A Tale of Urban Violence in Brazil. The Case of Marielle Franco

The paper aims to analyze the first months of the federal intervention in the state of Rio de Janeiro introduced by presidential decree on the 16 th of February 2018. The security crisis in the state, which started in 2017 and gained significant media coverage, led to Michel Temer’s decision to put the Armed Forces in charge of public security, for the first time since the country’s redemocratization. Deemed controversial since its announcement in February, the measure was further questioned after the assassination of Marielle Franco, city councilor of the Municipal Chamber of Rio de Janeiro known for her fierce opposition to the military intervention. The paper examines whether the dysfunctionality of the Brazilian state in the sphere of public security can be tackled by hardline policies such as the federal intervention and questions the validity of this measure in the given context. It claims that harsh security policies are inefficient in combating violence in Brazil and stresses that the dominating discourse on crime and violence, which contributes to the public opinion’s support of counterproductive hardline measures, requires a redefinition.


Introduction
President michel temer announced on the 16 th of february 2018 that the national Armed forces would be put in charge of public security in the state of rio de Janeiro, a decision authorized by presidential decree and approved by both houses of congress the following week (câmara dos deputados 2018; senado federal 2018). This was the first time a federal intervention -as this constitutional measure is called -was implemented in Brazil after the end of the military rule in 1985 and the adoption of the constitution in 1988. The decision was justified by the argument of a security crisis in rio de Janeiro and its peak during the 2018 carnival, which supposedly witnessed an unprecedented number of gunfights and mass robberies. criticized by human rights activists, representatives of the academia and part of the state Assembly and federal congress members (lima 2018, Zambrzycki dutra 2018;ramos, Paiva 2018;lobato 2018), the decision has been seen more as an attempt to improve the president's favorability ratings, which at the beginning of 2018 hit rock bottom. 1 however, a crucial blow for the implemented measure was the assassination of marielle franco, city councilor of the municipal chamber of rio de Janeiro, human rights activist and ardent critic of the military intervention, on the 14 th of march 2018. This event reminded the public opinion about the inefficiency of extreme measures aimed at confrontational policies when tackling crime and violence in Brazil and in the state of rio de Janeiro in particular. 1 The disapproval rate of michel temer's government reached 70% in January 2018 and only 6% of respondents rated his government as "very good or good" according to a public opinion poll conducted by datafolha. These numbers dropped further in June, when the results of a new datafolha poll showed that 82% of respondents disapproved of the President whereas his approval rate was of 3%, thereby michel temer is the least popular head of state since Brazil's redemocratization (datafolha 2018b).

Monika Sawicka
The paper examines whether the dysfunctionality of the Brazilian state in the sphere of public security can be tackled by hardline policies such as the federal intervention. Also, it questions the validity of this measure in the given context. to this aim, definitions of state dysfunctionality will be highlighted in the first section, followed by an analysis of intentional homicide rates in the 21 st century in Brazil and its comparison with global intentional homicide rates, justifying the argument of Brazil's dysfunctionality. The next section will focus on violence rates in the state of rio de Janeiro between 1991 and 2017 measured by the number of intentional homicides. The last two sections will describe the tragic event of the 14 th of march and its aftermath, and expose the toll of the first seven months of the federal intervention in rio de Janeiro. The conclusion will highlight the main argumentation line of this paper: the inefficiency of harsh security policies in combating violence in Brazil and the necessity to modify the dominating discourse on crime and violence which contributes to the public opinion's support of counterproductive hardline measures.

The dysfunctional state
The category of dysfunctional states has experienced a significant popularity increase after the collapse of the bipolar system (Kłosowicz, mormul 2014: 18) which contributed to the increase of definitions of dysfunctionality, including new terms aimed at capturing different degrees of dysfunctionality: collapsed states, failed states, crisis states, fragile states, quasi-states, or weak states. one of the well-known definitions of state dysfunctionality was proposed by stephen ellis in his article published in "foreign Affairs" in 2005. According to ellis, dysfunctional states are unable to fulfill two main obligations: they cannot guarantee the rule of law and order on their national territory and are unable to fulfill their international commitments (ellis 2005). Another frequently quoted definition of dysfunctional states, proposed by robert I. rotberg (2003: 3), stresses that state dysfunctionality applies to countries that are unable to provide public services to their citizens within the national territory. In a hierarchical system of public services, security and in particular human security, is the most important. Providing security encompasses a broad set of state services, including preventing cross-border invasions, loss of territory as well as efficient combat of organized crime and other threats to domestic human security. rotberg's definition does not limit state dysfunctionality to the state's obligation to protect its citizens from external and internal menaces, but also mentions the duty to provide other public services, such as healthcare or education. It should be stressed that the dysfunctionality of the Brazilian state is not restricted to the area of public security. however, as the next section will show, the level of violence, expressed by the intentional homicide rate, makes the question of security the main argument in favor of classifying Brazil as a dysfunctional state.

