In the past two decades, several democracies have slipped into democratic recession. Faced with economic or security crises, democratically elected executives in Latin America, the Middle East, Eastern Europe, and Africa have used their popularity to push for legislation – particularly constitutional amendments – that, over time, destroys systems of checks and balances, hinders free and fair elections, and erodes political rights and civil liberties. Across the world, these heads of government have found ways to subvert democratic norms while simultaneously maintaining a democratic façade. Using and abusing elections and institutional reform, they are turning new and old democracies alike into competitive authoritarian regimes.
Throughout the second half of the twentieth century, Venezuela was one of the most stable and prosperous democracies in Latin America. It had regular elections, strong institutions, and more wealth than any of its regional counterparts. Unlike other countries in the region, it did not succumb to right-wing dictatorships or protracted guerrilla warfare. Venezuela had forty years of democracy when Hugo Chávez was elected in 1998. Once in power, Chávez introduced constitutional amendments that strengthened his hold over congress, courts, and oversight agencies, and extended his time in office. He used that power to distort the electoral playing field to such an extent that it became almost impossible for the opposition to defeat him. In twenty years, Venezuela went from being one of the most promising democracies in the continent to one of its most egregious authoritarian regimes.
Not all countries, however, have suffered this fate. In Colombia, Alvaro Uribe tried to erode democracy but failed. In line with the Chavista playbook, between 2002 and 2010, Uribe introduced constitutional amendments to undermine the independence of the legislature and the courts, enhance the powers of the executive, and extend his time in office. Uribe was popular, populistic, polarizing, and willing to undermine democratic institutions to achieve his preferred policies. His government harassed journalists and members of the opposition and the courts, and worked in tandem with illegal armed actors to systematically intimidate those who criticized his administration. War-ridden, Colombia’s democracy had not been as stable as Venezuela’s, yet Uribe was not able to impair democratic institutions. Despite his attempts to undermine checks and balances and thwart the fairness of elections, Colombia’s constitutional order remained fairly strong. Uribe was unable to reelect himself for a third term and stepped down, giving way to a new democratically elected president.
Venezuela and Colombia are not the only countries where executives with hegemonic aspirations sought to entrench their rule. Between 1978 and 2019, Latin America has seen the rise of at least twenty-five leaders willing to undermine democratic institutions in order to fulfill their policy agenda. Fourteen of them – Carlos Menem (1989–99) in Argentina, Alberto Fujimori (1990–2000) in Peru, Jorge Serrano Elías (1991–3) in Guatemala, Evo Morales (2006–19) in Bolivia, León Febres (1984–8) and Rafael Correa (2007–17) in Ecuador, Arnoldo Alemán (1997–2002) and Daniel Ortega (2007–present) in Nicaragua, Roberto Suazo (1982–6), Juan Manuel Zelaya (2006–9), and Juan Orlando Hernández (2014–22) in Honduras, Nayib Bukele (2019–present) in El Salvador, Chávez and Uribe – tried to change or circumvent their nation’s constitution in order to remove checks on their presidency and hold on to power. However, only six of them – Fujimori, Chávez, Morales, Correa, Ortega, and Hernández – successfully transformed their countries’ democracies into competitive authoritarian regimes. The other seven did not.
Why do some potential autocratsFootnote 1 successfully erode democracy while others fail? This is the central question of this book, and one I will explore by comparing the dynamics of erosion in Colombia and Venezuela. Unlike traditional democratic breakdowns (e.g., military or civilian coups) the erosion of democracy happens over time. It takes years for an executive with hegemonic aspirations to succeed in eroding democracy. In the discussion that follows, I conceptualize this phenomenon in two stages. The first focuses on the factors that increase the likelihood of electing these hyper-ambitious leaders. The second focuses on the circumstances that, once in power, help or hinder these executives’ success in their attempts to erode democracy.
Weakly institutionalized party systems, weak states in crisis and – to a lesser extent – weak economic performance are critical to understanding where and when we are likely to see executives willing to undermine democratic institutions in order to advance policy goals. These factors, however, cannot fully explain why some of these leaders are successful in their attempts to erode democracy while others are not. In order to better understand this puzzle, Opposition at the Margins brings attention to the opposition, an often overlooked actor. Because the erosion of democracy happens sequentially, those who oppose executives with hegemonic aspirations have ample opportunities to respond. Their goals and strategies, I argue, are crucial to account for the success or failure of potential autocrats.
Some exceptions notwithstanding (Gamboa Reference Gamboa2017; Cleary and Öztürk Reference Cleary and Öztürk2020; McCoy and Somer Reference McCoy and Somer2021;), when analyzing the erosion of democracy, the spotlight often falls on the executive: the circumstances that brought her to power and the institutional context in which she started her term, her access to resources, her popularity, and/or her domestic or international support (see, e.g., Mazzuca Reference Mazzuca, Domínguez and Shifter2013a; Svolik Reference Svolik2015a; Haggard Reference Haggard2016a; Handlin Reference Handlin2017a; Corrales Reference Corrales2018a; Weyland and Madrid Reference Weyland, Madrid, Weyland and Madrid2019a; Weyland Reference Weyland2020). These accounts often obscure the role of those out of power. They either assume that once these would-be autocrats are in office, there is little that can be done against their hegemonic aspirations, or maintain that only weak potential autocrats fail, and do so because of their own shortcomings or mistakes. Without disregarding the balance of power between the government and the opposition during the processes of democratic erosion and how this context may shape the choices available to those fighting the potential autocrat, Opposition at the Margins complements existing accounts by emphasizing the importance of opposition agency. As I will show, using the right combination of goals and tactics, even weak oppositions have opportunities to defeat strong incumbents. The decisions they make are as important as the resources they possess.
