Beyond decision making: What kinds of powers do courts have?

Judicial decisions tend to monopolise scholarly attention as the main, prototypical expression of judicial power. The power to decide a controversy is indeed the most visible manifestation of power by judges. However, decision making provides us with an incomplete account of judicial power, in two ways. First, judges can influence the behaviour of other institutions and private parties without making an actual decision. The threat of a judicial decision in the near future can already shape the strategies of actors outside the court.^{Footnote 6} Any indication that the threat of interference is more or less likely can therefore, in principle, affect their behaviour. Second, within a given court, there are institutional steps that need to be taken so as to prime the case for a decision. The power to decide must therefore be enabled by other internal mechanisms before it can be used. If these mechanisms are instead employed to disable the possibility of a decision, the court’s power is not to be feared.

A full picture of the configuration of judicial power thus requires us to account for all mechanisms that can emulate the power to decide cases in terms of its effects on the world outside the court, or that can modulate the effects of a decision if and when this decision-making power is used. In this section, we discuss three such powers, by means of which Justices within collegiate judicial bodies can influence the legal status quo and the behaviour of actors outside the court, either directly or indirectly: (i) agenda setting (deciding what the court will decide); (ii) position taking (speaking on behalf of the court, thus signalling a potential judicial decision in specific directions); and (iii) decision making (resolving cases, controversies, and matters of dispute brought to the court).

Agenda setting involves enabling the court (or a fraction of its members) to decide a case or issue – or, conversely, preventing the occasion for such a decision. Selecting the agenda is a consequential aspect of judicial power. Courts trying to establish a politically relevant role for themselves need to decide the right cases at the right time.^{Footnote 7} Even for well-established courts, timing can shape the very substance of the decision. Timing may affect the likelihood that losing parties will retaliate against the court or its specific decision (for example, by passing a constitutional amendment).^{Footnote 8} Moreover, depending on the appointment mechanisms in place, the passage of time itself can change the composition of the court, allowing a different set of judges to make the decision and therefore changing the potential outcome of the case.^{Footnote 9}

But how can courts actually control the timing of their decisions? Many different mechanisms and design features influence agenda setting for this purpose. Comella, for example, distinguishes between two broadly considered ‘models’ of apex courts, each with its own clusters of agenda-setting mechanisms and implications for overall timing control. On the one hand, there are ‘Supreme Courts’, which typically have very limited mandatory jurisdiction, expansive docket control mechanisms, strict standing rules, and limited access to the court by litigants. In contrast, ‘Constitutional Courts’ lack formal docket control mechanisms, have expansive mandatory jurisdiction and tend to be very open to potential litigants and inconvenient or untimely cases.^{Footnote 10} Fontana, however, sees in discretionary docket control the ultimate mechanism for timing control. When a court simply refuses to hear a case, this decision will draw less public attention and make criticism by the losing party less likely, as losers in the status quo will have a harder time blaming their defeat on the Court.^{Footnote 11}

However, judges can reap these same institutional benefits in the absence of formal discretionary docket control, even with expansive standing rules and mandatory jurisdiction, and without resorting to any indirect strategies of avoidance. The STF, for example, has developed mechanisms for remaining completely silent on any given case or controversy, independent from any formal power to decide what to decide.^{Footnote 12} It does so by not including a case in the agenda in the first place, and, when it has already started to decide a case that suddenly becomes ‘inconvenient’, it can remove it from the agenda (even if not from the docket) and keep it on hold indefinitely. These lasting silences are achieved by a combination of positive (the power to include something on the agenda) and negative (the power to veto the inclusion of a case or issue in the agenda) agenda-setting powers within the court. Although the use of such powers in itself is not a judicial decision, it can have lasting effects on the status quo, in the same fashion as a decision would, but with much less exposure for the Court.

