Variations in Homicide in the Triângulo Mineiro (Minas Gerais): Factional Conflict and Analytical TimeAbstract
Over the last two decades, homicide rates in Brazil have varied greatly across time and space. However, we observe little variation in the preferred victim profile. Based on the case of the Triângulo Mineiro region, we propose to interpret this phenomenon of variations in homicide rates by considering the analytical time of sociological transition between different orders of regulation of lethal violence. The preferred victim profile does not vary because these deaths are subject to strong regulation of lethal violence. We used a multi-method approach with analyses of summaries of lethal conflicts in relation to rates disaggregated by sociodemographic profiles. Our results reinforce the hypothesis that factional conflict directly impacts fluctuations in aggregated rates.
Keywords:
homicides; factional conflict; multi-method; urban violence; quantitative series
Resumo
Nas últimas décadas, as taxas de homicídio no Brasil variaram pouco em relação ao perfil preferencial da vítima. A partir do Triângulo Mineiro, buscamos compreender a variação das taxas de homicídio considerando o tempo analítico de transição sociológica entre diferentes ordens de regulação da violência letal. O perfil da vítima preferencial não varia porque essas mortes respondem a uma forte regulação da violência letal. São analisadas sinopses de conflitos letais em relação às taxas desagregadas por perfis sociodemográficos. Os resultados reforçam a hipótese de que o conflito faccional impacta diretamente na flutuação das taxas agregadas.
Palavras-chave:
homicídios; conflito faccional; multi-métodos; violência urbana; séries quantitativas
Introduction
Over the past two decades, homicide rates in Brazil have varied considerably over time and across geographic space. At national level, aggregate homicide rates suggest that violent deaths increased persistently and steadily until 2017, after which they followed a consistent downward trend. Examining specific states and cities, aggregate homicide rates suggest rapid fluctuations in lethal violence over the same period. This article tests a recent hypothesis concerning the variation in these rates (FELTRAN et al., 2022) and proposes the notion of analytical time as a means of articulating qualitative and quantitative methodologies.
While we recognise the numerous conceptual and bureaucratic issues involved in the production of crime statistics in Brazil, the homicides examined in this article are understood as acts of intentionally perpetrated lethal violence - that is, as Weberian social actions (WEBER, 2019). Bureaucratically, these homicides are classified as “deaths from external causes” and constitute avoidable, socially determined, predictable and preventable deaths. They also show a high degree of regularity in terms of the typical profile of the victims, who are predominantly young black or brown men and, as the literature demonstrates, occupy the lower echelons of the operation of illegal markets (CORDEIRO, 2023; FELTRAN et al., 2022). This regularity means that, even though the levels of homicide rates vary across Brazil’s states and regions, the lethal victimisation of young black men directly impacts aggregate variation in the quantitative data. Since rates are bureaucratic representations of the phenomenon of homicide rather than its empirical manifestation,2 here we analyse the fluctuations in homicide curves through the production of synopses of urban conflict dynamics identified through qualitative studies.
The sociological literature on homicide has demonstrated that power disputes surrounding illegal markets directly impact lethal violence in Brazil (RATTON; DAUDELIN, 2018; RODRIGUES, 2020; ZALUAR, 2002, 2019), with the lower-level operators of these illegal markets constituting precisely the predominant profile among homicide victims (HIRATA, 2018; HIRATA; GRILLO, 2019). Feltran (2019) demonstrates that violent deaths related to the “crime world” and its associated networks account for approximately 75 to 80 per cent of homicides in Brazil. More recent studies combining qualitative and quantitative approaches have further supported this argument (CORDEIRO, 2023; FELTRAN et al., 2022), demonstrating that, in addition to conflicts surrounding criminal dynamics and the sociodemographic factors mentioned above, being embedded in illegal markets as a lower-level operator considerably increases the likelihood of becoming a homicide victim.
This concentration in the victim profile of violent deaths demonstrates that lethal violence in Brazil is not chaotic, but is to a large extent strictly regulated by armed groups (FELTRAN, 2018). Far from being diffuse throughout the national social fabric, contemporary lethal violence is the development of organised forms of exercising local power, which reveal the particularities of conflicts over wealth and expose “the traces of histories, institutional failures and the routine violence of everyday life” (DAS, 2020, p. 188). However, although there is a clear regularity in the profile of preferred victims, homicide-rate trends have varied very quickly in contemporary Brazil, as a result of multifaceted sociological processes that warrant closer investigation.
The reconstruction of empirical synopses based on qualitative research, combined with the disaggregation of homicide rates by sociodemographic profiles, were two complementary procedures that enabled us to argue that disaggregated historical series reflect conflict and pacification processes among armed groups involved in illegal markets, which occurred at distinct temporalities in each of the three municipalities studied, all located in the Triângulo Mineiro region in the west of Minas Gerais state. Without a knowledge of the senses of justice that order this lethal violence, and relying solely on deductions from aggregate data, it would have been much more difficult to understand the sociological mechanisms that affect the variation in regional homicide rates. Over the past two decades, the Brazilian literature has made considerable progress in understanding the social dynamics and everyday contexts in which lethal violence has predominantly been produced and rendered plausible. This qualitative knowledge, necessary to comprehending the sociological mechanisms that produce lethal violence, has been recognised as fundamental to advancing the international literature on homicide, violence and urban conflict.
Pursuing this direction, we compare three municipalities from the same region, each a different size, in which variations in homicide rates initially appeared chaotic and without any regularity. While the rates in one city were rising very rapidly, neighbouring cities were experiencing a substantial decline in numbers, with these trends altering the following year. But what were the mechanisms of homicide regulation in these three disparate urban contexts in the interior of the state of Minas Gerais over the past two decades (between 2000 and 2020)? What have the consequences been of the power struggles surrounding illegal markets for official homicide statistics, and what does this violence - a category central to political theory - leave us with as a principle for thinking about urban order in Brazil today?
Our findings demonstrate that, in the Triângulo Mineiro, the early arrival of an increasingly organised external crime group - the Primeiro Comando da Capital (PCC) - was the primary factor that directly impacted local crime dynamics. Disaggregated homicide rates track the conflicts that followed, as we shall see in the qualitative synopses of the municipalities studied. Before turning to these, however, it is worth reviewing the literature and the methods employed.
Different theoretical and methodological approaches to the study of homicide: combining qualitative and quantitative methods
Homicides are multicausal, relational and complex phenomena. Accordingly, the hypotheses developed in the literature, whether through qualitative or quantitative approaches or through studies combining the two, are neither definitive nor entirely opposed to one another. Studies employing exclusively quantitative methods, particularly given their deductive character, often lack descriptions of either the histories that contextualise lethal violence in particular localities or the ways in which this violence is regulated. The data is presented as though violence did not involve deliberation and/or interdiction, as though it were decided upon instantaneously and individually, rather than organised by armed groups, whether police or criminal. Yet it is precisely this organisation that jumps out in qualitative studies.
On the other hand, these exclusively qualitative studies, which manage to capture the mechanisms organising lethal violence, lack the capacity to demonstrate - at regional, national and even international scales - when these mechanisms are operating. Comparison helps to make such inferences, while quantitative analysis is necessary to enhance the generalisability of these findings.
There are still important issues to be resolved in the international debate. Gerring and Knutsen (2022) have recently sought to outline a general theory of homicide, situated within the field of sociological and political theory. This theory operates at transnational and historical levels, taking the history of the state at its epicentre, and argues that variation in aggregate homicide rates can be explained by long-term social and political processes - such as (i) the existence of a government; (ii) the proportion of territory covered by that government; and (iii) whether this government was indigenous or imposed externally through colonisation. However, in addition to disregarding short-term variations in specific contexts because it relies solely on aggregate data, this theory also overlooks the concentration of homicides in time and space, across different sociodemographic profiles, or in contexts where violence is regulated by established criminal or police groups. The theory helps us understand structural and long-term variations in lethal violence, but impedes us from interpreting what happens in cases of rapid, local and short-term variation in homicide rates. This requires disaggregation by sociodemographic profiles and by the motivations underlying violent deaths, as well as situated analyses. If we start from the assumption that the conditions under which lethal violence is produced and reproduced are multiple and variable according to local configurations, and cannot be explained exclusively by structural factors, this theory presents evident limitations. It cannot, for example, explain what happened in the Triângulo Mineiro between 2000 and 2020.
The sociological literature on lethal violence in Brazil has advanced diverse interpretations of the phenomenon of homicide. Recent research demonstrates that political disputes surrounding illegal markets directly impact the occurrence of homicides (CORDEIRO, 2023; FELTRAN et al., 2022; KAHN, 2021; RATTON; DAUDELIN, 2018; RODRIGUES, 2020). Relationships established around illegal markets, whether via political commodities (MISSE, 2006) or via relations between the “crime world” and the forces of law and order, have also been invoked to explain fluctuations in homicide rates (CANO; DUARTE, 2012; FELTRAN; MOTTA, 2021; HIRATA; GRILLO, 2019; RATTON; DAUDELIN, 2018).
