Resisting Displacement: Leveraging Interpersonal Ties to Remain despite Criminal Violence in Medellín,Colombia

Abstract

Although civilians across the globe are fleeing conflict in record numbers, the reality is that far more remain behind. In addition to traditional wars, people stay in territories governed by criminal organizations. How might individuals threatened with displacement by a criminal gang manage to resist? Drawing on intensive participant observation and interviews in marginal neighborhoods of Medellín, Colombia, I argue that the urban residents most likely to remain despite being at risk of displacement are the “well connected.” Despite threats, they leverage ties to a community figure or member of the armed group to stay. I test a number of related hypotheses using an original survey and survey experiment. Unlike other work stressing that residents are trapped by scant resources or remain only by joining local associations or belligerents, my theory reveals residents’ agency and neutrality as they seek safety and security in conditions of state absence.

Keywords

migration survey design Latin American politics conflict processes trust social capital

Violence and conflict have forced record numbers of civilians from their homes in recent years. According to the United Nations, there are more refugees now than at any time since World War II (Edwards, 2017). Although many individuals fleeing violence make it across international borders, becoming refugees under humanitarian law, many more remain within their home country and are referred to as internally displaced persons. In fact, internally displaced persons outnumber refugees two-to-one (Internal Displacement Monitoring Center [IDMC], 2017). Today, displaced persons not only flee wars waged by traditional armies, they also leave behind militias, terrorist groups, and organized crime (Cantor, 2014; IDMC, 2017). And they not only escape violence in the countryside but also in cities (IDMC, 2017; Lindley, 2010).

In the peripheries of many Latin American cities, criminal organizations govern over communities, from providing goods and services to regulating residents’ behaviors, from collecting “taxes” to defending territory (Arias & Barnes, 2017; Duncan, 2014). Forced displacement is a tool some criminal organizations wield to achieve these governance aims (Bada & Feldmann, 2019; Cantor, 2014). Despite the violence in these gang-controlled territories, most people manage to remain in their homes and communities. Oftentimes, one family flees while neighbors—who share nearly identical histories and socio-economic traits—remain behind. Remarkably, some people even stay despite receiving direct threats that they “get out, or else” from a criminal gang. How? That is the central question of this paper.

I examine this question in the context of Medellín, Colombia. A city of roughly 2.5 million, Medellín officially registers an average of 5,000 to 15,000 people fleeing each year from one neighborhood to another within the city (Corporación Region, 2017; Personería, 2015; Sánchez, 2016). Criminal gangs, and the violence they produce, are responsible for most of the city’s displacement (Personería, 2015, p. 213; Sánchez, 2016, pp. 95–97). Organized crime has a long history of regulating communities in the city, as well as producing (and dampening) violence to achieve its aims (Abello-Colak & Guarneros-Meza, 2014; Duncan, 2014).

“When I left, it was because they killed my sister . . . ,” a woman who fled conflict in one of Medellín’s peripheral neighborhoods says.¹ In contrast, a man who endured similar violence in that same neighborhood states, “We had paid a very high price, the death of my son, there was no longer a reason to leave.”² Yet another resident stayed despite receiving direct threats: “They threatened me many times, and they tried to force me out and take my house, but . . . this is my house and I am not going to let anyone take it.”³ The aim of this paper is to explain the variation in responses to similar violence: the first neighbor fled while the second stayed. This paper addresses how some people remain even despite being threatened by an armed group, as the third resident quoted above illustrates.

Based on extensive field research and participant observation conducted while living in a displacement-prone neighborhood in Medellín, Colombia, I theorize that residents leverage their interpersonal connections to remain in place. The individuals and families who remain, despite being targeted by the armed group, are not those with the most connections, but rather those with the right connections—to an important community figure or a member of the criminal group. This holds even if those connections are weak. Extending Granovetter’s (1973) foundational work on social ties to the arena of violence, I theorize that interpersonal connections increase the likelihood a resident stays through two mechanisms. In the first, an individual given several hours to leave “or else,” reaches out to an important community figure who then intervenes on her behalf with the armed group to short-circuit the threatened displacement. The community figure might be the president of a local association, neighborhood founder, or religious figure. A resident may also draw on his long-standing ties to a community figure to remain in the neighborhood despite actions—like rights activism or breaking community norms—that would normally lead to expulsion were he anyone else. These ties lead the criminal group to refrain from making a threat in the first place, even though he has the same profile as displaced persons from the area. The second mechanism works similarly, but the resident at risk of displacement instead has ties to a member of the criminal group. Either the resident leverages his ties to a gang member (other than the one making the threat) who shields him from displacement, after the threat was made, or, the criminal group refrains from making the threat, despite the resident’s norm violations or rights activism (which would typically provoke expulsion), due to the resident’s long-standing ties to the gang.

I test hypotheses from this theory using an original survey and survey experiment conducted in three Medellín neighborhoods. Through these tests, I demonstrate that the mechanism I propose has an independent effect, yet one that complements other theories drawn from the literature. In particular, these tests reveal that one of my hypotheses—that ties to a community figure may short-circuit displacement after a threat is made—works alongside an explanation from the civil wars literature: that targeted civilians are more likely to remain when a local community association intervenes on their behalf with the armed group (e.g., Kaplan, 2017; Steele, 2017). Rather than stay only by joining a local association or armed group (e.g., Kalyvas & Kocher, 2007), or due to limited resources (Bohra-Mishra & Massey, 2011), as the literature maintains, I show that residents also leverage interpersonal connections to remain despite violence and targeted threats.

This paper addresses an important empirical phenomenon that to date the literature has largely ignored. While several scholars of comparative politics consider criminal governance (e.g., Arias, 2017; Arias & Barnes, 2017), few study an important governance strategy employed by criminal organizations, namely forced expulsion. And among the exceptions (e.g., Bada & Feldmann, 2019; Muggah, 2015), none—to this author’s knowledge—are studying how residents resist displacement at the hands of these criminal gangs. This paper identifies a new mechanism—namely, interpersonal connections—for how residents threatened with credible violence remain in place, making an important contribution to the broader literature on civilian agency in conflicts (e.g., Wood, 2003). The implications of my findings are that, rather than flee violence, some individuals resist a much stronger criminal group using their agency and while maintaining their neutrality. These findings link to broader themes in the political science literature, for instance, how individuals seek safety and security where the state does not enjoy a monopoly over the use of force.

The theory presented here may hold relevance beyond the city of Medellín. Displacement due to criminal violence occurs throughout much of Latin America. Indeed, scholars warn that flight due to organized crime has reached the level of humanitarian crisis in Central America and Mexico (e.g., Albuja, 2014). Meanwhile, on a global scale, conflict is increasingly urban (Staniland, 2010), and, conflict has long had criminal elements (Andreas, 2008; Barnes, 2017). As such, a theory of residents resisting criminal groups developed in Medellín may gain traction in cities outside the region. Finally, while the current numbers of refugees and internally displaced persons are staggering, the reality across conflict zones is that far more people stay in their homes than flee (Ibáñez & Moya, 2016; Steele, 2009; The World Bank, 2017). A theory of how residents remain despite credible threats of violence, then, is relevant not only for organized crime-driven conflicts, but also for more “traditional” conflicts such as civil wars.