Prolonged judicial silence can be used to create social and political faits accomplis, enabling certain strategies by actors outside the court, in the short run, and making it less likely, in the long run, that a judicial decision would force a reversion of the status quo. Consider, in the Brazilian case, the several lawsuits filed by President Dilma Rousseff before, during, and in the wake of her impeachment trial by the Senate. In these cases, Rousseff challenged the constitutionality of the conviction on many grounds.^{Footnote 13} President Rousseff was officially removed from office on 31 August 2016 but, as we write, the STF is still silent on these lawsuits. In the meantime, we have had municipal elections all over the country, a couple of watershed constitutional amendments have been approved by the Temer government, and a new STF Court justice has been appointed – yet, the STF itself has not said anything about Rousseff’s claims. A judicial overruling of the Senate conviction was unlikely from the outset; after all this time, it has become a practical impossibility.

Regardless of the merits of Rousseff’s claims, as time passes and the political environment settles around her impeachment, it becomes more and more costly for political actors to obey a judicial decision reinstating her in office.^{Footnote 14} Over time, therefore, judicial silence has skewed the outcome of a future ruling; if the case is decided at all, the STF will most likely dismiss it on a technicality (e.g., the new elections in 2018 will have made the case moot). This is a similar strategy employed by the court when dealing with other politically sensitive cases that could have been a political lose-lose situation for the court. If it decides against the current government, there is a high likelihood that its decision will invite retaliation; otherwise, however, it will legitimise a politically controversial measure it might actually consider unconstitutional. Why decide, then? Regardless of a Justice’s position on the merits, it is tempting to simply act as if the court’s trigger was never triggered at all. Examples of agenda setting and judicial silence being used to deflect inconvenient cases like this abound in Brazil.^{Footnote 15}

When judges are not under an obligation to provide some answer to a given case within a certain period of time, the distinction between different timing control mechanisms, including docket control, becomes less important.^{Footnote 16} Their time is theirs to control, and, while the refusal to rule on a case is still a decision that might draw attention and therefore leave the court exposed, it is comparably harder to keep track of the judicial power to simply remain silent. But, whatever the precise mechanisms by which the court’s agenda is set, they can all be used in ways that modulate the final decision of the court, or prevent it from taking place and thus keeping the status quo undisturbed. These agenda-setting powers can work, in the world outside the court, as functional equivalents to the actual power to decide the case.

We turn now to position taking. Courts can influence the outside world by taking a public stance that signals, to different actors, a potential favourable or unfavourable decision in the future. If we assume that courts operate in a strategic environment, then the behaviour of different institutional actors is influenced by what they anticipate as the future behaviour of the other actors (judicial or otherwise) with whom they now engage. Studies of the judicialisation of politics show political actors adjusting their behaviour to the fact that their decisions today might be challenged before the constitutional court tomorrow.^{Footnote 17} Future threats, actual or imagined, can therefore affect behaviour today. It follows from this assumption, in principle, that any mechanism by which a court can signal what its decision would be on a given issue can influence behaviour by other actors around it. In particular, by taking a public stance on an issue before it is challenged in a lawsuit, the court provides information for actors outside it of the possible legal consequences attached to their planned lines of action.

In its clearest manifestation, such a position-taking power could take the form of an advisory opinion issued by the court on an issue being debated in Congress at that very moment, or a contemporary political event – like the Suprema Corte de Justicia de la Nación did in Argentina in 1930, in the infamous ‘acordada’ by which, in the absence of any actual case or controversy triggering its jurisdiction, the Justices issued a formal document acknowledging and normalising the rupture of the constitutional order.^{Footnote 18} But similar effects can be achieved in more subtle ways. A comment by a Justice in a lecture or interview might be read as signalling sympathy for a given cause or interpretation of the constitution, thus encouraging or discouraging new lawsuits; an official speech by the Chief Justice in the opening of the judicial year, or a side remark made to journalists after a court session, might signal to politicians that a certain bill will face at least some judicial resistance if approved.^{Footnote 19} In all these examples of position taking, the court or its members ‘borrow’ the authority and the threat of future decisions to provide incentives, in the present, for the behaviour of other actors.