How dangerous is Brazil?
IPeA, the Brazilian Institute of Applied economic research, together with the Brazilian forum of Public security (fBsP), a think tank based in são Paulo, published in June 2018 the newest edition of Atlas of Violence 2 , an empirical study presenting detailed crime rates and the profile of the victims. According to the 2018 study, the Brazilian homicide rate 3 was 30 times higher in 2016 than the european one (IPeA, fBsP 2018: 3). A recent study published by the Brazilian forum of Public security, Anuário Brasileiro de Segurança Pública 2018, presents the homicide rates for 2017. According to the document, Brazil broke once again its homicide record: 63 880 intentional homicides, corresponding to a homicide rate of 30.8 per 100 000 habitants. Another very important finding of crime rate analyzes refers to the racial profile of the victims: whereas the homicide rate of non-white citizens reached in 2016 40.0 deaths per 100 000 habitants, the rate was 16.0 for white citizens. This allows to claim that the risk of assassination is 2.5 times higher for black and pardo 4 Brazilians. The profile of the homicide victims becomes clear when taking into consideration their sex and age. The homicide rate for young people aged 15 to 29 is 142.7 per 100 000 and it further rises to 280.6 per 100 000 when narrowing the group to young men. The statistics show that the typical homicide victim is a black young man (fBsP 2018). 2 The study presented data for the year 2016. 3 Brazilian research institutions, including IPeA and fBsP, classify as "intentional violent deaths" 4 incidents: intentional homicides, theft following homicide (latrocínio), bodily injury followed by death and persons killed by military or civil police. see: fBsP, Anuário Brasileiro de segurança Pública 2018, http://www.forumseguranca.org.br/publicacoes/anuario-brasileiro-deseguranca-publica-2018/, access: 10. 09.2018. 4 Brazilians of mixed ethnic ancestries. Another distinctive feature of crime and violence in Brazil is the number of victims of police brutality. In 2017, 5 144 people were killed by the military or civil police, which corresponds to 8.1% of all homicide victims. In comparison, the same year in the usA police killed 1 147 people (mapping Police violence 2018).

Intentional homicides in Brazil
The unodc database allows to position Brazil's homicide rates in a broader context. According to un data, in 2016 5 Brazil was the country with the highest number of intentional homicides (61 283), followed by India (42 678) and mexico (24 559). A more exact picture is obtained, however, when comparing the homicide rates: in 2016 Brazil, with a homicide rate of 29.5, was the world's 8th most dangerous country (unodc 2018). The abovementioned data allows to justify the claim of the dysfunctionality of the Brazilian state, unable to protect its citizens and provide security.

How dangerous is Rio?
stating his reasons to implement federal intervention, President temer compared organized crime in the state of rio de Janeiro to a metastasis spreading across Brazil, which nearly took over in the state and required a "tough, firm answer" (temer 2018). The idea of engaging the military in issues of state security in rio de Janeiro and other federal units is not new. According to data published by the ministry of defense, between 1992 and 2017 Brazil witnessed 130 operations of guarantee of law and order (Garantia da Lei e Ordem, glA) -a measure foreseen in the Brazilian constitution of 1988 and implemented by presidential decree (ministério da defesa 2018). The glA, seen as federal support for the states in the matter of public security 6 , was applicable in cases of exhaustion of states' security forces and major public security crises. The military's role was to support the police in their operations, a support limited to specified initiatives and territory within the state. most importantly, this measure did not shift the responsibility for public security from the state to the union, as it happens in the case of a federal intervention. The latter measure provides the union with the power to interfere in public matters under the states' responsibility (garcia 2009: 58-61, scalquette 2003: 46-54). In case of the federal intervention in rio de Janeiro, the state's autonomy is suspended until the 31 st of december only in the sphere of public security. The decree nominated a military, general walter Braga netto, to hold the position of the 'intervener' (interventor) in charge of all public security forces including the military and civil policy and reporting directly to the President. concerns raised by specialists included the problem of the position of the intervener being offered to a general as well as the accountability of soldiers and policemen in cases of abuses of power (dw Brasil 2018; souza, oliveira 2018; lima 2018). It should be also stressed that the measure was implemented by a president who came to power due to a highly controversial impeachment process of his predecessor President dilma rousseff (taub 2016). most importantly, however, the question emerges whether the state of rio de Janeiro faced unprecedented crime rates which would justify the extreme measure.
Brazilian main television channel globo called the 2018 carnival "the carnival of violence", highlighting a situation of chaos and lack of police control, with tourists falling prey to an increased number of thefts and shootings taking place close to tourist attractions (globo extra 2018). Images of an extremely violent carnival once again reminded the public opinion of the surge in violence since early 2017 (uol 2018). media stressed that the levels of violence were reaching peak points, comparing the situation in 2017 with 2010 numbers (teixeira 2018). It is undeniable that rio de Janeiro experienced a spike in violence in 2017, the number of homicides increased by 7.5% compared to the year 2016 (IsP 2018). however, to understand the security situation in rio, a broader picture is required. data provided by the Institute of Public security (IsP), a rio-based think tank, shows that despite a significant increase in homicides in 2017, the number (6 749 deaths) and homicide rate (40.0 deaths per 100 000 habitants) is still significantly lower than murder rates from the last decade of the 20 th century (the highest number of intentional homicides being 8 638 deaths in 1995 and the highest homicide rate: 64.8 deaths per 100 000 habitants in 1994) (IsP 2018).