Some studies have also sought to explain the reduction in homicides, particularly in light of the consistent decline observed in São Paulo.3 Quantitative approaches have been used to test hypotheses that changes in Brazil’s age pyramid may have led to a decline in homicide rates (MELLO; SCHNEIDER, 2007; NERY et al., 2019), as well as to analyse other factors, such as falling unemployment rates (PERES et al., 2011), the presence of direct state interventions such as the Lei Seca (legislation on drink-driving: JUSTUS et al., 2018), the disarmament campaign (DREYFUS et al., 2008) and rising incarceration rates (KAHN, 2013).
There are also qualitative studies, particularly ethnographic research, arguing that the reduction in homicide rates in the state of São Paulo, especially among young people and adults occupying lower-level positions in illegal markets and living in peripheral neighbourhoods, is due to the emergence and stabilisation of mechanisms for regulating lethal violence by the PCC criminal faction (FELTRAN, 2010, 2012, 2018; MANSO, 2019; MANSO; DIAS, 2018; RUOTTI et al., 2017). This faction, through its own ways of instrumentalising security policies, is said to have strengthened its position as a conflict-regulating authority in the “crime world” through debates - a mechanism specific to the faction that operates as a space for dialogue in which opposing parties defend their arguments and positions, asserting what they consider to be certo (right) and errado (wrong). As has been observed with the PCC in São Paulo and elsewhere in Brazil, we have argued that political disputes over illegal markets, waged by armed groups, directly affect homicide indicators (FELTRAN et al., 2022; MARQUES, 2023).
In this sense, a deductive, long-term approach, like the theory developed by Gerring and Knutsen (2022), can be productive when articulated with a short-term inductive approach that takes into account the local urban-order context, helping us shed light on other dimensions of the phenomenon under study. By investigating the contexts in which violent deaths occurred, we composed qualitative synopses of the histories of lethal violence and decomposed aggregate homicide rates by victim profiles. Qualitative investigation makes it possible to map factors such as the expansion of crime networks and the incorporation of actors into transnational illegal markets, which become legitimate actors in the management of local urban order, producing rapid, locally situated variations.
In Brazil, the territorial, temporal and victim profile concentration of homicides makes evident the need for theories that identify causal mechanisms on the basis of these variations. Quantitative analysis enables us to identify trends, detect aggregate patterns and build inferences from historical homicide series. Empirical fieldwork makes it possible to comprehend the contexts and trajectories that make lethal violence possible, plausible, often expected and ultimately routinised, as well as the meanings attributed to this violence, which are produced and transformed locally. Taken together, these approaches enhance our analytical capacity to understand homicide, generating hypotheses that bring together long-term factors and situational analyses.
Methodological procedures
We analyse a two-decade period, from 2000 to 2020, combining ethnographic techniques with the simple disaggregation of homicide rates. First, we characterised the contexts in which homicides occurred in each of the three cities in the Triângulo Mineiro included in the study: Uberlândia, Uberaba and Cidade da Saúde. These cities were selected primarily because of the variation in homicide rates, which initially appeared chaotic and irregular, as well as rumours of a factional presence in Uberlândia. At the outset, we were unaware of the factional dynamics in the other two cities and, moreover, we were interested in comparing three different cases within the same region, an approach absent from the canonical debate on homicide in Brazil. On the strength of this comparison, and in dialogue with the ethnographic and quantitative literatures on homicide, we believe it is possible to advance a theoretical debate on variations in these crimes in Brazil. The comparison also brings together cities of different sizes and degrees of regional influence, thereby enhancing the explanatory power of the mechanisms identified in the study.
Cidade da Saúde, the fictitious name given to the smallest city, has approximately 100,000 inhabitants and is located near the border with Goiás state; Uberaba, a medium-sized city, has approximately 360,000 inhabitants and is located near the border with the state of São Paulo; and finally, Uberlândia, the largest city in the study and the second most populous municipality in the state of Minas Gerais, has approximately 725,000 inhabitants. In addition to possessing different geographical characteristics, the three cities have very distinct criminal dynamics, as we were able to observe through our research.
To this end, through intensive observation of violent contexts and an analysis of rumours and media coverage of homicides in each city, dozens of life-trajectory interviews were conducted with residents of peripheral neighbourhoods in these locations, family members of prisoners and/or former prisoners, and people involved in the criminal world at the time of the research or previously (such as former gang members and members of the PCC).4 These were our principal interlocutors, particularly in the Turquesa neighbourhood in Uberaba, Moji in Cidade da Saúde and Tatuí in Uberlândia,5 over a period of four months in each city.
The combination of two factors - a history of armed conflict and the feasibility of conducting fieldwork - led to the selection of one neighbourhood in each municipality. Communication with interlocutors was established through connections made while accompanying support groups for family members and friends of prisoners in the prison system; through facilitating film discussions organised by the Social Inclusion Programme for Former Prisoners (PrEsp); and through contacts established prior to the research. During and after the research phase, findings were presented to some interlocutors from different neighbourhoods and institutions in the cities, as a technique for data triangulation and validation. Interviews with a Civil Police chief and investigators in two of the cities studied also contributed to a better understanding of the phenomenon, although their material is not presented here. The interviews and observations were recorded in field diaries after each meeting. Where possible, conversations were recorded and subsequently transcribed.
Still on the question of the qualitative analysis, we also used a range of formal and informal documents, including news reports and court records, as well as a letter written by members of the PCC. The letter was recovered during a Civil Police operation at one of the prisons in Uberlândia and subsequently accessed by the researchers as a document from the case proceedings. This letter was transcribed and, through triangulation with sources indicating that it contained a plausible version of the local history, we reconstructed a timeline to facilitate a clearer understanding of its narrative. The qualitative synopses of the criminal dynamics in each municipality were then developed from the body of observations, interviews and documents collected. The conflicts, motivations and negotiations that led to violent deaths, and the effects generated by their regulation, were systematically organised and described in detail.
Only after reconstructing these synopses did we disaggregate homicide rates in the municipalities, region and state under study and cross-referenced them with the local chronologies of urban conflict in each municipality. To ensure consistency of the data, this study used only quantitative data generated by the Mortality Information System (SIM) of DataSUS. The data generated by SIM/DataSUS were obtained using the microdatasus6 package in R software, version 4.3.1. In this study, homicide data comprise the broad groups X85-Y09 (deaths resulting from assaults) and Y35-Y36 (deaths resulting from legal interventions or operations of war) in the International Classification of Diseases (ICD-10). For demographic data, we used table 2093 of the Brazilian Institute of Geography and Statistics (IBGE).7
We considered the data from SIM/DataSUS to be the most reliable available at the time of the research for two reasons: the data is processed in accordance with the World Health Organization’s ICD-10 classification; and the homicide data include intentional deaths recorded on death certificates issued by medics who certified deaths resulting from assaults by third parties. In addition, this data allowed for disaggregation by sociodemographic profile, something not possible with earlier data, which lacked reliable coverage and consistency and, for the municipalities studied, often contained no information on violent deaths by age, colour/race and sex. This limitation also explains our decision to begin the quantitative analysis in the 2000s, even though the memories of those interviewed extend back to the 1990s.
Although only homicide data were analysed in this article, we do not overlook the fact that underreporting of homicide rates remains a widespread phenomenon in Brazil, particularly because of inconsistencies in the details recorded of disappearances. In addition, important studies have observed that one of the most common ways of disguising these figures is to use other categories (such as disappearances or deaths from undetermined causes) to identify cases of violent death. Suspicion further increases when the curves for different categories suddenly follow opposite trajectories while remaining at proportionate levels (CANO; DUARTE, 2012; MUSUMECI, 2020).
After a series of tests taking into account the findings from the literature on the predominant profile of homicide victims (FELTRAN et al., 2022; HIRATA, 2018; MISSE, 2006; RATTON; DAUDELIN, 2018) and from our own collective research, as well as assessing the quality of the available data, we were able to disaggregate homicide rates into three profiles: (i) “general population”; (ii) “young black men,”8 that is, males aged 15 to 29 who self-identified as black (preto) or brown (pardo); and (iii) “population excluding young black men.” Contrasting qualitative situations with quantitative data was essential to understanding the criminal dynamics and armed regulation in the municipalities under study. This was the path we followed to make quantitative inferences from historical series of homicides, both aggregating and disaggregating the data according to sociodemographic profile.
Homicide prevention policies in the Triângulo Mineiro
Since 2003, following the reorganisation of its public security policy and the creation of the Social Crime Prevention Policy, the state of Minas Gerais has operated Fica Vivo! (Stay Alive!)- a public policy for homicide prevention structured through the implementation of Crime Prevention Units (UPCs) in selected municipalities.9 The programme is recognised nationally and internationally as an important initiative for homicide control (BEATO FILHO et al., 2017), and has generated a substantial academic literature evaluating its outcomes, particularly in urban clusters in the Metropolitan Region of Belo Horizonte (SILVEIRA; SILVA; BEATO FILHO, 2008; SILVEIRA, 2012).
