Institutional change and accountability: procedural ecology and defiance of the Brazilian case ⁴

Abstract

This article analyzes the process of institutional change, focusing on endogenous, exogenous and systemic processes. Scholars usually analyze institutional change considering endogenous and exogenous explanatory factors to talk about the direction and outcomes of the changes. In addition to these two factors, this paper proposes a third one. This paper proposes that the process of institutional change should be understood in terms of systemic variables and should take into account the interactions between institutions as a relevant factor in explaining the veto or promotion of institutional change. Based on qualitative evidence gathered from interviews with key accountability actors in Brazil, the paper discusses the incremental changes that occurred in the Brazilian accountability institutions. We demonstrate that, despite these incremental changes, the final result in terms of institutional change was not a broader systemic change with greater cooperation and coordination of control activities between accountability institutions. Accountability processes are understood as part of a dynamic system, under which competition and cooperation are part of the action repertoire of accountability actors and used differently depending on the situation.

Keywords:
institutional change; accountability; procedural ecology; incrementalism; Brazil

Resumo

Este artigo analisa o processo de mudança institucional nas instituições de accountability no Brasil, com foco em processos endógenos, exógenos e sistêmicos. Os estudiosos geralmente analisam a mudança institucional considerando fatores explicativos endógenos e exógenos para falar sobre a direção e os resultados das mudanças. Além desses dois fatores, este trabalho propõe um terceiro. Este artigo propõe que o processo de mudança institucional deve ser entendido em termos de variáveis sistêmicas e deve levar em conta as interações entre instituições como um fator relevante na explicação do veto ou promoção da mudança institucional. Com base em evidências qualitativas coletadas de entrevistas com os principais atores das instituições de accountability no Brasil, o artigo discute as mudanças incrementais ocorridas nestas instituições. Demonstramos que, apesar dessas mudanças incrementais, o resultado final em termos de mudança institucional não foi uma mudança sistêmica mais ampla, com maior cooperação e coordenação das atividades de controle entre as instituições de accountability. Os processos de accountability são entendidos como parte de um sistema dinâmico, segundo o qual a concorrência e a cooperação fazem parte do repertório de ação dos agentes e são usados de forma diferente dependendo da situação.

Palavras-chave:
mudança institucional; accountability; ecologia procedimental; incrementalismo; Brasil

Introduction

The paper aims at analyzing institutional change theories to offer a theoretical contribution regarding the explanatory elements of institutional development. The main point of the paper is that the explanation of institutions’ incremental changes must take into account both endogenous and exogenous factors, and also systemic variables that consider the interaction between institutions a key element.

In order to analyze that hypothesis, the paper conducted a case study of institutions of the Brazilian accountability system. This case study is based on qualitative evidence gathered to analyze the recent changes under which federal accountability institutions underwent in Brazil. Leaders of the institutions at the center of the federal accountability system in Brazil were interviewed, including members of the Public Prosecutor’s Office (MPF), the Federal Police (PF), the Office of the Comptroller General (CGU), the Federal Court of Accounts (TCU), the Federal Justice (JF) and the government Ministries. The interviews helped to map the procedural ecology in the Brazilian case and to understand the direction of change in the institutions, according to how they performed in the control of the public administration and the fight against corruption.

In the first section, the paper reviews the literature on institutional change, focusing on the discussion of explanatory variables of change. In the second section, the paper discusses the literature on the Brazilian accountability institutions and their recent process of institutional change. In the third section, the paper presents the methodology and the qualitative evidence collected to explain the change process under way. In the last section, from the conclusion about the Brazilian case, the paper analyzes the importance of the concept of procedural ecology in the explanation of the institutional development process.

Institutional change and incrementalism

The issue of institutional change has constituted a crucial element of political analysis. Institutions play a primordial role in society and, consequently, influence the results of government actions and policies. Institutions represent the rules of the game and they may be formal or informal (Hall and Taylor, 2003; Olsen, 2009). The progress made by the Political Science field includes the recognition of how important institutions are to development and social change processes. Understanding development processes means understanding that the fundamental analytical problem involved is how to constitute institutional causal mechanisms that can replace the exogenous factors associated to change (Rezende, 2010). By studying institutional change, the paper embraces a contemporary controversy in the specialized literature - there is no consensus as to what can and should be considered change’s causal factors.

Lindblom’s classic work states that incremental changes must be understood as a result of the rational nature of the decision making undertaken by political actors. The processes are incremental due to the high institutional costs associated to abrupt ruptures in the accumulated knowledge and in the institutions internal practices. Institutional change is incremental because the context of change determines that small additional values will start to be added to the policies so that non incremental changes are, rationally, considered by the decision makers to be irrelevant or without any practical or contextual applicability (Lindblom, 1959). This concept of incremental change does not mean that a policy is transformed small step by small step. Institutional changes may take positive or negative directions whereby strategies can be reviewed in the course of their unfolding, or the context may impose defeats on the policy makers (Weiss; Woodhouse, 1992).What is fundamental is that institutional changes occur in contexts of uncertainty, which means that the decision making processes tend to try to transform policies by incremental amounts avoiding abrupt ruptures and with the aim of effectively achieving incremental changes in practices and results (Lindblom, 1959).

Incremental change processes represent marginal adjustments to complex rules, norms and voluntary obedience structures without which an institutional rupture would occur (North, 1990). Analyzing the institutional change process would mean then understanding the catalyst of exogenous factors - the mechanisms that trigger institutional change. Understanding these mechanisms is essential to any analysis of institutional change processes and results. Institutional changes are imperfect, without a definite course. A well-designed institution is not an established solution, but a development process that can be activated under certain conditions. The problem of Political Science regarding institutional change is the need to identify the processes and determinants that increase or hamper the orderly effect of institutions (Olsen, 2009).

The study of institutional change would open up four analytical pathways, namely: (1) the theory of critical junctures, focusing on a development process; (2) theories that presuppose the existence of institutional malleability, which is explained by the actions of losing coalitions in the political process (3) studies that focuses on multidimensional aspects and the intentional and non-intentional effects of interactions in the explanation of change; (4) the action of reformers, whereby, in the face of collective action problems, political agents prove themselves capable of undertaking reforms that generate change (Pierson, 2004). From this perspective, it is important to understand the institutional development process in an incremental way, in which the marginal gains emerge from the exogenous factors unleashed by events external to the institutions themselves.

