1
The Wrong Label
The Social Construction of Ladraois and the
Implications for the Use of Violence by the Brazilian
Police
Name: Nikkie Menting Bachelor thesis Interdisciplinary Social Sciences Student number: 10362002 Date: 20 July 2016 Supervisor: Martijn Dekker Words: 17.857
2
Table of contents
Abstract ... 4 Acknowledgments ... 6 Introduction ... 7 Theoretical Framework ... 9 Street-level bureaucrats: The mediators in the system ... 9 Policemen: their place, their roles and their discretionary space ... 13 The Police force in Brazil ... 17 Problem definition ... 22 Introduction: ... 22 Main Question: ... 23 Subquestions: ... 23 Methods ... 24 Research Strategy ... 24 Respondent groups ... 25 Gaining access ... 26 Research Methods ... 27 Issues encountered ... 28 Data analysis ... 30 Ethical Justification ... 30 Results ... 31 Part I: The police ... 31 Identification with the organization ... 32 Actions outside their mandate ... 33 The normalization of arms ... 35 Part II: The ‘good citizen’ versus the ladrão ... 37 Ladroais ... 37 Good citizens ... 38 Part III: The game of ‘Routine Killable Subjects’ ... 39 Resisting the label ... 40 Violating the law but not the norm ... 41 The wrong label ... 43 The labelled killed ... 45 Conclusion ... 48 Discussion ... 50 Literature list ... 513 Appendices ... 55 Appendix I: Research proposal for the military police in English ... 55 Appendix II: Research proposal for the military police in Portuguese ... 59 Appendix III: Support letter ... 63
4
Abstract
This research discusses the use of discretionary space of police officers in the state of Bahia, Brazil trough the means of storytelling. Firstly, participant observation was conducted to understand the daily life of a police officer in Bahia. Secondly, there were stories gathered about the experiences of the police from young men that are vulnerable to misconduct by the police. The results shows that discretionary space is not so much shaped by the managers that develop policy for public security, but by an implicit social system. The police in Bahia work in a violent environment, and they have to estimate the danger of citizens in a split-second. On the basis of appearance and behaviour, the police as well as the community frame citizens as a ‘good citizen’ or a ladrão. The good citizens are safe for the police and cooperate with the police, but the citizens labelled as ladrão run the risk of being marginalized and therefore become the victim of misconduct of the police.5 “I don’t feel that it is the necessity to know exactly what I am.” Michel Foucault
6
Acknowledgments
The text in front of you is a product of 6 months of blood, sweat and tears. After over a decade of wishing and waiting, I finally had the courage to go to Brazil, this weird country, which is so beautiful and ugly at the same time. This thesis would not have been here without the help of dozens of people. First and foremost, I have to thank Mestra Omara. Thank you for taking me into your family, helping me gaining access and kicking my ass when I needed it. Thank you Joelma, Joseane (Gel) and Jarilma for transcribing my interviews. Ilona, thank you for taking the beautiful picture that is at the title page. Thanks to all the visitors of the house of Omara for explaining everything about Brazil and about the police. I would like to thank my parents for always having faith in me that I actually could finish this thesis. Thank you Merel Bredeveld, for reading my thesis over and over again to correct my sometimes-hilarious mistakes. Thank you Lisette for teaching me that I should follow my dream, no matter what happens. Thank you Sylvia for helping me find back the common thread in my story, and many thanks to Martine for making all the tea it took to finish this project. Thanks to all the friends and family for their empathy and moral support during this project. Lastly, I have to thank all respondents and my contacts in Brazil who allowed me to do the fieldwork. Special thanks go out to the tourist guide who introduced me to the youngsters who talked about their relation with the police. And to the youngsters I would like to say: “Thank you for opening up to me”. Thanks to all the police officers who talked to me, explained their daily work, gave me access to the scene and guaranteed my safety. You were all awesome, and I could not have written this story without you.7
Introduction
“Brazil is more than just exotic women and beaches”1, said one of the police officers at the head office of the military police in Bahia, when I visited them for the first time to inform them about the possibilities for participant observation. On the other hand, the image of exoticness, the nice beaches and the big parties are part of a strong marketing strategy that has to convince the world that Brazil is a nice place to visit (Reid, 2014). With international events like the Football World Championship of 2014 and the Olympic Games in 2016, Brazil claims to be a new important player in the international arena. However, it has a lot of internal struggles to overcome before being the world power it claims to be. There are problems with social security, the amount of high-skilled citizens and problems with public security (Reid, 2014; Roett, 2010). Public security becomes increasingly more important with the organization of big events like carnival and the Olympic Games, but also in terms of regular tourism. The high levels of armed robbery, assault in public transport and pickpocketing are very problematic for tourism (McCoy, 2014). Brazil decided to spend high levels of public funding to preventing crime policies around the early 2000’s (Cano, 2006, p. 134). Although public security has been a top priority for years, the situation is still problematic. Before the World Cup in 2014 Brazil was criticized heavily in the media for their crime fighting strategy. The favelas2 were swept with highly militarized police squats to create a ‘false sense of security’, according to the locals (Ormiston, 2014). Violence conducted by the police is normal. In 2006, the Brazilian police killed more people only in Rio de Janeiro than in the entire United States (Pereira in Reid, 2010). This image is also confirmed in the Brazilian films that people inside and outside of Brazil watch. Films as The City of God (Ribeiro, 2002), Elite Squad (Padilha, 2007) and Trash (Bevan, 2014) show the dark side of the partying exotic country, constructing an image of the police as violent, corrupt and dangerous. In Latin-American-studies, the police of Brazil are also a popular subject. The police are often studied in their relation to human rights, state legitimacy and racism. All in all, the police are criticized by their own inhabitants, by people abroad and by academics. 1 Police officer, Personal communication, April 2016. 2 Brazilian Slums.8 This research explores what is going on between the image of the parties and the beautiful beaches and the image of bad policemen suppressing the inhabitants of the favelas. The police need some level of consent of the population to be able to do their jobs, and besides sweeping favelas they are also the people that you ask for directions or that you ask for help when you are the victim of a crime. Also, the population is not a helpless receiver of help or misconduct of the police, but they have some level of agency in every meeting with them. On the other hand, a policeman has some freedom to execute his work as well. This research is not focussing on the side of the functioning of the police, nor the side of the victims. It is about the interactions and the ideas behind these interactions that form the relation between the police and the community they serve or should serve. With participant observation with the military police in Bahia, I gathered stories from several policemen about their interaction with the community and their explanation of their task. Talking to youngsters who are affected by the police, I gathered stories about how to deal with the police if you risk being mistreated by them. With these stories, I tried to shed some light on the grey area, that is between the black-and-white image of the exotic beaches versus the corrupt police.
