¹ Brazilian Constitution, art 61, para 2: ‘The initiative of the people may be exercised by means of the presentation to the Chamber of Deputies of a bill of law subscribed by at least one percent of the national electorate, distributed throughout at least five states, with not less than three-tenths of one percent of the voters in each of them.’

² See e.g. Fernanda Castro, ‘Câmara retira seis propostas do MPF e desfigura pacote anticorrupção’ G1/Globo (30 November 2016) available at <http://g1.globo.com/politica/noticia/2016/11/camara-dos-deputados-conclui-votacao-de-medidas-contra-corrupcao.html>.

³ The Lava Jato (‘Car Wash’) operation started in 2014 as a money laundering investigation by the federal police. Since then, it has grown in size and scope, as it uncovered a cartel of construction firms that were granted public contracts, from the state-run oil company Petrobras, in exchange for bribes and campaign contributions to parties from the governing coalition.

⁴ See e.g. Tsebelis, G, Veto Players: How Political Institutions Work (Princeton University Press, Princeton, NJ, 2002).CrossRefGoogle Scholar

⁵ Our argument focuses on individual powers within constitutional courts or supreme courts. In ‘decentralised’ systems of judicial review, like in the US, Argentina and Brazil, a single lower court judge is empowered to perform judicial review. In these scenarios, however, it is still assumed that all such individual decisions will be eventually (and often immediately) subjected to oversight by higher, collective judicial bodies. In this sense, lower judges performing judicial review present fewer problems than the form of individual judicial review discussed in this piece.

⁶ This idea is present in many varieties of social science studies on courts. See e.g. A Stone Sweet, Governing with Judges: Constitutional Politics in Europe (Oxford University Press, Oxford, 2000); KE Whittington, ‘“Interpose Your Friendly Hand”: Political Supports for the Exercise of Judicial Review by the United States Supreme Court’ (2005) 99(4) American Political Science Review 583; Epstein, L and Knight, J, The Choices Justices Make (CQ Press, Washington DC, 1998);Google Scholar Taylor, M, Judging Policy: Courts and Policy Reform in Democratic Brazil (Stanford University Press, Stanford, CA, 2008).CrossRefGoogle Scholar In these approaches, the threat of future judicial intervention can be very specific in terms of the expected outcomes; see, in contrast, the broader idea of ‘bargaining under the shadow of the law’ as developed in R Mnookin and L Kornhauser, ‘Bargaining under the Shadow of the Law: The Case of Divorce’ (1979) 88(5) Yale Law Journal 950.

⁷ See L Epstein, J Knight and O Shvetsova, ‘The Role of Constitutional Courts in the Establishment and Maintenance of Democratic Systems of Government’ (2001) Law and Society Review 117; Fontana, D, ‘Docket Control and the Success of Constitutional Courts’ in Ginsburg, T and Dixon, R (eds), Comparative Constitutional Law (Edward Elgar, London, 2011).Google Scholar

⁸ We are assuming here that judges are strategic decision-makers – that is, their preferred position on a given case will be modulated (and therefore not sincerely expressed) – given the possibility of undesired reactions by other institutions against the court or the decision itself. In this sense, decision timing might determine what judges perceive as a feasible or prudent exercise of the powers at their disposal.

⁹ On agenda setting in the US Supreme Court, see e.g. Epstein, L, Landes, W and Posner, RA, ‘The Best for Last: The Timing of U.S. Supreme Court Decisions’ (2015) 64 Duke Law Journal 101Google Scholar; Benesh, SC, Brenner, S and Spaeth, HJ, ‘Aggressive Grants by Affirm-Minded Justices’ (2002) 30(3) American Politics 219CrossRefGoogle Scholar. For one analysis of the Brazilian case, see Werneck Arguelhes, D and Hartmann, IA, ‘Timing Control without Docket Control: How Individual Justices Shape the Brazilian Supreme Court’s Agenda’ (2017) 5(1) Journal of Law and Courts 105.CrossRefGoogle Scholar

¹⁰ Ferreres Comella, V, ‘The Consequences of Centralizing Constitutional Review in a Special Court: Some Thoughts on Judicial Activism’ (2004) 82 Texas Law Review 1705.Google Scholar

¹¹ See Fontana (n 7).

