On the Unconstitutionality of the Central Bank’s Autonomy: The Case of Brazil ()
1. Introduction
As is well known, the autonomy of Central Banks is one of the main pillars of the mainstream approach in economics, whether following a general neoclassical or new-classical framework1. To support this recommendation, they rely on some fundamental premises, such as the existence of a potential GDP—based on a given technology that remains unchanged—and the dichotomy between the real and monetary sectors of the economy, that is, the neutrality of money.
Based on these highly questionable assumptions2, they claim a monopoly over the scientific nature of economic analyses, as if all other approaches were less scientific and more ideological. For this reason, they explicitly advocate for economic matters to be removed from political debate and placed in the hands of experts who align with the economic approach they present.
With this set of fallacies, they have, whether intentionally or not, supported numerous initiatives over the past decades around the world that threaten the normative core of the Democratic Rule of Law, embodied in ideals such as popular sovereignty, fundamental rights, and the separation of powers.
One of these many initiatives was Brazil’s Complementary Law 179 of 2021, which grants autonomy to the Central Bank of Brazil.
The present text aims to discuss the autonomy of the Central Bank of Brazil from the perspective of constitutional law and constitutional theory. In other words, it seeks to question if this autonomy, as it is currently legally structured in Brazil, is or is not compatible with the normative framework of the 1988 Constitution. More directly, it involves questioning, once again, the constitutionality of Brazil’s Complementary Law 179 of 2021.
The discussion begins by examining the relationship between self-determination in modern law and the problem of legal interpretation. It then addresses the nature of economics as a social science and, more directly, the legal structure of popular sovereignty, internally connected to the exercise of fundamental political rights and the separation of powers. Finally, it explores the institutional duty of the Brazilian Federal Supreme Court in a case such as this and the possibility of the Court revisiting its previous decision on the matter.
The theoretical and methodological framework adopted is Jürgen Habermas’s discourse theory of law and democracy. The main theoretical advantage of this framework is that it highlights the internal connection between law and democracy within the context of modern law: this is the core argument against the autonomy of the Central Bank of Brazil presented here.
2. Self-Determination in Modern Law and the Problem of Interpretation
Modern law is founded, ultimately, on a single fundamental premise: self-determination. As an element of what Jürgen Habermas refers to as the “normative content of modernity” (Habermas, 1987: pp. 336-367), this self-determination can be summarized by the idea that no one should be compelled to do or refrain from doing anything except by virtue of a norm that they impose upon themselves.
In complex societies, such as modern societies, this idea can only be realized through a series of mediations. More immediately, a decision made by a judge in a specific case must, in order to preserve its legitimacy, be understood by the parties involved—and by society at large—as a decision that enforces the law in the drafting of which the convicted or acquitted parties in each case also participated.
For this understanding to be even minimally plausible, it is necessary that fundamental rights and political participation procedures be institutionalized, ensuring individuals have the effective ability to determine who will be responsible for drafting laws. These laws, in turn, only acquire their legitimacy if they demonstrate coherence with the highest-ranking positive norm: the Constitution.
This complex chain of legitimation, which could be further detailed into numerous stages and explored through endless intricacies, is undoubtedly one of the elements that make the Democratic Rule of Law such a challenging institutional arrangement. After all, drawing once again from Habermas’s work, this chain of legitimation reveals a constant tension between facticity and validity—that is, between the everyday realities of political and judicial processes and the normative expectations these very processes inherently carry (Habermas, 2021).
Be that as it may, if the Democratic Rule of Law remains, for us, an achievement to be preserved, there is no alternative: only this legitimizing chain, with all its inherent complexities, ensures that popular sovereignty is more than a mere illusion or an idealistic naivety. And there can be no Democratic Rule of Law without popular sovereignty (Habermas, 2018: pp. 419-434).3
It is not by chance, therefore, that the Constitution assumes such a significant role in the political and legal disputes of modern life. It is with the Constitution that the self-determination of a sovereign people is initially manifested, and it is in the Constitution that the pathways for maintaining this self-determination are defined—guiding it through the complex chain of legitimation until it reaches even the simplest judicial decisions in individual cases. In this way, the Constitution ensures the equiprimordiality between private and public autonomy—co-original dimensions in which autonomy, or self-determination, is divided within the legal-political sphere (Habermas, 2006: pp. 113-128; Cattoni de Oliveira, 2023a: pp. 72-80).