Our research did not seek to assess the impact of this policy in the Triângulo Mineiro, nor did our interlocutors make direct reference to it. However, given the theme of this article and the centrality of homicide as a public policy issue, we believe it important to engage with the literature on evidence-based preventive public policies. Among the municipalities studied, Uberlândia is the only city with both municipal UPCs, implementing the Central Programme for Monitoring Alternative Penal Measures (CEAPA) and PrEsp, and territorial UPCs, implementing Fica Vivo! and Conflict Mediation, since 2005, with distinct implementation timelines. The Specialised Policing Group in Areas of Risk (Gepar) was implemented rapidly in 2005, in contrast to the much later introduction of the Conflict Mediation programme in 2012. The literature reports that this resulted in the predominance of “qualified repression” over “social protection,” generating policing in “areas of risk” that was disconnected from the preventive methodology (BONESSO, 2020) and accelerating the incarceration of young people living in peripheral neighbourhoods, particularly black men (JESUS, 2023).10
The public policy reached Uberaba in 2008 and the city currently has only one municipal UPC, running both the CEAPA and PrEsp programmes. From 2009 to 2014, the municipality operated the Fica Vivo! and Conflict Mediation programmes, but these were discontinued when funding for their implementation in the municipality was suspended. Around 400 young people had been attended by these programmes. Finally, in Cidade da Saúde, which for several years was among the most violent cities in the region, the public policy was introduced only in 2014, with the implementation of a municipal UPC that operates only the CEAPA programme.
The importance of developing public policies aimed at homicide prevention - like the Fica Vivo! programme or Pernambuco’s Pacto pela Vida (Pact for Life), subsequently replicated in other states - seems evident to us. Institutional programmes such as Fica Vivo! may have had some impact on the variations in violent-death rates in the region, particularly in Uberlândia, the only municipality among those studied to have a territorially based intervention, while the other programmes focus on monitoring alternative penal measures . In Uberlândia, which has had the “complete” public policy in place since 2012, the coexistence of two forms of interventions aimed at young people in peripheral neighbourhoods is notable: state mechanisms for conflict mediation and mechanisms operating within the “crime world” (BONESSO, 2018), the combination of which may also have contributed to the decline in violent deaths during some of the periods analysed. A different research protocol would be required to investigate these possible effects systematically.
In our qualitative research, however, these impacts did not appear to be directly related to the rapid fluctuations in homicide rates, which occurred in divergent directions in each city. Other factors proved to be more directly implicated in these variations, as we shall see. In short, the results of Fica Vivo! in the Triângulo Mineiro do not appear to be comparable to the effects identified in the literature for the Metropolitan Region of Belo Horizonte. Our data point to an explanation of local variations in aggregate homicide rates that centres primarily on sociological mechanisms, which, as we discuss below, could also be considered evidence on which to base preventive public policies or could be incorporated into them.
Power disputes and fluctuations in regional homicide rates
The Triângulo Mineiro is fundamental to the national and international flow of goods, functioning as a major transport and commercial hub connecting the South/Southeast of Brazil with all the other regions of the country. The study of three disparate contexts in the region contributed to an understanding of the heterogeneity of lethal violence in an area that contains no less than 56 per cent of the population of Minas Gerais, a state that accounts for 10 per cent of Brazil’s population. The Triângulo Mineiro is a key logistical node in the country, connecting Brazil’s largest and most dynamic metropolis with the national capital, as well as bordering Mato Grosso do Sul to the west (a state bordering Paraguay and Bolivia), the state capital Belo Horizonte to the east and Goiás state to the north, “all orbiting within an approximate radius of just 500 km” (JÚNIOR; FERRAÇO, 2019, p. 29). Moreover, the road connection between Brazil’s Southeast and Northeast and North is much more efficient via the Triângulo Mineiro than along Brazil’s coast.
Our research sources demonstrate that the PCC expanded into the Triângulo Mineiro as early as the start of the 2000s, much earlier than previously thought, therefore, and that this expansion was already being driven by the São Paulo government’s strategy of transferring prisoners identified as faction leaders to prisons in other states (FELTRAN, 2018; MANSO; DIAS, 2018). In 2012, the Minas Gerais government disputed a report released by the Federal Government indicating the faction’s presence in the state, which was only officially acknowledged in 2014, when the Public Prosecutor’s Office of Minas Gerais charged 41 prisoners at the Uberaba penitentiary with participating in the PCC.11
From 2017 onwards, the faction’s presence became undeniable to public officials in Minas Gerais due to a change in the type of criminal activity in which it was involved: there was an expansion of the illegalised drug market, robberies of cash machines and money laundering through the sale of car parts and vehicles (DUARTE; ARAÚJO, 2020). But, given the limited scale of its activities up to that point, the faction’s operations in the state amounted to little more than sensationalist speculation (DUARTE; ARAÚJO, 2021). In June 2018, the faction’s activities in Minas Gerais gained increased media attention following a series of attacks on buses, police bases and public buildings in an offensive that lasted five days, with 101 coordinated attacks in more than 30 cities across Minas Gerais, concentrated in the Triângulo Mineiro, Alto Paranaíba and southern Minas regions. Among the documents we collected in this research, we highlight the aforementioned letter from PCC members, which states that “the prisons are a powder keg,” claiming that the “oppression suffered in the prison units” was the reason for the offensive.12
Given that the effects of this expansion are multiple and unique to each locality,13 as a result of their interaction with their own criminal and urban histories, our interest lies in understanding how this expansion manifested itself in homicide rates in the Triângulo. To do so, we test the hypothesis developed by Feltran et al. (2022), which considers transformations in the local “crime world,” such as the transition to factional conflict, to be a central factor in explaining variations in homicide rates, both aggregated and disaggregated. Figure 1 presents the homicide rates of three cities in the Triângulo Mineiro region, demonstrating this rapid fluctuation in the curves.
Homicide rates per 100,000 inhabitants in Uberlândia, Uberaba and Cidade da Saúde, by year (2000-2020)
The rates suggest a long and sustained period of growth in homicides, followed by a downward trend over the past decade, at different times in each city. From 2012 onwards, the rates suggest a sharp and sustained decline in Uberlândia, while they increased in Cidade da Saúde and, in 2015, were rising sharply in Uberaba. In Uberlândia, the city in which the rates began to fall first, from 2012 onwards, the downward trend remained stable through to 2020, with the same later occurring in Uberaba and Cidade da Saúde. Qualitative research and analysis of the disaggregated historical series help us understand these fluctuations. Figure 2 presents the homicide curves for each municipality studied, disaggregated by sociodemographic profile, from 2000 to 2020.
Homicide rates per 100,000 inhabitants in Uberlândia, Uberaba and Cidade da Saúde, by sociodemographic profile (2000-2020)
The discrepancy in homicide rates between the three sociodemographic groups analysed in the figures is clear. Homicide rates among young black men are much higher and display more rapid and intense fluctuations. By contrast, rates among the general population and the population excluding young black men remain relatively stable, with more limited variation. Even so, over the past two decades, all three cities have experienced considerable fluctuations in homicide rates, both upwards and downwards. To understand these rapid fluctuations, the concentration of homicides among young black men and the impact of power struggles between armed groups, we shall now examine the histories of conflict in each municipality in detail.
The 2000s
In 1999, the PCC began its expansion into Uberlândia through the presence of “brothers,” the name given to the baptised members of the faction transferred to one of the city’s two penitentiaries, as revealed in a letter found in a wing of the Professor João Pimenta da Veiga Penitentiary. At that time, the city’s retail drug trade was organised around territorial boundaries fragmented into bocas (literally “mouth”: a drug salespoint) and their respective owners. As a result of this organisation, certain streets or neighbourhoods came to be regarded by illegalised drug-trade networks14 as belonging to the owner of a boca, preventing third parties from conducting commercial activities there, under threat of violent punishment. Territorial control was thus fragmented and governed by logics dependent on the individual decisions of the owners.
The arrival of new criminal actors in Uberlândia generated severe tensions among these local criminal actors, but also resulted in strong alliances. As early as the first years of the 2000s, around 2004, these new actors became immersed in a violent dispute over the city’s illegal markets, in what our research identifies as a period of tomadas de bocas: killing the former owners of bocas or brecando (blocking) their activities,15 they took possession of their points of sale, under threat of punishment for failure to comply with the order. These points of sale now tended to operate without weapons, as the aim was to monopolise the retail drug trade. It is important to note that this did not involve the armed and overt seizure of territories within neighbourhoods, settlements or occupations, but rather the use of violence to establish control over key points in the value chain. Controlling the distribution chain would consequently represent a further step in the attempt to monopolise the markets.
