Regulatory Oversight for Better Regulatory Governance: The Role of the Federal Court of Accounts Following the Mining Dam Collapses in Brazil ()
1. Introduction
Democracies in developing countries, such as Brazil, are recognized for experiencing specific constraints in their public governance (O’Donnell, 1994) . It is from this very realistic perspective that the Brazilian Federal Court of Accounts (FCA) has over the years taken on a prominent role in overseeing bureaucratic activities. With over 130 years of existence, the FCA has broad constitutional authorization to oversee the expenditure of federal public resources, both from a legal regularity perspective and in assessing the legitimacy of these expenditures. Since the promulgation of the Brazilian Constitution of 1988, the FCA has received an expansion of its formal competencies to now encompass the assessment of the legitimacy and economic rationality of public spending (Dantas & Dias, 2018) .
This oversight by the FCA can be interpreted through various theoretical lenses, which will be briefly discussed. But the fact is that in Brazil, the FCA plays a very significant role in overseeing regulatory activities. In practice, FCA’s oversight has been likened to what international experience identifies as “Regulatory Oversight Bodies”, which are meta-regulatory structures with institutional designs that vary from country to country. The mission of these entities is to oversee, control, and coordinate the state’s regulatory activity, imposing a burden of justification for the acts or omissions of regulators and challenging, or in some cases vetoing, the policies of entities that do not meet the pre-established burdens. Regulatory Oversight Bodies are directly related to governmental programs for regulatory improvement. Therefore, they can also provide guidance, training, and support for the entities they oversee, promoting administrative expertise and efficiency.
This article aims to present the contribution of the Brazilian Federal Court of Accounts as a regulatory oversight body targeting improvements in regulatory governance, which is a new role, complementary to the formal control of public expenditures. The specific scenario that will be used as an explanatory cut-off for this new role over regulatory governance will be the supervision and eventual control exercised by the FCA over the Brazilian National Mining Agency (NMA). The time frame will encompass 12 years of public audits, allowing the sample to highlight the FCA’s behavior both before and after the disastrous mining dam collapses in 2016 and 2019 in the Brazilian state of Minas Gerais, which have become the biggest environmental disasters of recent decades in Latin America. The article, which is eminently descriptive rather than normative, will also show that the changes in the Brazilian Court of Accounts arise from demands of the regulatory community itself, so institutional learning relates to contributions made by experts in public auditing and control. In this specific sense, the article demonstrates that it would be possible to claim that the FCA has been exercising a responsive oversight, with improvements in its own structure and operation.
The article is organized into three chapters. The first addresses the theoretical foundations and presents some empirical notes about Supreme Audit Institutions at the international level and the Brazilian Federal Court of Accounts at the national level. The second chapter delves into the rationale for addressing Courts of Accounts as Regulatory Oversight Bodies. The third chapter is eminently empirical, demonstrating the potential of a Regulatory Oversight Body regarding the institutional governance of administrative agencies. The structuring of the theoretical framework was mainly based on literature about public accountability, responsive regulation, and regulatory oversight.
2. Supreme Audit Institutions and the Brazilian Court of Accounts
This section aims to elucidate the role of institutions specifically designed for the external oversight and control of the State. To do this, it’s essential to outline the core concepts and theoretical foundations that, when applied, justify the existence of institutions like the Federal Court of Accounts and their international counterparts.
The FCA aligns with a globally recognized institutional design termed Supreme Audit Institutions or Superior Audit Institutions (SAI). These refer to high-ranking organizational entities whose primary goal is the external oversight of the State (OECD, 2016) . While SAIs are governmental bodies, they conduct external supervision as distinct entities with their procedures, activities, and resources. They are not part of the structures they oversee, and their mission is to monitor, assess, and, in some cases, correct actions that deviate from specific standards if necessary. Given these attributes, it’s widely accepted that external oversight requires a high degree of independence and protection against potential pressures from audited entities, which may occasionally adopt an adversarial stance.
SAIs represent the contemporary institutional embodiment of the persistent drive to curtail arbitrary use of scarce public resources. Their role ensures that governments remain accountable and use the public budget to benefit citizens optimally. In republican regimes in which governance is realized through political representation, officials, whether administrators, legislators, or judges, are prohibited from acting in self-interest. Given their role as representatives and their management of collective assets and interests, these officials act on behalf of the public. Any action is only legitimate if persuasively aligned with the broader community’s interest.
Beyond a mere abstract notion, this defining trait of republican governance systems underscores the necessity for a public entity that continuously monitors the nation’s political elites. Political science has, for some time, recognized that periodic elections are, at best, a flawed mechanism to ensure those in political power act in their constituents’ best interests (Przeworski, Stokes, & Manin, 1999) . As Andrew Arato astutely noted, nothing in the political representation model inherently prevents chosen representatives from betraying the very group they are meant to serve (Arato, 2006) . Elections by itself are no guarantee of republicanism.
Hence, republican principles advocate for the implementation of proactive mechanisms to track state functions. If power is derived from the people, then the people must consistently oversee their representatives, who, after all, are fallible humans, not selfless entities (Madison, 1788) . This perspective offers a realistic view of representative democracy, highlighting potential governance challenges, especially those stemming from human imperfections and unequal access to political power (Dahl, 1998) .
In political economy discussions, issues related to political representation often fall under the agency theory (Fearon, 1999) . In this paradigm, an “agent” (the elected politician) makes decisions on behalf of a “principal” (the voter). Complications arise when the rational politician might not always act in the voter’s best interests, even if those choices are publicly touted as being for an abstract public interest.
In this context, characterized by healthy democratic contestation or even profound mistrust, it becomes essential to emphasize institutional channels that allow citizens to ensure republican vigilance (Pettit, 2002) over the proper functioning of state structures. This is to prevent budgetary or asset misappropriations, as well as the malicious manipulation of strategies guiding public interest activities.
The increasingly evident shortcomings of delegative democracy have led to more sophisticated reflections on how to ensure this republican vigilance. With the partial erosion of popular voting as a mechanism to ensure governmental responsiveness, various institutional mechanisms have emerged in societies. These mechanisms, with the declared purpose of constantly scrutinizing the exercise of political power, serve as tools of monitoring and institutional restraint on national leaders. In political science doctrine, these mechanisms are typically studied under what is commonly termed public accountability (Bovens, Goodin, & Schillemans, 2014) . For the purposes of this study, the definition of accountability in the public sector is as follows: the legal or political ability to ensure that public agents, whether elected or not, are responsible and responsive in their actions, subjecting themselves to demands for justification and information to the recipients of this assurance regarding their positions and the decisions they make (Pettit, 2009) . Additionally, they are held to judgments either based on their good performance or due to deviations or misconduct, ultimately resulting in the application of sanctions.
