Relational regulation in the Brazilian Ministério Publico: The organizational basis of regulatory responsiveness


Salo Vinocur Coslovsky, Robert F. Wagner School of Public Service, New York University, 295 Lafayette Street, 3rd floor, New York, NY 10012, USA. Email:


Recent research on regulatory enforcement has been showing that the best way to enforce protective regulations is to thoroughly adapt the remedy to the problem at hand. Unfortunately, this is far from easy, as organizations and work environments all too often encourage street-level bureaucrats to standardize and simplify their practices. In an attempt to bridge this gap, many scholars equate responsiveness to preprogrammed escalation. This article analyzes how the Brazilian Ministério Público (MP) promotes a more ambitious form of responsiveness, here called relational regulation. While most of the formal features of the MP encourage routinization, groups of reformist prosecutors identify important cases, recruit external allies, and jointly devise innovative and context-specific solutions to problems of non-compliance. Crucially, this mode of action is not an anomaly, the result of individual proclivities or specific and formal job assignments. Rather, it is systematically produced by a mostly parallel and covert organization that operates within and expands beyond the MP itself. These two logics – routinized and custom-made responses – compete, cooperate, interpenetrate, and find common ground within the same organizational umbrella, and through their interaction they create an enforcement agency that is more robust, reliable, and responsive than allowed by current theory.

1. Introduction

Over the past several years, research on regulatory enforcement has been showing that the best way to enforce protective regulations is to thoroughly adapt the remedy to the problem at hand; that is, enforcers ought to threaten the diffident, punish the recalcitrant, and educate the uninformed. In its different guises, this type of engagement has been called responsive (Ayres & Braithwaite 1992; Braithwaite 2005), flexible (Bardach and Kagan 1982), tit-for-tat (Scholz 1984, 1991) and creative (May & Burby 1998). Although easy to describe, this type of tailoring can be quite difficult to achieve, as organizations all too often routinize, simplify, and standardize their procedures (Lipsky 1980; Wilson 1989; Feldman 1992).

To overcome this difficulty, some scholars have suggested preprogrammed escalation as embodied in the “enforcement pyramid” (Ayres & Braithwaite 1992). Unfortunately, this solution falls short in important ways. As pointed out by Nielsen and Parker (2009), regulators rarely have the communication and relational skills as well as technical and emotional intelligence to escalate and de-escalate appropriately. Moreover, compliance often requires that enforcement agents transcend the issuing of fines, warnings or advice to one infringer at a time (Coslovsky 2008; Pires 2009). As indicated by Sparrow (2000), effective enforcement requires that agents identify important problems and solve them, and this may require the coaxing of collective action (Scholz & Gray 1997) and the elimination of root causes that hide at the end of meandering and idiosyncratic causal chains (MacDuffie 1997). Finally, even if regulators and the regulated agree on the course ahead, they may lack the mandate, resources, and skills to produce actual results. Confronted with so many difficulties, can an enforcement bureaucracy truly adapt the remedy to the problem at hand?

This article examines how prosecutors associated with the Brazilian Ministério Público (MP) strive to overcome this challenge when enforcing labor and environmental regulations. The MP is composed of approximately 9,000 prosecutors1 endowed with significant discretion. As described in the literature on prosecutors, discretion cuts two ways. For some scholars, it is the raw ingredient of abuse, corruption, and injustice to be controlled through external review, increased transparency, financial rewards, and managerial change (Alschuler 1968; Davis 1969, 2000–2001, 2007; Vorenberg 1981; Bibas 2009). For others, it is a fact of life; discretion never disappears, only moves around (Remington 1993). If so, the challenge is to create “the good inspector” (Bardach and Kagan 1982), who uses discretion to enact a contextual and effective problem solving approach (Goldstock 1991; McCoy 2003; Berman & Feinblatt 2005; Garrett 2007; Levine 2007; Barkow 2009).

Studies of the Brazilian MP fall into similar camps. While some scholars examine the challenges presented by prosecutors' excessive discretion and limited accountability (Arantes 2002; Sadek & Cavalcanti 2003; Hochstetler & Keck 2007; Kerche 2007; Nóbrega 2007; Taylor & Buranelli 2007), others take discretion for granted and call attention to prosecutors' ability to make the law matter (McAllister 2008) and break the endless regress (Mueller 2006).

Moving beyond the debate about the merits of discretion, and reporting on a classification that is native to members of the MP itself, Silva (2001) divides prosecutors into two ideal-types: “fact-oriented” versus “office-bound” prosecutors. Office-bound prosecutors strive to process every legal case that comes their way and are adamant about preserving their independence from external influences. Conversely, fact-oriented prosecutors see themselves as social problem-solvers (p. 138). Instead of processing individual cases from a distance, and as reported also by Cavalcanti (1998), Hensler (2005, unpublished data), and McAllister (2008), these prosecutors establish close relationships with non-governmental organizations (NGOs), social movements, public bureaucracies, and other entities to address important social problems.

The identification of office-bound and fact-oriented practices is unique to neither prosecutors nor Brazil. It is characteristic of broader organizational life, and has been anticipated in the literature on the performance of organizations. For instance, the case-processing approach of office-bound prosecutors corresponds to Lipsky's (1980) description of the overwhelmed street-level bureaucrats forced to routinize, simplify, and standardize their procedures. It resonates with Silbey's (1980–1981) description of case processing in the attorney general's office in Massachusetts. Alternatively, the problem-solving approach of the fact-oriented prosecutor corresponds to Sabel's (2006) characterization of the robust organization, endowed with search routines and error-detection and feedback mechanisms to customize its responses to the particularities of the problems at hand. Some of the features associated with robustness – particularly the systematic preservation of organizational slack – can be identified in Johnson's (2001) study of Japanese prosecutors.