According to the accounts obtained in our research, the tomadas de bocas began with three drug traffickers from Ribeirão Preto who moved to Uberlândia in the early 2000s, claiming to be “brothers”16 of the São Paulo faction PCC. In the following years, the tomadas de bocas intensified in different parts of the city, some of which were successful. As Figure 3 suggests, this political dispute over illegal markets, pursued violently, affected young black men in particular. The homicide rate among this group rose by 97% in the space of just three years, from 2005 to 2008, rising from 45.28 to 89.21 per 100,000 inhabitants. During the same period, the rates suggest that homicides among the general population and the population excluding young black men remained at constant levels. Among the total population, the rates rose from 17.37 per 100,000 inhabitants in 2005 to 17.48 per 100,000 inhabitants in 2008; among the population excluding young black men, they fell from 15.54 to 12.61 between 2005 and 2008.
Homicide rates per 100,000 inhabitants in Uberlândia, by sociodemographic profile (2000-2010)
In Uberaba, homicide rates over the period show little variation among the general population, with a sharp decline in 2001 and a substantial increase in 2007, followed by a downward trend. For the rates by sociodemographic profile, however, those for young black men are markedly higher in these years. Moreover, throughout the first decade of the 2000s, the figures suggest substantial variation in lethal victimisation, particularly among young black men.
The criminal dynamics of this period in Uberaba provide an insight into the marked variation in the municipality, which is also concentrated among the same sociodemographic profile. The first decade of the 2000s was marked by conflicts aimed at monopolising the illegalised drug trade and by rivalries that, like the bocas, were organised territorially. Until 2006, these conflicts were driven by people involved in crime from two neighbourhoods: Independência, one of the city’s largest and most historic neighbourhoods, located near the city centre; and Dores, a neighbourhood situated on the far eastern periphery.17 The clash between the two areas marked the popular imagination, whether because of the influence of Independência or because of the years when news reports of violent deaths occurring in these neighbourhoods were prominent.
In 2006, a neighbourhood on the southern periphery of Uberaba, Floresta, joined the violent disputes between the two neighbourhoods, exacerbating the conflict, subsequently described in local newspapers as a “war.” This immediately led to a 135% increase in the homicide rate among young black men, from 61.81 per 100,000 inhabitants in 2006 to 145.75 in 2007, as shown in Figure 4. Homicide rates among the general population also increased, from 17 to 25.45 per 100,000 inhabitants between 2006 and 2007, representing a 49.7% rise in lethal violence over the period.
Despite the sharp increase in lethal violence, the period of war did not last long. Soon afterwards, Floresta allied itself with Independência, although we do not know whether this was for purely commercial reasons or because the two shared a rivalry with Dores. In 2008, in the wake of this alliance, there was a decline in lethal violence in the city, with a 49% reduction in homicides among the general population, from 25.45 to 12.88 per 100,000 inhabitants between 2007 and 2008. In the following years, the homicide rates suggest that there was little fluctuation in lethal violence, as shown in Figure 4. Residents recall that, during this period, many gerentes de boca were killed or imprisoned, which they perceived as contributing to a diminution of the war. Subsequently, this weakened conflict combined with the predominance of the alliance between the Floresta and Independência neighbourhoods.
In the early 2000s, Cidade da Saúde experienced a transition in the criminal order, moving from a period in which gang groups with violent rivalries predominated to a more fragmented organisation of criminal dynamics, without gang ties and oriented towards the illegalised drug trade. The criminal order that was marked by gangs in the 1990s was composed of groups of up to 20 people, formed around the neighbourhoods where they were born and lived. At the time, the gangs related to one another through violent rivalries, which took the form of physical fights or confrontations involving objects like chains, stones and knives. These rivalries came to structure personal and group relationships, entangled in networks of personal and family conflicts and in the very consolidation of the neighbourhoods. Sometimes these fights generated expanding cycles of vendettas,18 established among rivals, friends and family relatives of rivals.
Even so, the “gang fights” helped consolidate rivalries that later intensified with the expansion of illegal drug and arms markets in the interior of Minas Gerais state. This change in the criminal order reverberated in how conflicts were managed and in the motivations for lethal conflicts. In this second phase, with rivalries between gangs as their antecedent, management was taken over by the gerentes de boca and was based on killing as punishment. In these situations, people regarded as rivals or as having committed serious transgressions - such as members who caused losses to collective illicit businesses, informers, indebted customers, talaricos19 or those who committed other serious offences, whether or not involving questions specific to the criminal world - were subject to lethal punishment.
In this way, the gerentes de bocas were able to impose their order through their capacity to determine who lived and who died. The respectability of the power exercised by the gerentes in the neighbourhoods follows from the protection of the community that characterised the gang era in the 1990s, when they prohibited robberies and punished those who committed them. In the 2000s, in Cidade da Saúde, the rates suggest substantial variation in homicides, both upwards and downwards. For the period from 2001 to 2006, the DataSUS database does not provide the number of homicide deaths among young black men, with the exception of 2003 and 2004 (Figure 5). This does not mean that no homicides occurred among this group, but rather that such cases may have been overlooked and that there may also have been more underreporting during these years. Even so, the empirical context of the 1990s and 2000s presented here certainly includes this sociodemographic profile.
Homicide rates per 100,000 inhabitants in Cidade da Saúde, by sociodemographic profile (2000-2010)
As we have seen, the homicide rates in the three Triângulo Mineiro cities studied here suggest markedly different patterns of lethal victimisation across time and space, which, nevertheless, show an overwhelming concentration among young black men. More than this, the qualitative research demonstrates that this was not a case of higher victimisation among this profile in general, but rather of individuals fitting this profile who occupied the lower echelons of the operation of illegal markets, particularly those involving drugs and robbery.
We turn now to examine the empirical context and homicide rates of the three cities in the following decade, analysing the fluctuations in the homicide curves and the histories of the crime world during this period relationally.
From 2010 to 2020, much changed
Following the tomadas de bocas, the expansion of the PCC in Uberlândia was marked by a violent struggle over the illegalised drug markets. From the beginning of the decade until 2014, the homicide curve in Uberlândia showed upward fluctuations, suggesting a trend towards increasing violent conflict. By this point, the three brothers from the interior of São Paulo competing for the city’s illegalised drug markets, claiming to be PCC brothers, had built alliances and rivalries around the tomadas de bocas, which were subject to constant negotiation (BARBOSA, 2005), escalating the conflict that had begun on a localised basis.
This political dispute for drug markets, that is, for monopoly over bocas - which is claimed when a boca is taken, through the use of the faction’s name for their own benefit - is a particular characteristic of the faction in the region under study. The tensions generated by the faction’s expansion into Uberlândia resulted, during this first decade, in more violent conflicts than “pacification.” These characteristics reveal not only the actions of local actors in the criminal world, but also the inherent plasticity of the faction’s organisation (BIONDI, 2010; FELTRAN, 2011, 2018).
Until 2014, violent disputes over the territories in which the bocas operated were still frequent. The letter cited earlier describes the conflicts surrounding illegalised drug markets, followed by a stabilisation of the bocas - that is, a ceasefire between gerentes. This was made possible by the division of territories, a relationship described in the document, and by the setting of standard sale prices for drugs, as observed during fieldwork. This shift displaced violent conflicts away from business-centred disputes, contributing to a reduction in deaths in these contexts, even though other factors determining the lethal resolution of conflicts persist.
With the establishment of the PCC as hegemonic in the city’s illegal markets, and through its presence in the prison system, the ceasefire was reinforced in subsequent years by the implementation of debates,20 the faction’s mechanism for conflict resolution. Thus, each situation brought before the Disciplina, the brother responsible for the faction in each neighbourhood, is judged together with other brothers and those involved, who hear the parties and arbitrate a fair punishment on a case-by-case basis, aiming to repair the harm suffered by the parties involved while avoiding death as a form of punishment wherever possible (BIONDI, 2018; FELTRAN, 2010; MARQUES, 2009; MANSO; DIAS, 2018).
Already established as an ordering force in the Uberlândia crime world, the Comando was also present in other cities in Minas Gerais, as demonstrated by the PCC attacks in Minas Gerais in 2018. While the attacks were taking place, lethal violence in Uberlândia was declining.21 In 2018, the rates suggest a sustained reduction in homicides, which had been declining since 2014. Among the general population, the reduction was 56.9%, from 25.76 homicides per 100,000 inhabitants in 2014 to 11.08 in 2018. Figure 6 shows that the rates suggest a consistent reduction in homicides, beginning in 2014 and continuing through to the most recent period analysed. Over this period, rates fell by 73% among the general population, dropping from 25.76 homicides per 100,000 inhabitants in 2014 to 6.95 in 2021. The decline is even more evident when we consider the profile of young black men: from 2014 to 2021, the rates suggest an 84.5% reduction in lethal victimisation among this profile, from 120.76 homicides per 100,000 inhabitants in 2014 to 18.61 in 2021.
