Blog of the International Journal of Constitutional Law

The Return of Lula in Brazil: New Challenges for Comparative Presidential Studies

Juliano Zaiden Benvindo, University of Brasília and National Council for Scientific and Technological Development

[Editors’ Note: This is one of our biweekly ICONnect columns. For more information on our four columnists for 2021, please see here.]

A recent column published in The Economist titled The Problem of Latin America’s Proxy Presidents raises the argument that, as “a result of term limits and partly a consequence of the commodity boom of the 2000s,” there has been a proliferation of “proxy” presidents in the region. Examples, according to that magazine, are Luis Arce as Evo Morales’ proxy in Bolivia, Iván Duque as Álvaro Uribe’s proxy in Colombia, Alberto Fernández as Cristina Fernández de Kirchner’s proxy in Argentina, and Dilma Rousseff as Lula da Silva’s proxy in Brazil. In Ecuador, Andrés Arauz would be Rafael Correa’s proxy, but he ended up losing the elections last April. This scenario is depicted as problematic because “a proxy risks being a weak president, carrying the can for decisions inspired by a sponsor who exercises power without responsibility.” Another interesting occurrence in Latin America are presidents who leave office but run again for presidency once the opportunity arises. Chile is a great example: Michelle Bachelet (2006-2010; 2014-2018) and Sebastián Piñera (2010-2014; 2018-2022) have switched their positions quite synchronically, reflecting a stable electoral division in that country between the center-left and center-right political spectrums. Brazil, where its rather stable coalitional presidentialism for over twenty-five years was disrupted by the victory of Jair Bolsonaro in 2018,[1] may elect Lula da Silva (henceforth Lula), who governed the country for two terms (2003-2006; 2007-2010), in the next 2022 presidential elections. Brazil could thus experience both movements in its recent history: a “proxy” president represented by Dilma Rousseff, who was impeached in 2016 on very flimsy legal grounds, and the return of her “sponsor” Lula for another term in office. What does the return of a “sponsor” president mean for comparative presidential studies?

The Economist’s column endorses a common but controversial argument in presidential studies. The dichotomy between “proxy” and “sponsor” presidents may sound plausible, but under what circumstances does such a relationship actually differ from normal politics? The analytical frontier here might be quite slim. Presidents may be stronger or weaker than the previous one because of a series of complex variables – either subjective (leadership, charisma, intelligence, etc.) or objective (congressional support, institutional dialogue, economy, international environment, etc.). Presidents may also have interest in returning later to office or preserving their policies for long by keeping an eye on their successors, but whether this means a personal or an institutional interest – for example, of their party – seems an artificial divide. Naturally, personality traits matter and should be interpreted as fundamental “independent variables” for presidential studies,[2] but it would be misleading to downplay the whole institutional settling and other environmental elements in politics. Moreover, even if we hypothetically considered that former presidents sponsor in some ways the new president exclusively for his or her personal gain and “not necessarily those of the country,” it would not represent in itself an atypical political behavior.

Game theoretic models, which are largely adopted in political science, would fail if they needed to count on the best intentions of politicians or if they disregarded their personal interests. More difficult still, behavioral analyses in the political realm have to deal with the “collective nature of politics.”[3] Paul Pierson, in his groundbreaking Politics in Time: History, Institutions and Social Analysis, contends that “a crucial feature of most collective action in politics is the absence of a linear relationship between effort and effect,” which means, in other words, that “in politics, the consequences of my actions are highly dependent upon the actions of others. What I get depends not just on what I do, but (mostly) on what others do.”[4] Therefore, even though it seems at first sight plausible to draw a linear connection between a “proxy” and a “sponsor” based on some historical events in Latin America, some such conclusions – for instance, the direct association of a “proxy” with weakness – seems to overlook the warps and wefts of normal politics.

Despite that, that piece reminds us how Latin American countries have historically struggled to stabilize their presidential systems and may be still under continuous risk of seeing such an equilibrium collapse. One of the attempts to overcome past instabilities appeared in the wave of new constitutions in the region during the 1980s and 1990s, which raised barriers and constraints for re-election. This strategy did not last long, though. Argentina (1994), Brazil (1997), Colombia (2005), Ecuador (2008), and Bolivia (2009), for example, changed their rules, previously allowing uniquely a re-election after sitting out one term – or no re-election in Colombia – to accept immediate re-election, but limited to a consecutive two-term limit. Proposals to go even further in such relaxations were successful in Peru (2000) and Venezuela (2009), where the rules allowing for the immediate re-elections with a consecutive two-term limit were altered to accept multiple re-elections, or, even more radically, as it occurred in Honduras (2015), from banning re-election to no term limits at all.[5] The region has also featured a number of charismatic and powerful leaders who have taken advantage of their personality traits to promote institutional transformations for their benefit, such as the relaxation of term limits and strengthening of presidential powers.[6] The history of charismatic leaders in Latin America, and their association with instabilities, is well documented in the literature.[7] It is thus logical and even expected that tying the knots of instability with the new model of presidential constraints, even if relaxed, would appear as a powerful argument: if such leaders cannot extend their time in power, “proxy” presidents are the way out to keep their influence in politics and to pave the way for their return.