A Tale of Urban Violence in Brazil. The Case of Marielle Franco
Monika Sawicka further to a diachronic comparison of rio's homicide rates, homicide rates across Brazilian states in 2017 should also be highlighted when analyzing President temer's decision to proceed with the intervention. According to the data, in 2017 rio ranked second among the most violent Brazilian states measured by the number of homicides, with the state of Bahia leading the ranking (6 915 deaths). The picture changes, however, when homicide rates are compared: rio ranked only eleventh, with states from the Brazilian northeast region leading the list (rio grande do norte -68.0 deaths per 100 000, Acre -63.9 deaths per 100 000 and ceará, 59.1 deaths per 100 000) (fBsP 2018).
Available data shows that the federal intervention is, therefore, questionable not only due to its timing, but also its location. Both journalists and scholars have highlighted that the presidential decree was signed shortly before a planned voting on the main project of temer's administration: the pension source: IsP, Indicadores Gerais da Segurança, http://www.isp.rj.gov.br/conteudo.asp?ident=155, access: 2.09.2018.

Graph 3. most dangerous states in Brazil
Graph 2. Intentional homicides in the state of rio de Janeiro system bill. The chances that the unpopular bill would pass in congress were close to none and the sudden turn of the president's attention towards the security crisis was interpreted as a way to avoid the failure of the administration's trademark project. The pension bill project would have been a constitutional amendment, on which the congress is not allowed to vote during a federal intervention (constitution of the federative republic of Brazil 1988, title Iv, chapter I, section vIII, subsection II, Art. 60, § 1).
Another possible explanation for temer's decision were the president's favorability ratings at the beginning of the election year. temer's decision to focus his attention on security matters seen by many Brazilians as the most pressing problem of the country (cardoso 2018) can be interpreted as a move to save the compromised image of his government and secure followers. In Brazil, a tough stance towards public security challenges is commonly used to assure supporters, as a significant part of the public opinion is traditionally in favor of militarization of pubic security policy and a confrontational approach towards what is seen as 'the war on crime' (malaguti Batista 2003: 35-37; ramos, Paiva 2007: 57; caccia Bava 2013:3; lima, sinhoretto, Bueno 2015: 125). The endorsement of the federal intervention has been measured by datafolha, a polling institute of grupo folha. The poll results show a signi ficant yet decreasing support for the military's involvement in combating violence in rio de Janeiro, with 79% of respondents being in favor of the measure (datafolha 2018a).
despite the measure's popularity, the tragic event on the 14 th of march in rio de Janeiro questioned the efficiency of the federal intervention and reminded Brazilians of frequently overlooked aspects of the public security challenge.

Who killed Marielle Franco?
marielle franco was a 38-year old favela-born politician who got elected by over 46 000 votes to the city council of rio de Janeiro in 2016. Known for advocating for women's, lgBt and favela residents' rights, she was regularly denouncing cases of police and army brutality -last time on the morning of her death. Being a fierce opposer to the militarization of public security, including the federal intervention, franco also became head of a council commission created to monitor the intervention two weeks before her death. There are no doubts that the assassination of marielle franco was linked to her political activism. franco and her driver Anderson gomes were fatally shot on their way home from an event dedicated to black women mobilization, held in central rio de Janeiro. The assassins must have been following the councilwoman as they knew her location in the car which had tinted windows. After firing nine bullets they escaped without stealing any belongings from franco's car. The civil Police of rio de Janeiro, in charge of the investigation, from the start ruled out robbery as a motive, indicating that

