Geografía humana


Analysing the geographies of the 'transnational' gangs of Central America: the changing spaces of violence


Analizando las geografías de las pandillas 'transnacionales' en Centroamérica: los espacios cambiantes de la violencia


Ailsa Winton*


* El Colegio de la Frontera Sur, Carretera Antiguo Aeropuerto km 2.5, 30700, Tapachula, Chiapas, México. E–mail: ailsawinton@hotmail.com


Recibido: 1 de julio de 2011.
Aceptado en versión final: 9 de abril de 2012.



The aim of this paper is to contribute to knowledge of 'transnational' youth gangs in the Central American region, through an analysis of the mutually constitutive processes of identity, space and place production. It is argued that insights into gangs gained through analyzing their spatial dynamics and practices, and discussing ways in which these dynamics and processes connect the local and global scales, offer useful knowledge concerning the functioning of these gangs in a field still lacking in–depth academic research. Drawing on over a decade of direct research with young people in the region, the paper finds that poor understanding of gangs inevitably leads to ineffective, counterproductive interventions, and demonstrates that the geographies of maras are a fundamental –and still neglected– aspect of their development and transformation.

Key words: Youth gangs, transnational, Central America, violence, space.



El objetivo de este trabajo es dar a conocer nuevo conocimiento sobre las pandillas 'transnacionales' en la región de Centroamérica, a través del análisis de los procesos (mutuamente constitutivos) de la producción de identidad, espacio y lugar. Se argumenta que el análisis de las dinámicas y prácticas espaciales, y la discusión de las diferentes maneras en que estas dinámicas y procesos se conectan a la escala local y global, nos proporcionan nuevo conocimiento 'útil' en este campo de investigación aún naciente. Basado sobre más de una década de investigación directa con jóvenes en la región, este trabajo argumenta que entendimientos pobres de las pandillas, resultan en intervenciones ineficaces y contraproducentes, y demuestra que las geografías de las maras representan un aspecto fundamental –y todavía bastante descuidado– de su desarrollo y transformación.

Palabras claves: Pandillas juveniles, transnacional, Centroamérica, violencia, espacio.

'MS13 is a name you might have heard, and need to remember. This brutal street gang somehow grew from a local
Los Angeles problem into an international threat, and it may already be in a neighborhood near you ...
It's going international, crossing borders at will, leaving its bloody mark from Central America to the American heartland.'

(Explorer, National Geographic Channel, 12/02/06).



For decades now, youth gangs in the Americas have attracted significant −if uneven– academic, political and popular attention, yet significant empirical changes in gang structures, activities, and formation brought about by a range of global–local transformations have not, until very recently, led to any fundamental shift in how we think about gangs. This paper is an attempt to explore the various ways in which youth gangs have themselves become 'glocalized', both as subjects and actors within shifting political, economic, and social landscapes.1 It of course goes without saying that globalization is not a uniform process of change, that global processes are always filtered through local conditions, and it is this channel from the global to the local (the so–called glocalization) which informs this work. Focusing on the emergence and spread of so–called 'transnational' gangs in the Americas, exploring the conditions, networks, and flows involved in creating and sustaining such a transnational organization (if indeed it can be represented as such), the paper aims to explore the shifting relationship between violence, space and territory with respect to the youth gangs known as 'maras'. It is argued in this paper that it useful to try and gain further insights into gangs through the lens of their spatial dynamics and practices.


Understanding global–local spaces

The historical dominance of US gang research, and within this criminology, has profoundly shaped both understandings of, and responses to this type of youth organization in much of the region. While the pioneering work of Thrasher (1927) on immigrant groups in 1920s Chicago notably did not include criminality as a defining feature of youth gangs, since then, gangs became seen as unilaterally criminal, deviant, pathological, at the expense of more nuanced readings of their motivations, values, practices and organization. While the Chicago School was a significant contribution to social science, the entrenchment of human ecology, but particularly the ideologically–loaded notion of 'so cial disorganization' in the study of gangs has been problematic, especially in the late–modern, post–industrial era when models of assimilation are no longer even theoretically viable (see Hagedorn, 2007c).

Yet, there are ideas here which can and have been rescued. In principle, its focus on the spaces of the city as key to differential urban processes was highly significant, even if the relevance of intentional, political structuring of these spaces was overlooked (Hagedorn, 2007a). Rather, then and this applies broadly to many contemporary approaches to gangs, be they cultural theories of gangs as a lower class subculture, political approaches to gangs as a form of resistance to blocked opportunities, economic approaches to gangs as entrepreneurs, or psychological interpretations of gangs as the result of individual pathology (Rodgers, 2006) −the challenge now is to adapt theory to the late–modern, globalized, post–industrial era.

It is curious to note that the rather more philosophically and theoretically holistic criminology of Latin America has not, to date, engaged to any significant level with the issue of gangs in Central America. After an initial stream of sociological epidemiological studies, some more detailed analyses have emerged recently (particularly, Cruz 2010; Valenzuela et al., 2007; also Reguillo, 2005; Rodgers, 2006, 2007; Zilberg, 2004), although not yet on the scale of research in the US. This is noteworthy given the elevated public and political profile of the gang issue in Central America in recent years, although of course this itself might well have encouraged the production of more descriptive, perhaps even polemical work at first.