1.1 The Argument in Brief
I define the erosion of democracy as a type of regime transition from democracy to autocracy that happens over time. Like classic democratic breakdowns (i.e., civilian or military coups), democratic erosion entails a regime change. In this type of democratic backsliding, incumbents introduce decrees, legislation, or constitutional amendments that enhance their powers and increase their time in office. Individually, these reforms strengthen some of the executive’s powers but fail to fully capture state institutions or provide budgetary powers large enough that would seriously unbalance the electoral playing field. Over time, however, these alterations accumulate to a point in which not only do they hinder horizontal accountability (O’Donnell Reference O’Donnell1994, Reference O’Donnell2007, 49–78) but also skew the electoral playing field by thwarting electoral accountability (O’Donnell Reference O’Donnell2007, 49–75). These alterations increase the executive’s hold over courts, congress, and oversight agencies and allow her to extend her time in office. This enables the head of state not only to run for two or more terms (in presidential systems) but, more importantly, to manipulate the electoral process to such an extent that it becomes extremely difficult for the opposition to defeat her. A democracy that has undergone erosion, therefore, is no longer a democracy, but a competitive authoritarian regime (Levitsky and Way Reference Levitsky and Way2010, 5–16).
Contrary to classic breakdowns, the erosion of democracy happens gradually. In civilian or military coups, authoritarian leaders, once in office, quickly dismiss elections, ban opposition parties, and/or close state institutions. Executives with hegemonic aspirations however, want to keep a democratic façade (Schedler Reference Schedler2013). Although they could close congress and the courts, in the post–Cold War environment these tactics are likely to trigger an adverse response from the international community and/or jeopardize domestic support (Lührmann and Lindberg Reference Levitsky and Way2018). Accordingly, potential autocrats instead chose to introduce constitutional amendments or legislation that expand their hold over congress, courts, and oversight agencies slowly. It takes years before they can successfully skew the electoral playing field to such an extent that it becomes impossible to defeat them.
In other words, the rise of an executive with hegemonic aspirations and the breakdown of democracy do not happen simultaneously. Therefore, I study these events independently. I conceptualize the erosion of democracy in two stages. The first focuses on the likelihood that a country elects a potential autocrat. Following existing literature, I argue that countries that have weak states with governance problems (Mainwaring Reference Mainwaring2012a; Diamond Reference Diamond2015a; Handlin Reference Handlin2017a), weakly institutionalized party systems (Carreras Reference Carreras2012a; Mainwaring Reference Mainwaring and Mainwaring2018a), and poor economic performance (Svolik Reference Svolik2015a; Haggard Reference Haggard2016a) are more likely to see the rise of these leaders. These characteristics cannot entirely explain why some of these executives successfully erode democracy while others fail, however. To understand this puzzle, the outcome of the second stage, I focus on the role of the opposition. Because democratic erosion happens over time, even after a leader with hegemonic aspirations assumes office, the opposition has institutional and noninstitutional resources it can use against the incumbent. How it uses these resources, and what it employs them for, is critical to understand why some executives can more easily erode democracy than others.
The degree of toleration for a government’s actions domestically and abroad is contingent upon the nature of the opposition’s challenge (Gartner and Regan Reference Gartner and Regan1996). Because executives with hegemonic aspirations come to power in democracy, opposition strategies and goals deemed unacceptable in democratic politics are risky gambles. While these strategies could potentially stop autocrats in the short term, they increase the government’s incentives to repress (i.e., violate freedoms of speech, assembly, or association, legally or verbally harass opponents, or violate integrity rights [Davenport Reference Davenport2007]) and decrease the costs of doing so. Such opposition tactics and goals enable the incumbent to paint the opposition as “radical” or “undemocratic,” endow her with “legitimate” reasons to remove opposition leaders from office, prosecute or jail them, and enhance the incumbent’s ability to rally citizens around the flag and push for more aggressive antidemocratic reforms.
Extrainstitutional strategies with radical goals – that is, tactics that use noninstitutional repertories like coups, protests, boycotts, or strikes to remove the president before the end of her constitutional term – fit these criteria. They convey a rejection of the established institutional mechanisms for redress and create a zero-sum game (McAdam Reference McAdam1999, 57–8). This combination of strategies and goals presents an existential threat to the incumbent – boosting the appeal of repression – and will likely be seen as unwarranted domestically and abroad – curbing the costs to repress. If successful, this combination of strategies and goals can halt the process of democratic erosion, but at the cost of breaking democracy altogether, further polarizing society, or martyrizing the potential autocrat. However, if failed, this combination of strategies and goals will likely jeopardize the opposition’s legitimacy inside and outside the country, providing the executive with more leeway to remove opposition leaders from office, and prosecute or jail them, as well as galvanize enough support to push for more aggressive antidemocratic reforms that, weakened, the opposition will be ill equipped to stop.
Strategies and goals that are considered part of “normal” democratic politics, on the other hand, are a safer gamble. While unlikely to bring the process of democratic erosion to an immediate halt, they increase the costs and reduce the incentives to repress. They protect the opposition’s legitimacy, hindering the executive’s ability to credibly label it as “undemocratic” or “radical,” reducing the incumbent’s capacity to harass or repress the opposition while keeping a democratic façade and taming her ability to rally around the flag in support for more aggressive antidemocratic reforms.
Institutional strategies with moderate goals – that is, tactics that use institutions like elections, congress, or courts to stop or modify antidemocratic reforms – fall in this category. This combination of strategies and goals convey an implicit acceptance of the proper channels of conflict resolution in a democracy and are less threatening for the ruling elite (McAdam Reference McAdam1999, 57–8), thus increasing the costs and decreasing the incentives to repress. Using electioneering, litigation, legislative tactics, or lobbying to modify specific reforms not only represents a low threat to the executive but also fails to provide her with “legitimate” reasons to attack the opposition or support a push for more aggressive reforms. These tactics are overall less costly to the opposition. If successful, they allow the opposition to buy time, keep some presence in the legislature, and be better equipped to repeal more radical antidemocratic reforms further down the road. If failed, the erosion of democracy may continue but the opposition lives to fight another day.
Following this logic, moderate extrainstitutional strategies – that is, strategies that use noninstitutional repertories to stop or modify antidemocratic reforms – and radical institutional strategies – that is, tactics that use institutions to remove the president before the end of her constitutional term – fall somewhere in between. Extrainstitutional strategies with moderate goals decrease the incentives for repression but also the costs of doing so. Although potentially effective to exert pressure on the incumbent and call attention to her abuses, these tactics can easily turn illegal or violent, in which case they could provide the head of government with “legitimate” reasons to crack down on the opposition. Institutional strategies with radical goals increase the incentives for repression but also the costs of doing so. Although they can successfully remove an executive from office while protecting the opposition’s legitimacy, they can also threaten the incumbent’s existence and thus – if failed – risk a more aggressive response.