The case-processing approach of office-bound prosecutors and the problem-solving approach of fact-oriented prosecutors were modeled by Repenning and Sterman (2001) as “work-harder” and “work-smarter” loops. However, these authors portrayed these loops as self-reinforcing and mutually exclusive. According to this view, agents or organizations will sustain one mode or the other, but not both. March (1991) advanced a different perspective, pointing out that organizations ought to combine “exploration” and “exploitation.” The former includes actions such as search, risk taking, discovery, and innovation, and corresponds to problem-solving. The latter includes actions such as enforcement and implementation, and corresponds to case-processing. According to March (p. 71), “maintaining an appropriate balance between exploration and exploitation is a primary factor in system survival and prosperity.” Called ambidexterity by those who study private sector competitiveness, the sometimes precarious and uneasy coexistence of these abilities is widely believed to enhance organizational survival, flexibility, learning, and effectiveness (Tushman & O'Reilly 1997; Gupta et al. 2006; Raisch & Birkinshaw 2008; Raisch et al. 2009; Fang et al. 2010). This proposition is further supported by Sparrow (2008), who conditions the benefits of ambidexterity on the existence of proper managerial structures and controls (pp. 162–165).

In this article, I take Silva's (2001) findings of a divided MP a step further and examine how separate factions of prosecutors, representing the two modes of action described above, inhabit and foster separate organizational entities that coexist under the same organizational roof. As a result of 18 months of fieldwork conducted in Brazil between 2007 and 2008, and interviews with more than 50 prosecutors and 100 representatives from NGOs, private firms, government agencies, and unions,2 I found out that the MP is best understood as a composite organization, in which prosecutors associated with each of these factions build and maintain separate but overlapping organizational policies, processes, and structures. Under a unified banner, these entities interact, fight for resources, cooperate, monitor each other, establish alliances with external actors, and learn how to minimize their conflicts through the careful use of language. Instead of detracting from overall performance, this dynamic fosters ambidexterity and makes for a more responsive and resilient organization.

The remainder of this article describes these two factions, their respective organizational logics, and how they interact. The next section, “The Brazilian prosecutor in action,” introduces the Brazilian prosecutors and uses two examples to illustrate the practice of social problem-solving, or, as will be called here, relational regulation. Section 3, “Prudent conformity: Formal structures, incentives, and constraints of the MP,” describes how most of the formal features of the organization encourage case-processing. Section 4, “Relational practices that sustain creative enforcement,” explains how a faction of prosecutors joins forces with NGOs, unions, social movements, and other entities to build and maintain a mostly parallel, covert, and overlaid organization that supports and encourages this different mode of action. Section 5, “The interaction between the two factions,” analyzes how the dynamics between these two modes of action creates an organization better able to confront the challenges faced by the MP. Section 6 concludes.

2. The Brazilian prosecutor in action

Brazilian prosecutors are a corps of approximately 9,000 prosecutors and 10,000 support staff (Ministério da Justiça 2006) who work out of field offices spread throughout the country. They are associated with one of 26 state or four federal procuracies that operate under a unified body of laws and guidelines. According to these laws, and in addition to prosecuting felonies and misdemeanors and arguing cases in open court, Brazilian prosecutors can conduct inspections and investigations, subpoena documents, organize public hearings, dismiss cases, and initiate (and settle) civil lawsuits in a range of areas, including environmental preservation, minority rights, public health, and more. Based on this range of prerogatives, I claim that Brazilian prosecutors are comparable to regulatory enforcement agents. Not surprisingly, and as will be examined in Section 3, most of the time they adopt one or more of the enforcement styles already identified in the literature, including sanctioning, deterrence, persuasion, rule following, retreatism, case handling, bargaining, and cooperation (Silbey 1980–1981; Bardach & Kagan 1982; Silbey & Bittner 1982; Hutter 1989; Gormley 1998; Winter & May 2000; May & Wood 2003). However, and as will be discussed in Section 4, on some occasions some prosecutors do something else altogether: they engage in a kind of root-cause analysis and social problem solving that goes beyond what Sparrow (2000) commends as the best regulatory craft, and Bardach and Kagan (1982, pp. 152) describe as the “ideal to which a regulatory agency can only aspire.” The following two vignettes illustrate this practice.

2.1. Affordable housing

To attend to an immense demand for affordable housing, unscrupulous developers sometimes create parceling schemes in land that is publicly owned, environmentally protected, or contaminated. Prospective dwellers move in anyway, and start lobbying government agencies for public services such as water, sanitation, electricity, and transportation. Naturally, each improvement ameliorates the lot of occupants, who are often destitute and in dire need of these services. However, improvements also attract more people and make the illegal occupation ever more difficult to resolve.3

The intractability of the problem is obvious and prosecutors are often at a loss about how to proceed. Not surprisingly, many adopt a routinized approach: they request documents, try to indict some people (but actionable defendants are hard to find), and by and large push paper around while these communities grow and establish ever more permanent roots.