Thus, it is argued, the present approach may have something to offer this emerging field of research. Certainly, there are few studies of gangs related to different scales of geographical analysis, surprising in the case of the maras given their supposedly 'transnational' characteristics (excepting Zilberg, 2004) and few dedicated to the spatial aspects of gangs in relation to their evolution (Cruz, 2010, being a notable exception). Much early work on gangs in Central America was dedicated to understanding what caused young people to turn to gangs, and what structural conditions allowed these configurations to take hold in marginal Central American communities, together with suggestions as to how they may be prevented (see for example, Cruz and Portillo, 1998; ERIC et al., 2004; Rocha, 2000, 2006; Santacruz–Giralt et al., 2002; Smutt and Miranda, 1998; also Winton, 2004). Subsequently, attention became focused on the pernicious effects of the extreme anti–gang policies of the last decade (see for example Hume, 2007; Savenije, 2004a, 2004b; Thale, 2006), while more recently, and in part as a consequence of these developments, debate has hinged on the extent to which gangs may be considered transnational organizations, or whether they remain locally embedded.

This interest in the transnational aspects of these gangs is related at least in part to the perceived threat to Central and North American states that comes with the alleged scaling up of gang organization and activities, and also to the associated political responses to these threats. It is worth noting that despite documenting a series of important transformations in the way gangs operate in recent years (see below), most academic research does not find significant 'mafia–style' organization at a transnational level among gangs. In part to offset this apparent hysteria over the gang take–over of the Americas, others have focused on the transnational aspects of gangs more in terms of deepening understandings of the socio–cultural significance of transnational gang identities (for example Valenzuela et al., 2007).

It is worth noting that a serious limitation to researching gangs in this or any context is lack of access to verifiable data on their internal structure and leadership. This is extremely closely guarded information, and the spread of false information is rife, just as gangs are diverse, such that any bold statements about gang organization in Central America ought to be viewed with some scepticism. There are simply too many contradictory accounts to be able to paint a complete empirical picture of gang structure in the region. What we can hope to do is enrich our understandings through careful reflection on research and knowledge. This paper hopes to contribute to this endeavor in contemplating the spatial and scalar aspects of transforming gang practices in the Central American context, specifically in terms of 'trans–local' dynamics.


Gangs in global–local spaces

To frame this discussion of the spatial and scalar dynamics of gangs, this first section attempts to outline their context in Central America, couching this in terms of the city as a space of alienation. If, as Joan Moore (1998:75) suggests, 'conditions in many nations [...] are ripe for the formation of American–style youth street gangs', what might such conditions look like? Why was it possible for the maras to gain such a strong foothold in urban Central America?


The city as a site of economic alienation

Urban settings, and more precisely urbanization, have long been associated with violence, and it has traditionally been posited that social relations are weaker in cities than in the countryside, with socially atomized urban dwellers being both more vulnerable to violence and to being violent (Rodgers, 2007).2 But, as Hobsbawm (2005, cited in Rodgers, 2007:7) has pointed out, violence does not emerge in cities because they are inherently alienating spaces, but rather because they are spaces inhabited both by the poor and by the political powers and elites that affect their lives, and thus 'violence is inevitably a function of the economic and political relations that exist within a city'. The city concentrates diversity. Similarly, for Mike Davis, (2004, 2006 cited in Rodgers, 2007) urban slums in the South are not evidence of cities as 'engines of growth' experiencing rapid economic expansion, as conventional wisdom would have it, so much as a dumping ground for those excluded from increasingly technological and informational production processes.

In such a situation, many people, particularly young people, face 'multiple marginalities' (Vigil, 2006), or 'fragmentation' (Davids, forthcoming). If the post–industrial city has become a site of alienation, it could be argued that the barrio (and perhaps by extension, the gang, Rodgers, 2006) 2 This notion borrows to some degree from social disorganization theory, which as mentioned above has itself played a significant (if contentious) role historically in gang studies, particularly in the US. has become the new sociological basis of collective social life, or 'the proximal place where a person can learn and exercise their capacity for control, and produce security' (Merlo and Milanese, 2000:16). Conflicts at this level, then, become a necessary way of managing marginalization by means of governing belonging (Ibid.:17), and rather than being indicative of social disorganization, constructions and representations of 'deviance' have the role of maintaining a specific kind of social organization. As Nateras Domínguez (2007:131) argues in the case of Latin America, 'structural, symbolic, political, and daily violence become language in the geography of the metropolis', thus violence is not aberrant (as the state is wont to present it) but in fact it makes sense within daily practices (Castillo Berthier and Jones, 2009).

Yet, as the marginality of the barrio mixes with the global–local flows associated with globalization, new spatial configurations and inequalities emerge. An important way in which borders (be they invisible/inferred/imagined/imposed) may be transgressed is through the cultural penetration of globalization: global communication networks facilitate, if always unevenly, relatively 'borderless' flows of information. Yet this information can itself heighten marginality, as the gap between young people's aspirations and their capacity to reach them grows ever wider.


The city as a site of political alienation

Elsewhere, it has been argued that the transformation and continuation of violence in contemporary Central America is strongly related to the particularities of the local institutional landscape, whereby a wide range of violent institutions and groups (state, private, civil, formal, informal, criminal), enter into relationships of conflict, collusion, or collaboration, resulting in a complex web of violent perpetrators (Winton, 2011). In addition to these aspects of the Central American context, it is useful to consider the institutional landscape at a national level, and state capacity more specifically.

A helpful way of thinking about state capacity, particularly in relation to gangs, is through the notion of 'vulnerability'. As Kirby (2004) argues, the strength of the notion of vulnerability is that it considers not only the threats to a system, but also focuses attention on the ability of a system to cope with these threats. There are thus two sides to vulnerability, which are mutually reinforcing: an increase in risk on the one hand (increased and diversified criminal activity), and the erosion of mechanisms to cope with those risks (state inefficiency, transnational nature of crime) on the other. By extension, therefore, a vulnerable state allows gangs to prosper in the first place, and then later becomes itself vulnerable to the pressures placed on the system by organized para–legal non–state actors, including gangs. Gangs also fit into an absence of legitimacy, creating a kind of para–legal order. Indeed, for Reguillo, (2005) the power of para–legality is far greater than illegality, since the former implies a disregard for the system and the beginning of an alternative pact.