A Tale of Urban Violence in Brazil. The Case of Marielle Franco
Monika Sawicka the case showed signs of an execution with a possible involvement of corrupt members of the military police or militias (cruz 2018). 7 The question which, therefore, arose was not only who killed marielle franco, but also who ordered the assassination.
on the 15 th of march several Brazilian cities witnessed the outcry against the assassination of franco and gomes as thousands of protesters gathered on the main squares of state capitals. In rio de Janeiro, hundreds of people were expressing their indignation not only because of the death of a human rights defender. franco's assassination was a rare occasion to remind the public opinion of frequently overlooked aspects of violence: the impunity of corrupt police officers commonly targeting underprivileged inhabitants of the favelas and growing power of the militias terrorizing poorer districts of the city. If everyday cases of crime and violence were used by the media to produce sensational front covers and fill in the violence section of journals, the incident from the 14 th of march induced a greater number of citizens to address deeper causes of the problem and reflect upon overlooked issues. Among the most important ones was the situation of favela residents, whose lives were put at stake every time the police entered a favela to crack down on gangs. This caused protesters to question the sense of "heavy fist" policies. The assassination of marielle franco and the outburst of citizen indignation, which followed instantly, created a rare opportunity to question the metaphor of war used in media and authorities' discourse. This metaphor followed by a matching set of images and stereotypes related to the profile of criminals (poor, black and living in favelas), and hardline solutions necessary to handle the permanent security crisis. This imaginary had also a predominant impact on the constant feeling of insecurity experienced by the Brazilian middle-class and the development of a culture of fear and violence in Brazil (malaguti Batista 2003: 35;souza e silva 2007: 94-96). franco's assassination provided a rare opportunity to remind Brazilians about simplifications in the public discourse which were constantly revealed by rio's councilwoman. Protesters in Brazilian cities in march 2018 focused their demands on curbing police and militias impunity and claiming effective investigation policies rather than supporting tough measures to reduce crime rates at all costs (deister 2018b).
The assassination of marielle franco was a serious blow for popular support for the federal intervention. however, its sense and efficiency are further questioned when analyzing the crime statistics for the first six months of the intervention.

How efficient is the federal intervention?
The center of security and citizenship studies at the candido mendes university in rio de Janeiro, one of the leading academic centers focused on crime and violence research in Brazil, created the initiative Intervention observatory (Observatório da Intervenção) in february 2018. The observatory follows and assesses the progress and projects undertaken in the state of rio de Janeiro under intervention, as well as publishes the statistics of the intervention on a monthly basis. Infographics presenting the numbers of assassinations, killed police officers, and victims killed by the police or number of robberies are published on the 16 th of each month.
In an interview for radio cBn rio, silvia ramos, the coordinator of the project, gave a very stern assessment of the intervention, emphasizing that the numbers were alarming and events which happened after the 16 th of february proved that the police and military neither regained control over the city and its suburbs nor managed to curb police brutality or corruption as declared at the beginning of the intervention. ramos also criticized the model of the intervention, embedded in the logic of war and confrontation, rather than an urgently required reconstruction of police forces, improvement of internal procedures aimed at strengthening transparency and accountability or capacity building with focus on training and intelligence capacities (cBn rio 2018).
regarding the numbers of the intervention, the data published by Observatório da Intervenção on a monthly basis shows that rio de Janeiro ten months after the intervention is a city even more violent, with a higher number of shootings (increase by 56%) than before the measure decreed by President temer (observatório da Intervenção 2018). The number of homicides and people killed by police or armed forces increased in the first six months of the intervention, as shown by data published by the Institute of Public security. Between february and August 2017, the number of homicides in the state of rio de Janeiro reached 3 810, the number for the same period in 2018 is 4 039 (IsP 2018). Particularly alarming are the numbers of people killed by the police or armed forces: 615 during seven months between february and August 2017 and 916 between february and August 2018. The increase by 50% is appalling, as already in 2017, the police was responsible for 16,7% of all violent deaths in the state of rio de Janeiro and for the city of rio de Janeiro the share was significantly higher, reaching 25%. According to the latest report published on the 16 th of december, during the ten months of the intervention the number of homicides slightly dropped (-6,7%) in comparison to the same period in 2017 but nevertheless remained alarmingly high (4127 deaths). The most striking outcome is however the 36,3% increase of deaths by police forces (observatório da Intervenção 2018). due to ongoing impunity of police and army representatives during the military intervention, these numbers could get even higher in 2018. street-theft numbers are another indicator of crime in rio de Janeiro which did not decrease with the implementation Among events which reverberated through Brazil after the 16 th of february the assassination of marielle franco was the most high-profile case which made headlines across the world. In June, another death brought rio's citizens to the streets: the death of fourteen-year-old marcos vinicius who was shot on his way to school during a police operation in the favela complex of mare. specialists on public security also criticized the use of helicopters by security forces which strafed densely populated areas of the favelas, as well as the inability of the intervener and his team to explain what happened in rocinha during a police intervention in march, which caused the death of at least six people, or another operation in city of god in may with a death toll of four (deister 2018a).
nine months after the assassination of marielle franco, the identity of the murderers remains unknown. It has been highlighted that the investigation takes longer than investigations of other famous cases of political murders. so far, media informed about two lines of investigation with little progress in both in recent weeks, and in August the investigation team has been replaced (cruz 2018). The lack of progress in franco's case can be yet another proof of very limited investigation capacity of civil police. It is questionable whether representatives of armed forces in charge of public security in rio de Janeiro will change this picture as they lack expertise in police intelligence, combating corruption or improving institutional transparency, which should be at the core of changes to be implemented within the state's military and civil police.
The problem of transparency also surfaces when analyzing the spend of resources during the federal intervention. several weeks after the presidential decree instituting the intervention, general Braga netto, the intervener, stated that the intervention would require Brl 3.1 billion ($ 770 million). At the end of march, the federal government informed that resources would reach Brl $1.2 billion. In the first two months of the intervention only about 3% of this sum has been spent to buy ammunition, 16 armored vehicles and 200 rifles (Pietricovsky, mororni 2018). At the end of it, the spend is still very low, reaching only 6% (observatório da Intervenção 2018).