On the other hand, institutional change processes can also be explained by endogenous factors - when causal elements of change stem from the behavior of the institutions’ internal agents in regard to the institutional forms and contexts (Mahoney and Thelen, 2010). The political conflict is the result of the distribution and mobilization of resources, generating the change process. It is important to understand the way in which the agents interpret the allocation of political resources and unleash, internally, the process of change directed at their institutions. This can be either an effort to maintain the status quo or to provoke a rupture. The conflicts regarding resource distribution generate a space for debate about the institutional rules. The greater the degree of ambiguity, the larger the space for contestation from the internal agents, thereby creating a favorable context for endogenous change in the institutions (Mahoney and Thelen, 2010).

According to this perspective, institutions are not necessarily stable and are susceptible to change. The agents act internally to modify the institutions and the way the political context defines their strategies is important. According to Mahoney and Thelen, understanding the agents’ strategies requires comprehend two elements: (1) whether the veto powers of those involved in the change process are strong or weak; (2) the degree of discretionary power agents enjoy in interpreting rules and procedures and applying them. These two variables - discretionary power and veto power - allows the combination of the political context with institutional forms. From the interrelation between these two variables, it is possible to classify the institutional change strategies in four categories: (1) - displacement; (2) - layering; (3) - drift; and (4) - conversion (Mahoney and Thelen, 2010).

Thus, the literature acknowledges that understanding institutional change involves exogenous and endogenous factors in the light of the formal and informal rules that regulate the institutions. In regard to the exogenous factors, critical junctures and political contexts are what matter in the change process. As for the endogenous factors, the behaviors and strategies adopted by the agents inside the institutions allow the grasping of the institutional change process. Both the studies of Pierson (2004) and Mahoney and Thelen (2010) understand the process of institutional change based on the relationship between agents and structures. However, this literature does not endeavor to understand institutions’ systemic dynamic.

Besides the political context and the context of formal rules that delimit the relations between agents and institutional structures, we must also interpret the interactions among institutions as a systemic dynamic in which the relations between them are interdependent - the change process leads to the formation of institutional coalitions. In other words, apart from analyzing the rules in the light of institutional ambiguity, agents’ strategies, critical junctures or political contexts, it is also necessary to consider institutions interactions, in a context of a complex system of competencies and organizations in which situations of institutional conflict and cooperation make a difference in determining the result of change. The changes may encounter a veto or cooperative situation, according to the performance of other institutions and the relations of interdependence among them.

DiMaggio e Powell (1983) identified the process of institutional definition - the process of structuring institutions - as being marked by the following elements: (1) the extent of interactions among organizations; (2) the emergence of domination and organizational coalitions; (3) an increase in the information load that organizations have to deal with; and (4) the development of mutual knowledge involved in a joint venture among the participants in the set of institutions. In other words, the institutional change process goes beyond the boundaries delimited by the organizations themselves and their respective competencies. The change can be interrupted or vetoed altogether, according to the interactions among the institutions and the impact of the institutional rules may be positive or negative, according to the institutional ecology.

Consequently, institutional incrementalism depends on two analytical keys: (1) the interaction between the agents and the institutional structures; and (2) the systemic interaction among institutions. Systemic interaction among institutions, as defined by DiMaggio and Powell (1983), means the formation of an organizational cluster within the sphere of the State that involves mutual knowledge concerning a common public venture undertaken by a set of institutions. The interactions can take the form of an organizational coalition whereby the institutions come together around a common objective, or they may stem from institutional competition that undermines the impact of the development process of each institution in the organizational cluster. In this perspective, the development of shared knowledge and the control of information become essential to the constitution of organizational clusters that can imply the solidification of the status quo - reducing the impact of the change - or even the vetoing of any change at all (Galvin, 2012).

We understand procedural ecology as the interdependent relations of the institutions in a system in which the actions undertaken by one institution depend on the actions of other institutions. This interdependency is fixed by formal and informal rules and procedures that involve a set of institutions focused on achieving a certain goal. Therefore, it is the system that fixes the ecology in such a way that the result depends on the institutions coordinated and cooperative action.

An autonomous organization of the institutions intervenes in the change process. The cooperation and coordination of the activities depend on fixed institutional rules and, at the same time, on the establishment of political ties so that the actors of the institutions understand that the cooperation and coordination are vital for the success of the actions undertaken by the system. The essential elements for a procedural ecology to function are the knowledge accumulated inside the institutions and the agents’ ability to control the information that results from their actions.

The chart below summarizes the discussion and displays the explanatory variables associated to institutional change:

Chart 1.
The explanatory variables of the institutional change process

	Causal factor of institutional change	Analytical focus of the change process	Explanation of institutional change
Exogenous variables	Critical junctures, dependence on the trajectory	Political and broader institutional situation	Shocks external to the institutions trigger institutional change
Endogenous variables	Factors associated with the distribution of resources among institutional agents	Organizational elements internal to the institutions (formal and informal)	Organizational changes lead to changes in the distribution of power among the agents and the resignification of institutional objectives and processes
Systemic variables	Interaction among the institutions of an organizational cluster	Procedural ecology	Formation of organizational coalitions, control of information, knowledge generation, institutional cooperation or conflict processes

• Source: elaborated by the authors

Based on this discussion, we advocate that the institutional change process, in addition to the exogenous and endogenous variables, depends on the procedural ecology and its system dynamics, whereby the interaction among institutions matters. These interactions are formal because they are embedded in the institutions’ procedures, competencies and functions, all directed at achieving a certain end. But it is also informal, because it depends on the shared knowledge between the institutions’ agents and on the processes of coordination and cooperation. In this case, the informal interactions depend on political conceptions, on the process of interpreting the rules and on the formation of organizational coalitions within the system of institutions.

In the next section we examine the case of the development of accountability institutions in Brazil, with particular attention to the systemic interactions and the institutional change process resulting from a recent modernization effort.

The brazilian case - changes in accountability institutions system

In the beginning of the democratization process, in Latin America in general and in Brazil in particular, it was broadly advocated that there was a lack of horizontal accountability and weak horizontal accountability institutions (O’Donnell, 1996).The Brazilian democratization process meant a return to basic polyarchic conditions, in the light of a transition process that was a response to the crisis of the authoritarian regime implanted in 1964 (Santos, 1998). In 1987, the Constituent Assembly faced the dilemma of making choices that would generate an institutional arrangement capable of efficiently aggregating and processing the growing pressures stemming from an extremely heterogeneous and plural social context (Abranches, 1988).