¹² See Arguelhes and Hartmann (n 9). The STF was granted limited means of formal docket control with the judicial reform amendment of 2004. In a specific kind of appeal (Recurso Extraordinário) that accounts for a large share of the court’s annual workload, two-thirds of the Justices can now decide that a given case lacks ‘general repercussion’ and should not, therefore, be decided by the Supreme Court. There was de facto docket control in the STF, however, before such a formal mechanism was created, and informal control is still applied through various other means.

¹³ MS 34193-16, MS 34371-16 and MS 34441-16. In addition to several alleged due process violations, Rousseff argued that the specific provisions of the ‘Impeachable Crimes’ law the Senate and the Chamber used to convict her were actually unconstitutional, and that the impeachment decision was wrong on the merits because her conduct did not amount to an impeachable offence according to the constitution.

¹⁴ On the relevance of case selection for compliance with judicial decisions and building judicial power in recent democracies, see Ginsburg, T, Judicial Review in New Democracies: Constitutional Courts in Asian Cases (Cambridge University Press, Cambridge, 2003) ch 3CrossRefGoogle Scholar.

¹⁵ See, for example, the Court’s silences on the privatisation of the telecommunications sector in the 1990s (MS 34562-16, still undecided), or on the Arbitration Act of 1996, which was only decided in 2001; or on several aspects of the ‘Plano Real’ stabilisation measures from 1995, which were decided only two decades after the country had made the change to a new currency. For more examples and general trends, see J Falcão, I Hartmann and VP Chaves, III Relatório Supremo em números: o Supremo e o tempo (FGV, Rio de Janeiro, 2014); VE Vieira and R Glezer (eds), A Razão e o Voto: Diálogos constitucionais com Luis Roberto Barroso (FGV, Rio de Janeiro, 2017).

¹⁶ Arguelhes and Hartmann (n 9).

¹⁷ Stone Sweet (n 6).

¹⁸ See Cayuso, SG and Gelli, MA, Ruptura de la Legitimidad Constitucional – La Acordada de la Corte Suprema de Justicia de la Nación de 1930 (Instituto de Investigaciones Jurídicas y Sociales Ambrosio L Gioja, Buenos Aires, 1988)Google Scholar.

¹⁹ See Davis, R, Justices and Journalists: The Supreme Court and the Media (Cambridge University Press, New York, NY, 2011)CrossRefGoogle Scholar ch 1, which discusses other kinds of motivations that can prompt courts to reach out to the public. The importance of the press as an arena in which judicial power can be promoted or constrained has been documented in several studies, in different ways. See e.g. Katon, J, Judicial Power and Strategic Communication in Mexico (Cambridge University Press, New York, NY, 2010)Google Scholar; R Davis and VJ Strickler, ‘The Invisible Dance: The Supreme Court and the Press’ (2000) 29(2) Perspectives on Political Science 85; Friedman, B, The Will of the People (Farrar, Straus and Giroux, New York, NY, 2009).Google Scholar In these studies, however, the focus is still on the court as a collective institution, and the press is usually seen primarily as an arena to strengthen the court’s legitimacy or to protect its decisions from retaliation or disobedience by other institutions.

²⁰ In the STF, for example, a decision that is provisional in theory often becomes final in practice. Between 1988 and 2013, preliminary injunctions (‘liminares’) stayed in place on average for more than two years before the court made a final ruling; in the main kind of abstract review procedure (Ação Direta de Inconstitucionalidade, ‘ADI’), the average duration was more than six years. When the researchers considered only the liminares that still had not been superseded by a final ruling, the overall average was more than six years – and more than 13 years for the ADIs. See Falcão, Hartmann and Chaves (n 15).

²¹ We will explore this in detail in the Brazilian case in section II: Designing a (de)centralised court.

