Street Urban Peace in Contested Informalities: The Hidden Face of Colombia’s War on Drugs

Law enforcement interventions in drug markets require policy coordination to prevent collateral outcomes that might harm vulnerable people under criminal control and spread crime across places. This paper analyzes street-level peace building in the inner city at the frontline of the ongoing urban war on drugs in Colombia. Building on the emergent literature on criminal governance and using crime script models, the paper argues that behind the chaos that in appearance prevails in open-air drug markets, illegal economies in Colombia’s urban centers are ruled by schemes that set parallel mechanisms of order, social control, and distribution. Furthermore, the paper addresses the ways in which street-level bureaucracies and therapeutic policing interventions can become a way of building urban proximity, connectedness, and trust in the context of contested informalities.


I. Introduction
In 2016 Colombian authorities cracked down the largest open-air drug market in the world's top cocaineproducing country. Known as 'El Bronx', the place hosted 3,000 crack homeless drug users every day, but it also was a hub for child sex trafficking, kidnap brokering, and a stolen goods trade. Using a therapeutic policing approach, the intervention was conducted to rescue vulnerable populations under illicit economy exploitation, regain control of a ' don't go' area in downtown Bogotá, and revitalize the crime-ridden neighborhood. The focus on social attention and urban revitalization rather than zero tolerance goals, such as imprisoning drug market participants or winning the war on drugs, was paramount in the sustainability and success of the intervention.
The paper poses two main contributions to the debate of illicit economies and urban peace. It states that criminal governance is not always a function of state weakness and that it can actually be established in low-crime equilibrium cities, such as Bogotá. Extralegal actors set parallel mechanisms of social control when crime is normalized and state-sponsored protection rackets perform adequately. On the other hand, it argues and shows how public health and citizen security approaches to the problem of drug markets can be situationally blended in urban ' do no harm' interventions. In this particular case, the main objective of the intervention was not to end drug supply in downtown Bogotá, but rather to rescue vulnerable populations and to regain territorial control. Therapeutic policing, a street level combination of policing and social work, was the way in which the state built local capabilities to connect social policy with drug users and homeless citizens while allocating resources towards urban revitalization.
The paper develops in three moments. In the first, I describe the social order of El Bronx that shaped the architecture of the illegal market and how different actors, including legal and extralegal ones, participated in it. The section is not only focused on the structure and roles of the market but, most importantly, in the patterns of behavior and mechanisms through which the illegal operation endure in downtown Bogotá.
In the second moment, I describe the logic of the intervention: local gangs' efforts to adapt to a completely new situation and to the street level model that was implemented. Here, I focus on the mechanisms through which state-sponsored protection rackets enhanced criminal governance and the way in which this type of governance was disrupted. Amid this process, contested informalities, defined as the disputed character of everyday economic hustle of people living under criminal control, remain a crucial and unexplored area of criminal governance.
Finally, the paper approaches therapeutic policing as a peace building process at the street level and as a mechanism to combine citizen security and social policy approaches into ' do no harm' interventions. The conclusion discusses the lessons learned from the El Bronx intervention and how they can inform the selected responses to tackle illegal economies in other urban contexts.

Data and Methods
The present paper is not a public policy or impact evaluation. It falls within the crime analysis field of applied research in criminology and criminal justice to address immediate policy needs. In doing so, crime analysis is a systematic, analytic process aimed to provide information about crime patterns that help to understand who is doing what with whom by focusing on the relationship between persons and organizations involved in illegal activities (Hagan 2018). The analysis proceeds as follows. Two years of participant observation were conducted in Bogota's central district of Mártires, including neighborhoods adjacent to El Bronx. One of these two years of participant observation elapsed after state interventions, assessing the aftermath and response of extra-legal organizations to the loss of territorial control. Also, semi-structured interviews were conducted with city officials, prosecutors, social workers, and community leaders following a combination of nonprobability quota and snowball sampling. Secondary analysis of open sources, such as official reports and statistics, as well as press articles, supplemented participant observation and interview data.

The urban face of the war on drugs in Colombia
For years and since the 1980s Colombia has been known as the top cocaine producer country in the world. Focused in eradicating coca leaf crops and combating trafficking, the US transferred $10 billion in bilateral foreign assistance between 2000 and 2016 to fund Plan Colombia, a strategy meant to combat left wing guerrillas, upgrade law enforcement capacities, and stabilize the security situation in a convulsed country (Beittel and Rosel 2017). Despite the recent spike in coca leaf crops registered in the last years, anti-narcotics efforts held in Colombia during the last decades largely affected main drug cartels. Anti-money laundering and interdiction efforts neutralized cartels' kingpins and struck their finances.
Following a global trend in which drug use has heightened in producing countries that have the dual condition of producer and consumer countries (UNODC 2019), drug consumption in Colombia quadrupled over the previous two decades (Camacho et al. 2016). The increment in local consumption met criminal networks' need for alternative profits after interdiction seizures incremented continuously during the same period until 2016 (Observatorio de Drogas de Colombia ODC 2017).
