Collusion and Cynicism at the Urban Margins

This article examines the clandestine connections between participants in the illicit drug trade and members of state security forces to understand how they impact everyday understandings of the law. Drawing on a unique combination of long-term ethnographic fieldwork in a poor, high-crime district in Argentina and wiretapped conversations drawn from a court case involving a drug trafficking group active in the same area, we find that traffickers use illicit relationships to maintain economic control of the territory, and that collusion fosters widespread cynicism about law enforcement among residents. This article expands the literature on the covert relationships between drug trade participants and agents of the state by detailing the inner workings of collusion. Furthermore, it analyzes residents’ perceptions of police complicity as an underexplored source of legal cynicism. Finally, it offers a methodological blueprint of how to access and analyze data that capture state actions usually hidden from public view.

During the first decades of the twenty-first century, most Latin American countries have witnessed an increase in urban violence, making Latin America the only region in the world where lethal violence (as measured in homicide rates) is still growing without being at war (UNDP 2013;Bourgois 2015;Koonings and Kruijt 2015;Cruz 2016;Menjívar and Walsh 2017;Santamaría and Carey 2017;Bergman 2018). Analysts agree that this violence is not evenly distributed socially or geographically but instead concentrates in the territories where the urban poor dwell, known as favelas, colonias, barrios, comunas, or villas in different countries (Moser and McIlwaine 2004;Wilding 2010;Rodgers, Beall, and Kanbur 2012;Penglase 2014;Larkins 2015;Salahub, Gottsbacher, and de Boer 2018).
Social scientific studies point to a number of factors associated with this increasingly ubiquitous character of violence in low-income neighborhoods: poverty, unemployment, inequality, the accumulation of structural disadvantages, the lack of social cohesion and informal social control ("collective efficacy"), and the twin influence of the illicit drug trade and the fragile legitimacy of the state's monopoly of violence (Sampson and Wilson 1995;Sampson, Raudenbush, and Earls 1997;Morenoff, Sampson, and Raudenbush 2001;Ousey and Lee 2002;Imbusch, Misse and Carrión 2011;Denyer Willis 2015;Cruz 2016;Durán-Martínez 2017;Bergman 2018). This article focuses on these last two factors to scrutinize the clandestine connections between drug traffickers and members of the state security forces and poor people's shared understandings of law enforcement. How does police-criminal collusion impact the ways in which residents of a marginalized area perceive the state's actions and legitimacy?
To answer this question, this article draws on a unique combination of ethnographic evidence collected over thirty months of fieldwork in a poor, high-crime district in Buenos Aires, and documentary evidence, including wiretapped conversations culled from the indictment of a drug-trafficking group operating in the same area. Based on this material, we document relationships of collusion between state agents and drug traffickers and how they impact everyday understandings of the law. First, we find that traffickers use their illicit relationships with police agents to seek an economic monopoly over the drug trade. Second, we demonstrate how police-criminal collusion fosters "legal cynicism" among residents of lowincome neighborhoods, cultivating the shared belief that law enforcement agents are not only "illegitimate, unresponsive, and ill equipped to ensure public safety," but also complicit with criminal groups Papachristos 2011, 1191). By turning our analytic attention from the overt to the covert actions of state actors, we expand our understandings of the inner workings of the opaque relationships that they develop with drug market participants. We also contribute to literature on the social responses to violence by specifying an additional source of legal cynicism: the illicit involvement of state actors in the local drug trade.

The State at the Urban Margins
When examining unprecedented levels of interpersonal violence in marginalized urban spaces, given that state structures "operate through, and in turn, transform the ordering of surrounding networks of social relationships and practices" (Clemens 2016, 94), political sociologists are compelled to ask: How is the state intervening at the urban margins? How does it police poor neighborhoods? And how do materially and symbolically dispossessed residents understand this policing?
In US urban sociology, answers to these questions have been dominated by two main images: the state's neglect of poor areas and its punishing presence. Two decades ago, social scientific analysis depicted marginalized urban areas as governance voids deserted by the state (Williams 1992;Anderson 1999; for a review, see Soss and Weaver 2017). More recently, many of these same areas have been described as militarized spaces firmly controlled by the state's iron fist (Goffman 2009;Rios 2010). In Latin America, answers to similar questions have been highly influenced by O'Donnell's notion of "brown areas": "neofeudalized regions" where legality is obliterated and where residents enjoy, at best, a "low intensity citizenship" (O'Donnell 1993(O'Donnell , 1359(O'Donnell , 1355. Shantytown dwellers are just one of the categories of people who are "often are unable to receive fair treatment in the courts, or to obtain from state agencies services to which they are entitled, or to be safe from political violence" (O'Donnell 1993(O'Donnell , 1361. At the Latin American urban margins, analysts inspired by this framework note, the poor meet a state that routinely breaks the rule of law (Brinks 2008;Koonings and Kruijt 2009;Goldstein 2012;Penglase 2014) and frequently violates human rights (CELS 2012(CELS , 2016. Although both sets of literature illuminate important aspects of the interactions between the state and its poor citizens, neither in the United States nor in Latin America has much attention been given to a mode of state intervention that engages with illegal actors and takes part in illicit actions as part of its daily operation. Those living at the urban margins are sometimes "abandoned" by the state and sometimes overpoliced (and brutalized) by it. But they are also subjected to what Arias describes as "police-criminal collusion": an "active political constellation" of state and illicit actors that not only erodes the rule of law, but also institutes "a separate, localized, order" (Arias 2006a, 324, 295; see also Arias 2006b; Arias and Barnes 2017).

