The Pandemic and the Crisis of Democracy in Brazil

Remy Maduit | Authors published

The Pandemic and the Crisis of Democracy in Brazi

Leonardo Avritzer is a Professor of Political Science at Universidade Federal de Minas Gerais (UFMG), Brazil, and a Researcher at The National Council of Scientific and Technological Development (CNPq), Brazil.
Lucio Rennó is a Professor at the Political Science Institute of the University of Brasília, Brazil, and a Researcher at The National Council of Scientific and Technological Development (CNPq), Brazil.

The erosion of democracy is taking place worldwide. Steven Levitsky, Daniel Ziblatt, and Adam Przeworski, among many others, established the main elements of a theory on the crisis of democracy as situations in which status quo institutions are in some kind of disarray: “… no change occurred, but it may… the current situation is in some ways threatening…”. [1] For Levitsky and Ziblatt [2], two elements are in jeopardy: institutions and legality.

There are, however, disagreements on how widespread the crisis of democracy is, which countries fit the theory, and how to reverse current threats. By considering issues of mass behavior, in particular perspectives on popular support for anti-democratic solutions and regime legitimacy, the current debate about the crisis of democracy can be better qualified. [3] This is especially true for countries that have undergone repeated episodes of turbulence and novel forms of instability in the shape of mandate interruption, constitutional impasses, and impeachments. [4]

We explore the current dilemmas of democracy in Brazilian, focusing on public opinion and mass political behavior. Anti-democratic attitudes have gained increased salience in the country, following two impeachment processes, repeated corruption scandals, and the election of an outsider—with clear authoritarian tendencies—in the 2018 elections. The weakening of political institutions created a fertile ground for populist attacks on democracy from the electoral system to the operation of counter-majoritarian institutions. Erratic behavior has marked the first eighteen months of Jair Bolsonaro’s administration by the president, unceasing conflict among political groups, threats of mandate interruption, volatile coalitions, and disdain for public policy. The Covid-19 pandemic—a clear external shock—has further deepened conflicts. The pandemic has brought governors back to the center of Brazilian politics, as a counterpoint to the faltering Federal Government’s position on the health crisis.

Therefore, Brazil has become an interesting case to explore the effects of the pandemic on the dynamics of the democratic crisis and attempted populist attacks on democracy. Bolsonaro has adopted radical stances toward the pandemic and is considered to be one of the most extreme negationists. [5] How have regime legitimacy and authoritarian attitudes, indicators of the extent of the crisis of democracy, and support for a populist, authoritarian leader interacted, and how have they been affected by the Covid-19 pandemic? In this article, we will discuss the crisis of democracy in Brazil using public opinion data from 2018 to 2020. Hence, we compare trends in attitudes before and during the pandemic. Has Bolsonaro’s extreme stances on the pandemic and democracy itself provoked more radicalism among the public?

The article is divided into four sections and this introduction. Next, we discuss influential theoretical contributions for identifying the dimensions of the crisis of democracy, incorporating the discussion on regime legitimacy and authoritarianism at the mass level. We then focus on the Brazilian, exploring current events, the Bolsonaro government, and populist behavior. We analyze public opinion data from Brazil, using the “Cara da Democracia” surveys for 2018, 2019, and 2020. Finally, we present our conclusions.

Theory

Democracy is an essentially contested concept involving endless disputes about its meaning, its proper uses, and a broadly accepted definition of it. [6] The crisis of democracy is also prone to contestation. Within the sphere of political science, there are several theories on how the development of democracy unfolds, and these influence how we may analyze the current crisis of democracy. Four important works have structured the debate in the last few years. [7] Each one draws on a conception of democracy.

Steven Levitsky and Daniel Ziblatt have set up the key elements of a theory on the crisis of democracy from a minimalist perspective of democracy. For Levitsky and Ziblatt, there are two elements at the core of democracy: institutions and legality. In their view, democracy’s frailty is its potential to break down from within, rather than through violent seizures of power. [8] They are concerned about legal efforts to subvert democracy “… in the sense that they [anti-democratic actions] are approved by the legislature and accepted by the courts”. [9] The authors establish their framework for understanding the crisis of democracy by breaking with a key tenet of democratic theory that a crisis of democracy leads to a coup. For them, it is internal degradation carried out by populists or outsiders—often under the law—that should be because of concern. In this sense, “crisis” is used to mean the internal degradation of the consensus on democracy at the elite level.