Despite these difficulties, Jessica Riveiro, a state prosecutor based in São Paulo state, with more than 15 years of experience and assigned to the collective affairs desk, found a way around this apparent dilemma. She decided that the municipal government should encircle the illegal settlement with a fence to arrest its growth. Needless to say, dwellers opposed this idea, and Ms. Riveiro faced protests and even death threats. Moreover, the municipal government refused to invest in such an unpopular intervention.4

To understand what happened next, one must realize that Ms. Riveiro is the prosecutor in an industrial town with a sizeable backlog of environmental infractions. From past experience she knew that firms avoid paying cash for reparations owed. Rather, they prefer to pay in-kind, which is cheaper and allows firms to reframe the legal settlement as an act of corporate social responsibility. Familiar with these preferences, Ms. Riveiro located within her docket a defendant about to settle and offered a deal: she would drop the ongoing suit if the firm agreed to provide hundreds of yards of sturdy chain-link fence. However, the squatters still opposed Ms. Riveiro's plan, and she knew that they could easily knock the fence down overnight. As part of this ongoing negotiation, she learned that water supply at this settlement remained unreliable, particularly during the hot summer months. So she proposed the dwellers a deal: agree to the installation of the fence, and even care for it, in exchange for improved water services. The next step was to bring the water agency into the equation. While its representatives agreed to extend a waterline to improve services in that settlement, the agency was unable to fund the construction. The mayor offered municipal laborers and tractors to dig the ditches and install the pipes, but Ms. Riveiro rejected the offer; in fact, she “forbade” the mayor from doing it. She wanted the dwellers, many of them recent migrants from other parts of the country, to “voluntarily” work in this construction as a way to increase their sense of ownership and identification with the place.

As a result of Ms. Riveiro's efforts, all of the pieces fell into place in a way that no one could have envisioned from the start. The ever-growing illegal parceling scheme was contained, its dwellers gained better access to water services, the large petrochemical firm settled its pending legal case, and a variety of public agencies performed the services that citizens expect of them.

2.2. Shrimp farming

In the early 1990s, entrepreneurs in Northeast Brazil found ways to cultivate shrimp at a relatively low cost and high productivity. Due to a crisis in other producing countries, the Brazilian industry quickly boomed, leading to increased income and employment in a region that needed both. However, this development was not without immediate environmental consequence. In order to access abundant free and clean water, and to use the tidal flow to save on pipes, pumps, and electricity, shrimp farmers often establish their farms directly on top of mangroves, which destroys these publically owned, legally protected, and environmentally sensitive areas. Ultimately, these social costs offset a large share of the benefits stemming from this industry and force law enforcers and policymakers to choose between a thriving industry and respect for environmental laws.

In many cases, the prosecutor in charge sides with the environment and starts legal proceedings requiring violators to close down or relocate. Yet Jairo Azevedo, a federal prosecutor with more than 10 years of experience and responsible for the environmental desk in a small state in Northeast Brazil, departed from this norm and did something else.

A group of approximately 100 small shrimp farmers had occupied a mangrove on the shore of a lake. Upon learning of this violation, Mr. Azevedo warned them of impending legal action. To his surprise, the farmers promised to ignore any court orders to leave and to resist if the police forced them out. Such a confrontation would be disastrous for all involved, and Mr. Azevedo started to search for alternatives. First, he needed basic data on these farmers: their population, income, productive capacity, and more. He asked the federal environmental agency for help. This agency responded that it would have to buy satellite pictures and the software to examine them and perhaps even sign a cooperation agreement with a local university to conduct the analyses. Unfortunately, it had no budget to cover these expenses. Mr. Azevedo insisted for approximately six months, but he also searched for support elsewhere. Eventually, he discovered that the state-level environmental agency had aerial pictures of that region, as well as staff to analyze the images. Moreover, this agency agreed to send census-takers to the farms to double-check the information, and in a few weeks the prosecutor had all the data he needed.

The next step was to locate an appropriate plot of land to accommodate the farmers. With the help of the state's land institute, Mr. Azevedo identified an unoccupied parcel of public land that was large enough to host all the farmers and located nearby so they would retain their connections to neighboring towns. He also found out that the governor had authority over this particular plot, and she agreed to donate the land. Still, the problem was not solved. Unlike the previous plot, in which the natural movement of the tide allowed farmers to fill and empty their tanks for free, the new plot required that they pay for pipes, pumps, and electricity. At the time of my visit, Mr. Azevedo was still struggling to find the money to pay for these interventions. He had met with different legislators to see if they would include an appropriation amendment in the state budget, but nothing was certain yet.

These two examples illustrate how far from writing legal opinions, memos, and briefs as well as arguing in open court a prosecutor may go when trying to “enforce the law.” Moreover, and instead of adopting any given enforcement style, the prosecutors used their legal powers, professional status, and public legitimacy to reach out to other institutions and reassign to public or private entities some of the costs, benefits, and risks associated with compliance in a way that made the change acceptable to all. Prosecutors refer to this practice as negotiating deals, building alliances, assembling networks, or stitching solutions together. This is the kind of practice that Silbey et al. (2009) call “sociological citizenship.” It is also related to what Winship (2006) calls “puzzle-solving.” In this particular setting, I call it relational regulation. Independent of the name that it takes, prosecutors who adopt this mode of action try to solve – or prevent – a whole class of important but seemingly intractable problems. This modus operandi is time-intensive and risky. Prosecutors cannot know in advance how much time and effort a stitching attempt will demand, and whether a solution will be reached at all. To survive in an environment of limited resources and competing priorities, this type of engagement must be rooted in, nurtured by, and actively promoted by an organization. The next two sections examine this organization.

3. Prudent conformity: Formal structures, incentives, and constraints of the MP

Why do some prosecutors engage in the deviant practice of relational regulation? The reasons are not immediately visible. In fact, and similar to many procuracies around the world (Silbey 1980–1981), the Brazilian MP encourages its prosecutors to deploy a cautious, reactive, and formalistic case-processing approach. This behavior emerges from three complementary forces. First, prosecutors are subjected to a heavy workload that allows for no slack and requires that they relentlessly simplify their routines just to meet mandatory deadlines. Second, most managerial policies adopted by the MP reward diligence and consistency rather than creativity, and cautious case-processing rather than creative problem solving. And third, most of the support structures, staff, and managerial devices provided by the organization help prosecutors manage the workload in a reactive, cautious, and formalistic way.