A Theory of Residents Leveraging Interpersonal Ties to Resist Displacement

How do individuals seek safety and security where the state does not enjoy a monopoly over the use of force? To explore this question, I consider residents who resist displacement from criminal organizations. Criminal gangs provoke displacement in cities across Latin America (Bada & Feldmann, 2019, p. 69; Cantor, 2014, pp. 35, 39; Muggah, 2015, pp. 225–226). To varying degrees, criminal organizations provide goods, services, and employment in some urban peripheries; they also regulate residents’ behaviors and punish norm violators, as well as collect “taxes,” maintain a semi-monopoly over violence, and defend territory (Abello-Colak & Guarneros-Meza, 2014; Arias & Barnes, 2017; Duncan, 2014; Gordon, 2019). Put simply, they govern over communities. O’Donnell (1993) contends criminal governance emerges at the urban margins due to state absence and a breakdown in the rule of law (p. 1358), while Arias (2017) counters that it results from state-criminal relations in the area, ranging from “criminal disorder” to divided, collaborative, and tiered forms of governance (pp. 7–8, 42–44). Illustrating the link between governance and displacement, some criminal organizations impose curfews and laws of silence over residents, forcing out those who break these norms (Cantor, 2014, p. 41); they also “tax” residents, sometimes provoking flight when families are unable to pay the extortion (Sánchez, 2016, p. 100). Skirmishes between criminal organizations—and the state—over territory further result in displacement (Cantor, 2014, pp. 49–50). The academic literature on criminal governance and displacement is embryonic (Bada & Feldmann, 2019, p. 62; Cantor, 2014, p. 35). How residents resist displacement in these conditions remains an open question.

Some insights into how residents resist fleeing urban criminal groups and their violence may be gleaned from the civil wars literature. Although armed groups in civil wars are generally thought to pursue political or ideological aims, setting them apart from profit-motivated criminal organizations, theories from the civil wars literature may still be fruitfully applied to contexts of criminal violence. This is particularly the case for certain areas of overlap: both civil war and criminal actors pursue profit, territorial and social control, and the accumulation of the means of violence, in many cases (for more, see e.g., Barnes, 2017; Kalyvas, 2015).

A strand of the civil wars literature reasons that civilians are generally safer and more autonomous where strong community associations are present to facilitate collective bargaining against the armed group, which is more likely to accept circumscribed intervention in local affairs under such conditions (Arjona, 2016).⁴ These local associations are most effective at deterring displacement when they harshly sever ties with any member who defects (Steele, 2017), clarify the “fog of war” by managing relations with the armed actor including clearing up misunderstandings between community members and the belligerent (Kaplan, 2017), and have connections to external supporters, such as churches or foreign support networks (Kaplan, 2017; Krakowski, 2017; Steele, 2017).

Another strand of the civil wars literature suggests that civilians remain in their communities and generally protect themselves by collaborating with or joining the controlling armed group (Kalyvas & Kocher, 2007; Steele, 2009, p. 424). Meanwhile, several scholars working on migration and conflict posit that the residents who remain are those without sufficient financial resources to leave, along with those who have no family elsewhere to receive them, or whose social networks are unable or unwilling to take them in (Adhikari, 2013; Bohra-Mishra & Massey, 2011; Ibáñez & Moya, 2016). On the other hand, the individuals most likely to flee conflicts are on average younger (Ibáñez & Vélez, 2008), and more likely to have received a violent threat or lost a family member (Adhikari, 2013).

I define violence-induced urban displacement as occurring when individuals are provoked, directly or indirectly, to leave their homes in a city for another location—be it in the same city or further afield—by criminal organizations, other armed groups, or state forces.⁵ Such flight results from diverse causes, from the presence of these groups and shoot-outs between them, as well as the extortion they may charge and their occupation and sale of residences; it also results from these groups targeting violence at particular individuals, for instance as a reprisal, harassment, or in an attempt to forcibly recruit adolescents.⁶ The focus of this paper is criminal gangs using targeted violence. Specifically, displacement results from selective targeting when an armed group threatens or commits violence against a particular individual, due to a (perceived) action or characteristic of that individual.⁷ The theory developed below is about staying despite selective targeting, a subject that has received little scholarly attention to date. As Ibáñez and Moya (2016) write, “While published research exists on the causes and consequences of displacement, much less is written on the decision to stay or on the consequences of that decision” (p. 252). In this paper, I develop a theory to address the sources of this individual-level decision to stay when faced with a credible threat of violence.

Participant Observation and Interviews

Over the course of 15 month’s fieldwork in Medellín, I conducted interviews and participant observation to develop a theory for how some residents remain despite conflict and despite being targeted by an armed group. I interviewed displaced persons and people I would have expected to flee but who managed to stay (e.g., rights activists, LGBT⁸ persons), as well as community leaders, non-governmental organization workers, government employees, and current and former gang members.⁹ Simultaneously, I conducted six months of participant observation at a local non-governmental organization that serves several of Medellín’s marginal neighborhoods where displacement takes place. After building the necessary rapport with the community, I moved to one of these neighborhoods where I lived for six months. Conducting participant observation in the organization and neighborhood allowed me to directly assess several factors relevant for displacement, such as community cohesion, police presence, and public service provision. Throughout, I also “hung around” talking to ordinary people, writing up fieldnotes as I went (Fenno, 1978).¹⁰ Based on this qualitative data, I inductively generated the following theory for how people resist displacement.

Forced to Flee: Displacement in Medellín

The marginal neighborhoods of the Cartucho Blanco district¹¹ lie high up the steep, verdant hillsides of Medellín. Above the older stretches of the district that are formally integrated into the city, lie neighborhoods where two-story residences are topped with corrugated iron sheets for roofing—the first story is often of solid construction, while the unfinished second floor may have exposed cinderblocks or wood. Public services are present but at times run irregularly and much of the neighborhood is only accessible by foot or motorcycle, outside the paved main street. Families first settled these areas about 40 years ago. At yet higher reaches of the hillsides, where the terrain is steep and erosion likely, lie neighborhoods founded in the last 10 years. Public services there are spartan, and the one-story, dirt-floor residences are often constructed of wood or found materials, such as plastic sheets. Recent migrants from the countryside, indigenous persons, or Afro-Colombian communities are more likely to call these high-up neighborhoods home.¹² These descriptions, and those throughout this section, are based on my participant observation and interviews.

These are generally close-knit communities. At 9,000 inhabitants on average, these neighborhoods are often filled by extended families. When safety reigns, few residents move away from them, even among those who have the means to leave, due to a sense of belonging. Neighbors help one another, chat in the street, play on the same fútbol team or attend the Catholic or Evangelical church together. Most neighborhoods have a main street, even if dirt, where residents can flag down an informal bus or buy the basics. The residences closer to this main street are more accessible, generally better off financially, and are more likely to have public services. Nearly all neighborhoods have a Residents’ Association (Junta de Acción Comunal), which is a government-affiliated organization with elected leaders from the community working to improve the neighborhood. While some of these associations are well-organized and law-abiding, others are weak, newly formed, or corrupt.¹³ A couple of neighborhoods have a small police station staffed by a handful of officers; however, many residents report that the police are unresponsive and corrupt.¹⁴ These neighborhoods suffer some poverty and undernourishment, as well as high unemployment. Under these conditions, domestic violence and fights between neighbors are frequent. Drug use is also common, specifically marijuana and an inexpensive cocaine paste that is smoked, called bazuco.¹⁵

