Prior to the Constitution of 1988, the Brazilian system of judicial review was closer to the American tradition of judicial review than to the Austrian model developed by Hans Kelsen, which concentrates the jurisdiction on constitutional matters in a single Constitutional Court. The declaration of unconstitutionality was merely a step to be followed by any ordinary the judge in the justification of her decision. Even though a dispute over the constitutionality of an Act could eventually reach the high courts by means of an appeal, there was no warranty that this would ever happen.

Although there was an abstract action for assessing the constitutionality of statutes before the Federal Supreme Court, the scope of this action was very strict and only the General-Attorney of the Republic (who was at the same time the Procurator of the Federal Government and the Chief of the Public Ministry) was empowered to bring it before the Supreme Court. In a legal system where the General-Attorney of the Republic was directly subordinated to the President of the Republic and where the Executive Administration was authorized to legislate in place of the Congress in a wide range of areas, like Brazil was at the time of the Military Dictatorships of the three decades which antedated the Constitution of 1988, that sort of action was of very little utility.³

These brief historical considerations help one understand the roots of the current Brazilian system of constitutional jurisdiction. They explain, for instance, the fact that any judge in the country is competent to decide a constitutional issue. The general competence to adjudicate on constitutional issues has its origins in such historical tradition. Nonetheless, the Constitution of 1988, in spite of keeping the incidental system of judicial review, was deliberately designed to break down with this tradition. The Brazilian Constitution of 1988 is partly inspired by the Kelsenian ideas that a decision which pronounces the unconstitutionality is not declaratory, but rather constitutive; that there can be no “unconstitutional” statute in the sense of a null and void piece of legislation, but only a statute “contrary to the Constitution” which can be derogated by the Constitutional Courts through a special procedure that is different from ordinary legis la tion; and that as a general rule the Constitutional Court pronounces erga omnes decisions creating a derogatory rule which eliminates a previous norm incompatible with the Constitution (in such a way that the court is a negative legislator). These ideas have a deep influence in the institutional settings of the Brazilian state and the doctrines about the relationship between the Legislative and the Justiciary.

The Constitution of 1988, although without naming the Federal Supreme Court a “Constitutional Court”, placed that court in a special position and established a set of privative competences to exercise the judicial review. The court is now competent to decide abstract actions of unconstitutionality against most statutory provisions, regardless of any specific situations to which they are applied. Such direct actions can be brought before the court by a relatively vast group of entities which represent general sectors of the society.

This concentrated system of judicial review coexists with the historical model of incidental and diffused constitutional adjudication. The simultaneity of the two traditional models is one of the distinctive features of the Brazilian system of judicial review. Even though the judicial ideology that dominates the discourse of the Supreme Court remains faithful to Kelsen's image of the Constitutional Court as no more than a “negative legislator”, our analysis will show that there are occasions in which it effectively acts as a positive law-making agency, albeit strictly bound to the constitution and sensitive to its judicial role. In fact, one can easily agree that the Brazilian Constitution expressly requires the Court to lay down general and abstract norms which sometimes are hierarchically ranked in the same level of the ordinary legislation. In order to justify this assertion, we will analyse the Brazilian system of judicial review by separating the incidental and diffused review from the abstract and concentrated one.

1. THE DIFFUSED AND INCIDENTAL SYSTEM OF JUDICIAL REVIEW

The model of incidental and concrete judicial review remains applicable in Brazil and constitutes one of the central features of the Brazilian legal tradition. It is a concrete form of judicial review because the unconstitutionality of a norm (no matter where it can find its sources: in a statute, in an international treaty, in an administrative decree, in a conventional norm within a contract etc.) is argued by one of the parties in the course of an ordinary legal dispute. As opposed to most European countries which have a Constitutional Court, the judge must decide the constitutional issue herself. The decision about the (un)constitutionality of the norms is a necessary step that the judge has to take before reaching her conclusion. Further to being a concrete system of judicial review, this is also a system of diffused judicial review because the constitutional jurisdiction is spread out through the court system. Every court in Brazil has constitutional jurisdiction. When the unconstitutionality is argued before a first-instance judge, there is no need for any special formalities, except for giving the other party the chance to counter the arguments advanced for the declaration of unconstitutionality. The Constitutional Court does not have a say on the issue unless the case reaches it by an extraordinary appeal that only is admissible after a final second-instance decision is pronounced.

When the unconstitutionality is argued before a court of appeal or any other highcourt, the constitutional issue can only be decided in a plenary session of the court. The reporter judge must suspend the judgment of the case until a decision on the constitutional query is achieved by the majority of the full house.¹

Nevertheless, the rule which establishes this privative competence to the plenary sitting of the courts does not avoid contradictory decisions. The Brazilian judicial system is heavily fragmented, for several reasons. First, Brazil is a federal state constituted by 26 States and one Autonomous District where the headquarters of the Federal Government are situated. Each State has its own courts and very rarely a precedent from a different court of appeal of the same hierarchy is quoted in a state court. Second, there are different and completely autonomous court systems whose competences are determined according to the subject matters of the cases: there is a Federal Court system (which decides issues of interest of the Federation or of its subsidiaries, such as autarchies, public enterprises, public foundations, Federal Universities etc.), an Electoral Court system (which has administrative competences for organizing the general elections; normative authority to issue general norms supplementing the electoral legislation and jurisdictional competences over a vast range controversies around the application of electoral law - for instance, cases about limits of the freedom of expression, unlawful political advertisements, abuses of economic power in the elections etc.), a Labour Court system (which decides labour cases and is informed by the principle of the protection of the employees against the dominant position of their employers), a Military Court system (for Military Crimes), and the General State Courts (which have jurisdiction on all remaining subjects, such as civil disputes in general and criminal cases in which the Federation is not a victim or an interested party).