3.1. The overwhelming nature of the work

The São Paulo Ministério Público publishes an annual report with aggregate data on its activities (Corregedoria Geral do Ministério Público 2006). Even if these figures are underestimated, due to prosecutors who fail to submit their reports on time, the numbers remain staggering. Assuming that 1,400 prosecutors work 200 days per year and devote 4 days a week to case processing, in 2005 each prosecutor processed 39 cases, attended 4 hearings and trials, and wrote 4 briefs and motions per day. If each of these tasks took 12 minutes, prosecutors had to work for 10 uninterrupted hours a day for 4 days a week just to keep up. This leaves them with one day a week to perform other mandatory activities, such as visiting jails, prisons, halfway houses, and police precincts; attending school board meetings; reporting back to headquarters; attending training sessions; and meeting with lawyers, citizens, and other interested parties (in 2005, on average, prosecutors met with seven people per day).

Altogether, this means that unless prosecutors are exceedingly efficient in processing their cases (one prosecutor compared judicial files to a novel, “but with the pages shuffled”), delegating tasks to well-trained interns, and working long hours themselves, they will be permanently swamped. To help minimize this burden, the MP provides prosecutors with two predetermined devices. First, clerks attach a colored tape to the cover of every legal file to indicate an approaching deadline stipulated by the court. Second, the organization gives each prosecutor a printed manual and a CD-ROM with templates for the various arguments, briefs, and motions they must write. These templates, so prized that newly hired prosecutors cheered loudly once they received their copies, embody the MP's legal philosophy and indicate which violations are worth prosecuting and how. By using them, prosecutors routinize their procedures, avoid the risks of trailblazing, and deflect any negative reaction toward the organization.

3.2. Self-selection

Who decides to be a prosecutor? Contrary to the popular image that Brazilian prosecutors are activists who choose this career to add legal power to their social vision (Paul 2007), a large proportion of young prosecutors join the organization straight from law school and bring limited professional experience to the job.5 Moreover, my interviews indicate that many choose this career not out of a deep-seated conviction to serve the public, but rather through a process of elimination, often attracted by the prestige, high starting salary, and benefits associated with public legal careers in Brazil.

3.3. Admission criteria

In Brazil, aspiring prosecutors are selected through several rounds of written and oral examinations. Throughout this process, candidates are assessed not only for their detailed knowledge of Brazilian law, but also for their aptitude for the job. In theory, examiners could value characteristics such as public mindedness, creativity, social consciousness, street wisdom, and the ability to work well with others. However, my interviews revealed that examiners mostly care whether candidates dress and behave conservatively, hide political affiliations, demonstrate unwavering respect for authority and “lawyerly values” such as addressing examiners with the proper formalities and honorific titles, have flawless Portuguese, and pronounce Latin legal jargon with utmost accuracy. In short, candidates are admitted based on their disposition to conform, not trailblaze.

3.4. Formal training

Before being assigned to their first posting, newly hired prosecutors undergo an intensive orientation course in which experienced colleagues teach new recruits how to face the demands of the job. In São Paulo, this course – which I attended in full – lasts two weeks and the directive offered most frequently was not “be brave” or “be creative,” but rather “do not create any unnecessary trouble for yourselves” and – above all –“do not do anything that could jeopardize the stellar reputation of the agency you have just joined.” In brief, “be cautious.” Exemplifying this sense of caution, newly hired prosecutors immediately took to calling me – first behind my back and then openly –espião da Corregedoria (the spy from Internal Affairs). At the end of each day, I would join large groups of newly hired prosecutors who congregated at a nearby pub to unwind, and they repeatedly asked me not to record or report any loud behavior or less-than-moderate drinking. Even if they would have to be convicted of a felony confirmed at appeal to be expelled from the organization, these prosecutors feared being thrown out and understood that reticence was the best way to blend in.

3.5. Apprenticeship

For approximately the first three years on the job, prosecutors are considered alternates. As such, they travel throughout the jurisdiction of the agency filling vacancies and replacing senior colleagues who go on vacation, take leaves of absence, or need an extra hand. These initial years serve as an apprenticeship, in which novice prosecutors experience a variety of jobs and work environments and interact with different senior colleagues. During these years, prosecutors from Internal Affairs keep track of the novices' work. They read copies of all their memos and briefs and conduct surprise inspections to check, in loco, whether their legal arguments are solid, use of language accurate, deadlines met, and even whether offices are physically well kept and the prosecutors are wearing proper attire. These issues are all formally reported and tracked. At the end of the apprenticeship, and in the absence of a truly egregious violation, alternates are awarded life tenure and assigned to a post in a small town.

3.6. Assignments and promotion

In São Paulo state, the statutory career of tenured prosecutors is composed of three rungs: (i) entrância inicial (small town); (ii) entrância intermediária (medium town); and (iii) entrância final (big town or capital). To fill these vacancies, prosecutors apply to posts as they become available. Formally, half of the vacancies are to be filled on merit and half on seniority, but experienced prosecutors readily admit that, in practice, merit is too controversial and difficult to measure.6 In São Paulo, all postings are filled on seniority (Macedo 2010). Also important, the organization keeps separate seniority lists for each of the three rungs and newly promoted prosecutors return to the bottom of the new list. This system minimizes intraorganizational conflict, but it also practically ensures that talent will be misallocated. There is no system or central authority capable of identifying institutional priorities, talents, and opportunities and then matching the best person to the job. Rather, the existing system often requires that prosecutors choose between a legal theme they favor (such as environmental protection, public education, or citizens' rights) and a place in which they want to live; they rarely have both.