Amid local drug market proliferation in Colombia's main cities, inner-city retail drug markets often gained a top place in the public safety agenda. In the long run, public pressure lead to the intervention of some inner city drug markets by law enforcement and justice authorities opening a new chapter in the story of Colombia's war on drugs.

Urban violence and contested informalities
Urban violence is one of the problems of contemporary societies that demand the most resources and efforts from countries around the world whose political and economic life revolves around cities (Atkinson and Millington 2019;Kilcullen 2013). Being home to the outcast and most marginalized sections of urban populations, impoverished, unruly, and crime-ridden areas, known as skid rows, ghettos, and inner cities or slums, often lodge different types of illegal markets and criminal activities.
Paradoxically, in these places, chronic poverty and addiction tend to appear intertwined with some of the most profitable illegal markets around the world, such as drugs, street sex trade, or stolen goods. Informal and illicit economies overlapped in El Bronx for decades, converting it in the open-air drug market by antonomasia in Colombia.
By contested informalities, I refer to the struggle and challenges of people to afford living in stigmatized areas controlled by criminal groups such as gangs or cartels that depend on informal markets. A vast majority are law abiding citizens tied to jobs from the off-the-books economy in which a structure of unwritten standards define who can trade what with whom, who can work on one street corner or the other, what the prices are, and so on (Venkatesh 2006). Informal markets in these communities emerge as a shady and contested economy given that while actions are technically illegal they are not often criminalized or seen as a threat to public safety (Boel 2016;Konove 2018).
The social disorder and plight that characterizes these places remain a challenge for public authorities affecting the design of policing strategies that are deployed (Bittner 1967;Braga and Weisburd 2010;Wilcox et al. 2018). To solve the surge in crime, governments opted for the implementation of tough-on-crime policies. Following Felbab-Brown, the zero tolerance approaches to crime that become popular in the '80s brought other problems, such as human rights violations and mass incarceration (Felbab-Brown 2013).
Comprised of minimum mandatory sentencing, the stiffness of criminal codes regarding drug-related crimes, and the diffusion of crime control techniques in lower income communities, tough-on-crime policies led to systematic confinement and mass incarceration (Hinton 2016). Hence, the narrative of urban informality as a problem of criminality and lawlessness unique to the urban ghetto prevailed as the underlying conception of urban decay (Sharkey 2018).
As a result, law enforcement interventions followed by urban renewal projects gained legitimacy as a well regarded and technical solution to urban violence fueled by illegal markets, receiving vast quantities of public and private resources (Lehrer 2000). Yet, in some cases, state intervention plans have failed in their purpose to shut down illegal markets and restore governance, which poses a variety of public policy challenges related to illegal economies, policing, and urban peace. In addition and regarding the present case study, in 2015 Colombia's Constitutional Court sentenced that civil authorities must respect mendicancy when it is the expression of an autonomous decision made by any person. For that reason, homeless people in a condition of problematic consumption or addiction, can be offered treatment as far as it is consensual by the future patient. No person in Colombia can be forced into shelters or drug treatments, even if the person is under homelessness or exploitation. This legal restriction is critical in how social and drug policies overlap and shape urban drug markets in Colombia, and in the path through which the country has slowly advanced in harm-reduction urban programs despite being one of the main epicenters of the enduring global war on drugs.

Previous Literature
The first strand of relevant literature for the present study is the one on illegal markets and criminal governance. Illegal markets, particularly drug ones, not always breed violence (Andreas and Wallman 2009;Beckert and Dewey 2019;Curtis and Wendel 2007;Reuter 1986). Evidence suggests that criminal organizations exist as a form of protection given the extended uncertainty of drug markets where information problems often trigger violence and the need for increased reputation, rather than as a way of controlling offer and demand (PAHO 2010). However, recent research by Bergman (2018) shows how some cities in Latin America, such as Bogotá, can be classified as Low Crime Equilibrium cities. These are places where property crime rates might fluctuate from low to high, yet featuring low violence rates that are usually maintained by local authorities.
Similarly, in high-risk drug markets such as cocaine-based drugs like 'basuco' or crack and heroin, illegal operators have plenty of incentives to hide systemic violence (Durán-Martinez 2018;Gambetta 1993;Varese 2017). However, when things go wrong violence is used to enforce agreements (Grillo 2016;Skarpedas 2001). A moral economy of violence emerges as an encrusted feature of places where the law is rarely enforced or enforced selectively and thus everyday violence is exerted discretionally against disposable subjects (Scheper-Hughes 2015; Willis 2015).
