media since 1978 is considered a recent development. This trend reflects the transition to democracy, which was associated with strong guarantees for freedom of expression and other civil liberties. Furthermore, rotation of power created a sense of uncertainty regarding who will remain in power, thus weakening the incentive of any news agency to develop long-term commitments with certain factions or political parties (Perez-Liñán, 2007). In this new environment, political scandals became a weapon used by political actors to tarnish the reputation of their opponents. However, the promising economic performance of some administrations rescued the president's popularity and legitimacy, even when he was perceived by public opinion to be implicated in actions of corruption and abuse of power, such as Carols Menem in Argentina and Alvaro Uribe in Columbia. In other words, public opinion tends to punish the president for corruption when the economy is weak (Carlin et al., 2015).

In Brazil, President De Mello's ouster from power was directly influenced by his involvement in a major political scandal, which involved receiving illegal contributions via Paulo Farias, his financial consultant during the presidential campaign. Leaks about the president's involvement in illegal financial transactions stirred cries of anger among civil society and the political class in general. Hundreds of thousands of Brazilians took to the streets in black calling for the impeachment of De Mello. In addition, the Congress initiated a series of investigations to examine the validity of the claims and found him guilty (Avritzer, 1995).

The negative consequences of the president's involvement in allegations of corruption were present in the case of Dilma Rousseff as well. In 2014 the Council for Control of Financial Activities in Brazil found out that there were huge illegal financial transactions between a number of private corporations and the state-owned company Petrobars. Rousseff served as the executive director of Petrobras from 2003 to 2010. According to allegations, some executive directors of the company were involved in receiving bribes and using Petrobras’s financial assets for electoral campaigns. Although official investigations failed to find any reliable evidence supporting the direct involvement of Rousseff in this corruption scandal, the latter proved damaging to her career as many figures in her party were accused of corruption.

While the impeachment of De Mello was widely celebrated as a victory for the rule of law and accountability in the Brazilian political system, Rousseff's impeachment, on the other hand, led to severe polarization within the Brazilian society. Rousseff was impeached in a context of declining popularity owing to the Petrobras scandal, World Cup demonstrations and deteriorating economic conditions. Despite the involvement of politicians from several political parties in the Petrobras scandal, right-wing news media concentrated solely on the alleged involvement of the labor party and it demonized Rousseff although she was not directly implicated in the scandal (Onbasi, 2019).

Concerning Argentina, political scandal added to the delicate status of the ruling coalition. In August 2000, rumors began to circulate about the involvement of the ruling party in offering bribes to some senators in exchange for their approval of some labor-related bills. The scandal took place amid the severe economic crisis. Despite the president's denial of the accusations, a Gallup poll found that 85% of the respondents believed that the accusations were true. Furthermore, while 30% of the respondents held favorable views about the government's performance, this percentage dropped to only 14% after the scandal. De la Rua’s handling of the allegations proved catastrophic, as he promoted two accused politicians to be his close consultants.

This scandal added to the tense relationship within the Alianza, the ruling coalition that included the Radical Party and Frepaso Party. Carols Alvarez, the leader of the Frepaso, saw this scandal as damaging to the government's credibility. He called for serious investigations of the corruption allegations but his demand was met with refusal. Consequently, he presented his resignation on October 6, 2000, fearing that the allegations will negatively impact his reputation. The consequences of the Senate scandal were catastrophic for de la Rua, as it further undermined his legitimacy and popularity, and led to the actual collapse of the ruling coalition (Martínez, 2015).

As for Ecuador, corruption is endemic, and it leads to low levels of trust in the political class in general. The three deposed Ecuadorian presidents were implicated in allegations of corruption and abuse of power. President Abdala Bucaram was widely accused of nepotism, as he appointed many of his family members and close friends in important government posts. He was also accused of embezzlement of financial donations raised originally for poor school pupils and also of stealing $3m. After his forced exit, a poll found out that a majority of Ecuadorians consider corruption the primary problem that faces their country, and that Bucaram is the most corrupt president since the democratic transition. His successor, Jamil Mahuad, was also accused of corruption when the head of an Ecuadorian bank declared that he paid $3m to finance Mahuad's electoral campaign, and that this sum of money was eventually diverted to Mahuad’s brother's account. The scandal created an impression that Mahuad's policies were the outcome of pressures from the economic elite (Conaghan, 2012). In the case of Lucio Gutierrez, he was accused of packing the Supreme Court with his supporters, a measure that further led to a decline in his popularity.