3.7. Support structures and staff

Prosecutors' offices are often located within or next to the courts, and each prosecutor is entitled a team of legal interns, administrative assistants, and judicial clerks to help them analyze judicial files, write briefs and motions, and manage the workload. Even the layout and furniture of the buildings, with small individual offices and carts equipped to transport legal files, favor the fast disposition of cases. Conversely, a prosecutor who wants to leave his or her office, conduct field inspections, acquire technical (non-legal) knowledge about violations, and conduct negotiations leading to compliance has no discretionary budget, equipment, or institutional memory and limited internal sources of technical support.

In brief, the task environment and the organizational structure, policies, incentives, and constraints put forth by the MP promote internal harmony and encourage prosecutors to work on a case-by-case basis, process files in the order in which they arrive, and interact closely with the courts but few or no other institutions (cf. Silbey 1980–1981). Not surprisingly, many pursue prudent conformity and become office-bound (Silva 2001). Moreover, all prosecutors readily admit that their primary duty is to baixar a pilha (lower the pile of cases) that slowly accumulates on their desks, and those prosecutors who are particularly good at it are called “tractors” or “cow catchers,”7, meaning that nothing can stop them from moving forward on their designated tracks.

Nevertheless, and as illustrated by the cases of Jessica Riveiro and Jairo Azevedo, some prosecutors deviate sharply from this norm. If the work environment, prevailing human resources' policies, and official support structures do not explain this deviance, what does? The next section describes the features that encourage and support prosecutors to engage in social problem-solving and relational regulation.

4. Relational practices that sustain creative enforcement

Since Zald and Berger (1978) called attention to coups d'état, bureaucratic insurgencies, and mass movements within organizations, deviance and “covert political conflict” (Morrill et al. 2003) have become a well-noted feature of organizational life. Drawing from this literature, and building on the typology proposed by Silva (2001), I claim that the Brazilian MP is composed of two factions of prosecutors that interact with each other on a daily basis. On one side, one finds the office-bound prosecutors whose behavior is encouraged by most of the formal and visible features of the organization. Next to them, one finds prosecutors whose deviant mode of action is captured by the expression to stitch solutions together. Their mode of action requires a different set of organizational structures, policies, incentives, and support than the one the formal MP readily provides. To encourage the stitching of solutions, these prosecutors have built a parallel, overlaid, and covert organization that, even if sometimes embryonic and difficult to discern, exists within and extends beyond the MP itself. This section discusses three central features of this organization.

4.1. Creating slack

Prosecutors who want to deviate from case-processing must find ways to avoid being overwhelmed by their caseloads. According to my interviews, and contrary to the popular perception that young prosecutors are more outgoing and risk-prone than their senior colleagues, a prosecutor needs significant experience to know how to create and maintain slack. At a lecture to newly hired colleagues, an appeals-level prosecutor from São Paulo confessed, “when I started, I was overly eager and prosecuted everything, even the theft of a duck from somebody's backyard.” Eventually, those prosecutors intent on stitching solutions together must learn to process cases fast, use templates and canned answers, avoid extraneous commitments, delegate some functions to others, and minimize personal involvement with trivial cases and dismiss others outright. Basically, and somewhat paradoxically, they must learn to over-routinize some aspects of their job to free up time for other, more consequential practices.8

Still, slack is not sufficient, as it can be used in different ways, including leisure. Prosecutors will engage in relational regulation only if they are recruited, trained, supported, and rewarded for this kind of action.

4.2. Recruitment, training, and support

Prosecutors devoted to stitching solutions together reach out to junior colleagues the day they enter the organization. A senior prosecutor from São Paulo who is now retired explained how this recruitment works:

Half the prosecutors do not really care [about solving social problems], but the other half are very engaged. . . . This consciousness is not something they bring to the job; rather, they acquire it after they start. . . . When a new prosecutor arrives, she may be recruited by those with a social vision and acquire one herself. But there are others who are concerned about punishing crime and that is what they do for the rest of their career. This recruiting happen in various ways, in some cases it is through study groups, in other cases it is more ad hoc: a new prosecutor needs help with a case, consults with a senior colleague, establishes this connection, and it goes from there.

Experienced prosecutors who favor this alternative practice often mentor and guide their younger colleagues. A state prosecutor from São Paulo, with more than 15 years of experience, described his personal development and evolving commitment:

It was not in law school that I learned how to negotiate these deals, not in the orientation course for new prosecutors, or in the early years of my career as an alternate prosecutor either. My past experiences did not help. I learned it on my current posting, because I needed it. At that time, I called the environmental support office (CAO) and asked for help, and I talked to the head-prosecutor and asked that he assign my predecessor, Cardoso, who had moved on, to come back every so often to coach and tutor me. Both Cardoso and I were admitted to the MP in the same class, but he always worked with collective affairs so he had this experience that I lacked. . . . He spent almost two years tutoring me, and it was great . . . he would come to important meetings, and would help me with certain files, he is the one who taught me [how to conduct these negotiations].

Prosecutors who identify with this mode of action also strive to create formal and permanent channels, structures, and resources within the MP to help them recruit, train, and support their colleagues. The most visible of these structures are the internal offices that provide assistance, both legal and technical, to prosecutors responsible for enforcing collective rights.9 These offices are staffed with senior prosecutors and other technical experts, all of them available to answer questions and help field-based prosecutors handle their respective caseloads. While not all prosecutors assigned to these offices support social problem solving and relational regulation, the existence of these posts both enables and strengthens this mode of action.

Along similar lines, prosecutors assigned to other leadership posts within the organization often use their perch to organize congresses, training courses, and workshops that further reinforce their favorite mode of action. One such prosecutor said:

Environmental prosecutors form a tight-knit community, and we interact with each other. The main forum of exchange is our annual meeting. Not all environmental prosecutors favor settlements and negotiated solutions, but all those who attend this meeting do. We even joke that the same people come back every year.