Conclusions
The strategic Plan of the federal intervention was published only four months after the intervener had taken power over public security in rio de Janeiro. Among its objectives are reduction of crime rates, recovery of the operating capacity of police forces, coordination of different segments of public security, strengthening of the technical character of these forces, as well as restruc-turing of the management of the penitentiary system. The past ten months have shown that little progress has been made to meet these objectives and the federal intervention has been described as chaotic, unprepared and focused on short sighted punctual operations. no progress was noticed in the improvement of coordination between the military and civil police forces or crime investigation focused on dismantling criminal networks, including militias. The fundamental crime indicator, the number of homicides, shrank insignificantly, contrary to the number of deaths caused by security forces which increased by over 30%. According to silvia ramos, the employment of the federal intervention, which surprised Brazilians, raises the question of the next steps in case of a probable failure of the ongoing intervention: "After the intervention you have the state of emergency, then the state of siege and then you drop an atomic bomb" (dw Brasil 2018). datafolha in cooperation with the Brazilian forum on Public safety published in march 2018 the results of a public opinion poll about fear, risks and victimization in rio de Janeiro. According to the poll, 92% of the respondents, residents of the city of rio de Janeiro, feared stray bullets, also 92% were afraid of being hurt or dying during an assault and of standing in the middle of a gunfight between the police and criminals. despite the small sample of respondents (1 012) limited to only one Brazilian city, the poll reflects the culture of fear that is a core social phenomenon observable among the Brazilian society (datafolha, fBsP 2018: 6). The common sentiment of dread and menace was further expressed in the high approval rates for the intervention and army engagement in security matters in rio de Janeiro: 83% in october 2017 (datafolha 2018a).
however, the outcomes of the first six months of the intervention, assessed negatively by specialists, seem to also have an impact on the public opinion: the deaths of marielle franco and marcos vinicius mobilized thousands of Brazilians who took to the streets to protest against police brutality and biases against black and poor residents of rio's favelas. The number of contestants of armed forces engagement in public security matters in rio de Janeiro also grew between october 2017 (15%) and August 2018 (27%), although the number of supporters of the federal intervention (66%) and its prorogation (72%) remains high (datafolha 2018a; estarque 2018).
As expected by specialists, the intervention did not provide a significant turn of the security situation in rio de Janeiro. As the federal intervention has been the most far reaching and drastic measure aimed at combating organized crime in recent Brazilian history, it is crucial to hold the federal government as well as the intervener and his team accountable for it. In-depth analyses should assess outcomes of the measure, indicate the reasons of its inefficiency and continue to stress necessary policies to be implemented. The most important is the role of the media in the dissemination of this assessment, yet it is not certain that the media will take such role. so far, images of crime and violence robustly exposing assault cases on representatives of the Brazilian middle