5. The absence of a legislative shield

Starting from the 1990s, Latin America saw a proliferation in the use of impeachment as a constitutional mechanism by the opposition to depose unpopular and uncompromising presidents. Judging that the call for the military to intervene is no longer a valid strategy in their struggle with an adamant administration, legislators resorted to impeachment or declaration of incapacity in several presidential crises. The president's ability to maintain majority support in the parliament proved crucial in withstanding calls for ousting him, whereas the collapse of the governing coalition was a prerequisite for deposing the president in the three cases.

In Brazil, president Color de Mello belonged to a junior party which was quickly formed to enable him to run for presidential elections. This party held only 3% of the seats in the Brazilian parliament. De Mello invested no time or energy to build a cohesive governing coalition. His leadership style was based on giving orders instead of negotiating with major parties. These parties received very little in terms of policy concessions or patronage (Pérez-Liñán, 2007). Unsurprisingly, when the president's involvement in a financial scandal was revealed by news media, and public opinion began to turn against him and his economic policies, it was too late to build a legislative shield to protect him.

Almost all organized civil society groups resisted the continuation of De Mello in power. These groups launched a nationwide campaign calling for his impeachment. This campaign was called Ethics in Politics. In addition, the parliament quickly referred the accusations against him to an investigation committee. Found guilty by the parliament, De Mello was removed from office by a majority of 76 against 5 in the House of Senate (Weyland, 1993). De Mello's demise seems to be the outcome of several factors, prominent among them is his neglect of the huge difference between the assets needed to win elections (such as political eloquence) and those necessary to continue in power (such as stable legislative majority support). Furthermore, the pattern that the president pursues early in his interactions with the legislature greatly impacts his ability to seek support from parliamentarians when he finds himself in a political crisis that may impact his survival in office (Valenca, 2002).

Dilma Rousseff suffered the negative consequences of the absence of a legislative shield as well. Rousseff faced a severe political impasse during her second term owing to the dismal economic performance and the scandal of Petrobras. Compounding the problem was the fact that Brazil has had the most fragmented parliament since it transitioned to democracy. The ruling coalition consisted of ten parties. Furthermore, Rousseff lacked the negotiating skills enjoyed by Lula da Silva, her predecessor, to keep the coalition members together. The major coalition partner, Partido do Movimento Democrático Brasileiro (PMDB), expressed its dismay over the distribution of portfolios and considered it unjust. With the election of Eduardo Cunha as the speaker of the House of Deputies in 2015, the relationship between the Labor Party and PMDB became sourer. Cunha played a decisive role in accepting the request to impeach Rousseff, a tantamount to his intention to undermine the influence and power of the ruling party and ending the coalition with it (Macaulay, 2017). In fact Cunha retaliated against Rousseff's decision not to drop allegations of corruption against him. It is evident that Rousseff’s declining approval ratings encouraged many legislators from the opposition and even some ordinary citizens to initiate resolutions to impeach her (Llanos and Pérez-Liñán, 2020). Rousseff was found guilty of manipulating the federal budget to conceal a growing deficit. Nearly 54% of those who voted against Rousseff came from the ranks of her governing coalition. Therefore, Rousseff's loss of majority support in the parliament was the decisive factor behind her removal (Macaulay, 2017). The major political crisis Rousseff faced revealed an inherent problem in the Brazilian political system. Proportional representation in Brazil leads to the proliferation of several small parties. It is almost impossible for one party to win the majority in the parliament. Therefore, the president finds himself obliged to form a governing coalition and to share his power with minor parties (Neto, 2002). This task proved difficult during Rousseff's second term owing to difficult economic conditions and the latent desire of PMDB and right-wing parties to end the hegemony of the labor party over Brazilian politics that lasted for 13 years (Mello and Spekter, 2016).