9
Theoretical Framework
This theoretical framework will explore the existing theory of the Brazilian police officer as a street-level bureaucrat. The first chapter will cover the concept of street-level bureaucrats in general. The second chapter will consider the special features of a police officer as a street-level bureaucrat, especially discussing it’s discretionary space. The third part of this theoretical framework will inform about the police in Brazil and the current problems they face, as well as their problematic work environment. Street-level bureaucrats: The mediators in the system This chapter is an overview of the broad term ‘street-level bureaucrat’. The vital role of a street-level bureaucrat in an urban democracy will be explored. Street-level bureaucrats are implementers of policy and therefore have an important role when policy is designed. On the other hand, they have to deal with the outcome of policy plans in practice. This unique mediating role that street-level workers have in the system, will be explored in the next paragraphs. Lipsky defines the police officer as a ‘street-level bureaucrat’. He considers street-level bureaucrats “Public service workers who interact directly with citizens in the course of their jobs, and who have substantial discretion in the execution of their work” ([1980] 2010, p. 3). Street-level bureaucrats maintain a key role in policy implementation, as they are the broker between the upper level of the policy makers and the civilians that are the subject of policy (Lipsky, [1980], 2010, p. 4). As they are the last step in the chain of implementation, they directly interact with civilians. Civilians perceive the effects of policy in their daily lives, not by the policymakers that work behind their desk, but by the street-level bureaucrats that they encounter in daily life. Civilians do not experience new policy programmes in an abstract way like policy is designed. “As they experience oppression and deprivation, in a concrete setting, and they will not see it as an end product of large and abstract processes “(Piven and Cloward in Lipsky, [1980], 2010, p. 10). Therefore people will protest against institutions instead of against policy. Police violence in Brazil is for example enhanced by the far-reaching mandate on the use of violence by the police (Husain, 2007, p. 69). However, when policemen presumably killed civilians in Bahia,10 their families started to protest against the police and not against the designed policy (Associação de Familiares e Amigos de Presos e Presas do Estado da Bahia, 2016). The prosecution of citizens after they committed a crime and other duties of the state are carried out by the state through street-level bureaucrats (Baynes, 2010, p. 15; Lipsky, [1980] 2010, p. 4). If policy would be the brain of the state, then street-level bureaucrats would be its arms. Street-level bureaucrats are the practical form of policy, and therefore vital for the relation between state and citizen. As Lipsky states: “Thus, in a sense street-level bureaucrats implicitly mediate aspects of the constitutional relationship of citizens to the state. In short, they hold the keys to a dimension of citizenship” ([1980] 2010, p. 4). On one side, street-level bureaucrats have more information and expertise about the welfare system that provides certain benefits for citizens (Lipsky, [1980] 2010, p. 90). In other words, they are experts in a certain policy area. Policemen know how they have to execute the security-policy in their community. When new policy is designed to for example improve the security or to secure a higher level of human rights, the street-level bureaucrats will be informed or retrained (Husain, 2007, p. 127). However, they also operate directly in society, so they use their experience that allows them to predict how to implement policy in the best possible way (Lipsky, [1980] 2010, p. 4). They are in fact, the mediators of the governmental policy system: “These street-level bureaucrats, mediate between the rules and regulations set by bureaucracies and the fuzzy everyday experience at the street-level” (Verloo, 2015, p. 34). Because they are the mediators in the system, they are the last part in the chain of policymaking that decides if policy will fail or succeed. Street-level workers also have to be considered as policy-makers. Lipsky describes this the following way: [P]ublic policy is not best understood as made in legislatures or top-floor suites of high-ranking administrators, because in important ways it is actually made in the crowded offices and encounters of street-level workers.” ([1980] 2010, xii). In this quote of Lipsky, the ‘daily encounters of street-level workers’ is really important, because the street-level workers are the so-called ‘arms of the state’. They are the visible part of designed policy for citizens. By daily encounters of street-level bureaucrats, people make sense of the social reality that is shaped on a more abstract level by policy
11 makers. In situations of conflict, street-level bureaucrats can be important actors in the chosen dominant frame (Verloo, 2015). To function in their always-changing environment, street-level bureaucrats use knowledge based on experience. This knowledge is called ‘tacit knowledge’; knowledge that can only be learned from other people in a social context, based on frequent experiences (Goffin and Coners, 2011, p. 301-302). To allow street-level bureaucrats to use this ‘tacit knowledge’, they have a wide range of ‘discretionary space’. This is the freedom they have to make decisions ‘on the spot’ and to develop strategies using their ‘tacit knowledge’. To implement policy, government officials give street-level bureaucrats a set of rules and laws. However, these laws and rules have a certain margin in which the street-bureaucrats can decide how to act in a specific situation. Lipsky describes this margin as the following: “Street-level bureaucrats have discretion because the nature of service provision calls for human judgment that cannot be programmed (…). Street-level bureaucrats have responsibility for making unique and fully appropriate responses to individual clients and their situations” (Lipsky, [1980] 2010, p. 161, italics added). All in all, street-level bureaucrats have discretionary space because a top-down structure of policy does not work for policy implementation. According to Lipsky, street-level bureaucrats generally have the intention to improve society (Lipsky, [1980] 2010, p. xiv). Because the nature of their work does not allow direct supervision, there is a certain level of confidence in street-level workers to accomplish the task at the best of their knowledge. However, discretionary space is also very often exploited for personal gain or the gain of state institutions, as proven by the uncountable examples of police scandals in Brazil (Willis, 2015; Globo Bahia, 2016; Correio, 2010). When street-level bureaucrats make decisions at the cost of civilians, the civilians will try to resist the policy carried out by the workers. In this case discretionary space could prevent conflict, or be the cause of more conflict: “At best, street-level bureaucrats invent modes of mass processing that more or less permit them to deal with the public fairly, appropriately, and thoughtfully. At worst, they give in to favouritism, stereotyping, convenience, and routinizing – all of which serve their own or agency purposes” (Lipsky, [1980] 2010, p. xiv). If street-level bureaucrats indeed have the good intentions described by Lipsky,