The return to polyarchic conditions fostered a process of political openness that was essential for the democratic re-construction of the country. Parallel to that, the return of elections set in motion an incremental process of developing State’s institutions and, among them, the accountability institutions. Gradually, the established conditions enabled the consolidation of the democratic regime, especially in regard to the conditions of governability (Limongi, 2006). Since then the political competition, the presence of oppositions and clearer institutional rules have all contributed to the development of horizontal accountability institutions motivated by the fight against corruption and the expansion of checks and balances mechanisms (Melo, Pereira and Figueiredo, 2009).

Political change in Brazil in the last fifteen years has led to a process of incremental development of accountability institutions (Taylor and Praça. 2014). However, the constitution of a political regime that guarantees governability conditions has had an associated cost related to the conditions of democratic governance in the public sector. An antagonism has arisen between the exercise of political authority, on one hand, and the exercise of public authority in the field of government administration, on the other. The institutional change process has taken place incrementally, with the gradual introduction of new rules and routines into the functioning of institutions and the gradual increase and availability of additional resources. Furthermore, those incremental changes in the accountability institutions in Brazil empowered them to reveal various corruption scandals. This has led to a series of rule changes that have boosted the role played by those institutions in the Brazilian society (Taylor and Praça, 2014). A growing spiral of institutional incrementalism has stemmed from this process, aggregating new functions to the accountability institutions in the light of the corruption scandals that they have revealed.

In regards to exogenous factors, the 1988 Constitution was the first critical juncture, because it determined the competences of a series of institutions responsible for the accountability system. The framework of institutional competences is linked to a theoretical discussion that understands that it is up to the government the exercise of the internal control over its own activities and it is up to the Legislative branch the external control (in Brazil, exercised with support of the Federal Court of Accounts). Moreover, the Constitution delegates to the Public Prosecutor´s Office powers to control the public administration civilly and criminally and to propose judicial and extra-judicial solutions.⁵ Another critical juncture was the Fiscal Responsibility Law (LRF) which attributed to the accountability institutions greater powers to control the actions of Brazilian State public bureaucracies, extending the powers of audit, account control, and control processes.⁶

In regards to endogenous factors, the institutions of the horizontal accountability system in Brazil have been undergoing changes since 1988, and these changes have led to a new behavior pattern from their agents (Olivieri, Teixeira, Loureiro and Abrucio, 2013). Endowed with a greater degree of autonomy, the accountability agents adopted processes for interpreting the norms that turned them into political agents of fundamental importance to Brazilian democracy. Institutional change also derives from changes in the institutions´ routines, with the participation of the accountability agents. They influenced the implementation of the new rules, mobilized resources to assure that the rules were enforced, controlled important public sector information and selected the way in which institutional choices could be made (Galvin, 2012).

Besides exogenous and endogenous factors, we will also consider the interdependences between the institutions to better explain the recent changes that affected Brazil. It is a solid consensus in the literature that horizontal accountability depends on the existence of institutions authorized and disposed to control, supervise, rectify and punish actions performed by public agents that are considered to be illicit by other agencies and that are divergent with the public interest. Therefore, horizontal accountability concerns the checks and balances mechanisms of the Republic’s three branches of government - the Legislative, Executive and Judicial branches. It is expected that their functioning is effective and that it does not result from the isolated performance of each one but, instead, from the complex network of institutions (O’Donnell, 1999).

The institutions of the accountability system must be autonomous and well empowered to inspect, control, correct and punish any illicit acts in order to defend the public interest. The system of accountability institutions means a complex network within which the institutional competences are specific, but the actions of each institution depend on a process of institutional sequencing and cooperation with other institutions (Mainwaring, 2003). In the Brazilian case, the network of institutions is extensive, embracing the three branches of government and the media. Notwithstanding, at the very center of the network there are only four institutions, namely the Federal Court of Accounts, the Federal Public Prosecutor´s Office, the Federal Police and the Office of the Federal Comptroller General. These four are responsible for inspecting, controlling, correcting and instructing legal actions taken against public administrators and politicians in cases of corruption, misappropriation, embezzlement or any other divergence from the public interest.

The Federal Court of Accounts is a body of the National Congress and was founded in 1891. Since then Federal Court of Accounts has advised several decisions of the National Congress and established actions of control of the administration. In the 1988 Federal Constitution, it is the responsibility of this Court to exercise external control over the public administration by undertaking audits and informing the Legislative branch about the correct use of public resources on the part of public agents. Recently, the Federal Court of Accounts has left aside its more legalistic attribution and has strengthened its authority to audit the performance and spending of the public administration (Speck, 2000). This was a critical juncture for the Court, because the LRF changes in the entrenched practices within its administration. Another critical juncture that brought greater changes for the Federal Court of Accounts was the enactment of the already mentioned Law of Fiscal Responsibility. It added administrative practices and auditing to the Court´s responsibilities and brought new regulations and the need to modernize its internal practices and invest in technology (Loureiro, Teixeira, and Cacique, 2009).

In the case of the Federal Public Prosecutor´s Office, the 1988 Constitution brought another critical juncture, insofar as the constitutional provisions guaranteed its functional and administrative autonomy, freeing it from political influences and empowering it to act in the public sphere. The Office takes on one of the functions of the Judicial branch, since it is charged with defending collective and diffuse social interests. It is also charged with defending the democratic regime, in particular the protection of the nation’s public and social patrimony. The competences of the Office’s have a broader outreach, and, for the purpose of the present research, the essential function performed by the Office is the one of horizontal accountability: it has the authority to take legal action against public administrators and elected politicians. In the case of the Office, a strong dose of political will on the part of its members has been added to its institutional incrementalism change and the socialization of that volition occurs in defense of a society that has been rendered apathetic in the face of a State faced by a myriad of corruption scandals (Arantes, 2011a).

The same process occurred with the Federal Police. Given its role of police force of the Federal State, it conducts criminal investigations in cases of penal infractions contravening the public and social order to the detriment of goods, services or interests of the Federal government or of its autarchies and public corporations as well as investigates other crimes with inter-state or international repercussions that need to be uniformly suppressed. The Constitution recognized it as a permanent government body and formally structured its careers. The permanent characteristic meant that, even though it was subordinate to the Executive Branch, it could not be dissolved by the government. In the second case, the structuring of its careers meant that there are established rules governing access to the respective positions and that there is a settled internal hierarchy with clear rules regarding ascension and the sub-divisions of the body. In the internal socialization of Federal Police agents and senior officers there is a strong component of repressive action against organized crime and corruption associated to the strengthening of democracy (Arantes, 2011b).