Other prosecutors, unable to occupy leadership posts within the organization, find a platform elsewhere. For instance, in São Paulo, a group of prosecutors took over the research department of the prosecutors' professional association (Associação Paulista do Ministério Público[APMP]) and used their position to publish several books and articles defending their mode of action (see, for instance, Fabbrini 2002 and Ferraz 2003).

One of the most significant structural resources that helps prosecutors recruit, train, and support those interested in relational regulation is the Ministério Público Democrático (MPD), a self-funded membership-based organization open solely to prosecutors. Modeled on the Italian Magistratura Democratica, the MPD was founded in 1991 and has more than 300 active members spread throughout Brazil. A state prosecutor from São Paulo, with more than 20 years of experience and assigned to the homicide desk, explains:

Not all prosecutors are engaged or passionate about politics, but many are, and some of us have even been members of the partidão[colloquial name for the Communist Party]. We all meet through the Ministério Público Democrático (MPD). The MPD was created to oppose a particular group that had been leading the MP for many years. We succeeded right away and placed our candidate as head prosecutor, but then the MPD lost its focus and some of its purpose. Eventually it found a new mission, to support those prosecutors who are engaged in social causes; it houses the leftist MP. Many prosecutors don't really understand and do not like the MPD. They say “why is there a ‘Ministério Público Democrático' when the MP is all democratic?” But the MPD fulfills this function, of bringing together people who think alike.

4.3. The role of civil society

These structures notwithstanding, the MP does not provide prosecutors with all the staff, budget, or technical expertise they need to engage in relational regulation. To fill this gap, and as noted by Cavalcanti (1998), Hochstetler and Keck (2007), and McAllister (2008), some prosecutors establish alliances with NGOs, social movements, labor unions, neighborhood associations, church groups, and like-minded factions of professionals within relevant government agencies.

In some instances, these alliances are formal and visible. For example, in the countryside of São Paulo, prosecutors have partnered with city-level health surveillance agencies to make sure that migrant sugarcane workers are lodged in appropriate dorms with functioning bathrooms and kitchens. A city councilor integral to this alliance explains:

The health surveillance agency has almost no enforcement power. Conversely, the prosecutor has a lot of power and can inflict the heaviest fines. . . . She is very stern, but she can not be everywhere and know everything. So they struck an alliance. Nowadays, there is a lot of exchange going on, and the MP has been learning a lot from the Health Surveillance Agency.

In other cases, the alliance is informal and involves factions of sympathetic officials within an otherwise hostile public bureaucracy. A state prosecutor from São Paulo, with more than 15 years of experience, provides an illustration:

You know, there is a lot of corruption in the granting of environmental licenses, and a lot of political pressure. But I have an excellent relationship with bureaucrats from the Forest Service, and the Environmental Agency, and they are my sources. Sometimes they bring me copies of their impact assessments and tell me “this is what I wrote. Save it, because this original may change as it makes the rounds.” Others call me to tell me they are being pressured [to reverse a technical decision].

An officer with the state-level environmental agency tells essentially the same story, but from the point of view of the beleaguered bureaucrat:

Whenever the pressure on us grew, we called the MP . . . we know the prosecutors. We have been giving courses to them, and every so often we organize field trips for them to visit the parks. Técnicos like myself have done it on our own. Do you think the head of the environmental agency cares for this? They hate us. In many cases we were fighting against the government itself, the governor wanted to please some mayor, and we were pressured to grant a license. To resist, we leaned on the prosecutors.

Crucially, many of these alliances are much more than short-lived quick fixes. Rather, these prosecutors and their allies often take purposeful action to strengthen their joint organizational capacity, institutionalize their achievements, and make their own jobs easier in the future. In practice, this means that a prosecutor may agree to settle a case in exchange for a new building, cars, computers, or specialized equipment that will strengthen the prosecutors' allies.10 A state prosecutor from São Paulo, with more than 10 years of experience and assigned to the collective affairs desk, describes how she helped empower the local environmental protection agency, and how this agency now helps the prosecutor:

On one occasion, there was this massive toxic spill from a chemical industry in my jurisdiction. I started an inquiry, and the managers of this industry came to me with an assessment of the damage that went only to the boundary of their land. Very funny! No way, I told them, you will assess the damage on the whole region, as indicated by the technical expert the environmental agency has assigned to the case. Fortunately, the environmental agency is very competent and well equipped; it has all the equipment and resources, because in all my settlements I get the firms to pay for equipment, training, to refurbish their buildings, and thus I help strengthen and empower the regulatory and monitoring agencies.

Finally, it is important to clarify that formal assignment does not necessarily correlate with type of engagement. For instance, during my fieldwork, I encountered experienced prosecutors assigned to collective affairs who favored case-processing and prosecutors assigned to typical case-processing jobs (such as the assessment of injunctions known as mandados de segurança) who engaged in relational regulation. Moreover, and contrary to the myth that Brazilian prosecutorial assignments are clear and determined in advance, skilled prosecutors can often redesign their jurisdictions to acquire the mandate they need to stitch solutions together. Taken together with the evidence presented in this section, these patterns suggest that engagement in relational regulation is not caused by individual features, formal career incentives, or situational conditions. Rather, the existence of this mode of action is better explained by the presence of a range of formal and informal structures, processes, and incentives built for this purpose and that exist within and expand beyond the MP itself.

5. The interaction between the two factions

Prosecutors who favor case-processing and their colleagues who favor problem-solving have responded to the presence and actions of the other by developing a series of interactional patterns and behaviors that strengthen the performance of the larger organization. I identify four patterns of relations, which I name competition, collaboration, interpenetration, and reconciliation.