An additional obstacle arises at this point. A Constitution cannot, nor is it able to, foresee and explicitly establish everything that reflects the sovereign will of the people who enact it. It is up to that same people, on the appropriate occasions and in accordance with the constitutionally established procedures, as well as to the constituted powers likewise regulated by the Constitution, to unfold that sovereign will throughout the years of the Constitution’s life.
To “unfold,” in this context, is nothing other than to interpret. However, this does not mean confining the Constitution within the strict boundaries of the Judiciary. Even a misconception of hermeneutics, like that of Hans Kelsen, recognized that the task of interpreting the law extends beyond the actions of judges and representatives of the essential functions of justice (Kelsen, 2003). The judiciary interprets the Constitution, and so do the executive and legislative branches, as well as the population as a whole, through the various channels available for public participation, ranging from voting to freedom of expression.
Naturally, these interpretations occur in different ways, serve distinct functions, have varying effects, and materialize through diverse means—a judicial decision, a law, a public policy, or an electoral choice. Nonetheless, within the framework of the Democratic Rule of Law, all these instances must be understood as moments of constitutional interpretation.
The prominence that judicial interpretation has gained in recent decades in Brazil has overshadowed the constitutional interpretative role played by the other two Branches of government. This, beyond being a serious issue in itself, also renders the very understanding of judicial interpretation much more incomplete—if not outright flawed—particularly regarding its possibilities, duties, and limits.
Although there is no such thing as a purely “literal interpretation” of any text, the 1988 Constitution establishes a series of obligations and prohibitions in provisions that leave little room for interpretative efforts. The Federal Court of Accounts, for instance, must issue a prior opinion on the annual accounts submitted by the President of the Republic within 60 days from their receipt (Art. 71, I). Similarly, political parties are prohibited from receiving financial resources from foreign entities or governments, as well as from being subordinated to them (Art. 17, II).
There is, however, a significant portion of constitutional provisions that invite a deeper hermeneutic exercise. Building a free, just, and solidary society, ensuring national development, eradicating poverty and marginalization, and reducing social and regional inequalities are fundamental objectives of the Federative Republic of Brazil (Art. 3, I, II, and III). As constitutional mandates, they hold normative force. They are not mere statements of intent; they are binding norms that must be fulfilled, and non-compliance entails some form of sanction. But how should they be fulfilled? The pursuit of full employment is one of the guiding principles of Brazil’s economic order. It is a binding norm that must be respected. But how should this be achieved?
In cases like these, the centrality of economic policy decisions becomes clearly evident. And this is where our problem begins.
3. Economics as a Social Science, Popular Sovereignty and the Separation of the Powers
Economics is typically a branch of the social sciences. Even someone like Frank Knight, one of the founders of the Chicago School and a professor who influenced figures such as Milton Friedman and James Buchanan, boldly acknowledges this in his critique of positivist claims within the field of economic science:
“The fundamental propositions and definitions of economics are neither observed nor inferred from observation in anything like the sense of the generalizations of the positive natural sciences, or of mathematics, and yet they are in no real sense arbitrary. They state “facts,” truths about “reality”—analytical and hence partial truths about “mental” reality, of course—or else they are really “false.” Economics and Other Social Sciences deal with knowledge and truth of a different category from that of the natural sciences, truth which is related to sense observation—and ultimately even to logic—in a very different way from that arrived at by the methodology of natural science. But it is still knowledge about reality” (Knight, 1940: pp. 5-6).4
Asserting that economics is a social science in no way diminishes its scientific nature, nor does it prevent decisions based on its scientific principles from being understood as technical decisions. As Knight puts it, “the fundamental propositions and definitions of economics…are in no real sense arbitrary” (Knight, 1940: p. 5).