Adam Przeworski provides a distinct conception of crisis. Przeworski draws on a more standard theory of crisis: for him, a crisis occurs when “the current situation is in some ways untenable, that some threats to democracy have already materialized, yet the status quo democratic institutions remain in place”. [10] Thus, Przeworski remains in the realm of a well-established theory of crisis as an intermediary situation in which elements of the old order no longer function, while crises are untenable in the long term. For him, when democracy is in crisis “… incumbents make it impossible for the opposition to remove them from office… or the opposition does not recognize the legitimacy of the government and the government defends itself by repression…”. [11] Again, the emphasis is on elite disputes over the regime. Yet, one element is missing in Przeworski’s analysis, and that is the attempt by those in power to degrade political institutions.

The comparison of the two theories opens up an avenue to study the current crisis of democracy. Each vision of the crisis is based on a theory of democracy. Przeworski’s version is strongly grounded in the Huntingtonian view of democracy, the concept that Anglo-Saxon democracy is a one-way street. [12] According to this view, there is a core of democratic countries belonging to the Anglo-Saxon world—in particular, Britain, the United States, and several European nations — that are not at risk of de-democratization. In this sense, Przeworski privileges the idea that a crisis in Anglo-Saxon democracies does not lead to a breakdown. [13]

Levitsky and Ziblatt do not have an exact theory on the crisis of democracy. If they had one, they would have to show the direction in which the crisis would progress. However, they have a different perspective from the Huntingtonian/Przeworskian argument. For Levitsky and Ziblatt, there is a possibility of internal degradation of democracy that will lead to the end of democracy without a regime breakdown. This possibility is rooted in the fact that important political leaders within the major democracies do not respect the elements that make democracy strong and stable in institutional terms. The authors point out that primary elections in the United States have broken with a conception of gatekeeping, which kept historically populist leaders away from the process of formation of government and governance. [14] Thus, degradation from within breaks with the Huntingtonian element and incorporates Anglo-Saxon countries in the crisis of democracy framework.

What neither of the above approaches considers are the reverberations of elite dissent on mass politics. Does the crisis of democracy, which emerges more clearly through elite behavior, reflect and dialogue with mass attitudes? Yascha Mounk’s theory on the erosion of trust and David Runciman’s conception of a democratic deficit call attention to changes in the structure of support for democracy at the mass level. For Runciman, democracy has a delivery problem that affects public support. Democracy was based on the delivery of dignity and long-term respect, and it cannot accomplish both nowadays. [15] Mounk develops a similar diagnosis based on the decline of trust in government: “Government legitimacy, these scholars admitted, had declined: citizens have become much more willing to challenge their current rulers. But regime legitimacy, they insisted, had remained stable… This is an appealing story. But over the recent years, it has looked less and less plausible”. [16]

It is necessary, therefore, to understand deconsolidation from the perspective of mass regime support. As Foa and Mounk [17] point out, when citizens cease to believe that democracy is the “only game in town” and start accepting authoritarian alternatives, deconsolidation is on its way. They argue that to measure deconsolidation it is necessary to consider popular support for democracy, openness to non-democratic alternatives, and support for anti-system parties and candidates.

Hence, we must look at the literature on regime legitimacy, especially institutional trust and satisfaction with the functioning of the system, to better comprehend how these elements work in practice. [18] Curiously, the literature on the crisis of democracy has rarely entered debates on mass regime support and authoritarian attitudes.

The two major forms of legitimacy—diffuse and specific—date back to Easton’s [19] original systems theory approach, arguing that citizens can position themselves critically toward both the general, latent ideas of how the political system is organized and towards outcomes generated by performance. Trust in institutions captures diffuse forms of system support, whereas satisfaction with democracy relates to the results offered by the system in its broadest scope. In this article, we will measure the crisis of democracy by looking at oscillations between these two variables: latent and specific system support. It is important to point out that citizens’ lack of trust in democracy opens up space for criticisms advanced by populist incumbents.