Lastly, the creation of the Office of the Comptroller General in 2003 was the result of a long process of incrementalism in the internal audit of the federal government. Prior to the creation of the CGU, the internal control was achieved by a decentralized system, spread by various government administration entities and coordinated by the Internal Control Federal Department attached to the Ministry of Finance (Olivieri, 2010).⁷ The Office of the Comptroller General was created by Act 10683/2003, concentrating all Federal Government’s internal audit activities as well as initiatives directed at combating corruption. Prior to its creation, the Executive Branch´s decentralized internal audit was disorganized and ineffective, both in its performance of audits and in the evaluation of management practices (Balbe, 2013).

The 1998 Constitution and the return of democracy in Brazil constituted a trigger to a process of incremental change in the accountability institutions. Firstly, because it placed at the institutional center a considerable and complex chain of competences, rules and processes. Secondly, because it inevitably led to organizational and practical changes in the interior of those institutions. Thus, from a situation of a complete lack of horizontal accountability, Brazil achieved incremental changes in accountability institutions that led to the development of new practices in the public administration and the gradual empowerment of the accountability institutions. The overall result has been the development of a system of accountability institutions with exclusive and superimposed competencies directed at preventing, investigating and punishing illicit acts in the sphere of the political system and public administration (Taylor and Praça, 2014).

Institutional changes came about as a reaction to the critical junctures imposed by corruption scandals and the new Constitution as the exogenous factors, as well as from organizational and socializing changes of the institutions agents, as the endogenous factor. The scenario unfold anti-corruption policies in the following ways: (1) - the dissemination of anti-corruption policies in view of the fact that Brazil had become a signatory of international conventions against corruption; (2) - the layering of new organizational formats of public administration, such as the creation of the CGU and the greater autonomy of the Federal Police, with the co-existence of new and old institutional formats; and (3) - the conversion of institutions, such as the Federal Court of Accounts, which took on more powers and autonomy, established by the Constitution and by the LRF.

The constitutional organization of accountability institutions catalyzed a gradual enhancement of the State’s capacity to control and supervise the bureaucracy and this has brought about incremental changes in each institutions. However, these incremental changes did not result in any incentives for institutional cooperation (Taylor and Buranelli, 2007). Both the democratization process in Brazil and the unveiling of corruption created critical junctures that brought about considerable changes in each institution of the accountability system. As for the endogenous factors, incrementalism led to considerable learning on the part of the agents and generated a lot of information and many control actions.

The accrued learning endowed the agents with greater capacity for action (Olivieri, Teixeira, Loureiro and Abrucio, 2013), but it also gave them control of information and actions that were used as a political resource by the accountability institutions (Gailmard and Patty, 2007). The changes in the institutions of the Brazilian accountability system did not result in a more systemic development that would involve the process of interaction, interdependence and complementariness that each institution maintains with each other in the light of the procedural ecology of their activities of control, audit, investigation and punishment of corruption.

The interactions among horizontal accountability institutions, in the light of the agent’s broad volition, reduced the impacts of control actions insofar as the procedural ecology created incentives for institutional disputes and barriers to sequenced control activities. In Brazil, the interaction process is marked by the interdependence of accountability institutions, since the institutional rules overlap. Increasing the interaction between the institutions of the accountability system does not lead to a cooperative or coordinated process of joint actions, since the organizations rationally struggle for their recognition by the public opinion and have quite strong political interests. In this sense, the information that the accountability institutions began to hold served as a strategic resource to ensure the realization of their interests and a favorable public opinion. In other words, the institutions began to withhold information and control actions that responded to exogenous and endogenous factors determining institutional change. On one hand, in the case of the exogenous factors, the information served to address the critical junctures of corruption scandals, in which the institutions fought for a favorable public opinion. On the other hand, information was a strategic resource that served the interests of the institutional actors, in respect to their broader political interests. The information acquired by bureaucratic organizations is instrumentally used. The agents decide how the information is used and passed on (Gailmard and Patty, 2013).

The change in the system of accountability institutions in Brazil was not accompanied by an environmental change in the interactions between these institutions and was not able to produce coordinated and cooperative actions or ensure rational sequencing of activities aimed at improving accountability. Incremental changes did not affect the systemic variables, being restricted to a logical closure and differentiation of each of the accountability system elements. The interactions between the institutions of the accountability is bound by rigidly defined procedural competences and largely uncoordinated procedural processes, driven by political forces stemming from the critical junctures of corruption scandals and the political interests of agents.

The next section explores this hypothesis about the procedural ecology and interactions between accountability institutions in Brazil.

Methodology, evidences and analysis

In order to understand the current scenario of procedural ecology among Brazilian accountability institutions, the paper analyzes interviews with the civil servants of the institutions under study. The purpose of the interviews was to understand how the system of accountability institutions works from the point of view of its agents: which points they consider to be critical in the interactions between the institutions, whether the coordination between them exists or whether it is still a frontier to be surpassed and what has been the direction of institutional change since the democratization process began. The reasons for the choice of the interview technique are several. Not only it makes possible to get to know the effective functioning of the institutions, starting from the assumption that that functioning does not derive from institutional norms and rules alone, but also that there are certain spheres of practice and organizational culture that can influence the forms of interpretation and application of the rule. Interviews are particularly revealing if the goal is to capture what people did or planned to do, such as the decision to establish interactions with other accountability institutions.

The selection of respondents followed the criteria for institutional placement and work experience in the area. We attempted to interview people who have an important role in the formation of the control and accountability policy, who are trained and strategically positioned to provide opinions and views on the procedural ecology of the accountability institutions system in Brazil.

The interviews were conducted from May 2014 to May 2015. Fifty-one federal civil servants who work in the four selected institutions - the Office of the Comptroller General (CGU), the Federal Court of Accounts (TCU), the Office of the Public Prosecutor´s Office (MPF) and the Federal Police (PF) - who hold key positions within their institutions. The civil servants interviewed hold management and coordination positions, being responsible for the core activities of their respective bodies. Briefly, it can be said that they are mostly civil servants with a long history in the Brazilian public service. Many of them, before entering their respective institutions, had already held other public office - which indicates that they have an extensive knowledge and experience in the field of public administration. In addition to that, they have several years of career within their own institution, which means that they have a broad notion of the inner work and challenges.