However, this technical-scientific character requires acknowledging its specificities. As with any other social science, the level of consensus is significantly lower than that typically found in the natural and exact sciences. This discrepancy is linked to the distinct types of postulates inherent to these fields and the corresponding methods available for their testing.
As a consequence, in general, a given economic theory may be valid and may suggest specific technical interventions; however, it will be just one among several equally valid formulations from a scientific perspective—even if these alternative theories propose the exact opposite of the initial economic formulation and advocate for entirely different technical solutions.
The Central Bank of Brazil is one of the topics where such divergence becomes evident. Serious research, equally supported by empirical evidence, leads to opposing conclusions. The Brazilian Federal Supreme Court (FSC), when ruling on Direct Unconstitutionality Action (DUA) 66965, rightly acknowledged this: “It is an undeniable fact that the issue of the Central Bank’s autonomy divides opinions.” (Supremo Tribunal Federal, 2021a: p. 3) However, the Court erred in concluding that this autonomy is a matter “that does not fall within the scope of constitutional interpretation” (Supremo Tribunal Federal, 2021a: p. 3).
I insist that in the Democratic Rule of Law, constitutional interpretation extends beyond what is carried out by the courts6. Its proper understanding requires attention to the relationship between constituent power and constituted powers, as well as to the different forms that popular sovereignty assumes during the constituent moment and in the moments it reserves for itself during the validity of the constitutional order. This means also recognizing political-electoral processes, carried out in the full exercise of fundamental political rights, and legislative and executive decisions as acts of constitutional interpretation.
In more concrete terms, the choice of the general approach to the country’s economic policy is one of the most significant aspects of an election for the head of the federal Executive Branch. After all, it is primarily the Executive’s responsibility to lead this policy, albeit with the support and oversight of the Legislative Branch. A constitutionally sound reading of political-electoral processes must, therefore, view them as specific acts of constitutional interpretation: in light of the relevant constitutional norms, which political project does popular sovereignty consider, in each periodic exercise of suffrage, to correctly fulfill these norms?
This constitutional interpretation carried out by popular sovereignty through voting cannot be disregarded without undermining the very foundation that sustains the legitimacy of the Democratic Rule of Law. In 2022, the general approach to economic policy chosen by popular sovereignty was that represented by the current government. Preventing this government from implementing its economic policy strikes a fatal blow to that sovereignty and, consequently, let it be repeated endlessly also to fundamental political rights, constitutional democracy, and the Democratic Rule of Law.
It is precisely this kind of obstruction that has been demonstrated by the Central Bank’s actions over the past two years, particularly through its decisions regarding the economy’s benchmark interest rate. Since this opposition by the Central Bank to the federal government’s intended economic policy is only possible due to the autonomy granted to it by Complementary Law 179 of 2021, the conclusion is clear: the autonomy7 of the Central Bank, as it currently stands in Brazil, is unconstitutional8.
This conclusion, of course, requires further clarification. It is not being argued that choosing to grant autonomy to the Central Bank is necessarily illegitimate as an approach to economic policy. The initial petition in DUA 6696 errs by directly linking this autonomy to a disregard for the “constitutional goal of ensuring everyone a dignified existence” (Partido Socialismo e Liberdade & Partido dos Trabalhadores, 2021: p. 34). Beyond global institutional actors typically associated with the advancement of neoliberalism, such as the World Bank and the International Monetary Fund, there are many serious scholars and individuals committed to the pursuit of social and economic justice who conduct their intellectual work under the premise that Central Bank autonomy can serve as an important mechanism for achieving this goal in a stable and long-term manner9.
That is, there is nothing preventing a particular government, aligned with this economic conception and relying on the forecasts it offers, from choosing to grant autonomy to the Central Bank as part of its specific approach to economic policy, using regulatory instruments within the Executive’s sphere of competence. What is not constitutionally permissible, however, is to impose this autonomy externally on a government that does not share the same economic vision—a government for which the autonomy of the Central Bank is incompatible with the economic policy it presented when it submitted itself to the electorate. And this is precisely what has been happening in Brazil.
