Movement and Death: Illicit Drug Markets in the Cities of São Paulo and Rio De Janeiro

This article analyzes the relationship between violence and the illicit drug market by comparing retail drug trade in the favelas and peripheral neighborhoods of the cities of Sao Paulo and Rio de Janeiro. To do so, we carried out ethnographical field work in both cities and a literature review on the topic. Considering illegal markets and forces of order in charge of suppressing them as part of one single object of study, this article explores the mechanics of the relationship between drug trafficking and the police and how it governs the lives and deaths of the poor population in those cities. Rio and Sao Paulo showcase different scenarios in terms of how the drug trade is structured and practiced: Whereas Rio de Janeiro lives in a state of ‘war’ due to the disputes among rival drug factions and with the police (who is thus more lethal), Sao Paulo is believed to live ‘at peace’, as trafficking is controlled by one single comando and the city showcases higher incarceration rates. We shall argue that such differences also influence the way the retail drug market operates in each city. This comparison focuses on the intersection of three dimensions: The marketplaces of drug retail sales; the dynamics of criminal collectives; and the different power dynamics among drug dealers and forces of order. O presente artigo reflete sobre a relacao entre mercado ilegal de drogas e violencia, a partir da comparacao do mercado varejista de drogas em favelas e bairros de periferias das cidades de Sao Paulo e Rio de Janeiro. As reflexoes aqui apresentadas baseiam-se no trabalho de campo etnografico realizado nas duas cidades, bem como a revisao de estudos sobre o tema. Tomando os mercados ilegais e as forcas da ordem encarregadas de combate-los como parte do mesmo objeto de pesquisa, o presente artigo explora a mecânica das relacoes entre trafico de drogas e policia, envolvida na gestao da vida e morte da populacao pobre. Em face as guerras concorrenciais entre faccoes cariocas e a ideia de paz associada a hegemonia de um unico comando em Sao Paulo, bem como a maior letalidade da acao policial no Rio de Janeiro e o maior encarceramento em Sao Paulo, sera argumentado que essas diferencas se relacionam tambem ao modo como e operada a venda varejista de drogas em cada cidade. A comparacao concentra-se no cruzamento de tres dimensoes: os lugares de mercado do varejo de drogas; as dinâmicas dos coletivos criminais; e os distintos arranjos de poder estabelecidos entre os traficantes e os agentes das forcas da ordem.


Introduction
The most common representation of Rio de Janeiro at national and international levels nowadays is that of armed conflicts between policemen and armed drug dealers. Given the great visibility of the country's former capital, these images shape the way violence is generally perceived in Brazil. However, this type of relationship between police and drug traffic is a very specific characteristic of Rio de Janeiro and cannot be observed in other large Brazilian cities, such as São Paulo. The way the drug trade is organized, and the means of interaction among trafficking groups and between these groups and the forces of order, vary from one place to another. Our aim is to highlight such differences by comparing the illicit drug market in the favelas (poorly urbanized, low-income neighborhoods, notorious for their lack of infrastructure) and peripheral areas in the cities of São Paulo and Rio de Janeiro.
This article is the result of a series of discussions between the authors, based upon ethnographic research about the criminal dynamics in the urban peripheries of São Paulo (2006)(2007)(2008)(2009)(2010) and favelas of Rio de Janeiro (2009-2012. We also performed a review of several studies carried out by peer ethnographers to whom we are indebted for the privilege of learning from them and discussing with them. 1 Our starting point is the set of images that determine the representation of 'urban violence', outlined as a 'public problem' (Boltanski and Thévenot 1991) in Brazil since the late 1970s. According to Machado da Silva (1999,2004,2008), the perception of constant increase in urban violence selects and indicates a complex of practices considered threats to personal physical integrity and to material property. We may also add that such representation is directly associated with the inception of a drug market, which, according to Machado da Silva, is seen to provide material bases for the reproduction of violent modes of interaction.
Antonio Rafael Barbosa (2005) argues that-more in Rio than in São Paulo-the drug trade was able to organize the field of popular illegalisms around itself, bringing all other criminal activities to its orbit and overcoding the entire crime market. In addition, the drug trade surfaced as the main manifestation of the so-called ' organized crime' since the beginning of the 1980s. This contemporary image of the 'social enemy' is imbued in a series of militarized social control policies (Graham 2010), foments the military markets and the fusion between National Security and Defense (Bigo 2000), bringing together Brazil's past and present governmental forms (Sanjurjo and Feltran 2015).
In Rio, throughout the 1980s, the drug trade in the favelas-known as movimento (movement)-becomes associated with a high volume of violence, uncommon in other large cities around the globe (Misse 2006). The frequency and intensity of armed conflicts between drug dealers and the police, as well as the competition among drug factions 2 for the control over territories in Rio de Janeiro, does not find parallel even in other large Brazilian urban centers like São Paulo. It is noteworthy that other Brazilian cities show much higher homicide rates than Rio; São Paulo itself was very violent in the 1990s. However, it is in Rio de Janeiro, a city torn between the favelas and the asphalt 3 (Ventura 1994), that the 'metaphor of war' (Leite 2000) became the rule of thumb to explain urban conflict. Assuming that a war is indeed under way, the advocates of this perspective defend the use of lethal police force in the favelas and against their dwellers on the pretext of fighting drug trafficking.