Clandestine Links and the Drug Trade
Historians and historical sociologists (Gunst 1995;Roldán 2002;Astorga 2005), political scientists (Brass 1997;Wilkinson 2004;Arias 2006a), and anthropologists (Das 1990;Kakar 1996) have documented the variety of clandestine links between state officials and criminal organizations. The Italian notion of intreccio best captures these illicit liaisons, referring to the "dense interweaving" between Mafiosi and the official state actors and political elites who take measures to protect them. The term draws attention to "more than a simple reciprocity between the mafia and the state," but rather "a vast gray area where it is impossible to determine where one leaves off and the other begins" (Schneider and Schneider 2003, 34; see also Arlacchi 1983;Schneider 2018).
Recent scholarship on the relationships between drug trafficking, state actions, and urban violence also highlights collusion as a key factor in the growth and consolidation of illicit market organizations (Gambetta 1993). Most of this literature has focused on a particular form of collusion that provides illicit protection from state action. What Snyder and Durán-Martínez (2009) call "state-sponsored protection rackets" are key to understanding the relationship between illegality and violence (Varese 2004(Varese , 2014. A state-sponsored protection racket is an informal institution "through which public officials refrain from enforcing the law, or alternatively, enforce it selectively against the rivals of a criminal organization, in exchange for a share of the profits generated by the organization" (Snyder and Durán-Martínez 2009, 254). By constructing these informal institutions, state actors offer extralegal actors protection from both competitors and the state itself in exchange for a share of their profits (Snyder 2006; see also Dewey 2012Dewey , 2017b. While these clandestine connections were previously theorized as outright corruption or the result of the state's inability or unwillingness to enforce the law, research suggests that collusive relationships are complex and multifaceted. For example, while state-sponsored protection rackets may imply a unidirectional protection relationship (i.e., the state protects), Varese (2011Varese ( , 2014 finds that political authorities and criminal networks develop strategic interactions to utilize each other's unique access to resources. While police maintain the ability to enforce the rule of law, extralegal actors enjoy greater freedom to act outside that framework. Overall, this literature suggests that collusion is reciprocal and contingent (Staniland 2012). Yet few studies focus on how authorities tinker with law enforcement in real time and space and how these relationships may impact perceptions of law enforcement in communities that may suspect or bear witness to collusion.

Methods
This article draws on a unique combination of ethnographic and archival data. 1 First, ethnographic fieldwork was conducted over thirty months in Arquitecto Tucci, a poor and violent district located on the outskirts of the city of Buenos Aires. Details about the field research can be found in Auyero and Berti (2015). In 2017, a research assistant conducted fourteen additional interviews with five political brokers, the local priest, and eight other residents in the neighborhood. These new interviews examined personal experiences with violence, addiction, rumors about corrupt police, and the ways individuals navigated nearby drug-selling operations.
Second, we accessed the judicial proceedings of a court case indicting members of a drug trafficking group and two police agents working in the same district as our ethnographic fieldwork. 2 We paid particular attention to transcriptions of wiretapped phone conversations nested within court records. According to Campana and Varese (2012, 27), these constitute a rich but "relatively neglected source of data" (see also Natarajan 2006;Campana 2011;Berlusconi 2013). As they assert, "data drawn from wiretap records have the advantage of capturing conversations as they occur in their 'natural' setting and may yield a fuller picture of the group, including conversations involving lower-level and upper-level actors" (Campana and Varese 2012, 15). And, we would add, between extralegal actors and members of state security forces. 3 Campana and Varese (2012) outline three prerequisites for minimizing bias and potential sources of error with wiretapped conversations: coverage, duration, and self-censorship. The wiretapped conversations we analyze meet all but one of these conditions. Our data include a reasonably wide coverage of actors and activities over a six-month period. However, given that members of the drug-trafficking organization suspect they are being surveilled, the wiretappings do not fulfill the prerequisite of no self-censorship. Although some traffickers seem aware of surveillance and censor their conversations accordingly, we interpret this as evidence to better understand the microdynamics of collusion itself. Moreover, while recording conversations without consent is inconsistent with the ethical standards of qualitative data collection, we use these transcripts not as data produced through in-depth interviews but as a legal archive produced by the state. To our knowledge, we are the first to combine ethnographic data with evidence culled from judicial proceedings and wiretappings, and we aim to show, in practice, the virtues of such a methodological combination.