When there is no binding precedent applicable to a case, there is no procedural mechanism enabling a party o r a j udge to submit the constitutional issue to the Supreme Court before a decision is reached in the ordinary courts. Apart from exceptional cases which will be dealt with later in this report, the technique of avocamiento, which is admitted in some Latin American countries, does not find an equivalent in Brazil. Only final decisions from the ordinary courts can be challenged by an “extraordinary appeal” (recurso extraordinary) to the Federal Supreme Court. These extraordinary appeals are admitted to decide constitutional controversies. They can neither re-examine the evidence or any question of fact nor adjudicate on a question of interpretation of infra-constitutional laws.

The issue of the unconstitutionality of any statute, international treaty, legislative decree or administrative resolution which establishes a general norm can be brought before the Supreme Court via a recurso extraordinario, although there are some procedural barriers aimed at filtering the number of appeals to the Federal Supreme Court.⁵

The amount of discretion of the court in choosing the appeals it will decide is relatively low. Until relatively recently there was nearly no discretion (all the cases which fulfill all procedural requirements where submitted to the court), but a recent Constitutional Amendment has determined that only cases of “general repercussion” (repercussao geral) can be submitted to the court.⁶ The basic idea is that only cases which reflect upon the status of positive law in a relevant way should be decided by the court. The main purpose of the appeals to the Supreme Court is not to protect individual situations, but rather to unify the interpretation of valid law. To decide whether or not an issue is of “general repercussion” it is necessary a judgment of the plenary session of the Federal Supreme Court (which is constituted by eleven Ministers). It is in the case law of the Court that one will find the criteria for identifying such cases. Nevertheless, once a party demonstrates that her case fits the constitutional definition of a case of “general repercussion”, she has a constitutional right to see her case decided by the court.

As a rule, the incidental declaration of unconstitutionality of a provision or an act neither is strictly binding nor has erga omnes effect. According to the wording of the Constitution, or an incidenter tantum decision to become universally efficacious the Federal Senate must pass a resolution derogating that norm.⁷ The Federal Supreme Court, when reaching a definitive decision recognizing the unconstitutionality of a legal provision, notifies the Senate, which will have discretion on whether or not the norm should be formally abrogated. Nonetheless, in spite of this constitutional provision, the practice of the Senate is to avoid eliminating particular legal provisions declared unconstitutional by the Supreme Court.

This does not mean, however, that the final decisions of the Supreme Court are not a uthoritative. Even though as a rule the decisions of the courts have inter partes efficacy, constitutional precedents are of fundamental importance in the

Brazilian legal system. If we stick to the classification of the institutional force of judicial precedents adopted by one of the authors of this report in a previous writing, who distinguishes three categories of normativity for judicial precedents (precedents “binding in a strong sense”, precedents “binding in a weak sense” and precedents “merely persuasive”), we can place this sort of case law in the intermediate category and characterize it at least as “binding in a weak sense.”⁸

As we will see later in this report, there is a clear trend of increasing substantially the normative powers of the Supreme Court in the decisions of unconstitutionality, regardless of whether they are pronounced in the course of a legal dispute or in a Direct Action of Unconstitutionality.

Some signs of this trend are the recent constitutional and legislative reforms that enhanced the binding character of the decisions of the court. However, before examining these law reforms we will outline some of the aspects of the concentrated system of constitutional review.

2. THE CONCENTRATED SYSTEM OF JUDICIAL REVIEW

2.1. Concentrated constitutional jurisdiction by direct actions⁹

The concentrated system of judicial review in Brazil was inspired by the systems from European countries like Austria, Germany, Spain, Italy and Portugal. Nevertheless, there are some specific features which distinguish the Brazilian model.

There are four basic types of direct and abstract actions of unconstitutionality in Brazil: the Direct Action of Unconstitutionality (Agao Direta de Inconstitucionalidade);^w the Declaratory Action of Constitutionality of a federal law or normative act (Acao Declaratoria de Constitucionalidade);¹¹ the Direct Action Against an Unconstitutional Omission (Agao Direta de Inconstitucionalidade por Omissao)¹ and the Claim Against the Disrespect to a Fundamental Precept (Arguigao de Descumprimento de Preceito Fundamental).'¹³ All these actions have a limited group of authorities or entities that can figure as claimants, which will be specified below. In any direct form of constitutional review, the General Advocate of the Union (Advogado-Geral da Uniao) will be heard in defense of the normative act. By the same token, the General-Attorney of the Republic will have a chance to present a memorandum when he is not the author of the action.¹⁴ Let us outline some of the features of these actions.

a) The Direct Action of Unconstitutionality;

The direct action of unconstitutionality is the main instrument by means of which the Supreme Court adjudicates on a general impugnation of the validity

the Governing Boards of the Federal Senate or of the Chamber of Deputies, the Governing Boards of the Legislative Assemblies of the Federal States, the Governors of the States, the General-Attorney of the Republic (Procurador Geral da República), the Federal Section of the Brazilian Association of Advocates (Ordem dos Advogados do Brasil), any Political Party represented in Congress and any national Union or class-representing association.²⁰

b) The Declaratory Action of Constitutionality

The Declaratory Action of Constitutionality (henceforth aDc, which stands for the Portuguese sentence “Agao Declaratoria de Constitucionalidade’) was not included in the original wording of the Constitution of 1988. It was introduced by the 3rd Amendment to the Constitution, dated from 1993. Only federal normative acts can be in the object of an ADC, although the same authorities who can bring an ADIN before the Supreme Court are also competent to bring an ADC.²¹

The distinctive feature of the ADC is that its purpose is to demonstrate not the unconstitutionality of a normative act, but rather its compatibility with the Constitution. Since every court and every first instance judge has constitutional jurisdiction, sometimes the lower courts are overwhelmed with identical claims arguing against the constitutionality of a particular statutory provision. In Fiscal matters and issues related to State Pensions, the number of identical cases contesting the constitutionality of tributes or the criteria used by the Government to update state pensions can be alarming. In the year 2008, for instance, 46.94% of the new cases submitted to the Supreme Court (that is, 34.394 of the total of 73,221 cases submitted to the Supreme