Decades have gone by and alternatives to this security policy model were tested, but Rio remains hostage of a conflict understood as a 'war' by all parties involved. In fact, the state was even subject to a federal intervention: Government administration was yielded to an Army general and the state of Rio de Janeiro's police force was backed up by a contingent of the National Army. Since the beginning of the intervention, in February of 2018, the number of deaths due to law enforcement's use of lethal force increased dramatically (up to four deaths per day), whereas there was no significant decline in criminal occurrence. According to Rio de Janeiro's Public Security Institute (ISP/SESEG-RJ, in the Brazilian acronym), 4 there was a 33.6% increase in deaths due to law enforcement's use of lethal force between February and December of 2018 in comparison to the same period in 2017. The number of homicides, in its turn, decreased 8.2%; registered robberies decreased 2.6%. 5 A significant part of Brazilian society perceives urban violence as a major public concern. Therefore, it does not come as a shock that the country's far right mobilized the fear of violent crime to triumph in the last elections. More rigorous punishment, increase in incarceration rates, softening of gun laws and even the extermination of criminals took center stage in victorious campaigns across Brazil-including the presidential campaign. In Brazil, over 60,000 people are murdered every year, and yet voters chose a program to combat violence that proposed to arm the civil population and increase the use of lethal force by the police.
This turbulent moment and the setbacks in public debate about violence sets the stage for this article. Our aim is to discuss the relationship between the illicit drug market and the government of deaths in favelas and urban peripheries in the cities of São Paulo and Rio de Janeiro in comparative perspective. Our premise is that violence is not a mere question of law enforcement, as the criminalization of illicit markets by the State apparatus is a constitutive part of the object and thus must be factored in the problematization. In this sense, we understand that there is a dynamic correlation between these markets' means of operation and the agents in charge of their control and repression (Veen 1999). We understand them as interrelated research topics, as the policies and their effects are part of the object, and not external to it (Barry 2002).
The representation of urban violence seems to bring forth a mechanic that organizes the functioning of criminal markets in an inevitable relationship with the forces of order. Therefore, it affects the management of deaths of the poor in the two cities: a device of government that is both concerted and conflictive, produced by weaves of decisive interactions between law enforcement and criminal agents (Hirata 2014). A dynamic in which 'forms of government beget crime and dynamics of crime beget government' (Feltran 2011).
To outline this relationship and devise parameters of comparison between Rio and São Paulo, we built upon a few suggestions made by Gabriel Feltran. 6 In terms of public security policies, the state of São Paulo's main initiative over the last decades has been mass incarceration. 7 The state's prison population (240,061 prisoners) is the largest in the country, its incarceration rate is one of the highest (536.05 prisoners per 100,000 inhabitants; DEPEN-MJ, June 2016). Rio de Janeiro, in its turn, saw an increase in incarceration rates, but is still below national average (301.09 prisoners per 100,000 inhabitants), and, thanks to frequent police armed incursions carried out in the favelas, the state has the country's highest death rates due to law enforcement's use of lethal force (Cano 1997;Misse et al. 2013). When it comes to criminal collectives, we see opposing segments segregated inside prisons and competing for control over territories for drug sales in Rio de Janeiro, whereas São Paulo has one single hegemonic comando in both prisons and territories across the city.
The organization of prison inmates gave rise to the inception of criminal collectives known as comandos or facções (factions), both in Rio and São Paulo. These collectives were formed within the prisons with the aim to regulate relations among inmates and demand better life conditions. It is worth noting, nevertheless, that the advent of the Comando Vermelho (CV) in Rio de Janeiro-known for giving rise to new forms of organized crime-dates back to 1975, 8 whereas Primeiro Comando da Capital (PCC), in São Paulo, was allegedly founded in 1993 9 (almost 20 years later). In both states, there was a temporal gap between the advent of comandos and the expansion of their influence across illicit markets in the favelas and low-income neighborhoods.
CV saw its popularity grow throughout the 1980s, when major heists were attributed to its members. It was only at the end of the decade, however, that that the drug trade in the favelas of Rio de Janeiro came to be controlled by factions from within the prisons (Misse 2006). In its turn, PCC just became known in 2001, when simultaneous rebellions took place in 29 prisons across the state of São Paulo. Five years later, PCC demonstrated their influence out of prisons in an episode known as 'PCC Attacks' (Adorno & Salla 2007)-or 'May Crimes' (Mães de Maio 2011), when considering exterminations carried out by the police in response to the attacks.
The comandos' common roots and their ties to the drug trade (their main source of income) suffice to determine certain consistencies between the cases of Rio de Janeiro and São Paulo. Although the concrete abstraction of ' organized crime' allows us to find similarities between factions from both cities, they seem far apart and empirically incongruent when looked upon in close detail.
For the purpose of this article, it is interesting to analyze convergences and divergences strictly in relation to the retail drug market-called movimento (movement)-and we refer particularly to sales points situated in poor neighborhoods, which are the object of our empirical research. Our field data indicates that retail markets in the cities of Rio and São Paulo are organized into a 'game of scales', local and translocal (Revel 1998), by collectives that showcase complex political and economic dynamics and mobilize horizontal and hierarchical relations.