We analyzed our data using open and focused coding (Emerson, Fretz, and Shaw 1995). For our ethnographic data, we recoded our field notes with attention to residents' perceptions of and interactions with police, their collective responses to violence, and their descriptions and evaluations of illicit relationships between police agents and participants in the drug trade, or what they called the arreglo or arrangement. For the judicial documents, we concentrated on both the activities of the trafficking group and its relationships with security forces. When violent episodes were mentioned, we triangulated these references with news articles for confirmation and contextual details. Finally, the court case contains a plethora of information that is not central to our substantive focus. In what follows, we present an analytic reconstruction that focuses on the clandestine connections between police and drug traffickers in Arquitecto Tucci. 4

Arquitecto Tucci
Arquitecto Tucci (population 180,000) sits in the southern part of metropolitan Buenos Aires, the capital of Argentina. Nearly two-third of residents lived in poverty, more than double the average poverty rate of the state (33 percent in 2014). The informal labor market contributes to most household incomes in the area. Jobs in construction, domestic service, street commerce, textile manufacturing, and scavenging are the occupations most frequently reported by the local population. Overall, extreme levels of infrastructural deprivation characterize the area: unpaved streets, open-air sewers, broken sidewalks, scarce lighting, and random garbage collection.
Arquitecto Tucci is an increasingly violent place. Homicides have quadrupled since 2007. There were 17 recorded murders in 2007, 32 in 2009, 54 in 2011, 59 in 2013, and 65 in 2015 (population growth was roughly 5 percent). Other equally poor urban areas in Buenos Aires have not experienced such skyrocketing rates of violence. 5 The murder rate (38.6 per 100,000) in the district is five times that of the state of Buenos Aires. Concurrently, news articles and qualitative fieldwork suggest that groups that store, prepare, and distribute drugs have expanded their operations into Arquitecto Tucci.
Criminal activity and violence are two of the main concerns cited by residents of Arquitecto Tucci. In a series of one hundred short interviews (see Auyero and Berti 2015), an overwhelming majority of residents report crime, safety, robberies, and drug dealing as their main preoccupations. We now turn our attention to one of these trafficking groups and its clandestine connections to the local police.

Los Vagones
In 2016, eleven people were indicted for a series of drug-related charges in the district of Arquitecto Tucci. These included nine civilians (three women and six men) and two members of the Buenos Aires police force who were working with a drug-trafficking group that we call Los Vagones (the train cars). 6 Court documents confirm that at least five more people were likely involved in the illicit activities but not indicted in the case.
Los Vagones is a drug trafficking and distribution organization that operates in the southern parts of the state of Buenos Aires. Although it occasionally smuggles marijuana from Paraguay, the organization primarily obtains drugs from nearby slums. According to court documents, the organization sells ten kilograms of marijuana and thousands of doses of paco 7 each day from a network of selling points called bunkers. The most important ones and the target of this court case are located in an area of Arquitecto Tucci near a train station known locally as the Tracks.
Los Vagones is led by three men-Pepe, Fifí, and Chato-and has a hierarchical structure with clearly differentiated roles. The group consists of those in charge of transportation (preparing vehicles and moving drugs), storage (of drugs, weapons, and other assets), and sales (to other gangs, dealers, and consumers). Leaders also control a network of security agents made up of lookouts called "spies" or "satellites" (mirillas, satelites, punteros) and guards called "hitmen" (sicarios), who threaten and in two documented cases even kill their competitors. 4 We are the first to admit that court cases do not afford an unobstructed view into the secrets of state power. Much critical data remains inaccessible. We are also deeply aware of the fact that some court cases, or parts of their evidence, might be the very outcome of a jockeying for power among rival security agencies, politicians, and actors in the judiciary. Although shadowy turf wars within the state are beyond our purview, it would be naïve for us not to consider the obvious fact that many a case could be the result of a dispute between "pieces of the state" (Schneider and Schneider 2003, 34). This does not make our evidence irredeemably "fabricated." Rather, it points to its inevitably partial character. Having reviewed several cases in different parts of Argentina and involving different branches of the judiciary and various security forces, we are confident in the validity of our evidence of the patterned ways in which clandestine connections operate. 5 On the geography of homicides in Argentina, see Sol Amaya and Bianca Pallaro, "El mapa de los homicidios en la Argentina," La Nación, November 16, 2017, https://www.lanacion.com.ar/2082534-el-mapa-de-los-homicidios-en-la-argentina. 6 Multiple different security forces operate in the state of Buenos Aires. For detailed descriptions of the structure of the state police known as la bonaerense, see Eaton (2008) and Saín (2004Saín ( , 2008. 7 Known as bazuco in Colombia, baserolo in Ecuador, and mono in Chile, paco is a cheap, highly addictive form of cocaine. On the way in which the spread of paco relates to the transformation of Argentina from a country of transshipment into one of consumption and production of cocaine, see the interesting report by Guy Taylor (2008).