²² See e.g. K Stack, ‘The Practice of Dissent in the Supreme Court’ (1996) 105(8) Yale Law Journal 2235; Afonso da Silva, V, ‘Deciding without Deliberating’ (2013) 11() International Journal of Constitutional Law 557CrossRefGoogle Scholar.

²³ Waldron, J, ‘The Core of the Case Against Judicial Review’ (2006) 115(6) Yale Law Journal 1346CrossRefGoogle Scholar.

²⁴ See Cohen, M, ‘Ex Ante Versus Ex Post Deliberations: Two Models of Judicial Deliberations in Courts of Last Resort’ (2014) 62(4) American Journal of Comparative Law 951CrossRefGoogle Scholar for a discussion of the role of the case reporter in different systems.

²⁵ See Cross, FB and Lindquist, SA, ‘The Scientific Study of Judicial Activism’ (2006) 91 Minnesota Law Review 1752Google Scholar.

²⁶ Spriggs, JF, Maltzman, F and Wahlbeck, PJ, ‘Bargaining on the US Supreme Court: Justices’ Responses to Majority Opinion Drafts’ (1999) 61(2) The Journal of Politics 485CrossRefGoogle Scholar.

²⁷ Cross and Lindquist (n 25): ‘The authority to speak first conveys an agenda-setting power that may be quite important because it enables the Chief to “direct discussion and frame alternatives”.’

²⁸ Ibid.

²⁹ In one recent instance in which a Justice has spoken out to discuss current affairs, for example, backlash and backtracking ensued. See MD Shear, ‘Ruth Bader Ginsburg expresses regret for criticizing Donald Trump’ The New York Times (14 July 2016).

³⁰ Toma has argued that politicians use this power to signal their broad approval/disapproval of Supreme Court decisions in a given year; the Chief Justice’s actions and words regarding the judicial budget would be another step in this signalling game. See Toma, EF, ‘A Contractual Model of the Voting Behavior of the Supreme Court: The Role of the Chief Justice’ (1996) 16(4) International Review of Law and Economics 433CrossRefGoogle Scholar.

³¹ See Edward White, G, ‘The Internal Powers of the Chief Justice: The Nineteenth Century Legacy’ (2006) 154(6) University of Pennsylvania Law Review 1463CrossRefGoogle Scholar. Contrast Ruger, TW, ‘The Chief Justice and the Institutional Judiciary: Foreword’ (2006) 154(6) University of Pennsylvania Law Review 1323Google Scholar. Ruger observes that, in recent decades, the office of the Chief Justice has acquired what he calls extramural powers that can be deployed independently from in-court majorities, such as ‘presiding over the important Judicial Conference, which helps set judicial policy, appoint[ing] key managerial personnel in the federal courts, and select[ing] the judges who sit on various specialized federal courts’. Ruger argues that, although administrative in character, these (individual) powers can have an impact on how federal courts decide cases and therefore should be subject to the same requirements as other judicial decisions (i.e., reason giving and collective deliberation/decision making).

³² Lane Scheppelle, K, ‘Guardians of the Constitution: Constitutional Court Presidents and the Struggle for the Rule of Law in Post-Soviet Europe’ (2006) 154(6) University of Pennsylvania Law Review 1757CrossRefGoogle Scholar. In the case of the German Constitutional Court, since the 1990s the President has engaged with the press to represent and defend the Court much more often than what would be the case in the US. See PE Quint, ‘Leading a Constitutional Court: Perspectives from the Federal Republic of Germany’ 154(6) University of Pennsylvania Law Review 1853: ‘President [Juta] Limbach (…) displayed a willingness to engage in extrajudicial discussion and explanation that went considerably beyond anything of the sort that has been seen in a Chief Justice of the United States – in recent times at least.’ However, such appearances would take place after the relevant decisions had been made, thus limiting their relevance as signalling mechanisms.