Two central aspects can be drawn from the concept of public accountability. The first refers to the idea that, at least in societies that value republican vigilance, it is the continuous duty of the leaders in power to provide credible information and justify it to the governed, so that they are subject to judgment. The second refers to the idea that there must be political or legal consequences that guarantee the fulfillment of the obligations or expectations assumed. In the language of political economy, this second aspect refers to the provision of sufficiently strong incentives (punishments, rewards) to the rational politician to ensure compliance with the desired behavior. Accountability without consequences is not accountability (Schedler et al., 1999) .
With that said, it’s important to understand how this abstract concept can be operationalized in social practices, at least generally speaking. In political science, literature organizes mechanisms of public accountability into dimensions or axes, namely: vertical, horizontal, and diagonal.
The mechanisms of the vertical axis represent the most fundamental public accountability methods. They pertain to direct control by society and are often associated with electoral mechanisms, plebiscites, and referendums. But they extend further: they also allude to political pressures stemming from the dissemination of information by society’s communication mediums. The press, for example, is a vital instrument of public accountability. It represents civil society organized to spread information about the government and demand stances. Due to specific dynamics, vertical axis mechanisms are explored through subdivisions between electoral and social accountability (Smulovitz & Peruzzotti, 2006) . From a public administrator’s perspective, these pressures mainly relate to political consequences.
In turn, horizontal public accountability mechanisms aim to address the weaknesses of the vertical axis. In this axis, relations are conducted within the state itself and are formally institutionalized within the legal framework. Horizontal accountability is executed through the structuring and effective training of state organizations that are legally tasked with overseeing the routines and sanctioning unlawful actions or omissions by public agents or other state institutions (O’Donnell, 2007) .
Horizontal mechanisms harken back to and derive from the traditional checks and balances system. However, these constitutional checks and balances are designed for the mutual control at the pinnacle of the State’s powers. They are often implemented intermittently and reactively, and because they lead to sporadic political friction among high-visibility state decision-making bodies, they are often seen as high-stakes institutional disputes. They might not be suitable for various situations, as they can sometimes be perceived as a direct affront to the harmony between powers and institutional stability (Mainwaring & Welna, 2003) .
Thus, institutions with specific public accountability missions have subtle differences compared to the Judiciary. The latter operates in a distinct dynamic, more reactive than proactive, with a broader mission concerning various societal ideals of justice and not limited to public accountability. On the other hand, institutions focused on specific horizontal accountability tasks are more proactive and less reactive, designed expressly to handle complex issues involving information auditing and accountability. Consequently, they are engineered for constant interactions with bureaucracy and can demand information that would otherwise be provided incompletely, imperfectly, or deceptively. Additionally, the vectors and action criteria of accountability institutions tend to be more technical than political, and the possibilities for implementing their decisions are outlined in the legal framework to foster legitimate expectations. The continuous focus on highly technical themes concerning financial resource utilization tends to diminish the impression of political disputes within the state, which, it should be noted, might not necessarily be the case.
Institutions of the horizontal axis, therefore, serve as continuous and specialized supervision mechanisms over public management, legally able to produce adverse consequences for detected misconduct. This capacity to generate such consequences is an indispensable element for an institution to be considered a genuine accountability mechanism, even if the consequences are achieved indirectly, such as when the overseeing institution has the capacity to mobilize the competent sanctioning authority (Mainwaring & Welna, 2003) .
Lastly, the diagonal axis refers to the mutual interaction between the vertical and horizontal axes, preventing state control instances from becoming isolated. Diagonal accountability is realized through programs or institutions capable of promoting citizen engagement in the control activities performed on the horizontal axis. In essence, the diagonal dimension is guided by citizen participation, by ensuring access to information, transparency, and by the interaction between these citizens and control institutions in a logic that expands democratic channels. In the diagonal axis, the concept of accountability seems to emphasize the understanding that control activity is not an end in itself but a tool intended to give effect to republicanism (Pelizzo & Stapenhurst, 2011) .
Diagonal accountability seeks to hold control institutions accountable to republican oversight and to streamline judgments made by other entities. This implies that the control mechanisms within the horizontal dimension must effectively communicate their information to the broader public and other specialized entities, rather than just a limited technical or political elite. It’s essential for the general populace to grasp and access the intricacies of the horizontal interactions. While the diagonal approach primarily emphasizes widening citizen participation in public resource management rather than direct oversight, there’s a widely accepted notion that these control mechanisms should use their legal authority when needed to diminish the ambiguity often overshadowing state activities. Essentially, it broadens the scope of collective vigilance. This mutual engagement between the citizenry and multiple control bodies enhances coordination, amplifies efficacy, and boosts the legitimacy of the oversight process.
Considering the various mechanisms through which accountability is achieved, it becomes even more evident that there’s a myriad of state and non-state actors involved in scrutinizing collectively owned activities. There’s a dense network of controls overseeing state actions, each with its operational nuances. Specifically, regarding Superior Audit Institutions, like the Courts of Accounts, it can be inferred that they are institutions firmly linked to promoting public accountability in both the horizontal and diagonal dimensions. This is because they are institutions established in the legal system, usually rooted in the Constitution, with the explicit mission of monitoring governmental activities. They possess the legal authority to investigate any potential discrepancies, and to foster dialogue and mobilize various entities of public accountability or communities of experts through their responsibilities.
The international experience shows that there’s a significant variation in the organization and functioning of different Supreme Audit Institutions around the world. However, this study could not find any references to another Supreme Audit Institution with autonomous responsibilities as broad as the Brazilian Court of Accounts, which currently incorporates various elements from both the Anglo-Saxon and Roman-Germanic models (Speck, 2011) . Thus, these two models will be the focus.
From a broad perspective, it’s important to recognize that any typology given for SAIs’ models will hardly suffice to cover the present complexity they have assumed in modern states. Originally, SAIs had roles with distinct characteristics and dynamics, but nowadays, most combine several of these roles, resulting in intricate institutional arrangements.