Table 1.
Respondents per institution, Brazil, 2014-2015

Institution	Frequency
Office of General Comptroller (CGU)	13
Federal Court of Accounts (TCU)	15
Public Prosecutor´s Office (MPF)	13
Federal Police (PF)	10
Total	51

• Source: CRIP, 2016.

In this paper, the interviews were analyzed and grouped in two thematic block, that related to the two dimensions of procedural ecology - cooperation and information. We sought to capture the kind of logic involved in the interactions between the accountability institutions (i.e. whether it is competitive or collaborative logic) and, secondly, how institutions treat and handle their information (including the informational obstacles institutions build and how they act as inhibitors of higher levels of interaction). Interviews were conducted using a semi-structured data collection instrument. The choice of this instrument was justified because the agents interviewed represent a bureaucratic elite. In elite interviews, the semi-structures questionnaire allows the respondents to present their answers within their own words, delineating the general framework about the role and interactions among accountability institutions, as well as the issue of corruption control in Brazil (Aberback and Rockman, 2002).

The logic behind the interactions: competition or cooperation?

The analysis of the interviews tended to reinforce the assumptions outlined in the literature review: in general, there is a visible lack of coordinated interactions and the predominance of a competitive relationship between the institutions - “it is a job done more or less arbitrarily without coordination among the institutions”(Respondent 2, MPF, 2014). The interactions do occur, but by way of a predatory competition between the accountability institutions, which are greatly concerned about media attention and usually do not focus on symbiotic cooperation and exchanges. Respondents, in general, reaffirmed that significant changes did occur in each of the institutions of the system. However, the system still lacks coordinated and joint action, thereby hampering the progress of cases.

At first, the causes for the competitive logic were sought in personal explanations: vanity and reluctance to share the merit of investigations. Furthermore, the interactions between the institutions responded to the political juncture. Depending on the political juncture, the body can be open to the interactions or averse to them.

However, apart from personal limitations and political reasons, the answers allow us to draw a scenario of institutional crystallization, characterized by inflexibility and a moving away from each other. In the words of a Prosecutor, the accountability institutions in Brazil are very firmly attached to their own role and competence. The Brazilian system of accountability is typified by the respondents as an assembly line. Each institution fulfills its role and sequences the activities. Thus, the institutions are not perceived as being aimed at coordinating. Depending on the institutional origin, the respondents portray their role as something unique that can be resolved internally, without the need to coordinate with other accountability institutions. For instance, this is the case of the interactions in the case of Special Audits, involving the Federal Courts of Accounts and the CGU.⁸

This excessive independence at least is perceived as something harmful. For the most part, Brazil lacks an “institutional culture” focused on coordination. The worst possible outcome would stem from institutions that do not communicate; they do not know what each one is investigating and monitoring - something that they call “institutional blindness”. This appears, for example, in the relations between the Federal Courts of Accounts and the Office of the Comptroller General and within the sphere of the MPF itself. The structure of the relationship between accountability institutions in Brazil was perceived by 80% of the respondents, who affirm that they are small islands, dispersed and without a more solid joint action.

Why do I say blindness? Sometimes it happened that a colleague who works in the civil arena was investigating a situation and the colleagues in the crime arena did not even know. […] Sometimes people in crime were hitting each other with a case of corruption in which there was already a civil investigation and we did not even know it, and the civil investigation already had all the necessary evidence, but since there was no centralized anti-corruption coordination, sometimes internal contact was difficult (Respondent 12, MPF, 2014).

Hence, the coordination among institutions in the Brazilian accountability system is associated to an institutional blindness, with an overlay of various institutions acting in the same case. Respondents perceive that coordination is not sufficiently fluid and this would negatively affect the outcome of the accountability system’s procedural ecology.

The institutions do not perceive themselves as part of the same organizational cluster, the same procedural ecology around an accountability system. This lack of collective identity, as being part of the same process, would be what leads to a duplication of structures, with the creation of parallel procedures and even parallel organizations, instead of deepening the interactions and coordination with the institutions that already exist. The problem is often narrated as a matter of the institutions indiference of its own functions and powers in the accountability system. Without this prior self-knowledge, interactions are severely damaged, because it is difficult to discern the complementary areas. For example, the public prosecutors many times do not understand the role of the other institutions of the system: “demands from the prosecution can justify a specific audit or not, for example, a trip to one municipality. The vast majority does not justify this because it would be impractical. We receive about a thousand demands per month” (Respondent 25, CGU, 2014).

The difficulties for institutional cooperation exist, but they do not mean that the interactions between the institutions do not occur. Respondents report that these interactions within the accountability system occur informally, especially between the Comptroller and the Federal Police. But, in the absence of a broader structured accountability system, the interactions occur more due to informal exchanges. All respondents pointed out that informal contacts matter more than formal rules.

There is no effective coordination, especially when the prerogatives and competences clash, leaving it up to the volition of agents within the institutions of the system to establish cooperation efforts. The problem seems to be the lack of a “strategic integration”, designed from above, with clear objectives. Different institutions do not know their own roles, even less the role of others and how they can beneficially help their work. There is an absence of a strategic definition of system priorities - the rule is that each institution sets its own priority and reluctant to answer the calls and requests coming from other parts of the accountability system.

The best examples of coordination occur exactly when there is mutual knowledge and information sharing. For example, between the Comptroller and the Police, they call each other to help in investigations and Special Operations - the CGU offers expertise in analyzing government data and the PF offers physical protection for the auditors. This proximity allowed them to understand each one´s role in the accountability system.

The other crucial point for the procedural ecology to take place in the direction of coordination efforts is the level of information exchange, analyzed in the next section.

Exchange of information

The other difficulty highlighted in relation to the coordination between the institutions comes from the absence of a systematic exchange of information. Each institution has its own database, which is not usually shared, at least not in an effective way. It is impossible to exercise accountability without the integration of information and the principle of publicity. In this sense, the first dimension is closely linked to the second one: the major obstacle to an accountability system in which the institutions act in a coordinated and integrated manner is the lack of integration between the information generated by each institution.

This means that the institutions engage in a competitive logic, fighting for space in the public sphere, especially with regards to control and publicity of information about corruption. This logic prevents fruitful exchanges of information and the sharing of databases, making it impossible to achieve an integrated network. Without this space for informational intersection, the final result in terms of accountability would be jeopardized.