this is still no guarantee of the successful implementation of policy. The work of street-12 level bureaucrats is characterized by a shortage of recourses in terms of time and in terms of money. They do not have the resources to inquire all their clients, and therefore they develop strategies of coping and processing large amounts of people. Because of the large amount of people, the people are dealt with as ‘clients’, which takes away a part of their humanity and the particularity of their personal case (Lipsky, [1980] 2010, p. 61-62). These coping-mechanisms are often on the basis of stereotypes and dominant framing, which can cause exclusion of groups from policy. “Their [Street-level bureaucrats’] work involves the built-in contradiction that, while expected to exercise discretion in response to individuals and individual cases, in practice they must process people in terms of routines, stereotypes, and other mechanisms that facilitate work tasks (…) They regard their adaptations to the job not only as mechanisms to cope with resource limitations, but also as functional requirements of doing the job in the first place” (Lipsky, [1980] 2010, p. 62). Street-level bureaucrats can for example decide whether or not they give their clients information, causing inequality. “They [clients] experience it [giving or withholding information] as confusing jargon, elaborate procedures, and arcane practices that act as barriers to understanding how to operate effectively within the system.” Therefore Lipsky states that street-level bureaucrats are also policymakers, making policy through daily interaction with civilians (Lipsky, [1980] 2010, p. 86). “As a consequence, in their day to day work, street- level bureaucrats have to work out practical versions of public policy that can often look quite unlike official pronouncements” (Evans et al., 2004, p. 876, italics added). These ‘practical versions of policy’ can be a cause of conflict, or the resolving mechanism of conflict. The daily actions street-level bureaucrats perform, can give salience to certain frames, where other frames become dominant within the working-culture of street-level bureaucrats. Therefore, the coping mechanisms as well as the social reality that is made by street-level bureaucrats are an important actor to be studied in the field of conflict studies. Verloo (2015, p. 312) concludes that street-level bureaucrats in the Netherlands are often not listening to the civilians, and do not give salience to the different frames used by those marginalized communities. She states that situations of conflict are a missed opportunity in policy-making, and that policy-makers
13 should use moments of conflict as learning moments. Lipsky raises a similar problem stating that the constructed social reality of street-level bureaucrats is often different than the constructed social reality of the civilians they work with. The most important difference is often that the clients see their individual case, where street-level bureaucrats see a case within a broader policy category (Lipsky, [1980] 2010, p. 60). It is therefore important to research how the reality of street-level bureaucrats differs from the civilians they work with, and how these frames are interacting. There has to be special attention for the way they view and they use discretionary space, as a powerful tool to control their subjects of work, namely civilians. Policemen: their place, their roles and their discretionary space “Policemen decide who to arrest and whose behaviour to overlook (….) Rules may actually be an impediment to supervision. They may be so voluminous and contradictory that they can only be enforced or invoked selectively (….) Police behaviour is so highly specified by statutes and regulations that policemen are expected to invoke the law selectively. They could not possibly make arrests for all the infractions they observe during their working day.” (Lipsky,[1980] 2010, p. 13-14). This quote indicates that the police are a special street-level bureaucrat. This chapter gives a general overview of the function of this street-level bureaucrat and the conditions that affect the successful interaction of the police and citizens. As most of the publications mentioned here were carried out with large datasets and were conducted in the United States, it must be noted that the results because of the different cultural backgrounds must be interpreted prudently. However, it gives a first indication to explore general problems of the police and their discretionary space. This chapter will start with a brief description of the special place of the police and their specific role in the state. Secondly I will give an overview of categories of police behaviour as designed by Sherman (1980) as well as the result of this behaviour. This is a first exploration of the way discretionary space is used by police officers. Thirdly, important factors of trust in the police will be discussed. Police are, being a street-level bureaucrat, the mediators between civilians and policy makers. For successful mediation, trust in the mediator is needed (Kydd, 2006). Therefore it can be
14 concluded that trust in the police is one of the key factors needed for the system to work. At last, the possible explanations for misconduct are explored, because bad experiences with the police are an important predictor for low levels of trust (Horowitz, 2007). Police officers have a specific role in the sovereignty of the state. Their function is to keep the state system working by arresting or fining those who are not behaving according to the laws designed by the state (Foucault, 1980, p. 178). Holding the monopoly of force3, they are one of the basic pillars in the construction and maintenance of the state (Schwartz, 2010, p. 14-32). They are, together with the military system, the only people legally allowed to use force on behalf of the state. They should be the upholders of the law, and are an important part of the welfare system (Cao & Zhao, 2005, p. 403). The police also function as a broker between civilians and policy makers. As an anthropologist specialized in Brazilian police puts it: “We may consider the police to be crime fighters but it seems we turn to them more often for our daily needs, such as asking for directions, than to report criminal offences” (Husain, 2007, p. 26). Sherman (1980) distinguished three often-used categories to explain police behaviour namely service, arrest or the use of force, and detection. Policemen have discretion in all of these categories. The first category ‘service’ contains generally all the positive things that policemen do to help civilians. This can differ from mediating small conflicts up to conducting security checks. In Brazil, the Military police is responsible for segurança publica, or public security (Willis, 2015). The second category ‘arrest’ and the use of force contains any form of physical force used by policemen in the broadest way, so this category also contains deadly violence (Sherman 1980). Arrest, as well as the use of violence, is also the main task of the military police. Thirdly ‘detection’ is merely questioning people, investigating and everything related to research (Sherman, 1980). This investigative branch of police work is the main occupation of the Civil Police (Willis, 2015). Halfway the 20th century, the police were just supposed to be the “primary legal enforcers [of the law]” (Hinton & Newburn, 2008, 214). When civilians broke the law, the police arrested them (Kelling, 1999). As
Lipsky already stated for street-level bureaucrats in general, the current thinking about decision-making by the police force in 3 In Zwartz is this monopoly of force carried out by the military, but this function is currently within the borders of the nation-state almost totally replaced by the police (O’Malley, 1997; Raines, 2000), the police are currently in Brazil the primary enforcer of the law (Hinton & Newburn, 2009, p. 2014).