With these considerations in mind, instead of delving into a functional typology that might get lost in such complexity, Speck (2000) suggests understanding SAIs based on two primary concerns that historically motivated their creation: 1) the managerial concern of efficiently managing public resources, and 2) the liberal apprehension about limiting Executive Power. These concerns are very distinct, leading to different dynamics of control. Although both concerns are present in modern states, they appear with varied intensity, at different times, materialize in different institutional formats, and relate in a specific manner in various political systems. In this light, it’s possible to construct two ideal models of SAIs around these fundamental concerns, thus explaining the origin and dynamics of these institutions.
The first motivation, the managerial concern for good governance of public resources, is associated with the rationalization process of administration that Max Weber discussed in his writings. The interest in maintaining a specialized institution for auditing administrative processes and government programs arose within the Executive branch itself, aiming to scrutinize the use of its financial flow. By examining the data, the goal was to modernize and optimize administrative action, ensuring support, power, and shielding from potential criticism. When SAIs focused on good management arose within the Executive, they merely gathered and processed information to aid decision-making. In this configuration, the control parameter pertained only to the performance of administrative action.
The second motivation, the liberal concern to limit the power of the Executive and hold public agents accountable for abuses, traces back to representative bodies that approved and allocated resources extracted from the community. The precursors of SAIs more aligned with this liberal concern are the Parliamentary Committees specialized in oversight and control tasks. These committees later delegated their tasks to external advisory institutions or assigned them to specialized courts. Over time, these institutions gained more autonomy, even reaching positions of significant autonomy from the Parliament. In this institutional setup, the control parameter originally referred to compliance with laws generated in parliament.
These two fundamental concerns can be associated with the two main general models of SAIs in developed countries: 1) the Anglo-Saxon model of Comptrollerships, and 2) the French model of Courts of Accounts. Commonly, all models value independence in their roles, serving as protection from potential external pressures. However, internal organization and control dynamics vary substantially based on the guiding principles of each institution (ultimately, the two fundamental concerns mentioned above).
In general, the role of Comptrollerships is to assist in overseeing the Executive, ensuring that top authorities are adequately equipped with the increasingly complex information needed for decision-making. Their findings typically come in the form of guidance to other powers. Their staff often consists of professionals trained in applied social sciences, particularly economics, management, and accounting. Their primary concern is improving state performance. The parameters to be evaluated are usually related to efficiency, effectiveness, and efficacy. Comptrollerships originate from countries like the United States, Canada, the United Kingdom, Germany, the Netherlands, Sweden, Mexico, Chile, and Argentina (Willeman, 2020) .
In turn, the French Court of Accounts model differs significantly from the Comptrollership-type SAIs. In this model, the SAI has the authority to issue judgments and enforce them effectively, assigning responsibilities and issuing commands and sanctions. In this framework, the powers of judgment may or may not be deemed final, but they undoubtedly resemble the characteristics of judicial decisions. Precisely because they possess attributes so similar to those of decisions issued by the Judiciary (definitiveness and coercion), the SAIs in the Court of Accounts model historically prioritize the technique of normative subsumption, focused on legality control. Their practices, therefore, lean towards legal compliance audits, the benchmark is the verification of compliance with the regulations related to public resource management. Consequently, these institutions are generally headed by individuals with a legal background, not managerial.
The strong autonomy of the Courts of Accounts to issue binding decisions to other powers raises serious questions about their suitability to the tripartite separation of powers model advocated in many Constitutions around the world. Although in its country of origin, France, the formal function of the Court of Accounts is to assist the Parliament in overseeing the public budget, its ability to act autonomously seems to be more in line with Bruce Ackerman’s contemporary theory (Ackerman, 2000) , which calls for a new reading of the separation of powers that adapts to modernity and overcomes Montesquieu’s outdated model. For Ackerman, autonomous SAIs would be part of a different branch from the traditional organic powers of the State: the branch ensuring governmental integrity.
The French Court of Accounts was the institutional design that originally inspired the Brazilian institution when Brazil also adopted a Court of Accounts, which is understandable given that Brazilian Administrative Law is heavily influenced by French law in general (Jordão, 2019) . Besides France, other countries adopting this model are Spain, Italy, and Portugal, which are nations deeply marked by the Roman-Germanic civil law tradition.
With the recent convergence of international experiences, SAIs in the Court of Accounts model have undergone profound reforms, given that since the beginning of the 21st century, they also began to host performance evaluations of public policies. This is due to the privileged constitutional position of SAIs to conduct performance audits and contribute to inducing good management in Public Administration.
In this context, the Brazilian Federal Court of Accounts (FCA), with its history spanning over 130 years, has traditionally centered its efforts on monitoring and evaluating the financial regularity of the State, addressing illegalities associated with the public budget. However, the institution did not remain static over the years. Today, the FCA holds the responsibility of exerting broad external control over the State, armed with a strengthened constitutional mandate to supervise all branches and entities of the Federal Union. While its foundational mission, its raison d’être, is still to ensure accountability to the citizens, this accountability has taken on a broader connotation. It’s no longer limited to examining accounting balances derived from formal numeric spreadsheet verifications. Instead, the FCA’s purview extends to any activity with potential implications for the public treasury, be it directly or indirectly, branching out into a series of specific competencies (Brazil, 1988) .
Certainly, the 1988 Brazilian Constitution initiated significant changes, broadening external control over Public Administration. This expansion arises from the addition of operational auditing among the FCA’s responsibilities and the opportunity for control grounded in the criteria of legality, legitimacy, and economic rationality of public expenditures. With this extensive constitutional mandate to autonomously supervise all state activities impacting the public treasury, the FCA occupies a crucial and unique position to access information that might otherwise remain hidden. This is especially pertinent concerning independent regulatory agencies, whose institutional structures amplify the potential for a deficit in democratic legitimacy and responsiveness to societal demands.
While the Brazilian state is organized around a triad of Three Powers, Executive, Legislative, and Judiciary, this structure is often challenged in practice. Given the verbose, unclear, and ambiguous wording of the Brazilian Constitution of 1988, there are various theories about the nature of the Federal Court of Accounts, whether it’s Judicial, Legislative, or even Administrative. This distinction has practical implications. In Brazil, the Judiciary has the prerogative to review all administrative acts. Thus, if the FCA is viewed as an extension of administrative justice, its actions would be under stricter judicial scrutiny. On the other hand, if it holds the final adjudication prerogative, akin to the Judiciary, the dynamics would change. This topic is too extensive to delve into here, given its deep roots in the Brazilian legal system. For the purposes of this article, it’s sufficient to understand that the FCA occupies a privileged oversight and control position in the Brazilian Constitution. It has a myriad of functions, adjudicative prerogatives, it is even called a Court, but is traditionally associated with a supporting role to the Legislative Branch.