One important fact "leaks out"-as historian Carlo Ginzburg (1989) would say-of the formal judicial proceedings: with the help of their two contacts in the police, these traffickers spend a considerable amount of time and effort avoiding detection by state security agents. Members of Los Vagones regularly confront attempts to surveil their activities and disrupt their drug shipments and sales, attesting to the fact that the group does not enjoy full police protection. If it did, members would not have to spend so much time and energy cultivating their few trusted contacts within the police.
As in cases documented in Russia, China, Mexico, and Brazil, police officers sell protection from the actions of the same police force to which they belong (Arias and Barnes 2017;Flores Pérez 2014;Stephenson 2017). They also sell protection in the form of what Dewey (2012, 5) calls "non-enforcement, or de facto suspension of the law." In addition to protection and nonenforcement, our data show that these actors develop a "complex web of interdependencies" that impacts life in the district (Stephenson 2017, 422). For example, police officers sell traffickers guns and ammunition, both of which are highly regulated in Argentina. Police also provide information to the leadership of Los Vagones. The exchange of information can be grouped into two general types: protective information serves to shelter traffickers from police actions and competitive information provides insights into the operation of illicit markets.
According to wiretapped conversations, members of Los Vagones mostly discuss drugs, distribution, payment, and security among themselves. Money is also a common topic of conversation, including how much they make and what they spend it on. During the period of surveillance, the group made between ARS$50,000 (US$3,000) and ARS$100,000 (US$6,000) per day. These mundane conversations also detail the organization's everyday expenses. In one conversation, a trafficker called a leader to report on the previous night's work. He said, "[We gave] $300 (gambas) to each of the seven lookouts (mirillas), $1,000 to Shrek and Tati, and I kept $2,000." Payments made to police officers are another common and recurring expense. One trafficker documented all his expenses to a leader over the phone: "I have $43,150 [left] because I used $400 yesterday … for the cops [cobanis]." In another conversation, a trafficker said, "I need to take $10,000 and send Cesar and Gorda to the place where we give the thing to the cops. I'll send them now because they are waiting." While many conversations are retrospective, traffickers also discuss future bribes. For example, Fifi told Chato that "the cops called" and wanted to know if they could "do it today" (i.e., get their bribe money). He then asked Chato about the amount: "Do I give them ten pesos [referring to $10,000]?" Chato said yes. Fifi then told him to meet La Morocha, the female police officer indicted in the case, at 5 p.m. at the "usual place." Three findings are critical to understanding how collusion shapes life in Arquitecto Tucci, particularly in relation to what analysts call "systemic violence." 8 First, the case clearly documents the violent tactics used by Los Vagones to acquire and maintain dominance over the area by forcibly eliminating competition. Second, it shows how the police engage in and facilitate this violence. Finally, it exposes the negotiated and sometimes volatile nature of these illicit relationships by documenting a limiting factor of collusion.

Making their presence known
Like other drug trafficking groups around the globe (Hazen and Rodgers 2014), members of Los Vagones spend significant time evading state security agents. They regularly modify their work routines and sometimes completely suspend activities to avoid interactions with state agents. In a wiretapped conversation, a trafficker told a supplier that he had to temporarily suspend his delivery because the National Guard was on the highway. This interaction highlights an important point: traffickers can be protected by one arm of law enforcement and pursued by another arm of the same state. While different state security forces sometimes work together, they can also act against each other (Barrera 2013;Hathazy 2016). 9 8 Systemic violence refers to the violence that arises "from the exigencies of working or doing business in an illicit market-a context in which the monetary stakes can be enormous but where the economic actors have no recourse to the legal system to resolve disputes" (Goldstein 1985, 116). In principle, participants in the drug trade cannot rely on authorities to manage their grievances or to solve their disputes, so they engage in informal (and sometimes violent) forms of social control (Andreas and Wallman 2009;Reuter 2009). Yet scholars agree that violent outcomes are not inherent to illegal actions, commodities, or markets (Friman 2009;Naylor 2009;Snyder and Durán-Martínez 2009;Williams 2009). Rather, the relationship between illegality and violence is mediated by if, when, and how the state intervenes (Durán-Martínez 2015; Vargas 2016). 9 In Argentina, there are four national security forces: the Federal Police, whose purpose is the prevention and repression of federal crimes according to the penal code; the National Guard, with police functions at national borders and on interstate roads; the Argentine Naval Prefecture, with police functions over rivers and borders, and the Airport Security Police, with control over aircraft and airports. Given the federal nature of the Argentine Constitution, each state (province) has the power to organize its own security force to police ordinary crimes (i.e., those not defined as "federal" by the penal code) (Barrera 2013). During the past two decades, the field of policing has experienced some important changes (Hathazy 2016); prominent among them is the militarization of the National Guard and its growth at the expense of both the federal police (la federal) and the police of the province of Buenos Aires (la bonaerense).