For the individual, the state is no longer able to make citizenship meaningful, which has long been considered the most basic form of societal membership, hence: 'the project of a national society of citizens ... appears increasingly exhausted and discredited' (Rodgers, 2003). Social fragmentation and polarization can be countered by the development of an alternative societal membership, whereby gangs become 'a resource with which to obtain an acknowledged identity' (Briceño and Zubillaga, 2002:27). So it is, that in the context of a situation of such manifold alienation and deprivation, some young people may find in a gang (among other factors) 'social recognition and prestige, economic reward, wellbeing and power' (Nateras, 2007:139). As one Honduran gang member in Tapachula, Mexico, expressed: 'I wanted to be someone in the gang. I wanted to be respected... and I wanted to have some kind of say in things, because when you're nobody, they treat you like... well...'.3

Thus, in the absence of an effective democracy, space becomes the battleground for (re)claiming any kind of position and belonging in life (Davids, forthcoming). This speaks not only to the legitimization or institutionalization of gangs, but most importantly in the present case, to the importance of locally based territorial structures such as gangs, whereby space becomes an overt source of power (see below). In this sense, the spaces of the barrio become the site not only of the accumulation of structural forces creating conditions necessary to sustain an alternative social structure such as a gang, but they also become the site for these very structures to be challenged.

These factors, of course, can be only part of the explanation of the spread of youth gangs in Central America. We must not forget that most young people living in the conditions described do not find it necessary to join a gang (see also Rodgers et al., 2008). If young people who get involved in gangs find something there that they were missing, most young people have the resources (social, cultural, and economic) necessary to prevent them joining. What is particularly important about these structural issues, however, is that they create the conditions that allow the emergence of gang structures in certain contexts. Of concern here is the interaction of gangs with their environment. Having briefly discussed some important globallocal aspects of the context in which gangs are enmeshed, I now turn to the spaces which they themselves actively produce.


Gangs as global–local spaces

It is argued here that it useful to try and gain further insights into gangs through their spatial dynamics and practices, and to consider the extent to which they represent the local enactment of 'global' processes. 4 It is to these themes that the remainder of the paper is directed, through considering the possibility of the gang itself being a 'glocalized' space.


Tracing a transnational gang: the rise of the maras5

The expansion of the youth gangs known as maras in the US, Mexico and Central America in recent years provides an interesting example not only of the increasing relevance of 'translocality' in gang formation, but also how counter–productive the misguided (transnational) institutional interventions can be. What makes the case particularly interesting is the extent to which transnational networks and flows have so profoundly transformed and scaled–up what started as a local, Los Angeles phenomenon. Across the region, maras have become a spectacular symbol of deviance, defiance and delinquency, yet as the gangs encounter local conditions in Central America, they transform according to local environments. Despite their transnational origins, therefore, I would argue that the gangs remain fundamentally local institutions, although strongly influenced by external/global processes (see also Cruz, 2010; Santamaría, 2007).

The origins of the two gangs making up what are commonly referred to as maras can be traced to gang formation in Los Angeles. First to emerge was the 18th Street Gang, or Barrio 18, which was originally formed mainly of Mexican immigrants, but then became relatively heterogeneous, and later the mara Salvatrucha (or MS13), which was formed mainly of Salvadoran migrants in order to provide some resistance to the Barrio 18 and other local gangs. Although these national affiliations are strong, both gangs in Los Angeles later contained a significant number of Latino youth from other countries, with their affiliation based on territory, a notional Latino identity, but above all attachment to the gang itself. In the words of a gang member from Honduras:

It doesn't matter where we're from. You could be from here [Honduras], from any Department, or from El Salvador, Guatemala, or from the US. As long as you're a Salvatrucha, you're one more member of the family (Savenije, 2004a:2).

While the basic structure and norms of these gangs were born according to local conditions in Los Angeles, international migration flows have caused them to expand into (arguably) transnational organizations, yet operating in a translocal way. In other words, international migration flows have meant that cultural practices and values originally attached to socio–spatial conflicts in the United States have been transmitted to, and transformed within local contexts in many cities in Central America.6

Although maras were reported in Central America during the 1980s (see AVANCSO, 1996), and there have always been bi–directional flows of migrants, legal and illegal, between the US and Central America, it was the vast upsurge in deportations and repatriations during the 1990s which is largely thought to have allowed the phenomenon of the maras to spread to such an extent in Central America (see Cruz, 2010 for a detailed account, also Liebel, 2004; Rodgers et al., 2008; Santamaría, 2007).7 First, the end of decades of civil war in much of Central America in the 1990s meant that there were mass–repatriations of Central American citizens who had fled the conflict during previous decades. Second, concerns among US authorities that Latino gangs were rapidly growing in size and influence led to mass deportations of gang members from 1992 (Savenije, 2004a), further to which changes in immigration laws in 1996 meant that non–citizens, and in some cases foreign–born citizens, who were sentenced to one of more years in prison (among them many gang members) could be repatriated to their country of origin. As a result of these stricter laws relating to the repatriation of felons and illegal immigrants, between 1998 and 2005 the US deported nearly 46 000 convicts to Central America, in addition to 160 000 illegal immigrants. Three countries, El Salvador, Guatemala and Honduras, received more than 90 percent of the deportations from the US, many of whom were gang members (Rodgers et al., 2008). So these young people, many leaving behind entire families in the US, and with no memory of their 'home country', were left to effectively fend for themselves (Zilberg, 2004). Perhaps unsurprisingly, with few or no apparent alternatives, for many continuing gang life was a quite logical decision.