On one hand, we find that the performance of the accountability actors takes place in the dimension of institutional autonomy combined with an internal learning that creates little incentive for institutional cooperation or for the procedural ecology efficiency. On the other hand, as respondents pointed out, the major obstacle to a more integrated and coordinated accountability system is the absence of a systematic exchange of information between the institutions.

The problem of information exchange has three dimensions within the accountability system (1) - among institutions; (2) - the problem of secrets and secrecy foreseen in the Brazilian legislation; (3) - information as a strategic resource to further the interests of the agents. Not only the institutions do not share data between them, but they also state the difficulties in obtaining data even when they have a warrant (especially from banks and the Federal Revenue Office).

Moreover, some institutions build a strong relationship with the media, leaking information in an intended way. Regarding the exchange of information among institutions, we tested the network of contacts established between the institutions under analysis. Surprisingly, the vast majority of respondents said that they never had any denied contact: “denying a contact with another institution at the federal level is an arbitrary form of behavior, no one has the courage to do it” (Respondent 26, TCU, 2014). But if denying is not the rule, slowness and delays are. There are complaints from all parts of the accountability system about the lack of agility in information sharing. There are excessive delays in obtaining crucial information for investigations, for example.

When respondents elaborate more on their responses, it becomes clear the existence of communication barriers beyond the denial: “one thing is having the access denied, another is when you are trying to communicate and that communication does not evolve, there is no answer, there are communication barriers.” (Respondent 48, CGU, 2014). The reality is that institutions do not clearly and openly deny contacts or information (which would be a break in constitutional principles), but they do set difficulties and obstacles to bar the access to the information. It is important to note that the contacts and this process of information exchange often occur informally, without any institutional provision for the use of strategic information by the institutions of the system. These informal exchanges and rules do not create binding decisions, grounded in a solid process and with a defined flow. They strengthen the discretionary power of the agents who have this information as a strategic resource.

Regarding the second issue, Brazilian law rigidly determines tax confidentiality and legal secrecy, imposing strict rules for the financial system to protect the confidentiality and privacy of its clients.⁹ Specially at the Federal Court of Accounts, there is a widespread criticism of the difficulties in accessing tax and financial secrets. Its members highly regret the absence of institutional procedures that are allowed only to other institutions of the accountability system, such as the Federal Police and its power to breach the confidentiality rule. Far from being seen as potentially beneficial, the seals of secrecy are questioned all the time as major obstacles to the internal work of the institutions and, in particular, to the interactions between them.

In this aspect, the TCU shows itself to be a different institution - it is the one that most questions the existence of the secrecy and the impossibility of breaking it. In this case, there is a competition for police instruments. The competitive logic is so entrenched that accountability institutions in Brazil fight for top positions in the system and dispute others’ institutional procedures. This type of dispute reflects a system of institutions that cannot establish strategic actions and processes of exchange of information and data. It is necessary to define better and more clearly the roles of each institution, establishing coordination mechanisms and a stronger procedural ecology.

The Brazilian Federal Revenue and the banks are systematically placed as the villains of the access to information and the establishment of interactions. In the case of the banks, they systematically disobey court decisions about sharing banking information. For example, CGU and MPF face enormous obstacles when they need to access the bank accounts of mayors: “If you ask me if we have had any negative contact of any federal entity that has prevented us from our work against corruption, I answer: Yes, federal banks, alleging, in my view unjustifiably, bank secrecy.” (Interview 25, CGU, 2014). In the case of the Federal Revenue, even with judicial authorization, it does not usually deliver the data requested by other institutions. In this sense, access to information appears as a powerful bargaining tool. Many institutions staunchly protect themselves behind secrecy rules. These rules are important in many cases but, unfortunately, they are so impregnated in the control bureaucracies that it becomes truly impossible to investigate malfeasance.

Finally, the control of information retained by accountability agents is presented as a key strategic resource. The logic behind the dispute between the institutions is not only about the procedures, but it is also about public opinion. Given the context in which the institutional capacities to unveil corruption were increased, accountability institutions start to play to the audience in order to consolidate their interests in the political arena. Combining the strongly proactive attitudes of the agents with a structure that does not create incentives for institutional cooperation, information becomes a strategic resource, both to pressure the political system and to increase the legitimacy of their actions in the eyes of society.

The result of this procedural ecology is that in the context of a structure with few incentives for institutional cooperation, based mostly on the proactive actions of its agents and on informal ways of information exchange, the Brazilian system of accountability institutions ends up challenging the institutional change. Within these institutions, taken individually, there were incremental changes that increased state capacity to exercise control of public policies and unveil corruption cases both in politics and public administration. However, the challenge is to consider that, in addition to incremental changes that empower the accountability institutions in the Brazilian state, we must also consider the highly informal and less systematic procedural ecology that encourages competition, the absence of solid procedures for control activity sequencing and of dynamic data exchange, the deliberate creation of obstacles and the political use of information.

Discussion and Conclusion

Systems are mutual interactions between complex elements, providing knowledge and a common cause to its elements, which interact to achieve an end (Bertalanffy, 1968). Systems theory proposes that systems can be open or closed. The advancement of theories of social systems, since Parsons (1969), considers that the systems are open, emphasizing the elements of exchanges with regard to knowledge, information, people and resources. However, as pointed out by Luhmann (2000), the condition for the establishment of exchange between systems is that each of them should be closed so that they can establish differentiation with respect to other systems. The closure is in relation to its own operative basis, differing from all the rest and creating its operating limits. By doing that, the systems create conditions for establishing their openness to the environment, opening them also to external demands that arise from the internal conditions they created (Luhmann, 2000). Our theoretical advance is to consider that in addition to closing and logical response to the demands, the process through which exchanges smoothly and interdependently take place should be considered in the analysis. In this case, the concept of procedural ecology does that, by looking at systemic variables in the process of institutional change.

Cooperation or conflict between institutions of a given system is a necessary condition but not sufficient to explain the process of institutional change. Conflict, as well as cooperation may be the elements that explain the change or determine a process of change. On their own, cooperation and conflict are situational variables. However, cooperation or conflict between institutions is essential to explain the change brought about by systemic bias, since they can alter the results of the interaction between system elements as well as indicate the direction of change. Conflict can result in positive changes, as well as cooperation between institutions can result in negative changes. In any case, the condition for thinking about the influence of conflict or cooperation on the process of institutional change is the existence of a procedural ecology.