Homicide rates per 100,000 inhabitants in Uberlândia, by sociodemographic profile (2011-2021)
In Uberaba, meanwhile, the rates at the start of the 2010s suggest that homicides were beginning to increase gradually. The decline that began in 2008, following an intense conflict between three neighbourhoods, went into reverse from 2012 onwards, with some fluctuation. At this point, the PCC was making its first moves to expand onto the city’s streets, favoured by its proximity to Uberlândia. But the presence of PCC brothers in Uberaba’s prison system dates back further in time to the mid-2000s, when prisoners were transferred to the city from the interior of São Paulo. According to fieldwork accounts, Dores, a neighbourhood located close to the Uberaba penitentiary - separated from the latter by a stretch of highway - was already considered one of the first neighbourhoods in the city to “be PCC,” that is, a neighbourhood in which the faction had established itself in the illegalised drug market.
The faction’s expansion onto the streets of Uberaba generated suspicion among some members of the local crime world because, although embedded in the dynamics of the illegalised drug trade - supplied by the PCC through shipments passing via Uberlândia and along the strategic supply routes for goods that run through the Triângulo - they claimed not to want to follow the discipline of the Comando. This resistance generated localised conflicts surrounding the drug markets, as the commercial expansion of the brothers did not recede.
The homicide rates for the period clearly demonstrate that these conflicts primarily affected young black men. From 2013 to 2014, the rates indicate a sudden increase in homicides, whose rapid variation suggests a temporary disruption in the dynamics of conflicts over the drug markets. Among young black men, there was an increase of 155%, rising from 43.71 homicides per 100,000 inhabitants in 2013 to 111.49 in 2014. This period of conflict lasted several years. The rates suggest a sustained period of elevated levels. Among the general population, between 2013 and 2015, there was a 197% increase in lethal victimisation in the municipality, from 9.99 homicides per 100,000 inhabitants in 2013 to 29.67 in 2015, the year when the series reached its highest value for the general population.
Through the commercial backing provided by brothers from Uberlândia and its activities in the Uberaba penitentiary, the faction gained strength, achieving political hegemony (GRAMSCI, 1987) marked by the active consent of a portion of the population and by the coercion directed against any systematic opposition. As the conflict between members of the local criminal world and PCC brothers diminished from 2015 onwards, there was a decline in killings resulting from these conflicts, although it would take time for the faction’s role as a regulator of lethal violence in peripheral neighbourhoods to become consolidated. Homicide rates began to fall from 2015 onwards, as the PCC maintained its presence in Uberaba’s illegalised drug market. This decline was even more pronounced among young black men, falling by 74% over three years, from 108.42 homicides per 100,000 inhabitants in 2015 to 28.61 in 2018 (Figure 7). The decline that began in 2015 has been consistent, with rates remaining at low levels to this day.
The case of Cidade da Saúde is very particular among the municipalities studied. Containing approximately 100,000 inhabitants, the city experienced unstable criminal dynamics over short periods during the past decade, as reflected in the substantial variations in the homicide rates in the historical series analysed (Figure 8). The expansion of the PCC into neighbouring cities in the Triângulo Mineiro, such as Uberlândia and Uberaba, the cases described above, also contributed in singular ways to the increase in lethal conflicts in Cidade da Saúde. As the faction became established in the region’s largest city, Uberlândia, narratives suggest that the group sent firearms, drugs and, occasionally, personnel to smaller cities such as Cidade da Saúde. This helped reactivate earlier disputes and intensify new conflicts within the illegalised drug market, which was becoming increasingly specialised.22
Homicide rates per 100,000 inhabitants in Cidade da Saúde, by sociodemographic profile (2011-2021)
By the mid-2010s, accounts began to circulate in the city that a brother from Uberlândia had been sent to Cidade da Saúde with the aim of “organising crime.” He was regarded by brothers from Uberlândia as the municipality’s Disciplina. While the city was already well connected commercially with the brothers from Uberlândia, the aim at this moment was for the two cities to be in syntony with one another. Nevertheless, violent deaths continued to be part of everyday life in the city, reported weekly in local newspapers, on social media and in WhatsApp groups, including deaths of people who self-identified as “cousins” or “brothers.”
It was only towards the end of the decade, in 2018, that lethal violence began to decline, following a commercial agreement between brothers and the local criminal world. By 2022, the limit of our fieldwork data, it was being said on the streets of the city that there were demands for payment of “overdue ideas.” The rates suggest a corresponding reduction in lethal conflicts within the criminal world: from 2018 to 2021, homicide rates among the general population fell by 59.3% from 25.18 to 10.24 per 100,000 inhabitants. Cidade da Saúde experienced substantial variations throughout the historical series analysed, but particularly during the 2010s.
During this decade, data on the racial and age profiles of homicide victims in the municipality became available. The rates indicate that the predominant profile follows the pattern observed in the other cities: young men aged 15 to 29 who self-identify as black or brown. Homicide rates among young black men underwent rapid and substantial fluctuations, with abrupt declines in 2013, 2016, 2018 and 2021, more pronounced than those among the general population and the population excluding this profile. We do not discount the underreporting of data (Musumeci, 2020), particularly in Cidade da Saúde, given the absence of data on racial profiles in several years during the 2000s.
This makes it clear that the downward trend we observed in all three cities towards the end of the 2010s closely mirrors the criminal dynamics encountered during the qualitative research, conducted with interlocutors who had no access to the homicide curves or trends in their variation. It seems clear to us that the sharp rises and falls in aggregate rates are directly driven by variations in the profile of young black men (and, as we know qualitatively, this profile is in fact much more specific, referring to a very small minority of this population, directly involved in illegal markets). More than this, it is clear that the homicide series reflect dynamics internal to the criminal world, which are constantly being negotiated and transformed. The most recent situation, which we observed towards the end of the decade in all three cities, indicates that the PCC’s hegemony has become consolidated across the three municipalities and that a pax monopolista23 has been established, stabilising the homicide rates at low levels, although it remains subject to continual and fragile negotiation.
Analytical time
Based on a qualitative analysis of the criminal order in each city over two decades, together with quantitative inferences based on homicide rates, we developed an analytical model that incorporates the dynamics of criminal conflict as a temporal marking of variations in lethal violence. Thus, this temporal marking does not refer to chronological time and its dates, but rather to another chronology, that of analytical temporality. Along the timeline organised according to the sociological regularity of moments of transition in the criminal order, with their corresponding curves of lethal violence, we consider these moments to constitute analytical times.
As we have seen from the qualitative synopses of the cities studied, the empirical context shows different regimes of criminal order across time and space, organising lethal violence. Our argument is that these regimes produce different arrangements for regulating order in the criminal world and follow a sociologically regular temporal sequence, in which the temporal marking is the process through which the PCC arrives, becomes established and consolidates its hegemony in each territory. On one hand, therefore, we have a chronological temporal arrangement, in which events occur on specific dates, for example, in the 1990s or in 2013, in each city. On the other hand, we have an analytical temporal arrangement, which considers the sequence of events involved in the consolidation of the PCC as the fundamental temporal marker for analysis. In this temporal arrangement, we no longer have dates as markers, but rather analytical processes, such as the periods before, during and after the consolidation of the PCC. Viewed through dates, we see cities in the same region experiencing distinct situations in terms of variations in their homicide rates; viewed through the proposed analytical temporal arrangement, however, these variations become readily intelligible as regular manifestations of a sociological pattern.
We developed a visualisation showing these similar criminal processes in each city as they occurred at distinct dates. This visualisation highlights how specific criminal events and dynamics influence homicide rates in different cities, allowing for a comparative analysis of patterns of lethal violence. In this way, organising the data according to analytical time rather than chronological time facilitates our understanding of the shift towards factional conflict in the case studied. Figure 9 shows the homicide rates when ordered according to analytical time.
Homicide rates per 100,000 inhabitants in Uberlândia, Uberaba and Cidade da Saúde, by analytical time
Figure 9 presents homicide rates at the moments marking the transition in the criminal order, occurring in different years in each of the municipalities under study. This visualisation relates homicide rates to the dynamics of conflict and pacification, examining the responsiveness of lethal-violence indicators at each specific moment of transition in the criminal order. In the first analytical time, “before the PCC,” the rates correspond to the period before the Comando had begun its expansion onto the streets of the cities, even though it had already reached the prisons, because, in all cases, until its expansion onto the streets, the prevailing criminal order was that of the local criminal world. The rates correspond to the years 2000 to 2003 in Uberlândia, 2002 to 2005 in Uberaba and 2003 to 2006 in Cidade da Saúde. We can note that the rates follow very similar patterns, particularly in Uberaba and Uberlândia, with a curve suggesting an increase in lethal violence, but with greater variation in Cidade da Saúde, as we observed in the synopses. In this first analytical time, rates in Uberaba display considerable variation: 192%, rising from 5.74 per 100,000 inhabitants in 2002 to 16.77 per 100,000 in 2005, as a result of rivalries surrounding the illegalised drug trade, as we saw in the municipality’s qualitative synopsis. The variation in rates in Uberlândia and Cidade da Saúde is smaller: 51.1% in Uberlândia, rising from 10.57 per 100,000 inhabitants in 2000 to 15.97 per 100,000 in 2003, and 17.4% in Cidade da Saúde, rising from 14.37 per 100,000 inhabitants in 2003 to 16.87 per 100,000 in 2006.