Court) dealt with Fiscal Law and Administrative Law matters.²²

As established by federal law, the proponent of the ADC must indicate in the bill of complaint “the existence of relevant judicial controversy over the application of the provision which constitutes the object of the Declaratory Action.”²³ The purpose of the action, as summarized by Minister Gilmar Mendes in an academic writing, must be the preservation of the presumption of constitutionality of the legislation, which can be threatened if there are a large number of decisions departing from the norm. ²⁴ It is a burden of the claimant to demonstrate that there is relevant judicial controversy over the constitutionality of the provision by quoting a substantial number of contradictory decisions from different judicial bodies.²⁵ In this sense, the court has held that the claimant on the ADC must, at the time of filing the claim, demonstrate a “relevant proportion” of judicial disagreement. This view was justified by a reductio ad absurdum: such judicial disagreement must be strong enough to install a state of general lack of legal certainty, for otherwise the anticipatory rulings of unconstitutionality would mischaracterise the jurisdictional nature of the activity of the Supreme Court by turning the court into an organ of consultation, and thereby opening the way to the risk of undermining the balance of powers between the Legislative and the Judicial Branches of Government.²⁶

In line with the purpose of increasing the degree of legal certainty and unifying the interpretation of the Constitution, there are some additional powers conferred to the Court. When assessing the procedural requirements of the ADC, the reporter judge may, if not satisfied with the decisions quoted by the claimant, request additional information to

any other court or create a commission of experts to analyze evidence on the impact of jurisprudential disagreements in the current state of the law.⁷

Furthermore, the Court may, by a collective judgment of the absolute majority of its members (that is, 6 out of 11 Ministers), issue restraint orders determining other jurisdictional bodies of the Republic (State Courts, Federal Courts, Labor Courts, Electoral Courts, Military Courts and first instance judges) to suspend the judgment of all cases regarding the application of the law or the normative act under discussion.⁸

c) The Direct Action against Unconstitutional Omissions

While the ADIN challenges a positive norm and is aimed at derogating unconstitutional legislative acts, the Direct Action against an Unconstitutional Omission (Omissive ADIN) deals with the inertia of the law-giver, that is, an “unconstitutional gap” in the legal system.⁹ The Constitution is deprived of its cogency by the lack of a normative act that should have already been passed by Congress.

There are, of course, several dilemmas which remain unsolved by the case law. If what characterizes the legislative activity is its continuous and intermittent character, how can one recognize an unconstitutional omission? If the court is prohibited by the democratic principle and by the principle of the rule of law to substitute the legislator when he remains negligent in concretizing the Constitution, what kind of efficacy can be attributed to a decision which recognizes an unconstitutional omission? If there is no procedure for coercively executing this decision, how can one expect the decision which recognizes an unconstitutional omission to be authoritative?

These questions are still unanswered and probably will remain for a long time. The Brazilian doctrine of the unconstitutional omissions follows the pattern of the case law of the German Constitutional Court.¹⁰ In the academic literature, for instance, two types of unconstitutional omissions are recognized: complete and partial omissions. While a complete omission takes place when the legislator does not produce any law in spite of the fact that there is a genuine constitutional obligation of regulating some constitutional issue, a partial omission occurs when the legislative authority regulates a situation in an unconstitutional way because it does not cover a set of situations that should have been included in the statute. The classical case of partial unconstitutional omission is the concession of a benefit in detriment to the principle of equality. The law is unconstitutional because it fails to cover situations that should have been included in its general hypothesis.¹¹

The number of omissive ADINs in Brazil is relatively low, and the tendency is that this number drops even further. As the Constitution gets older, a smaller number of matters referred to in its text are left completely unregulated. The trend is that the court should deal more with partial omissions than with complete omissions.¹²

As it happens with the ADIN, one may bring and omissive ADIN before the court not to uphold her own right, but rather to protect the juridical order as a whole. The object of the action is the integrity of the legal order, not any specific right.

In general, the Supreme Court's case law on constitutional omissions can be classified as conservative. The court attributes heavy weight to the princip le of democracy and to the principle of division of powers. The classic idea of a system of “checks and balances” recommends one to avoid any type of judicial activism when it comes to providing a remedy for unconstitutional omissions.

The court seems to incline itself towards the view that, in case of a complete absence of a regulation that is required by the constitution, the court should declare its unconstitutionality without pronouncing the nullity of any act and without issuing a direct order to the legislator. In this sense, the Court has held, in an omissive ADIN which intended to establish that the value of the minimum wage was unconstitutional because it could not supply for the satisfaction of the basic needs of a person, that while deciding an omissive ADIN the Supreme Court can do no more than notify the legislative body competent to remediate the omission, in order to it aware of the unconstitutionality and to enable it to regulate the matter required by the Constitution without interference from the Justiciary.¹³

In case of normative acts within the competence of the Executive, however, the Constitution establishes that the Court should give the Administrative authority a deadline of 30 days to eliminate the omission.¹⁴

As mentioned before, one of the difficulties of omissive ADINs is to determine when an unconstitutional omission is characterized. The Court has held that the general rule must be that if the legislative process has already initiated there can be no unconstitutional omission of the legislator.¹⁵ Nevertheless, more recent decisions hold that if the intertia deliberandi is unequivocally characterized and the delay amounts to negligence, the Court may pronounce an unconstitutional omission in spite of the existence of a bill under discussion.¹⁶ In a leading case in which it was demonstrated that there was a 10-year delay in enacting an act which was expressly required by the Constitution for the creation and the redefinition of territorial boundaries of Municipalities, the court held that there had been a breach of the duty to legislate and that the inertia deliberandi could be challenged by an omissive ADIN.¹⁷ This recent leading case is also important because it inaugurated the possibility to set a deadline for Congress to legislate: the Court ruled that the lawgiver should enact a statute, within 18 months of the publication of the decision, to eliminate the omission and to regulate retrospectively the facts occurred (and the political and administrative acts practiced) between the promulgation of the Constitution (or, to be more specific, the Amendment to the Constitution which established the duty to legislate) and the coming into force of the upcoming regulation.¹⁸

d) The Claim against the Disrespect to a Fundamental Precept

The Claim against the Disrespect to a Fundamental Precept of the Constitution (henceforth ADPF, for Argüiçâo de Descumprimento de Preceito Fundamental) is subsidiary to the ADIN and can be brought before the court by the same entities which are legitimized to bring the other types of direct action.¹⁹ Even though the ADPF was mentioned in the original wording of the Constitution of 1988 (and not merely introduced by an Amendment to the Constitution, as it happened with the ADC), it was n ot regulate d until the coming into force of a law which was passed in December 1999. Because of its subsidiary character, it is only admitted when there is no alternative remedy to protect the fundamental precept of the Constitution.²⁰