³³ According to Davis and Strickler (n 19) 90, ‘despite the lack of formal personal encounters (…) the justices shape press coverage by directing the press to their written work, by being selective in their public appearances, by providing background information, by shutting off other points of access, and by avoiding issues of contention, focusing instead on minor matters such as working conditions. (…) They usually refuse to discuss current cases, and reporters know better than to ask. But the justices do discuss their roles in past decisions and offer insight on the Court’s inner machinations.’ See, however, Davis (n 19) (arguing that US Supreme Court Justices are becoming less averse to appearing in the press in recent years).

³⁴ Arguelhes and Hartmann (n 9).

³⁵ See ibid (n 9) for a detailed analysis of this mechanism and the lack of limits in its use, including examples of the strategic use of vistas.

³⁶ Ibid. See also D Dimoulis and S Lunardi, ‘Definição da pauta no Supremo Tribunal Federal e (auto)criação do processo objetivo’ (Anais do XVII Congresso Nacional do CONPEDI, Brasília, 2008), 4.357–77.

³⁷ D Werneck Arguelhes and L Molhano Ribeiro, ‘O Supremo Individual: Mecanismos de atuação direta dos ministros sobre o processo político’ (2015) 46 Direito, Estado e Sociedade 121.

³⁸ These two Justices were Gilmar Mendes and Marco Aurélio, both of whom have made some of the most consequential public appearances in the last few years (see ibid). When President Dilma Rousseff first launched the idea of political reform via an exclusive constituent, STF Ministers spoke out on the issue. In particular, Ministers Roberto Barroso and Gilmar Mendes presented different views (cautious acknowledgement and frontal rejection, respectively) on the constitutionality of reforming the constitution through this type of mechanism.

³⁹ In a few occasions, Justices have even used this power to try to influence the behaviour of their own colleagues in pending cases. See Arguelhes and Ribeiro (n 37) for examples and discussion.

⁴⁰ Ibid; J Falcão and D Werneck Arguelhes, ‘O invisível Teori Zavascki e a fragmentação do Supremo – Uma retrospectiva de 2015’ in J Falcão, D Werneck Arguelhes, and F Recondo, O Supremo em 2015 (FGV, Rio de Janeiro, 2016) 21. The STF’s Chief Justice does have some exclusive, institutional opportunities to speak on behalf of the court – for example, the Chief Justice must speak at the opening of the judicial year, at the beginning of each semester. In practice, however, associate Justices access the press and the public so often that the potential value of the CJ’s exclusive opportunities to speak out are diluted.

⁴¹ For a comparison of data on the court’s caseload over time, contrast e.g. MP Veríssimo, ‘A Constituição de 1988, vinte anos depois: Suprema Corte et ativismo judicial “à Brasileira”’ (2008) Revista Direito GV 407; J Falcão, P de C Cerdeira and D Werneck Arguelhes, ‘1o Relatório Supremo em Números: O Múltiplo Supremo’ (FGV, Rio de Janeiro, 2011); and Hartmann, IA and Ferreira, L, ‘Ao relator, tudo: o impacto do aumento do poder do ministro relator no Supremo’ (2015) 13(17) Revista Opinião Jurídica 268CrossRefGoogle Scholar.

⁴² See Veríssimo (n 41).

⁴³ Data obtained from the STF’s official website: see <http://www.stf.jus.br/portal/cms/verTexto.asp?servico=estatistica&pagina=decmonocraticas>.

⁴⁴ Ibid. According to data from the Supremo em Números database at FGV Direito Rio, the same proportion can be found when we look just at abstract review cases. Between 2012 and 2016, there were 883 individual decisions on provisional injunctions – an average of 80 per Justice per year. Moreover, in the last decade, individual judges made more than 90 per cent of all provisional rulings on abstract review. See also Falcão, Hartmann and Chaves (n 15)

⁴⁵ Data obtained from the Supremo em Números database.

⁴⁶ This is the time between the issuing of the decision and the next collegiate decision in the case. We assume that the next collegiate decision will typically either be about the previous, provisional ruling, or be in itself a decision on the merits that ends the dispute and supersedes the individual ruling. That might not be the case, however, and even longer periods of time may have passed, in the cases included in the data, before the court reviews the individual ruling or decides the case.