For the purposes of this article, it’s sufficient to understand that recent analyses, whether critical or supportive, identify the FCA as an entity partly responsible for Brazil’s regulatory landscape. The FCA’s oversight of other state entities can be interpreted through various theoretical frameworks, primarily falling into three categories.
The first perspective postulates that external auditing and control should be carried out expansively based on criteria like legality, legitimacy, and economic rationality, regardless of its nature, whether formal compliance control or performance control. This involves oversight on financial matters (in a broad sense) or on all other matters. External control is, in this perspective, understood as a monolithic competency that encompasses the joint oversight of all these elements (legality, legitimacy, and economic rationality) (Furtado, 2015) . The significant risk of this stream is that, in practical terms, it turns the FCA into an entity combining legal control with operational control to have total control powers over any state action or program.
A second theoretical stream, more moderate, believes that all activities of public entities are unequivocally covered by the Brazilian Courts of Accounts’ oversight. However, only control over intermediary activities is broadly and unrestrictedly applied. These refer to the support and auxiliary tasks that make it possible for the administration to achieve its main purposes. They are essentially the backend operations, administrative in nature, that are not the end goal but are essential for smooth functioning. Examples might include procurement processes or human resources management. In sum, direct spendings.
Regarding the primary activities, the activities the entity is created to carry out, there would be a general principle of self-restraint by the Courts of Accounts, with a very strong prima facie preference for the technical reasons provided by the agency (Mendonça, 2012) . These primary activities can be defined as the core tasks for which the administrative entity was established. They represent the ultimate objectives or primary goals that serve the public interest or fulfill the principal mandate of the public body. However, this self-restraint by the FCA can be overridden in situations of procedural flaws, lapses in social participation, evident and unambiguous regulatory omission, or strictly literal illegality.
The prevailing view is that there’s room for external control over regulatory decisions, as long as it’s exercised with due respect to the agencies. The problem, however, is that this stance heavily relies on a theory of deference that is both consistent and operational in less-than-ideal professional circumstances. Yet, this approach is consistent with the FCA’s precedents. Presently, the Court embraces the notion of second-order control: a control that is always applicable to agencies but never directly to the essence of regulation. In terms of the primary activities of the regulatory agencies, the FCA firmly believes it should not usurp the role of the regulatory body, as doing so would contradict the Federal Constitution (FCA, 2004) .
A third theoretical stream, resulting in more restrictive interference, believes that the 1988 constitutional movement expanded the duties of external control but did not turn the FCA into a general review instance of Public Administration. It was a Court designed to have direct jurisdiction limited to examining the legality of budget execution, while the new generic authorization to conduct operational oversight is in fact limited to indirect jurisdiction. This indirect jurisdiction would only culminate in collaborative acts or in triggering other controlling institutions that could then act coercively. In this view, the Judiciary, not this administrative Court, would be responsible for the correction of general irregularities. The FCA would be responsible for addressing irregularities solely concerning direct public expenditures, such as in the case of procurements.
This third, more skeptical and critical theoretical stream regarding the FCA’s limits, claims that the Court of Accounts uses imprecisions in the norms governing it to expand its competencies. However, they believe this expansion doesn’t merely signify a formal dogmatic problem (Sundfeld & Camara, 2019) . These authors warn that the lack of constitutional or legal provisions that clearly authorize the FCA to act means a lack of democratic backing since there’s no assent from the institutions responsible for modeling the exercise of power. The phenomenon observed today in the FCA, in this view, would be the undemocratic self-proclamation of highly interventional competencies in individuals’ sphere. It would be an illegitimate and anti-legal action, even if based on a social convenience rhetoric related to the realization of fundamental rights.
Regardless of the theoretical perspective adopted by the norm interpreter, empirical research has shown that, in Brazil, the Court of Accounts plays a consolidated and highly significant oversight role in the daily operations of the state (Marques Neto, Palma, Rehem, Merlotto, & Gabriel, 2019) . Although the Court adheres to a deferential discourse (aligned with the second theoretical stream), there are instances where its oversight closely resembles the first theoretical stream, which endorses coercive measures in Public Administration, even amidst operational challenges. The FCA’s approach has been proactive in pursuing enhancements in the Brazilian regulatory environment. The subsequent chapter provides a more in-depth exploration of the regulatory oversight phenomenon.
3. Regulatory Oversight Bodies
Although there is an extensive literature justifying the need for state regulation, this type of intervention can also impose its own problems. Even highly sophisticated regulations can impose high compliance costs, inhibit innovations, generate barriers to entry, create ancillary risks, and generate income for mobilized interest groups. At a lower level of sophistication, regulators can make mistakes, choose to issue poorly crafted regulations, have little time and resources to regulate, and neglect certain important social objectives that are not clearly contemplated in their own narrow mission. Regulators can also be captured, choosing to maximize their personal interests to the detriment of collective benefit, lack access to underrepresented interests in the regulatory process, or need to make tough choices in risk versus risk trade-offs in a real world of limited rationality.
This means that invariably, wherever there is regulation there will be a need for oversight of the regulatory system. This need arises from the demand to reduce the side effects of intervention, promote efficiency in the definition and implementation of instruments, encourage consistency and transparency, ensure accountability, enhance organizational culture, and seek the optimization of overall results. Regulatory oversight exercised by a centralized and high-ranking governmental body has been increasingly explored by the international community, particularly within the OECD, which aims to foster an agenda for improvements in the regulatory environment of nations.
Since 2008, the OECD has been promoting debates on so-called Regulatory Oversight Bodies (ROBs). In 2008 the debate was still nascent, but it gained momentum, with consistent reports produced over the years. In 2012, the OECD Regulatory Policy and Governance Council issued a Recommendation in which Principle 3 requires countries to “establish mechanisms and institutions to actively oversee regulatory policy procedures and objectives, support and implement regulatory policy, and thus promote regulatory quality” (OECD, 2012) .