In other words, from a constitutional standpoint, the issue is not the Central Bank’s autonomy itself, but the fact that it is established through legislation. Thus, alongside the matter of popular sovereignty, the separation of powers comes to the forefront of the discussion.
Much of the debate within the Brazilian Federal Supreme Court during the ruling on DUA 6696 focused on whether Complementary Law 179 of 2021 was formally unconstitutional. The prevailing view was that there was no formal unconstitutionality, either because it did not require the exclusive initiative of the President of the Republic or because this requirement had been fulfilled10.
However, this is not the core issue. The fundamental question is to what extent the President of the Republic may or may not have the authority to initiate such legislation with this specific content, and, if not, to what extent the Legislative Branch can impose this content—the autonomy of the Central Bank—on the Executive.
The initial petition in DUA 6696 asks about this limit, arguing that it constitutes a “renunciation of competence” prohibited by the Constitution” (Partido Socialismo e Liberdade & Partido dos Trabalhadores, 2021: p. 20). This argument was not directly addressed by the FSC, and although framed differently from the terms used in the initial petition, it seems to me to hold the core of the correct answer to this case. Even if the head of a given government intends to grant autonomy to the Central Bank, and can legitimately do so in line with the economic policy they plan to implement, they cannot take advantage of a temporary parliamentary majority to establish this autonomy through legislation, as this would improperly bind future governments.
The argument that a future government could, if it wished, negotiate with the Legislative Branch to repeal the existing law or draft a new one is insufficient. Given the time such negotiations typically require, this would impose a disproportionate burden on the future government, which, in the extreme, could entirely undermine the implementation of its intended economic policy.
But isn’t the Legislative Branch also elected by popular sovereignty? Wouldn’t the drafting of a law always respect this crucial pillar of the Democratic Rule of Law? Not necessarily. In its constituent moment, popular sovereignty not only defines how it will continue to manifest itself during the validity of the Constitution it has established but also sets the boundaries within which the constituted powers—elected during the constitutional order—must operate.
Therefore, any actions by these powers that exceed the limits established by the Constitution violate popular sovereignty, even if those powers have been democratically elected. Reinforcing the need for a constitutionally sound interpretation of political-electoral processes, it is essential to understand that democratic elections select representatives who are bound to act in full compliance with the provisions of the Constitution.
The responsibility for conducting economic policy lies with the Executive Branch through its ministries and other agencies. The Legislative Branch plays a vital role in providing support, conducting oversight, and helping to establish the necessary legal framework. However, it cannot intervene to such an extent that it undermines or distorts the Executive’s authority over the conduct of economic policy.
Undoubtedly, it is the role of the Legislative Branch, with presidential sanction, to regulate financial, foreign exchange, and monetary matters, as well as financial institutions and their operations (Art. 48, XIII). Yet, this does not in any way imply that the Legislature can strip the Executive of its ability to define the administrative, concrete, and day-to-day aspects of economic policy. This is precisely the issue at stake when the question is whether or not to grant autonomy to the Central Bank through legislation.
This is the case because, in stark contrast to what was asserted by Justice Luís Roberto Barroso in the judgment of DUA 6696, the Central Bank is anything but a “neutral arbiter” (Supremo Tribunal Federal, 2021a: p. 52). Its decisions have direct—and often nearly immediate—impacts on fundamental matters within the jurisdiction of the Executive Branch, such as economic growth, social inequality, employment levels, credit supply, public debt, and inflation.
All this reasoning, of course, does not imply advocating for a blank check for the Executive Branch to implement any economic policy it desires. Once again, economic policy must be understood as a policy that interprets the Constitution and enforces its norms. An economic policy that fails to do so can and should be subject to oversight, depending on the specific case, by the Legislative Branch, the Judiciary Branch, and the population through the various forms in which civic engagement is possible.