On arrival, many such deportees established clikas which either supplanted existing gangs (pandillas) or absorbed them (Rodgers et al., 2008). As Cruz (2010: 388) finds, 'the maras did not expand because of a premeditated and centralized process, but through a social imitation process based on migration and networking'. Migration is, after all, as much a flow of ideas as it is a flow of people. It is now overwhelmingly the case in cities in El Salvador, Guatemala and Honduras, that the clikas of the mara Salvatrucha and the Barrio 18 are the dominant local gangs, even in some cases where more established gangs still remain. It is important to note, therefore, that deportations had a massive effect not so much on the numbers of young people involved in street gangs per se, but rather on the practices and organization of gangs. Conditions in urban Central America were such that the physical dissemination of 'American' gang structures precipitated important changes in the local (Central American) youth gang landscape. It has been argued in terms of the impact of migration on local gang structures in the US, that gang members can contribute to proliferation 'if they introduce new and exciting cultural distinctions from those existing gangs', whereby they act as 'cultural carriers of the folkways, mythologies and other trappings of more sophisticated urban gangs' (Maxson, 1998:3).

Certainly, the 'imaginary' of the mara is as (if not more) powerful as their physical presence, both as an appealing identity for members, imitators, and sympathizers, as well as a generator of fear and suspicion among the general population. Thus, the shared cultural symbolism of the maras, their practices and rituals (although varied) unite them under a shared identity that may be based more on sense of 'imagined' community, than on real physical connections through space. Indeed, Santamaría (2007) finds that the transnational character of maras in the region is limited to the cultural and symbolic influences shaping their identity and behavior, rather than any more formally organized transnational structures. However, it is certainly the case that these cultural norms and practices morph according to the context in which they are appropriated, thus forming the basis for 'glocal' gang structures. Thus, as micro– and macro–level conditions change over time, so do the structures and functions of gangs. Gangs are constantly in process, evolving, transforming, displacing, being displaced. So how do these new structures affect local geographies? How is the 'gang' spatially constituted?


Rewriting the streets: space, identity and violence

An obvious but nonetheless important point to note is that while deportees did provide the impetus and leadership that brought about the spread of the mara gang structure, the majority of gang members –then as now– were local youths. Of particular note is that the levels of violence used by maras in Central America far exceed the violence of their founding gangs in the US: in a culture of normalized violence such in Central America (see Winton, 2011), violence which is 'exceptional' far surpasses that which would be normatively understood as such in less violent contexts: the bar is set much higher in already–violent contexts. Conversely, it is also true that the mara presence in Southern Mexico, while problematic, has not been associated with the levels of brutality seen in Central America. Thus, local conditions are an important filter for gang structures and practices.

Once established, the dominating presence of the maras created an intriguing 'translocal' geography in many cities in Central America, such that later, deportees or returnees would be immersed in a new geography of conflict based on their identity as a gang member in the US, 'rewriting the geography' of the Central American streets (Zilberg, 2004). The gang identity relocates, moving with them through space, across borders and over territories, as they become enmeshed in the implications of their identity in these new unfamiliar spaces.

There are pertinent issues, therefore, regarding attachment to place and the meaning of territory for maras. For Reguillo (2005:79), 'translocal migrants, mara members do not confine themselves to any particular territory because they were discarded long ago. It is this uprooting which is their main strength'. Certainly, the implication that gang members have a mobile identity is an interesting one, and one which will be explored at greater depth below: the very fact that maras have been formed according to transnational flows means that their identity is a unique combination of places and translocal influences. Yet, I do not think it wise to take this too far and detach maras from their locality, not least because maras now are heterogeneous amalgams of a whole range of different groupings, many (if not most) of whom have never even been to Los Angeles.

What is particularly interesting though in terms of the connection between identity, space and territory, in the case of the maras is that although clikas function according to territoriality, the final attachment is to the gang, and defense of territory is merely the vehicle through which to display and perform this identity, such that the barrio has taken on a symbolic rather than physical significance. This notion is borne out by the way territory is sometimes divided at a sub–barrio level. Certainly in the case of Guatemala, one neighbourhood may be host to various clikas of the MS13 and the Barrio 18, with each territory sometimes not stretching to more than a few blocks. The 'barrio' therefore is imagined, a social space synonymous with 'gang' in its meaning, which in this sense does dislocate the gang from the neighbourhood.

This ambiguous relationship with territory in turn impacts upon the relationship between the gang and the wider community, since the symbiotic relationship of mutual protection so often reported as a key characteristic of gangs (see for example Rodgers, 2006) becomes impossible. Rather than protecting the barrio, gangs create a violent disorder for community members. As one young woman aged 17 noted in an interview:

Just outside [my house] is the boundary of two gangs −the 18s and the 13s. When they come out to fight they start doing it in the street with machetes, and stones −they throw stones which hit the houses, and so people have to stay where they are. I can hear it all from my house, when they start to fight, to kill.8

It is notable first of all, that since that interview the weaponry of gangs has changed significantly to now consist largely of knives and firearms, and second that the imposition of gangs' geographies directly impinges upon the mobility of others. It is worth remembering that in any locality, there are as many geographies as there are inhabitants.

