Over the last several months, much of American and Canadian reporting on organized criminal violence in Latin America has been oriented around a single topic: the potential threat of northward-moving migrants and asylum-seekers as they attempt to escape their violence-ravaged countries. Not only has this prejudicial portrayal of migrant caravans as disguised MS-13 gang members served to justify the adoption of policies in the United States that have separated families and tear-gassed civilians on the “wrong side of the border,” so too has it been dangerously myopic in a separate capacity. In shifting the nature of victimhood away from asylum seekers and donning it ourselves, we risk concealing and overlooking important variation in the source and scope of the threats posed by criminal groups directly to citizens of Latin American countries. Figuring out where, when, why, and how violence is produced ought to take precedence over concerns regarding the indirect and imagined externalities of that violence.
Yet developing effective policies aimed at both protecting civilians and combatting the immediate security threats posed by criminal groups first demands that we properly identify who the victims of criminal violence are and this has proven to be rather difficult. One reason for this is that scholars of criminal violence continue to struggle with developing satisfying theories linking the territoriality of criminal organizations—the degree to which criminal groups make claims over the areas in which they operate—and the geographic spatiality of both the violence they produce and the civilian victimization associated with such violence. While we have learned a lot in the last couple years about how criminal groups operate, why they produce violence, and why they persist in the face of government repression, getting the question of criminal territoriality right is crucial. In order to predict the degree to which civilians are threatened by the criminal groups with which they share space, we must first understand criminal organizations’ relationship to that space and likely disposition to produce violence therein.
Theories of criminal territoriality have come in three forms to date. The first school of thought argues that criminal groups deeply resemble rebels and insurgents in their desire to control territory held by the state. While criminal organizations like FARC or ISIS may not have clear political ambitions, this line of reasoning argues that their need to control the flow of contraband, run extortion rackets, and/or prevent the state from disrupting their accumulation of illicit rents encourages criminal organizations to construct territorial enclaves within which they seek to maintain a monopoly of power. The second school of thought argues that criminal groups, as primarily economically-motivated entities, are not fundamentally interested in replacing or conquering the state; therefore, they have are believed to have no motivation to establish areas of exclusive territorial control. Rather, while they may be motivated to maintain operational monopolies within their turfs vis-à-vis criminal competitors, this body of literature argues that criminal groups can, and often do, share space with the state peacefully should government forces prove willing and able to cooperate. Bridging these bodies of literature, the third school argues that criminal groups operating in different industries are likely to vary in their degree of territoriality; while gangs may be very territorial—and therefore likely to combat the state in the areas where they operate—drug-trafficking organizations are believed to be more transactional and less grounded in the particular areas where they work.
In seeking to identify the merits of each of these theories of criminal territoriality, I began mapping criminal homicides, a generally agreed upon metric of civilian victimization by criminal groups, in several Latin American countries and found something interesting. A discussion of two countries, Mexico and El Salvador, helps to illustrate the point.
Mexico 2009In Mexico, criminal conflict and violence has typically been characterized by a shifting series of turf wars between hierarchically-structured drug-trafficking organizations. While Mexican cartels had previously succeeded in peacefully divvying up the criminal turf, the cocaine boom of the 1980s and the transition to democracy had the dual effect of increasing criminal competition while destabilizing networks of government corruption and patronage. By the early 2000s, the two largest criminal organizations—the Gulf and Guajajara cartels—had started to fragment, producing a wave of turf wars that spanned across the country. After President Calderon launched his War on Drugs in 2006, violence exploded as the government sought to crack down on criminal groups and cartels looked to take advantage of newfound weaknesses in their competitors’ operations. Yet, while certain areas of the country have come to be connoted with criminal conflict, it is interesting to note that the center of gravity of criminal violence in Mexico has shifted significantly over time.
[Above are images of Local Moran’s I cluster analyses (using queen weights with two orders of inclusive contiguity) of the Empirical Bayes drug-related homicide rate (homicides associated with criminal turf wars or conflicts between the state and a drug cartel) from 2007 to 2011. These data were collected by the Empirical Studies of Conflict (ESOC) project at Princeton University.]
As the above cluster maps demonstrate, violence in 2007 was concentrated in Tierra Caliente in the South around the Lazaro Cardenas port, in Culiacan, Sinaloa, and around two, small border crossings, Nogales and Agua Prieta. In 2008, violence erupted in Sonora and Chihuahua as the Sinaloa Cartel launched a turf war against its competitor in Ciudad Juarez. By 2009, this cluster appears to have split into two distinct turf wars—one on the border and one in Durango. In 2010 there was a major shift as this latter turf war subsided and violence moved east to Nuevo Laredo and Reynosa on the Texas border, coinciding with the split between the Gulf Cartel and its enforcement wing, Los Zetas. By 2011, violence had concentrated more heavily on the east coast and a new flareup had emerged in the west.
What these maps effectively illustrate is a story of shifting alliances, group fragmentation, selective government offenses, and the emergence of continuously new war fronts in a complicated, country-wide, criminal conflict. Had the data extended beyond 2011, one would see the development of new hot-spots across the country and the disappearance of others, shifting on a yearly basis in a way that does not coincide clearly with any geographic or structural predictors of turf war vulnerability, like shipping plazas or processing areas, rendering it exceptionally difficult to predict where violence is likely to emerge and how bad it will be.