Accordingly, oppositions that resort to radical extrainstitutional strategies and, depending on the circumstances, radical institutional or moderate extrainstitutional strategies, run a high risk of curtailing their ability to prevent the erosion of democracy. Likely to fail (Baykan, Gürsoy, and Ostiguy Reference Baykan, Gürsoy and Ostiguy2021), these tactics can undermine their legitimacy and resources while also increasing the incumbent’s incentives and support to advance more aggressive antidemocratic reforms that a weakened opposition will be less suited to stop. On the contrary, oppositions that resort to moderate institutional strategies alongside some types of moderate extrainstitutional or radical institutional strategies have a good shot of enhancing their ability to prevent democratic erosion. Although moderate in their achievements, these tactics can safeguard the opposition’s legitimacy and resources, protect courts and oversight agencies, and hinder aggressive antidemocratic reforms.Footnote 2
The argument outlined earlier draws from approaches that highlight agency and international factors in processes of regime change. Some scholars suggest that transitions from and to democracy are elite-driven (Linz Reference Linz1978; O’Donnell and Schmitter Reference O’Donnell and Schmitter1986; Berman Reference Berman1998; Capoccia Reference Capoccia2007a; Mainwaring and Pérez-Liñán Reference Mainwaring and Pérez-Liñán2013). They argue that democratic breakdowns are the outcome of elites’ strategic choices in response to crises. Like these theories, I emphasize the importance of actors and their decisions. Unlike these theories – and in line with more recent work on democratic backsliding and polarization (Gamboa Reference Gamboa2017; Levitsky and Ziblatt Reference Levitsky and Ziblatt2018a; McCoy and Somer Reference McCoy and Somer2019, Reference McCoy and Somer2021; Cleary and Öztürk Reference Cleary and Öztürk2020) – I pay attention to elites’ choices even after an executive with hegemonic aspirations has attained power.
Elite decisions, of course, do not happen in a vacuum. International factors are also essential to understand transitions from and to democracy (Brinks and Coppedge Reference Brinks and Coppedge2006; Levitsky and Way Reference Levitsky and Way2010; Bunce and Wolchik Reference Bunce and Wolchik2011; Mainwaring and Pérez-Liñán Reference Mainwaring and Pérez-Liñán2013; Schenoni and Mainwaring Reference Schenoni and Mainwaring2019). International factors shape the balance of power between regime coalitions and constrain what is deemed acceptable or unacceptable in pursuit of political change. My theory highlights the importance of international and domestic audiences in shaping the executive’s and the opposition’s strategic choices. Contingent on the assumption that these audiences have a normative preference for democracy that motivates authoritarian leaders to keep a democratic façade, I argue that some goals and strategies are more useful than others at preventing the erosion of democracy.
1.2 The Importance of Understanding the Erosion of Democracy
My argument has three relevant implications for the study of democratic politics. First, this book offers an updated and broader understanding of democratic breakdowns. Following the trend of scholars who have theorized democratic backsliding in recent years (Bermeo Reference Bermeo2016a; Dresden and Howard Reference Dresden and Howard2016a; Cameron Reference Cameron2018; Waldner and Lust Reference Waldner and Lust2018a), I distinguish democratic erosion from declines in the quality of democracy and civilian/military coups. I analyze it as a type of democratic breakdown that happens sequentially, as a process rather than a one-shot game.
This approach distinguishes the rise of an authoritarian leader from her ability to undermine democracy. Conventional theories to the study of democratic breakdowns have mostly focused on the factors that bring to power leaders with a normative preference for dictatorship or a very weak preference for democracy. With classic democratic breakdowns in mind, they assume that once these leaders are in office, there is little that can be done to prevent a democratic reversal and, as a result, they fail to consider what happens afterward.
Common approaches to regime change fall in this group. Low economic development (Przeworski et al. Reference Przeworski, Alvarez, Cheibub and Limongi2000; Svolik Reference Svolik2008), economic performance (Svolik Reference Svolik2015a; Haggard Reference Haggard2016a), governance problems (Fortin Reference Fortin2011; Mainwaring Reference Mainwaring2012a; Andersen et al. Reference Andersen, Møller, Rørbæk and Skaaning2014; Diamond Reference Diamond2015a; Handlin Reference Handlin2017a), and weak institutions (Carreras Reference Carreras2012a; Pérez-Liñán and Mainwaring Reference Pérez-Liñán and Mainwaring2013; Mainwaring Reference Mainwaring and Mainwaring2018a; Ginsburg and Huq Reference Ginsburg and Huq2019a; Weyland and Madrid Reference Weyland, Madrid, Weyland and Madrid2019a; Weyland Reference Weyland2020), they claim, increase the likelihood of democratic breakdown. Economic and security difficulties lead to legitimacy crises that undermine popular support for democracy and unsettle democratic institutions. Feeble institutions, in turn, make electoral politics unpredictable, increasing the probability that leaders willing to circumvent democracy attain power.
As Chapter 3 will show, these theories are good at explaining why executives with hegemonic aspirations are elected in the first place, but they are less satisfactory when it comes to explaining why some of these leaders successfully erode democracy while others fail. Economic recessions and weak states with governance problems can shake an executive’s hold on power. Once this incumbent has been replaced by the potential autocrat, these variables should decrease her support and her ability to erode democracy, not the other way around. In other words, they should not necessarily overdetermine the concentration of power in the hands of the new president. By conceptualizing the erosion of democracy as a process, not only do I move away from accounts that see regime backsliding as inevitable once a hyper-ambitious leader comes to power, but I also specify the effect that structural, institutional, and state-centered variables have on democratic erosion.