In Argentina, President de la Rua faced opposition even from within the ranks of his party. In 1997 the Radical Party formed a coalition with Frepaso, giving rise to what became called the Alinaza. The Alinaza won the presidential elections of 1999 after promises of just economic policy and clean politics. However, the Alianza disintegrated in October 2000 when the president, Fernando de la Rua, refused to carry out serious investigations about the scandal of the Senate's bribery. Also, the scandal discredited the government's claim to clean politics. In contrast to Carlos Menem, who managed to pass almost all of his economic reform bills owing to the majority status his party enjoyed in the parliament and the ad-hoc agreements he reached with small parties, the Radical Party suffered defeat in parliamentary elections in October 2001 and lost half the number of its seats compared to the 1999 elections. The more serious issue was that 22% of the electorate cast null votes, a shred of evidence that an increasing number of Argentinean citizens opposed the whole political class and signaled their exit from the electoral game. Furthermore, the governing coalition suffered internal splits over the austerity measures applied by the government, and it lacked any genuine linkages to organized labor or the business sector (Levitsky and Murillo, 2005).

As for Ecuador, and until 1996, five elected presidents were able to complete their terms. Despite their minority status in the parliament, those presidents could avoid a clash with the legislature owing to clandestine coalitions with the opposition. Major parties in the parliament used to lend their support to the president according to secret deals in exchange for major policy concessions or patronage. Starting from 1996, however, the president was no longer able to allocate financial expenditures to specific provinces, depriving him of a major asset in building and maintaining the governing coalition (Acosta and Polga-Hecimovic, 2010).

Bucaram's status as a minority president forced him to form a coalition with the largest opposition party in the parliament, the Christian Socialist Party. This coalition quickly fell apart due to disagreements over economic policy. The erosion of the governing coalition was followed by the formation of a national front by major opposition parties to oppose Bucaram’s austerity measures. This front led the efforts to depose the president, accusing him that he is mentally incapable to assume the responsibilities to rule. The front used the nationwide anger against Bucaram to legitimate an otherwise dubious measure to remove him from office. In the wake of the financial collapse that hit Ecuador, president Mahuad, Bucaram's successor, sought to form a governing coalition, first with leftist parties and then with the Rolodista Party. These efforts failed due to the lack of political assets needed to attract new coalition partners and his low approval rating among public opinion. In the absence of any legislative shield to protect him, Mahuad was vulnerable to public outrage that culminated in his ouster by a civil-military coup on January 21, 2000 (Acosta and Polga-Hecimovich, 2011). Similarly, President Gutierrez was ousted by the parliament in 2005 amid massive social mobilization. The Ecuadorian Congress declared his abandonment of office after he lost the support of the Pachakutik Party, due to his betrayal of electoral promises and his neoliberal policies (Martínez, 2015).

6. Military behavior during presidential crises in Latin America

Historically, military coups represented one of the major manifestations of political instability in Latin America. Moreover, till the third wave of democratization, the army used to enjoy extraordinary prerogatives over whoever holds official power, and it enjoyed total impunity from prosecution for human rights abuses. However, military disengagement from politics took place in several Latin American countries since 1978, reflecting the failure of military regimes in achieving economic development, and the massive social mobilization against stark abuses of human rights committed by military leaders (Messas, 1992). The Ecuadorian crisis that took place in 2000 testifies to the hostile environment against military coups in Latin America. On January 21, 2000, a civil-military coup took place to topple President Jamil Mahuad. Although the coup managed to topple the president, it failed to establish a new political order as senior military leaders opposed the military's return to direct rule. Several factors contributed to aborting the coup attempt. A poll conducted shortly after the coup indicated that 79% of Ecuadorians support maintaining the constitutional order and restoring democratic rule. Besides, senior military leaders considered the move of junior officers as equal to breaking the hierarchy within the army. Finally, the USA and the Organization of American States condemned the coup attempt and threatened that they would impose economic sanctions on Ecuador if the democratic rule was not restored (Barracca, 2007).

Generally, civil-military relations in Latin America are currently characterized by the military's obedience to civilian government and a huge decrease in the army's political influence. In the majority of presidential crises that took place after the third wave of democratization, the army preferred to remain quartered and neutral, refraining from quelling public demonstrations or declaring its support to one competing faction over another. This new pattern of military behavior has contributed to the persistence of democracy in Latin America and signaled the new political context that prevailed in the continent after the third wave of democratization (Pion-Berlin, 2008).