This pattern of violence was not replicated in El Salvador. Over the course of that country’s civil war, thousands of refugees fled to the United States seeking asylum. In American prisons, young, marginalized, Guatemalan, Honduran, and Salvadoran men learned from their US counterparts to organize themselves into the two major street gangs we know today—the Mara Salvatrucha (MS-13) and the Eighteenth Street Gang (Barrio 18). After their members were extradited back to their respective countries in the mid to late 1990s, both gangs came to establish dense criminal turfs in poor, peri-urban communities across Central America where they engage, to this day, in the extortion of local businesses and civilians, collecting “war taxes” in exchange for protection from the very threats they themselves pose. Recent estimates suggest that there are tens of thousands of mareros (gang members) in El Salvador today, about a third of which are incarcerated, including the gangs’ current leadership.
After producing wave after wave of violence in the mid-2000s in response to a rather militant and indiscriminate policing strategy referred to as Mano Dura, both gangs signed peace agreements with the government, which led to a dramatic decrease in the homicide rate in 2011 and 2012. By 2014, however, these agreements had fallen apart as the government could no longer sustain the optics of sanctioning the continued operations of these two, major criminal organizations. Although not captured in this dataset,1 violence has skyrocketed since then. It has, however, maintained a similar spatial pattern.
[Above are images of Local Moran’s I cluster analyses (using queen weights with one order of contiguity) of the Empirical Bayes mara-committed homicide rate from 2010 to 2014. These data were collected by the Policia Nacional Civil (PNC) of El Salvador.]
In exploring the geographic distribution of criminal violence in El Salvador over the time period in question, one notices that gang-committed homicides remained remarkably concentrated in and around the capital of San Salvador, irrespective of the national politics and temporary ceasefire. Due to the notable consistency in the spatiality of criminal violence, the population vulnerable to criminal conflict has been rather easy to identify: socially, politically, and economically marginalized communities in the urban periphery, primarily around the capital city.
This latter description differs radically from what we see at the national level in Mexico and, I believe, bringing these two cases together facilitates a new way of thinking about criminal territoriality—and civilian victimization by criminal groups—that fails to be captured fully by the existing literature.
To begin, the very fact that criminal operations and violence take place within prescribed geographical spaces suggests that all criminal groups must be territorial to some extent, irrespective of their political motivations. Thus, while it goes without saying that drug-trafficking organizations behave differently than street gangs, referring to the latter as territorial and the former as transactional seems to miss an important reality: that it is the nature and dynamism of criminal groups’ territoriality that varies, not the extent to which they are territorial. All criminal organizations are fundamentally preoccupied with managing and maintaining control over their turfs. What differs, however, is the degree to which these turfs are likely to move or change location over time.
In Mexico, the fluid nature of alliances, government offensives, and novel narcotic processing and shipping methods produced a series of transitory flare-ups between cartels in a dynamic turf war that has changed radically since 2006. Shipping routes shifted, opportunities emerged, and the geographical turfs that criminal groups sought to maintain and expand changed in a corresponding manner. In El Salvador, by contrast, criminal groups extract their illicit rents almost exclusively from a stationary civilian population, leading the turf wars that rage between gangs to more closely resemble wars of attrition. Given that the source of illicit rents in that country is stationary, so too must be the territoriality of its criminal groups. Criminal organizations in both Mexico and El Salvador seek to establish dominion over their turf; it just so happens that what constitutes their respective turfs varies considerably.
Just by being able to map instances of criminal violence in both countries, we gain critical insight into thinking about who suffers at the hands of criminal groups, as well as the challenges associated with preventing or reducing that suffering. In Mexico, while some areas have certainly been hit harder than others, the shifting nature of the Drug War suggests that, without a national solution, the spatiality of violence is likely to continue shifting as cartels seek to exploit new opportunities to smuggle drugs into the United States and eek out advantages against their competitors. Protecting civilians, in this context, becomes incredibly difficult as the populations likely to be vulnerable to criminal violence cannot easily be identified and protected before the onset of hostilities. In El Salvador, there is a different problem: while the threats that criminal groups pose are highly localized and, thus, easier to predict, maras are deeply entrenched within their criminal turfs and are prepared to combat the state and their rivals no matter the costs and consequences to both themselves and to the civilians on which they prey.
There are no easy solutions to managing either of these conflicts. There remains, however, little advantage to be gained by being fatalistic about how endemic they are or by prioritizing the second-hand threats we face over the first-hand victimization of those who suffer the violence of criminal groups directly. By improving our understanding of how these organizations operate within the spaces that contain them, we can better conceive of, and design policies that incorporate more refined and accurate understandings of the localized nature of criminal groups’ territoriality and the potential threats they pose to the communities that reside within their respective turfs.
If you liked this article, please consider becoming a Patron and contributing to the work we do here at The Mantle.
- 1. Information regarding general homicide rates can be found on the PNC site at https://www.transparencia.gob.sv/institutions/pnc/documents/estadisticas.