A certain common sense-informed sometimes by the media, some other times by the academy-suggests a different picture: That of a 'mafia' or a ' cartel', where drug trade across the city would be ' controlled' by a hierarchical, centralizing structure, and their leaders' calculations and authoritative actions. 10 This image of 6 We greatly appreciate Gabriel Feltran's insights provided during conversations, which resulted in this excerpt. Nonetheless, we must point out that we are fully liable for any misconceptions that arise out of it. 7 For a historical review of the Brazilian case, see Salla (2006Salla ( , 2007; for a current analysis, see Godoi (2015) and Minhoto (2002). 8 Foundational year of 'Falange da LSN' (in reference to National Security Act -LSN, in the Brazilian acronym). The group was renamed as 'Falange Vermelha' in 1976 and saw its popularity grow from 1979, when the press started using the name 'Comando Vermelho' (Lima 1991). 9 Karina Biondi emphasizes that the massacre that took place at Carandiru (a former penitentiary in São Paulo) in 1992 is critical to understand how PCC came about and that there are several versions for the advent of PCC, albeit all overshadowed by the happenings of August 31, 1993. For a more detailed description, see Biondi (2010) and Dias (2013 ' organized crime' is more aligned with the idea of an enemy (whether rational or irrational) to be eliminated than with actual practices verified by more detailed studies.
In this study, we aim to understand how the local configuration of drug trade can help us go beyond these stereotypical images of urban violence. Our starting point was the empirical difference found while crossanalyzing three dimensions: (1) forms of organization and territorialization of the retail drug trade; (2) differences between factions in each city; and (3) drug dealers' relationships with police officers. Our aim was not breaking down each dimension, but rather, understanding how they connect to one another in order to understand such particular type of urban economy as well as its social and political consequences.

The Drug Trade Marketplaces
Overall, drug networks can be described through a 'geometry of variable scales' that articulate wholesalers and retailers. They consist of several groups of different sizes and varied articulations in terms of profitability, risk, and negotiation methods, aimed at ' circumventing' the laws and thus allowing the circulation of merchandise (Barbosa 2005). Drug trafficking takes place both in poor and rich neighborhoods, but its dynamic varies significantly depending on the circumstances. 11 For this reason, we opted for focusing our analysis on the retail trade in the favelas and urban peripheries of São Paulo and Rio.
To some extent, the networks present in both cities are somehow alike. The territorialized sales points, easily identifiable by drug users, are called biqueiras in São Paulo and bocas de fumo in Rio. In both cities, they need to obtain a 'business licence' (Barbosa 2005;Hirata 2018) from forces of order and to develop security strategies that tend to be quite different in each city. Up to a certain point, one can observe that, in both contexts, local trafficking enterprises (known as firmas, i.e., ' companies') showcase hierarchical frameworks that emulate traditional employer-employee relationships and job descriptions. Both in Rio and in São Paulo, firmas closely resemble the 'business model' described by Johnson et al. (1992) in their typology of crack distribution networks in the United States.
Broadly speaking, we see similarities in the ways roles are distributed and in the job descriptions of each role. The business owner is called the 'boss'. The 'managers' are the boss' henchmen and hold a coveted position within the group. They are in charge of organizing sales activities that, in its turn, are carried out by vapores, who earn a fixed weekly wage and/or a percentage of the sales (Malvasi 2012;Grillo 2013;Batista 2015). In São Paulo, Hirata (2018) also mentions the existence of campanas, who occupy lower positions in the biqueira's hierarchy 12 and have similar roles as olheiros and fogueteiros (scouts and informants) in Rio de Janeiro (Barbosa 1998): Warning the group that the police are nearby.
There is one position, nevertheless, that can be found in Rio de Janeiro but not in São Paulo: The soldados (soldiers). They are in charge of the defense-or contenção (containement) (Barbosa 1998)-of the boca, 13 and use gun power against law enforcement and dealers from other factions. The use of armed guards at sales points in Rio de Janeiro sets it apart from the practices in São Paulo in terms of their relationship with territories and law enforcement and even in terms of the roles performed by the boss and the manager in each city.
In order to have a better grasp of the movimento's configuration in both cities, it is critical to know that PCC is hegemonic in São Paulo, 14 whereas large criminal factions-Comando Vermelho (CV), Terceiro Comando Puro (TCP), and Amigos dos Amigos (ADA) 15 -have been fighting one another to control traffic in the favelas in Rio for over three decades. In Rio, the conflicts among factions resulted in the creation of armed defense around drug sales points and increasing treason prospects. This situation led to the proliferation of weaponry at bocas and the need for drug dealers to claim surveillance and punishment rights within their domains. This military administration of territories-exemplified by the existence of soldados-is the core of significant differences between internal forms of organization of factions in Rio and São Paulo.