However, if the rules of the game are followed, would such a move be illegitimate or undemocratic? Comparative presidential studies have dealt with various dimensions of this discussion in very fascinating ways. The leading premise is that government turnover[8] and political competition[9] are positive for democratic stabilization and the rule of law, though such a conclusion may be less straightforward if the country features independent oversight institutions,[10] and other markers such as institutional fragmentation, level of ideological alternation, and period of alternation are taken into consideration.[11]Moreover, presidential systems may better stabilize themselves if, instead of adopting a permanent ban on reelection, they adopt a “weaker ban on reelection for consecutive terms”, such as setting a consecutive two-term limit or even accepting an “eventually reelected [president] to a non-consecutive term.”[12] Based on such a premise, it is a positive development that Latin America has reached a level of discussion according to which, if relaxation may be permitted in some circumstances, it would be detrimental to the rule of law in others, as the Colombian Constitutional Court said in adopting the “substitution doctrine” to set limits in the famous Uribe case.[13] Despite that, how much of a stretch would it be to extend this premise to the narrative of “proxy” and “sponsor” presidents? Not that the The Economist developed this argument directly, but the idea is all there: “sponsor” presidents behave as strategists of their own interests, which are “not necessarily those of the country,” and will do whatever it takes to keep their political clout alive, even using a “proxy” president to “keep the presidential seat warm” for them. In addition to downplaying the warps and wefts of politics, this argument seems to suffer from methodological challenges of comparing realities that are not easily translated into a generalized Latin American pattern.[14]

The Brazilian case of the “sponsor” Lula reveals how intricate these translations may be. Lula left office in 2010 with an approval rating of 87%. His government thus featured the “best predictor of success”[15] to push for change in presidential term limits. Yet, Lula himself did not push for a constitutional amendment authorizing him to run for a third consecutive term. Congress, however, drafted proposals with this goal, even because they could reach governors and mayors. What is particularly striking is that the proposal that went the farthest was buried by Lula’s Worker’s Party (PT). Representative José Genoíno (PT), the rapporteur, claimed that a second consecutive re-election would violate “the implicit unamendable clause” that is found in periodical elections and the republican principle laid down in the Constitution. This scenario demonstrates that the very “collective nature of politics” may lead to other movements that point towards a more “institutional” presidency,[16] even though Lula’s charismatic personality would not disappear and would continuously play a big role in the following elections. President Dilma Rousseff (PT), who succeeded him, was certainly weaker, but her government was very popular in most of her first term (reaching 63% approval in March 2013) and would only see a slump in popularity after the protests of June 2013[17] and, more drastically, right after her reelection in 2014, when the so-called CarWash Operation”, a massive corruption probe that also implicated PT, intensified.

The years have passed, and Brazil has since experienced the impeachment of President Dilma Rousseff in 2016, the imprisonment of former President Lula in 2018 for 580 days on corruption charges related to the CarWash Operation, and the election of the far-right President Jair Bolsonaro in 2018. The CarWash Operation was a cataclysm in all areas, but it was particularly disruptive for the political system to the point of paving the way for the election of Jair Bolsonaro as President, who would not possibly be elected otherwise. It was also highly disruptive for the system of justice. In a series of decisions this March and April,[18] the Brazilian Supreme Court finally ruled that the judge presiding over that operation, Sergio Moro, had no authority whatsoever over the cases involving former President Lula, but, more seriously, that he behaved with partiality towards him. In an impressive decision that should serve as a guidance in comparative law for other cases of lawfare, Justice Gilmar Mendes, the rapporteur, pointed out, with details, how Judge Moro colluded with federal prosecutors and violated the most basic constitutional guarantees of due process of law for personal and political gains. One result of this problematic process was that Lula could not run again for president when he was leading the polls in the 2018 elections. Sergio Moro, in turn, became Bolsonaro’s Minister of Justice. In broad terms, that decision concluded that, if it is to combat corruption, the very system of justice cannot itself be corrupted so profoundly.