Besides system support, authoritarian attitudes are important for understanding adherence to anti-democratic alternatives. [20] There is a long tradition of measuring authoritarianism on an individual level based on psychological traits related to values ingrained during socialization and ideas received from parents. [21] Authoritarian attitudes based on regime preference, contrasting democracy with other alternatives, and focusing on values associated with order and liberty have also been studied. [22]

Independent of the measurement strategy, studies find that citizens with authoritarian attitudes support candidates who espouse authoritarian rhetoric or who were active in previous periods of authoritarian rule. In contemporary Latin America, the authoritarian vote is particularly connected with right-wing candidates. [23] Therefore, it is important to know the scope and range of authoritarian attitudes to understand the potential for popular support of democratic breakdown or corrosion.

We operationalize authoritarian attitudes in the Brazilian public through support for military intervention, given its role in regime breakdown in Latin America. [24] In Brazil, views favorable to military intervention have gained significant mass support in recent years—as expressed in political protests and demonstrations—and are in line with the declared and manifest sympathies of the current president.

In this article, we will use the four theories to explain the crisis of democracy in Brazil. Our point of departure is the Levitsky/Przeworski approach regarding the elite level. We show how Jair Bolsonaro has become a champion of authoritarianism among Brazilian political elites, openly defending rhetoric that undermines the idea of democracy as “the only game in town.” Bolsonaro’s open defense of the Military Regime in the past and of current military intervention, along with a discourse of intolerance, radicalized ideological cleavages, undermining democracy from within. Next, we focus on the masses, discussing how the decline of trust in political institutions—besides authoritarian attitudes — is related to the crisis in democracy.

We also show that restorative moves occur within the erosion process. If, on the one hand, Brazil has a president who publicly defends authoritarian avenues once the risk of regime interruption looms high on the political horizon, corrective forces from classical political actors and control institutions mobilize to resist the threats and push authoritarians to retreat. How this plays out in public opinion is as yet unknown, especially in light of the external shock brought on by the Covid-19 pandemic.

The Deterioration of Democracy, Bolsonaro’s Election, and the Covid-19 Pandemic

By several standards, Brazil had a successful process of democratic development until 2014. [25] There is plentiful evidence to support this statement: a non-contentious process of handover of power in all elections between 1989 and 2014; overlapping different policies, leading the Workers Party (PT, Partido dos Trabalhadores) to accept the fiscal policies of the Social Democrats (PSDB, Partido da Social Democracia Brasileira) and then to accept the social policies of the PT. Parties in government increased the power of the Brazilian counter-majoritarian institutions and accepted a new role in civil society and participation in politics. [26]

To some, Brazil seemed to have broken a cycle of poverty and exclusion, mostly by building political institutions that led to economic institutions being inclusive. [27] Acemoglu and Robinson wrote about Brazil during a moment of hope and excitement in the country. The Economist, a few years earlier, had passed the verdict: “Brazil takes off!” [28] A few years later, everything seemed to have gone off-course, and the ruling was revised: “Has Brazil blown it?” [29] Had Brazil moved back to its traditional exclusionary pattern, showing how vulnerable moments of expansion and inclusion are? What happened to the exemplary case that broke with historical patterns of exclusion?

In 2014, or perhaps in June 2013, Brazil fell into an acute and ongoing economic and political crisis. A long series of corruption scandals tainted all major parties—PT, PSDB, and the Brazilian Democratic Movement (MDB, Movimento Democrático Brasileiro). [30] In addition, the economy was derailed, with the worst recession in a century. Bolsonaro rose to power with a heavily conservative and anti-systemic discourse. [31] Finally, at the beginning of 2018, calls for military intervention emerged at the public level. Street protests by right-wing movements first appeared in 2013, an interval of decades from the conservative marches of the early 1960s.

In 2018, Brazil elected a politically reactionary president who does not believe in negotiation and coalition-building, democratic politics, or counter-majoritarian institutions. His coming to power created a new populist disequilibrium between the presidency and counter-majoritarian institutions that were strengthening their control capacities during the New Republic.

The first year of Bolsonaro’s administration was successful in advancing his campaign proposals. Bolsonaro carried out a strategy of institutional disruption in three policy areas: human rights, the environment, and education. He discontinued key, successful policies, facing little opposition. In Congress, he could push through significant pension reform.

However, the Covid-19 pandemic introduced additional elements of political turbulence in Brazil. Bolsonaro’s positions regarding the pandemic are at the core of the current instability.