In Rio de Janeiro, the 'boss' is known as dono do morro, 'the hill's owner.' (We may use only 'dono' for short.) There is an emphasis on the possession of territories at the favelas, since each firma's hierarchic structure 11 For an analysis of middle-class drug dealers, see Grillo (2008); for drug networks in Copacabana (an upscale neighborhood), see Pereira (2003). 12 Name given to a drug sales point in São Paulo. 13 Name given to a drug sales point in Rio de Janeiro. We may also use its short version, boca. 14 PCC is also confronted by other rival groups, such as Comando Brasileiro Revolucionário da Criminalidade (CRBC) and Terceiro Comando da Capital (TCC), but few are the prisons controlled by these two groups and their influence in the outskirts is negligible (Dias 2013;Marques & Biondi 2010). 15 There are also the milícias (militia), criminal groups that charge protection fees, control the market of basic services and build electoral bases for political candidates in low-income neighborhoods. For details about the milícias, see Cano and Duarte. is based on the donos' inalienable right over the exploitation of the drug trade in a certain area, warranted by their alignment with the drug faction (Grillo 2013). The persona of the dono do morro is part of Rio de Janeiro's urban history and precedes the drug trade. It is already present in the very first reports about the city's favelas, as written by Costallat (1995Costallat ( [1924), where 'bold' Zé da Barra is described as the 'favela's unquestionable boss.' This persona simply does not exist in São Paulo. The biqueira's bosses have control over business in one or more sales points, but not over the favela or neighborhood where the trade takes place. They are certainly prominent figures in any peripheral neighborhood, but their position is only valid in respect to their working relations with subordinates.
The favelas in Rio de Janeiro tend to accommodate several sales points owned by the same dono, whereas in São Paulo, each firma is in charge of one single sales point. Some donos have just a few sales points, while certain bosses control multiple points. However, the most common scenario is a larger territory accompanied by greater organizational complexity in Rio de Janeiro (Grillo 2013) in opposition to dynamic, fluid, fragmented trade in São Paulo (Malvasi 2012). This difference in scale requires distinct means of operationalization and originate different marketplaces (Braudel 1996) of the retail drug sales in each city.
Since the firmas in Rio de Janeiro control larger territories, they also showcase a more diverse corporate structure in terms of job allocation. Donos do morro look to building political alliances at local level to ensure ownership over drug sales points and revenue. While the large number of battering, banishments, and executions carried out by dealers are a convincing way to deter whistleblowers, betrayals, and lack of payment, power is not sustained only through the use of force. Donos do morro distribute multiple managerial positions-known as responsas 16 (a short-term for 'responsibilities')-among the most prestigious dealers and offer them a share of the sales profits. As many donos are incarcerated or do not live in the favelas under their control, they appoint a general manager of their utmost trust, known as responsável or frente do morro (the one in charge or hill's front), and this person takes over all their duties in that given area (Grillo 2013).
In larger firmas, there are different ranks of managers or responsáveis below the frente do morro. They are assigned diverse tasks, such as drug acquisition and bagging, scheduling of sales teams on duty and sales accountancy. Bocas de fumo trade drugs in batches. There are different price tiers for each drug type and batch size (marijuana at R$2, R$5, or R$10; cocaine at R$10, R$15, or R$20; and so on). Different managers are in charge of each drug type and/or price tier, but the batches are all sold by the vapor on duty under the protection of armed soldados. These two roles have predefined rotating schedules and are not considered positions of trust (Grillo 2013).
In São Paulo, on the other hand, hierarchy is less complex and consists of the following positions: Boss, manager, vapor, and campana. This structure also allows for more flexibility-bosses may agree to take charge of trading just one drug type and the roles of manager and vapor are less well defined than in Rio (Hirata 2018;Batista 2014). Fewer managers are necessary in São Paulo, as local business operates at a smaller scale. Vapores work for one single manager, who is their employer and to whom they are to report the sales they performed while on duty. Each drug has only one standard market price. In contrast with what happens in Rio de Janeiro, where the multiplication of managerial positions is seen as a strategy to maintain donos' sovereignty over their territories, the appointment of managers at biqueiras in São Paulo is solely based on their business profile and competences, despite their having a somewhat important role in the formation of alliances.
We could say that, from its morphological aspect, the drug distribution is very distinctive. In Rio, the circulation of merchandise is based on an organizational and territorial extension that multiplies intermediate (managerial) positions by putting drugs to sale in batches of different predefined sizes and prices, whereas in São Paulo, dealers work with a standard amount and price of each drug and seek to increase the volume of sales by concentrating the trade within circumscribed territories. For these reasons, it is possible to buy the same drug type in batches of different prices and sizes in Rio de Janeiro, but not in São Paulo; and the quality-based competition can occur intra-territorially in Rio de Janeiro and only inter-territorially in São Paulo.
In both cases, the good flow of drugs and money is ensured by means of centralized coercion-however, in Rio, it stems from the distribution of 'responsibilities' carried out by donos do morro (Grillo 2013), whereas in São Paulo, it stems from the distribution of competencies and attributions by bosses to their employees (Hirata 2018). This means to say that the relationship between donos do morro and their subordinates is very different from the one between São Paulo's bosses and their subordinates. Firstly, because internal mobility at the firma takes places in different ways, albeit based on similar factors. The so-called caminhada (one's 'walk' in ' crime') (Marques 2014) and consideração (consideration, repute) (Barbosa 1998(Barbosa , 2013 built around the valorization of the dealers' personal histories and collective memories (Hirata 2018).