To explain the new pattern of military behavior, it is important to note that the military is a rational actor that weighs the expected gains and losses from pursuing a certain course of action. One of the important objectives of the military institution is to protect its members from legal prosecution for human rights abuses. Another objective is to maintain the cohesion of the institution and prevent any probable split within its ranks. Moreover, the army has a distinct identity and self-image about its proper role, which is determined, partly, by the constitution and other relevant legal provisions (Pion-Berlin et al., 2014). In most presidential crises in Latin America, the army reasoned that suppressing popular uprisings would lead to several political and legal predicaments. It would have meant that its members would face the danger of legal prosecution. In countries that managed to litigate members of the military regime for abuses of human rights, the desire to avoid this unpleasant scenario becomes a top priority. Besides, the law in several countries denies the army any significant role in keeping internal security. Furthermore, during a presidential crisis, the influence and authority of the president become weaker due to his deteriorating approval rating and his inability to secure a legislative majority, while the nature of the democratic system means that the president will not remain in power forever. Accordingly, the president's ability to retaliate against the army's refusal to carry out his orders and quell public demonstrations is very limited (Pion-Berlin and Trinkunas, 2010).

These factors were present in the Argentinean army’s decision not to intervene in the presidential crisis that ousted President Fernando de La Rua. In December 2001, the clashes between the demonstrators and security forces resulted in the killing of more than twenty citizens. Failing to contain the demonstrators' rising anger, president de la Rua declared a state of siege for thirty days and met the army leaders, requesting their help in quelling the uprising. The military refused to carry out the presidential order, precipitating the demise of de la Rua. The army has concluded that siding with the president would pose a real threat to its cohesion and corporate interests. Argentina was one of the first countries in Latin America which managed to prosecute leaders of the military regime for human rights abuses. As self-restraint would be almost impossible in containing unarmed demonstrators, disobeying the presidential order seemed the safer option. Moreover, Argentinean law assigns the mission of keeping internal security exclusively to police forces. The clear distinction the law establishes between internal and external missions further legitimized the army's decision not to intervene in the presidential crisis (Pion-Berlin and Trinkunas, 2010).

7. Conclusion

Episodes of presidential instability have challenged long-held assumptions about the dynamics of political systems in Latin America. They proved that the destiny of democracy in the continent has nothing to do with the fate of certain elected presidents. While the simmering conflict between the president and the legislature proved detrimental to democracy before the third wave of democratization, in none of the three countries has the removal of an elected president led to a return to authoritarian rule. The persistence of democracy during presidential crises in Latin America can be attributed to several factors; chief among them is political learning by civil and military elites that resorting to the military option during an existing presidential crisis is no longer a proper strategy, as it might lead to dangers of international isolation and political and economic catastrophes. Political elites became more attentive to the new rules of the game as they seek solutions to their conflicts within the existing constitutional order rather than outside its boundaries (Perez-Liñán, 2001). Moreover, presidential instability has proved that the president is no longer in a position of hegemony over other constitutional bodies. The decisive role of the parliament, public opinion and mass media in ousting undesirable presidents testifies to the decreasing influence of the chief executive and the deepening of political and legal accountability.

Throughout its modern history, Latin America has suffered from the recurrent intervention of the military in politics and the successful toppling of several presidents. However, the third wave of democratization ushered in a new pattern of political instability in the continent as the presidents in many countries could not complete their term due to the hostile behavior of opposition political parties or social movements. The decisive factor behind this change is the military's decision to withdraw from politics. Judging that its corporate interests are best served by remaining neutral during presidential crises, the outcome of these episodes of conflict between the president and the parliament is determined by the relative strength of each side.

Through the systematic comparison of Brazil, Argentina and Ecuador the study finds out that presidential instability is largely determined by the president's ability to maintain a legislative majority in the parliament and how he handles his relationship with public opinion. It is evident that inability to maintain a legislative shield is the outcome of several factors, such as the high fragmentation of the party system and the president's lack of negotiation skills. Legislative behavior in the three cases is influenced, partly, by the state of public opinion. Presidents are likely to complete their constitutional terms when they fulfill their electoral promises, refrain from enacting economic policies that marginalize large sectors of the citizenry, and maintain a cordial relationship with members of their party and members of the ruling coalition in general.

Presidential instability might be considered a corrective mechanism as it may represent an inevitable measure to overcome the conflict and gridlock between the chief executive and the legislature or to get rid of a president whose behavior contradicts prerequisites of political or legal responsibility (Marsteintredet, 2014). Meanwhile, in some instances it may be synonymous to a legislative coup, and public opinion may be easily manipulated by mainstream mass media. While presidential instability does not necessarily affect the procedural dimension of democracy, its effects on the quality of democracy, long-term social harmony and continuity of public policy are major issues that need further scrutiny (Hochstetler and Samuels, 2011).