Neighborhoods are under the control of street-level gangs (combos) and, in some cases, one gang will control several neighborhoods. The gangs are comprised mostly of young men in their teens and 20s who grew up in the neighborhood and were gradually recruited into the criminal organization, often starting as errand boys (carritos). Having grown up in the neighborhood they control, gang members know the residents, who include their parents, store owners, and family friends who looked after them when they were young.¹⁶ In addition to controlling access to the neighborhood and patrolling for outsiders who might rob residents, gangs buy uniforms for school children, resolve disputes between neighbors, and make loans. They also regulate community norms, for instance, preventing spousal or child abuse, enforcing where drugs are consumed, and promoting conservative values like traditional gender roles. Some residents have a strong sense of loyalty to the group, willingly contributing “taxes” toward neighborhood security (colaborarles con la seguridad), storing drugs or arms, or helping in other ways such as cooking a meal. While outside gangs will occasionally conquer new territory, most have long tenures, so the residents generally view one gang as hegemonic.¹⁷ To keep in business, the gangs sell drugs in the neighborhood (from spots referred to as plazas de vicio), as well as extort (cobrar vacunas) local buses and shops.¹⁸ These are generally hierarchical groups, for instance, rank-and-file members typically receive orders to expel a resident from above.¹⁹

The street-level gangs are beholden to criminal super-structures (estructuras), several of which control the city.²⁰ While street-level gangs had ties to civil war actors in the 1990s and early 2000s, such ties have faded over time. Indeed, the civil war moved away from Medellín in the mid-2000s, peace accords were signed in 2016, and the paramilitaries (to which many groups had ties) have increasingly morphed into criminal enterprises. Rather than being politically motivated actors in a civil war, then, these gangs are largely criminal.²¹

Much of the forced displacement occurring in the peripheral neighborhoods of Medellín is targeted, meaning the street-level gang in control of the neighborhood threatens specific residents that they must leave or face death.²² Government reports and my interviews show that the gangs force residents out of the neighborhoods over (perceived) grievances, such as calling the police (sapear)²³ or loyalty to a rival gang. Having violated a rule or norm is also reason for expulsion, including being a rights activist, being a sexual minority, robbing within the neighborhood, spousal abuse, or consuming drugs in front of children. The gangs also expel households to gain territory or turn a profit, often by evicting a household then looting and renting it out to newcomers. Some residents also flee the repercussions of gangs exercising territorial control, such as being forced to store arms or drugs, pay a “tax” for protection, or join the gang.²⁴ Residents generally consider threats “to get out of here now” or “pay up, or else” to be real if made face-to-face or if the message is sent through a relative. “They confront you directly, it’s because you have to leave . . . there is not a negotiation,” one displaced person said.²⁵ Whereas threats made via rumor or menacing looks from gang members, while serious, are less likely to provoke flight.²⁶ But all residents know that, if a direct threat is left unheeded, then being attacked and most likely killed is the probable outcome.²⁷

Despite the targeted threats and occasional generalized violence, the majority of residents remain in their homes and communities. Some would prefer to leave but lack the financial resources to make the move; other residents have strong emotional ties to their neighborhood and community, so they remain; and yet others fear that affordable neighborhoods elsewhere in the city are equally dangerous (Marston, 2019). With the aim of remaining, many residents develop coping mechanisms to weather violence and avoid unwanted attention from the gang. For instance, they stay indoors, hide wealth, and keep their heads down (Auyero & Kilanski, 2015; Marston, 2019). Those residents who receive direct threats, however, must deploy other strategies to remain in place.

Leveraging Weak Ties to Resist Displacement

How are some residents able to remain despite being targeted by a criminal group? I argue that who they know matters. Connections to the right person—a figure with standing in the community or a gang member—reduce the likelihood of flight. The connections need neither be numerous nor strong. Rather, many individuals who resist displacement do so through one tie—but to the “right” person. As improbable as it may sound, otherwise “regular” community residents leverage their ties to resist the much stronger armed group and remain in their homes and communities. As theoretical grounding, I rely on Granovetter’s (1973) seminal work on social ties.²⁸

I argue that ties reduce the likelihood of displacement through two mechanisms. In the first, a resident has ties to an important community figure who shields her from displacement by intervening on her behalf with the gang. Once the gang has set a displacement event into motion—for instance, told a household they have 24 hours to leave—an important figure in the community or someone with sway over the gang steps in to short-circuit the flight. For this to work, the community figure intercedes with the gang to negotiate for the family to remain in the neighborhood. If the gang concedes, then the individual or family stays. Such an intervention is no guarantee, but it decreases the likelihood of displacement. The community figure could be a neighborhood founder, a well-liked elder, a priest or other religious figure, the leader of a local association, or anyone the gang members grew up respecting from the neighborhood.

To short-circuit displacement, the community figure must possess clout with either the gang or the wider community. In the case of clout with the gang, he is respected enough by the armed group that he can either reason with or shame them into stopping the displacement. To be clear, his clout is with the same gang threatening the displacement, the gang that is hegemonic in the neighborhood at the time. As one gang member explained, some community figures have sufficient sway with the gang to intervene on a resident’s behalf:

. . . someone who is not part of the gang but can give their opinion, there’s those too, here there is a man [un señor] who is older, he’s really respected . . . everything that has happened in this neighborhood, the neighborhood is his . . . all his time living in this neighborhood, he’s respected for that . . . he can give his opinion, “guys, this and that,” . . . we don’t mess with them [people like this man] and they don’t mess with us, they need us and we rely on them, it’s not in our interest to have enemies.²⁹

In the case of clout with the community, the important figure holds such sway with residents that going against her could erode the gang’s control over the neighborhood. While the gang can still force the family to flee, making enemies of such a respected community figure is unwise and costly, since she can sway community opinion against the gang. The gang’s receipt of “taxes” from residents, and the protection they provide through noncollaboration with rival groups and the police are contingent on such sentiments, so the gang works to preserve them.³⁰ Since the community figure’s leverage comes from her position in the community, she need not have a long-standing relationship with the gang (though she could) to intervene with them on behalf of the threatened resident. While male community figures also successfully intervene on the behalf of residents about to flee, women sometimes do so by leveraging the touchstones of motherhood and candor, so common to paisa (Medellín’s) culture, going so far as to “scold” gang members or directly confront them (in a manner far too dangerous for men to assume). When successful, the head of the gang acquiesces to the community figure’s demands and calls off lower-ranking members from expelling the resident.

The first mechanism works in another way, as well: a person who otherwise would be at risk of displacement has such close ties to a community figure that the gang refrains from targeting the resident. Here, the criminal group avoids targeting a resident whose actions or characteristics would otherwise provoke expulsion, such as rights activism, violating norms, or being lesbian, gay, bisexual, and transgender (LGBT). The gang refrains because it is nearly certain the community figure will swiftly intervene in the displacement and challenge their authority in front of the community. To avoid immediate and potential long-term problems with an important community figure, the gang simply does not threaten the resident in the first place.

In the second mechanism, a resident avoids displacement through connections to a gang member. In this case, the person threatened with displacement asks her contact, friend, or relation in the gang (who did not make the threat) to vouch for her with the criminal group. The contact then intervenes on her behalf, generally by speaking with the head of the gang, perhaps relaying that she is “one of them,” that she will not break the gang’s rules again, or that all she needs is more time to repay the debt.