The heart of the matter is that, in Rio de Janeiro, the comando gives donos de morro 'inalienable possession' (Weiner 1992) over traffic, in a way that any concessions of business exploitation go out and about the exchange system without losing its ties to its original owner (Grillo 2013). Donos cannot be forced out of their sales points unless through 'wars' 17 or 'coups d' Etát',18 which involves confronting the faction that supports them. In São Paulo, on the other hand, as bosses are employers, business arrangements are less stable, allowing compositions and recompositions of the firmas, sometimes without the need of confrontation (Malvasi 2012).
In both cases, it seems to us that what is at stake is the logic of a segmentary movement (Barbosa 2001). However, this can take on different forms depending on the dynamics between 'bosses' or donos do morro and comandos in each city.

Criminal Collectives
We are now ready to move on to the second dimension of our discussion. Comandos or factions are unique compounds made up of distinct matters. Hence, even though segmentation is present in both cities, it has different roots, that is to say, the process of breaking and building networks happens in the two cities, but they follow different steps.
Both cities showcase alliances between firmas and internal disruptions caused by treason within local criminal groups. Our hypothesis is that although internal ruptures and the formation of new alliances seem less frequent in Rio de Janeiro, they have a larger impact over the faction's dynamics, while they are more frequent in São Paulo, but cause a minor impact on the comando. This hypothesis offers a more nuanced version of the explanations that focus exclusively in the idea that peace in São Paulo is the result of PCC's monopoly in opposition to the frequent wars that take place between rival factions in Rio de Janeiro.
At first sight, one could say that the comandos in both cities work as horizontal networks of mutual protection (Misse 2006). That means to say, the adherence to these collectives and the ties of which they are made favor their members in situations of confrontation with their ' others'. The factions offer protection to their members-whether against mistreatment in prison, in confrontations with law enforcement, or in disputes against rival factions. Inmates organized themselves into comandos as a response to the precarious life conditions in prisons, where fear and distrust abound 'not only because of the violence practiced by correction officers, but also because of prison gangs that brought inmates together to rob, rape, and kill other inmates' (Lima 1991). The image of the faction's ' other'-alemão in Rio de Janeiro and coisa in São Paulo (Biondi 2010)-favors the production of a collective identity through processes of exclusion and opposition (Misse 1999). They function as devices to establish boundaries and contrasts that highlight the presence and adherence to the comandos.
However, the comandos in each city show distinct principles of inclusiveness and exclusiveness. In PCC, members go through a 'baptism' and must pay monthly fees to the comando 19 while out of jail (Dias 2013). Factions from Rio de Janeiro, on the other hand, have no clear association protocols. Whether or not engaged in local traffic, anyone who lives at an area controlled by a certain faction can evoke belongingness to it, as Natasha Neri (2009) showed with respect to adolescents detained in correctional institutes. It is worth noting that, despite PCC's 'baptisms' 20 , association with the comando is as fluid as in Rio de Janeiro. PCC's influence goes beyond its 'baptized' members, called 'brothers' 21 : It provides conduct benchmarks to the so-called ' cousins,' who claim 'to be sided with the comando' (Biondi 2010;Marques 2015). This is also evidenced by PCC's presence in correctional institutions for adolescent offenders, who, albeit not allowed to be baptized, often mime PCC's relationships and discourses (Mallart 2014). However, the difference between 'brothers' and ' cousins' -or baptized members and those who are only 'sided' with the comando-promotes distances and proximities (Marques 2010) that are not as demarcated in Rio.
At the favelas in Rio de Janeiro, everyone who works at a boca de fumo is automatically part of a faction and is authorized to boast it. Nevertheless, even though the sense of belonging is cardinal to the identities of those engaged in ' crime', 'bandits' (bandidos) sometimes switch their factional allegiance when the control over local drug sales points in the areas they live is taken by a rival comando. This is the case because the organizational structure of factions in Rio de Janeiro is closely knit with the retail drug market. The comandos are networks of alliances between donos de morro (Barbosa 2005); once they switch factional allegiance, all their underlings must follow suit. The deposition of a dono involves a change of command by means of a 'war' or a ' coup'-in the case of the latter that happens when power is taken over by the local firma's own members with the support of a rival group.