So while space remains a crucial source of power and control for the gangs, the spaces inhabited and created by gangs are malleable and fluid. As Cruz (2010) finds, the fact that gang identity in this case is not strictly restricted by territory, means that gang practices (including violence) are similarly borderless. This is not to say that identities are detached from space, but rather the gang identity remains attached to the individual as they move through the city. It could be argued, then, that their gang identity is not so much dislocated as it is relocated: it has meaning beyond the territory of their clika, in the different spaces of the city, and more widely in the region.9


'Transnational' syndicates or barrio gangs?

It has been established that the general functioning and identity of maras are rooted in transnational migration flows, but beyond their transnational roots, an important question is of course to what extent they are organized at this level. First, in an imagined sense, they have certainly formed a 'global' community. The Barrio 18 has had its own website, and a cursory search of YouTube indicates that members, affiliates and sympathizers of both gangs have created a virtual community which transcends national borders. So it is that through the creation of imagined communities, identity and belonging are constructed in a way that disrupts the connection between space and identity (Helvacioglu, 2000).

The extent to which the gangs are physically networked is more contested. Authors reporting a rise of complex and expansive criminal networks, for example journalists Lara Klahr, (2006) and Fernández and Ronquillo (2006), provide accounts which seem based on dubious and perhaps sensational empirical evidence (see Wolf, 2010).10 Indeed, as Rodgers et al. (2008:23) argue, 'in many ways, the federated nature of the maras is more of an imagined social morphology than a real phenomenon, based on the fact that the steady flows of deportees from the United States share a common language and reference points'. In other words, the fact that gang identity can be relocated, as noted above, has meant that 'mareros' are perceived as monolithic and omnipresent, such that it is assumed that they are the result of conscious and formal networking and organization, whereas it seems more likely that in reality, these connections where they do exist are loose and informal. As Santamaría (2007) finds, national and transnational links between gangs and gang members are neither formalized nor institutionalized, such that the maras operate rather as a disorganized criminal network.

Indeed, I have found that direct communications between the different neighbourhood clikas of gangs is rare, but it occasionally served two broad purposes: to facilitate migration, and to standardize or control gang practices and operations (particularly of new groups). This communication was said to take the form of coded letters sent with members travelling through Central America and Mexico, and perhaps maps of where other clikas are based who would then be obliged to receive and protect them. It seems likely that the majority of gang members do not have any links with clikas beyond their immediate locality, and that those links that do exist are temporary and sporadic, and do not have an explicitly organizational or expansionist purpose.

Public interest in the transnational organization of youth gangs is more specifically related to growing concern over the extent to which maras have become involved with organized criminal networks, particularly the drugs trade, yet there is little evidence to support the idea that the maras are a regionally organized criminal operation.11 While it is certainly the case that some gangs have aligned themselves to some extent with organized crime, these associations seem to be sporadic and patchy, and in cases where they do exist, they appear to constitute a strategy of survival, rather than expansion. In large part, these strategies have emerged in the face of anti–gang policies across the region, which have dramatically changed the context in which gangs operate, both within neighborhoods, and within the prisons where many thousands of gang members are held. The remainder of the paper focuses on the nature of this relationship between gangs and the state, and the effects of increasing state repression on gang activity and violence.


State–gang relations and the changing (spatial) strategies of maras

We have already seen in part how state interventions affect the formation of gangs in the case of US immigration policy. I have argued elsewhere that currently it is institutional factors that have perhaps the most significant effect on the functioning of gangs in Central America (see Winton, 2011). These institutional factors begin with recent (regional) legislation in Central American countries affected by maras. Understanding the significance and impacts of gangs (and state responses to them) can be enhanced by considering the differing degrees of vulnerability in state political systems in the face of the challenge that gangs present.12 A vulnerable state whose already weak institutions have at once allowed, and been further degraded by the activities of gangs and organized criminal groups, a state which can no longer monopolize the legitimate use of violence, a state crippled by corruption, is a state lacking the fundamentals necessary to comprehensively deal with the threat that gangs present. While the regional spread of gangs in Central America springs from common experiences of fragmentation, insecurity, vulnerability, the formal regional response (overwhelmingly punitive and repressive) is inspired by panic, stemming from state weakness.

As part of a regional strategy to tackle the maras, national governments in affected countries in Central America and subsequently in Mexico, implemented anti–gang legislation (known generally as 'mano dura', or 'iron fist') which in effect made it illegal to belong to a gang. There has been some national variation in the specifics of the legislation, but essentially it became possible for anyone with a tattoo to be imprisoned, regardless of whether or not they have committed a crime. Indeed, the cultural stereotyping of the 'undesirables' is remarkable: illicit gang association has been defined in El Salvador as 'groups of people who act to disrupt public order, or attack decorum and good principles, or who mark their body with scars and tattoos' (Proceso, 2003). This approach gained wide popular support, but has also received strong criticism both nationally and internationally, has had a range of pernicious effects (see below), and significantly, it has failed to reduce violence. While more recently there have been some moves to offset these punitive approaches through preventative and rehabilitative measures (driven in large part by foreign agencies including the UNDP), in reality there is little change on the ground in the way gangs are dealt with (Rodgers et al., 2008).

Ironically, if unsurprisingly, the more repressive regional gang policy becomes, the greater the regional spread of the maras (being more mobile now than ever), and the greater their involvement in other criminal activities for survival. Heavy handed policies in Central America have had the indirect effect of aligning gangs with organized crime to a greater extent than in the past, including low–level drug dealing, trafficking of migrants, extortion and contract killings. In addition, the very fact that the maras have recently linked with other crimes and have 'diversified', has meant they have ceased to be just a 'barrio' issue, and has contributed to their becoming something of a regional obsession (see Reguillo, 2005). Also important has been the US identifying these gangs as a matter of national security, which has raised the profile of the issue at a regional level. All of these factors speak to the scaling up, or regionalization of the problem. As a phenomenon, it becomes not only translocal (under which guise it operates across borders, but remains a fundamentally local problem, and thus more forgettable at a regional level), but really transnational in its threat (Ibid.).