Furthermore, the Recommendation highlighted the importance of maintaining a permanent entity in charge of regulatory oversight, established close to power centers, to ensure that regulation is truly useful to public policies holistically. The high rank and proximity to power are also justified by the need to have authority over Ministries or other bodies subject to oversight. The political accountability of the ROB can derive from different types of horizontal relationships among State institutions, but must always respect neutrality and maintain independence from the partisan political environment, which has the potential to undermine efforts towards technical expertise (Wiener & Alemanno, 2011) . Moreover, entities without institutional missions specifically aimed at regulatory oversight, or which only supervise on an ad hoc basis, cannot be classified as ROBs.
The main functions of Regulatory Oversight Bodies vary from country to country, with no single formula. Nevertheless, as debates matured, the OECD published a study identifying five main functions: 1) scrutinization of the quality control of regulatory instruments; 2) identification of policy areas where regulation can be made more effective; 3) systematic improvement of regulatory policy; 4) promotion of co-ordination; 5) guidance, advice and support (Renda, Castro, & Hernández, 2022) .
Quality control relates to the scrutiny of ex ante impact assessments and/or ex post evaluations of legislation, scrutiny of the use of regulatory management tools such as Regulatory Impact Assessments and public hearings and challenge if deemed unsatisfactory, and monitoring compliance with better regulation guidelines. While ROBs might undertake ex-ante or ex-post impact assessments and evaluations, it’s not mandatory. They might also assess the quality of legal drafting or champion the implementation of particular new regulatory measures. In essence, this function deals with the oversight of the influx of new regulations.
Identifying areas of policy where regulation can be made more effective is the scrutinization of the stock of existing regulations. Here, core tasks include coordinating the planning and execution of reviews spanning entire policy areas, supervising the initiation and application of review and sunset clauses, orchestrating baseline measurements of administrative burdens, applying stock-flow linkage rules, and overseeing and enforcing targets for reducing the costs of current regulations, often referred to as regulatory budgeting. This second primary function mainly concerns the review of existing regulations and reduction of compliance costs and doesn’t necessarily correlate with the introduction of newer, perhaps less onerous, regulations.
Systematic improvement of regulatory entails regular monitoring and reporting on the performance and progression of administrative activities associated with the regulatory governance cycle. It also involves coordination with other oversight bodies, if present. Such reports may lead to evaluations of the regulatory governance framework, accompanied by relevant proposed changes. Alternatively, other institutions could directly undertake these evaluations and proposals, but the ROB would be responsible for monitoring and reporting.
Guidance, advice, and support primarily refer to capacity building within the Administration. This function pertains to the formulation and execution of guidance and guidelines on utilizing regulatory management tools. Through these activities, the central ROB establishes the governing rules, which are then applied consistently throughout the oversight of the regulatory governance cycle. Guidance, advice, and support serves as the backbone for regulatory development and reform. By emphasizing capacity building within the Administration, Regulatory Oversight Bodies ensure that there is a continuous enhancement in understanding, skill sets, and expertise among administrators, policymakers, and other key personnel. Well-articulated guidelines offer a roadmap for stakeholders, minimizing ambiguities and promoting consistent application of regulatory principles across various sectors.
OECD member countries have consistently invested in regulatory oversight. In 2014, 33 countries had at least one institution responsible for promoting regulatory policy, as well as monitoring and reporting regulatory reforms and regulatory quality at the national administration in a holistic and constant perspective. By 2017 this number had reached all 39 jurisdictions assessed. Most are institutions independent of traditional branches of government, and therefore are not embedded in Ministries or Secretariats. Still, most admit the assistance of government public servants when support is strictly necessary.
In general, ROBs have a very solid legal anchoring. Since early 2018, five young ROBs had their missions renewed by new rules, and seven more became permanent institutions. Many others expanded their governing rules to encompass more detailed aspects of oversight functions. Additionally, in 2021 most of the OECD Oversight Entities reported an increase in their staff and budget. OECD countries have bet their resources on the inspection and analytical capacity of these institutions as a way to optimize the functioning of the entire state apparatus through evidence, which ultimately implies greater overall trust in elected representatives.
Among the central oversight functions, the most widespread is the quality control of Regulatory Impact Analyses: about 75% of all ROBs, in all jurisdictions, have this responsibility. Commitment to the systematic improvement of regulation through advice to the Government, the promotion of best practices, the regulatory capacity of institutions, and inter-institutional dialogue comprises about 70% of OECD ROBs. Guidance on the use of regulatory instruments falls within the competence of about 70%. Systematic assessment of regulatory policy comprises about 55% of all ROBs. About 45% of all ROBs are responsible for quality control of stakeholder engagement activities. About 40% scrutinize the legal regularity of regulation. Finally, only about 30% are tasked with ex post evaluation of regulation.
To summarize, Figure 1 shows the evolution of veto powers over the years, while Figure 2 presents the main reasons for such vetoes (OECD, 2021) .
Figure 1 and Figure 2 demonstrate that OECD countries strongly adhere to regulatory oversight for a myriad of reasons. However, the 2021 report on regulatory oversight highlights that the OECD continues to emphasize the necessity for reforms to bolster these institutions. Notably, the OECD underscores the importance of its Oversight Entities broadening their ex-post evaluation of existing regulations, which currently stands at only 30%.
Such responsibilities intersect with those carried out by traditional institutions like Central Government Ministries, Secretariats, Parliament’s audit units, and Supreme Audit Institutions. The latter encompasses Brazil’s Federal Court of
![]()
Figure 1. Scrutiny of RIA quality (in number of jurisdictions where each option applies).
![]()
Figure 2. Reasons for returning RIAs (in number of jurisdictions where they are applicable).
Accounts. The past 20 years, however, have also witnessed the establishment of non-traditional institutions assigned specific tasks to advance the better regulation programs agenda (OECD, 2018) .
Nevertheless, there isn’t a universal formula. Relying on traditional institutions with overlapping duties has the merit of leveraging already deployed human, material, and informational resources. For instance, in the UK, the Regulatory Oversight Entities comprise the National Audit Office and the Public Accounts Committee. These entities aim to bolster accountability and maximize the utilization of public resources, closely mirroring the role of the Brazilian Court of Accounts.