Members of Los Vagones also discuss quotidian ways to avoid detection; for example, the need to change cell phones and numbers to avoid surveillance. Wiretapped conversations reveal that they not only work to evade state officials, but they also avoid other illicit actors who target their profitable business. Late one night, a trafficker decided not to deliver drugs because he didn't feel safe: "I'm scared because it's late," he admitted on the phone, referring to his fear of what Contreras (2012) calls "stick-up kids," or groups of young people who attempt to rob drug dealers.
Traffickers also carefully monitor their territory in Arquitecto Tucci, including nearby competitors and residents. For instance, one dealer informed Fifi that a man named Raulito was selling drugs in their area. Fifi instructed, "Go to him and tell him I said that he has fifteen minutes to go somewhere else, otherwise his legs will be broken. Tell him Tío and I said that. Call me back in fifteen to let me know he left." As this conversation suggests, leaders do not hesitate to threaten the use of violence to maintain their economic control over the territory. In addition to surveilling state agents, petty criminals, and competitors, members of Los Vagones also sometimes intimidate residents to keep them from talking to authorities about the traffickers' illicit activities. In one phone conversation, a trafficker threatened a man who he suspected had passed information to the police. The trafficker said, "We don't want to have any problems with you." He then offered him a job with the organization, making clear he would eventually discover the snitch.

A violent business
Los Vagones not only verbally threatens residents and competitors but also arms its members, attacks rival bunkers, and even kills other drug traffickers. Wiretapped conversations attest that members of Los Vagones are regularly armed with lethal weapons. For example, one trafficker informed Chato: "At the train station there's a group [un junta]. I think they are selling." Chato decided to send hitmen to intimidate the group. He directed the trafficker to send Saverio, a hitman, with a .38 (handgun), and if needed, to let Japonés and Tigre know about the incident so that they could "do what they gotta do." In another conversation, a trafficker told a leader about his plan to confront a competitor at a rival bunker: "They are there until 11 p.m. We'll visit him around nine or ten." The leader replied, "Yes, that's good … fire some shots [mándenle cohete], break his knees." The trafficker informed the leader that he also planned to loot the bunker: "They are in the house. We'll go in and take whatever they have." The leader, adamant about the use of physical force, reiterated, "Yes, but break his knees. If not, tomorrow he will be back doing the same thing" (our emphasis).
In the most extreme example documented in the court case, Mateo, an informant in witness protection, provided information about the involvement of Los Vagones in a double homicide. According to his testimony, Pepe sent four hitmen to kill members of another trafficking group planning to do business in Arquitecto Tucci. That night, they attacked a bunker run by what Pepe called "addicts" (fisuras) from Paraguay. Two men were killed and Pepe forcibly took over bunker, expanding his business to the new selling point by eliminating his competition.