The second analytical time, by contrast, refers to the period during which there was an active attempt to establish the PCC in peripheral urban areas, especially through disputes over illegalised drug markets in the studied municipalities. Homicide rates increased considerably, especially as an outcome of lethal conflicts arising from political struggles over these markets. In Uberlândia, rates rose by 22.2% during this period, from 25.26 per 100,000 inhabitants in 2009 to 30.85 per 100,000 in 2012. In Uberaba, from 2012 to 2015, rates suggest an upward variation of 69.3%, from 17.49 per 100,000 inhabitants in 2012 to 29.67 per 100,000 in 2015. In Cidade da Saúde, rates during the period corresponding to the PCC’s advance into the markets, despite some fluctuations, also suggest an increase: 26.3%, from 18.89 per 100,000 inhabitants in 2012 to 23.86 per 100,000 in 2015.
In the third analytical time, we observe the most advanced period reached by our investigation, the moment when the PCC has become established and possessed hegemonic control over the illegal markets, particularly the illegalised drug markets, in the peripheral regions of the cities studied. The consolidation of the faction in these municipalities resulted in a reduction in killings regarded as commonplace, as well as in other forms of violence in peripheral urban areas and prison units. This process is narrated as following a long period of conflict, as we observed in the qualitative synopses and in the other analytical times, and is represented quantitatively by the progressive decline in homicide rates. For this analytical time, which is the most recent, we consider the years following the transition to the factional order in the municipalities under study. In Uberlândia, there was a 47.8% decline in the rates corresponding to this third analytical time. From 2015 to 2018, rates fell from 21.24 to 11.08 per 100,000 inhabitants. In Uberaba, the corresponding period is from 2016 to 2019. Over four years, there was a 49.4% reduction in lethal victimisation, from 20.51 homicides per 100,000 inhabitants in 2016 to 10.38 in 2019. In Cidade da Saúde, the period corresponds to the final years analysed in this article, the most recent period: from 2018 to 2021. In the city, the indicators suggest an even greater decline, of 59.3%, with rates falling from 25.18 per 100,000 inhabitants in 2018 to 10.24 in 2021.
This analytical temporality of the transition from one criminal order to another is our interpretive hypothesis for variations in homicide rates in the Triângulo Mineiro. It is evident that this succession of analytical times - the “before”, “during” and “after the PCC” - refers to the analytical processes specific to the case of the Triângulo Mineiro and cannot be generalised to other localities. Precisely for this reason, because it derives from situational analyses, we suggest that analytical times also be considered in other settings when constructing comparisons of homicide rates. Sociological processes need to be related to dates. The temporality of the transition from one criminal order to another, as in the case studied here, would provide a much better explanation of trends in homicide variation throughout the country.
In this sense, the empirical elements marking the transition from one criminal order to another in different cities and states - such as the arrival of increasingly powerful firearms to factions, changes in illegal drug markets in the region since the late 1990s, the expansion of factions such as the PCC and Comando Vermelho into disparate criminal contexts and their struggles over illegal markets - although they did not occur simultaneously, demonstrate a process of transition between forms of homicide regulation. Understanding the empirical context and drawing inferences from quantitative indicators form the basis of our analytical model, which proposes analytical temporal marking as a means of interpreting lethal violence through a qualitative-quantitative approach. Analytical temporal marking helps us to interpret variations in homicide rates, whereas a comparative view of the homicide rates of the three cities in chronological order merely leads to confusion.
Concluding remarks
Based on a study of criminal dynamics over the past two decades in three cities in the Triângulo Mineiro, we have demonstrated that the periods of the PCC’s arrival, subsequent fight for hegemony in the regulation of illegal markets, and subsequent establishment as the hegemonic faction in the region produced systematic variations in homicide rates among a specific population profile, which drives the variations observed in aggregate rates. This constitutes a demonstration case, now produced through a qualitative-quantitative analysis, of the relevance of considering the criminal order and its conflict dynamics in understanding variation in lethal-violence rates in Brazil.
In sum, the article tests the analytical hypothesis (FELTRAN et al., 2022) that factional conflict is a fundamental explanatory element for understanding variations in homicide rates in Brazil, without precluding explanations based on other supplementary causes. By studying the “before,” “during” and “after the PCC” in the Triângulo Mineiro, our findings corroborate theoretical hypotheses already developed in the specialised literature that factional dynamics directly impact homicide curves, both aggregated and disaggregated, particularly affecting young black or brown men living in peripheral neighbourhoods and recruited by illegal markets.
The analytical focus on investigating the causal mechanisms that shape the regulation of homicides in three disparate urban contexts in the interior of Minas Gerais enabled us to conceive of three analytical times, based on the transitions between criminal orders that structure a large part of the regulation of lethal violence. Working with a sociological timeline enabled us to combine multiple and distinct criminal narratives, describing them in detail so as to understand the development of the region’s criminal history in all its heterogeneity. In these cases, we sought to present histories and cases that captured elements of each period and municipality, with the aim of not obscuring the common elements described by the analytical times through the specificity of each instance.
In the case studied, the establishment of the PCC’s hegemony contributed to a strong downward trend in average homicide rates in the three localities towards the end of the period analysed, through the establishment of a “third mediator of conflicts” (ROCHA, 2019): the faction itself, through the mechanism of debates (BIONDI, 2018; FELTRAN, 2010, 2012). The PCC’s policies are said to have established a degree of predictability in everyday criminal life, thereby mitigating, to some extent, vendettas and the fatal outcome that otherwise remains on the horizon as a means of resolving everyday problems. This constitutes a normalisation of conduct and conflict resolution that does not exclude a violent outcome, but regulates and organises its implementation, resulting in a reduction in lethal violence, which had increased during the wars associated with the absorption of local gangs by organised criminal groups. Today, it is the organised actors responsible for perpetrating violence in the territories where they are hegemonic who are called upon when a serious conflict arises.
However, the actors involved in regulating the criminal order and, consequently, in regulating homicides vary over time, as we have seen in the case studied through our analysis of the three analytical times. In some of Brazil’s territories and states, police lethality is much higher than in the Triângulo Mineiro, although the region itself also exhibits extremely high levels that warrant detailed analysis. We understand the regulation of lethal violence to result from provisional arrangements between criminal and state actors, which are neither totalising nor definitive, but which have undergone regular processes of institutionalisation over recent decades in different neighbourhoods, cities, states and across the country.
Thus, PCC hegemony over illegal markets in peripheral neighbourhoods of the cities investigated here does not entail the production of peace through harmonious relations with other actors but, as Richmond (2021) demonstrates, much more the suppression of the possibility of conflict by those who have the capacity to engage in it. Hence, these arrangements between criminal actors (and state actors who operate extralegally) are undesirable from a republican perspective, institutionalising criminal orders that compete with the state for sovereignty over territories - initially in an unstable and provisional manner, and subsequently in a more stable form, as occurred in São Paulo. These are normative power regimes that coexist with the state and whose mechanisms for regulating violence could be seized by the state.
In this study, we found that the regulation of the criminal order in the cities we studied is carried out by actors who are constantly shifting, which does not prevent us from identifying logic and regularity in the trends they have produced. The actors who exercised regulatory authority over the criminal order in the analytical times examined here impose their authority and coexist with other actors who may also use their access to the means of violence to assert their control. This regulation directly and centrally impacts aggregate variations in homicide rates, even if it is discernible only in specific homicides that victimise young black and brown men, residents of peripheral neighbourhoods and operators in illegal markets. The war between these actors has been so intense in Brazil that, despite involving a very small minority of the population, its demographic traces are evident. Viewing other contexts with the notion of analytical time in mind, focusing on the dynamics through which criminal orders are instituted, would seem a promising avenue for understanding variations in homicide rates and their levels, by relating empirically grounded knowledge to inferences made from quantitative data.
Notes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Declaração de disponibilidade de dados
Os conjuntos de dados relacionados a este artigo estarão disponíveis mediante solicitação aos autores correspondentes.
References
- BARBOSA, Antônio Rafael. Prender e dar fuga. Biopolítica, Sistema Penitenciário e Tráfico de Drogas no Rio de Janeiro. Tese (Doutorado em Antropologia Social) - Universidade Federal do Rio de Janeiro, Rio de Janeiro, 2005.