This mechanism also works when the gang refrains from making a threat in the first place due to a resident’s long-standing connections with a gang member. Being the relative of a gang member, dating one or being his friend, or having gone to school together or played as children often grants residents prestige and favors, as well as latitude to transgress rules and norms.³¹ Nearly every neighborhood has human rights activists who challenge the control of the armed group or other people who violate community norms. By transgressing community norms, I mean persons who break rules imposed by the gang or who go against the neighborhood’s expectations, for instance, persons who consume drugs in front of children, rob in the neighborhood, or are LGBT. Norm breakers are often forced to flee. If they stay, it is because ties to the gang—especially to the gang’s leader—lead its members to refrain from threatening such residents in the first place. In a counterfactual scenario where these ties were not present, the gang would force out these residents since they share the same profile as other displaced persons from the area. In practice, leveraging these ties is often as simple as one of the gang members saying in passing “oh, that guy’s OK,” to his associates. Both men and women can rely on these ties; they generally do so in similar manners—for instance, highlighting relatives in common—with the exception that some women may benefit from dating male gang members (who are over-represented in these criminal groups).³²

Take, for instance, the experiences of one rights activist who ran a youth LGBT organization advocating for the visibility of sexual minorities throughout the Cartucho Blanco district. He had attended grade school with the leader of his neighborhood’s gang. Although he received minor threats from the gang for his activism, none were serious enough to prompt flight. His position grew more precarious, however, when a new gang drove out the old one. The new gang intimidated the LGBT organization members by flashing knives outside one of their meetings and shooting at their building, then told the leader: “We don’t want to see you here anymore.” He reported the threats to the local Ombudsman’s Office (Personería) but fled immediately to a government-run shelter after receiving a phone call from his mom who relayed that the gang had come by her house looking for him. After leaving the shelter some weeks later, the leader rented a place in a different neighborhood of the same district where he stayed for several years. He then hesitantly returned to his old neighborhood, considering it his home, after hearing the old gang had returned to power. On a quick trip to first check things out, el fuerte, or the leader of the gang, saw the rights activist and said, “I know you, you’re OK.” The leader also asked that if the activist returned and resumed work that he would notify the gang before outside rights organizations or the police came to the area, thus giving him tacit approval. Although the community leader was more strategic about the events he put on afterward, he resumed his advocacy work in the neighborhood.³³ What changed? Why was the same activist carrying out the same advocacy permitted to stay in the neighborhood under one gang but not the other? The critical mechanism determining whether he stayed or fled was the presence (or absence) of weak ties, which kept the gang from making threats (or not).

Unlike this LGBT activist, many rights activists, sexual minorities, or others who transgress community norms lack the necessary ties to the gang to keep it from making threats that they leave. I provide two examples from displaced persons. Referring to the gang that forced her out, one middle-aged community activist said, “I knew who they were, but I’d never even say hi in the street.”³⁴ Meanwhile, another LGBT person and activist who fled remarked, “It’s better not to get involved with the gang at all . . . come on, I say hi and everyone thinks I’m part of the group.”³⁵ Finally, the leader’s ability to remain was not tied to differences between the gangs’ policies; indeed, he received minor threats from the first gang, confirming they also had the policy of forcing out people who do not conform to the traditional values many gangs enforce. Instead, the leader leveraged ties to remain.

I derive the following hypotheses from this theory:

Hypothesis 1 (H1): Ties to a community figure reduce the likelihood of displacement. The community figure may shield the resident from displacement once threatened (H1a), or, the gang may show restraint from making the threat in the first place, despite the resident sharing the profile of persons typically expelled (e.g., norm violator, rights activist), due to the resident’s ties to the community figure (H1b).

Hypothesis 2 (H2): Ties to the gang reduce the likelihood of displacement. A gang member may shield the resident from displacement after a threat is made (H2a), or, the gang may show restraint from making a threat in the first place, despite the resident sharing the profile of people generally expelled, due to the resident’s gang ties (H2b).

Testing the Theory of Resistance: A Survey and Survey Experiment

The first half of this paper laid out the theory I developed for how residents resist displacement based on interviews and participant observation during field research in several marginal neighborhoods of Medellín. The second half tests hypotheses derived from this theory using an original survey and survey experiment carried out in displacement-affected neighborhoods distinct from those where the theory was generated. Following Thachil (2018), I used my qualitative research to inform the survey, making for more realistic questions and experimental vignette. In the next section, I introduce the survey and its experimental question. I then present the survey results: first from the experimental question, followed by the quantitative findings and respondents’ qualitative responses to open-ended questions.

Survey and Sampling Design

The survey was carried out from August to October 2017.³⁶ In all, 618 respondents participated, 530 of which were randomly selected. Respondents were selected using a multi-stage sampling procedure. First, I purposively selected three neighborhoods in the periphery of Medellín, all within the same district (comuna), based on comparability and accessibility, in consultation with a local partner non-governmental organization.³⁷ Enumerators approached the appointed households then selected a household member (over the age of 18) to participate at random.³⁸ Approximately 530 participants were included through this random method; roughly 70 men and 20 urban displaced persons were brought into the sample using a screening procedure to include a higher representation of these populations. The survey’s response rate was approximately 93%.³⁹ Since the survey covered sensitive topics, the team went to every reasonable effort to minimize respondents’ re-traumatization, including calling the on-team psychologist to make house visits as necessary.⁴⁰ I personally hired, trained, and supervised the team of five enumerators that carried out the survey.⁴¹

The survey and subsequent data analysis compare the characteristics and experiences of people who fled urban violence (i.e., displaced persons) with urban residents who have not fled violence (i.e., non-displaced persons). The survey collected data on individuals who fled violence in the countryside (i.e., rural displaced persons), but they are not included in the statistical analyses here. Displaced persons were asked about their displacement, as well as their lives, communities, and public services at the time they fled (which was between a month before the survey and 20 years ago, with many centered at 2011/2012). Non-displaced persons were asked comparable questions about their life, community, and public services in 2011/2012, a time of high violence and displacement in Medellín (Personería, 2015). Survey respondents also answered standard demographic and economic questions, as well as questions about gang control and related topics, for both the past (at the time of displacement, or 2011/2012) and the present. Survey respondents answered both closed- and open-ended questions. In all, 157 urban displaced persons participated in the survey. Respondents’ average age was approximately 42; about 20% of respondents self-identified as White, while others reported having mixed heritage, or being indigenous or Afro-Colombian.

An ideal survey design would capture respondents who stayed in and fled the same neighborhood, thus controlling for conditions of violence and other place-based effects. Due to the cyclical nature of migration in Medellín, approximately 20% of respondents fled one of the sampled neighborhoods then later returned to it, thus reaching this ideal. The majority of survey respondents, however, instead fled from other neighborhoods in Medellín (or even another city in a few cases) then resettled in the sampled neighborhoods. This means they fled neighborhoods and conditions of violence different from those that the non-displaced comparison group lived through (in the sampled neighborhoods, in 2011/2012). As such, results from regressions and descriptive statistics must be interpreted with care, specifically variables related to neighborhood characteristics and respondents’ social relations. While means exist to follow migrants from sending to receiving communities for tighter comparison groups (Massey, 1987), they were not practical for this survey, because many displaced persons refuse to be tracked due to safety concerns.

Additional details on the survey are available in the Supplemental Appendix, including screening and sampling procedures, post-survey weights applied, and the safety measures taken for respondents and enumerators.

Experimental Survey Question

The survey included an experimental question to test one of the hypotheses for how residents resist displacement. Specifically, the survey experiment addresses the questions: (i) Does the intervention of a community leader, or other important figure in the community, impact displacement? and (ii) To the degree successful, is it the person as such who impacts displacement, or is it that he represents a local organization, for instance the Residents’ Association? The first component (i) tests H1a: whether a tie to a community figure reduces the likelihood that a resident targeted for displacement flees. The second component (ii) tests H1a against the primary explanation in the literature for how residents remain in place when threatened: that they are more likely to stay when a local community association intervenes on their behalf to negotiate with the armed group (Kaplan, 2017; Krakowski, 2017; Steele, 2017).⁴² In other words, this test is meant to determine whether community figures have their own effect on displacement, or whether the organizations they represent have the real impact.