According to the latest polls, Lula would beat Bolsonaro in a second round for the presidency by 52 to 34 percent, an impressive figure. There are many months ahead until October 2022 and the scenario can rapidly change. There is also no doubt that President Bolsonaro will do whatever he can, and cannot, do to stay in power. Still, the likelihood that Lula, a pragmatic center-left politician, can become President again recalls those fantastic narratives that only a country where “even the past is uncertain”[19] can tell. It also challenges common wisdom and even traditional preconceptions that are continuously repeated in comparative presidential studies. Considering the political disruption that led to President Bolsonaro, the democratic erosion that he has since promoted in the country, and the redemption narrative that comes with the rise of Lula after having been victim of a lawfare, it seems that traditional tools for presidential studies might not be enough. The scenario goes far beyond the dichotomy between “sponsor” and “proxy” presidents or whether a third term – even if not a consecutive one – is or is not detrimental to democracy.[20] There was, after all, governmental turnover – and a radical one – and a charismatic leader like Lula might be needed, as he may prove the only one who could gather enough support from distinct democratic fronts to beat Bolsonaro and push for change.[21] It is still too early to call, but it looks like Brazil can once again become an important reference for new agendas of comparative presidential studies.

Suggested citation: Juliano Zaiden Benvindo, The Return of Lula in Brazil: New Challenges for Comparative Presidential Studies, Int’l J. Const. L. Blog, Apr. 28, 2021, at:

[1] See Sergio Abranches, et al., Democracia em Risco? 22 Ensaios sobre o Brasil Hoje (Companhia das Letras 2019)

[2] See Ignacio Arana Araya, ‘The Personalities of Presidents as Independent Variables’ (2020) Political Psychology

[3] Paul Pierson Politics in Time: History, Institutions, and Social Analysis, (Princeton University Press 2004) 32

[4]  Ibid. 32

[5] See Javier Corrales, ‘Can Anyone Stop the President? Power Asymmetries and Term Limits in Latin America, 1984–2016’ (2016) 58 Latin American Politics and Society 3, 8

[6] See Araya supra 2, 14 (showing how “treating the personalities of presidents as independent variables allows us to answer pressing questions”… such as “can the traits of powerful leaders explain institutional transformations such as the relaxation of term limits that, in turn, change the constraints that presidents face?…)

[7] See Carlos De la Torre, ‘The Resurgence of Radical Populism in Latin America’ (2007) 14 Constellations 384, 384-97; Carlos de la Torre, ‘Populism and the Politics of the Extraordinary in Latin America’ (2016) 21 Journal of Political Ideologies 121, 121-39; Carlos De la Torre, ‘Populism in Latin America’ (2017) The Oxford handbook of populism 195, 195; Michael L Conniff, et al. Populism in Latin America, (The University of Alabama Press 2012)

[8] See Samuel P Huntington The Third Wave: Democratization in the Late Twentieth Century, (University of Oklahoma press 1993)

[9] See Joseph Wright, ‘Political Competition and Democratic Stability in New Democracies’ (2008) 38 British Journal of Political Science 221, 221-45.

[10] See Shale; Hoff Horowitz, Karla; Milanovic, Branko, ‘Guvernment Turnover: Concepts, Measures and Applications’ (2009) 48 European Journal of Political Research 107, 107-29.

[11] See Ibid. 111

[12] Rosalind Dixon and David Landau, ‘Constitutional End Games: Making Presidential Term Limits Stick’ (2019) 71 Hastings L.J. 359, 417.

[13] Colombian Constitutional Court, Decisión C-141 of 2010. See Manuel Jose Cepeda Espinosa and David Landau Colombian Constitutional Law, (Oxford University Press 2017) 352-360; Yaniv Roznai Unconstitutional Constitutional Amendments: The Limits of Amendment Powers, (Oxford University Press 2017) 65-68; Carlos Bernal, ‘Unconstitutional constitutional amendments in the case study of Colombia: An analysis of the justification and meaning of the constitutional replacement doctrine’ (2013) 11 International Journal of Constitutional Law 339, 339-57.

[14] See Juliano Zaiden Benvindo, Carlos Bernal, and Richard Albert, ’Introduction: Facts and Fictions in Latin America Constitutionalism’ in Richard Albert, Carlos Bernal, and Juliano Zaiden Benvindo (ed.), Constitutional Change and Transformation in Latin America (Hart Publishing 2019) 1-18.

[15] Corrales, supra 5, 3.

[16] Araya, supra 2, 4.

[17] See Juliano Zaiden Benvindo, Brazilian Elections and Demonstrations of June 2013: The Rise of Conservatism?, Int’l J. Const. L. Blog, Nov. 1, 2014, available at:; Juliano Zaiden Benvindo, The Mass Protests of March and April 2015 in Brazil: A Continuation of June 2013?, Int’l J. Const. L. Blog, Apr. 29, 2015, at:

[18] STF, Habeas Corpus (HC) 193726, HC 164493

[19] This saying “in Brazil, even the past is uncertain” is normally credited either to Pedro Malan, a former Minister of Economy, or to Gustavo Loyola, a former President of Brazil’s Central Bank.

[20] See Dixon and Landau, supra 12.

[21] See Araya, supra 2.


Leave a Reply

Your email address will not be published. Required fields are marked *