- BEATO FILHO, Cláudio Chaves; RIBEIRO, Ludmila Mendonça Lopes; OLIVEIRA, Valéria Cristina de; PRADO, Sara Carla Faria. “Reducción de homicidios en Minas Gerais: Un análisis del programa «Fica Vivo!»”. Revista Cidob d’Afers Internacionals, Barcelona, vol. 0, n. 116, pp. 129-157, 2017.
- BIDERMAN, Ciro; DE MELLO, João M. P.; DE LIMA, Renato S.; SCHNEIDER, Alexandre. “Pax monopolista and crime: The case of the emergence of the Primeiro Comando da Capital in São Paulo”. Journal of Quantitative Criminology, vol. 35, pp. 573-605, 2019.
- BIONDI, Karina. Junto e misturado: uma etnografia do PCC. São Paulo: Terceiro Nome. 2010.
- BIONDI, Karina. Proibido roubar na quebrada: território, hierarquia e lei no PCC. São Paulo: Terceiro Nome, 2018. (Coleção Antropologia Hoje, 1 vol.).
- BONESSO, Márcio. Prevenção à criminalidade: arte e esporte na segurança pública de Minas Gerais. Curitiba: Appris, 2018.
- BONESSO, Márcio. “Economia das drogas e políticas de segurança no Triângulo Mineiro: o controle do crime entre mercados do atacado e do varejo”. Revista Brasileira de Segurança Pública, São Paulo, vol. 14, n. 2, pp. 82-99, 2020.
- CANDOTTI, Fabio; CUNHA, Flávia; SIQUEIRA, Ítalo. “A grande narrativa do Norte: considerações na fronteira entre crime e Estado”. In: MALLART, Fabio; SIQUEIRA, Ítalo. BR 111: as rotas das prisões brasileiras. São Paulo: Veneta, 2017, pp. 35-47.
- CANO, Ignácio; DUARTE, Thais. No sapatinho: a evolução das milícias no Rio de Janeiro [2008-2011]. Rio de Janeiro: Fundação Heinrich Boll, 2012.
- CIPRIANI, Marcelli. Os coletivos criminais de Porto Alegre entre a “paz” na prisão e a guerra na rua Dissertação (Mestrado) - Pontifícia Universidade Católica do Rio Grande do Sul, Porto Alegre, 2019.
- CORDEIRO, Ricardo. Morte matada Curitiba: Appris, 2022.
- COZZI, Eugenia. “‘Nosotros éramos una cooperativa de distribución’: Algunas transformaciones en el mercado de drogas ilegalizadas en un barrio popular de Rosario, del cuenta-propismo a una comercialización a mayor escala”. Dilemas, Rev. Estud. Conflito Controle Soc., Rio de Janeiro, vol. 13, n. 2, pp. 463-484, 2020.
- DAS, Veena. Vida e palavras: a violência e sua descida ao ordinário. São Paulo: Unifesp, 2020.
- DREYFUS, Pablo; GUEDES, Luis Eduardo; LESSING, Ben; BANDEIRA, Antônio Rangel; NASCIMENTO, Marcelo de Sousa; RIVERO, Patricia Silveira. Small arms in Rio de Janeiro: the guns, the buyback, and the victims. Geneva: Small Arms Survey, 2008.
- DUARTE, Thais; ARAÚJO, Isabela. “Caminho sem volta? Faces da expansão do PCC a Minas Gerais”. Tempo Social, São Paulo, vol. 32, n. 3, pp. 173-196, 2020.
- DUARTE, Thais; ARAÚJO, Isabela. “PCC em pauta: Narrativas jornalísticas sobre a expansão do grupo pelo Brasil”. Dilemas, Rev. Estud. Conflito Controle Soc., Rio de Janeiro, vol. 13, n. 2, pp. 505-532, 2021.
- EM DEZ ANOS, HOMICíDIOS CAEM 70% NO ESTADO DE SãO PAULO. Câmara Municipal de São Paulo, São Paulo, 15 May 2014. Available at saopaulo.sp.leg.br/blog/em-dez-anos-homicidios-caem-70-no-estado-de-sao-paulo. Consulted 6 February 2024.
- FELTRAN, Gabriel. “Crime e castigo na cidade: os repertórios da justiça e a questão do homicídio nas periferias de São Paulo”. Caderno CRH, Salvador, v. 23, n. 58, pp. 59-73, 2010.
- FELTRAN, Gabriel. Fronteiras de tensão: política e violência nas periferias de São Paulo. São Paulo: Unesp, 2011.
- FELTRAN, Gabriel. “Governo que produz crime, crime que produz governo: O dispositivo de gestão do homicídio em São Paulo (1992-2011)”. Revista Brasileira de Segurança Pública, São Paulo, vol. 6, n. 2, pp. 232-255, 2012.
- FELTRAN, Gabriel. Irmãos: Uma história do PCC. São Paulo: Companhia das Letras, 2018.
- FELTRAN, Gabriel. “Homicídios no Brasil: Esboço para um modelo de análise”. In: BUENO, Samia; LIMA, Renato Sérgio de (coord.). Anuário Brasileiro de Segurança Pública 13. ed. São Paulo: Fórum Brasileiro de Segurança Pública, 2019. pp. 30-35.
- FELTRAN, Gabriel; MOTTA, Luana. “Policía y ladrón”. Runa, Buenos Aires, vol. 42, n. 1, pp. 43-64, 2021.
- FELTRAN, Gabriel; LERO, Cecília; CIPRIANI, Marcelli; MALDONADO, Janaina; RODRIGUES, Fernando de Jesus; SILVA, Luiz Eduardo Lopes; FARIAS, Nido. “Variações nas taxas de homicídios no Brasil: Uma explicação centrada nos conflitos faccionais”. Dilemas, Rev. Estud. Conflito Controle Soc., Rio de Janeiro, vol. 15, pp. 311-348, 2022.
- GERRING, John; KNUTSEN, Carl Henrik. “Homicide and state history”. American Sociological Review, Thousand Oaks, vol. 87, n. 6, pp. 950-980, 2022.
- GRAMSCI, Antonio. A questão meridional São Paulo: Paz e Terra, 1987. (Coleção Pensamento Crítico, 72 vol.).
- HIRATA, Daniel. Sobreviver na adversidade: mercados e formas de vida. São Carlos: EdUFSCar, 2018.
- HIRATA, Daniel; GRILLO, Carolina. “Movimento e morte: o tráfico de drogas nas cidades de São Paulo e Rio de Janeiro”. Journal of Illicit Economies and Development, London, vol. 1, n. 2, pp. 122-133, 2019.
- JESUS, Andreia Sousa de. Fica Vivo! O racismo na política pública de prevenção à criminalidade em Minas Gerais. Tese (Doutorado em Ciências Sociais) - Universidade Estadual Paulista, Marília, 2023.
- JÚNIOR, Renato; FERRAÇO, Laurejan. Guerra Federal: retratos do combate a crimes violentos no Brasil. 2°ed. aum. Brasília, DF: Fundação Biblioteca Nacional, 2020.
- JUSTUS, Marcelo; CERQUEIRA, Daniel; Kahn, Tulio; MOREIRA, Gustavo. “The “São Paulo Mystery”: The role of the criminal organization PCC in reducing the homicide in 2000s”. Economia, vol. 19, n. 2, pp. 201-218, 2018.
- KAHN, Túlio. “Crescimento econômico e criminalidade: Uma interpretação da queda dos crimes no Sudeste e aumento no Norte/Nordeste”. Revista Brasileira de Segurança Pública, São Paulo vol. 7, n. 1, pp. 152-164, 2013.
-
KAHN, Túlio. “Confronto entre facções e homicídios: Teste empírico”. Fonte Segura: Revista do Fórum Brasileiro de Segurança Pública, São Paulo, n. 127, 29 set. 2021. Disponível em: https://fontesegura.forumseguranca.org.br/wp-content/uploads/sites/2/2021/10/Ed_107_Analises_criminais_Confronto_entre_faccoes_e_homicidios_teste_empirico.pdf 2021. Acesso em: 3 jun. 2026.
» https://fontesegura.forumseguranca.org.br/wp-content/uploads/sites/2/2021/10/Ed_107_Analises_criminais_Confronto_entre_faccoes_e_homicidios_teste_empirico.pdf -
LINHARES, Carolina. “Apesar de queda em mortes, ataques a ônibus e crimes assustam MG”. Folha de S. Paulo, São Paulo, 11 June 2018. Available at https://www1.folha.uol.com.br/cotidiano/2018/06/apesar-de-queda-em-mortes-ataques-a-onibus-e-crimes-assustam-mg.shtml Consulted 5 June 2026.