5.1. Competition

As anticipated by March (1991, p. 71), both modes of action “compete for scarce resources.” In the case of the Brazilian MP, this competition is not well balanced. As suggested by Lipsky (1980) and Silbey (1980–1981), the task environment and the characteristics of the job favor case-processing. As a result, office-bound prosecutors represent the status quo and do not feel they have to defend or explain themselves. For this exact reason, their power should not be underestimated. A state prosecutor from São Paulo, with more than 30 years of experience, describes how the rank and file reacted to a proposal suggesting the systematic creation of slack:

Our goal was for the MP to step back from acting in too many fronts and restrict itself to acting on what counts, but resistance was tremendous. A colleague defended selectivity in a speech at the annual prosecutors' congress and was almost pelted with stones.

Conversely, prosecutors partial to relational regulation are not demure. They often call themselves promotor de fato or bom promotor (real or good prosecutors) and disparage their colleagues as office-bound or funcionário público (public servant). In one revealing incident, a prosecutor referred to a case-processing colleague as horácio, a local cartoon character that depicts a Tyrannosaurus rex. When I asked for clarification, she explained, “they have short arms and cannot reach very far.”

Interestingly, this competition has disciplinary effects and ensures that prosecutors who engage in relational regulation are careful with regard to ethics, legality, transparency, and political neutrality.11 A federal prosecutor with more than 20 years of experience and assigned to São Paulo state explains this disciplining effect:

Prosecutors like us, if we did not have job tenure and other professional prerogatives, we would be out. The political pressure is just too intense . . . . There are people within the MP who dislike us, are jealous of us. So we must be careful not to make any mistake, because they keep their eyes on us.

5.2. Collaboration

As examined in section 4, those prosecutors intent on engaging in relational regulation must over-routinize some parts of their practices to create slack. One way to accomplish this goal is to share the case-processing burden with office-bound colleagues. This is exactly what the prosecutors in an eight-prosecutor office devoted to jury trial in São Paulo did. A few years ago, and thanks to a two-week strike in the Judiciary branch, three prosecutors found the time to leave their office and connect with community leaders interested in preventing crime in their neighborhood. To allow for this continued exploration, the eight prosecutors decided to split responsibilities: while five would stay in the office to ensure that all incoming cases were processed on time, the other three would engage with the community to try to prevent crime.12 In an interesting twist, crime rates dropped so precipitously that the head prosecutor, attuned to the decline in the case-processing workload but oblivious to the reasons behind it, tried to transfer some of the prosecutors to another office. To preserve their arrangement and to expand their way of acting to other communities, these prosecutors resorted to their allies in civil society, who protested – successfully – against the proposed move. The overall arrangement was so successful that the three prosecutors who engaged in relational regulation dedicated their jointly authored masters' thesis on the topic to their case-processing colleagues.13

5.3. Interpenetration

As indicated earlier in the article, it would be inaccurate to claim that all formal and visible features of the MP favor case processing and that all informal and covert features favor relational regulation. Rather, the structure of the MP reflects the division in its ranks. Prosecutors associated with both factions continuously fight, often in opaque legal jargon, over the proper form and function of the organization. These disputes are not a winner-takes-all situation. Rather, they frequently result in small but concrete changes in the way prosecutors define their jobs and the MP formally organizes itself. The existence of technical support offices covering a variety of topics, official congresses, training sessions, and workshops, all funded by the MP, illustrate this point. The organizational structures created by this interpenetration may strike some as illogical or idiosyncratic. However, they let the larger organization be different things to different people and attend to conflicting demands for procedural, substantive, and preventative justice.

5.4. Reconciliation

Finally, prosecutors associated with either faction often adjust their language to conform to generalized expectations and preserve unity within the MP. This means that many prosecutors devoted to social problem-solving and relational regulation publicly claim to be just enforcing the law, even as they are stitching solutions together. Likewise, many who favor case-processing claim to be solving important social problems even if they rarely leave their offices. In the end, prosecutors associated with both factions share a language filled with references to concepts such as enforcement, rights, and constitutional principles that are generic and ambiguous enough to ensure internal unity and a unified external front. As pointed out by Crawford (2008), Brazilian prosecutors like to tell a cogent and unified story about themselves, and this unity helps them protect the MP from frequent attacks on their legal powers and professional discretion.

6. Conclusion

Even though the Brazilian Constitution defines the MP as a unitary and indivisible organization, I found that beneath a veneer of cohesion and homogeneity coexist two different organizations with different sets of ambitions and practices. A sizeable faction of prosecutors embraces a procedural vision of justice and adopts a cautious, routinized, ritualistic, and reactive case-processing approach. Representing the status quo, they work primarily in their offices and manage the case flow. By and large, they confirm Lipsky's (1980) generalization about the overwhelmed street-level bureaucrat forced to cope. Next to them one finds prosecutors who resist the pressures of conformity and strive to use their professional status, public legitimacy, and legal powers to identify important societal problems and solve them. They make every effort to create and institutionalize the kind of robust organization that Sabel (2006) describes.

The existence of these distinct modes of action within the same institution suggests that responsiveness should be divided into two types. One type is responsive regulation, which Ayres and Braithwaite (1992) conceive as preprogrammed escalation. According to them, regulators are responsive when they respond to the actions and reactions of the regulated firms by moving up and down a fixed sequence of deterrence and cooperative stimuli. This formulation assumes that responsiveness can be routinized (Heimer 1998), that all engagements are dyadic (i.e. involve only one regulator and one entity being regulated at a time), and that regulators have a limited repertoire of enforcement tools. According to this view, the challenge of enforcement is one of categorization; that is, enforcement agents ought to attach clear legal labels to multifaceted real-life occurrences.