15 the field has changed. In the late 1950’s, researchers started to recognize that policemen actually have a lot of discretion in their daily patrols on the street. Discretion is found on every part of the institutional ladder: “This discretion existed not only at executive levels but also at every level of the police organization, especially with the practicing police officer” (Kelling, 1999). However, police discretion is not an easy concept to grasp or to define: “If discretion is properly linked to complexity (…) it will become increasingly clear that discretion is (…) shaped by professional police and departmental knowledge and skill, not personal inclination.” Police discretion is in this thesis therefore defined as “the institutional space a policeman has to make professional decisions in a specific setting”. Police discretion is a much-studied subject (Mastrofsky, 2004; Herbert, 1998; Van Maanen, 1978; Sherman, 1980 and others). As Lipsky already stated, this discretion is needed because there is no ‘one size fits all’-model to the problems the police encounter on a daily basis. However, the downside of discretion is that policemen can misuse their authority. Misuse is defined in this thesis as: “the abuse of the formal as well as the informal authority a policeman has for their own gain”. When a policeman acts in good faith but the effect is negative on the community, it does not fall under the definition of misuse in this thesis. The causes of misuse of this discretion can be divided in two broad categories. According to Raines (2000), there are two factors that cause police officers to commit unethical or criminal acts: “First, the role the police officer plays in society shapes the job he or performs. The role an officer plays contributes to the second factor – the autonomy of the job” (Raines, 2000, p. 6). She states that more control will not automatically lead to less misconduct, but that the misconduct will just be harder to find. Also, there are several control mechanisms that do not take away the discretion from a policeman like control mechanisms of peers, supervisors and civilians they work with (Raines, 2000). She states that the best way to decrease abuse of discretionary power is ethics training. However, this ethics training has to correspond with the needs of the officers involved. Secondly, the training has to correspond with the their world view (Raines, 2000). Even equipped with ethical training, the practical implementations in daily life remain problematic (Raines, 2000). The anthropologist Van Maanen (1978, p. 309), focuses more on the judgement of policemen on a micro level. He states that the police divide citizens in 3 types of
16 categories, namely the ‘know nothings’, the ‘suspicious persons’ and the ‘assholes’. The know nothings are mostly coming to the police for help, and they respect their authority. The suspicious persons are mostly interrogated or stopped on the basis of their appearance. They are suspected to have committed a serious offence, and the police treat them as the official procedures prescribe. The police are carrying out their professional function according to the book in an objective manner, without any secondary interests or involving personal emotions (Van Maanen, 1978, p. 309). However, the relation between the assholes and the police is more conflictual. The assholes “are stigmatized by the police and threatened harshly on the basis of their failure to meet police expectations arising from the interaction situation itself” (Van Maanen, 1973, p. 309). Because these people do not behave as the patrolman would like to see, they are more vulnerable to ‘street justice’. Street justice is a physical attack to justify the personal insult or disrespect of the asshole in question (Van Maanen, 1978, p. 309). The asshole is seen as ‘the other’, and is often not granted the status of human being. The asshole must be stupid, wrong or irrational in their nature; the label asshole takes away the necessity to explain the behaviour in a rational way. The asshole is seen as a symbol of all those ‘others’ out there that remain unpunished within the framework of law enforcement. ‘Street justice’ is therefore also a symbolic ‘pay-back’ for all the people that frustrate the police, but could never be punished using the pathway of the ‘suspicious person’ (Van Maanen, 1979, p. 323). Now that we have explored some of the factors that influence the way discretionary space is filled in, we will explore some features that cause citizens to trust the police. Trust is determined by a lot of factors. Surprisingly, there are no indications that misconduct of the police directly erodes trust in the police (Weitzer, 2002). However, after incidents of misconduct, there is a dramatic increase of negative attitudes towards the police (Weitzer, 2002). Although many action groups as well as scholars state that the police in the United States as well as in Brazil are racist (Associação de Familiares e Amigos de Presos e Presas do Estado da Bahia, 2016; Guimarães, 1995, p.224; Willis, 2014), race does not seem a significant predictor for trust in the police. Encounters with the police, or second-hand stories about encounters of the police are better predictors for the amount of trust (Horowitz, 2007). Secondly, safer neighbourhoods have higher confidence in the police, because they give them credits for their safety. Places with a higher crime rate experience the
17 police as less trustworthy, because their demands are not met (Horowitz, 2007). As with street-level bureaucrats, daily interactions with the police are again important. As Lipsky stated for street-level bureaucrats in general, the literature on police officers also states that ‘fairness’ is an important concept. People trust the police on the basis of the ‘fairness’ of their actions: “People evaluate the police against standards of justice, with a particular focus on the fairness of the procedures by which they exercise their authority. Three aspects of procedures are found to be important: neutrality of decision making, respectful and polite interpersonal treatment, and providing opportunities for input into decisions” (Tyler, 2005 , p. 339). Tyler confirms about police trust what Lipsky states about street-level workers in general. The success of the street-level bureaucrat is decided in day-to-day interactions, where interpersonal treatment on a micro-level is important for the policy to succeed (Lipsky, [1980] 2010, p. 151-152). On the other side is again the neutrality of decision-making, which demands that the designed policy is carried out without privileging some people or a group of people (Lipsky, [1980] 2010, p. 161). The Police force in Brazil This chapter informs about the police force in Brazil in a broad way. The first paragraph discusses the history of the police in Brazil. After this, the decentralization of the security will be discussed. The third part will discuss some of the main problems within the police force. The third part will discuss the organisation of the Military Police and the Civil Police, as well as the division of their tasks. Because Brazil has been a dictatorship from the mid-sixties until the mid-eighties, the police have a different place in society than in Western countries. In 1964, the military took over the regime after which Brazil became a military dictatorship. This military dictatorship seemed to be with the consent of the population, because the state was very organized and carried out its state functions very well (Hagopain & Mainaring, 1987, p. 486). However, the police and the military executed their work on the basis of clientelism, because they were one of the most important means of the elite to maintain the inequality that they profited from. The massive amount of people living in extreme
18 poverty were excluded and coerced by the use of violence (Husain, 2006). There is no clear date when the governmental structure changed from totalitarian to democratic, because there was no sudden regime change. In 1984, the military allowed some democratization, but the remaining leading figures stayed in power. This resulted in a mixture of representation by the people with many leaders of the former military regime on important positions of power. This resulted in the former regime obstructing policies that would make the country more democratic (Hagopain & Mainaring, 1987, p. 486). This legacy still impacts the working culture of the police in Brazil, as the police are traditionally seen as a tool that fosters the regime (Leeds, 2007, p. 27). Koonings and Kruijt (2006, p. 27-28) even state that barely anything changed: “In the period of approximately 20 years since Brazil’s progress of re-democratization (…) the sector that has made least progress is criminal justice and, in particular, the police”. When the new constitution was designed for Brazil, the part that discussed the police force remained the same as before (Koonings and Kruijt, 2006, p. 28). The problems of excessive violence were mainly seen as problems that were inherent on the dictatorship. The problem of urban violence was seen as ‘just a policing problem’. When the country became a democracy, the left-wing politicians expected the problem to disappear naturally (Cano, 2006, p. 133). Brazil is a very decentralized nation because it is a Federal Republic (Reid, 2015; Bateman, 2012). Public Security is the primary responsibility of the separate states, (art 144 Brazilian Constitution) and it is therefore not possible to generalize any of the information that is drawn from case studies in other states. Each state makes policy for the different types of police4, and has it’s own office of justice (Art 144, Brazilian Constitution; Cano 2006). Municipalities play on a lower level mainly a role in prevention, but if necessary they also perform repression tasks. In exceptional cases, they also take the right to kill citizens (Correio, 2016). The police-policies therefore differ per state, and therefore caution is needed to generalize the police culture from state to state. Also the cultural as well as economical differences differ enormously from state to state (Reid, 2015). Having said this, their basic organization of the police is the same in every state, the public security problems are similar in the 3 biggest cities of the 4 For a specific explanation about the difference between the civil and the military police, see the separate paragraph about this subject on page 21.