Furthermore, the existence of monopolistic players that have enough power and violent capacity in order to display territorial control and to enforce a criminal order explains low levels of violence in illegal markets (Bowden 2014;Jacobs 1999;Levi 2007;Reuter 1985Reuter , 2009). Nevertheless, low levels of violence in the context of competitive organized crime operations, illegal markets, or places with the absence of strong statehood can also be explained by a) arrangements between extralegal groups (Idler 2019;Shortland 2019) or b) arrangements between extralegal groups and the state (Barnes 2017;Sobering and Auyero 2019;Willis 2015). This relation means that less violence does not necessarily imply less criminal activity (Aziani et. al 2019;Bergman,2018).
For instance, Snyder and Duran-Martinez (2009) argue that one of the mechanisms by which illegal markets maintain low levels of violence is the existence of state-sponsored protection rackets. Statesponsored protection rackets are informal institutions through which law enforcement and civil authorities refrain from enforcing the law or enforce it selectively against the rivals of a criminal organization, in exchange for a share of the profits generated by the criminal business (Snyder and Duran-Martinez 2009: 254). Despite the common belief that only criminals participate in criminal networks, organized crime schemes often provide incentives to institutional participants, from law enforcement to civil agencies (Salcedo and Garay 2016).
Recently, an emerging body of literature has looked at the ways in which criminal groups govern and the incentives they have to set parallel forms of governance to contend violence, deliver public goods, regulate informal economics, establish community arrangements, and impart justice (Desmond Arias 2007;Feltran 2010;Lessing 2018Lessing , 2020Trejo and Ley 2018). Some literature on criminal governance, or governance without the state in the context of organized crime, often understands criminal orders as a function of state weakness, especially in peripheral geographies where weak state institutions are not able to deliver public goods (Duncan 2014;Felbab-Brown et al. 2018;Risse 2011). In the case of Colombia, Abello and Guarneros (2014) found that criminal actors mediate the roles they played in local governance through strategies of deinstitutionalization.
The paper also builds on the concept of street-level bureaucracies (Lipsky 2010) to highlight the relationship between street-level peace building and urban transformation at the macro level. As Lipsky states, there is a significant gap between planned policies and their actual implementation at the street level. In the implementation process, public servants deal with unexpected interactions and problems that demand constant decision-making, discretion and accountability.
In the policing world, everyday policing or patrol work implies interactional definitions of deviance depending on the labeling, typifications, or perceptions that police officers have of people living in crime ridden areas (Sharp 1982;Van Maanen 1978). The nature of patrol work and the everyday interactions between drug users, criminals, and the state demands a local and rooted approach to the dynamics of urban violence and how the city is permanently negotiated through the territorial production of the politics of care and crime control (James Smith and Hall 2018).
Finally, researchers have also strained the importance of a local turn in peace building policies. They understand the local turn as a vehicle to design more effective interventions but also to empower local communities towards emancipation and transformative action (Leonardsson and Rudd 2015).

II. The Social Order of El Bronx
The set of norms that organized the architecture of the black market, as well as its primary operating illegal economies, was not the result of monopolistic actions to impose arrangements in terms of price and social order to consumers, dealers, or other participants in the market. Instead, the social order of El Bronx was a function of constant bargaining in everyday interactions between criminals and street-level bureaucracies, comprised of law enforcement agents and social workers from the city.
Furthermore, while drug trafficking was the central element of the illegal operation at El Bronx, other driving factors of crime make this place different from the conventional street corner selling spot. Similar to other cases, factors such as legal cynicism, waste accumulation, youth out of school, and social disorganization contributed to the insecurity of the place (Sampson 2012;Shaw and Mckay 1942). Entrenched in the territorial control gained in this sector of the city and in state sponsor protection rackets, local gangs (called ganchos 1 ) created a sort of liberated zone where they set up a diversified criminal portfolio strengthened by the concentration of hundreds of drug users in the same place.
In fact, Figure 1 illustrates the criminal operation at El Bronx between the years 2012 and 2016. Because 2012 was the last war among local gangs that led to gunfire exchange inside and outside of the open-air drug market, this period was considered (RCN 2012). It was also the last year in which authorities conducted a significant crackdown attempt before the 2016 comprehensive intervention (El Tiempo 2012). These events might have changed the architecture and patterns of the criminal operation that prevailed until 2011.
The illustration of the illegal operation was made with data from the Police Sectional Intelligence unit (SIPOL), field reports from local officials in charge of managing security incidents at the neighborhood level between 2012 and 2015, interviews with street-based social workers at El Bronx, and secondary data available in open sources. I organized the data using crime script models applied to analysis in criminal justice policy. Crime scripts are used to understand the process that occurs prior to the commission of a crime, specifically those classified as rent seeking that demand a social organization of resources such as drug trafficking (Chiu et al. 2011;Leclerc 2016).