Nevertheless, this possibility and necessity of oversight does not justify the prior constraint of a government to an economic framework and a configuration of the Central Bank that is entirely opposed to the economic policy it presented to popular sovereignty through the electoral process, above all when such an economic framework and Central Bank configuration represent merely one among several legitimately possible options, and the one endorsed by the government is likewise a valid alternative according to the principles of economic science.
4. The Institutional Duty of the Brazilian Federal Supreme Court
The mention of oversight raises the question of what role the Judiciary should play in the current scenario. In its ruling on DUA 6696, the Brazilian Federal Supreme Court erred not only by stating that, in substantive terms, the issue of the Central Bank’s autonomy does not fall within the scope of constitutional interpretation but also by concluding that this is the case because it is an “essentially political matter” (Supremo Tribunal Federal, 2021a: p. 3).
It is certainly important for the Judiciary to refrain from interfering in political matters. But it is equally crucial that the Judiciary does not neglect its institutional role by mischaracterizing it as an essentially political issue something that, in fact, is not.
Under the guise of an “essentially political matter”, framed as a dispute between supporters and opponents of the Central Bank’s autonomy lies a serious legal-constitutional issue concerning the violation of popular sovereignty and the separation of powers. Ideally, this violation would be addressed by the democratically elected Branches themselves through dialogue between the Legislative and Executive Branches. However, in the absence of such a path, if the Brazilian Federal Supreme Court is entrusted with the judicial dimension of constitutional interpretation, it must, in the face of this normative violation, declare the unconstitutionality of the provisions that currently make the Central Bank’s autonomy legally mandatory in the country.
The fact that the Brazilian Federal Supreme Court has already ruled on the matter in its judgment of DUA 6696 does not preclude further review. Although the causa petendi in Direct Unconstitutionality Actions is open—meaning that the FSC is not limited to the arguments presented in the initial petition—this does not imply that a previous ruling completely bars a new challenge to the same norm. A new challenge remains possible if it is based on a causa petendi distinct from that of the original petition and also different from the reasoning effectively employed by the FSC in its prior judgment.
The FSC itself, in fact, recognizes this possibility. An important precedent in this regard is Direct Unconstitutionality Action 5579 (Supremo Tribunal Federal, 2021b), also ruled in 2021, with Justice Cármen Lúcia as the rapporteur and a unanimous decision among the Justices. In that case, the Court explicitly stated that the declaration of the constitutionality of a normative provision does not preclude its re-examination, provided that the causa petendi in the two actions are distinct. In this case, the distinction lay between material unconstitutionality, alleged in the new action, and formal unconstitutionality, raised in the previous action.
Given that the causa petendi in a DUA is open, the only way to remain consistent with this principle and to permit the re-examination of constitutionality is if the FSC did not, in its earlier ruling, effectively address arguments of material unconstitutionality, whether or not they were presented in the corresponding initial petition.
An even more significant precedent is DUA 5081 (Supremo Tribunal Federal, 2015), ruled in 2015, with Justice Luís Roberto Barroso as the rapporteur and a unanimous decision. In the rapporteur’s vote, it is stated:
“The issue of the constitutional illegitimacy of the loss of mandate in cases involving majoritarian electoral positions, which is the subject of the present action, was not raised at any point, either in the initial petition, in the vote of the Rapporteur Justice, or in the other statements made in Plenary. Since the causa petendi in concentrated control of constitutionality actions is open, nothing would have prevented this issue from being addressed in DUAs 3.999/DF and 4.086/DF. However, this did not occur. 4. In such cases where this Court has not expressed itself on the specific constitutional issue, I believe it is appropriate to re-examine the norm previously deemed valid by the Court, especially when the analysis of the constitutionality of the normative act occurred only from a formal perspective. (...) It cannot realistically be assumed that the Court will anticipate all possible material unconstitutionality flaws in these hypotheses” (Supremo Tribunal Federal, 2015: pp. 9-10).