At a local level, as a result of these policies there have been many reports of worsening police brutality towards suspected gang members throughout Central America, and frequent accusations of social cleansing and vigilantism. In one colonia in Guatemala City alone, it was reported that 85 young people were killed in this way in one year, often brutally.13 Such policies also further reduce state capacity to govern effectively, in obscuring other insidious forms of violence and crime.14 As one retired gang member in Honduras notes,

they say that it's only gang members who rob, who kill, who rape, but there are a lot of people doing these things who are nothing to do with gangs. They say they're all gang members. It's not like that. We're not like people to them −they blame us for everything.15

While it is important not to lose sight of the fact that fear of gangs is often grounded in real experience of crime and violence perpetrated by gang members, reliance on repressive policies is deeply counterproductive, not least since the rehabilitation of gang members becomes virtually impossible. Young people living as part of a gang generally rely entirely on that gang for economic and social survival, and if there is no viable alternative livelihood, there is little incentive to leave a gang. Moreover, rehabilitation means little when they are still very much at risk from persecution simply because of bearing the marks of the gang. In the words of a Barrio 18 leader, 'there are some who have a family, children and maybe work, but just for going around with tattoos they could be banged up, or even killed... it's a violation of human rights'.16

The logical, and often only option for gang members in such a situation is to reduce their visibility. Visibility has in the past been crucial to the performance of gang identity, but there is the sense now that their very essence is changing. As one representative of the Institute of Human Development (IDH) in Tapachula, Chiapas commented: 'now the kids go around well–dressed, they're just checking things out to see where things are calmer so they can carry on'17. Many gang members no longer use tattoos, no longer dress in the same way, or hang out in large groups; they have become an 'invisible' target.18 In addition, it is possible that this new way of operating is more expensive for gang members (requiring the acquisition of private spaces, vehicles, etc.), which could be linked in part to increased criminality on their part to sustain the organization (Cruz, 2010).

Repressive policies in Central America have also physically displaced gang members, as Zilberg (2004:777) sums up: 'efforts to reassert national sovereignty through zero tolerance policing strategies only, and most ironically, reproduce transnational flows and formations'. As a spokesman for the Secretary of Security in Honduras enthuses: 'in the face of our offensive, thousands of gang members have fled to other Central American countries, and to the US and Mexico'.19 The most significant of these movements has been recent 'incursion' of gangs into southern Mexico, where illegal border crossing and trafficking is said to have provided them with plentiful opportunities to survive. Police crackdowns in the border region of Mexico seem to have repeated this pattern of regional displacement, with gang members drifting north into other cities such as Oaxaca, and according to some accounts, even further north to the US.

If, as noted above, gang themselves slot into an absence of legitimacy and hegemony, (challenging legality in terms of confronting an absence rather than a presence) an authoritarian response, which is really an attempt to 'fill the absence of legitimacy with a double dose of legality' (Reguillo, 2005:77) is ultimately ineffectual. First, as Reguillo (2005:80) notes, 'when death, instability, uncertainty, hopelessness and detachment become rooted as everyday experiences, punishment by example is irrelevant', and second, prisons become an extension of gang life, a normalized experience (Hagedorn, 2007b).

In fact, it has been argued that the most significant sites of organization and operations for gangs in contemporary Central America are its prisons. As Cruz (2010:392) finds, prisons have created alternative spaces of organization for gangs, in which gang members from different areas of the country come together, and are able to become a 'sort of standing assembly where they could debate, make pacts and decide on structures, strategies and ways to operate'

Rather than rehabilitation, (increasingly overcrowded) prison may often reinforce gang allegiances, and strengthen links between different clikas of the same gangs, and between these and criminal groups. Mass incarceration serves not to dismantle gangs, but rather to reinforce and transform the way in which they operate: it becomes a space crucial for gang survival, a site away from the territory of the barrio, but which is nonetheless intricately tied to its spaces. Although it is certainly true that the extent to which gangs are organized in this way in prisons varies significantly,20 it certainly marks a new and important development in the way in which gangs organize, with prisons becoming a key space to which gang identities and activities have been relocated.


Concluding remarks

In this paper, we have seen how the global and local interact both in terms of constructing the context in which maras have come to operate, and also in terms of how gangs themselves create and transform the spaces in which they operate. It is of key importance to recognize, therefore, that transnational processes are produced and enacted at a local level. The fluidity and complexity of gang identities in the case of the maras is due in large part to their dynamic spatiality, emerging as they have from transnational processes, being enacted translocally, and possessing a fluid and dynamic territoriality, resulting in gang identities that are flexible, mutable, and mobile. In such a scenario, it becomes possible for gangs to transform rapidly, while maintaining their unity and identity, making them particularly difficult for the authorities to contain and control.

To ignore the importance of these geographies of gangs is to misunderstand the way they operate, to underestimate their capacity to transform and survive in the face of harsh repression. We know that poor understanding of gangs inevitably leads to ineffective interventions. It seems that there is still much work to be done to enrich our knowledge of gangs throughout the region.



I would like to thank Charles Wood and Cathy McIlwaine for comments on an earlier draft of this paper, and to the two anonymous reviewers for their helpful observations.