The functioning of drug trafficking in Rio de Janeiro is coordinated by alliances that articulate local groups in a permanent space of negotiation. There is no authority ' above' the donos de morro; the hierarchy only rules over business activities circumscribed to each group's territory (Barbosa 2006). Thus, donos rule hierarchically over their underlings at a local level while engaging in horizontal relations with allies at a supralocal level. The support from these allies authenticates their right to exploit the drug trade in given territories (Grillo 2013). Barbosa (1998) highlights the importance of 'friendship' (amizade) for the drug traffic in Rio de Janeiro. The distribution of operation areas and the circulation of drugs, weapons, and men is based on dynamics of alliances and segmentarity that permeate all disputes and negotiations within the city. Hierarchy in local sales points is vertical and topped by donos de morro, but its organizational structure also depends on lateral alliances between high rank members (Barbosa 1998). Movement and segmentation processes take place through composition/recomposition by activating/deactivating old friendships. Thus, once a local group obtains enough prestige and power to break up with the faction they belong to, the comando summons lineages to reactivate their local presence, which can lead to the following scenarios: (1) The inception of a new comando; (2) A shift in the relationship arrangements between factions; or (3) The abating of the emerging power center. 22 We could say that the relationships between alliances and hierarchies at local and supralocal levels in São Paulo are somehow similar but for one aspect: The myriad of drug sales points and bosses ensures a more fluid, sparse conformation. This is significant. There are many more bosses in São Paulo than there are donos de morro in Rio; hence, the groups are more mobile and translocal in São Paulo, whereas hierarchy and territories are more centralized in Rio.
It is noteworthy, however, that PCC is not exactly a drug trade faction. It is indisputable that, as the hegemonic comando at prisons and urban peripheries of São Paulo, it has a critical role in the drug market's regulation and that the drug trade is one of its members' main source of income. However, biqueira's bosses are not necessarily PCC 'brothers'-sometimes, they can be only ' cousins' (Malvasi 2012). The comando's organizational structure differs from its counterparts in Rio de Janeiro to the extent that it does not directly reflect the traffic's entrepreneurial hierarchies. Many of the alliances of which PCC is made of are aimed at the practice of robberies; in Rio, nonetheless, robbers work more independently and do not feature in drug factions' payrolls (Lyra 2013). This is the reason why the term responsa, refers to managerial positions occupied by dealers in drug retail firmas in Rio de Janeiro (Grillo 2013), while in São Paulo the term designates 'political positions' (Biondi 2010) or leadership roles (Marques 2014) occupied by PCC members in charge of ensuring ' discipline within the comando' 23 in prisons and quebradas 24 (Biondi 2018).
In this respect, the reconfiguration of alliances and rivalries among 'bosses' in São Paulo does not jeopardize PCC's influence in neighborhoods and quebradas. In contrast to the large political blocs in segmentary opposition that dispute the control of territories in Rio, PCC mitigated the disputes among the myriad of gangs that controlled the fragmented retail drug market in the city. In view of the drastic 80% reduction in homicide rate throughout the 2000s in the state of São Paulo, urban ethnographers argue that the decrease in violent deaths was associated with PCC's role in regulating markets and mediating conflict-what became known as the 'PCC Hypothesis' (Lima 2009).
Karina Biondi presents a sophisticated analysis of 'tuning' (sintonia), a critical concept to understand the differences between realities in São Paulo and Rio de Janeiro (Biondi 2018). She proposes that PCC is a 'movement' organized by means of ' connections' and ' acquaintances' between those that are 'sided with one another', 'sharing the walk'. According to her, 'tuning' is the link that connects the comando, 'thieves' 22 Power arrangements between factions have gone through significant changes in Rio lately. In response to a ' coup' followed by war in the favela of Rocinha, reconquered by CV, there was an attempt of fusion between ADA and TCP, creating a new faction called Terceiro Comando dos Amigos (TCA). Some donos de morro from ADA did not accept the fusion and, failing to maintain their autonomy, switched sides to CV. 23 Strict compliance to the criminal group's code of conduct. 24 Extremely poor neighborhoods in São Paulo.
(ladrões), prisons, and quebradas. However, she argues that this is more than just a synonym for connection-'tuning' generates a different 'movement' (movimento). Biondi's definition of movimento greatly diverges from the one Misse (2006) used to describe the local drug market in Rio de Janeiro.
She states that, within PCC, being 'in the movement' means to be 'in tune' with the comando; to align with ideas that echo in PCC and ' embrace' responsas (responsibilities) in it (Biondi 2018), regardless of one's role in traffic and other crimes. According to Marques (2010), it means to be, at once, a product and a producer of ' crime'. Crime and criminal are not legal categories in this sense: It is possible to be 'sided with crime' without actually engaging in criminal action, as it is possible to be alien to ' crime' while performing criminal acts (as it is the case of policemen who face criminal charges).
It might be useful to think of how the differences in the movimento's configuration interweave with the way connections and segmentations are used to give rise to criminal factions in Rio de Janeiro and São Paulo. This does not imply that these collectives determine or establish the singularity of marketplaces. The urban history of marketplaces and criminal collectives does not allow us to ascertain this link. Nevertheless, it seems reasonable to associate a market defined by donos' 'friendships' and alliances and another consisting of 'bosses' 'in tune'. Such is the case because political and economic relations (if ever possible to separate these terms) seem to produce, on one hand, a more extensive, densely organized space, impacted by segmentary blocs of alliances and rivalries; and, on the other hand, a more compact, scattered space, whose articulations show varied levels of convergence and divergence.