The modifications or enhancements many OECD countries have pursued are indicative of initiatives aiming to sustain more effective regulatory systems. Current efforts directed at regulatory oversight (or meta-regulation) are informed by multiple theoretical standpoints. Broadly, these endeavors are designed to refine regulatory procedures and organizations, foster technical knowledge and dialogue within regulatory sectors, mitigate regulatory constraints by proposing intelligent alternatives to direct command and control, and simplify regulations to focus on their core purposes, thereby minimizing detrimental effects (Lodge & Wegrich, 2009) .
All these perspectives share a vision of Law and its institutions, viewing regulation as a tool to achieve the aims pursued by the legal system. In other words, they envision a Law that responds to societal demands and objectives. This functional view of Law may seem self-evident, but it’s not1. Increasingly, there’s political pressure for a pragmatic legal system that delivers substantial justice, making the Law a real facilitator in responding to societal needs. This phase is what Philippe Nonet and Philip Selznick refer to as the transition to the responsive legal system (Nonet, Selznick, & Kagan, 2017) .
Nonet and Selznick’s work emerged from the realization that a strictly legalistic Law, that is, autonomous and formal, no longer meets societal demands. The ideal of legality is now understood as progressively reducing arbitrariness in law and its administration in line with societal aims. Thus, the state should flexibly use its powers and intellectual and organizational resources to meet societal demands. If Law ignores reality, reality will ignore Law. This implies that Law needs to be guided by real-world inputs to steer society towards its goals.
This theoretical construction has significant implications in the institutional sphere, especially regarding the dynamics of public control, which is the subject of this article. In this new responsive legal order, which is more flexible and proactive in all aspects of state action, it is posited that “Powers without checks, and not the blurring of boundaries between them, constitute the real threat of tyranny and injustice” (Davis, 1959) . In other words, it means that “the dangers of arbitrariness must be managed in a way to facilitate, rather than hinder, the expansion of institutional competencies.” For the pioneers of Responsive Law, in a world where the Law increasingly takes on more responsibilities, and new decision-making nuclei are created, the inability to intervene becomes an increasingly evident source of arbitrary power.
This is not to say that control should always be exercised inadvertently. One cannot ignore that control activity also has its costs. However, it implies that the interpretation of norms that allow public control should be carried out to enable the expansion of control possibilities, even if it is exercised with deference to decision-making instances. The application of the Law becomes the art of conferring legitimacy to the public act, i.e., ensuring that the purposes are taken seriously in the functioning of government institutions.
Administrative Law, as known in Brazil, has more characteristics of the traditional autonomous (formalist) French law, which is based on limiting administrative powers. The new Responsive Law, in turn, primarily aims to enable and facilitate the exercise of power according to the objectives. Public control is tasked with evaluating institutional problems associated with different contexts and administrative mandates and begins to point out mechanisms suitable to correct or moderate such issues (Nonet, Selznick, & Kagan, 2017) . Controlling bodies need to adapt to determine control techniques that harmonize the presented problem with the pursued objectives, avoiding the simplistic solutions of Autonomous Law (legal/illegal) for today’s complex issues (Stewart, 1915) . There is, therefore, a sophistication of control techniques.
Responsive law starts structuring solutions, keeping coercion as the guarantor of its structures. If one can speak of a paradigmatic function of responsive law, it is regulation, not adjudication, states Nonet and Selznick. This is because regulation is the process of devising and correcting policies, or actions, to achieve legal objectives. The same logic applies to regulatory oversight or meta-regulation. The latter can be defined as the process to ensure certain regulations meet minimum standards concerning the objectives they pursue and promote engagement and regulatory capacity to achieve quality parameters beyond the minimum established by the norms, to foster solutions for institutional purposes.
This context is not unfamiliar to Brazil’s Court of Accounts. Since the promulgation of the 1988 Constitution, control parameters have expanded from mere legality to also evaluate the legitimacy and economic rationality of public administration acts. Today, these courts can conduct performance audits, assessing public structures and policies in terms of their intended institutional objectives. This expanded dynamic of oversight developed by the Court of Accounts aligns with the responsive law literature.
On a practical level, responsive law literature seeks to establish a project of responsive regulation. This project has been steadily taking shape since the 70s, drawing upon a collection of theories attuned to evidence and fostering dialogue between academic and regulatory communities (Ayres & Braithwaite, 1992) . These theories recognize the challenges of effective state regulation, driven by a deep understanding of issues such as social complexity, normative multiplicity, and epistemic scarcity.
This understanding leads to the pursuit of regulatory interventions that are sophisticated yet less resource-intensive, in terms of state assets, human contribution, and epistemic costs, compared to traditional regulatory formats. Such an approach aims for active participation from the community involved in the regulatory process. Drawing on the idea of reflexivity, responsive regulation underscores the importance of learning through ongoing observation and self-reflection.
Baldwin and Black introduced the notion of “really responsive regulation” (Baldwin & Black, 2008) . They argue that regulators should be responsive not only to the performance of those being regulated but also to broader variables: their genuine receptivity to and engagement with the regulator; the overarching context of the regulatory regime, including historical or cultural aspects; the diverse rationales behind regulatory tools and strategies; the overall performance of the regime; and potential shifts in these elements.
As evident, the truly responsive regulator as envisioned by Baldwin and Black is exceedingly nuanced. In this backdrop, Oren Perez (2011) provides a pointed critique, highlighting that the regulator proposed by Baldwin and Black demands even more information than the traditional “omniscient” ruler rooted in a command and control paradigm. This is because the regulator would need to thoroughly comprehend the numerous rationales present within the regulatory sphere, devise intervention strategies attuned to these distinct rationales, and yet ensure a cohesive overarching approach. Moreover, this regulator would require a remarkable capability for both retrospective and prospective analysis, recalibrating strategies in response to the ever-changing context.
These intricacies of responsive regulation present a seeming paradox. While responsive regulation emerges as a pragmatic endeavor striving to achieve societal objectives at minimal costs, its epistemic and operational expenses seem steep, implying the need for substantial resource allocation. Addressing these challenges, Oren Perez suggests in his critique that they might be addressed through a deeper examination of responsive regulation, which he labels as second-order reflexivity (Perez, 2011) . This in-depth analysis involves stepping back from substantive debates to ponder the conceptual and practical implications of primary responsiveness theories, weighing the feasibility of actualizing these endeavors.