Lastly, we find decision-making power itself, which we most immediately associate with judicial power. In looking at how courts operate, it should be disaggregated in two smaller components: the power to issue preliminary decisions might be assigned in different ways, and according to different criteria, than the power to issue final decisions. The importance of this distinction in practice, however, must not be overstated. The decision mentioned in the Introduction of this article, for example (the suspension of congressional voting on the anti-corruption package), was formally a mere preliminary injunction, not a decision on merits.^{Footnote 20} The relevance of the formal distinction between preliminary injunctions and final decisions depends on how often (and for how long) preliminary injunctions can be allowed to remain in place so as to shape the status quo.^{Footnote 21}

In terms of design and operation, courts can greatly differ from each other regarding the scope and limits of the powers discussed above. Moreover, these powers can be limited in many ways – by law and legal doctrine, by the behaviour of other institutions, and by the social conventions and norms of professional conduct. But, whatever their limits in a given design, their possible existence should encourage us to look beyond decision making when we look at (and compare) how courts are designed to influence the outside world.

Who holds power within the Court?

How can these different powers be assigned within courts? To begin with, talking about a ‘court’ as single entity can be misleading. Dissents and concurrences remind us that the court is a sum of individuals with their own views on how cases should be decided, and whose opinions might not necessarily coincide with those of the court.^{Footnote 22} Multiple, conflicting views exist within the institution, and the decision of the court is obtained either through persuasion, deliberation and bargaining, or, when all else fails, through counting votes. We hope that the best (not the most popular) argument will prevail, but even so, as Waldron pointed out, a divided court must resort to the same mechanism we see in majoritarian, legislative institutions: the side with the most votes wins.^{Footnote 23}

But even these views on vote counting and dissents still assume that the final decision of the court – the decision that will directly affect the world outside – will be the product of some sort of collective decision-making procedure, by which the views of each individual judge will be considered, filtered, and finally subjected to a mechanism that ensures that the majority position will prevail. This assumption is still problematic. As a matter of institutional design, it is possible to confer powers of agenda setting, position taking, and even decision making to a sub-majority of judges, or even to individual judges within the court.

We begin by distinguishing between collective and individual allocations. In the former, some sort of collective decision-making procedure (requiring either a majority or a sub-majority of the votes in the court) must be adopted for a given power to be used; in the latter, the individual will of a single judge in the court is a sufficient condition for deploying it. Individual allocations, however, come in very different forms. Centralised individual arrangements allocate the power to a specific, fixed position/role within the court that is endowed with powers that are not available to regular Justices qua Justices. The typical example is the office of Chief Justice (CJ) – for this position to mean something in practice, there should be at least some issues its occupant decides alone. In decentralised individual arrangements, in contrast, a given power is equally assigned, in principle, to each individual justice, regardless of any specific institutional position defined beforehand.

Some courts adopt the figure of a ‘case reporter’, who might have some specific powers, over a specific case, which are not shared by her fellow judges.^{Footnote 24} In our classification, this would still count as an individual decentralised arrangement, provided that any judge in the court can, in principle, be assigned the role of reporter (that is, that the office of reporter is not a permanent position in the court’s structure). Still, within this category, we can consider that powers granted to the reporter represent a less centralised allocation than powers granted to all justices not just in principle, but also in the sense that any judge can exercise them at a given time.

We have, therefore, a spectrum, from the more collective/centralised to the more individual/decentralised: a majority of Justices; a sub-majority of Justices; the Chief Justice; each case reporter or equivalent position in the court; each and every Justice in the court. Different combinations of powers and modes of assignment will lead to overall more or less centralised courts. The US Supreme Court, for example, is more of a centralised court. The most relevant agenda-setting power is the decision to grant certiorari, which can be made collectively, by the vote of a sub-majority of four Justices. However, when voting on which cases should be accepted, Justices use a list of relevant options that is created by the Chief Justice. Although they can add cases if they want, the large volume of cases and the scarcity of time can make them deferential to the CJ’s initiative in this ‘deciding to decide’ stage.^{Footnote 25}

The Chief Justice or the most senior Justice in any given voting group chooses who will draft the opinion for that group. Their choice will not prevail, however, if the tentative drafter is unable to write an opinion that can command the support of the other members of the coalition.^{Footnote 26} Additionally, the Chief Justice is the first to speak during deliberations, an agenda-setting power that, although not comparable to actually deciding what to decide, can influence how cases will be decided.^{Footnote 27} Provisional injunctions (for example, for stays of execution) are exceptional. Even when possible, they require a majority of Justices. Overall, decision-making power is firmly dependent on the majority. Although skilful Chief Justices can control the debates to try and promote certain outcomes, the office carries little authority over the rest of Justices when it comes to actual voting.^{Footnote 28}