The ADPF is of significant importance because it makes it possible to protect the Constitution against acts which are left outside of the scope of the ADIN. The most obvious cases are municipal laws and normative acts which were in force before the p ro mulgation of the Constitution.²¹ At the time of the coming into force of the statute which regulates the ADPF, there was a genuine social pressure for the admittance of a direct action for adjudicating on the constitutionality of pre-constitutional and municipal laws.

Furthermore, the ADPF enabled the Court to assess the validity of acts that have always been excluded from abstract constitutional jurisdiction. Any normative act of a public authority can be challenged by an ADPF, including “interpretative directives” issued by a court of justice.²² In this sense, the sumulas of the Brazilian courts of justice might be submitted to a direct assessment of their constitutionality.²³ A sumula (from Latin, Summula) is an abridgment of law which is relatively authoritative because it is enacted by a court of justice to publicize its own case law. Perhaps the closest equivalent to the sumula that one can find in European law is the Italian massima, which can be defined as “a very abstract statement representing the core of meaning of a legal rule, as it is interpreted by the judgment considered.”²⁴ Sumulas are only edited after a set of repetitive decisions, and their enactment is preceded by a vote of the full panel of the court. Nearly all the courts in Brazil have repertories of their sumulas, and for obvious reasons jurists tend to attribute a gr eat deal of authority to them.²⁵

Finally, as Minister Gilmar Mendes has recently argued, the ADPF may be used to review administrative acts of regulation (regulamentos administrativos) which violate either the principle of Due Process of Law (which establishes “no one shall be deprived of her liberty or her property without the Due Process of Law”)²⁶ or the principle of Formal Legality (which establishes that “no one shall be obliged to do or to refrain from something unless by order of a law”).²⁷ In our opinion, however, the ADPF should be applicable to such normative administrative acts only when they are enacted by the Municipal administration, for in the case of federal laws or laws from the member-states of the Federation these acts are already covered by the ADIN.²⁸

An opened question for Brazilian constitutional lawyers is which constitutional norms can constitute a parameter for and ADPF. What should the Court understand by a “fundamental precept”? In this topic, there is no final answer in legislation or in the case law. It would be recommendable to pass a law establishing a complete list of subjects that can be protected by the ADPF. In the absence of such legislation, however, the Court developed its own directives to define the fundamental character of a constitutional provision. After recognizing that it is “very difficult to delimitate, a priori, the fundamental precepts of the Constitution”, the Supreme Court held that at least the following group of norms are protected by the ADPF: (1) the Individual Rights;²⁹ (2) those precepts which, by virtue of an explicit constitutional provision, can not be suppressed by an Amendment to the Constitution and therefore form the “immutable core” (clausulas pétreas) of such Constitution;³⁰ and the so-called “sensitive principles” of Federalism,³¹ that is, the set of norms which if disrespected would authorize the interference of the Union in the competences of

the States or the interference of the States in the competences of the Municipalities.⁵²

2.2. Concentrated Constitutional Review by Concrete Claims

Further to the four types of abstract claims of unconstitutionality highlighted above, there are two special writs that are relevant for constitutional jurisdiction. The first one is the Writ of Injunction (Mandado de Injungao), which establishes a specific remedy for unconstitutional omissions when such omissions prevent the exercise of a right already established by the constitution. The second is the Writ of Mandado de Seguranga, which is an action that can be brought against a public authority (be it a physical person or an administrative body) that by its illegal action keeps one from exercising a legal right.⁵³

These two actions, the Writ of Injunction and the Mandado de Seguranga (when used to challenge a bill under discussion in one of the Houses of Congress), are a form of concentrated, although not abstract, constitutional review. It is concentrated because the jurisdiction is privative to the Supreme Court, but concrete because the writs are used to protect a right of the claimant, rather than the general integrity of the legal order.

a) A Mandado de Seguranga (when used to contest an act within the legislative procedure)

The Mandado de Seguranga (henceforth MS) is a type of constitutional action that is brought against a public authority in order to obtain an injunctive relief in the form of a court order restraining such authority from performing a particular act. The MS can be grounded either on a violation of constitutional norms or on a violation of ordinary statutory regulations. In the vast majority of cases, the incidental declaration of unconstitutionality within a MS does not differ from the rest of the diffused and incidental forms of judicial review, but in one specific case it does have a particular feature which transforms it into an instrument of concentrated constitutional review: the proponent of an MS may challenge the production of a general normative act. Any interested party is legitimised to sue the Governing Board of the Federal Senate or the Governing Board of the Chamber of Deputies in order to prevent the House from passing an unconstitutional act which would affect a liquid right of the claimant.⁵⁵ By that procedure, the Supreme Court is empowered to interfere in the legislative process in order to knock down a bill by exercising a preventive form of constitutional jurisdiction.⁵⁶

b) The Writ of Injunction (Mandado de Injungao)

The Constitution of 1988 established a large number of rights to be regulated by ordinary legislation. Nevertheless, there is a constitutional provision stating that all fundamental legal rights are “immediately applicable.”³² The constituent law-maker, in order to uphold the efficacy of the Constitution, created a new constitutional remedy named Writ of Injunction (Mandado de Injungao, henceforth MI). The MI is also a form of concentrated although concrete judicial review. One may bring an MI before the Supreme Court not in defense of the general integrity of the legal order, but rather in defense of one's own rights.