A second type of responsiveness is here called relational regulation. Exemplified by Ms. Riveiro's and Mr. Azevedo's actions, this type of responsiveness is contextual, deliberative, and open-ended with regard to both means and ends. Enforcement may involve a multitude of participants, and the repertoire of enforcement tools is limited only by the regulator's imagination and ability to recruit allies. The challenge of enforcement revolves around search; that is, regulatory enforcement agents ought to acquire context-specific information about the problem at hand and then search for and recruit allies willing to assume some of the costs, benefits, and risks associated with the proposed change until compliance is acceptable to all.

These two modes of action correspond to March's (1991) exploitation and exploration, and they compete for scarce resources within the organization. While the case-processing approach is thoroughly supported by the task environment and a number of formal features of the MP, relational regulation is supported by a looser set of devices, structures, and processes that its proponents have built and maintain. Whenever possible, these prosecutors reshape the formal MP in a way that supports their vision. But they often depend on the support of external allies, such as factions of sympathetic bureaucrats in other government agencies or NGOs and community groups, to engage in their favored mode of action. Interestingly, and in a self-reinforcing cycle, these partners often use the resources they obtain through their joint action to strengthen each other. In a broader canvas, this dynamic challenges the belief that civil law systems are inherently hostile to the systematic use of prosecutorial discretion. Rather, it shows that discretion can create the conditions for its own survival, and that proper organizational features can channel it toward effective enforcement and consensual problem solving.

It would have been ideal if this article had also measured the size of each of these factions and examined whether the MP has found the perfect balance between them. Unfortunately, the data are not available. As mentioned earlier, all prosecutors engage in case-processing some of the time. What differentiates the groups is that some prosecutors over-routinize their case processing so they can engage in relational regulation on the side. For this reason, any estimate of the relative size of each faction would have to code not individual prosecutors, but their countless (and often ambiguous) enforcement acts. And then, not all relational regulation initiatives succeed at the same rate, and some may not succeed at all.

Rather, this article characterizes each of these factions and identifies some of the organizational features that support and encourage their different ways of acting. Also, it contends that the simultaneous harboring of both case-processing and relational regulation – an ambidexterity of sorts – fosters a level of internal competition, collaboration, interpenetration, and reconciliation that strengthens the organization as a whole. Even more, this duality preempts public criticism that prosecutors pick and choose their caseloads and allows for the creation of positive feedback loops, in which the resolution (or prevention) of certain infractions frees up resources for prosecutors and their allies to tackle other pressing concerns. The case of the jury trial division described in section 5 is particularly revealing. The five prosecutors within that office who favored case processing ensured that all cases moved speedily through the system (as required by law). And the three prosecutors who favored relational regulation devoted themselves to crime prevention and contributed toward decreasing the caseload for the office as a whole.

At the end of the day, these facts, when examined under the light of the literature on organizational performance and the merits of ambidexterity, suggest that the coexistence of case processing and relational regulation creates an MP that is more reliable, responsive, and robust than a homogeneous entity could ever be.


I would like to thank Judith Tendler, Susan Silbey, Richard Locke, Michael Piore, Suzanne Berger, Roberto Pires, Mansueto Almeida, Matthew Amengual, Michael Lipsky, the editors of Regulation & Governance, and four anonymous reviewers for their always insightful comments and prompt feedback. This research was funded by the Massachusetts Institute of Technology and was part of the project “The Rule of Law, Economic Development, and Modernization of the State in Brazil”, coordinated by Judith Tendler and funded by the Department for International Development (DFID-UK) and the World Bank.


  • 1

    I use the generic term “prosecutor” to refer to all those officials who, in Brazilian Portuguese, and depending on actual attributions and institutional affiliation, start their careers as promotor de justiça or procurador de justiça. All names are pseudonyms, some other identifying details (such as gender) may have been altered, and all translations from Portuguese are mine.

  • 2

    This research was composed of two tracks: in one track, I traveled throughout the country to interview institutional actors involved in cases of compliance and non-compliance with labor and environmental laws. Through this method, I noticed the existence of social problem solving within the organization. To understand the organizational basis of this practice, I spent several months in São Paulo conducting participant observation and interviewing prosecutors at different stages in their careers.

  • 3

    This vignette has already been discussed in another publication (Silbey et al. 2009).

  • 4

    In 2009, the municipal government of Rio de Janeiro suggested a similar intervention in one of its favelas, and the general public reaction was overwhelmingly negative (Chade 2009).

  • 5

    A 2004 constitutional amendment requires that applicants possess three years of legal work experience. This provision has been put into practice only recently. The majority of young prosecutors I interviewed either possessed no prior experience or worked as private lawyers part-time while studying for the MP's entrance exam.

  • 6

    There are recognizably more intraorganizational controversies in those states where the procuracy awards promotions and transfers on merit.

  • 7

    A cow catcher is a device attached to the front of a train to clear obstacles off the track.

  • 8

    This duality – the fact that the same individuals alternate between a fact-oriented and office-bound orientation – makes it particularly difficult to measure the relative size of each faction.

  • 9

    In São Paulo, these offices are named Centros de Apoio Operacional (CAO) (Operational Support Centers).

  • 10

    This is not exclusive to prosecutors in Brazil. In the US, the Environmental Protection Agency (EPA) encourages businesses to adopt Supplemental Environmental Projects (SEPs), which often involve the private provision of public goods (Environmental Protection Agency 2001).

  • 11

    One could expect that case-processing prosecutors (and even some citizens) would resent the risks taken by social problem-solving colleagues, particularly if results are not immediately forthcoming. In practice, prosecutors who engage in this deviant practice admit that, because of formal organizational controls, all prosecutors, no matter their orientation, must process their respective caseloads before engaging in any other activities. By carrying their weight, they preempt this type of conflict and ensure internal and external harmony.

  • 12

    This experience is analogous to Boston's Ten Point Coalition, as described by Winship and Berrien (1999).

  • 13

    Document on file with the author.