Second, analyzing the erosion of democracy as a process allows me to highlight the importance of the opposition’s tactics and goals after an executive with hegemonic ambitions has attained power. Few scholars have analyzed the role of the opposition in processes of democratic backsliding (for exceptions, see Gamboa Reference Gamboa2017; Abi-Hassan Reference Abi-Hassan, Hawkins, Carlin, Littvay and Rovira Kaltwasser2019a; Cleary and Öztürk Reference Cleary and Öztürk2020; McCoy and Somer Reference McCoy and Somer2021). Recent studies of democratic erosion have mostly emphasized the role of the executive and/or the structural or institutional landscape these leaders face. These scholars argue that the erosion of democracy is the outcome of the incumbent’s ex ante popularity (Levitsky and Loxton Reference Levitsky and Loxton2013a; Corrales Reference Corrales2018a; Weyland and Madrid Reference Weyland, Madrid, Weyland and Madrid2019a), institutional strength (Batory Reference Batory2016a; Stoner-Weiss Reference Stoner-Weiss, Bunce, McFaul and Stoner-Weiss2010), strategic choices (Carlos de la Torre and Lemos Reference Torre and Lemos2016; Balderacchi Reference Balderacchi2017a; Handlin Reference Handlin2017a, Reference Handlin2018), resources (Hidalgo Reference Hidalgo2009; Hawkins Reference Hawkins2010a; Mazzuca Reference Mazzuca, Domínguez and Shifter2013a), ideology (Weyland Reference Weyland2013), domestic and/or international support (Carlos de la Torre Reference Torre2013; Corrales Reference Corrales2015), and/or the strength of the institutions she is trying to co-opt (Weyland Reference Weyland2020). Few of them look at the opposition, and when they do, they think of its choices as heavily constrained by feeble institutions and/or resource asymmetries.
The balance of power between the government and the opposition at the outset of erosion or the institutional setting it plays out in does not overdetermine the outcome. Variables such as economic growth, mineral resources, weak institutions, and popularity cannot fully distinguish between successful and failed cases of erosion. Indeed, Hugo Chávez had access to limitless oil revenues that helped him enhance his powers and extend his time in office, but neither Daniel Ortega nor Juan Orlando Hernández, who also succeeded in eroding democracy, had a similar advantage. The Colombian economy during Alvaro Uribe’s government did much better than Ecuador’s during Rafael Correa’s administration, yet the latter eroded democracy, while the former did not. Rafael Correa came to power in the midst of unstable institutions that helped him erode democracy; Chávez, however, did not. The average popularity of Correa was high, but so was Uribe’s. In fact, it was higher than Hernández’s, Chávez’s, or Evo Morales’s. Similarly, political polarization was higher in Colombia during Alvaro Uribe’s government than in Honduras during Juan Orlando Hernandez’s government. Yet the former failed to erode democracy while the latter did not.
This is true if we look more closely at the two cases analyzed in this book. It is often tempting to succumb to the fallacy of retrospective determinism, whereby we judge an institutional framework or an opposition relatively weak because of its ultimate demise or failure to prevent the erosion of democracy and relatively strong because of its ultimate resilience or success in protecting democratic governance. To avoid this problem, Chapters 4 and 5 provide detailed accounts of the balance of power between government and opposition in both of my cases and the institutional landscape in which they deployed their strategies. I show that at the beginning of his government, the opposition to Chávez was not only stronger than the opposition to Uribe but also capable of selecting among a wide set of institutional and extrainstitutional strategies.
Instead of focusing on the executive and/or the vulnerability of the institutions she takes over, I join scholars that argue we ought to recognize the agency of the opposition as well (Lindberg Reference Lindberg2009; Bunce and Wolchik Reference Bunce and Wolchik2011; Gamboa Reference Gamboa2017; McCoy and Somer Reference McCoy and Somer2019; Cleary and Öztürk Reference Cleary and Öztürk2020; Jiménez Reference Jiménez2021; Ong Reference Ong2021). The resources available to the head of government, as well as her ability to use them, I posit, can be curbed or enhanced by the opposition’s strategic choices. Whereas moderate institutional strategies alongside some types of moderate extrainstitutional or radical institutional strategies can help the opposition protect the resources it has (while frustrating the government’s attempts to co-opt them), radical extrainstitutional strategies alongside some types of moderate extrainstitutional or radical institutional strategies, can hinder the opposition’s ability to do the same. Consequently, the first set of strategies and goals can hamper the incumbent’s ability to increase her leverage vis-à-vis those who oppose her. The second set can enhance it.
This latter point is crucial. The third wave of democracy ended most military and single-party dictatorships. Between 1974 and 1999, eighty-five countries democratized (Geddes Reference Geddes1999), and classic coups d’état, executive coups and election-day frauds declined (Bermeo Reference Bermeo2016a). Unfortunately, however, this democratic awakening was not durable. Not only did some regimes fail to fully democratize (Brownlee Reference Brownlee2007; Levitsky and Way Reference Levitsky and Way2010; Schedler Reference Schedler2013a), but, even more concerning, new and old democracies began experiencing democratic setbacks. According to Freedom House’s 2021 report, since 2005, the number of countries with democratic declines has outweighed the number of countries with democratic gains every year (Repucci and Slipowitz Reference Repucci and Slipowitz2021).
The erosion of democracy has become an increasingly common type of democratic reversal. It has proven to be pervasive and hard to prevent. The cases of democratic erosion have increased since the 1990s. Initially circumscribed to nations of the Global South in Africa and Latin America, democratic backsliding has spread to economically more developed and/or “poster boy” democracies like Poland and Hungary. Today, even fully consolidated democracies, like the United States, have seen threats of backsliding via executive overreach (Levitsky and Ziblatt Reference Levitsky and Ziblatt2018a; Roberts Reference Roberts, Weyland and Madrid2019).
Few of the mechanisms established to protect and advance democracy across the world have succeeded in constraining executives with hegemonic ambitions. These incumbents have found ways to subvert democratic norms without breaching international democratic standards. The international community (standard-bearer of democratization during the 1980s and 1990s in the West) has few means of sanctioning them (Meyerrose Reference Meyerrose2020). By studying how these leaders succeed (or fail) in their attempts to erode democracy, I provide insights not only on what countries need to do to prevent potential autocrats from coming to power but, more importantly, on what the opposition and the international community can do to improve their chances of stopping the processes of democratic erosion.