Bolsonaro epitomizes the denial and downplaying of the seriousness of the pandemic. As early as March 2020, in a national network speech, he minimized the disease, calling it a gripezinha (“a light flu”) and suggesting that those who were fit would be immune. He contested social distancing measures, claiming they would hurt the economy. [32] And as late as August 2020, after having contracted the virus, Bolsonaro vetoed legislation that made mask use mandatory nationwide. He also stated publicly on 19 August 2020 that face masks had zero efficacy, which contradicted expert opinion. [33] Finally, Bolsonaro adopted anti-scientific positions, favoring the use of controversial medical treatments. Bolsonaro’s position is of defiance and distrust toward specialists’ recommendations and science-based views and data, using conflict as his dominant form of political engagement.

It is surprising that, despite his defiant behavior towards Covid-19, Bolsonaro’s administration was initially well-placed to successfully tackle the pandemic. Bolsonaro’s Health Minister was a doctor with a great deal of experience. Luiz Henrique Mandetta is a moderate politician with experience in public health administration. He organized what could have been a successful response to the pandemic, with daily briefings and inter-federation coordination, creating trust in the state’s response to the health crisis.

However, the president himself derailed Mandetta’s efforts. Bolsonaro shared fake information on social media and broke social distancing laws by going to anti-democratic rallies without a mask. He was also a leading voice in weakening compliance with social distancing. His nationwide speech in March 2020 had a strongly demoralizing effect on the population, decreasing levels of social distancing. [34]

The pandemic generated a dispute regarding the adequacy of the president’s policy. Bolsonaro’s approval rates for handling the public health crisis sank, and the president’s response to decreasing popularity and increasing political protest was to threaten democratic institutions. For seven straight weeks between March and late May 2020, he attended rallies in which military intervention was defended and democratic institutions sidelined. Thus, Bolsonaro’s attacks on majoritarian institutions such as Congress and counter-majoritarian institutions such as the Supreme Court ratified a mistrust of checks and balances. Bolsonaro reveled in the idea of a military coup with him in power or what was called by demonstrators intervenção militar com Bolsonaro (“military intervention with Bolsonaro”).

The politicization of the coronavirus crisis in Brazil began with a clash between Bolsonaro and the governors/mayors who adopted social distancing policies. In the federal government, disagreements between Bolsonaro and Mandetta escalated until they fired the minister.

In addition, Bolsonaro’s relationship with Congress has been turbulent. In the first year of his administration, Bolsonaro vetoed more legislative proposals than prior presidents in the same period and had five times more vetoes overturned by Congress. He cast sixty-two vetoes, compared to forty-six by his predecessor, Michel Temer. Congress overturned seventeen vetoes of Bolsonaro’s, compared to three of Temer’s. No other prior president had had vetoes overturned in their first year of government. Bolsonaro also had the lowest legislative success rate of any prior president and the second-highest number of decrees enacted, losing to Collor de Mello in 1990. [35] In dealing with the pandemic, Congress and the Chamber of Deputies, in particular, took a leading role in promoting social security policies to assist poor populations during periods of lockdown.

Courts have also become pockets of resistance against Bolsonaro. The Supreme Court has ruled against the Bolsonaro government frequently. Bolsonaro’s defeated attempts to take from governors’ and mayors’ responsibility for social distancing policies is a paradigmatic example.

The people also rebelled against Bolsonaro. Self-defined “anti-fascist” soccer team fan clubs in São Paulo protested on 31 May 2020 against pro-military intervention demonstrations. [36] Similar protests inspired the recent anti-fascist protests in Brazil in the United States, where riots and demonstrations sparked by George Floyd’s death positioned the defense of democracy and civil liberties at the forefront of social protest.

Ever since the reactions of the elite and the masses, pro-Bolsonaro rallies have significantly diminished, and Bolsonaro has made space in his government for the traditional catch-all clientelistic parties. He appointed to the Supreme Court an unknown judge backed by these groups and critical of the Lava-Jato operation. [37]

Thus, Bolsonaro’s first eighteen months in government fit well into the discussion of the crisis of democracy and the populist erosion of democracy. There is a continuous process of institutional disruption taking place, a frustration of economic growth and inclusion, and an open attack on counter-majoritarian institutions through the threat of a military coup to be carried out by Bolsonaro himself. But how does this reflect mass political behavior and attitudes? How has the Brazilian public reacted to these movements and countermovements?