In this last citation, I added the italics emphasis on the word “especially”. It carries significant meaning: “especially” is not “exclusively”. Moreover, the strength of this precedent is even greater because the fundamental argument for the declaration of unconstitutionality in DUA 5081 was the violation of popular sovereignty.11
5. Conclusion
The arguments outlined here were not specifically discussed in the judgment of DUA 6696, as they were, for the most part, absent from the initial petition of that action. The small portion of these arguments that does appear, under the heading of “renunciation of competence” constitutionally prohibited, was not addressed in the ruling and, even then, is presented in that petition in a manner different from what is outlined above.
This, then, is the decisive point here: the arguments I offer in this paper were not specifically addressed by the Brazilian Federal Supreme Court. And they must be; it is essential for Brazilian democracy that they are. Once examined, it seems to me that their validity cannot be denied.
In recent years, we have witnessed the exponential growth of literature on the crises of democracy and constitutionalism12. The warnings presented in these studies are undoubtedly valuable. However, it is essential to understand that the authoritarian threat does not stem solely from so-called populist leaders or the reactionary political culture that often accompanies them, it also arises from our routine difficulty in accepting what constitutional democracy demands of us.
In the specific case discussed here, it demands something simple: respect for popular sovereignty, fundamental political rights, and the separation of powers in Brazil through the acknowledgment of the unconstitutionality of the Central Bank’s autonomy as established by Complementary Law 179 of 2021.
Acknowledgements
The author would like to thank Almir Megali Neto, Henrique de Queiroz, and Marina Pompermayer for their valuable comments and suggestions on the first manuscript of this text.
Funding
This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior—Brasil (CAPES)—Finance Code 001.
NOTES
1For an introduction to this approach in contemporary economics, focused on macroeconomic issues. See: Romer, 2011.
2For an introduction to alternative economic approaches, generally known as heterodox in contrast to the orthodox approach of neoclassical and new-classical schools, see, also with a focus on macroeconomic issues. (Blecker & Setterfield, 2019)
3In more precise technical terms, whose specifications make sense within the overall trajectory of his work, Jürgen Habermas speaks of the “internal link between Rule of Law and democracy” and the “mediation between popular sovereignty and human rights”.
4Italics added by me for emphasis.
5For a detailed criticism of the decision rendered in this judgment. See: Bercovici & Morais, 2023.
6See the seminal reflections of Menelick de Carvalho Netto (Carvalho Netto, 2020; Carvalho Netto, 2021; Carvalho Netto & Scotti, 2011). See also Cattoni de Oliveira, 2023b.
7This autonomy encompasses the fixed terms of the Central Bank’s board, which do not coincide with the term of the President of the Republic, but it is not limited to this aspect. For details regarding the scope of such autonomy and the mechanisms that implement it. See: Bercovici & Morais, 2023.
8For a similar conclusion. See: Streck, 2023.
9An example is the research conducted by Lízia de Figueirêdo, Full Professor at the Department of Economic Sciences at UFMG and Cedeplar. Although based on neoclassical premises and employing models that ultimately trace back to the paradigm introduced by Robert Solow, it is impossible to deny her concern and sincere engagement with the issue of poverty. For more details, see her profile in the Academia.edu: https://independent.academia.edu/LiziaDeFigueiredo.
10See, again, the criticism presented by Gilberto Bercovici & Viviane de Morais (Bercovici & Morais, 2023).
11Another noteworthy case involving controversial models and policy choices related to economic issues, and raising questions similar to those addressed here, was Constitutional Amendment 95 of 2016, which established the so-called “fiscal spending ceiling” in Brazil. Unfortunately, following Constitutional Amendment 126 of 2022 and Complementary Law 200 of 2023, the Direct Unconstitutionality Actions 5658, 5680, 5715, and 5734 were dismissed without a ruling on the merits.
12See Meyer, 2021. My critical dialogue with this book can be found in Gomes (2022) and Gomes & Equipe de Pesquisa (2024). A comprehensive introduction to this literature on crises, especially the crises of democracy, is offered in Moreira (2020). My divergent position, which seeks to offer a different diagnosis of these crises, at least in the Latin American context, is available, among other texts, in Gomes (2021).