Judicial proceedings provide clear evidence that this violence is supported by collusive relationships between traffickers and local police agents. First, traffickers obtain their guns (called toys, tools, or girls) and ammunition from members of the police force. In the double homicide described above, guns registered with the police department were used to commit the murders. Second, Los Vagones regularly receives protective and competitive information from the police. For example, police officers provide the group with information about other law enforcement agents who are surveilling them. On a wiretapped cell phone, Lucho (one of the police officers indicted in the case) spoke with Fifi. Lucho asked, "Are they [police officers] bothering you?" Fifi replied, "Yes, [they were] the other day but not yesterday or today." Lucho then promised to help Fifi with the problem: "They are someone else's," he explained, indicating that the police officers were from another agency. "They are from the other side of the tracks. I will find out who they are." In a subsequent conversation, Lucho warned Fifi, "You guys are being marked [surveilled] by the street patrol." He gave Fifi more details: "Write down the license plate number of the car … XL5-C94. It's a white Nissan. They were told that people were bringing stuff around where you are." As these conversations attest, protective information is regularly passed between police officers and traffickers. This information exchange is reciprocal: police officers also receive information from traffickers about other drug dealers in the area. For example, a little after noon one day, Fifi called Lucho. "What's up?" Lucho said when he answered the phone. Fifi replied, "The same as usual … those fucking smurfs [pitufos de mierda, referring to police officers] have been bothering us for a while." Lucho explained what he thought was happening: "Okay let me explain. It [the problem] is not going to be for long, you see? They opened a new police station [habilitaron ahi la base] … and they sent them to put on a show [hacer un poco de espamento]. But they aren't going to fuck around … for long." Fifi responded, "Okay, we can wait. Hopefully that happens so we can get back to work. But there is also competition." "Leave that to me," Lucho assured him. "I'll deal with them. I'll go fuck with them a bit." Fifi then expressed the importance of Lucho's intervention: "It will be a huge win [un golazo] if you can mess with them." When Lucho mentioned he was unfamiliar with the area, Fifi went on to provide specific details about where competitors were doing business. As this conversation shows, traffickers not only use violence to eliminate competition, but they also rely on their contacts in the police to provide competitive information and even take out competitors. Put another way, traffickers pay not only for goods and information but also for direct police intervention on their behalf-in this case, to maintain an economic monopoly on their territory with police help.

Profits and payments: The limits of protection
The clandestine relations between traffickers and police agents are not established and negotiated without conflict. Although Los Vagones makes large sums of money selling drugs, these profits fluctuate based on many factors, from issues of supply and demand to the variable costs of labor. One issue stood out for the traffickers: their ability to actually do business. The court case shows that traffickers often respond to protective information not only by modifying their work routines but by stalling shipments, suspending sales, and ultimately reducing time spent making money. As a result, traffickers negotiate their weekly payments for protection based on their cash on hand. In a wiretapped phone conversation, a leader of Los Vagones told Lucho: "When your friends [referring to the police officers] call you … your little friends who want to be paid … tell them that we are not going to give them what we usually give them. Tell them that we are going to give them only eight [thousand]." In another conversation with a police officer, a leader of Los Vagones explained his decision to reduce the weekly bribe: "Why?" he asked hypothetically. "Because your friends are not letting us work during the day or at night. We can't work like this. If you want to earn the money you used to, well, I have to earn mine and those people are not letting me do my job. See what you can do about it." As this conversation suggests, traffickers regularly negotiate the terms of their clandestine relationships with police officers. Although we do not know the outcome of this interaction, this collection of wiretapped conversations shows that clandestine relationships are negotiated affairs susceptible to shifts, renegotiations, and sudden breakdowns.

Perceptions of Collusion
Residents of Arquitecto Tucci are widely aware of the relationships of collusion between traffickers and police. Over the course of our ethnographic fieldwork, many expressed a general feeling that the police were not working for them. For example, Cintia (eighteen) recalled a time when her house was robbed and then asked rhetorically: "Why would you call the police? What for? They don't do anything. You make the complaint. They take the complaint and they file it away." Melissa (sixteen) put it this way: "You call them [the police] fifty times … They never come. And when they come, they ask you a bunch of questions and don't do anything." Mabel (thirty-seven) summarized police ineffectiveness like this: "The cops are always late … to collect the body if someone was killed or to stich you up if you've been raped." Residents do not simply perceive the police as incompetent but also as highly extortive: "I was driving my motorcycle and the cops stopped me," Juan (eighteen) said one day. "They [police agents] took all my money. You have to give them money otherwise they take your motorcycle. They want free money [quieren unos mangos de arriba]," he explained.