Military Intervention, Legitimacy Crisis, and Support for Bolsonaro: Data and Analysis

We use “A Cara da Democracia” national surveys of March 2018, April 2019, and June 2020 to discuss mass politics in Brazil. Sample sizes and interview mode vary between the years, but all represent urban populations and are therefore comparable. [38] In addition, although questionnaires varied between the years, items were repeated in all three waves and the question order of the items we analyse here was identical. We limit our analysis to the items that were repeated each year to increase comparability.

This is the only dataset available (with representative national samples) that permits the short-term comparison of responses in years of intensified conflict over democracy in a recently democratized country, including a wave that occurred during a pandemic. Therefore, the data capture how the initial months of the pandemic may have affected public opinion. Hence, we rely on a unique dataset to examine (1) popular reactions to issues associated with the crisis of democracy and (2) how the socially tense and politically convoluted response to the pandemic in Brazil may have affected popular opinion.

We focus first on an important element of democratic deterioration that reemerged in 2018: military intervention. The idea of military intervention was a key component of Brazilian political behavior during the fifties and early sixties. [39] The 1988 Constitution kept a door open for military intervention in politics in its Article 142. Introducing the GLO (Law and Order Guarantees) in 1997, allowing the army to deal with criminal issues in large Brazilian cities if called upon to do so by one of the political branches (Lei Complementar 97/1999), also helped bring the military back into political life. Last but not least, MINUSTAH, the Brazilian mission in Haiti, associated the concept of military intervention with actions against criminality.

Table 1. shows significant support for military intervention in politics, but at a steadily declining rate. [40] The political atmosphere leading to the 2018 elections was receptive to Bolsonaro’s critical position toward democracy and mainstream politics. However, support for military intervention oscillated over time and is waning. It ranges from a favorable majoritarian position towards military intervention under conditions of crime and corruption in 2018 to a minority position in 2020. Hence, the pandemic seems to have maintained the rate of decrease in popular support for openly anti-democratic positions that were already present in 2019.

A second important factor is the historically low levels of institutional trust in Brazil. [41] The situation leading up to the 2018 elections, given the high level of exposure to political corruption, may have significantly increased mistrust. Table 2 shows great stability in institutional mistrust, with more variation regarding the Judiciary, reaching a high point in 2019, but receding to the levels of 2018 in 2020. We also see a decrease of mistrust in Congress, of about ten percentage points, in 2020.

The improvement in Congressional evaluation, in particular, can be attributed to its response to the pandemic and its critical position in the Bolsonaro government’s handling of the health crisis. Congress has continued to be operational during the pandemic, working remotely, and has kept very active in promoting supportive and protective health and economic policies. Congress did not just approve, but it increased the amount of emergency aid to poor families proposed by the Executive Branch during the pandemic. This was one of many occasions in which Congress, through its leadership, parted ways with the Bolsonaro government, in defense of a more moderate and active response to the pandemic.

Dissatisfaction with democracy also varied over time, but at high levels. This indicator shows a critical feeling of the population towards the level of functioning of democracy in Brazil, which opens up space for outsiders, populists, and authoritarian alternatives. The data show a structural mistrust of democratic institutions and dissatisfaction with democracy. However, the pandemic, together with Bolsonaro’s reaction to it and his position on democracy, may have had the unexpected effect of attenuating radical authoritarian attitudes in the public and spurring specific forms of support for the system.

Finally, support for Bolsonaro—a leader with populist and authoritarian inclinations—is important for understanding the current risk of democracy in Brazil. Core Bolsonaro supporters made up 13 percent of the Brazilian electorate in early 2018, increasing to 20 percent and 21 percent in the years following the election. [42] Hence, Bolsonaro has not lost the momentum he gained during the 2018 campaign in his first years in office, despite the recurrent crises and his controversial positions regarding the Covid-19 pandemic. However, it is possible to speculate that Bolsonaro has not gained supporters during the pandemic, something he might have done if he had adopted moderate positions.

Finally, we look at the socio-economic and political profile of authoritarian citizens and Bolsonaro supporters. For the sake of brevity, we combined the above traits associated with the deconsolidation of democracy—support for military intervention, mistrust in institutions, and dissatisfaction with democracy—into a single indicator with a value of one (1) if the respondent holds all the above attitudes. Hence, we use a composite index for authoritarianism based on attitudes toward the democratic regime. [43] This is an extreme measure of support for anti-democratic solutions, and it represents a small portion of the Brazilian population. In 2018, this was 9 percent of the population, falling to 4 percent in 2019 and 2020. Hence, data shows a very low rate of concurrent adhesion to several dimensions of authoritarianism, and this rate also decreases in the brief span of three years.