An experimental survey question is useful because it allows for data collection regardless of a respondent’s status as a displaced person (or not) or having ties to community figures (or not). Rather than ask respondents to hold in their head and rank information about types of leaders and their impact on displacement, this question was designed to randomly assign the information to respondents and then compare the mean responses across groups. To ensure the scenario was realistic and reminded respondents of their own neighborhoods, I wrote it based on evidence collected during my fieldwork, consulted local contacts, and piloted the question. The experimental vignette asks respondents to imagine that, after mediating a dispute between neighbors, the local gang decides one family must leave the neighborhood. The respondent then hears one of the four randomly assigned treatment scenarios in which someone intervenes with the gang on behalf of the threatened family. The treatment scenarios are the intervention of (a) an important figure in the community, (b) the same with explicit ties to the armed group, (c) the official leader of the Residents’ Association, and (d) the same with explicit ties to the armed group. The control is no intervention. After random assignment to the control scenario or one of the four treatment scenarios, respondents are asked how likely the family is to flee.

As it is worded, the question is a fair test of H1a since there is little a priori reason to think that respondents would automatically answer that a community leader/figure mitigates displacement. On the contrary, respondents could answer that a community leader/figure interceding makes the household more likely to flee—for instance, by encouraging the gang to expel the family immediately or helping the family relocate. And while the responses are based on respondents’ intuitions about a hypothetical displacement rather than firsthand experience, survey participants lived in displacement-prone neighborhoods for years if not decades, and some are themselves displaced persons. The experimental survey question reads as follows.

Experimental Survey Question
Control	Imagine that in a neighborhood like your own, two neighbors get into a fight. As may happen with your neighbors, this fight started because one of them was playing music too loud, late into the night, after having drank a lot [of alcohol]. The next day, the gang [los muchachos del barrio] come by to sort out the dispute, which had turned into a fist-fight. They asked details about what happened from both sides. They decided one family was more at fault, had disturbed the peace of the neighborhood too much, and so said to the family, “Mijo, you have to go . . . ”
T1	[Control text] After half an hour, an important member of the community comes to intercede on the family’s behalf. He is respected by almost all and has lived in the neighborhood for a very long time, but he is not an official leader of any community organization.
T2	[Control text] After half an hour, an important member of the community comes to intercede on the family’s behalf. He is respected by almost all and has lived in the neighborhood for a very long time, but he is not an official leader of any community organization. He has worked with the gang [los del combo] on community projects and has a relationship with them.
T3	[Control text] After half an hour, the president of the Residents’ Association [Junta de Acción Comunal] comes to intercede on the family’s behalf.
T4	[Control text] After half an hour, the president of the Residents’ Association comes to intercede on the family’s behalf. He has worked with the gang on community projects and has a relationship with them.
Question	On a scale from 1 to 5, 1 being very unlikely and 5 being very likely, how likely is it that the family leaves?

Ties to a Community Figure Reduce Displacement: Results From the Survey Experiment

The key finding of the experiment is that, in the perception of residents living in neighborhoods where displacement occurs, the intervention of a community leader reduces the likelihood of flight by 16 percentage points. Or from the chances of fleeing being close to “likely” to being “neither likely nor unlikely.”⁴³ Figure 1 presents the results from the survey experiment; to make this figure, I plotted the mean responses of the control and treatment groups, along with a 95% confidence interval.⁴⁴

Figure 1.

Leaders’ impact on displacement.

Next, an important member of the community who has no formal role in any local organization is perceived to be equally as effective at mitigating displacement as the formal leader of the local Residents’ Association (see Figure 1).⁴⁵ Indeed, a difference-in-means test shows that community members perceived them to reduce the likelihood of displacement at nearly indistinguishable rates (see Figure 2). This suggests that the power of a leader or important figure to reduce the likelihood of displacement resides in the person rather than their role as representative of a local organization. A prominent theory in the field suggests that civilians are more likely to remain despite armed group threats where a local association is present to intervene on their behalf (e.g., Kaplan, 2017; Steele, 2017). This survey experiment is designed precisely to test whether a community figure has an independent effect in reducing the likelihood of displacement when compared to the impact of local associations. Indeed, a critic might suggest that the intervention of a community figure is only successful because she represented a local association and, as such, the association has the impact, not the person. However, this test lends confidence that, in the eyes of neighborhood residents, the community figure by herself matters for reducing displacement.

Figure 2.

Cross-table of differences among groups with t-test.

While caution must be exercised when extrapolating from attitudes expressed in surveys to actual behavior, the results strongly suggest that community leaders can prevent some cases of displacement. Indeed, the mechanism I theorized based on participant observation and interviews held when tested in other neighborhoods using a different data collection method. Results also show that community figures can mitigate flight, not just leaders representing a local organization. In other words, this out-of-sample test provides support for H1a.⁴⁶

Gang Ties Reduce Displacement: Quantitative Survey Results

Next, my theory suggests that ties to the gang reduce the likelihood of displacement. As an out-of-sample test of this hypothesis (H2), statistical analyses are run on observational survey data collected in neighborhoods distinct from theory generation. For the regression analyses, I use linear probability models with a binary outcome variable, displacement (displaced = 1, non-displaced = 0). For ease of interpretation, regression results are presented graphically as margins plots for the variable of interest. The full regression table can be found in the Supplemental Appendix, along with logistic regressions, run as robustness checks (Suppmental Appendix, Figures 22–23).

Regression results and descriptive statistics show the broad correlation expected: residents with ties to the gang are more likely to have remained. From the regression results, having connections to the gang is associated with a nearly 17 percentage point decrease in the probability of displacement, holding all else constant. Demonstrated in Figure 3 below, these findings are significant at the p < .01 level. These numbers are obtained with a linear probability model, in which the outcome variable is fleeing (or staying).

Figure 3.

Predictions of the impact a connection to the gang has on displacement.

Results are based on the survey question, “In the twelve months before leaving/At that time [2011/2012, for non-displaced persons], if there were a crime in the neighborhood, who would you first turn to?” I take the answer of “the group at the margins of the law”—a term used for the gang—rather than other responses (“the police,” “no-one,” “the residents themselves,” etc.) to proxy for a preexisting connection to the gang. If a respondent would turn to the gang in the case of a crime, then the respondent likely knows at least one member of the gang and how to contact them. Considering the sensitivity of asking about gang ties in the context, this proxy is adequate. Among displaced persons in the sample, 14% reported they would have turned to the armed group were a crime committed, while 28% of non-displaced persons reported the same.

Descriptive statistics show a similar trend. If my hypothesis (H2)—that connections to the gang matter for mitigating displacement—holds, then displaced persons in the sample should not have gang ties before fleeing and persons who remained despite violence should have these ties. That prediction is met when tested with proxies for connections other than those used in the regression model. As Figure 4 shows, most people who fled report having no connections to the gang before leaving. On the other hand, a slight majority of persons who stayed report such a connection (during 2011/2012, when displacement was high in their neighborhood), as shown in Figure 5.⁴⁷

Figure 4.

Displaced persons: Connection to gang (at time of displacement).

Figure 5.

Non-displaced persons: Connection to gang (in the past).

In sum, regression results and descriptive statistics show that, for residents, there is a correlation between having ties to the gang and remaining through a period of heightened displacement.

Leveraging Ties to Reduce Displacement: Qualitative Survey Results

I turn now to an analysis of the open-ended survey responses to assess qualitative support for my hypotheses. I first consider whether gangs refrain from making threats against people who have ties to an important community figure, despite actions or characteristics that would otherwise generally lead to their expulsion, such as breaking community norms or being a rights activist (H1b). The following quotation provides support for this hypothesis, it comes from a 43-year-old mother, whose son avoided displacement for several years through ties to his father, a man well-known in the community:

My son did some damage in the neighborhood [hizo un daño en el barrio]. He started to rob and that’s why they threatened him. They told him he had to join the army [i.e., leave the neighborhood] . . . or they’d kill him. It was the neighborhood gang. Initially they didn’t do anything to my son because everyone in the neighborhood knows my husband, since he was young.