1This article is the product of reflections based on a period of research on youth marginalization, gangs and violence in the region spanning over 10 years, particularly in Guatemala, and later Mexico, but also in Honduras and to a lesser extent El Salvador. This research has comprised qualitative research with young people in a range of marginal communities, and also in different institutional contexts (including NGOs, schools, and prisons), through interviews, group discussions, photography projects and observation in a range of contexts.

2 This notion borrows to some degree from social disorganization theory, which as mentioned above has itself played a significant (if contentious) role historically in gang studies, particularly in the US.

3 Interview in detention centre in Tapachula, Chiapas, CERESO 4, December 2006.

4 I recognise that the implied uniformity of the term 'gang' and 'gang member' is highly problematic, due to the fluidity of young people's identities in relation to gang affiliation. In reality, gangs are hugely diverse structures, and young people's relations with them are complex and time–space dependent.

5 It should be clarified that the term mara was used by gangs in Central America prior to the development of the 'transnational' gangs which it later came to represent. It is now a term used popularly throughout Central America to refer to youth gangs, but interestingly it is rarely used by gangs themselves, who more commonly refer to themselves as 'pandillas', the generic name for gangs. Confusingly, banda, pandilla and mara, far from being interchangeable labels, have very specific connotations in different countries.

6 It has been argued that there is some similarity between this pattern and the Mexican Cholos, first established in Los Angeles during the 1940s among children of Mexican immigrants, also in the face of discrimination and vicitimization. Both are directly related to (disadvantageous) migration, and the gangs themselves share similar broad traits in terms of the norms that shape their identity, and the codes of conduct that rule the way they behave (see Nateras Dominguez, 2007). Yet it must be noted that there are very significant differences between them, due largely to vastly different local conditions in Central America.

7 Interestingly, the maras have not spread to Nicaragua, for which a partial explanation is that young Nicaraguans historically have not migrated to the US to the same extent as their Central American neighbours, largely since it is more convenient to travel to neighbouring Costa Rica, where in 2001, 53 percent of emigrants from Nicaragua arrived (Rocha, 2003). Where they have migrated to the US, this has historically tended to be to Miami, which has a very different gang landscape to that in Los Angeles.

8 Interview in Guatemala City, January 2001.

9 Another important space which is strongly imbued with these relocated gang identities is the prison (see Cruz, 2010; and discussion below).

10 It is notoriously difficult to obtain accurate information about the organization of gangs, since this is the information guarded most closely by gangs themselves.

11 Perhaps more important in terms of criminality than whether or not they are key actors in the drug trade, is the more locally pernicious development of protection rackets, which have become an increasingly common strategy among maras in recent years (see Cruz, 2010).

12 Thanks to Charles Wood for this insight.

13 Interview in Guatemala City, July 2007.

14 In fact, gang members are much more likely to be killed than to kill. Of young people under the age of 30 murdered between 1998 and 2001, 34 percent were gang members (YCARE, 2007).

15 Interview in San Pedro Sula, Honduras, August 2004.

16 Interview in San Pedro Sula, Honduras, August 2004.

17 Interview in Tapachula, Mexico, November 2006.

18 Tattoo removal has become common in Central America, although many young people have to resort to rudimentary and brutal techniques where removal is not available or is expensive.

19 El Universal, 05/07/05.

20 As one young female gang member noted in an interview in a prison in Chiapas, the conflicts of outside do not always crossover into the prison: "it's a barrio thing". Interview in CERESO 3, Tapachula, Mexico, December 2006.



Appadurai, A. (1996), Modernity at Large. Cultural Dimensions of Globalization, University of Minnesota Press, Minneapolis/London.

AVANCSO (1996) (3rd ed.), Por Sí Mismos: un Estudio Preliminar de las 'maras' en la Ciudad de Guatemala, Cuaderno de Investigación No.4, AVANCSO, Guatemala.

Briceño León, R. and V. Zubillaga (2002), 'Violence and globalization in Latin America', Current Sociology, vol. 50, no. 1, pp. 19–37.

Castells, M. (1997), The power of identity, the information age: economy, society and culture, vol. II, Blackwell, Cambridge, MA/Oxford, UK.

Castillo Berthier, H and G. A. Jones (2009), 'Mean streets: gangs, violence and daily life in Mexico City', in Jones, G. A. and D. Rodgers (eds.), Youth Violence in Latin America: Gangs and Juvenile Justice in Perspective, Palgrave–Macmillan, New York.

Cruz, J. M. (2010), 'Central American maras: from youth street gangs to transnational protection rackets', Global Crime, vol. 11, no. 4, pp. 379–398.

Cruz, J. M. and N. Portillo (1998), Solidaridad y violencia en las Pandillas del Gran Salvador: más allá de la vida loca, UCA Editores, San Salvador.

Davids, T. (forthcoming), 'Urban Youth, Poverty and Gang Violence in San Salvador: ambiguities of security and violence between the local and the global', in Suárez de Garay, M. E. and J. C. G. Aguiar (coords.), La Seguridad entre lo Local y lo Global: los entornos latinoamericanos, ITESO/CEDLA, Mexico/Amsterdam.

ERIC, IDESO, IDIES & IUDOP (2004), Maras y Pandillas en Centroamérica, UCA Editores, San Salvador.

Fernández Menéndez, J. and V. Ronquillo (2006), De los maras a los Zetas: los secretos del narcotráfico, de Colombia a Chicago, Random House Mondadori, Mexico.

Hagedorn, J. M. (2007a), 'Introduction: Globalization, Gangs and Traditional Criminology', Hagedorn, J. M. (ed.), Gangs in the Global City, University of Illinois Press, Ubana/Chicago.