Like the Omissive ADIN, the MI is a constitutional writ whose function is to break down the inertia of the legislator. The writ is admissible whenever the lack of a regulatory infra-constitutional norm “makes it impossible to exercise the civil liberties and fundamental legal rights or the prerogatives related to nationality, sovereignty and citizenship.” ³³

The first leading case on the admissibility of the MI was decided merely one year after the promulgation of the Constitution of 1988. In this case (MI 107), the court held that the efficacy of a decision delivered in an MI is similar to that of an Omissive ADIN: the MI is an action which intends to obtain from the Justiciary a declaration of unconstitutionality of an omission in regulating a right, with a view of notifying the entity responsible for that regulation to take action, as it happens with the Omissive ADIN.³⁴ The court was very firm in its conviction that it could not act as a “positive legislator”, in the lines of the classical liberal view of Hans Kelsen.³⁵ No additive decisions either to integrate or to amend the legislation were admitted.

There are, however, some cases in which the court gave a broader scope to the procedural remedy of the MI.

In the MI 283, for instance, the court recognized a state of negligence of the Congress in regulating a norm established by the Temporary Provisions of the Constitution.³⁶ That norm stipulates a duty to provide compensation for the victims of abuses committed by the military dictatorship via Secret Acts of the Ministry of Defense which banned a large number of people from exercising certain economic activities. Since the temporary provisions required the passing of a federal law to regulate the particulars of such compensation, the victims could not exercise their constitutional rights. In the light of this specific situation, the Supreme Court not only ruled that there was an unconstitutional omission, but also established a deadline of 45 days for the Congress to legislate. The Court determined, moreover, that in case the state of parliamentary negligence remained after that deadline, the applicant would be automatically entitled to claim compensation against the Union, in the form of the general rules of the Civil Code.³⁷

In another relevant case, the Constitution guaranteed a fiscal privilege to beneficent institutions of social assistance, excluding them from taxation by contributions to the social security, “as long as these entities complied with the conditions established in law.”³⁸ The Constitution has left to the ordinary legislator the task to discipline the conditions with which those entities shoul d comply in order to claim immunity from the contributions.

Nevertheless, the Federal Government understood that such entities could claim no fiscal immunity until

Congress passed a law listing such conditions. The Supreme Court, after holding that there was an unjustifiable legislative omission, fixed a deadline of six months for Congress to pass a law eliminating that omission. Furthermore, it determined that if no law was passed until that deadline, the claimant would be automatically entitled to claim the fiscal benefit.⁶⁴

In these two cases, the court took a step towards judicial legislation, albeit only with inter partes effect. That step, however, was of limited significance. In both cases the Constitution is very clear about the rights that are protected by its provisions. There is no doubt about the semantic meaning of the constitutional norms and it is very easy to understand the scope of the right which is determined by the Basic Norm. The reference to ordinary legislation can mean no more than the assertion the law-maker may, within a certain margin of appreciation, restrict or extend the protection of such rights. One could even say that the MI was not really needed in those cases, on the grounds that the Constitutional provisions in issue were immediately applicable, in spite of the fact that the legislator might restrict them.⁶⁵

c) Additive Decisions and the Writ of Injunction

In some recent decisions, the Supreme Course has made substantial changes in its case law by recognizing the possibility of additive decisions within a Writ of Injunction. In a claim (an MI) filed against the absence of a law regulating strike actions by public servants, the Court overruled in part its leading case (MI 107) concerning the limits of the judicial powers of filling in unconstitutional gaps. The Brazilian Constitution expressly contemplates among the fundamental labour rights the right to come out on strike, and has a specific provision requiring a federal law to lay down the particulars of strike actions within the public services. Nevertheless, nearly twenty years after the promulgation of the Constitution, no law had been passed regulating these matters. While the Government stated that its servants were not allowed go on strike until a statute fixing the limits and the conditions for exercising this right was enacted, the unions of workers and many leaders of labour organizations interpreted the aforementioned constitutional provision as establishing an unlimited or unconstrained right. In times of tension between the Government and its servants, the situation has reached a level where serious losses have been suffered by the population. In administrative bodies like the National Institute of Social Security, a multimonth strike has lead thousands of pensioners to suffer into lerable delays in receiving their pensions. Admini strative claims of new benefits have been suspended and a large part of the population have been unable to claim benefits such as maternity leave or the allowance paid by the Government (in place of the salary) to people away from work for health reasons. As the Court held, the absence of a regulation on this issue has led to a sort of “state of nature” which has “serious consequences for the Rule of Law.” As Minister Gilmar Mendes expressed in his opinion, to leave the issue unregulated would amount to a sort of “judicial omission” in protecting the Constitution.⁶⁶

On the face of this context, the Supreme Court pronounced the first additive decision - or at least the first admittedly additive decision - in the history of its case law. As opposed to the decisions on the MI 283 and the MI 232 - where it was ruled that a Constitutional right may be directly applicable in spite of the possibility of restrictive legislation only if its content can be directly determined by the interpretation of the constitutional text -, the Court decided to make positive regulations for a situation which was leading to serious social conflicts. After analyzing in detail the practice of additive

decisions in the Italian tradition, the Court made express reference to the works of the Portuguese Professor Rui Medeiros, who admits additive decisions integrating legislation or yet when the regulation adopted by the court is “constitutionally obligatory.”⁶⁷

The court decided thus to analogically apply the ordinary labour laws which regulate strike action in private labour contracts. Until further legislation is passed, public servants are subjected at least to the same rules that apply to ordinary workers as to the abusive forms of strike action.⁶⁸

Nevertheless, one should not overestimate the impact of this new case in the state of the law. Arguments by analogy are a central feature to any legal system, and there is nothing original or particular to developed forms of constitutional review. The sole distinctive feature of this type of case law is that it constitutes an analogical decision with erga omnes effects, since it establishes a rule to be generally observed until legislative acts are passed.