In this case, a person who we would expect to have received an immediate threat to leave the neighborhood—due to probable vandalism or robbery—remained for some time due to his connections to a well-known community figure.

Next, I consider whether residents remain by leveraging their ties to gang members—other than the one making the threat—to shield them from displacement (H2a). Responses to open-ended survey questions reveal residents availing themselves of precisely these ties to remain in their homes and communities. For instance, a 38-year-old man says: “ . . . I wouldn’t join [the gang] so they told us we had to leave the neighborhood . . . it just so happened I had a vocal friend who was with them [the gang] and he stood up for me.” As further evidence, an 18-year-old male survey respondent reports,

At the moment we’re being displaced from this neighborhood, because I owe some money to the gang . . . they told me that I have to leave [me abriera del parche], they didn’t want rats here, they gave us three days to leave. My mom went and talked to another guy, a heavy [con otro man un duro] . . . so that he’d help us and they’d give us more time . . . that man came and helped us and they gave us 15 days.

In sum, several residents threatened with displacement had contacts in the criminal organization who interceded on their behalf allowing them to remain in the neighborhood. This lends confidence to H2a.

Finally, I turn to residents who remain in the neighborhood without ever receiving threats, despite having the profile of other displaced persons (e.g., rights activists, drug abusers, LGBT persons), due to their ties to gang members. Testing this hypothesis (H2b) is difficult since demonstrating the absence of threats is a challenge. However, qualitative data from the survey are suggestive. For instance, a mid-50s rights activist and community leader, who has lived in her neighborhood more than 20 years, says,

I know the guys in the gang, there is a good relationship with them, they take care of the neighborhood . . . [but] they threatened me and my family with an anonymous pamphlet, they said in there to get out. [I] immediately went to the head of the gang to look into the threat, they said it was a mistake and we didn’t have to leave.

As the quotation shows, a resident the gang would be expected to force out (i.e., a rights activist and community leader), but who had a long-standing tie to gang members, received the threat to leave only by mistake. As such, it may be assumed that she and her family would not normally receive such threats. This evidence provides support for H2b.

In conclusion, qualitative data from the survey provide further evidence that some residents leverage interpersonal ties to resist displacement. These results are in line with the experimental and quantitative results above. Taken together, interpersonal ties—between the person at risk of displacement and a community figure or gang member—are shown to decrease the likelihood of forced flight. Either the criminal group refrains from making the threat in the first place due to the resident’s ties, or the community figure or gang member intercede with the gang on the resident’s behalf to short-circuit the displacement. These out-of-sample tests afford confidence in the hypothesized mechanisms. However, I also consider whether they continue to have an effect when tested alongside other explanations from the literature for how civilians remain despite violence.

Alternative Explanations

Several theories other than my own exist to explain how some individuals and families resist displacement. These alternatives mostly avoid the informal interpersonal ties that I emphasize, however. Rather than outright refute any of these other explanations, my theory has an independent, yet complementary impact on reducing the likelihood of flight, alongside the others. In addition, the mechanisms I propose operate in contexts and at times where the alternatives are absent.⁴⁸

First, several scholars in the civil wars literature suggest that residents avoid displacement by joining a local association which then bargains on their behalf with the armed group (e.g., Kaplan, 2017; Steele, 2017). At base, these theories claim that the organization, rather than any one leader, matters when bargaining with the armed group for residents to remain in place. However, the survey experiment shows that neighborhood residents believe independent community figures are equally as effective at preventing displacement as those with institutional backing. Since community associations send a person to intervene in potential displacement events, the experimental survey question disentangles whether it is the community figure, as person acting alone, that stops flight, or the local association, represented by a person, that reduces displacement. In the sampled neighborhoods, the results from the survey experiment make it clear that the community figure has an independent effect on mitigating flight, in the eyes of neighborhood residents.

Establishing the independent impact of community figures is important because a local association may not be present, or it may be ineffectual or exclusionary. Where associations are either absent or ineffective, residents wishing to remain must seek out other means. Next, local associations may be corrupted by criminal groups (Abello-Colak & Guarneros-Meza, 2014) or exclude certain residents based on ethnicity, political leanings, or socio-economic status (Bell-Martin & Marston, 2019, p. 10). A resident threatened by the criminal group, then, would not be able to appeal to the local association, nor would excluded residents, all of whom would need an alternative means to remain in their homes. Such a means might be leveraging their personal ties, as my theory elaborates.

Another strand of the civil wars literature suggests that citizens protect themselves only through direct collaboration with or joining the controlling armed group (Kalyvas & Kocher, 2007; Steele, 2009, p. 424). The difference between these hypotheses and my own is that collaboration requires actively helping the gang in exchange for remaining in the area, while my theory shows how residents stay without giving up their neutrality or aiding the group. According to my theory, when a resident at risk of flight remains, she does so through her ties to a community leader who intervenes on her behalf with the armed group without ever providing material or other support for the belligerent, let alone joining it. Similarly, a resident may be threatened precisely for not collaborating with the gang but may still manage to stay because she has a connection to a gang member other than the one making the threat.

While some residents undoubtedly collaborate with or join the armed group to remain in the neighborhood, many more stay without forfeiting their neutrality. First, the responses to open-ended survey questions above reveal that despite receiving threats to join or get out, some residents successfully use their interpersonal ties to stay in the neighborhood while still avoiding gang recruitment. Moreover, if the best means to stay after receiving a threat is to collaborate with the gang, then we would expect to see everyone who received threats also storing things for the gang (a proxy for collaboration). In the survey, among those who stayed through the years of high violence, 8.3% report having received at least one threat; and, 5.8% stored something—guns, money, or drugs—for the gang. However, the overlap between the two groups is a mere 14% (see Supplemental Appendix). It is highly likely then, that some people stayed without ever collaborating with the gang, even though they received threats.

Third, several migration scholars posit that the residents who remain are those without sufficient financial resources to leave, along with those who have no family elsewhere to receive them, or whose social networks are unable or unwilling to take them in (Adhikari, 2013; Bohra-Mishra & Massey, 2011; Moya & Ibáñez, 2016). As my theory predicts, however, the effect of gang ties remains statistically significant in the regression analysis even after controlling for finances (see Figure 8 in the Supplemental Appendix). Moreover, some people without family or acquaintances elsewhere still leave. Indeed, over 35% of urban displaced persons surveyed reported not having a place elsewhere to stay at the moment they left and, in open-ended questions, some respondents mention spending their first nights in the streets or a shelter.⁴⁹ Finally, for many, the decision to stay or go is one made with agency,⁵⁰ as was the case with a 31-year-old survey respondent who says she remained through the years of high violence, “Because I loved the neighborhood, despite us having a place to go to.”

A fourth alternative explanation is that individuals who remain in their homes may have preferential access to the state. While state capacity is widely recognized as limited in many of the peripheral neighborhoods that skirt cities in the Global South (O’Donnell, 1993), those residents who enjoy greater access to the state may leverage it for increased protection. This is not reflected in the data, however. First, in the linear probability models used to determine factors associated with displacement, the effect of gang ties—the mechanism theorized here—remains statistically significant after controlling for state capacity (see Figure 22 in the Supplemental Appendix). Second, survey respondents who fled and those who stayed report similar levels of (dis)trust in the police during the time of flight (or time of violence), which I take as a proxy for state access. Were state access to matter, we would expect significantly more people who remained to report trust in the police (see Supplemental Appendix, Figures 19 and 20). Last, one resident who had exceptional access to the state was reticent to use it. Referencing the crosshairs of a gun he says calling the police “exposes you to the gang . . . I’m going to be in their sights.”⁵¹

Finally, rather than ties to a gang member limiting flight (as H2 suggests), the inverse might hold. Indeed, some displaced persons in the survey report fleeing precisely because they had ties to someone in or close to the gang (see Supplemental Appendix, Figure 21). However, as Figure 6 shows, most displaced persons report having minimal or no relationship with the gang that forced them to flee. And, as shown in the regression analyses, ties to the gang are a statistically significant predictor of staying rather than fleeing.