For illustration, Olsen notes that responsiveness can be institutionalized within a legal system either through the formal commitment of regulators or via a central meta-regulatory body overseeing the regulatory process, serving as a hub for reflection and the enhancement of regulatory quality. Centralized monitoring offers the benefit of pinpointing intrinsic shortcomings and mutual discrepancies in suggested regulations, which perhaps could elude detection by individual agencies owing to institutional, temporal, and epistemic limitations.
It’s essential to recognize that both the overarching theory of Responsive Law and the pursuit of responsive (and reflective) regulation not only align but also champion the incorporation of institutional oversight mechanisms centered on responsiveness and reflexivity. Various institutional frameworks can be applied to Regulatory Oversight Bodies, and numerous institutions globally have matured to match the sophistication evident in OECD regulatory watchdogs2. Among the many entities that have evolved to embrace the regulatory oversight role are Supreme Auditing Institutions, exemplified by Brazil’s Federal Court of Accounts.
4. The Federal Court of Accounts and the National Mining Agency
The decisions under review total 177. All of them were read in full and organized into a table to distinguish which cases exercised (I) strict control over the state’s financial regularity, and which exercised (II) broader legality and performance control, that is, regulatory oversight. Many of these decisions did not issue direct commands to the Agency but only internal determinations to maintain monitoring, for example. However, 34 decisions were categorized as controlling financial regularity, and 33 decisions were categorized as regulatory oversight.
It is observed that in the case of inter-institutional relations between FCA and NMA, the external control body has a significant role as a regulatory watchdog. There is a large margin of forwarded deliberations (almost 50%) where the FCA decided to exercise coercive and immediate control outside the scope of strictly financial legislation (accounting, financial, budgetary, and patrimonial matters). On the contrary, the FCA was interested in acting coercively to: 1) address performance issues; 2) ensure NMA’s compliance with general legislation and mining legislation; 3) even if sometimes without precisely indicating which legal norm or performance parameter was being violated by NMA; and 4) reformulate the internal governance and management structure of NMA based on the FCA’s own assessments.
It is possible to observe that the deliberations, in general, have become less generic, abstract, or self-evident over the years. In the first three years of the 2010-2020 decade, for instance, there were directives such as: 1) “recommend to NMA to make efforts to equip its inspection area with a structure suitable for the relevance and materiality of mineral exploration activity”; 2) “be aware of the need to adequately provide material resources to the inspection areas of its superintendencies” and that “conduct, whenever possible, on-site inspections to validate the information presented in the Annual Mining Reports”; and 3) “inform the Ministry of Mines and Energy (MME) and the NMA of the fragility of controls and the current supervision exercised by the agency over mining enterprises involving rare earth elements”. These are quite generic directives.
It was observed that after 2014 the directives sent to NMA became more specific. A possible explanation is that this period coincides with the internal restructuring and specialization of the FCA. Conversely, the surge in more specific directives, which in practice have a mandatory character, represented an attempt by the FCA to enhance its directive power, supplanting the regulator in the exercise of its primary activities. The Judgement of the Performance Audit n˚ 2440 of 2016, for example, marks a moment when the FCA seeks to hold NMA directors personally accountable for operational or performance deficiencies on the Brazilian mining sector, in addition to continuing the controversial practice of demanding action plans for the recommendations issued.
The main conclusion from the audits and decisions of the FCA along the decade was the incompatibility of NMA’s organizational structure, the number and training of its staff, and its budget, given the complexity of the responsibilities of the regulatory body for the Brazilian extractive industry. In the Audit Judgment n˚ 657 of 2012, an investigation presented more detailed data on the insufficiency of human and material resources in the NMA. It was concluded that despite the efforts of the Public Administration, there was no immediate possibility of solving the problems. There were severe performance issues in all NMA’s decision-making processes. To address this insufficiency, the mobilization of the top leadership of the Executive and Legislative branches was seen as a solution, as they were responsible for providing resources to enhance decision-making quality in the NMA. In turn, in Judgment n˚ 1734 of 2012, the FCA explicitly recognized that despite identifying a performance deficit of the NMA concerning its institutional duties, the problematic decisions were within the agency’s administrative autonomy and, ultimately, within the Executive’s administrative oversight power over independent agencies.
After the collapse of the Mariana mining dam in 2015, which are now among the biggest industrial disasters of the century, the FCA’s Judgment n˚ 2863 of 2015 deemed it appropriate to alert the Executive and Legislative branches so they could address NMA’s challenges with a larger budget allocated for training and inspections. Judgment nº 2440 of 2016 emphasized the seriousness of the studies presented by the NMA itself regarding its budget proposal to carry out inspection activities and to meet the demand for human resources. It was proven that mines in Brazil were inspected, on average, once every 114 years, which encouraged illegal mining and other dysfunctionalities, such as serious environmental impacts, as well as operational structures with inadequate technical specifications (e.g., occupational safety structures and mining dams).
The debate became more intense after the collapse of the Brumadinho mining dam in 2019. In both cases, the entrepreneurs had presented inspection reports and stability declarations attesting to the safety of the structures. However, it seems safe to say that the regulator did not fulfill its institutional mission of checking the veracity of these reports. Since then, institutional controls over the agency have seemed increasingly intense and strict.
In this context, Judgment n˚ 1116 of 2020 resulted from the only inspection explicitly qualified as a compliance audit, and not a performance audit. This inspection sought to understand the legal regularity of the measures adopted by NMA to monitor and inspect the mining waste dam that broke in the city of Brumadinho. Incidentally, the audit ended up analyzing the legal regularity of ANM’s actions on mining dams all over the country.
In brief, it was evident that the responsibility for developing technical details in mining projects was transferred to the entrepreneur and their responsible engineer. Therefore, the audit team of the FCA concluded that it would not be possible to control the regulatory entity regarding the techniques and methodologies adopted by the entrepreneur in the construction and execution of dams. The regulator’s role would only be to manage the declarative data from the entrepreneurs and validate them. However, given the budgetary constraints faced by ANM, the entity prioritized inspections in projects where there was a higher indication of danger. In the case of the Brumadinho dam, the documents attested to stability, which classified the dam as non-priority. Thus, the audit team concluded that ANM followed the procedures provided in the regulations, according to the risk category, potential associated damage, and other classification parameters concerning the inspection of the safety of the Brumadinho dam.