Finally, position-taking powers seem to be very limited in the U.S. Supreme Court in general. Justices do not typically give interviews; even when they lecture, they typically talk about ‘bland’ topics.^{Footnote 29} The Chief Justice does have a few position-taking tools at her disposal if she wants to send such signals. The annual reports on the federal judiciary, for example, can function as position taking on issues bearing on the administration, organisation, and funding of federal courts, which are a matter of congressional decision.^{Footnote 30} But such powers are still very limited in comparative terms.^{Footnote 31} Consider, for example, how the Presidents of the Russian and Hungarian Constitutional Courts frequently and directly accessed the media and the public in the 1990s to defend their respective Court’s power, sometimes even going so far as to take a stance on current political controversies.^{Footnote 32} In any event, to the extent that position-taking powers exist and are relevant in the U.S. Supreme Court’s design, the CJ seems to have a more significant share of it, especially considering that associate justices still rarely discuss current affairs in the press.^{Footnote 33}

Designing a (de)centralised court: The case of Brazil

The case of the STF stands in stark contrast to the examples sketched above. First, agenda-setting powers are fragmented and distributed on many different levels. A case is only sent to the Court for a decision after the case reporter says it is ready and after the Chief Justice includes it in the agenda. Even after the court begins to decide the case, however, in practice each individual justice holds a veto power that can be activated at any time: she can request for the judgment to be halted, and the case files to be sent to her chamber for further study (‘pedido de vista’, or simply ‘vista’). Although the court’s rules of procedure limit such ‘vistas’ to a few weeks, they last more than one year on average, with some lasting for a decade or more. There is no institutional mechanism to force the case to be sent back for judgment.^{Footnote 34}

Unsurprisingly, Justices have used this mechanism to indefinitely stall a final judgment in several cases, even when a majority of votes has already been formed.^{Footnote 35} For a given decision to happen, therefore, (i) the case reporter and (ii) the Chief Justice must positively select a case, and (iii) the remaining Justices must refrain from using their veto-like power. If any individual Justice objects to the Court deciding the case, she can delay a decision for years or even indefinitely, since the STF and its members have shown not to be bound, in practice, by any deadlines.^{Footnote 36} Moreover, given the STF’s vast backlog, it is very costly for the press, for the parties involved, and for the other Justices to monitor which cases are being delayed and for what purposes.

Second, although this would seem to be forbidden by existing rules, in practice STF Justices talk to the press or the public about cases pending for a decision by the court.^{Footnote 37} In 2013, for example, in the days after some congressional committees began to discuss a political amendment that would significantly limit the STF’s powers, four Supreme Court Justices spoke to the press about the amendment. Of the three that criticised it, two actually stated that such an amendment would be unconstitutional.^{Footnote 38} Shortly afterwards, Congress stopped discussing the amendment. Position taking by the three Justices was not the only factor that led to this outcome, but such public statements undoubtedly provided congressmen with information on the likelihood, direction, and intensity of future judicial reactions, while also providing incentives (and perhaps specific legal arguments) for actors considering taking their political defeats to court.^{Footnote 39} While the actual effectiveness of this kind of signalling depends on many variables, including the standing of the speaking Justice, there seems to be no mechanism to punish or control Justices who talk to the press about pending and future cases. Just as it happens with the agenda-setting powers, then, the use of position-taking powers is entirely a matter of individual inclination and/or strategy.^{Footnote 40}

Third, in practice if not in theory, individual justices in the STF can perform judicial review by themselves. Although formally decision-making powers are allocated to the majority of the Justices, in the plenary court or in one of the chambers, there are in practice several mechanisms by which individual Justices can review and suspend the actions of the other branches without persuading a majority of their colleagues. As a formal matter, several different laws grant the case reporter – who is randomly assigned to the case as it enters the docket – the power to make individual decisions in exceptional circumstances, without necessarily giving the merits of the case a full analysis. These circumstances typically include preventing irreparable harm to rights, or irreversible consequences that would make a future decision moot. In practice, however, individual decisions are often neither exceptional, nor provisional.