Who are the authoritarians and how are authoritarian attitudes related to Bolsonaro’s support? As seen in Table 3, it is important to highlight that there is a two-way relationship between being a core Bolsonaro supporter and authoritarianism. In 2018 and 2020, the likelihood of these two attitudes converging was extremely high. We don’t disentangle the direction of causality, but we show that bolsonarismo and authoritarianism go hand in hand.

Still, the characteristics of authoritarian and core bolsonarista differ from each other. PT partisanship is a strong negative predictor of bolsonarismo but not of authoritarianism. The perception of corruption as the main national problem is a significant predictor of bolsonarismo, but not of authoritarianism. Men are more likely to be both authoritarian and core bolsonaristas; those self-identified as black or brown (pardo) skin color are less likely to be both authoritarian and core bolsonaristas. Evangelicals are much more likely to be core bolsonaristas, but not authoritarians. The effect of the age variable on bolsonarismo changed over time, with younger citizens being more likely to support Bolsonaro in 2018, but not so much in 2019 and 2020, when older people were more likely to adhere more strongly to bolsonarismo. [44]

The decrease in size of the authoritarian group has made its relationship with bolsonarismo stronger–as these variables present larger odds of influencing each other. The pandemic does not seem to have reversed, deepened, or profoundly changed patterns that were already developing. There is the crystallizing out of a group with conservative positions in Brazil, with some anti-democratic inclinations, that were not altered significantly by the pandemic. This group currently makes up about 20 percent of the population.

Conclusion

We studied dimensions of the democratic crisis as reflected in public opinion in an interesting contemporary case that overlaps an outsider president with populist and authoritarian inclinations and the significant impact of the Covid-19 pandemic. Brazil has undergone dramatic waves of turbulence and instability since 2014, with election results questioned, an impeachment process, the election of a right-wing extremist, and continuous and recurrent episodes of political crisis coupled with the most serious economic downturn in Brazilian history.

The year 2020 has been especially tumultuous, given President Bolsonaro’s downplaying of the Covid-19 pandemic and several concurrent crises that struck his administration. Hence, Brazil is an important case for studying the depth and reach of the so-called crisis of democracy during the pandemic.

Data from public opinion polls show that mistrust of democratic institutions and dissatisfaction with democracy are structural components of Brazilian mass politics, which confirms prior studies. [45] However, we also saw a decline in support for military intervention from 2018 to 2020 and an improvement in the evaluation of Congress. Therefore, the pandemic did not contribute to the deepening of a democratic crisis among the Brazilian public. When we condense anti-democratic attitudes into a single indicator, a tiny portion of the population encompasses all the distinct positions simultaneously, and does so at a decreasing rate, further raising doubts about the depth of the crisis of democracy.

Finally, support for Bolsonaro increased during the 2018 campaign and stabilized in 2019 and 2020, with one in five Brazilians strongly aligning with the president. The pandemic has not had a significant effect on the dynamics of these variables in the short run.

There are two perspectives on the stabilization of support for Bolsonaro and the decrease in support for a coup: the first is the Przeworskian view of how the crisis was halted. The other possibility is to look only into the deterioration results. However, the reversion of the process of deterioration also needs to be singled out. Foa and Mounk’s interpretation of de-consolidation appears plausible. In a young democracy, a reversal in the processes of democratic consolidation is much more likely than in consolidated ones. Mistrust in institutions and dissatisfaction with democracy have been structural traits of Brazilian politics and help facilitate deconsolidation in younger democracies.

Finally, the pandemic has not decreased support for Bolsonaro or increased authoritarianism but may have halted the former. Bolsonaro did not try to promote the union of Brazilians against a common threat embodied by the pandemic. Instead, he further advanced his agenda of conflict and polarization. Still, Bolsonaro feeds on authoritarianism, and backed by anti-democratic forces, he personifies a current threat to democracy. If his popularity increases, which is plausible given his widely publicized actions on strengthening welfare policies, for which he has claimed credit, and the ideological adhesion of a significant contingent of “true believers,” he could become an even greater menace to democratic stability. If elected for a second term, Bolsonaro might further advance his anti-democratic and conservative agenda, as have done many other populists when they were in power. [46] What is momentary could become permanent and institutionalized.