19 country, São Paulo, Rio de Janeiro and Salvador.5 The next paragraphs will therefore discuss various case studies and the main problems that the researchers highlighted in their research. Many of the case studies problematize the excessive amount of lethal force that the police use. This is also explained in the form of ‘the killing consensus’ (Willis, 2015). Conducting a fieldwork research in the favela’s or comunidades6 among a gang and the police, he found that the killing of both parties is actually not random. There is an underlying consensus about who has the right to kill when and where. This results in a system with periods of high violence and periods where there is more peace. The regulations of killing are decided by ‘urban, spatial, and political conditions’ (Willis, 2015). The classical functions of the state, namely the upholding of their own laws and the insurance of the safety of the population, are taken over by the biggest criminal organization in the São Paulista periphery. Parallel on this logic of the killing consensus, there are problems with respecting human rights in general. Since the democracy was established in Brazil, the desire to lower crime rates was always higher than the desire to promote human rights. There is a widespread fear among the Brazilian middleclass to get murdered, so they call for heavy measures (Husain, 2007). There is a sharp difference about ‘us’ the middleclass, and ‘they’, the criminals. The criminals are bad people in nature, so they have to be killed. As Willis (2015, p. 30) describes this mentality of the police in São Paulo: “Homicide detectives navigate moral and situated notions of deservedness of killing by drawing on signals, queues, and social imaginaries (…) a differentiate social type, understood as inextricably and holistically criminal, and therefore the legitimate subject of deadly police action”. The greater emphasis the direct result, so preventing crime, causes a situation in which human rights are not respected. This phenomenon also occurred at the police force in the case presented in Rio de Janeiro (Husain, 2007). Except of the heavy focus on security, there are a lot more problems that cause violations of human rights. The military as well as the civil police are heavily underpaid, underequipped and inadequately trained (Ferranti et al, 2002; Willis, 2015). Because of 5 In the case studies of Husain (2007) in Rio de Janeiro and Willis (2015) in São Paulo, there are for example problems with upholding human rights, corruption and highly organized crime groups. 6 Favela is the traditional Portuguese word for the slums further away from the city center in Brazil, but this term is replaced by the Brazilian word for communities because of the negative connotation it had.
20 their bad pay, a lot of civil policemen work for private security companies when they are off-duty (Huggins, 2000; Husain, 2007; Willis, 2015). Off duty police officers are often the victims of ‘payback’ if the Brazilian police managed to catch important criminals or a great number of criminals of the organized crime groups. Because of this high risk, they often work 3 jobs to move to a middle class neighbourhood to be safe off-duty (Willis, 2015, p. 28). Another problem is racial profiling and discrimination among Military as well as Civil police officers. People are discriminated against for being young, poor, or black. The Military police stop racial minorities more often for random control (Ramos and Musumeci, 2004). More general, the police use racial profiling as a tactic dealing with crime. For instance, black people in Rio de Janeiro have a bigger chance to be body-searched than white people. All of the underprivileged groups have a bigger chance being threatened, coerced, or being the victim of psychological or physical violence (Ramos and Musumeci, 2004). This racial narrative is also common within the Civil Police. As Willis (2015, p. 30) describes the thoughts about killed citizens of the Homicide Detectives in São Paulo: “(…) the routine and unexceptional killable subject is: a man, often dark skinned, living and killed in the urban periphery, of a low socioeconomic background, understood to be associated with violence.” These forms of discrimination cause the general impression of distrust of the police. The media strengthens this image, and therefore all the minority groups have less confidence in the police and hold more often hostile attitudes towards them (Ramos and Musumeci, 2004). The whole topic of racial profiling is ignored within the military police of Rio de Janeiro. They claim that no one is really black or really white, but everyone in Brazil is a mixture. This claim is a legacy of the ‘fallacy of racial profiling’ that existed for a long time among scholars, calling Brazil a ‘racial democracy’ (Htun, 2004; Martins et al. 2004; Bailey, 2004). This concept states that there is no prejudice or discrimination towards black people, especially compared to the United States. Brazil has always been perceived as a country that successfully mixed all the races7, overcoming the differences between black and white. A much-heard argument is that Brazil has a very racially mixed population. In other words, there are not as many whites or blacks as in the United 7 The author is aware of the fact that race is a socially constructed concept based on existing psychical characteristics (Sauer, 1992). The perceivement of race is the discribed subject in this case.
21 States. The majority of the people is mulatto, a mix of the black and white or black, white and native Brazilian (Hasenbalg & Huntington, 1982; Pena et al. 2011). Bahia is one of the states with the biggest black population, because it was the economic as well as the political capital for a long time during the slave trade (Pena et al. 2011; Reid, 2015, p. 15) Because of the idea of a ‘racial democracy’ police officers, that are often quite black themselves, do talk about racial profiling (Bailey, 2004) . They keep denying the problem saying that there is no such thing as racial profiling, because there is no difference between black and white, there are just different shades of brown (Ramos and Musumeci, 2004, p. 11). On the other side, racial profiling also seems to be a problem partly caused by the media as well, giving the fact that television shows, news and telenovela’s8 are mainly dominated by white people. Black people are just shown in relation to violence and criminal acts on the news. In the telenovela’s they barely play lead roles, but always act as waiters or other personnel (Adkins, 2007). The organization of the Brazilian police is very different from the organization of the Dutch police force. There are different types of police for different levels. On the level of the whole country there is the federal police, that deals with national phenomena like terrorism, organised crime, immigration and border control and responsibility for the indigenous people (Art 144.1I-IV Brazilian Constitution). On the state-level there are the military police and the civilian police (Husain, 2007, p. 77, 80; art 144.4-144.6 Brazilian Constitution). The military police prevents organized crime and is organized in battalions. They are also keeping order in daily life and during big events like carnival (Husain, 2007, p. 74; Polícia Militar da Bahia, 2016). The civil police have more of a juridical function and carry out investigations, for example after a theft (Husain, 2007, p. 75). The civil police are not uniformed, although they are to be recognized in public because they have official police cars (Oliveira, 2013). There is rivalry between the military police and the civil police, also because they are paid from the same state budget (Husain, 2007, p. 75). In both forces, the salary is very low, and there is a big salary gap between higher officers and lower ones (Husain, 2007). 8 Defined as “A soap opera produced in and televised in or from many Latin-American countries” (merriam-webster dictionary, 2016).