The following actors were involved in the illegal operation: -Ganchos (Hooks): Denomination used by the local gangs behind the operation of the drug market and other illicit economies at El Bronx. Their name comes from getting hooked, to refer how regu-lar consumers were expected to keep loyalty and exclusiveness to a certain taquilla (selling spot). Each gang had a taquilla and its own branded basuco as a reputation marker for the quality of the drugs they distributed, mainly basuco. -Sayayines: Denomination is given to those gang-affiliated members in charge of the security of the ganchos. They were the ones who enforced social control and exerted violence within the open-air drug market and in its adjacent blocks. -Homeless Citizens: Their participation in the illegal operation was given in the role of victims. It was the product of ganchos' coercion that assigned homeless people the following tasks: a) Campanero (Whistleblower): Report to the sayayines any strange movement or person that approached the area, b) Halador (Puller): Find and take potential clients to the different taquillas depending on whom he was working for, connect people with other illegal services like buying stolen goods, escorting clients out of El Bronx to make sure they did not get mugged in the adjacent blocks, collect money and deliver drugs to people that were willing to buy but not to consume inside El Bronx, and c) Carritos (Cars): Logistic tasks of transportation of arms, money, guns, or even dead bodies.
Notwithstanding that the above mentioned were the leading operators of the El Bronx open air drug market, the whole process involved other problematic and occasional drug users, neighbor communities, and business owners, as well as state agents in the service of local gangs. Figure 1 illustrates the illegal circuit, not the geography of El Bronx. The different points of entrance to the circuit for a drug user or a client are in gray, namely a) buy drugs, b) buy or sell stolen goods, and c) access to entertainment services. Contrary to continued lines (___), the intermittent ones (….) mark the possible developments of interactions within the illegal market that, despite the evidence collected that they did happen, they can not be defined as a necessary or fixed pattern of the unlawful circuit. Given that drug trafficking was the main economy at El Bronx, buying drugs was the most common way to enter the circuit. Drugs were available for retail or wholesale in case there was a previous trusted relationship between the buyers and the ganchos. Wholesale was the way many local dealers in Bogotá stocked themselves to set up their lines of supply throughout the city.
Wholesale transaction was also the most concealed part of the illegal operation at El Bronx. Besides being a high-quality stock supply for small traffickers across the city, it offered 'safe spaces' of transaction in warehouses for storing, and provision that also worked as offices for decision makers and higher rank cadres of the ganchos. Most of the operation was coordinated in these places, including arms trafficking, bribes payments to corrupt police officers, details on selective homicides, and coordination with other local drug trafficking structures around the city. These structures functioned as satellite selling points from El Bronx.
In the case of retail sales, the ganchos offered not only 'high quality' products compared to other midrange local markets in the city but also safe spaces to use, which meant spaces out of reach of law enforcement. These places were crafted as entertainment places (pools, bars, liquors stores), cheap motels, or the public space in the case of impoverished homeless users. The payment could be made in cash or stolen goods (e.g., smartphones) or spare auto parts based on average prices established in somehow stable ways, allowing these goods to be swapped for a certain amount of bichas (basuco doses).
Regarding robbery and mugging, it is important to clarify that the ganchos controlled the handling of stolen goods inside El Bronx, not outside. While the handling of stolen goods affected the overall security in downtown Bogotá, local gangs didn't perform control over other groups dedicated to stealing in adjacent neighborhoods. Dynamics of supply and demand mediated the relationship between ganchos and robbers. For that reason, theft appears with a star (*) in Figure 1 to mark that it was a dynamic controlled by the ganchos only inside El Bronx, where it was explicitly prohibited.
In the case of places of entertainment (pools, bars, and liquor stores), infamous parties took place without any sort of legal restrictions on hours, underage participants, spurious liquor, slot machines, and gambling. These places constituted the contact point to the offering of sexual services with kids and teenagers who were victims of human trafficking or were held in chronic addiction as local gangs extorted them to perform sexual acts in return for drugs. Once the service was agreed, it took place in the adjacent motels.
The motels, known as paga diarios, were the preferred place for consumers from a higher economic status that spent extended sprees or even considered these places their home. The motels were also the place of a particular form of extortion and kidnapping called ' drugs for debt.' Under this scheme, the supply line retained many people against their will, including tourists, and extorted their families after the drug user had allegedly acquired a debt for the drugs and the motel room.
Several reports mentioned the use of exemplary physical punishments against homeless drug users. However, it remains unclear if these punishments were the result of not paying ' drugs for debt' to the ganchos or if it was a format in which the homeless could not met the expectations of selling drugs and give money back to the supply line. Exemplary physical punishments consisted of public beatings and degrading treatments by the sayayaines against homeless people and drug users. Also, ganchos permitted the vending of drug paraphernalia (*), secondhand goods, and street food in the public space under their control, as long as vendors paid an irregular tax for doing so.
Despite being known as a don't go area of the city for tourists and even local law-abiding citizens, El Bronx was always open to the public and clients in general. The normalization of crime can be seen in a saying of social workers on the frontline delivery of ground-based social services: "El Bronx was open to the city but closed to the state and its institutions." As the next section suggest, El Bronx endured as a result of a criminal governance system structured upon state-sponsored protection rackets. This form of protection racket enabled local gangs to achieve territorial control while further commanding a competitive adaptation process to socially control consumers.