---

[1] Przeworski, A (2019) Crises of Democracy. Cambridge: Cambridge University Press.

[2] Levitsky, S., Ziblatt, D (2018) How Democracies Die. New York: Crown

[3] Foa, RS., Mounk, Y (2019) Democratic deconsolidation in developed democracies, 1995–2018. In: Harvard Center for European Studies Open Forum Series 2018–2019. Cambridge, MA: Harvard University Press.

[4] Boniface, D (2007) The OAS’s mixed record. In: Legler, T., Sharon, FL., Boniface, D. (eds) Promoting Democracy in the Americas. Baltimore, MD: The John Hopkins University Press.
Llanos, M., Marsteintredet, L (2010) Presidential Breakdowns in Latin America: Causes and Outcomes of Executive Instability in Developing Democracies. New York: Palgrave MacMillan.
Perez-Liñan, A (2007) Presidential Impeachment and the New Political Instability in Latin America. Cambridge: Cambridge University Press.

[5] https://www.theatlantic.com/politics/archive/2020/03/bolsonaro-coronavirus-denial-brazil-trump/608926/

[6] Collier, D., Levitsky, S (1997) Democracy with adjectives. World Politics 49(3): 430–451.
Gallie, WB (1956) Essentially contested concepts. Proceedings of the Aristotelian Society 56(1): 167–198.doi:10.1093/aristotelian/56.1.167

[7] Levitsky, S., Ziblatt, D (2018) How Democracies Die. New York: Crown.
Mounk, Y (2019) The People Versus Democracy. Cambridge, MA: Harvard University Press.
Przeworski, A (2019) Crises of Democracy. Cambridge: Cambridge University Press.
Runciman, D (2018) How Democracy Ends. London: Profile Books.

[8] Levitsky, S., Ziblatt, D (2018) How Democracies Die. New York: Crown.

[9] Ibid.

[10] Przeworski, A (2019) Crises of Democracy. Cambridge: Cambridge University Press.

[11] Ibid.

[12] Huntington, S (1991) The Third Wave: Democratization in the Late Twentieth Century. Norman: University of Oklahoma Press.

[13] Przeworski, A (2019) Crises of Democracy. Cambridge: Cambridge University Press.

[14] Levitsky, S., Ziblatt, D (2018) How Democracies Die. New York: Crown.

[15] Runciman, D (2018) How Democracy Ends. London: Profile Books.

[16] Mounk, Y (2019) The People Versus Democracy. Cambridge, MA: Harvard University Press.

[17] Foa, RS., Mounk, Y (2019) Democratic deconsolidation in developed democracies, 1995–2018. In: Harvard Center for European Studies Open Forum Series 2018–2019. Cambridge, MA: Harvard University Press.

[18] Booth, J., Seligson, M (2009) The Legitimacy Puzzle in Latin America: Political Support and Democracy in Eight Latin American Nations. Cambridge: Cambridge University Press.
Dalton, R (2004) Democratic Challenges, Democratic Choices: The Erosion of Political Support in Advanced Democracies. Oxford: Oxford University Press.
Norris, P (1999) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press.

[19] Easton, D (1965) A Systems Analysis of Political Life. New York: John Wiley.

[20] Cohen, MJ., Smith, AE (2016) Do authoritarians vote for authoritarians? Evidence from Latin America. Research & Politics 3(4): 1–8.doi:10.1177/2053168016684066
Seligson, A., Tucker, J (2005) Feeding the hand that bit you: voting for ex-authoritarian rulers in Russia and Bolivia. Demokratizatsiya: The Journal of Post-Soviet Democratization 13(1): 11–44.doi:10.3200/DEMO.13.1.11-44

[21] Cohen, MJ., Smith, AE (2016) Do authoritarians vote for authoritarians? Evidence from Latin America. Research & Politics 3(4): 1–8.doi:10.1177/2053168016684066

[22] Seligson, A., Tucker, J (2005) Feeding the hand that bit you: voting for ex-authoritarian rulers in Russia and Bolivia. Demokratizatsiya: The Journal of Post-Soviet Democratization 13(1): 11–44.doi:10.3200/DEMO.13.1.11-44