In spite of the general effect of its analogical decisions, the Court expressly insists that it is not acting as a “positive legislator”, but merely as the “guardian of the Constitution.”⁶⁹

3. EFFECTS OF THE JUDICIAL DECISIONS ON CONSTITUTIONAL MATTERS

3.1. Inter partes and erga omnes decisions

The distinction between inter partes and erga omnes effects of constitutional decisions provides the key criterion to distinguish the decisions of constitutional questions in concrete cases (that is, as an incidental question within a legal dispute) and in abstract constitutional actions (that is, claims detached from any case of application of the provision challenged by direct actions).

As a rule, decisions of unconstitutionality - including the cases of “partial annulment of a legal norm without textual reduction” of a statutory provision (declaragao de inconstitucionalidade parcial sem redugao de texto) and the “interpretation in accordance with the Constitution” (interpretagao conforme a Constituigao) - pronounced in abstract or direct actions are erga omnes efficacious and have both a derogatory effect on the unconstitutional provision and a repristinatory effect on the legislation which was abrogated by it.⁷⁰ Decisions on abstract constitutional actions do not resolve concrete disputes, but rather eliminate the statutes or the provisions pronounced as “unconstitutional.”

On the other hand, decisions of unconstitutionality within a legal dispute have inter partes effect and thus lack authority to derogate statutory legislation. Notwithstanding this, there is an ongoing discussion in the Supreme Court about the possibility of attributing erga omnes efficacy to incidental constitutional decisions that the Court might have adopted in concrete constitutional review. In a case in which only four of the eleven Ministers of the Federal Supreme Court have already delivered their judgments, some Ministers of the Court have argued that the system of constitutional review in Brazil has suffered a “constitutional mutation.” The Constitutional provision which requires that a Resolution of the Federal Senate should be passed in order to attribute erga omnes efficacy to a

final decision of unconstitutionality laid down by the Supreme Court⁷¹ would have lost its juridical relevancy: once the system of judicial review developed in the direction of permitting the Court to lay down a group of binding sumulas,⁷² there would be no point in requiring a resolution of the Senate to do something that the Court can now do by its own authority.⁷³ If this interpretation prevails, a constitutional mutation implicitly derogating a particular Constitutional provision will be explicitly recognized.⁷⁴

3.2. Binding and Non-Binding decisions

Further to the distinction between inter partes and erga omnes decisions; there is another relevant classification in Brazilian law with regards to the efficacy of judicial decisions about the (un)constitutionality of a provision. The 3rd Amendment to the Brazilian Constitution, from 17 March 1993, established the rule that the decisions of the Federal Supreme Courts in ADINs and ADPFs have not only erga omnes efficacy, but are also binding upon all the judicial and administrative bodies of the Federal Union, the States and the Municipalities.⁷⁵

More recently, another Amendment to the Constitution empowered the Federal Supreme Court to enact, by a decision of two thirds of its members and after a series of reiterated decisions about a constitutional issue, a special type of sumulas with binding efficacy over all the bodies of the Justificary and all the administrative authorities of the Federal, State and Municipal Governments.

T hese sumulas differ from the ordinary sumulas of the Court and are explicitly named sumulas vinculantes (binding sumulas).⁷⁶ The Court may decide to issue these sumulas on its own authority, but any of the parties legitimised to bring an ADIN before the Court can also request the promulgation, revision and annulment of a binding sumula.⁷⁷ These sumulas, as their own nomen iuris reveals, are also strictly binding.

Finally, a recent Federal Law attributed binding effects not only to the decisions of the Federal Supreme Court in ADINs, ADCs and binding sumulas, but also to any decision in an ADPF.⁷⁸

The key difference between the erga omnes and the strictly binding effects of judicial decisions is that the latter means not only that a decision is applicable to all juridical relations, but also that there is a constitutional writ to guarantee the efficacy of such decisions. This constitutional writ is named “Complaint to Preserve the Competences of the Federal Supreme Court and the Authority of its Decisions” (henceforth Reclamagao, for the Portuguese words “Reclamagao para preservagao da Competencia do Supremo Tribunal Federal e Garantia da Autoridade de suas Decisoes).⁷⁹ Any party, in the course of any judicial or administrative dispute, may bring a Reclamagao before the Federal Supreme Court. The Supreme Court may issue constraint orders and annul any judicial or administrative decision of any body of the Republic.⁸⁰ In this case, and only in this case, there is a Brazilian equivalent of the technique of avvocamiento.

Here we have another situation where the Court clearly acts as a positive legislator. In effect, it can be argued that the efficacy of the binding decisions of the Supreme Court - and particularly the efficacy of a binding sumula - is equivalent to the general efficacy of the legislation, for these decisions - and especially the binding sumulas - establish the final interpretation of the Constitution.⁸¹

3.3. The scope of the binding effects of the decisions of the court

One of the most controversial problems of Brazilian Constitutional Law is whether it is admissible or not to challenge, by means of a Reclamagao, a judicial decision which violates not only the ruling of a binding decision of the Supreme Court, but also its “justifying reasons” (fundamentos determinantes). In a case decided in October 2003, the Federal Supreme Court decided that the writ would be admissible in order to compel the lower courts to respect not only the concrete order which pronounces the unconstitutionality of a particular provision or establishes that it shall be interpreted in a particular sense, but also the motivation or the ratio decidendi of a binding decision.⁸² Nevertheless, in spite of this decision of the plenary sitting of the court, there is still an ongoing discussion among its members over the thesis that the binding efficacy of a decision transcends the particular order laid down by the court and thus encompasses the reasoning or the general principles formulated to justify the conclusions of the court.⁸³ There is no objective method for determining the ratio decidendi of the decisions of the court in ADINs, ADCs and ADPFs.