III. Protection Rackets and Criminal Governance
According to the typological distinctions of illegal markets proposed by Beckert and Dewey (2017), El Bronx economy and its adjacent ecological urban settings showed an ongoing process of contested informalities. A process in which the whole palette of illegal economies provided a living for thousands of people based on the following markets: 1) markets in which the good or service is illegal (i.e., drugs, sex with minors), 2) markets of stolen legal goods (i.e., automobile spare parts and smart phones), 3) markets of falsified, counterfeited or forged goods (i.e., forged or spurious liquor), 4) legal products traded in illegal ways (i.e., adult entertainment goods and services sold to underage consumers), 5) legal markets in which regulations are not observed (i.e., gambling and slot machines without required permissions). Controlling the overlap of criminal activity with thriving illegal and informal economic activities was one of the major challenges for the rule of law at El Bronx since organized crime cannot run illegal markets for long periods without a certain level of cooperation from state officials.
However, the criminal governance system emerging at Colombia's urban drug war is more of a street-level criminal order where crime is managed instead of fought. According to Dewey (2015), in a street-level clandestine order, the presence and action of the state are not replaced or nullified but instead transformed into a type of governance explicitly or implicitly shared with criminals to deliver both public goods, such as social services, and public bads, such as coercion and disciplinary violence.
Inevitably, Bogotá is the city with the most prominent public budget, the strongest statehood, and the lower and most stable homicide rate of the three biggest cities of the country in the last three decades (see Figure 2). For years, organized crime in Colombia's capital was managed rather than fought using state-sponsored protection rackets to keep violence low amid thriving criminal markets. Although hosting the biggest open-air drug market of the country, it still featured a low homicide rate and some violent but contended conflict resolution.
Taking advantage of the strong budgetary and firepower that the Bogotá Police Department can deliver, officers at different hierarchical levels assigned to the central district started a private protection business in El Bronx. To emerge and persist over time, state-sponsored protection rackets need a) the state's capacity to enforce the law, which is key in the business of selling non-enforcement or enforcement against competitors, and b) the criminal organizations' capacity to comply with regular paying and with particular behaviors, such as containing violence (Snyder and Duran-Martinez 2009).
In El Bronx, the result was a highly diversified thriving public drug market, the deepening of the criminal control over vulnerable populations, and the erosion of civil authorities that had to implicitly bargain their presence and social service delivery with local gangs. From traditional corruption mechanisms to more sophisticated shared governance schemes, the offending model in El Bronx embedded state-sponsored protection rackets as follows: -Law enforcement agents continuously leaked privileged information about crackdowns and seizures that were about to be taken in the district. -When pressured by political authorities or public opinion, supply lines and local gangs made positive results available to local enforcement agents. Small amounts of drugs, arms, and petty criminals that were most likely not prosecutable were presented as proof of criminal justice system efficiency. -Law enforcement agents coordinated the perimeter security of the black market by organizing people at the entrance and exit of El Bronx. Security coordination was a central feature of the criminal order that lasted for years as it implied everyday coordination with the private security scheme that local gangs installed on site. -Law enforcement agents whistle blew about infiltration attempts by other agencies. This type of cooperation ended up in the killing of one police officer that was shot to death while undercover back in 2012 and the kidnapping and torture of two general prosecutor's office investigators in 2015. -Law enforcement agents performed disciplinary violence against vulnerable individuals that, for any reason, will fall in disgrace on behalf of the local gang.
Within criminal governance schemes emerging in the new local-urban war on drugs in Colombia and Latin America, interactions between criminal and law enforcement agents are creating parallel power structures specialized in controlling populations rather than territory. This is a crucial distinction as some of the literature on criminal governance is based on the notion that territorial control is a necessary and previous condition for criminals putting into practice of norms, rules, and parallel forms of justice (Duncan 2014;Koonings and Kruijt 2009;Lessing 2020;Willis 2015). However, evidence collected prior and after the intervention of El Bronx suggests that organized crime groups and local gangs can adapt to the loss of territorial control while still being able to enforce norms and parallel forms of justice, deliver 'public goods' from the perspective of consumers in the form of secure spaces to use drugs, and most importantly, provide order to the underworld amid sustained law enforcement operations. The conditions and mechanisms of this criminal resilience are beyond the scope of this paper and demand further future research on the theoretical distinctions between forms of criminal control. Furthermore, the urban face of the war on drugs in Latin America, and Bogotá in this case, shows that illegal markets can perform social control and tiered territorial one in the face of stable political and administrative institutions (Arias 2017). Following Arjona (2016), shared expectations are understood here as a marker of emerging social contracts among parties. Shared expectations and their enforcement both by state-sponsored rackets in El Bronx that facilitated, and even encouraged, social control imposed by local gangs, created patterns and regular interactions that shaped a steady criminal governance structure.