1.3 Research Design
There are many ways in which scholarship can benefit from using a mixed methods approach (Seawright and Gerring Reference Seawright and Gerring2008; Tarrow Reference Tarrow, Brady and Collier2010; Goertz Reference Goertz2017). In this case, I use quantitative and qualitative analytic techniques in order to answer two different types of questions. The first – what variables increase the likelihood of electing executives with hegemonic aspirations? – focuses on the average effect that a variable, or set of variables, has on a given outcome. I wish to know under what conditions countries elect potential autocrats and to what extent these and other related variables influence the autocrat’s ability to erode democracy, not the causal path by which they come to power. Therefore, quantitative multivariate analysis is appropriate to answer this type of question.
In Chapter 3, I use descriptive and analytic quantitative techniques to assess the average effect of economic development and growth, perceptions of the economy, state capacity, state performance, trust in institutions, and party system institutionalization on the probability that a country will see the rise of a potential autocrat. I identify presidents with hegemonic aspirations using an updated version of Mainwaring and Pérez-Liñán’s (Reference Mainwaring and Pérez-Liñán2013) regime preferences database and an original dataset of constitutional amendments introduced by democratically elected Latin American presidents between 1978 and 2019. Unlike existing datasets on institutional reform (e.g., Comparative Constitutions Project, Latin American Constitutional Change Database), my dataset records attempts to amend the constitution regardless of whether they were successful or not. This allows me to distinguish between presidents who do not try to erode democracy, presidents who try to erode democracy and fail, and presidents who try to erode democracy and succeed. I measure the independent variables using existing databases such as the World Bank’s for economic development and growth, mineral exports, and state performance; Hanson and Sigman’s (Reference Hanson and Sigman2021) for state capacity; Latinobarómetro (1995–2017) for perceptions of the economy and trust in institutions; Mainwaring’s (Reference Mainwaring, Bizarro, Petrova and Mainwaring2018) for electoral volatility; Helmke’s (Reference Helmke2017) and Elkins and Ginsburg’s (Reference Elkins and Ginsburg2021) for institutional strength; the Executive Approval Project (2019) for president approval ratings; and Varieties of Democracy (2021) for political polarization.
In the first stage, I identify twenty-five cases of presidents with hegemonic aspirations in Latin America (for details, see Table 3.1). Two of them – Fujimori (1990–2000) and Jorge Serrano Elías (1991–3) – launched coups. Twelve introduced reforms or legislation that circumvented or changed the constitution in order to enhance their powers and extend their time in office. Five of them successfully eroded democracy, six failed, and one is too early in the process to determine success or failure.
The second question – why, once elected, some potential autocrats successfully erode democracy while others fail? – is concerned with a very specific event. I do not wish to understand the average effect of a variable or set of variables in a president’s ability to erode democracy, but rather trace the set of interrelated causes, the process by which she transforms, or fails to transform, a democracy into a competitive authoritarian regime. Accordingly, qualitative methods are more appropriate to answer this question (Goertz and Mahoney Reference Goertz and Mahoney2012). In Chapters 4 and 5, I use comparative historical analysis (Mahoney Reference Mahoney, Mahoney and Rueschemeyer2003; Mahoney and Thelen Reference Mahoney and Thelen2015) to assess the conditions that enable some presidents with hegemonic aspirations, but not others, to increase their powers and extend their time in office beyond a second term. The leverage of comparative historical analysis derives from within-case analysis (George and Bennett Reference George and Bennett2005; Bennett Reference Bennett, Brady and Collier2010; Bennett and Checkel Reference Bennett, Checkel, Bennett and Checkel2015). Consequently, I use process tracing to evaluate the causal mechanisms that connect the opposition’s strategic choices and goals with the executive’s ability to erode democracy in a case of successful erosion (Hugo Chávez in Venezuela) and a case of failed erosion (Alvaro Uribe in Colombia). These cases allow me to analyze the effects of opposition strategic choices and goals on the incumbent’s ability to undermine democracy at different points in time. I use detailed case knowledge to outline the government’s and the opposition’s resources, describe the strategic choices available, and connect the ones the opposition chose to its ability to protect (or fail to) and utilize (or not) these resources when facing more radical power grabs.
1.3.1 Case Selection
I chose to study Venezuela and Colombia using a qualitative set theory logic (Goertz and Mahoney Reference Goertz and Mahoney2012; Goertz Reference Goertz2017) and a most similar systems design (Gerring Reference Gerring, Box-Steffensmeier, Brady and Collier2008; Seawright and Gerring Reference Seawright and Gerring2008). The qualitative set theory logic for case selection prioritizes causal mechanisms (Goertz Reference Goertz2017). It dictates that we should select cases that allow us to assess both the causal mechanisms and the constraint causal mechanisms. That is, the mechanisms by which the outcome happens (Y=1) whenever that causal condition is present (X=1), and the mechanisms that prevent the outcome from happening (Y=0) whenever that condition is absent (X=0) (Goertz Reference Goertz2017). Table 1.1 shows how the eleven cases of presidents with hegemonic aspirations who tried to erode democracy distributed across the main independent and dependent variables.
Note: Opposition strategies for these cases (in italics) are harder to categorize.
Out of these eleven cases, eight align with my argument. The opposition used mostly radical extrainstitutional strategies against Hugo Chávez (Venezuela, 1999–2013), Evo Morales (Bolivia, 2006–19), and Rafael Correa (Ecuador, 2007–17) – though opposition strategies in Ecuador are harder to categorizeFootnote 3 – enabling these presidents to erode democracy; the opposition avoided radical extrainstitutional strategies and used mostly moderate institutional strategies against León Febres (Ecuador, 1984–8), Carlos Menem (Argentina, 1989–99), Roberto Suazo (1982–6), Arnoldo Alemán (Nicaragua, 1997–2002), and Alvaro Uribe (Colombia, 2002–10), hindering their ability to erode democracy. Daniel Ortega (Nicaragua, 2007–present), Juan Orlando Hernández (2014–22), and José Manuel Zelaya (2006–9) deviate from the theory. Their cases, however, do not invalidate the argument, but help refine it and establish scope conditions.