Accounts of extortion and incompetence by the police often merged with descriptions of their illicit relations with criminals and traffickers. For many residents, police officers are not that different from drug dealers. For example, Rosa and Cata had both sold drugs in the past. Reflecting on their experience, they confidently explained: "Being a dealer [transa] is less risky than being a thief. You always have a cop protecting you. It's quite easy. They come to pick up the bribe. Cops go to a pizza place and they get two pizzas for free. They go to the butcher and they get two chickens for free. And they go to the bunker and they get 500 or 600 pesos every night." Police officers, they said, also charge to release people from jail, levying an under-the-table bail to profit off those they arrest. "For the petty thieves, they charge you 500 pesos," Rosa explained. "For the important ones, at least 1,000 pesos … only those who don't have money go to jail." As these accounts attest, residents in Arquitecto Tucci share a general perception that police do more to facilitate the drug trade and to line their own pockets through extortion than to protect residents. "Half a block from my house," a longtime resident said, "there's a bunker where they sell paco. Everybody knows that you can find it there. The cops know, but nobody does a thing. Neither the neighbors nor the cops do a thing about it … Nobody wants to get in trouble." Not missing a beat, a friend added, "The cops have their deals. They know everything. But they only do something when it is in their favor, or when someone else benefits … you know what I mean?" Shared understandings of collusion were clearly on display at a community meeting that residents organized in February 2013. In a small room with twenty-five other neighbors, Nerina, a grassroots activist, opened the meeting and summarized the impetus for the gathering: "There have been many deaths in the neighborhood recently and the police are nowhere to be found … there are a lot of kids who are consuming paco, and lots of drug dealers [transas]." Some of the residents in attendance were also members of the local chapter of Mothers against Paco, a grassroots organization that gathers parents of those addicted to paco. Isabel, the most outspoken leader of the group, shared: "Things are quite messed up. I don't want more police, or mano dura … I want the police to do their job. We all know where the dealers are … we have insecurity because the kids who are on drugs steal in order to buy." Alicia, another leader of Mothers against Paco, brought up the issue of nonenforcement: "They [the police] know where they [the dealers] are but they don't do anything." Other attendees expressed similar distrust towards the police: "Everybody knows where the dealers live, and everybody knows that the police are in cahoots with them … and we are afraid that if we report the dealers, we'll suffer the consequences." Despite fears of retaliation, residents attend community meetings, participate in groups like Mothers against Paco, and occasionally participate in public protests. But the widespread perception of collusion can also translate into inaction such as the refusal to speak up or to go to the police to make a complaint or report a crime. ("It's better if you don't make a complaint at the local precinct. You will get in trouble." "You will be forced out of the neighborhood.") At other times, this cynicism with respect to law enforcement motivates residents to take matters into their own hands. Rosa is a case in point. She recounted an episode that took place a few months before we interviewed her involving a group of young people who had squatted in a house on her block and were using it to sell drugs. "What could we do about it? Those kids were quite aggressive," Rosa said, "and the cops knew them. There is no point in reporting them to the police." She then explained what happened at a community meeting called to address the squatters and their activities: "We had meetings with the neighbors about the drug problem. The police chief came. And so did the son of the trafficker, the wife of the trafficker … All those who attended the meeting got threats. One of the neighbors told the chief the location of the selling point and she received a death threat. Her house was stoned. That meeting was a sham, the chief and the street cops are complicit with the dealers … If they put 10,000 police cars on the streets, half are going to be complicit with them." Sensing the lack of legal alternatives, Rosa organized a group of local youth to forcibly remove the squatters. She remembered, "It was pretty rough … we took all the things from inside the house … it was violent … but we got rid of them." A few months later, her daughter's purse was snatched by a kid riding a motorcycle. Rosa knew the thief ("he sells drugs for a dealer who has an arrangement with the cops"). Rather than call the police, she went with her husband to his house to demand the perpetrator return the purse. As Rosa's experiences suggest, residents resort to alternative ways to maintain social control, sometimes organizing collective action to solve community problems (as in the small gang she gathered to evict the dealers) and other times relying on self-help to address their grievances (as in threats made to retrieve the purse).
We do not mean to suggest that all residents share the same understandings of the illicit relations between drug traffickers and the police. The suspicions and concerns that we have described are necessarily interpreted through an individual's particular experiences. While suspicions about the form and extent of collusion may vary, our long-term fieldwork confirms that a general perception of police incompetence, extortion, and collusion with drug traffickers was widespread among residents in Arquitecto Tucci.

Discussion
What does all this tell us about the workings of the state and its clandestine relationships with drugtrafficking organizations? How do these relationships shape residents' views of law enforcement? We contend that set of interactions between Los Vagones and members of the state repressive apparatus give concrete shape "to what would otherwise be an abstraction ('the state')" (Gupta 1995, 378). The state that emerges out of these interactions is neither "weak"-as in descriptions of poor neighborhoods as areas abandoned by the state (Anderson 1999;Venkatesh 2008) or "governance voids" (Koonings and Kruijt 2007)-nor "strong," as in descriptions of poor neighborhoods as highly militarized spaces firmly controlled by the state's iron fist (Rios 2010;Müller 2012). Rather, the set of clandestine interactions described above show that the state is a deeply ambivalent organization. 10 Judicial proceedings illustrate the actions of an ambivalent state with two coexisting and seemingly contrary qualities, acting simultaneously as an enforcer of the rule of law and an accomplice to criminal acts (for similar arguments, see Arias 2006b;Rodgers 2006;Denyer Willis 2009Arias and Goldstein 2010;Jaffe 2013).