22
Problem definition
Introduction: As described in the theoretical framework, the police are, being street-level bureaucrats that often deal with conflict, an important link between society and policy makers as well politicians. Trust is a necessary and central concept for a street-level bureaucrat to carry out their mediation function. However, it can be stated that the amount of trust in the police by the population in Brazil is generally very low. In situations of conflict, citizens are not used to go to the police but they have more trust in the local organized crime groups for seeking restitution (Willis, 2015, p. 58). Lipsky (1989) describes various tactics citizens use to seek help from street-level bureaucrats. Within seemingly unimportant daily interactions between citizens and the police, conflicts are framed and policy is made (Lipsly, 1989). Actions of officers are a combination of policy regulation and discretionary space, in which the officers have agency in the discretionary space. In a situation of conflict, different people have a different perception of the situation. The moment of conflict between street-level bureaucrats and citizens can be used to adapt policy to include more citizens and make policy in that way more successful (Verloo, 2015, 296). In this story, citizens are not only a passive receiver of policy but they also participate in policymaking. The moment of conflict is a chance to include potentially marginalized citizens and is therefore a chance to design a ‘negotiated democracy’ (Verloo, 2015). The general view among scientists as well as journalists is that the Brazilian police fail to include all citizens in their policy objective. There is little research about how daily interactions between the police and citizens affect the power relations between them. The research that exists is primarily focussed on human rights or the excessive amounts of violence that is used by the police (Husain, 2007; Willis, 2014; Koonings and Kruijt, 2015). During non-violent daily activities of the police, the urban democracy as well as security policy is already determined. Non-violent daily interactions already decided who is included and who is excluded in certain places, situations and policy. These daily routines of inclusion and exclusion are a potential cradle of violence (Çankaya, 2016). Studying this last step of policy-implementation from the perspective of a negotiated urban democracy can help to understand why the current policy for the police is so unsuccessful.23 In this urban democracy, there are different actors that interact with or need the police, like the policymakers and citizens discussed in the first part of the theoretical framework. These parties have a different view on the actions of the police, and the role of the police in the social reality as well as the negotiated urban democracy. To make sense of the social reality, they handle several different and sometimes conflicting frames. The media tell a certain dominant story, and victims or the police might tell another story. In the end, the people that conquer the dominant frame are traditionally the most likely to influence policy (Verloo, 2015). However, the Brazilian police are framed in the media as highly corrupt and dysfunctional, with a nature to kill innocent people when they are at the wrong place at the wrong time (Willis, 2015). On the other hand, public security is still considered an important subject for Brazilian policy makers and politicians (Mesquita Neto in Silva, 2014). This research studies both the story of the citizens that problematize the police as well as the stories of the people that work for the police. Comparing these different frames gives insight in the themes that are causing conflict among the stakeholders. Differences in values or different perceptions of the situation can cause conflict, and the police officer that is supposed to be the mediator in the system becomes the enemy. Discovering perceptions of the police culture as well as the culture of people that often interact with the police is the first step in restoring the relation of trust between citizens and the police. Main Question: How is discretionary space used by police officers in the interaction between them and the community in daily life? Subquestions: 1. What form has the discretionary space of an officer? 2. Who influences this discretionary space? 3. Which values/concepts are important when interacting with citizens (human rights, personal values, protocol)? 4. What is the vision of citizens on the use of discretionary space of the police? 5. Which tactics do citizens use when interacting with the police (to secure themselves, or to reach their goal of getting service of them)?
24
Methods
Research Strategy This research has a constructivist ontology. This implies “that social phenomena and their meanings are continually being accomplished by social actors” (Bryman, 2010, p. 33). Discretionary space as well as trust in institutions are both social phenomena that would not exist if individuals do not believe they exist. The social reality is constructed by constant, often conflictual, interaction between the police and the citizens they serve. This research might also have post-modernist features, because it accepts the constant changing of the field in which the data is collected. In an anthropological case study, the social reality goes from the ‘filter’ of the respondent through the ‘filter’ of the author of this thesis and through the ‘filter’ of the connection to existing theory. It is therefore inherently not objective. I fully agree with the epistemological position of the ethnographer Van Maanen (1988, p. 35): “(…) I regard the relation between the knower and the known to be a most problematic one and anything but independent in cultural studies (.…) Ethnographies (…) are always subject to multiple interpretations. They are never beyond controversy or debate.” This research is a combination of a confessional tale with a realist tale (Van Maanen, 1988). The functioning of the police is a sensitive subject and the culture gap influenced the data gathered by participant observation. The realist tale is chosen as the leading form of writing, to maintain a certain sense of objectivity. However, my presence as a foreigner has had such a big influence on the studied situations, that it is impossible not to include my behaviour and reasoning without compromising the data. The confessional tale allowed me to doubt some aspects of the fieldwork, allowed doubting strategies of gaining access and allowed more space to describe the problem of conducting a research abroad (Van Maanen, 1988, p. 75). This research has an interpretative epistemology, because the visions and the stories of the respondents are the leading data in the research (Bryman, 2010). There is a lot of research conducted on the realist basis, like crime statistics and death toll by the police. This research is built on interpreted stories, trying to explain the police culture and attitudes and the strategies of the subjects of the police. Lipsky (1989 [2010], p. 40) states that performances of street-level bureaucrats are extremely difficult to measure, so this would require an interpretative epistemology.25 There is a close relation between inductive and deductive strategies in this thesis. The theoretical framework is adapted after the data collection. This approach is defined as ‘grounded theory’ (Bryman, 2010, p. 570). The concepts and the theory about the use of discretionary space emerged from the collected data. The main question is an open, broad question because the intention of the research is to explore the subject. There is a lot of media coverage about the police in this region, and there are several anthropological researchers on the police in other states of Brazil (Husain, 2007; Willis, 2015; Ahnen, 2007). However, these researchers are all somehow related to the high level of violence of the police, as well as used by the population they serve. Because of the open question, there is more room for new information and unexpected angles. The stories, the opinions and the framing of the police officers, as well as the people that encountered the police are central in this research. The respondents say what they want to say on the subject, and present what they want to present. The advantage of this strategy is that the researcher cannot miss any information that the respondents think is important in understanding their use of discretionary space. On the other hand, the respondents presented just their side of the story, and might be ‘hiding’ any information that would damage their or the organizations image. In the literature studied, discretionary space is misused which results in unjustified violence, corruption and even connections with organized crime. However, these topics are not the main focus of this research. Also, the direct behaviour of the police is studied in participant observation, which makes it possible for the researcher to observe the relation between the community and the police directly. The citizens in the field were not informed about my presence, so their reaction was more natural. Respondent groups The military police of Itaparica Island is interviewed, as well as the police of Salvador. The military police and not the civil police is the main focus of this research, because they are the more likely to get into violent conflict with citizens.9 They are the ones that prevent crime as well as they must deal with the outbreak of violence in a certain area. Itaparica is a semi-urban area that did not have a lot of tourists at the time 9 This will not say that the Civil police and the Municipal Police do not kill or do not have permission to kill (See for example Globo 1 Bahia, 2016a; Correio, 2016). However, because providing public security is mainly the function of the military police, are responsible for the majority of killed citizens (Willis, 2015).