The STF issues a massive amount of decisions each year, in the range of dozens of thousands.^{Footnote 41} Historically, an average of less than 10 per cent per year of these decisions have actually been taken by a collective of judges, in the plenary court in one of its two chambers. Individual judges, then, make the vast majority of these decisions. Scholars have observed that this mass of individual rulings can operate as an informal mechanism for case selection, by which the most controversial, novel or important cases are sent for collective deliberation.^{Footnote 42} This is, however, only a partial view of what individual rulings represent in the STF. While they can function as informal docket control mechanisms to filter out lesser or ‘easy’ cases, the data available suggests that this mass of individual rulings enables individual action – including judicial review – even on important cases that would deserve, by any measure, collective deliberation by the STF.

Between 2010 and 2017, the STF issued 20,830 decisions on individual provisional injunctions – an average of 2,603 (260 per Justice) per year.^{Footnote 43} In the same period, the plenary court and two chambers combined decided less than 180 provisional injunctions.^{Footnote 44} In such a large number of individual decisions, it is unsurprising that the plenary court and the chambers have not been able to effectively monitor how individual judges use their decisional powers. Between 2007 and 2016, on average 797 days have passed between an individual ruling granting an injunction on abstract review cases and the first meeting of the plenary or the chambers for which that injunction was on the agenda.^{Footnote 45} Beyond abstract review, between 2012 and 2016, the average is 617 days.^{Footnote 46}

Technically, such injunctions can only be granted in exceptional circumstances. Granting them often means suspending laws, administrative acts, or deliberations on a bill. The only mechanism to identify and neutralise a wrongly issued provisional ruling (i.e., one outside the limits and exceptional conditions prescribed by law) is a vote by the plenary court or one of its chambers. However, with so much time lapsing between the individually issued injunction and the collective opportunity for oversight, what we see in practice is an effective opportunity for individual judges to reshape the legislative status quo, with no interference or oversight by their colleagues.^{Footnote 47}

At least in part, this scenario is created by one individual agenda-setting mechanism: namely, that the case reporter herself – the one who issues an individual ruling in the first place – gets to decide, on her own terms and under no deadlines, when the case is ready for judgment before the court. In practice, although not officially, submitting a case to the plenary court is entirely optional for the case reporter, and the passage of time can turn the provisional suspension of a law or administrative act into a social or political fait accompli.^{Footnote 48} This ‘loop’ between agenda setting and decision making is crucial in understanding how the STF works and how its members influence politics around it: a Justice can decide individually; and that same Justice decides when to take the case to the rest of the court.

We opened this article with the example of Justice Fux suspending, by himself, congressional voting on a key bill. This should not be read as an isolated episode. In the last decade, and especially during the political crisis of 2015–16, it would be easy to find other examples. By himself, Justice Gilmar Mendes granted an injunction suspending former president Lula’s appointment to president Rousseff’s cabinet. It took him a few weeks to clear the case for judgment by the plenary court, and in that period of time President Rouseff was suspended from office on a vote in the House.^{Footnote 49} By himself, Justice Marco Aurélio determined that the president of the Chamber of Deputies had to reconsider his summary dismissal of an impeachment petition against then-interim President Temer.^{Footnote 50} By himself, Justice Fux granted an injunction guaranteeing to all judges in Brazil a de facto salary bonus of dubious legality (‘auxílio moradia’).^{Footnote 51}

These are not minor issues. When decided, they were amongst the most heated political controversies of the day. In none of these high-profile cases, however, did the plenary court have the opportunity to review the individual, ‘provisional’ ruling. In the meantime, politics kept moving forward. In this fashion, supposedly provisional rulings become final in practice, even in high-stakes cases to which the press, the legal community, politicians, and the rest of the court were paying close attention. Whatever the law says about how these individual injunctions should be provisional, they can be placed outside the scope of the plenary sessions and have irreversible effects on the political landscape around the court. Individual rulings are included in the system as a mechanism to protect and guarantee, in an emergency, the plenary court’s possibility of making a decision before irreparable harm is done.