I select my cases from Cells B and C of Table 1.1. Following the selection logic outlined earlier, I chose: (a) a case from the cell in which neither extrainstitutional strategies nor democratic erosion are present (X=0 and Y=0) and (b) a case in which radical extrainstitutional strategies and democratic erosion are present (X=1 and Y=1). I then use a most-similar systems design to select the case of Alvaro Uribe and Hugo Chávez within these two cells. Venezuela and Colombia are similar in many ways. They democratized at the same time, had similar, very stable, two-party systems for most of the post-WWII era, and had traditional parties that declined roughly at the same time, in the 1990s. Both Hugo Chávez and Alvaro Uribe came to power in the midst of crises, and they were both very popular, populistic, polarizing, and willing to circumvent democracy. These similarities along theoretically relevant variables (i.e., those normally associated in previous work with the rise of potential autocrats and threats to democratic stability) make these cases comparable in the context of this book.
That being said, these cases are by no means identical; no two cases in comparative politics are ever an exact match. Crucially, Venezuela, unlike Colombia, owns very large oil reserves. Extraordinary oil revenues accrued during the commodity boom of the 2000s certainly helped Chávez erode democracy by allowing him to increase his military and civilian support. Uribe did not have those kinds of resources, but process tracing allows me to take this alternative explanation into account and show that Uribe’s comparative deprivation did not hinder his ability to secure the endorsements he needed to pursue a hegemonic project. The Colombian president was, and still is, beloved by the military, and his approval ratings were actually higher than Chávez’s throughout his government. Moreover, Chávez did not gain automatic control over these resources when he attained power in 1999. As I show in Chapter 4 and in line with my theory, an opposition misstep enabled him to gain full control over the petroleum company four years into his first term.
There are also other differences between the cases, but far from undermining comparability, such differences make the comparison interesting and productive. First, Colombia had a fifty-year-long armed conflict that Venezuela did not have. However, the Colombian armed conflict biases the case against my theory. In general, civil strife should decrease the likelihood that democracy survives. In the case of Colombia, the armed conflict provided a clear focal point for the president to seek support for his hegemonic project in a classic “rally around the flag” fashion. This might explain, for example, why Uribe was always more popular than Chávez. Second, the power of the Venezuelan and Colombian oppositions vis-à-vis the government at the onset of the process of erosion was different. The opposition to Hugo Chávez started off stronger than the opposition to Uribe. It had significant influence over the petroleum company, the armed forces, and the major media outlets, one-third of the seats in congress (post-2000), some support inside courts and oversight agencies, and the ability to mobilize millions of supporters to the streets. On the contrary, when Uribe came to power, the opposition had no access to any outstanding resource, no influence over the armed forces or major media outlets, one-third of the seats in congress, some support inside courts and oversight agencies, and the ability to mobilize some – not quite as many – supporters to the streets. Still, and despite these differences, Uribe was not able to erode democracy, and Chávez was.
Beyond the differences outlined earlier, the other two important issues to consider are whether Uribe and Chávez were equally determined to erode democracy and the strength of institutions they faced. It is possible that, from the outset, Chávez was simply more willing to undermine democratic institutions than Uribe was. Unfortunately, it is impossible to know for sure. Our perception of an executive’s willingness to undermine democratic institutions is imperfect and often tainted by the fallacy of retrospective determinism. Looking backwards in 2021 – as Venezuela’s regime crisis deepens – it is easier to perceive Chávez as more willing to undermine democratic institutions than Uribe (or other unsuccessful potential autocrats). It is also impossible to disentangle Chávez’s drive to undermine democracy from the opposition he faced. As I show throughout Opposition at the Margins, Chávez resorted to highly polarizing tactics that, unlike in Colombia, were met with similarly polarizing moves from the opposition. Had the opposition exercised more restraint, it is possible – even if not certain – that Chávez might not have resorted to more aggressive antidemocratic reforms. Finally, even if Chávez’s authoritarian tendencies were stronger than Uribe’s, he did not narrow all the choices available to the opposition when he got to power in 1999 or even in the years immediately after. Although increasingly constrained as the process of erosion moved forward, the Venezuelan opposition faced several viable paths up until 2006, many of which did not involve radicalization. In Chapter 4, I carefully reconstruct these alternatives. I show that it was ultimately the opposition’s choice of radical extrainstitutional strategies that provided Chávez the cover he needed to destroy Venezuela’s democracy without having to show his true authoritarian colors until 2007 or 2008.
Something similar happens with institutional strength. It is possible that, from the outset, democratic institutions in Venezuela were just weaker than democratic institutions in Colombia, making it easier for Chávez to co-opt them as well as curtailing the opposition’s ability to use them. In Chapters 4 and 5, I discuss these countries’ distinct forms of institutional weakness in detail. In general, however, it is not clear that institutions in Colombia were any more or less vulnerable than Venezuela’s. Between Venezuela’s transition to democracy and the election of Hugo Chávez, the country had replaced its constitution once and amended it twice. Between Colombia’s transition to democracy and Alvaro Uribe’s rise to power, this country had also replaced its constitution once, but amended it seventeen times. Colombia’s younger constitution might have been less prone to replacement but was by no means less prone to modification.
By the same token, there is no evidence that the institutional landscape at the onset of erosion inevitably skewed the Venezuelan opposition into extrainstitutional strategies. Which institutions to use and how to rely on them is certainly shaped by the institutional context. Oppositions in Colombia and Venezuela had access to a different set of institutional resources, but they both had powerful tools they could use. As I show in Chapter 4, up until 2005, the opposition in Venezuela was fairly strong in congress and – for a while – had some leverage inside the courts and oversight agencies. As I show in Chapter 5, throughout Uribe’s government, the opposition in Colombia was not particularly strong in congress but had more reliable courts to resort to. Both oppositions had the ability to use moderate institutional strategies.
1.3.2 What We Can Learn from the Cases That Do Not Fit the Theory
No theory is airtight. Suggesting a particular set of necessary or sufficient conditions does not eliminate the possibility that there might be falsifying cases: cases in which in the absence of the necessary condition, the outcome still happens, or in the presence of a sufficient condition, the outcome does not happen. Cells A and D in Table 1.1 fall in this category. Although the opposition relied mostly on moderate institutional strategies, and avoided extrainstitutional ones, Ortega and Hernández successfully eroded democracy. And in the case of Zelaya, the opposition resorted to radical extrainstitutional strategies that successfully put an end to the president’s hegemonic project (although breaking democracy altogether). These three cases help outline the scope conditions of the theory (Goertz Reference Goertz2017, 90–122), highlighting important assumptions underlying the argument. I will outline these conditions in more detail in Chapter 2. Later, however, I discuss the mechanisms that, absent radical extrainstitutional strategies, led to democratic erosion in Nicaragua and Honduras (2017), as well as the mechanisms that, despite the use of extrainstitutional strategies, prevented the erosion of democracy in Honduras (2009) (see Table 1.2 for more details).