The systemic variables need be consider when analyzing the process of institutional change, in which the interactions between institutions matter not only for the results, but also for internal processes, rules and organizational routines inherent to the institutions. According to the procedural ecology argument that we have developed here and the case study of Brazil, the process of institutional change should consider the systemic variables, along with the endogenous and exogenous factors. Thus, understanding the systemic factors of institutional development involves considering: (1) - the flow and sequencing of processes internal to the system; (2) - the agents behavior with regard to institutional interactions; (3) - the establishment of joint organizational goals for the system; (4) - the control and use of information by the system elements; (5) - the response to internal situations and demands on the part of each of the elements of the system; (6) - the response to external situations and demands on the part of each of the elements of the system.

By considering these analytical factors whenever appropriate, the process of change can be understood in terms of the interdependence between institutions. These factors affect the institutional development process, favoring either change or the status quo. Systemic variables thus influence institutional incrementalism, in a way that the agents still prefer incremental changes in routines and processes in response to endogenous and exogenous environmental factors. However, this incrementalism may result either in a differentiation and greater closing of each of the system elements, favoring institutional conflict, or it can lead to an incremental process that is capable of, through cooperation and coordination of system elements, driving the process of interdependence between overlapping and complementary institutions.

The theoretical contribution of this study focused on the concept of procedural ecology an aims at advancing the theories of institutional change. The case study of the Brazilian system of accountability institutions made it possible to point out that systemic variables, present in the interaction between institutions, are important into any study of institutional change processes. Systemic variables can create situations of change or reinforcement of the status quo depending on the context of cooperation or conflict between the institutions of an organizational cluster. Procedural ecology is then important when considering institutional development, in which the incremental changes are directed not only at the rules and routines internal to the institutions, but also at the institutions’ interactions.

The empirical contribution is to consider that the development of accountability institutions in the case of Brazil can be studied with reference to systemic variables. The fight against corruption and the control of public administration has advanced in the country due to incremental changes in the institutions, which enabled them to change internal rules and routines. This generated fundamental learning, advancing institutional capacity to exercise control over the bureaucracy and political institutions. However, the system still results in low efficiency, due to systemic factors, especially in the lack of sequencing and complementariness in control actions, breaking the accountability cycle exactly in the moment in which public officials are starting to be called to account for their acts. Unfortunately, this has lead to the persistence of impunity perception despite the notable recent advances of the Brazilian accountability institutions.

References

• ABERBACH, Joel; ROCKMAN, Bert A. (2002). “Conducting and Coding Elite Interview”. OS: Political Science and Politics, vol. 35, n° 4, p. 673-676.
• ABRANCHES, Sérgio Henrique H. (1988). “Presidencialismo de coalizão: o dilema institucional brasileiro”. Dados - Revista de Ciências Sociais, vol. 31, n° 1, p. 5-34.
• ARANTES, Rogério B. (2011a). “Polícia Federal e Construção Institucional”. In: AVRITZER, Leonardo; FILGUEIRAS, Fernando (orgs.). Corrupção e Sistema Político no Brasil Rio de Janeiro: Editora Civilização Brasileira.
• ARANTES, Rogério B. (2011b) “The Federal Police and Ministério Público”. In: POWER, Timothy; TAYLOR, Matthew (eds.). Corruption and Democracy in Brazil. The Struggle for Accountability. Notre Dame: University of Notre Dame Press.
• BALBE, Ronald da Silva (2013). Controle Interno e o Foco nos Resultados São Paulo: Editora Fórum.
• BERTALANFFY, Ludwig von (1968). General System Theory. Foundations, Development, Applications. New York: George Braziller.
• DIMAGGIO, Paul; POWELL, Walter (1983). “The Iron Cage Revisited: Isomorphism and Collective Rationality in Organizational Fields”. American Sociological Review, vol. 48, n° 2, p. 147-160.
• GALVIN, Daniel J.(2012). “The Transformation of Political Institutions: Investments in Institutional Resources and Gradual Change in National Party Committees”. Studies in American Political Development, vol. 26, n°1, p. 50-70.
• GAILMARD, Sean; PATTY, John (2007). “Slackers and Zealots: Civil Service, Policy Discretion, and Bureaucratic Expertise”. American Journal of Political Science, vol. 51, n°4, p. 873-89.
• GAILMARD, Sean; PATTY, John (2013). Learning While Governing: Expertise and Accountability in the Executive Branch. Chicago: University of Chicago Press.
• HALL, Peter; TAYLOR, Rosemary (1996). “Political Science and Three New Institutionalisms”. Political Studies, vol. 44, n° 5, p. 936-957.
• LIMONGI, Fernando (2006). “A democracia no Brasil. Presidencialismo, coalizão partidária e processo decisório”. Novos Estudos - CEBRAP, nº 76, p. 17-41.
• LINDBLOM, Charles E. (1959). “The Science of “Muddling Trough”. Public Administration Review, vol. 19, n° 2, p. 79-88.
• LOUREIRO, Maria Rita; OLIVIERI, Cecília; BRAGA, Ana Cristina (2011). “Bureaucrats, Parties, and Interest Groups”, in: FONT, Mauricio; RANDALL, Laura (eds.). Brazilian State: Debate and Agenda Lanham and New York: Lexington Books, p. 111-130.
• LOUREIRO, Maria Rita; TEIXEIRA, Marco Antônio C.; CACIQUE, Thiago (2009). “Democratização e Reforma do Estado: o Desenvolvimento Institucional no Tribunais de Contas no Brasil Recente”. Revista de Administração Pública, vol. 43, n° 4, p. 739-772.
• LUHMANN, Niklas (2000). Social Systems Stanford: Stanford University Press.
• MAHONEY, James; THELEN, Kathleen (2010). “A Theory of Gradual Institutional Change”. In ______. (Eds.). Explaining Institutional Change: Ambiguity, Agency and Power. Cambridge. Cambridge University Press.
• MAINWARING, Scott (2003). “Introduction: Democratic Accountability in Latin America”. In: MAINWARING, Scott; WELNA, Christopher. Democratic Accountability in Latin America Oxford: Oxford University Press.
• MELO, Marcus A.; PEREIRA, Carlos; FIGUEIREDO, Carlos M. (2009). “Political and Institutional Checks on Corruption: Explaining the Performance of Brazilian Audit Institutions”. Comparative Political Studies, vol. 42, n°9, p. 1217-1244.
• NORTH, Douglass C. (1990). Institutions, Institutional Change and Economic Performance Cambridge: Cambridge University Press.
• OLIVIERI, Cecília; TEIXEIRA, Marco Antônio Carvalho; LOUREIRO, Maria Rita; ABRUCIO, Fernando (2013). “Organizational Learning of Controllers and Controlled Agencies: Innovations and Challenges in Promoting Accountability in the Recent Brazilian Democracy”. American Journal of Industrial and Business Management, nº 3, p. 43-51.
• OLIVIERI, Cecília (2010). A lógica política do controle interno O monitoramento de políticas públicas no presidencialismo brasileiro. São Paulo: Annablume.
• O’DONNELL, Guillermo (1999). “Horizontal Accountability in New Democracies”. In: SCHEDLER, Andreas; DIAMOND, Larry; PLATTNER, Marc. The Self-Restraining State. Power and Accountability in New Democracies. Boulder: Lynne Rienner Publishers.
• O’DONNELL, Guillermo (1996). “Uma Outra Institucionalização: América Latina e Alhures”. Lua Nova, nº 37, p. 5-31.
• OLSEN, Johan P. (2008). “Change and Continuity: An Institutional Approach to Institutions of Democratic Governance”. European Political Science Review, vol. 1, n° 1, p. 3-32.
• PARSONS, Talcott (1969). Politics and Social Structure New York: Free Press.
• PIERSON, Paul (2004). Politics in Time. History, Institutions, and Social Analysis. Princeton: Princeton University Press.
• REZENDE, Flávio da Cunha (2009). “Analytical Challenges for the Neoinstitutional Theories of Institutional Change in Comparative Political Science”. Brazilian Political Science Review, vol. 3, n° 1, p. 98-126.
• SPECK, Bruno W. (2000). Inovação e Rotina no Tribunal de Contas da União. O Papel da Instituição Superior de Controle Financeiro no Sistema Político-Administrativo Brasileiro. São Paulo: Fundação Konrad-Adenauer Stiftung.
• TAYLOR, Matthew; PRAÇA, Sérgio (2014). “Inching Toward Accountability”. Latin American Politics and Society, vol. 56, n° 2, p. 27-48.
• TAYLOR, Matthew; BURANELLI, Vinicius (2007). “Ending Up in Pizza. Accountability as a Problem of a Institutional Arrangement in Brazil”. Latin American Politics and Society, vol. 49, n° 1, p. 59-87.
• WEISS, Andrew; WOODHOUSE, Edward (1992). “Reframing Incrementalism: A Constructive Response of the Critics”. Policy Sciences, vol. 25, n° 3, p. 255-273.