One thing seems to be certain. It is very unlikely that the Court will attribute binding effects to a decision which pronounces the constitutionality of a norm even when the parties offer new reasons which were not submitted to the court in the previous judgment. If one is able to put forward arguments which were not discussed by members of the court in a previous case that pronounced the constitutionality of a rule, there is no reason to believe that this rule will be “protected” from a claim of unconstitutionality on grounds not yet analyzed by the Court. The erga omnes and binding effects of the decision are thus merely of prima facie character. A decision that holds that a law X is compatible with the Constitution for the set of reasons R can always be challenged on the basis of the reason r1 if this reason does not belong to the set R. To put i t more simply, a decision which establishes the constitutionality of a legal provision is valid with a sort of clausula rebus sic stantibus, since some social changes and newer juridical understandings of the matter analyzed by the court in the past might justify a change in the court's doctrine.⁸⁴

3.4. interpretative and reductive decisions by the Federal Supreme Court

In the speeches of the Ministers of the Federal Supreme Court, one can find express and recurrent refere nces to many interpretative and reductive decisions, such as the technique of “interpretation in accordance with the Constitution”, the

“pronouncement of partial unconstitutionality without textual reduction” and the pronouncement of “partial unconstitutionality with textual reduction.” These are all interpretative techniques of infraconstitutional laws. The decisions of the court in these situations have erga omnes efficacy and are absolutely binding upon lower courts, as long as they are pronounced within the context of an action whose decisions are characterized by these types of effects.

The interpretative decisions of the Federal Supreme Court, when issued in the context of abstract and concentrated constitutional review, are strictly binding upon all judicial and administrative

bodies.³⁹

A pronouncement of “partial unconstitutionality with textual reduction” is perhaps the most frequent of these methods. If a legislative provision, in the same paragraph or sentence, contemplates two or more alternative hypotheses, it might be the case that only one of them violates the Constitution. The unconstitutionality of the provision is partial because among the multiple facts covered by the abstract description of the norm, only a few make this norm unconstitutional. A reductive decision is one which “eliminates part of the linguistic-uncontroversial core of the area of application of a norm.

»86

A pronouncement of unconstitutionality with textual reduction is thus one which eliminates some expression from the wording of a legal provision. In an abstract formula, the pronouncements of unconstitutionality with textual reduction of a norm can be described thus: “If a normative sentence S contemplates, in the conditions of application of the norm N, the facts C1, C2, ...and Cn; and if Cn is considered to be incompatible with the Constitution, the court may revise the sentence C by eliminating Cn from the set of the conditions of application of N.” To quote an example, the Federal Supreme Court

pronounced the unconstitutionality “with textual reduction” when a Federal Law regulating the activity of advocacy established that every advocate had “professional immunity” and thus “his speeches and manifestations in the exercise of his professional activities, either in or out of court”, did not amount to the crimes of “injury, defamation or contempt.”⁴⁰ It was held that this provision was unconstitutional while it immunized lawyers from being persecuted by the crime of contempt. The court held that if lawyers were excluded from the scope of the criminal provision which punishes the “contempt of court”, the autonomy of the courts and the authority of their decisions would be seriously endangered, and therefore pronounced the unconstitutionality of the expression “and contempt”, albeit keeping in force the rest of the legislative provision.⁴¹ In the same case, many other provisions of the same statute were reduced in the same way, for the Court held that the advocates were immunized to such extent that these privileges could not find a justifying reason and violated, among others, the principle of equality before the

law.⁴²

A “pronouncement of partial unconstitutionality without textual reduction", in turn, takes place when the legislative provision violates the constitution if interpreted in its literal or ordinary meaning. The court interprets a particular expression of the statute in a restrictive way in order to eliminate from the abstract norm that can be derived from the wording of the provision any sense which would collide with the Constitution. Instead of a principle of constitutional interpretation, this is considered to be a principle for the interpretation of the infraconstitutional legislation. Its function is to preserve the presumption of legitimacy of the legislation while avoiding maintaining in force unconstitutional readings of a statute.

Even though some scholars attempt to differentiate the “pronouncement of partial unconstitutionality without textual reduction” from the “interpretation in accordance with the constitution”, from the pragmatic point of view there is no difference between the two of them. To use a Kelsenian category, we can say that in both cases the court eliminates from the “frame” which defines the possible meanings of a lower-level norm those meanings that would make it incompatible with the constitutional norm which provides the basis of its validity. Whenever it is semantically possible, the court should interpret an infraconstitutional norm in a way that avoids the annulment of such norm.

These methods of legal interpretation might, however, represent a threat to the accepted principle that the Court should always act as a negative legislator. When interpreting a statute “in accordance with the constitution”, the court recognizes and reasserts that it is strictly forbidden to extend the scope of a legal provision in such a way that would create a general norm not established by the law-giver. As the court defines it, the interpretation “in accordance with the Constitution” is a “technique of constitutional review which can not lead to a particular interpretation that falls outside of the range of hermeneutic possibilities” left by the text a normative statement.⁹⁰ It is inadmissible to apply this interpretative method whenever it is impossible to choose, among the possible meanings of the infra-constitutional norm, one which would eliminate the unconstitutionality. One can not avoid the pronouncement of unconstitutionality when the meaning of the norm is undisputed.⁹¹

3.5. The temporal effects of the decisions on constitutional matters (on direct actions of unconstitutionality)

In Brazilian law, judges and courts generally consider themselves bound by the principle stated by Justice Marshall in Marbury vs. Madison: as a rule, unconstitutional laws are null and void and of no effect.⁹² The influence of the early decisions from the Supreme Court of the United States over the development of judicial review in Brazil is remarkable, probably due to the influence of the ideas of one of the greatest Brazilian jurists of all times, Rui Barbosa, who was a strong voice in defense of the civil rights in the country and one of the architects of the historical model of constitutional review.⁹³

As it was held in a relatively recent case, “the natural order of things” directs itself towards the view that a decision pronouncing the unconstitutionality of a norm retroacts to the date of the issuance of the norm considered to be unconstitutional.⁹⁴

Had the law-giver not passed statutory provisions explicitly authorizing the Supreme Court to lay down manipulative decisions, probably there would still be some resistance from the community of jurists against decisions with merely ex nunc or prospective efficacy.