Arrangements among Traffic, Territories, and Forces of Order
In order to have an ethnographic grasp of the agencements that allow the circulation of illicit drugs , the interactions between actors involved in this illegal market and agents in charge of controlling it is of critical importance. The fact that the commercialization of drugs is prohibited makes law enforcement a key element for the drug market regulation-one of the most profitable markets in the favelas and urban peripheries of both cities. Empowered by prerogatives granted by the State, the role of the police officers directly in charge of repressing drug markets goes beyond enforcing the laws-they are tacitly given autonomy and discretion to screen, select, and negotiate locally with those subject to their authority (Gazit 2009). Assembled [agencé] through conflicts and negotiations with police officers, traffic's power is not 'parallel', as commonly inferred in the images associated with the concept of ' organized crime'-rather, it is 'tangential' (Barbosa 2005).
The decisive interaction between law enforcement agents and those involved in drug trafficking is popularly known as acerto (agreement) in São Paulo and arrego (surrender) in Rio de Janeiro (Hirata, 2014). These dynamics are best understood through the concept of 'political commodity'. According to Michel Misse, the 'political assets' market is a parasite market that regulates informal and illicit markets, focused on services whose prices are not solely defined by market laws, but also by 'strategical assessments of power, potential resource to violence and balance of power' (Misse 2006).
The native/analytical concept of 'sales point' may help us understand how Misse's formulation relates economic and political forms. Briefly, we can say that all commerce (legal or illegal) sets a sales point in relation to an urban agglomeration, i.e., its business value stems from its strategical position within urban agglomerations (Villaça 1998). For this to work, one needs to get a (legal/illegal) business licence, which implies an authority that is, at the same time, coercive and consensually instituted and that has the power to authorize business activities in a certain area (Hirata 2014). This is shaped according to different logics of 'spatialization of exchanges' (Rabossi 2004) in both cities. We shall now describe how the sales point is negotiated in each city.
In São Paulo, local drug trafficking groups do not attempt to show military domain over territories. There are no longer places where law enforcement is now allowed in and that, ultimately, are not controlled by the police. In addition, sales strategies are usually discrete and negotiated with police officers. The relationship between police officers and drug dealers is built upon the common knowledge that State security forces have superior military power and no one attempts to 'lessen' such disparity. The agents' strategic force calculations assume that State forces are superior. Within the limits of each precinct's outreach, the police are potentially and effectively superior than the drug dealers, and they are allowed in neighborhoods where drug sales points are set whenever they like.
For this reason, the local police department acts as a drug sale regulating center and has supreme power over it. This allows law enforcement to charge for two different types of political commodity: A 'Business Licence' and ' extorsion by kidnapping' (Hirata 2014). The former is an authorization to operate at a certain place; the latter is the practice of invading sales points and 'kidnapping' dealers to demand a 'ransom' for their release.
In Rio, even though the State's armed forces are also clearly superior, drug traffickers resist police interventions within their territories. Intermittent and relatively unpredictable police incursions at the favelas are the backbone of a model of drug market regulation, which rests on the lack of regular patrolling in territories labeled as 'risk areas' and left under the control of armed criminals (Grillo 2016). These areas were already put aside by routine policing even before the emergence of armed traffic. This situation was aggravated in the last three decades as increased the capital flow in the favelas brought about by the cocaine trade (Misse 2006). The reproduction of retail networks came to rely on the use of armed 'soldiers' to protect drug dealers, weapons, drugs, and money from potential raids, thefts, arrests and seizures carried out by police officers or rival factions.
The payment of a business licence (called arrego) and the extorsion by kidnapping (or mineira) are also common practices in Rio de Janeiro (Barbosa 2005). However, the exchange of political commodity is never completely stabilized, and paying arrego does not ensure the ceasing of confrontations between drug dealers and the police. Police incursions are still the main method of crime 'regulation', as it measures up local traffickers' resilience and forces the need for negotiation, as well as serving as retaliation for murders and robberies in certain areas. As a result, strategical assessments of power and of potential resource to violence play a critical role in shaping the traffic's modus operandi.
The dramatized image of shirtless boys bearing rifles to protect the firmas is a more plausible scene in Rio than in São Paulo, where the trade is carried out with greater discretion, without exposing merchandise on the streets and displaying arms for asserting territorial control. Weapons (including rifles and grenades) are also found in São Paulo, but they are rarely used for protecting drug sales points against police operations. They are more often used for settlements or rented for the practice of heists.
It is wrong to assume that there are no violent conflicts between the Police and drug traffickers in São Paulo. They happen occasionally and may also end up in deaths. However, only 8% of homicides caused by police interventions are due to traffic repression (Grillo et al., mimeo), as opposed to Rio de Janeiro, where most of deaths due to police use of lethal force occur during conflicts with armed traffickers (Misse et al. 2013). In São Paulo, confrontation between the police and the drug traffic has not caused ruptures to the urban fabric; police cars are allowed across the entire metropolitan area.
What catches the eye in São Paulo is the frequency of slaughters allegedly perpetrated by police officers off duty. Dwellers at the peripheries see these slaughters as retaliation for the deaths of other police officers, supposedly murdered by PCC members. This was common practice in Rio de Janeiro throughout the 1990s. 25 In the twenty-first century, however, retaliation has been often disguised within formal operations resulting in several deaths; they integrate the strategic corpus of public security policies and may count on the legal device of "unlawfulness exclusion", which render most deaths legitimate when perpetrated by officers on duty.