Note: Highlighted are the cases for which that particular scope condition does not hold.
1.3.2.1 Time
One of the key factors underlying my theory is time. If democracy quickly breaks down, the opposition has few opportunities to fight back. The case of Daniel Ortega in Nicaragua underscores that point. Ortega came to power in Nicaragua thanks to an alliance between the Frente Sandinista de Liberación Nacional (FSLN) and the Partido Constitucionalista Liberal (PCL) led by Arnoldo Alemán (1997–2002).Footnote 4 Together, these parties co-opted courts and oversight agencies before Ortega ran for office in 2006 (Pérez-Baltodano Reference Pérez-Baltodano2010). Although Alemán did not erode democracy himself, he helped set the stage for Ortega to do so. Unlike the oppositions in Colombia, Venezuela, Ecuador, Bolivia, or Argentina, by the time the Sandinista leader came to power, the Nicaraguan opposition had few of the resources it needed to put up a fight. He controlled all branches of power from the first day of his government. Whereas in other cases, the oppositions had several resources and strategic choices available immediately after a potential autocrat came to power; in Nicaragua, the opposition found itself without much capacity and time to react.
1.3.2.2 Audiences with a Normative Preference for Democracy
The theory assumes that executives with hegemonic aspirations want to keep a democratic façade. They fear that overtly authoritarian moves will unleash domestic and/or international backlash, and therefore wait until the opposition gives them “legitimate” reasons to advance antidemocratic reforms. This only works, of course, if the potential autocrat faces domestic or international audiences with leverage and a normative preference for democracy. That was not the case of Juan Orlando Hernández in Honduras (2014–22). Even though the opposition used mostly institutional strategies with moderate goals to oppose him, the Honduran president was able to not only pack the court and change the constitution to extend his time in office (Landau, Dixon, and Roznai Reference Landau, Dixon and Roznai2019) but also reelect himself in highly questionable elections and enforce the outcome despite the criticism of domestic and international organizations (Malkin Reference Malkin2017; Rodríguez Reference Rodríguez2019). He was able to do so, in part, thanks to the quick endorsement of the US president Donald Trump (2017–21), who showed ample disregard for democratic processes and norms domestically and abroad (Carrillo Reference Carrillo2017; Levitsky and Ziblatt Reference Levitsky and Ziblatt2018a). Had Trump had a normative preference for democracy and reacted differently, Hernández might have been forced to repeat the elections following the demands of the Organization of American States (Levitsky Reference Levitsky2018).
1.3.2.3 Support for Executives with Hegemonic Aspirations
Built into the theory is the assumption that potential autocrats have military and/or civilian support. Without it, they are less likely to try to enhance their powers and extend their time in office, and if they do so, they are more likely to fail. That is the case of Manuel Zelaya (2006–9), whose term was cut short in 2009 when overnight the army escorted him out of the country. In 2008, the Honduran president tried to push for a referendum that would have allowed him to call for a constitutional assembly and rewrite the constitution in order to enhance his powers and extend his time in office (Ruhl Reference Ruhl2010; Weyland Reference Weyland2013). Defying Congress and the Supreme Court, which had decided against the project, Zelaya asked the army to go ahead with the ballot. In response, the opposition launched a coup. The move was not much different from the strategy used by the Venezuelan opposition against Hugo Chávez or Recep Tayyip Erdogan (Turkey, 2003–present) in 2002 and 2016, respectively. Unlike the Venezuelan or the Turkish incumbents who neutralized the coup attempts, the Honduran leader did not have strong military or civilian support to fight back. His approval rating at the time of the coup was 37 percent, and he did not have the organizational means to mobilize supporters to reverse the putsch. Democracy definitively broke down in Honduras in 2009, but it did not erode. The rupture in this Central American country represented a more traditional democratic breakdown.
To summarize, using an original database of institutional reforms introduced by democratically elected Latin American presidents, I am able to identify twenty-five presidents with hegemonic aspirations, eleven of whom introduced legislation or institutional reforms to undermine democratic institutions in a slow fashion. Five of them eroded democracy, six did not. Out of those eleven cases, I chose a case of successful erosion (Hugo Chávez) to evaluate the mechanisms that connect radical extrainstitutional tactics with an executive’s ability to erode democracy, and a case of failed erosion (Alvaro Uribe) to assess the constraining mechanisms that connect the absence of extrainstitutional radical strategies and the use of moderate institutional and moderate extrainstitutional strategies with an incumbent’s inability to erode democracy.
1.3.3 Qualitative Evidence
I collected the data to trace the dynamics of erosion in Colombia and Venezuela during twelve months of intensive fieldwork (2013–14). In the time I spent in both countries, I conducted eighty-eight interviews – fifty-two in Venezuela and thirty-six in Colombia – with politicians, journalists, congressional staffers, justices, clerks, members of advocacy groups and academics, and I did archival research in newspapers and congress (see Appendix A for more details).
In Colombia, I reviewed the weekly political magazine Revista Semana between 2002 and 2010. Using data from Congreso Visible alongside my own data on members of congress’ support for the government, I also identified and reviewed the congressional debates of sixteen bills introduced by the government coalition to increase the powers of the executive and/or extend Uribe’s time in office. Together, the interviews, newspaper, and archival research enabled me to analyze the danger posed by different antidemocratic bills, the strategies used by the opposition inside the legislative to stop them, and the extent to which these strategies helped curb Uribe’s attempts to erode democracy.