26 the research was conducted. There is a high social cohesion in the city because almost everybody knows everybody, and a lot of people are family of each other. This could influence the relation between the community and the officers. Salvador is a mega-city with 2,9 million inhabitants (Instituto Brasileiro de Geografia e Estatística, 2015). The police do not know the citizens they interact with, which might make the relation between them different. This urban environment is more like the environment studied by anthropologists in Rio de Janeiro and São Paulo (Willis, 2015; Husain, 2007), and it is therefore expected to have similar problems. To capture both sides of the story, I also talked to youth about their experiences with the police. They told several stories of misconduct, and I let them ‘frame’ their side of the story. This research will not focus on the truthfulness of the stories of the police officers nor the youth I talked to, but will show a shared reality and norm system within the two parties. The framing of the situation and the making sense of the stories is the main focus of the research. I talked to youth that I encountered on the streets in Plataforma, Pelourinho and San Antonio. This was merely convenience sampling (Bryman, 2010), but I only interviewed people that met the criteria of a “routine and unexceptional killable subject” that is “a man, often dark skinned, living (…) in the urban periphery of a low socioeconomic background (…)” (Willis, 2015, p. 30). I wanted to talk to people that might have had bad experiences with the police, so I choose to talk to a risk group. I interviewed dark-skinned males between 8 and 22, who are mostly enrolled in a local football project for lower socioeconomic youth. Pelourinho and San Antonio are not in the urban periphery, but Plataforma is. To avoid tension between the community and the police, and not to breach the relation of trust I built with the respondents, I decided not to interview officers and community members in the same bairro.10 Gaining access The respondents will be gained by ‘snowball sampling’. From the Netherlands I made one Brazilian main contact that speaks Portuguese and is known in the local community of Itaparica. He arranged the first contact with a base of the military police11 on the Island. To get permission, I wrote a research proposal for the military police and 10 Neighbourhood. 11 The police stations of the military police are called bases, and the police stations of the civil police are called delegacias.
27
for the communities that I wanted to talk to (see appendix I), that was translated for me in Portuguese. After talking to a Major12 on the Comando Geral da Policia Militar13, the
command centre of the Military Police in Bahia in Salvador, he gave me two official letters containing permission to conduct field research in one regular base and one community base. The capitão of the first community base mentioned the other base and was acquainted with the capitão that worked there, and I did not need separate permission. In the touristic neighbourhood and historic city centre of Pelourinho, I met a tourist guide that also has a social project playing football with children from the communidades. He brought me into contact with youngsters in the streets of Pelourinho and the neighbourhood next to it: San Antonio. I also talked to youngsters in the Suburbania14, specifically in the neighbourhood called Plataforma. Research Methods This research is a combination of interviews and participant observation. On each first visit of a new police base, I conducted an unstructured interview with the capitão. This was a good way to get some basic information about the base, and an easy way to build a relationship of trust between the researcher and the respondents. Giving the fact that I was the first foreign researcher in two of the three bases I visited, the respondents had a lot of questions about the goal and the public of my research. The military police in Brazil is very hierarchical, so the first contact with the capitão was important for the level of access that was allowed afterwards. I went to three different bases, and the nature of the activities I did differed substantially. On the first community base, I did an interview with the capitão. After this, I interviewed the police officer that gives the Judo lessons at the base. The third visit I conducted participant observation at the celebration of the 5th anniversary of the base, and the fourth visit I accompanied community visits with especially trained police 12 This is the rank directly above capitão. 13 The command centre of the Military Police of Bahia, here is also the department of social relations, the head commander of the Military Police and they realize the implementation of the public security policy of the state of Bahia. 14 The Suburbania are the bairros Plataforma, São Tomé de Paripe, Periperi, Alto da Terezinha, Coutos, Fazenda Coutos, Itacaranha, Nova Constituinte, Praia Grande, Rio Sena, São João do Cabrito, and the Islands de Bom Jesus dos Passos, de Maré e dos Frades that are further away from the center. The number of homicides is much higher in these bairros, there are problems with social security and the majority of the residents is black (Mendes, 2015).
28 officers. I also interviewed an important member of the community that manages the community library and sometimes organizes events together with the police. On the second community base, I started the day visiting an event of sexuality education, which was organized by the police. After this, I talked to the karate teacher that is also military police and to several people from the community that help the police in various ways. Lastly, I interviewed the capitão. On the regular base on the Island of Itaparica, I firstly interviewed the capitão as well. After this, I stayed on the street with two soldados que trabalham em pé,15 In Pelourinho and San Antonio I gathered 5 stories and I gathered 6 stories in Plataforma. I wrote the majority of the stories down in Dutch, as they were telling them. Sometimes this was complicated because of the fact that they speak differently from what I am used to, so my contact wrote down two of the stories because I was unable to understand them. Issues encountered Planning my fieldwork, I knew that gathering my data would be difficult. The difficulties described below influenced my chances of getting data, and might have influenced the outcomes of this research. The first and foremost problem I had during my stay was the cultural difference. Because I am from Europe, the people who encountered me treated me like ‘the other’. When I said I came from the Netherlands, some people said ‘aaah premeiro mundo’16, or asked me “a vida la é melhor, não é?”17 and I have been called gringa more times than I can count. Sometimes this was negative, because of the implicit hierarchical relation between the subject and researcher that always exists. Other times it was an advantage, because I could enter neighbourhoods where ‘rich Brazilians’ could not enter. Everyone knows not to engage in violence with a foreigner because the police will react with excessive repercussions to uphold their image of a safe country, especially with the upcoming Olympic Games. A second difficult thing was gaining trust and explaining my project to the respondents. A lot of people within the Military Police reacted surprised that I wanted to 15 Literally: Soldiers that work standing. These are policemen at the first year after their 6 months of education to become military police. They do a sort of internship of one year, patrolling the streets at busy places. 16 Aaah, first world. 17 The life there is better, isn’t it?