Figure 6.

Displaced persons: Connection to gang (at time of displacement).

Conclusion

In this paper, I ask how, in gang-controlled areas where their neighbors are targeted and fleeing, some residents manage to resist displacement. I inductively generated a new theory based on extensive qualitative field research in marginal, displacement-prone communities in Medellín, Colombia, then tested my hypotheses using an original survey and survey experiment. I show that the individuals and families most likely to remain despite being at risk of displacement are the “well connected.” Some residents leverage their ties to an important community figure, who shields them from displacement by intervening on their behalf with the gang. Other residents who, due to their activism or norms breaking, would be expected to have fled, instead manage to remain due to their ties to the gang, which keep the criminal group from targeting these residents in the first place.

My argument complements existing theories describing how residents resist displacement by adding a new mechanism to the literature, namely interpersonal ties. The other theories propose that residents stay in their homes by joining a local association or the armed group, or that residents who remain have preferential access to the state or lack the funds or social ties to leave. Instead, I show that individuals and families have agency in conditions of violence and, in some cases, can resist displacement even while maintaining their autonomy from local associations and armed groups. Granovetter (1973) first pointed out the strength of weak ties. I extend his work to violent settings by showing that residents leverage their interpersonal ties to resist displacement.

While evidence for this study is drawn from one city, the findings are relevant outside Medellín. Despite the recent civil war in Colombia, the street-level gangs of Medellín are driven by profits over politics. They are criminal organizations much like the clikas of Central American cities and local gangs of Mexican and Brazilian cities, where forced displacement also occurs (Cantor, 2014; Muggah, 2015). Medellín shares much in common with other Latin American and Global South cities, including factors relevant to my theory such as marginal neighborhoods, gangs, and local community leaders (Abello-Colak & Guarneros-Meza, 2014). If anything, Medellín has more state capacity than peer cities (Abello-Colak & Guarneros-Meza, 2014, p. 3277) but, as laid out above, my theory on interpersonal connections holds even in state absence. For these reasons, Medellín is a representative case for my theory.

The mechanism I uncover—interpersonal connections—may apply to residents remaining in contexts of criminal violence other than Medellín. One condition of my theory is that the criminal gang must have ties to the community, so that residents can leverage these to remain in place. Some criminal groups elsewhere, in fact, appear to be embedded in their communities and to show more restraint under these conditions (e.g., Arias, 2006, p. 73, 125; Arias & Barnes, 2017, p. 455). For instance, Cantor (2014) finds that members of local gangs (clikas) in Central American cities live within the gang’s territory and that, “Less violence is applied where inhabitants of a ‘core’ zone are family or are otherwise tolerant of the clika . . . (pp. 40–41).”

Does this mechanism also hold in civil wars? Rather than apply to all civil wars, I contend that civilians may be able to leverage interpersonal connections to stay despite threats in those conflicts where belligerents have ties to nearby communities, examples of which appear to be relatively numerous (e.g., Arjona, 2016, p. 1371; International Committee of the Red Cross, 2018, p. 54). For one, Muana (1997) finds that, in the civil war in Sierra Leone, the Kamajoi militia was responsible for far fewer civilian abuses than the Revolutionary United Front, in part because soldiers were

conscripted with the approval and consent of the traditional authority figures, maintained and commanded by officers loyal to those chiefs. This ensures . . . against atrocities on the civilian population on whom they rely for sustenance, legitimacy and support. (p. 88)

In conclusion, my theory requires, first, that an armed group has ties to the community and, second, that civilians and community leaders are able to leverage those ties to constrain the belligerent. While further research is necessary, these crucial conditions for the theory seems to be found in several criminal and civil war settings.

The forces that spark displacement in cities are unlikely to abate. Urbanization and urban conflict are increasingly common, while non-state armed groups continue to exercise varying degrees of sovereignty over urban territories. Continued study of forced displacement is worthwhile because it is likely to persist into coming years. But more so, it merits investigation because non-state armed groups are determining who is, and who is not, a part of a given polis in violation of individuals’ human rights to home, community, and participation.

Research Data

Marston_ResistingDispl_CPS_Analysis for Resisting Displacement: Leveraging Interpersonal Ties to Remain despite Criminal Violence in Medellín, Colombia

Marston_ResistingDispl_CPS_Analysis for Resisting Displacement: Leveraging Interpersonal Ties to Remain despite Criminal Violence in Medellín, Colombia by Jerome F. Marston in Comparative Political Studies

This article is distributed under the terms of the Creative Commons Attribution 4.0 License (http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us-sagepub-com-s.web.bisu.edu.cn/en-us/nam/open-access-at-sage).

Supplemental Material

Marston_ResistingDispl_CPS_CdbkInstr – Supplemental material for Resisting Displacement: Leveraging Interpersonal Ties to Remain despite Criminal Violence in Medellín, Colombia

Supplemental material, Marston_ResistingDispl_CPS_CdbkInstr for Resisting Displacement: Leveraging Interpersonal Ties to Remain despite Criminal Violence in Medellín, Colombia by Jerome F. Marston in Comparative Political Studies

Research Data

Marston_ResistingDispl_CPS_data for Resisting Displacement: Leveraging Interpersonal Ties to Remain despite Criminal Violence in Medellín, Colombia

Marston_ResistingDispl_CPS_data for Resisting Displacement: Leveraging Interpersonal Ties to Remain despite Criminal Violence in Medellín, Colombia by Jerome F. Marston in Comparative Political Studies

Footnotes

Acknowledgements

I would like to thank Peter Andreas, Catalina Arreaza, Nicholas Barnes, Aimée Bourassa, Robert Blair, Gustavo Duncan, Elizabeth Fussell, Jon Gordon, Giovanni Mantilla, Rebecca Bell-Martin, David Skarbek, Richard Snyder, Cadence Willse, and Rebecca Weitz-Shapiro as well as three anonymous reviewers.

Author’s Note

Versions of this paper have been presented at the Brown Political Science Department IR/CP Workshop and the Migration Working Group, Population Studies Training Center, Brown University. All mistakes are my own.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The author would like to acknowledge funding for this research from Fulbright-Hays DDRA, the Graduate Program in Development at the Watson Institute, the Cogut Center Brown-in-the-World Travel Grant, the Joukouwsky Dissertation Research and Travel Funds (Brown University), and the Populations Studies Training Center at Brown University.

ORCID iD

Jerome F. Marston Jr.

Supplemental Material

Supplemental material for this article is available online at the CPS website

Notes

Author Biography

Jerome F. Marston Jr. is a researcher at the Global Coalition to Protect Education from Attack and a global fellow at the Center for Human Rights and Humanitarian Studies at the Watson Institute, Brown University. He holds a PhD in political science from Brown University.

References

Abello-Colak

Guarneros-Meza

(2014). The role of criminal actors in local governance. Urban Studies, 51, 3268–3289.

Adhikari

(2013). Conflict-induced displacement, understanding the causes of flight. American Journal of Political Science, 57, 82–89.