In practice, however, they can be used for the opposite purpose: to make an individual decision and prevent the plenary court from having a say on the issue, or at least to guarantee that the plenary court will have to decide when the status quo has already been transformed in the direction desired by the individual judge.

The STF, then, has a highly individualised and decentralised design, in which individual judges can make their preferences prevail against both external (political) and internal (judicial) majorities. All Justices have the power to individually prevent a case from being decided, and can use the press to try to threaten other institutions by means of signalling future decisions. Finally, and more disturbingly, each case reporter can issue provisional decisions, even if this means suspending the execution of statutes or administrative acts, and then use her agenda-setting powers to control if and when such a provisional ruling will be taken to the full court, blurring the distinction between ‘provisional’ and ‘final’ decisions. Beyond the formal allocation of the power to decide on the merits of a case, though, this decentralised blueprint is created by the combination of and interaction between agenda-setting powers, provisional injunctions and, to a smaller extent, position-taking powers.

It should be noted that this system was not designed to work like this. In theory, most of the individual powers we described should be part of a delegation framework, with the individual judge working as an agent of the collegiate principal, so as to make it possible for the court to deal with the massive amount of cases entering its docket every year. This logic, however, does not capture how the institution works in practice. The mechanisms by which the ‘principal’ could oversee and correct the ‘agent’ are extremely limited, and have been ineffective in even the most important of cases. While a majority of justices could, in theory, change the court’s internal rules of order so as to limit most of these powers, nothing of this sort has been observed in practice, and it still unclear whether this majority would be willing or able to enforce the new rules to discipline individual justices.

Collective decision-making and normative defences of judicial review

Constitutional theorists often separate the questions of (i) how to best interpret constitutional texts and (ii) whose interpretation of the constitution should be binding. Theories that focus primarily on the first question do not deny that all institutions and their members can interpret and apply the constitution – but they do assume that, because the judiciary is insulated from politics, judges will have fewer incentives not to deviate from what the normative theory of interpretation prescribes. Dworkin’s influential theory of ‘law as integrity’, for example, provides both normative guidance to constitutional adjudication and a justification for judicial review, but it does so by establishing how the constitution should be interpreted. Insulation from politics and the requirement to give reasons for their decisions are enough to turn courts into ‘forums of principle’, and thus better suited to redeeming the promise of law as integrity. Still, a Dworkinian would not necessarily reconsider her support for judicial review, at least in the US context, if the Supreme Court were staffed by a single independent judge, as long as that judge was committed to ‘law as integrity’.^{Footnote 54} Dworkin’s and similar theories can be criticised for ignoring the collective nature of the judicial review enterprise, but, for the same reason, the individual exercise of judicial power would not seem to be an independent source of criticism for this theory.^{Footnote 55}

Such theorists could, of course, still acknowledge the existence of good, prudential reasons for collective decision-making that are not in themselves required by their normative theory of constitutional adjudication.^{Footnote 56} Indeed, independently from whatever ideal theory we adopt, we might still think that the power to strike laws down on the basis of constitutional interpretation can be designed in ways that make it more effective or safer – for example, because a single justice might more easily succumb to personal biases, external threats, or other extrajudicial temptations. Or, once we factor in appointment procedures, we might want to design this power in ways that will increase its legitimacy. People would be more suspicious of a single Hercules appointed by a single president or governing majority; in contrast, a multi-member body will most likely include appointees from politicians of very different persuasions.