However, criminal governance requires more than pure coercion (Arias and Barnes 2017). As criminal groups embed in the social fabric through family, kingship, and other everyday support networks, they are expected to deliver public goods, such as El Bronx safe heavens for drug using and informal justice. Three other types of norms showed shared expectations among local gangs, vulnerable drug users, and street-level bureaucracies.
-Social services were provided after prior coordination with local gangs in terms of when and how they were delivered. Concerts, self-care festivals, and even advanced posts of harm reduction attention had the acceptance of local gangs under particular conditions. Whenever these criminal networks felt disruption or detection risk increased, they mobilized clients against social programs. -Mugging was not allowed inside El Bronx. This pattern relates to the importance of consumers' safety, as the retail market was a public one. According to drug users, the 'olla is the place where all social classes meet.' 2 -Local gangs gave various types of informal permissions to work. Food vendors, unlicensed bars, recyclers, rooms for rent, sexual work, and so on were taxed as an additional source of profit. As the people making a living out of the El Bronx secondary economy were also drug users, informal permissions represented guaranteed revenue for the black market and a way to build trust and legitimacy among vulnerable populations.
The variety of social agents involved, the diversity of the interactions between this variety of actors, and the institutional effects of the parallel set of norms that governed contested informalities at El Bronx and adjacent locations indicates the complexity of illegal urban markets in Colombia.

IV. Therapeutic Policing as Street-Level Peace Building
In this section I show how two main aspects of state intervention converged into a sustainable form of drug market intervention. Dismantling state-sponsored protection rackets, complementing territorial control with vulnerable population focalized social programs, and building bottom-up governance capabilities at the street level enabled public authorities to shut down one of the biggest open air drug markets around the world. Rather than a traditional broken windows (Wilson and Kelling 1989) or zero tolerance approach (Orlando 1997), the intervention conducted at El Bronx falls more into a selective targeting and deterrence model for two main reasons. First, it was not focused on punishing all the offenses and all the offenders but those that represented a higher degree of harm, in this case, violence specialists. Secondly, the model displayed a continued targeting over prioritized organizations-and places-considered responsible for serious offenses, such as homicide and human trafficking.
However, according to Felbab-Brown (2013: 8), for an intervention to be considered focused deterrence, the ability of punishment has to be credible. In El Bronx, state-sponsored protection rackets led to the collapse of deterrence. As previous research on urban drug-related violence variation in Mexico in Colombia shows, fragmented states send the message that they cannot protect or dismantle criminal organizations (Durán-Martinez 2018).
In February 2016, three months before the comprehensive intervention in El Bronx, 14 active police officers were captured for charges of cooperating with criminal networks behind the El Bronx drug-dealing operation. All of them were assigned to the same police beat, including the commander of the Immediate Attention Command (CAI) next to the illegal market. Due to the high reliance and dependence that the criminal operation had on the state-sponsored protection rackets, disrupting the collusion scheme was the real no return point of state intervention. According to a police officer testimony filtered to the press, They state that everything is set up with authorities and with people at top levels of the Bogota Police Department, like a sort of overall shield for the people behind "El Bronx." That is the reason why they have first-hand real-time information on when warrant searches are going to happen, when an intervention is going to be conducted. They know when they are followed or monitored. They even have a camera monitoring El Bronx 24/7 (Noticias Uno 2016).
Allegedly, at least 45 police officers-including a colonel-cooperated with the drug operation. As nonenforcement or protection of the state was no longer available with corrupt law enforcement agents out of the equation, it was a matter of time until local gang rule was dismantled.
In May 2016, more than two thousand units entered El Bronx and disrupted the black market. Among the different numbers of the outcome of state action, 136 teenagers under criminal exploitation by local gangs were rescued. The following days, several riots in downtown Bogotá fueled by the ganchos affected public transport and commercial businesses, raising public concern about the El Bronx expansion to other parts of the city.
The unrest supposed a policy entanglement for city authorities. With an inelastic demand for drugs in adjacent streets and neighborhoods, the ganchos rapidly changed territorial control for intermittent presence while keeping and further strengthening population control. Social control was crucial to assure criminal rents and the only way local gangs had to achieve equilibrium and functionality once again. This is the point where the building of local governance capacity in the form of therapeutic policing was key to address criminal adaptation.
Following Stuart (2016: 15), therapeutic policing operates as a form of outreach social work that aims to transform and reintegrate residents as productive self-governing citizens. Although, in his study of Skid Row in Los Angeles, Stuart understands this form of policing as a moral one and as a street level mode of poverty governance, in places where criminal governance thrives and impedes police and social workers providing different services, such as shelter, counseling, and primary healthcare, therapeutic policing can be a main asset for local capacity.