For the audit team, it was clear that there were significant improvements to be made in the regulation of the mining sector, including the inspection phase. However, these were more related to weaknesses in NMA’s institutional capacity than to outright non-compliance with the legislation. It was understood that NMA was trying to optimize its scarce institutional resources, allocating its resources where there was greater danger. For this reason, the technical unit only proposed monitoring the situation.
The Minister of the FCA responsible for the case n˚ 1116 of 2020 disagreed with this position held by the team of auditors. In her decision, the Minister highlighted from the report a set of highly dangerous events that were not reported to the regulator or were described more mildly in the information system. Therefore, no immediate alert was generated for the regulator. She then drew a parallel to the Federal Revenue Service’s checking methodologies used to validate information, considering that in both cases there are strong incentives for the entrepreneur to falsify their data. From this parallel, she concluded that NMA did not act preventively to obtain reliable information, which would be unacceptable. From this conclusion, she believed that the instructive unit should open a process against the directors of ANM due to the alleged omission in regulating the details of mining activities and the omission in carrying out inspections to obtain reliable information. In the same case, another Minister asserted that:
“I believe it is time for this Court to contribute more decisively to changing the status quo, which is, unfortunately, a total disregard for the safety of enterprises of this nature, as well as an apparent attempt to attribute to others the responsibilities for tragedies that degrade our natural resources and worse, claim many lives”.
It would be beyond the scope of this paper to describe in detail all the activities carried out by the FCA that had an influence on the NMA in this decade. The examples provided above were illustrative for context. For a concise overview of the relevant cases, a list of the primary regulatory oversight rulings is provided below (Table 1).
Between years 2021 and 2022, FCA initiatives decreased in their oversight intensity on NMA’s primary duties. However, there are still cases in which there’s a lot of pressure put on independent regulators. For instance, Judgement n˚ 1.430 of the year 2021 fined the director of the NMA for systematically failing to
![]()
Table 1. FCA’s regulatory oversight primary rulings over the NMA.
regulate the sector. The fine was later overturned in light of new national legislation and the NMA’s chronic structural deficiencies. Even so, the FCA considers itself competent to exercise superior oversight over the board of directors of independent regulatory agencies.
5. Conclusion
As mentioned, the rulings were selected where the scope and depth of the inspection allowed the FCA to construct a systemic view of the NMA. In this regard, this article aimed to describe and demonstrate that there are various interrelated reports and decisions about NMA within the FCA. Over the past decade, the accumulation of repeated findings in audits led the board of Ministers to issue increasingly explicit and categorical statements of dissatisfaction with NMA’s organization and operation. All FCA reports indicated that the agency’s chronic deficiencies were linked to insufficient budget allocations, sharp budget cuts, and an extremely limited staff. In the cases of dam collapses, the public audits indicated that NMA had not adequately inspected because it was impossible to operate satisfactorily with its limited institutional resources. With the little they had, the regulator optimized inspections by directing them to dams classified as more dangerous.
From 2018 onwards, FCA’s board of Ministers began to note in their votes that the NMA operated in a state of complete neglect, with a high risk of fraud and corruption, and institutional “misgovernance.” For this reason, they believed it was the FCA’s responsibility to change this reality and drive improvements, and they began to propose increasingly assertive intervention strategies. This suggests that the FCA sees itself as a higher governance institution for mining regulation. For instance, in 2016 and 2020, there were cases where the board of Ministers disagreed with the findings of their technical unit of auditors and decided to initiate punitive proceedings against the NMA’s board of directors due to alleged systematic neglect in adequately regulating and inspecting the mining sector.
The accumulated experience of various Supreme Audit Institutions (SAIs), at least within the OECD, showed an expansion of their activities, scope, and tools. However, there is broad recognition of the side effects of traditional punitive control activities (Halachmi, 2014) , and for this reason the new expansion primarily focuses on adding more sophisticated functions to promote good governance, something also called meaningful accountability. In this context, the OECD suggests tapping into the vast potential of Supreme Audit Institutions to provide solid informational contributions to the government to address systemic and complex issues in the public policy cycle and enhance the effectiveness of state programs. In this new paradigm, SAIs play a strategic role in driving improvements in governance or public management. In this regard, the benefit of this article is to empirically test the FCA’s performance on these new paradigms and offer a significant illustration of the potentialities and limitations of this institutional governance strategy to other jurisdictions.
Acknowledgements
This research was independently conducted with the backing of the Ministry of Education in Brazil, specifically through the Coordination for the Improvement of Higher Education Personnel (CAPES) under Funding Code 001 associated with the PS-PPGD 2021 public notice from FGV Law Rio. Additional support was graciously provided by the Getulio Vargas Foundation.
NOTES
1For skeptics, the Law is, above all, repression. In a similar vein, the Law is also seen as rules that legitimize and protect political and social hierarchies. For others, the Law is a tool for liberation and social progress, a realm where legal institutions manage to subject even the powerful to the scrutiny of the law and social justice. Or, viewed more skeptically, this progressive and transformative legal environment can be understood as a form of antidemocratic tyranny by certain legal elites. On the other hand, some believe that a Law unresponsive to social demands becomes mere bureaucracy that hinders daily life and stifles innovation.
2“In the past, countries with a Roman law tradition set up forms of ROBs, as part of Councils of State, as in France and Italy. These bodies served as advisors to the government on the legality of regulatory decisions. They were also the superior level of the administrative courts, so they also exercised an adjudicative role meant to protect governments and avoid litigation in the civil courts regarding specific regulations. For example, in France after the Revolution, the Conseil d’Etat and the system of administrative courts were designed to shield the administrative state from being unduly constrained by the separate system of civil courts; the civil court judges were viewed as more sympathetic to the monarchy, while the administrative courts were meant to be more sympathetic to the legislature and to its efforts to redistribute power and wealth in France after the Revolution. Today, the Conseil d’Etat, acting as both a court of appeals for the administrative courts and a supervisory body for the administrative state, brings significant expertise to bear on the legality of regulatory decisions. However, it does not review impact assessments of proposed new regulations prepared by regulatory agencies. Modern ROBs, established since the 1970s, especially in common law countries such as the USA and UK, but also in other countries such as the Netherlands, and in the European Union, have a different origin. They were mainly created in response to stagnating economic conditions; a rising tide of regulation of health, safety and environmental risks; an accumulated array of economic regulation of sectors such as banking, communications, and transportation; and an academic literature on both the need for and problems with regulation” (Wiener & Alemanno, 2011: p. 313) .