Excessive concentrations of judicial power in a single judge, then, could still be seen as undesirable as a practical matter. But this problem is not related to the core of Dworkin’s and similar arguments in favour of judicial review. Other approaches, in contrast, defend constitutional review without directly requiring that judges approach the constitution in specific ways. Instead, they ask an institutional design question. If the goal is to enforce constitutional commitments, including fundamental rights, how should we design our institutions? A first set of theories in this group argues that certain properties in the design of judicial institutions might make them particularly good at fulfilling this task in a democracy, in ways that go beyond the mere independence from politics. In particular, as the idea of deliberation entered the scene, high courts were transformed ‘from a democratic anomaly into models of democratic decision-making’.^{Footnote 57}

From this perspective, these institutions’ procedural features, coupled with their insulation from electoral politics, gave them an edge in deliberating and providing an important good in a democracy: reasoned, principled arguments about our constitutional commitments. Indeed, in many recent defences of judicial review, constitutional courts are assumed to be institutions with a high potential, at least, for deliberation.^{Footnote 58} Courts may differ, of course, in their resources and incentives to become actual deliberators.^{Footnote 59} But theorists of deliberation assume that all courts share at least one such resource: judicial power is exercised through collegiate institutions.^{Footnote 60} Collective decision-making is a necessary, but insufficient condition for deliberative performance.^{Footnote 61} Still, when a single judge can decide at an apex court, the promise of judicial deliberation as a redeeming feature of constitutional review in a democracy is obviously lost.

Finally, there are arguments in favour of judicial review that, while also taking institutional design into account, do not rely on showing that courts are particularly good at ‘getting the constitution right’. Instead of focusing on the institutional expertise required by this task, they focus on what judicial review has in common with other deviations from majority rule, such as bicameralism. By requiring an increased level of political agreement between different institutions, such mechanisms promote positive outcomes like stability or the respect for minorities’ rights.^{Footnote 62} Such a ‘modest case’ for judicial review would not rest ‘on the idea that courts are more likely than legislatures to make correct decisions about how to define vague rights’ because they are ‘forums of principle’, but rather on the ground ‘that legislatures and courts should both be enlisted in protecting fundamental rights, and that both should have veto powers over legislation that might reasonably be thought to violate such rights’.^{Footnote 63} In such a view, giving courts this kind of power is justified as a measure to bias the political process towards minimising violations of rights by simply adding one extra veto point. If this veto is exercised by an institution that could be particularly responsive to rights claims, even better, but this is not required for the argument to work.

In principle, this line of reasoning would seem to justify even individual judicial review. If creating additional pro-rights checks is important in a democracy, why not give several independent judges, within a court, the power to void majority political decisions? However, the argument for judicial review as just another super-majority mechanism in a democracy needs a stopping point. Democracies might not be simply about majority rule, and vetoes and checks and balances, judicial or otherwise, might be a necessary component of constitutional democracy. But vetoes and checks cannot be multiplied endlessly or arbitrarily. The more we add veto points, the closer we approach a system in which law-making requires unanimity amidst conflicting views, which makes us risk abandoning, in practice, any resemblance of majority rule.^{Footnote 64}

To consider the implications of this point, we need to distinguish more clearly between two different versions of individual judicial review (IJR). In the first, we could have a truly individual court – a single judge with the power of judicial review. The second version of individual judicial review, by contrast, consists of a multi-member court in which each member can perform judicial review by herself (as we have argued is the case in Brazil). In the first case, there might be no problem at all for veto-centred approaches to judicial review. After all, in contrast to deliberative democracy theories, veto-centred theories hold no strong assumptions on how the judicial decision is formed, other than judicial independence and insulation from politics. In the second case, however, we have a problem. If judges in a collegiate court have individual judicial review powers, what we have then is a voting rule by which a single vote is enough to perform judicial review. For the veto-centred theories, this multiplication of vetoes might begin to harm the possibility of majority rule in democratic politics. IJR in a court of 11 might make it impossible for any minimally controversial measure to be adopted by a majority, while IJR in a court of three might not pose the same problem.

The second scenario is problematic for theories of deliberation as well. If IJR is actually a voting rule, it could still be the case that judges actually deliberate before using their individual powers. They could listen to each other’s arguments, and perhaps even try to reach a consensus so as to protect and foster their institution’s authority. But, if reaching a consensus and persuading each other is not possible, one vote would be enough to strike down a law (if decision-making power is allocated to a single judge), or to completely prevent that law from being struck by the court (if one single judge has the right agenda-setting powers). We have now moved from collegiality as an institutional feature to a finer question of the internal design of judicial decision-making procedures: why should a single vote in a multi-member court be enough (or not enough) for judicial review?