In Venezuela, the data collection process was a bit more convoluted. On the one hand, an excellent archive publicly available in El Nacional allowed me to review this newspaper between 1998 and 2010 with relative ease. On the other hand, government control over the National Assembly and its archive (Archivo Legislativo de la Asamblea Nacional) made legislative data difficult to obtain. With the help of some officials inside the National Assembly, I was eventually able to access a list of all bills introduced between 2000 and 2010, their author, the dates of their debates, and whether they were approved or not. I used that document to build a legislative dataset and identify congressional bills introduced by the government coalition in order to increase the powers of the executive.Footnote 5
Ideally, I would have reviewed the debates for each one of these bills. Unfortunately, I was denied access to the official transcripts. Luckily, the Legislative Information Office (Dirección de Información Legislativa [DIL]) – in charge of providing information to the public regarding new laws – has files of all bills approved since 2000, including their debates. In this office, I was able to review fifteen congressional antidemocratic amendments introduced and approved by the government.
1.3.3.1 Empirical Implications
I use the evidence outlined earlier to detail the pathway between opposition strategic choices and regime outcomes. Because each case serves a different purpose, I reconstruct the causal pathway differently in Venezuela and Colombia (and design the chapters accordingly). In Venezuela, I use the data collected for three purposes. First, I resort to interviews, newspapers, and secondary sources to demonstrate that the erosion of democracy in this country was not fait accompli in 1999. I do that by describing the various institutional and noninstitutional resources the opposition had and how they were, or could have been, useful to defeat Chávez. Second, I utilize a similar set of data to demonstrate that it was, in fact, the use of radical extrainstitutional strategies that jeopardized these resources. I describe specific instances of these strategies and reveal how each one of them undermined the legitimacy of the opposition while boosting that of the president. I also illustrate how these strategies enhanced Chávez’s ability to remove opposition members from state institutions, replace them with loyalists, and then use these institutions to further erode democracy. Together, the data collected provides enough evidence to show that the Venezuelan leader was able to undermine horizontal accountability without losing his democratic façade.
Third, using the legislative dataset, I show the viability of institutional and moderate extrainstitutional strategies, regardless of the context. The fact that the opposition had several seats in the National Assembly in 2000–5 but almost none in 2006–10 provides useful within-case variation to examine my argument. It allows me to compare the performance of congress with or without the members of the anti-Chavista coalition. Using descriptive statistics, I show that having members of the opposition in the legislature made a significant difference in the amount of legislation introduced and passed, as well as the overall delay of the legislative process, including pieces of legislation key to the erosion of democracy. I complement this evidence with in-depth analysis of a few pieces of legislation, where I show how the opposition used moderate institutional strategies to effectively delay processes of co-optation of courts and media outlets.
The burden of proof is slightly different in Colombia. Here I use evidence to perform a different set of tasks. First, I show that the survival of democracy was not fait accompli in 2002. For that purpose, I use interviews and newspapers to describe the threats that Uribe posed. I demonstrate that he was as populistic and polarizing, and as willing to erode democracy, as his Venezuelan counterpart. I also demonstrate that courts, congress, and oversight agencies were in real jeopardy.
Second, I connect the absence of radical extrainstitutional strategies to the persistence and enhancement of the opposition’s legitimacy vis-à-vis the president’s. I further tie that legitimacy to the opposition’s ability to keep its presence in congress, courts, and oversight agencies. I rely on newspapers and interviews to show not only that the opposition was legitimate but that legitimacy afforded it important leverage domestically and abroad as well. Relying on within and cross-case variation, I provide evidence to suggest that, given a chance, Uribe would have taken advantage of the opposition’s lack of legitimacy to undermine his opponents and remove them from office.
Finally, I outline how institutional and moderate extrainstitutional strategies helped prevent the erosion of democracy. Using congressional records, court rulings, and interviews, I show how legislative obstruction and targeted boycotts and demonstrations tamed Uribe’s antidemocratic reforms. Crucially, relying on within-case variation, I demonstrate that, without active input from congress, it would have been much harder for the Constitutional Court to rule against the president in a case about the legality of his attempt to expand presidential term limits.
1.4 Book Plan
This book is organized as follows. Chapter 2 outlines the theory. I start by defining the dependent variable. Doing so allows me to separate the rise of executives with hegemonic aspirations from their ability to erode democracy. I leave the theory and analysis of the first stage of democratic erosion for Chapter 3 and move to the second stage, where I describe the actors, their preferences, the strategies available to them, and their trades and payoffs. Within that framework, I outline the mechanisms by which oppositions that use radical extrainstitutional strategies betting on short term wins, end up curtailing their ability to protect democracy, the mechanisms by which oppositions that use moderate institutional strategies help prevent the erosion of democracy – even if not immediately – and the mechanisms by which oppositions that resort to radical institutional strategies and/or moderate extrainstitutional strategies can help or hinder their ability to shield democratic institutions from erosion. In this chapter, I also analyze some of the scope conditions and explore some of the potential drivers behind opposition strategic choices.
Chapter 3 focuses on the first stage of the argument. Based on the extant literature on regime and regime change, I highlight a set of variables that are essential to understanding the rise of presidents with hegemonic aspirations. Using an original dataset as well as data collected from several other sources, I show that weak states, weak party systems, and weak economic performance increase the likelihood of electing executives willing to circumvent democracy. I also show that these variables are not as useful to explain the variation between the Latin American potential autocrats who successfully eroded democracy and the ones who failed. This chapter assesses other alternative explanations as well. I show that, while favorable economic conditions and weak institutions at the onset of erosion as well as mass support and polarization during a hegemonic president’s term help potential autocrats erode democracy, they do not guarantee this outcome.
Chapters 4 and 5 use comparative historical analysis to address this puzzle. They trace the dynamics of erosion in Venezuela and Colombia, respectively. Chapter 6 expands that analysis to four other cases: Evo Morales in Bolivia, Recep Tayyip Erdogan (2003–present) in Turkey, Law and Justice Party (2015–present) in Poland, and Viktor Orbán (2010–present) in Hungary. The first three cases conform to the theory. I show how opposition’s radical extrainstitutional strategies jeopardized their ability to prevent the erosion of democracy in Bolivia and Turkey, and how moderate institutional and extrainstitutional strategies have helped delay the erosion of democracy in Poland. The last case, Hungary, does not conform to the theory. I use it to assess important scope conditions, in particular the importance of having time (and the institutional resources time affords) and an international community with a normative preference for democracy. Chapter 7, the conclusion, highlights the theoretical and empirical contributions of Opposition at the Margins.