• ⁴
  We are grateful to Professor Johan P. Olsen for his generous reading of the manuscript and comments. Clearly, responsibility for limits and problems lies solely with the authors.
• ⁵
  An out-of-court solution is a way of settling conflicts without going through a legal hearing. In Brazil, it is the prerogative of the Public Prosecutor´s Office to propose legal solutions in order to resolve conflicts. In order to exercise accountability and control the public administration, the prosecutors can propose recommendations and terms of conduct adjustment. The recommendation is solely the tool that prosecutors use to alert public officials about the need to resolve a situation that may be irregular or that can lead to one. The terms of conduct adjustment (TAC) is a legal instrument to obtain, from the causer of damage to diffuse and collective interests, an enforceable obligation of do’s and don’ts, by which whoever is responsible for the damage assumes the duty to align their conduct to the legal requirements, under penalty of sanctions laid down in the term.
• ⁶
  The Fiscal Responsibility Law, created by the Complementary Law 101/2000, sought to bring together the initiatives to achieve altogether the consistent expansion of state efficiency and fiscal accountability to control public finances. The most striking aspect of this Law, in addition to the fiscal adjustment, was that it established several initiatives in the realm of transparency of public accounts processes - its Article 48, and 59. Regarding the extent of control, the Fiscal Responsibility Law distributed the control powers to the Federal Court of Accounts, the internal control system to the Executive branch - exercised today by the Office of the Comptroller General (CGU) - and the Public Prosecutor´s Office. These institutions have the capacity to control the targets set in the Budget Guidelines Law, the public expenditure on personnel and the public debt. By reaffirming these empowerments, the Law created a favorable institutional framework for operation of the internal and external control instruments in the Brazilian government.
• ⁷
  It is important to emphasize that the centralization of the internal audit did not always occur only within the scope of the Ministry of Finance. In some moments these activities were transferred to the Ministry of Planning, Budget and Management.
• ⁸
  The Special Accounts Instrument (Tomada de Contas Especial) is used by the control institutions in Brazil to compensate for the losses regarding public resources and has its own formal rite. It can only be evoked after all other measures of damage repair have been exhausted and failed. As stated in the Instruction / TCU No. 71 of 11.28.2012, it is the responsibility of the CGU to issue the report about Special Accounts processes and give an opinion as to whether the facts have been properly investigated, which rules or regulations have been infringed. It has also to ensure the correct identification of the responsible and determinate the amount of the damage and of any possible repayments. The Federal Court of Accounts can also start this instrument and is the only responsible to judge the Special Account and take appropriate measures to redress the damage.
• ⁹
  In Brazil, tax confidentiality is regulated by the Complementary Law 105 of 2001, which establishes that all financial organizations are forbidden to provide any information about their customers or citizens of Brazil. The breakdown of tax and financial secrecy can only be achieved through a legal process and the express permission of the Judiciary. This Law is in opposition to the process of enhancing governmental transparency resulting from the Law 12.527 of 2011, which addresses the access to public information. The latter stipulated transparency procedures and the principle of publicity as the basic condition (the secrecy is the exception). It also stipulated the viability and necessity of transparency portals throughout the Brazilian public service, including the organizations from the three branches of power and the public prosecutors, in all three levels of the federation (at the federal, state and municipal level). Under the Law of Access to Information, tax secrecy remained a possibility but an exception to the norm of publicity. This was contrary to the interests of the Court of Accounts and of the Comptroller General, who wanted access to tax data to inform controlling processes and to tackle corruption.

Publication Dates

• Publication in this collection
  03 June 2019
• Date of issue
  Jan-Apr 2019

History

• Received
  08 Mar 2019
• Accepted
  30 Mar 2019