The current law is that decisions delivered in abstract and concentrated forms of judicial review of the constitutionality of laws normally have ex tunc or retroactive effects. Nevertheless, the Supreme Court may restrict the effects of the pronouncement of unconstitutionality of a law in order to deliver ex nunc or pro futuro decisions or even to determine that the pronouncement of unconstitutionality will produce effects only after a deadline to be set by the Court. The Court must comply with the following requisites while delivering such manipulative decisions: (i) there must be reasons of legal certainty or of (ii) exceptional social interest and, apart from that, (iii) the restriction or the exception to the retroactive efficacy of the decision must be established by a vote of at least two thirds of the members of the Court (in its plenary sitting).⁹⁵

There is no doubt, therefore, that manipulative decisions with regards to the temporal efficacy of the pronouncement of unconstitutionality of a law are admitted in the Brazilian legal system. Nevertheless, it is a consensus that the courts should manipulate the temporal effects of the decision of unconstitutionality not on the basis of purely pragmatic reasons, but rather on strictly juridical reasons.

3.6. The temporal effects of the decisions on constitutional matters (on diffused decisions on constitutional matters)

In spite of the lack of an express provision authorizing the courts to restrict the retrospective efficacy of the decisions of unconstitutionality in diffused and incidenter tantum judicial review, the Supreme Court has broadened the scope of the permission established by article 27 of Law number 9.868 of 19th November 1999. As a matter of fact, this provision, which authorizes the Court to limit the retrospective efficacy of decisions of unconstitutionality, is remarkably similar to the constitutional provision which, in the Portuguese Republic, authorizes the same measures.⁹⁶ The same order of questions that were raised in Portugal is now opened to the Brazilian debate. The court decided thus to rely on the doctrinal interpretation dominant in that country. In interpreting the Portuguese constitutional provision, Prof. Rui Medeiros claims that the article 282nd (4th) of the Portuguese Constitution applies not only to concentrated constitutional adjudication, but equally to diffused judicial decisions of unconstitutionality of statutes.⁹⁷

In consonance with this view, there are some decisions of the Federal Supreme Court admitting decisions pronouncing the unconstitutionality of laws with purely prospective efficacy even in th e diffused and concrete forms of constitutional jurisdiction.⁹⁸ The pronouncement of unconstitutionality in concrecto can be limited if “another constitutional principle justifies the denial of the application of the principle of nullity”, that is, of the general rule that decisions of unconstitutionality have retrospective efficacy.⁹⁹

In some exceptional cases, the Court has even admitted the so-called pronouncement of unconstitutionality without annulment of any concrete acts, on the grounds that the retrospective decision of unconstitutionality would itself violate the constitution to an extent even greater than would the maintenance of the unconstitutional acts (whose effects, in the particular case, could not be undone without serious losses to a large proportion of the society).¹⁰⁰

3.7 . Constitutional mutations

Constitutional mutations are also admitted in the jurisprudence of the Federal Supreme Court. A “mutation” is understood as a change in the interpretation of a constitutional provision, the meaning of which is altered in spite of the maintenance of the same wording of the Constitution. If it were not for the binding efficacy of some constitutional decisions, there would be nothing special about “constitutional mutations” to distinguish them from the general practice of overruling . Be cause of the general effects of such changes, in some very exceptional cases the Court has applied the technique of “prospective overruling” in order to avoid retrospective changes in the law that would cause social disruption.¹⁰¹ The general rule, however, is to avoid this technique and not to apply it unless there are very strong reasons advanced by the parties.

CONCLUSION:

THE IDEA OF THE NEGATIVE LEGISLATOR

The analysis of the Brazilian case law in this report has shown that the idea that the Supreme Court should act as a “negative legislator”, and not as a “positive” law-maker, is deeply entrenched in the ideology of the Brazilian legal system. We have seen, however, that as a matter of fact there is a significant range of situations in which the Court does act as a positive legislator in spite of its official discourse. Is this a contradiction? How should we reconstruct the ideal and the factual interpretations of the thesis that courts are “negative legislators”, with particular reference to the Brazilian Supreme Court?

In our view, the idea of the negative legislator can be reconciled with the normative powers held by the Constitutional Court. When the Court expressly asserts that its competences do not authorize it to lay down general norms and thus act as a “positive legislator”, it is making a normative claim and establishing for itself a general obligation to respect the authority of Congress. This obligation is connected to an ideal aspect of the judicial practice. This aspect can be characterized as a “regulative ideal.” The function of this ideal is the same function of the ideal assumption that there is always a “correct answer” to any legal dispute. This idea, suggested by Ronald Dworkin in Law's Empire, is relevant to legal practice because it amounts to the existence of an “interpretative principle” which requires judges to justify their decisions in the best possible way, “as if” there was always a single correct answer.¹⁰² It is a duty to judicial authorities which commands them to seek for the “correct answer” and to struggle to justify their decisions in the most rational way within the possibilities opened by the law. The ideal of the negative legislator has an analogous function. In effect, it is a theoretical construction closely connected to Hans Kelsen's strong democratic convictions. In the activity of adjudication, the judge should concretize the Constitution by individualizing its norms.

The idea of a negative legislator is by no means incompatible with the idea that the courts have authority to create concrete norms on the basis of the general norms laid down in the Constitution. To the Justiciary it is expressly recognized authority to enact interpretative norms in order to assure the impartial application of the Constitution. One could never question, however, that in several occasions this normative powers amount to a (partial) decentralization of the legislative function. Even though the norms produced by the judges are still concrete and individual if compared with the broad principles incorporated in the Constitution, they still hold a great deal of generality. Kelsen would hardly doubt that.

Nevertheless, the ideal of the “negative legislator” remains plausible. This ideal is a normative claim which establishes that judges should respect the competences of the ordinary legislator and adopt a sort of “judicial self restraint”, as long as it is possible to do so. If this is the correct interpretation of the idea that judges are “negative legislators”, then the Brazilian Federal Supreme Court is strongly committed to this ideal, in spite of its general normative competences.