⁴⁷ Dimoulis and Lunardi (n 36) have observed that, on several important issues, individual justices have adopted an activist stance in the last decade while the plenary court or the chambers remain passive or even completely silent.

⁴⁸ For extended discussion and examples, see J Falcão and D Werneck Arguelhes, ‘Onze Supremos, todos contra o Plenário’ in J Falcão, D Werneck Arguelhes and F Recondo (eds), Onze supremos: o supremo em 2016 (FGV, Rio de Janeiro, 2017) 20.

⁴⁹ T Pereira, ‘Lula Ministro e o Silêncio do Supremo’ in Falcão, Werneck Arguelhes and Recondo (n 48) 77.

⁵⁰ MS 34087. Justice Marco Aurélio’s injunction was issued on 5 April, and cleared for judgment by the plenary Court on 17 May. See L Scocuglia, ‘Marco Aurélio libera para julgamento pedido de impeachment de Temer’ JOTA (17 May 2017) available at <https://jota.info/justica/marco-aurelio-libera-para-julgamento-pedido-de-impeachment-de-temer-17052016>. At the time of writing, the Court has yet to rule on this individual injunction.

⁵¹ Since Fux made his decision within the AO n° 1.773, in 2014, each judge in Brazil has been receiving a de facto raise of approximately 4.300 BRL/month (1200 USD/month). According to one estimate, between the decision and November 2016, this has amounted to a total cost of 289 million BRL per year for the federal budget. See Redação JOTA, ‘Barroso libera para julgamento o fim do auxílio-moradia’ JOTA (14 November 2016) available at <https://jota.info/justica/barroso-libera-para-julgamento-o-fim-auxilio-moradia-para-juizes-14112016>. In December 2017, more than three years after his individual injunction, Justice Fux cleared the case for deliberation by the plenary court.

⁵² Silva (n 22); M Kumm, ‘Constitutional Courts and Legislatures: Institutional Terms of Engagement’ (2017) 1(1) Católica Law Review 55. See also Mendes, C, Constitutional Courts and Deliberative Democracy (Oxford University Press, Oxford, 2013).CrossRefGoogle Scholar

⁵³ Silva (n 22).

⁵⁴ Each judge is bound by ‘law as integrity’ individually, as she is ‘required to test his interpretation of any part of the great network of political structures and decisions of his community’. See Dworkin, R, Law’s Empire (Harvard University Press, Cambridge, MA, 1986) 245.Google Scholar

⁵⁵ See e.g. FI Michelman, ‘Foreword: Traces of Self-Government’ (1986) 100 Harvard Law Review 4.

⁵⁶ For a discussion of some of these prudential reasons, see Mendes (n 52).

⁵⁷ Cohen (n 24) 958–63 (emphasis added).

⁵⁸ See ibid, 958, fn 19 for examples.

⁵⁹ Silva (n 22); Mendes (n 52); Mendes, C, ‘Political Deliberation and Constitutional Review’ in Himma, K and Flores, I (eds), Law, Liberty and the Rule of Law (1st edn, Springer, Nova Iorque, 2013) vol 1 121Google Scholar.

⁶⁰ Mendes (n 52) 72: ‘Collegiality is a fact of constitutional courts, and constitutional theory needs to take that into account.’ The dimension in which courts really differ, in terms of their deliberative potential, is that ‘as any multi-member institution, a collegiate court needs to frame a procedure that allows for the conversion of the ‘‘many’’ into ‘‘one’’, to define what shape ‘‘the opinion of the court’’ will have’ (ibid 62).

⁶¹ According to Mendes, we should not assume that, because judges are unelected, the ‘internal dynamics of this conflictive multi-member institution’ will necessarily foster deliberation. See ibid.

⁶² See e.g. Freeman, S, ‘Constitutional Democracy and the Legitimacy of Judicial Review’ (1990) 9(4) Law and Philosophy 327, 364–5CrossRefGoogle Scholar.