» https://www1.folha.uol.com.br/cotidiano/2018/06/apesar-de-queda-em-mortes-ataques-a-onibus-e-crimes-assustam-mg.shtml - MANSO, Bruno. “A cena criminal brasileira mudou; compreendê-la ajuda entender as novas dinâmicas do homicídio”. In: BUENO, Samia; LIMA, Renato Sérgio de (coord.). Anuário Brasileiro de Segurança Pública 13. ed. São Paulo: Fórum Brasileiro de Segurança Pública, 2019. pp. 36-39.
- MANSO, Bruno; DIAS, Camila. A guerra: a ascensão do PCC e o mundo do crime no Brasil. São Paulo: Todavia, 2018.
- MARQUES, Adalton. Crime, proceder, convívio-seguro: um experimento antropológico a partir de relações entre ladrões. Dissertação (Mestrado em Antropologia) - Universidade de São Paulo, São Paulo, 2009.
- MARQUES, Thalia. A Ordem: tempo analítico, homicídios e PCC no Triângulo Mineiro (1990-2020). Dissertação (Mestrado em Sociologia) - Universidade Federal de São Carlos, São Carlos, 2023.
- MEDEIROS, Flavia. Linhas de investigação: uma etnografia das técnicas e moralidades numa divisão de homicídios da Polícia Civil do Rio de Janeiro. [S. l.: s. n.], 2021.
- MELLO, João; SCHNEIDER, Alexandre. Mudança demográfica e a dinâmica dos homicídios no Estado de São Paulo São Paulo: [s. n.], 2007.
- MELO, Juliana; RODRIGUES, Raul. Notícias de um massacre anunciado e em andamento: o poder de matar e deixar morrer à luz do massacre no Presídio de Alcaçuz, RN. Revista Brasileira de Segurança Pública, São Paulo, vol. 11, n. 2, pp. 48-62, 2017.
- MIRANDA, Ana Paula Mendes de; PITA, María Victoria. Rotinas burocráticas e linguagens do estado: “Políticas de registros estatísticos criminais sobre mortes violentas no Rio de Janeiro e em Buenos Aires”. Revista de Sociologia e Política, Curitiba, vol. 19, n. 40, pp. 59-81, 2011
- MISSE, Michel. Crime e violência no Brasil contemporâneo Estudos de sociologia do crime e da violência urbana. Rio de Janeiro: Lumen Juris, 2006.
-
MP DENUNCIA 41 INTEGRANTES DO PCC QUE ATUAVAM EM PRESíDIO DE UBERABA. O Tempo, Belo Horizonte, 24 April 2014. Available at https://www.otempo.com.br/cidades/mp-denuncia-41-integrantes-do-pcc-que-atuavam-em-presidio-de-uberaba-1.828155 Consulted 4 June 2026.
» https://www.otempo.com.br/cidades/mp-denuncia-41-integrantes-do-pcc-que-atuavam-em-presidio-de-uberaba-1.828155 - MUSUMECI, Leonarda. “Letalidade policial e pessoas desaparecidas no Estado do Rio de Janeiro, segundo os dados oficiais (2006-2018)”. Boletim Segurança e Cidadania, Rio de Janeiro, vol. 26, 2020.
- NERY, Marcelo; SOUZA, Altay; ADORNO, Sergio. “Os padrões urbano-demográficos da capital paulista”. Estudos Avançados, vol. 33, n. 97, pp. 5-36, 2019.
- PAIVA, Luiz Paiva. “‘AQUI NÃO TEM GANGUE, TEM FACÇÃO’: As transformações sociais do crime em Fortaleza, Brasil”. Caderno CRH, Salvador, vol. 32, n. 85, pp. 165-184, 2019.
-
PCC ORQUESTRA ATAQUES A CIDADES DE MINAS GERAIS. [S. l.: s. n.], 2018. 1 vídeo (2 min). Publicado pelo canal Brasil Urgente Available at https://www.youtube.com/watch?v=9v6BzdK5PQY Consulted 5 June 2026.
» https://www.youtube.com/watch?v=9v6BzdK5PQY - PERES, Maria Fernanda; ALMEIDA, Juliana Feliciano; VICENTIN, Diego; CERDA, Magdalena; CARDIA, Nancy; ADORNO, Sergio. “Queda dos homicídios no Município de São Paulo: Uma análise exploratória de possíveis condicionantes”. Revista Brasileira de Epidemiologia, São Paulo, vol. 14, n. 4, pp. 709-721, 2011.
- PLATERO, Karina; VARGAS, Joanna. “Homicídio, suicídio, morte acidental… ‘O que foi que aconteceu?’”. Dilemas, Rev. Estud. Conflito Controle Soc., Rio de Janeiro, vol. 10, n. 3, pp. 621-641, 2017.
- RATTON, José Luiz; DAUDELIN, Jean. “Construction and deconstruction of a homicide reduction policy: The case of pact for life in Pernambuco, Brazil”. International Journal of Criminology and Sociology, Ontario, vol. 7, pp. 173-183, 2018.
- RIBEIRO, Ludmila; OLIVEIRA, Victor Neiva; BASTOS, Luiza. “Pavilhões do Primeiro Comando da Capital: Tensões e conflitos em uma unidade prisional de segurança máxima em Minas Gerais”. O Público e o Privado, Fortaleza, vol. 17, n. 33, pp. 213-241, 2021.
- RICHMOND, Matthew Aaron. “The pacification of Brazil’s urban margins: Peripheral urbanisation and dynamic order-making”. Contemporary Social Science, Oxfordshire, vol. 7, n. 3, pp. 248-261, 2021.
- ROCHA, Rafael. “O papel do terceiro: Possibilidades de articulações de uma gramática moral e a gestão da violência nas periferias brasileiras”. In: ENCONTRO ANUAL DA ANPOCS, 43., 2019, Caxambu. Anais […]. São Paulo: Anpocs, 2019.
- RODRIGUES, Fernando. “‘CORRO COM O PCC’, ‘CORRO COM O CV’, ‘SOU DO CRIME’: Facções, sistema socioeducativo e os governos do ilícito em Alagoas”. Revista Brasileira de Ciências Sociais, São Paulo, vol. 35, n. 102, p. e3510216, 2020.
- RUOTTI, Caren; REGINA, Fernanda Lopes; ALMEIDA, Juliana Feliciano de; NASSER, Marina Mattar Soukef; PERES, Maria Fernanda Tourinho. “A ocorrência de homicídios no município de São Paulo: mutações e tensões a partir das narrativas de moradores e profissionais”. Saúde e Sociedade, São Paulo, vol. 26, n. 4, pp. 999-1014, 2017.
- SALDANHA, Raphael; BASTOS, Ronaldo; BARCELLOS, Christovam. “Microdatasus: pacote para download e pré-processamento de microdados do Departamento de Informática do SUS (DATASUS)”. Cadernos de Saúde Pública, Rio de Janeiro, vol. 35, n. 9, p. e00032419, 2019.
- SILVEIRA, Andréa Maria. Prevenindo homicídios: avaliação do Programa Fica Vivo no Morro das Pedras em Belo Horizonte. Tese (Doutorado em Sociologia e Política) - Universidade Federal de Minas Gerais, Belo Horizonte, 2012.
- SILVEIRA, Andréa Maria; SILVA, Bráulio Alves Figueiredo; BEATO FILHO, Cláudio Chaves. “Prevenção de crimes urbanos: O programa Fica Vivo!”. In: LIMA, Renato Sérgio de; PAULA, Liana de (org.). Segurança pública e violência: o Estado está cumprindo seu papel? São Paulo: Contexto. 2008. pp. 77-88.
- SINHORETTO, Jacqueline. Mapa do encarceramento: os jovens do Brasil. Brasília, DF: Presidência da República, 2015. (Série Juventude Viva).
-
STABILE, Arthur. “Bilhete indica que PCC teria ordenado série de ataques em Minas Gerais e Rio Grande do Norte”. El País, Madri, 5 June 2018. Available at https://brasil.elpais.com/brasil/2018/06/05/politica/1528156233_508936.html Consulted 6 June 2026.
» https://brasil.elpais.com/brasil/2018/06/05/politica/1528156233_508936.html - WEBER, Max. Conceitos sociológicos fundamentais Amadora: LeYa, 2019.
- ZALUAR, Alba. A máquina e a revolta: as organizações populares e o significado da pobreza. São Paulo: Brasiliense, 2002.
- ZALUAR, Alba. Nexos entre droga, violência e crime organizado. Revista Brasileira de Sociologia, Porto Alegre, vol. 7, n. 17, pp. 55-76, 2019.
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, and IBGE Table 2093.
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, CID-10, and IBGE Table 2093.
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, and IBGE Table 2093.
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, and IBGE Table 2093
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, and IBGE Table 2093.
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, and IBGE Table 2093.
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, and IBGE Table 2093.
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, and IBGE Table 2093.
Source: Authors’ own calculations based on DataSUS, ICD-10 groups X85-Y09 and Y35-Y36, and IBGE Table 2093.