[23] Cohen, MJ., Smith, AE (2016) Do authoritarians vote for authoritarians? Evidence from Latin America. Research & Politics 3(4): 1–8.doi:10.1177/2053168016684066

[24] Passos, A (2018) The military mystique: the war on drugs in Brazil and Mexico. PhD Thesis, Institut d’études politiques, Sciences Po, France.
Stepan, A (1971) The Military in Politics. Princeton, NJ: Princeton University Press

[25] Kingstone, P., Power, T (2008) Democratic Brazil Revisited. Pittsburgh, PA: University of Pittsburgh Press.
Montero, A, (2014). Brazil: reversal of fortune. Cambridge: Polity Press,.

[26] Avritzer, L., Marona, M (2017) A tensão entre soberania e instituições de controle. Dados [online] 60(2): 359–393.

[27] Acemoglu, D., Robinson, J (2012) Why Nations Fail: The Origins of Power, Prosperity, and Poverty. New York: Random House.

[28] https://www.economist.com/leaders/2009/11/12/brazil-takes-off

[29] https://www.economist.com/leaders/2013/09/27/has-brazil-blown-it

[30] Rennó, LR (2020) The Bolsonaro voter: issue positions and vote choice in the 2018 Brazilian presidential elections. Latin American Politics and Society 62(4): 1–23.doi:10.1017/lap.2020.13

[31] Ibid.

[32] https://noticias.uol.com.br/politica/ultimas-noticias/2020/03/24/leia-o-pronunciamento-do-presidente-jair-bolsonaro-na-integra.htm

[33] https://noticias.uol.com.br/saude/ultimas-noticias/redacao/2020/08/19/bolsonaro-mascara-eficacia.htm?utm_source=twitter&utm_medium=social-media&utm_content=geral&utm_campaign=noticias

[34] Fernandez, M., Bertholini, F., Santana, L (2020) Estados diante da Pandemia de Covid-19. Jota, 19 August. Available at: https://www.jota.info/opiniao-e-analise/artigos/estados-diante-da-pandemia-de-covid-19-19082020 (accessed 22 April 2021).

[35] Source: National Congress.

[36] https://www.brasildefato.com.br/2020/05/31/pm-avanca-contra-manifestantes-em-ato-pro-democracia-na-av-paulista-sp

[37] https://noticias.uol.com.br/colunas/tales-faria/2020/10/06/indicacao-ao-stf-e-fritura-de-guedes-revelam-o-verdadeiro-bolsonaro.htm; https://www.bbc.com/portuguese/brasil-54364957.

[38] In 2018, 2,500 face-to-face interviews were conducted in 179 municipalities. In 2019, 2,009 respondents were interviewed face-to-face in 151 municipalities, and in 2020, 1,000 interviews were conducted over the telephone in 69 municipalities.

[39] Carvalho, JM (2019) Forças Armadas e Política no Brasil. São Paulo: Todavia.

[40] See the Appendix for the wording of items and coding of responses.

[41] Rennó, L., Smith, A., Layton, M (2011) Legitimidade e Qualidade da Democracia no Brasil: Uma Visão da Cidadania. São Paulo: Intermeios; Nashville: LAPOP.

[42] In 2018, core Bolsonaristas were respondents who said they would vote for Bolsonaro in both electoral scenarios, with and without Lula da Silva. In 2019 and 2020, we code respondents as core supporters if they evaluate the Bolsonaro government positively and if they stated that they had voted for him in 2018.

[43] Seligson, A., Tucker, J (2005) Feeding the hand that bit you: voting for ex-authoritarian rulers in Russia and Bolivia. Demokratizatsiya: The Journal of Post-Soviet Democratization 13(1): 11–44.doi:10.3200/DEMO.13.1.11-44

[44] We included employment status on this equation as a series of dummies, but it had little effect on both dependent variables, so we do not present the results here.

[45] Rennó, L., Smith, A., Layton, M (2011) Legitimidade e Qualidade da Democracia no Brasil: Uma Visão da Cidadania. São Paulo: Intermeios; Nashville: LAPOP.

[46] Pappas, TS (2019) Populists in power. Journal of Democracy 30(2): 70–84.doi:10.1353/jod.2019.0026

.