⁶³ Fallon, RH Jr., ‘The Core of an Uneasy Case for Judicial Review’ (2008) 121(7) Harvard Law Review 1693 (original emphasis)Google Scholar.

⁶⁴ Fallon recognises this risk of proving too much with this line of reasoning. See ibid: ‘If multiple vetoes are good, why stop with the legislature, the President, and the courts? Why not establish other institutions with veto powers or insist on unanimous consent before any legislation can be enacted?’

⁶⁵ Waldron, J, ‘Five to Four: Why Do Bare Majorities Rule on Courts?’ (2014) 123(6) Yale Law Journal 1692Google Scholar.

⁶⁶ Ibid, 1710–11.

⁶⁷ We are not referring here to the mere fact that a randomly assigned case reporter has some advantages in the decision-making procedure. Many courts rely on individual reporters to draft a tentative first draft of the opinion: see generally Cohen (n 24). Depending on other factors like the court’s workload and the reporter’s reputation, the reporter’s initial assessment of the case might actually influence the outcome. This, however, would be one amongst many other factors influencing the collective decision, while the mechanisms we have discussed in the STF make the randomly assigned reporter’s preferences enough to determine the outcome. The problem arises not with randomness of the reporter per se (even if this position carries some weight in the decisional procedure), but with randomness combined with the reporter’s agenda-setting and decision-making powers.

⁶⁸ Waldron (n 65) 1712–18.

⁶⁹ Ibid, 1718–23.

⁷⁰ See e.g. Balkin, J and Levinson, S, ‘Understanding the Constitutional Revolution’ (2001) 87(6) Virginia Law Review 1045CrossRefGoogle Scholar. Fixed term limits can be conceived as a more specific attempt to treat all political coalitions and presidents alike, by making all electoral victories give them the exact same number of opportunities in influencing a court’s composition.

⁷¹ In considering, in a sympathetic light, the possibility of requiring a super-majority (such as two-thirds of the court) for performing judicial review, Waldron is indirectly discussing minority-voting rules. See Waldron (n 65) 1696–97 and 1730. Such an arrangement would enshrine, in practice, a deferential attitude towards legislative and administrative acts – not in interpretive approaches and jurisprudential positions, but built into the voting procedure itself. See Vermeule, A, Mechanisms of Democracy: Institutional Design Writ Small (Oxford University Press, New York, NY, 2007) 73–85CrossRefGoogle Scholar. But, as happens with all super-majority rules, this would also mean that a minority of justices has the power to prevent the court from using its power.

⁷² Mendes (n 52); L Kornhauser, ‘Deciding Together’ (2015) 1 Revista Estudos Institucionais 38.

⁷³ See e.g. Mendes (n 52); Michelman (n 55); RM Cover, ‘Violence and the Word’ (1986) 95(8) Yale Law Journal 1601; Kornhauser (n 72) (‘Normative theories of adjudication typically consider a single judge, deciding alone. Appellate judges in every country, however, sit in panels. They decide together. These theories thus implicitly suggest that a judge, sitting on a collegial court, should decide as she would decide were she sitting alone.’)

⁷⁴ The expression comes from Mendes’s reconstruction of legal reasoning practices in the STF. See Mendes, CH, ‘The Supreme Federal Court of Brazil’ in Jakab, A, Dyevre, A and Itzcovich, G (eds), Comparative Constitutional Reasoning (Cambridge University Press, New York, NY, 2017) 115CrossRefGoogle Scholar. On individualism and the limited scope of the exchanges between the Justices (and between their opinions), see also Silva (n 22); J Rodrigo Rodriguez, Como Decidem as Cortes? Para uma crítica do direito (brasileiro) (FGV, Rio de Janeiro, 2013); A Vojvodic, AMF Machado and EC Cardoso ‘Escrevendo um Romance, Primeiro Capítulo: Processo decisório e precedentes no Supremo Tribunal Federal’ (2009) 5(1) Revista Direito GV 21.