There were attempts to modify relationship patterns between the police and the traffic at the favelas in Rio de Janeiro, where dwellers' routines are carried out in the crossfire 26 (Machado da Silva & Leite 2008) between them. Since 2008, the implementation of UPPs (Pacifying Police Units, in the Brazilian acronym) in certain areas aimed to replace the policing model of one-off operations with stable occupations within the favelas, which led to morphologic changes in the drug trade. Traffic repression and the protection of bocas de fumo, previously administered through armed confrontation and the payment of arrego, now relied more on 'reciprocal monitoring' of the circulation of drug dealers and police officers (Menezes 2015). Despite the reduction on the ostensive display of fire arms and bocas' increased mobility, the traffic kept its modus operandi and remained territorialized. According to Palloma Menezes (2015), UPPs turned a ' crossfire regimen' into a 'minefield' one; there was a reduction in violent conflict, replaced by a rise in underlying tension.
Nevertheless, UPPs were only implemented in strategic areas aimed at securing FIFA's World Cup in 2014 and the Olympic Games in 2016. They collapsed even before the events took place. Since 2013, dwellers reported everything was 'back the way it was' (Menezes 2016); wars between rival groups and armed confrontation between police officers and drug dealers intensified across the city, including said 'pacified' areas. The violence in Rio, especially in terms of armed robbery, grew so much that culminated in the federal, military intervention mentioned at the beginning of the article.

Final Considerations
The drug trade is a very particular illegalism due to its extensive repression and intensive death risk. Repression is extensive because the drug trade combines two proscriptions-it is an informal (non-regulated) market and it trades illicit merchandise. Death risk is intensive due to a number of factors, including its contextual significance to public order; society's moral reaction to it; and its potential or imaginary ties with other criminalized practices (Misse 2006). The combination of extensive repression and intensive exposure to death risk makes the drug trade unique and constitutes the specific forms of its agencement.
As mentioned earlier on, the consequences of massive incarceration in São Paulo were very different from the ones caused by confrontation and military occupation policies in Rio. From a political standpoint, choosing between incarcerated or dead individuals does not seem a good solution for the drug issue; from an analytical standpoint, those are two different ways of governing those deemed 'ungovernable'. As Karina Biondi highlighted, massive incarceration (associated with slaughters in prisons and on the streets) stimulated PCC's growing influence. In its turn, confrontation policies in Rio de Janeiro only caused rivalry among factions to increase and, as a whole, made the comandos more important.
By comparing the drug trade in Rio de Janeiro and São Paulo, we aimed to investigate the different devices for the management of deaths by focusing on those that are subject to it. We combined three analytical dimensions that, combined in multiple ways, seemed relevant to discuss the theme: the drug sales marketplace, the criminal collectives' dynamics, and the relationship with forces of order. Rio and São Paulo differ in terms of how 'war and peace' are arranged within and between factions and with the Police. The articulation between ' circulation and blockades' of commodities is also very distinctive. These aspects are paramount to define and understand the retail drug market in those spaces.
The extensive, fractional, and dense drug trade in Rio does not find a counterpart in São Paulo, where the market is more confined, compact and fluid. Factions in Rio de Janeiro (CV, ADA and TCP) are built on local hierarchies and horizontal alliances, whereas PCC in São Paulo operates in horizontal, supralocal lines. While segmentations and allegiances are determined by 'friendships' among leaders in Rio, all partakers in ' crime' in São Paulo are 'tuned' in a same movimento. Last but not least, slaughters, murders, and prison negotiations appear differently in each city.
The result is that confrontations (between factions and with the Police) are more lethal in Rio, whereas PCC applies a 'white flag' policy and institutes the 'peace among robbers' inside and outside prisons in São Paulo. Internally, what seems decisive and alike in the comandos' compositions and policies is the tension between lines that seek to establish centralizing hierarchies and lines that tend towards more horizontal relations. However, both the conflagrated armed disputes over territory control in Rio and the disperse surreptitious conflicts in São Paulo have long been responsible for determining the mechanics of extermination of the poor population in the two cities.
The relationship with law enforcement is similar in both cities, but they incur more intra-factional and inter-factional conflicts in Rio than in São Paulo; they give rise to profit-oriented conflicts over territory that affect segregation by faction in prisons in the former, whereas are carried out through negotiations from inside prisons followed by demonstrations of power in the latter. To us, it is not important to figure out which police is more submerged in the game of political commodity, but rather, what different agencements are formed in these relations. Two aspects are important: On the one side, political markets are a way of allowing illicit markets to operate, as they warrant a fluid circulation of prohibited goods. On the other side, they enhance the segmentation of criminal collectives and their willingness to fight against police officers, raising prospective and effective tensions that are often captured by the media.
As explained above, there are differences of scale between drug retail markets in Rio and São Paulo and they are echoed in the political commodities market: political assets are negotiated at the retail level in São Paulo and at a wholesale level in Rio. 27 The disruption of these negotiations has different impacts in each city's political and economic 'balance'. It remains to be seen how these arrangements will be reorganized in light of the recent proposals -sanctioned at the election polls -of implementing barbarism as public security policies.