Building on work examining the relationships between criminal organizations and the state, our reconstruction of these clandestine connections shows protection and concealment at work as police officers provide goods and information to drug traffickers in exchange for money. In describing these contingent and reciprocal exchanges, our evidence further details the inner workings of state-sponsored protection rackets at the urban margins. We argue that traffickers use these clandestine connections to control a given territory to conduct their illicit business (preparation, storage, and distribution of drugs). They strive for this while attempting to go undetected by the police through their use of contacts inside the state repressive apparatus. For the traffickers, the state is thus both a resource to be used in their criminal enterprise and a force to be reckoned with and avoided.
We also find that residents' varying suspicions of and encounters with police-criminal collusion impact their understandings of law enforcement. Criminologists have long examined how perceptions of law enforcement shape citizen behavior, but not in the context of police-criminal collusion. Scholars developed the notion of legal cynicism to make sense of the shared belief that law enforcement agents are "illegitimate, unresponsive, and ill equipped to ensure public safety" Papachristos 2011, 1191). The term was first developed to challenge the notion that racial minorities in the US held subcultural values that tolerated deviance and thus accommodated higher levels of crime (Sampson and Bartusch 1998). Scholars have since moved from understanding legal cynicism as a form of anomie or "normlessness" that results from continual exposure to injustice, segregation, and insecurity (Anderson 1999;Nivette el al. 2015) to a cultural frame through which people observe and make sense of their social worlds (Lamont and Small 2008). Existing research shows that legal cynicism among residents has material effects on crime and violence. When people perceive the law as unavailable or unreliable, they are left to resolve their problems independently. Examining the case of Chicago, Kirk and Matsuda (2011) find that in neighborhoods with high levels of legal cynicism, residents are less likely to report crimes than in neighborhoods where people have more trust in the police. High rates of legal cynicism also help explain the persistence of homicides in Chicago despite declining rates of poverty and violence (Kirk and Papachristos 2011).
Different from this research, however, we find that residents in Arquitecto Tucci collectively develop legal cynicism not simply out of the perceived unavailability or bias of law enforcement agents, but also out of the felt (and as we presented here for the case of Los Vagones, very real) complicity between police officers and criminals. Much like the residents of Mexico City described by Davis (2006, 57), residents in Arquitecto Tucci see "only a fine line separting police from criminals," if any line at all. In other words, legal cynicism is not, as the literature suggests, simply the result of the state's inefficiency or ineptitude but also a product of its ambivalence.
Moreover, we find that poor people's legal cynicism stems not only from the suspected collaboration between police and dealers but also from the unstable features of collusion that make it so difficult for residents to figure out exactly what is going on. In other words, it is not simply the collusion but also its volatile character that makes residents believe that law enforcement agents are impassive and incapable and ultimately lack legitimacy. Given residents' (well-founded) suspicions about police officers, they cannot reliably call on them for protection from violence. As a result, those concerned with and/or directly affected by drug-related exchanges tend to avoid contacting the police. Paradoxically, drug dealers can call certain police officers to release areas from police protection in exchange for cash, thus obtaining the protection that is typically unavailable in illicit markets (Reinarman and Levine 1997;Ousey and Lee 2002).

Conclusions and Tasks Ahead
Research has well documented overt state action that polices, punishes, and disciplines the poor (Goffman 2009;Wacquant 2009;Rios 2010;Soss, Fording, and Schram 2011;Brayne 2014;Desmond, Papachristos, and Kirk 2016;Stuart 2016;Soss and Weaver 2017). Much less attention has been paid to the clandestine interventions of the state in what Wacquant (2007) calls "territories of urban relegation." Examining this form of illicit state action-which Arias (2006a) aptly calls "police-criminal collusion"-is imperative to better comprehend and explain the individual and collective responses to violence.
This article focused on one set of connections between police agents and drug dealers. As described above, Los Vagones developed ties with just two contacts in the local police force. Put in the context of our broader project on drug-trafficking organizations across Argentina, our research shows that other groups have far more extensive connections, relying not just on local police but on a patchwork of protection from different state agencies at the local, state, and federal levels (de los Santos and Lascano 2017; Tamous 2017). Further work should examine the common mechanisms of collusion as well as variation in the quantity and quality of clandestine relationships.
The combination of ethnographic and legal evidence, as used here, can be expanded to include multiple cases, scaled up to examine police-criminal collusion at the state and federal levels, and broadened to study other sorts of connections (i.e., between police officers, drug dealers, state officials, and politicians). For obvious reasons, evidence of the "actual links between state agencies and illegal actors or activities" is not easy to come by (Dewey 2017a, 15). Political ethnographers should collaborate with actors such as investigative journalists, attorneys, and state prosecutors who have access to (and are often willing to share) information on the clandestine dimensions of state action. Much like the inquisitorial trials that offered historians "an unexplored gold mine" (Ginzburg 1989, 142), judicial proceedings and investigative reports can provide a treasure trove of evidence on the actual workings of the state at the urban margins, especially when read in light of ethnographic material.