After the state intervention and subsequent demolition of the properties and "crack houses" at El Bronx, drug users dispersion threatened economic equilibrium and structure functionality for local gangs. From a stable concentration in between 1,500 and 2,500 drug users hosted at El Bronx, the city witnessed a reduction to concentrations of roughly 300 to 400 in the following weeks and from 250 to 100 in the following months. Further trajectories and movements and the possible local globe effect of these populations on other drug markets in the city are beyond this paper's scope.
As demand disruption by state action on drug users' agglomerations profoundly affects offers and increases competence, shifts in the use of disciplinary violence rose after the illegal drug market intervention. Paradoxically, affected local gangs decided to assess competition problems through the means of disciplinary violence targeting their longtime consumers for buying basuco from competitors instead of targeting competitors themselves. "Mosco solo sopla mosco" (meaning that people who bought from gancho mosco gang would stick to their line of supply and only smoke their product) was a constant response when homeless users were asked about why they kept roaming around the former black market.
Referring to homicide trends in the central district where the Bronx was, and adjacent areas where criminal networks tried to resettle, a local security liaison for the City Hall stated, The problem with this is that we have a new dynamic of homicides towards homeless people whose bodies are dismembered and then left in bags in very particular spots forming a sort of corridor between drug users' agglomerations. I think this is a consequence of the need for consolidation that these structures are experimenting; now, we have mobile and smaller drug markets that are operated by extended arms of the old ganchos that owned the business at El Bronx.
Two points are revealing about this excerpt. The first one is how in the absence of territorial control, violence against homeless that were frequent users in the illegal market turned into disciplinary violence regardless if it was to sanction unpaid drug debts or any other motive. The other one is how from the perspective of street-level security bureaucracies and decision-makers in the security secretariat of the city, population clustering was a crucial feature and marker for criminal governance.
Following this idea, as long as the city was able to disrupt population clusters in the adjacent areas, it could affect gangs' attempts to reach street control. For instance, a main drawback and challenge of this experience is the displacement of criminal activity to the city sewer system, pushing the drug market to the urban limit. It transformed it into a shadier and high-risk business for consumers where local gangs continue to display social control mechanisms. To maintain business after state action, supply lines might be forced to exchange territorial control for territorial presence but cannot waive the population and social control.
In another instance of adaptation, criminal networks behind El Bronx adopted a cuckooing practice, also known as trap houses in adjacent areas. Following Curtis and Wendel (2007), this adaptation falls into both the technical and the social organization of drug markets. From an open-air public drug market with extensive openness to the city, users can now reach the indoor format in which local gangs take over vulnerable people's properties, or even participate in the overnight lodging businesses in large properties where usually poor people stay for short terms trying to figure their way out to a stable place to live.
Street vendors, Venezuelan immigrants, sex workers, or just vulnerable people blend themselves with drug users and local gangs. These gangs accentuated the indoor selling in paga diarios (overnight lodging) to increase the security of the criminal network while keeping users clustering out of law enforcement's sight. In a context like Bogotá, where a third of the urban poor have depended on income from street vending and other informal businesses (Holland 2017), contestation of spaces of informality by criminals and noncriminals remains a key element for criminal resilience and urban peace.
On the one hand, the new role that police officers were to perform as guards of the social offer that the city was trying to deliver contrasted with the traditional mistrust between police officers and illegal market regulars, the corruption, the disciplinary violence, and intermittence that characterized law enforcement.
On the other hand, street-based social workers that for years had to negotiate their presence in El Bronx with criminal networks were now seen by vulnerable populations as a source of peril as long as state interventions were both preceded and followed by extensive social services offerings. The latter brought tension within social policy agencies and nonprofits that, for years, pursued a harm reduction approach at the cost of not reporting criminal activity to law enforcement and security agencies. As some of them expressed, 'Being seen as cops or snitches mined the confidence clients had in us. It also puts us into danger as folks were banned from talking to us or receiving any help at all.' This dynamic not only showed the intermittent, selective, and contradictory ways by which the state intervenes in crime-ridden areas through police action but also demanded alternative street governance mechanisms to replace a criminal order with a conventional one. The answer to this challenge posed both by the unintended consequences of state interventions in crime-ridden areas dominated by illegal markets' operators and by the efforts of criminal organizations to regain functionality and equilibrium was a therapeutic policing device.
In this case, homeless problematic drug users, the weakest link of the multiple interactions of the El Bronx open-air drug market but also the most numerous and hence most problematic subjects from the perspective of public opinion, are reached out to with actions ranging from therapeutic and educational programs to counseling and shelter, to self care services, to formal social control mechanisms. Figure 3 shows the formats presented by therapeutic policing efforts deployed to stem the concentrations of drug users around selling hubs.
The novelty of therapeutic policing stands in the notion of coercive benevolence that reconciles traditionally opposite punitive and rehabilitative insights to drug policy. Within this frame of intervention, law enforcement agents use social control to discourage deviant behavior and to make street living a costly and uncomfortable option for drug users, given the legal restrictions that Colombian constitutional law imposes. However, as Stuart points out in his study on Skid Row, Los Angeles, street bureaucracies' discretion and the high moral component of this model can generate perverse incentives to exert violence as a way of care (Stuart 2016).