Judicial power and majority rule in multi-member courts

In this section, we will consider individual assignments of judicial power – focusing on, but not limited to, individual judicial review – as a voting rule, by which the vote of a single judge in a multi-member court is a sufficient condition to exercise power over the outside world. We typically assume that courts should settle their internal disagreements and reach a decision by means of majority rule (MR). But are there good reasons behind this arrangement? Considering how arguments developed for political bodies could be applied to courts, Waldron maps three kinds of arguments in favour of MR: (i) efficiency (MR is the cheapest and fastest way to solve a dispute); (ii) epistemic (MR allows the institution to make better decisions and reach better conclusions); and (iii) justice arguments (MR treats all the parties involved as political equals).^{Footnote 65} Can IJR and other individual powers be justified by such arguments?

Efficiency, argues Waldron, cannot in itself be a decisive argument in favour of any voting rule, because there must always be an underlying concern with the legitimacy and the ‘decisiveness’ of the procedure. A ‘decisive’ procedure is one that would lead to the same outcome, over time, if employed again for the same reasons – otherwise, the answer we obtained from the institution now is indeterminate and unstable, and we would expect a different answer if we tried again. This is why we would not allow courts to reach a decision by flipping a coin – the cheapest, fastest possible decision-making mechanism.^{Footnote 66}

Using the ‘decisiveness’ standard to look at individual allocations of powers yields mixed results, depending on the specific configuration of the court. In the STF, each and every judge can veto judicial intervention by removing a case from the agenda – that is, each judge has a negative agenda-setting power. Whatever other problems this procedure might have, it is undeniably ‘decisive’: if Justice X wants to take the case out of the agenda, all other things being equal, she would do so again if we employed the same procedure to solve the same question. The same cannot be said, however, of the case reporter’s power to issue provisional (and potentially de facto final) rulings. The case reporter is assigned randomly as the lawsuit enters the court’s docket. If we consider the assignment of the reporter to be part of the full decision-making procedure – as we should, in order to understand how the STF actually works in practice –, then each re-run of the procedure would lead to a different outcome. Different reporters can have different views on the same issue, and depending on who gets the case we may or may not have a display of individual judicial review.^{Footnote 67}

Broadly speaking, epistemic arguments in favour of majority rule claim that this promotes better decisions. Once we assume that judicial decisions involve some degree of expertise and experience, when judges face internal disagreement the position that has the support of the largest number of these ‘experts’ should prevail.^{Footnote 68} But if, in a given court that is divided over this issue, a single Justice is able to force the court’s power in her preferred direction, we have an institutional problem on epistemic grounds. There is no guarantee that the court’s power is being used in a way that carries at least the support of at least a majority of the court. This would be a problem even if all judges were forced to listen to their colleagues’ reasons and arguments before using their individual powers; it is even more problematic and unjustified when they can do so before the discussion takes place.

Finally, we have a justice-based argument, which appeals to political equality: majority rule is the only way to ensure that every participant in the decision-making process has equal weight in the decision.^{Footnote 69} Waldron is somewhat sceptical of whether we must always treat the justices themselves as complete political equals in this sense. In some courts, the institution is designed so that certain offices (like the Chief Justice) have increased influence in the internal decision-making process, sometimes even by means of an extra vote in the case of a tie. The stronger case for treating judges in a multi-member court as equals comes from considering the democratic majorities behind them. Procedures of political appointment to judicial offices express an attempt and a commitment to make judicial interpretation open to influence by the cycles of politics – perhaps not in the short run, after a single election, but certainly after a majority coalition wins enough elections to make a certain number of appointments to the court.^{Footnote 70}

Treating judicial appointees as equals within the court, then, means giving equal weight to all the political choices made by elected representatives. Whatever the specific ideals, reasons, and calculations behind these attempts to regulate the translation of electoral victories into constitutional law, however, minority or individual rule in judicial bodies turns everything upside down. One appointment becomes enough to enable or disable judicial intervention in politics, even when a given political coalition has appointed several (or perhaps even a majority) of the justices of that court. An arrangement that gives each individual member of court, by herself, both the power to decide or to veto a decision would thus seem to fail on this standard as well.^{Footnote 71}