29 do research on them, and wanted to know what my research was exactly about and who would read it. The management of the Military Police told me that I needed to write a research proposal in order to get permission. I wrote the document in English, and a friend helped me to translate it into Portuguese. The document (see appendix I) stated that I wanted to understand the view of Military Police officers and to understand their views and dreams. I also had an official letter of my thesis supervisor. This documents showed that I was prepared and that the project was a serious and important thing for me, despite the fact that I sometimes had difficulties explaining my objective in Portuguese. At the first moments, they expected me to ask targeted questions or decide what they had to show me. I explained them that I wanted to do what they normally do, so after some time they relaxed and talked to me like I was an acquaintance that they had to show around. They normally asked me a lot of questions about the Netherlands, there was a true exchange of information for both parties. I never spoke about ‘discretionary space’, but spoke about the ‘the relation between the community and the police’ in Brazil. Most of the time I explained that the society in Brazil is more violent than in the Netherlands, and I wanted to understand how the police work in this situation of violence. Often people could relate to that, and it was a neutral explanation where neither the respondents nor the community are judged. It was hard to stay objective and critical during this research, because the subjects of research bond with the researcher (Bryman, 2010). There was an informal ambience during the fieldwork, and I befriended several police officers during this time. This was especially hard when I wanted to talk about the unjustified violence the police used according to the literature. Sometimes I mentioned this, other times the police brought it up. Because this research is built on several stories of people, including my stories, I did not want to exclude any point of view. Not to exclude any frame, I also listened to people who told me that the stories about violence of police officers were not true, and I decided to write down their part of the story as well. I also stated to all of the parties that I just wanted to collect stories in order to understand the relation between the community and the police better. I explained them that I also had to ask about the violent part of the police, because it was in the literature. The third obstacle was the language. The first month I did not do any field research or interviews, but spoke Portuguese with everyone around me. The first contact also always was in Portuguese, as there are few people that speak English. I
30 decided to do the interviews in Portuguese as well, because it was easier and it felt strange to talk in English when I could make myself understandable in Portuguese as well. The interviews were recorded and transcribed by native Portuguese speakers, as I was not able to transcribe the interviews myself. Data analysis The data will be analysed according to the logic of grounded theory. The first step is open coding ‘the process of breaking down, examining, comparing, conceptualizing and categorizing data (Strauss and Corbin, 1990 in Bryman, 2010, p. 569). Then these open codes will be linked to context, consequences and patterns of interaction, to make broader or different categories, which will be tested to use them again as open codes. These core codes give the researcher tools to make a connection between several concepts, establishing a theory about the subject or a relation between the concepts. The last stage is selective coding, in which the core codes can be found. These are the broad categories that will integrate all the concepts (Bryman, 2010). Ethical Justification In advance of the fieldwork, I already knew that the framing of the subject was essential, not breaching the relation of trust between the respondents, and me, as well as not to create tension between colleagues. Therefore, I decided to use the point of view of Verloo (2015), who conducted research on urban conflict in the Netherlands. She stated: “I took the problem of conflict as a given fact and decided to turn it around; I came into the field with a focus on my genuine curiosity about people’s dreams and worries (…) I stated that my research was about finding out how people are engaged in changes in their neighbourhood’. Everyone could relate to the fact that there were a variety of ideas about changes in their neighbourhood and that they could contest one another. This frame allowed diverse groups (…) to become a stakeholder in my research.” I put a similar phrase in my research proposal, which sometimes was met by sarcastic comments but other times build support for my research. The second risk was to damage the career of one of the officers interviewed. I witnessed several actions from which it is doubtful if it is in the mandate of the officers
31 conducting it. I also talked to police officers and residents that were critical about the functioning of the Military Police. As a lot of officers from different ranks wanted to read my final thesis, I knew that there is a substantial risk of me being criticized for the way I framed their stories, but also cause tension within the organisation. Because of the military character of the organisation, the officers are tightly monitored if the organisation has suspicion of misconduct. As I am a researcher from ‘the first world’, the stakes of maintaining a good image are high. Some officers stated to me that they are more scared to be punished by the Military Police than they are scared of getting shot.18 To minimize this risk, I decided just to tell the rank and function of the police officer I am talking about. As there are few female police officers in Brazil, I decided to use ‘him’ and ‘his’ regardless of the gender of the respondent. The stories about the two community bases I visited are mixed, as well as the interviews of the capitães.
Results
The results will show that the discretionary space of Military Police officers in the state of Bahia is very extensive. Lipsky ([1980], 2010, p. 15) stated: “Discretion is a relative concept, it follows that the greater the degree of discretion the more salient this analysis in understanding the character of the workers’ behaviour.” As the discretion in this case is very extensive, the analysis of understanding the workers’ behaviour is very salient. My analysis of police discretion will show that police officers use a social system to estimate if a person is good or bad, and which implications this has for the use of arms by the police. Part I: The police This first part of the results will discuss how discretionary space is seen and how it is used by police officers. Lipsky already states that the police need a lot of discretionary space, even for a street-level bureaucrat, as they are “constantly work[ing] under the threat of violence that may come from any direction at any time (Lipsky, [1980] 2010, p. 31). Firstly, it is important to state that there are a lot of differences between officers. Some told me that they felt really safe at work, others told me that they also walk with a 18 Community policeman, personal communication, May 201632 gun in the weekends because they are scared to get shot by armed people that they encountered at work. Also, a lot of different departments of the police have very different daily work. The theatre group and the capoeira group of the Military police are only working with the part of prevention and social work, where the forca táctica is only working with repression. I also spoke to officers in training that are just walking around for the sense of security, and are not encountering bigger problems than a collision. However, there are some important features that shape the discretionary space of a police officer in Bahia, namely the identification with the organization, actions outside their mandate and the normalization of arms. Identification with the organization Being a police officer was not a conscious career choice for most of the people that I met. A lot of lower ranked officers took the job because they needed a steady source of income to pay for private education and there are not a lot of jobs available in Brazil. Others kind of rolled into it, because family members are also police officers or because they went to a public school of the military police.19 The majority of soldados I talked to are also doing faculdade20, and this is a major point of identification. One policeman that was patrolling told me: ‘I am an officer, but I am actually a historian.’21 Other policemen told me that they were interested in this thesis because they were sociologist or studying psychology. They also use their skills they learned in faculdade in their work and in interpreting their work. The police officer that is also studying sociology is selected to be the capitão of a base comunitaria. He told me he understands discrimination of people from the neighbourhood where he works, using the theories of Foucault and Goffman.22 This result is surprizing, as the literature states that lower ranked policemen in Brazil are often not educated (Husain, 2007). At another base, the officer that studies psychology was mediating a conflict. The mother of a child came to the police base for help, because he had a conflict with the father about custody. When the officer introduced his colleagues, he said: ‘We are from the community police, this woman came to us because he wants to see her child and you are not allowing this. My colleagues here studied law, so therefore we can inform you 19 These are public primary and middle schools that give normal education, but also teach the values of the military police like discipline. 20 Litarally faculty, meaning a type of private part-time university. 21 Patrolling officer, personal communication, italics added, May 2016 22 Capitão, interview, May 2016