As an asymmetric response to organized crime, which included actions in the prevention, social policy, and urban planning spheres, therapeutic policing can be implemented as a peace-building tool only if it is oriented towards the golden thread of harm reduction, namely the implementation on people centric policy options (Shaw 2019). In other words, the therapeutic policing formats exposed in Figure 3 constitute a selective targeting effort at the street level to avoid vertical integration of criminal groups while striking their corruptive and coercive power (Felbab-Brown 2014). This is moving forward from random law enforcement amid distrust, violence, and addiction to the differentiation of illicit economies' participants and the delivering of social services towards second-chance programs for those who need to and will enroll.  In particular, the street-level bureaucracy approach can contribute to the challenge of urban peace building in the context of illicit economies in at least four ways: a) It understands policy not as a fixed construct but as a constantly changing and interactional one, b) it assumes that street-level practitioners' actions become policy, c) it gives a powerful tool to understand discretion and informal behaviors in a structured way, and d) it gives a bottom-up lens to the interface between government and the individual (Brodkin 2012). Furthermore, it can be argued that the street level bureaucracy policy approach is necessary in a context where criminal organizations' actions are displayed as a way to pursue control and stability through the involvement of extralegal actors providing services at the community level, involvement in local governance arrangements, and posing as beneficiaries and interlocutors of state programs (Abello and Guarneros 2014).

V. Conclusion, Discussion, and Policy Implications
The present study has made theoretical and empirical contributions to the way we understand criminal governance in urban settings and how therapeutic policing can be a sustainable intervention that on the one hand provides social services to vulnerable populations while refraining from escalating lethal violence. The criminal hub known as El Bronx in Bogotá, Colombia, operated not under an absence or weakness of the state but rather through an ongoing process of negotiation with primarily law enforcement and social workers. The intervention that provided services to vulnerable populations and regained territorial control was sustainable because it did not confront criminal actors involved in drug supply like it had been done in earlier years.
Still, criminals and law enforcement agents are not always on opposite sides of the law. Instead, under specific circumstances, street-level bureaucracies might further enforce criminal orders when trying to address vulnerable populations without destabilizing highly complex settings, such as crime-ridden areas. Sustained control in a particular block or neighborhood by criminal networks in a megacity such as Bogotá is expensive for local gangs. It increases the risk of detection once political authorities decide to shut down open-air drug markets.
Fixed demand guaranteed through population control might condition whether local gangs can perceive constant profits from illegal activity. In that sense, the loss of territorial control does not affect gang rule as much as the loss of population control. El Bronx was much more a social reality shaped by criminal-state interactions-whether they were oppositional or cooperative-than a geographical one. While this distinction and the different forms of criminal governance that emerge from it is a promising theoretical path of further exploration within the criminal governance literature, it goes beyond the scope of this study and is part of a soaring interdisciplinary research agenda.
Although the criminal organizations could never challenge the authority of the state over the city center, the impunity with which they operated represented an intolerable message to the public about the capacity of City Hall to provide security and justice. In this case, the power of the criminal structures consisted above all in deterring the mayor, as happened for many years, from disrupting the illegal market for an alleged social cost that such a decision would entail.
Therapeutic policing models directly affect the resilience efforts of criminal networks after state intervention, given that they erode mechanisms of population control by connecting social needs with social policy portfolios. Meanwhile, gender and race awareness is crucial when implementing harm reduction and care programs at the intersections of public safety, drugs, and social policies.
This research cannot conclude that interactions between social policy agents and local gangs before the intervention were the product of corruption, collusion, or the participation of city government officials in the informal institution of state-sponsored protection rackets. However, it found to what extent criminal networks behind the drug supply operation were comfortable and confident with state presence, mining city peacemaking efforts. The mechanisms and set of rules through which local gangs at El Bronx interacted with other criminal groups remain unclear and should be addressed in future research.
In terms of policy implications, El Bronx's case and the ongoing process of urban resilience backed by street-level peace building strategies may inform other experiences of urban decay and crime in terms of illegal markets in other urban settings across the global south. Urban peace building needs to happen at the street-level. Given the uncertainty and the perils that characterize unlawful urban life, flexibility, adaptability, and constant evaluation is required to respond to criminal governance challenges without harming vulnerable populations.
Law enforcement and public health approaches to the problems of local drug markets are not necessarily antagonistic. However, while interagency territorial work is a valuable asset in terms of street-level peace building, organizational cultures from police departments, first responders, and social services teams need to be addressed. Shared understandings of the problem of illegal economies become a crucial part of urban resilience.

Ethics and Consent
All state officials including law enforcement spoke on strict conditions of anonymity. Informed oral consent was obtained from all individual participants included in the study.