Albuja

(2014). Criminal violence and displacement in Mexico. Forced Migration Review, 45, 28–31.

Andreas

(2008). Blue helmets and black markets: The business of survival in the Siege of Sarajevo. Cornell University Press.

Arias

E. D.

(2006). Drugs and democracy in Rio de Janeiro: Trafficking, social networks, and public security. University of North Carolina Press.

Arias

E. D.

(2017). Criminal enterprises and governance in Latin America and the Caribbean. Cambridge University Press.

Arias

E. D.

Barnes

(2017). Crime and plural orders in Rio de Janeiro, Brazil. Current Sociology, 65, 448–465.

Arjona

(2016). Rebelocracy: Social order in the Colombian civil war. Cambridge University Press.

Auyero

Kilanski

(2015). Managing in the midst of social disaster: Poor people’s responses to urban violence. In Auyero

Bourgois

Scheper-Hughes

(eds) Violence at the urban margins (pp. 189–211). Oxford University Press.

10.

Bada

Feldmann

A. E.

(2019). How insecurity is transforming migration patterns in the North American corridor: Lessons from Michoacán. In Feldmann

A. E.

Bada

Schütze

(Eds.), New migration patterns in the Americas: Challenges for the 21st century (pp. 57–83). Palgrave Macmillan.

11.

Barnes

(2017). Criminal politics: An integrated approach to the study of organized crime, politics, and violence. Perspectives on Politics, 15, 967–987.

12.

Bell-Martin

R. V.

Marston

J. F.

Jr. (2019). Confronting selection bias: The normative and empirical risks of data collection in violent contexts. Geopolitics. Advance online publication. https://www-tandfonline-com-s.web.bisu.edu.cn/doi/full/10.1080/14650045.2019.1659780

13.

Betts

(2013). Survival migration: Failed governance and the crisis of displacement. Ithaca, NY: Cornell University Press.

14.

Bohra-Mishra

Massey

D. S.

(2011). Individual decisions to migrate during civil conflict. Demography, 48, 401–424.

15.

Cantor

D. J.

(2014). The new wave: Forced displacement caused by organized crime in Central America and Mexico. Refugee Survey Quarterly, 33, 34–68.

16.

Corporación Region. (2017, September 5). Editorial: El desplazamiento forzado dentro de la ciudad crece . . . y seguimos “Como si Nada.” [Editorial: Forced displacement inside the cities grows … and we continue “as if it were nothing.”] http://www.region.org.co/index.php/opinamos/item/247-editorial-el-desplazamiento-forzado-dentro-de-la-ciudad-crece-y-seguimos-como-si-nada

17.

Duncan

(2014). Drug trafficking and political power: Oligopolies of coercion in Colombia and Mexico. Latin American Perspectives, 41, 18–42.

18.

Edwards

(2017). Forced displacement worldwide at its highest in decades. United Nations High Commissioner for Refugees, 19 June. http://www.unhcr.org/news/stories/2017/6/5941561f4/forced-displacement-worldwide-its-highest-decades.html

19.

Fenno

R. F.

(1978). Home style: House members in their districts. Little, Brown.

20.

Gordon

(2019). The legitimation of extrajudicial violence in an urban community. Social Forces, 98, 1174–1195.

21.

Granovetter

(1973). The strength of weak ties. American Journal of Sociology, 78, 1360–1380.

22.

Granovetter

(1995). Getting a job: A study of contacts and careers. University of Chicago Press.

23.

Ibáñez

A. M.

Moya

(2016). Who stays and who leaves during mass atrocities. In Anderton

C. H.

Brauer

(Eds.), Economic aspects of genocides, other mass atrocities, and their prevention (pp. 251–273). Oxford University Press.

24.

Ibáñez

A. M.

Vélez

C. E.

(2008). Civil conflict and forced migration: The micro determinants and welfare losses of displacement in Colombia. World Development, 36, 659–676.

25.

International Committee of the Red Cross. (2018). The roots of restraint in war.

26.

Internal Displacement Monitoring Center. (2017). Global report on internal displacement 2017.

27.

Kalyvas

(2006). The logic of violence in civil war. Cambridge University Press.

28.

Kalyvas

S. N.

(2015). How civil wars help explain organized crime—And how they do not. Journal of Conflict Resolution, 59, 1517–1540.

29.

Kalyvas

S. N.

Kocher

M. A.

(2007). How “free” is free riding in civil wars? Violence, insurgency, and the collective action problem. World Politics, 59, 177–216.

30.

Kaplan

(2017). Resisting war: How communities protect themselves. Cambridge University Press.

31.

Krakowski

(2017). Resisting displacement amid armed conflict: Community-level conditions that make people more likely to stay. Journal of Peacebuilding & Development, 12, 68–84.

32.

Lindley

(2010). Leaving Mogadishu: Towards a sociology of conflict-related mobility. Journal of Refugee Studies, 23, 2–22.

33.

Marston

J. F.

Jr. (2019). To stay or flee? Displacement due to gang violence in Medellín, Colombia. https://doi.org/10.26300/3had-0v06

34.

Massey

(1987). The ethnosurvey in theory and practice. The International Migration Review, 21, 1498–1522.

35.

Mesa

G. S.

(2015). La cuadra times [The block]. Cámara de Comercio de Medellín para Antioquia.

36.

Muana

(1997). The Kamajoi militia: Civil war, internal displacement and the politics of counter-insurgency. African Development, 22, 77–100.

37.

Muggah

(2015). The invisible displaced: A unified conceptualization of population displacement in Brazil. Journal of Refugee Studies, 28, 222–237.

38.

O’Donnell

(1993). On the state, democratization and some conceptual problems: A Latin American view with glances at some postcommunist countries. World Development, 21, 1355–1369.

39.

Patillo-McCoy

(1999). Black picket fences: Privilege and peril among the black middle class. University of Chicago Press.

40.

Personería. (2015). Informe sobre la situación de derechos humanos en la ciudad de Medellín [Report about the situation of human rights in the city of Medellín]. Universidad de Medellín. http://www.personeriamedellin.gov.co/index.php/informacion-al-ciudadano/documentos/informes-derechos-humanos-ddhh/category/38-informes-ddhh-2015?download=161:informe-ddhh-2015

41.

Ramírez

J. G.

Rendón

A. J. C.

Duncan

G. C.

(2014). Nuevas Modalidades de Captación de Rentas Ilegales en Medellín [New modalities of illegal rent capture in Medellín]. Universidad EAFIT.

42.

Sánchez

B. M.

(2016). A silenced exodus: Intra-urban displacement in Medellín. In Cantor

D. J.

Serna

N. R.

(Eds.), The new refugees: Crime and forced displacement in Latin America. Institute for Latin American Studies.

43.

Sánchez-Jankowski

(1991). Islands in the street: Gangs and American urban society. University of California Press.

44.

Staniland

(2010). Cities on fire: Social mobilization, state policy, and urban insurgency. Comparative Political Studies, 12, 1623–1649.

45.

Steele

(2009). Seeking safety: Avoiding displacement and choosing destinations in civil wars. Journal of Peace Research, 46, 419–429.

46.

Steele

(2017). Democracy and displacement in Colombia’s civil war. Cornell University Press.

47.

Thachil

(2018). Improving surveys through ethnography: Insights from India’s urban periphery. Studies in Comparative International Development, 53, 281–299.

48.

The World Bank. (2017). Forcibly displaced: Toward a development approach supporting refugees, the internally displaced, and their hosts.

49.

Wood

E. J.

(2003). Insurgent collective action and civil war in El Salvador. Cambridge University Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.10 MB

8.60 MB