Hagedorn, J. M. (2007b), 'Gangs, Institutions, Race and Space: the Chicago School Revisited', in Hagedorn, J. M. (ed.), Gangs in the Global City, University of Illinois Press, Ubana and Chicago, pp. 13–33.

Hagedorn, J. M. (2007c), 'Gangs in Late Modernity', in Hagedorn, J. M. (ed.), Gangs in the Global City, University of Illinois Press, Ubana & Chicago, pp. 295–317.

Helvacioglu, B. (2000), "Globalization in the Neighbourhood", International Sociology, vol. 15, no. 2, pp. 326–342.

Hume, M. (2007), "Mano Dura: El Salvador responds to gangs", Development in Practice, vol. 17, no. 6, pp. 739–751.

Kirby, P. (2004), Is globalisation good for Us? Introducing the concept of vulnerability, Working Papers in International Studies No. 4, DCU Centre for International Studies, Dublin City University, Dublin.

Lara Klahr, M. (2006), Hoy te Toca la Muerte: el imperio de las maras visto desde dentro, Planeta, México.

Liebel, M. (2004), "Pandillas juveniles en Centroamérica o la difícil búsqueda de justicia en una sociedad violenta", Descactos, no. 14, pp. 85–104.

Massey, D. (1993), "Power–geometry and a progressive sense of place", in Bird, J., B. Curtis, T. Putnam and G. Robertson (eds.), Mapping the futures: Local cultures, global change, Routledge, London/New York, pp. 59–69.

Maxson, C. L. (1998), Gang Members on the Move, Juvenile Justice Bulletin, U.S. Department of Justice, OJJDP, Washington, D.C.

Merlo, R. and E. Milanese (2000), "La Construcción Social de la Juventud: desde la prevención de la exclusión social", in Merlo, R. and E. Milenese (coordinators), Miradas en la ciudad: métodos de intervención juvenil comunitaria, IMJ, México.

Moore, J. W. (1998), "Understanding youth street gangs: economic restructuring and the urban underclass", in Watts, M. W. (ed.), Cross–cultural Perspectives on Youth and Violence, JAI Press, Stamford, CT.

Nateras Domínguez, A. (2007), "Adscripciones juveniles y violencias transnacionales: Cholos y maras", in Valenzuela Arce, J. M., A. Nateras Domínguez y R. Reguillo Cruz (coords.), Las maras: identidades juveniles al límite, UNAM/El Colegio de la Frontera Norte/Casa Juan Pablos, México, pp. 127–155.

Proceso (2003), 'Plan "Mano Dura": Violencia Estatal Contra las maras', Revista Envio, 258 [http://www.envio.org.ni/articulo/1283].

Reguillo, R. (2005), "La mara : contingencia y afiliación con el exceso", Nueva Sociedad 200, pp. 70–84 [http://www.nuso.org/upload/articulos/3297_1.pdf ].

Rocha, J. L. (2000), "Pandillas: una cárcel cultural", Revista Envío, 219 [http://www.envio.org.ni/articulo/ 1012].

Rocha, J. L. (2003), "Why do they go? Theories on the Migration Trend", Revista Envio, 264 [http://www.envio.org.ni/articulo/2105].

Rocha, J. L. (2006), "Mareros y pandilleros: ¿nuevos insurgentes, criminales?", Revista Envío, 293 [http://www.envio.org.ni/articulo/3337].

Robertson, R (1995), "Glocalization: time–space and homogeneity–heterogeneity", in Featherstone, M., S. Lash and R. Robertson (eds.), Global Modernities, Sage, London, pp. 25–44.

Rodgers, D. (2003), "Youth gangs in Colombia and Nicaragua: new forms of violence, New Theoretical Directions?", in Rudqvist, A. (ed.), Breeding Inequality – Reaping Violence Exploring Linkages and Causality in Colombia and Beyond, Outlook on Development Series, Collegium for Development Studies, Uppsala, pp. 111–141.

Rodgers, D. (2006), "Living in the shadow of death: gangs, violence and social order in urban Nicaragua 1996–2002", Journal of Latin American Studies, vol. 38, no. 2, pp. 267–292.

Rodgers, D. (2007), Slum Wars of the 21st Century: the New Geography of Conflict in Central America, Working Paper no. 10, Crisis States Research Centre, LSE, London.

Rodgers, D., R. Muggah and C. Stevenson (2008), Gangs of Central America: causes, costs, and interventions, Occasional Paper of the Small Arms Survey, GIIDS, Geneva.

Santacruz–Giralt, M., J. M. Cruz and A. Concha Eastman (2002), Barrio adentro: la solidaridad violenta de las pandillas, OPS, Washington, D. C.

Santamaría Balmaceda, G. (2007), "maras y pandillas: límites de su transnacionalidad", Revista Mexicana de Política Exterior, núm. 81, pp. 101–123.

Sassen, S. (ed.; 2007), Deciphering the global: its scales, spaces and subjects, Routledge, New York.

Savenije, W. (2004a), "La mara Salvatrucha y el Barrio 18 St.: Fenómenos sociales transnacionales, respuestas represivas nacionales", Foreign Affairs: Español, vol. 4, no. 2, pp. 38–46.

Savenije, W. (2004b), "Las pandillas transnacionales mara Salvatrucha y Barrio 18 St.: una tensa combinación de exclusión, delincuencia y respuestas represivas", in Lesser, T., B. Fernández, L. Cowie and N. Bruni (eds.), Intra–Caribbean Migration and the Conflict Nexus, Human Rights Internet/International Organization for Migration/Association of Caribbean States/The University of the West Indies, Ottowa, pp. 205–228.

Smutt, M. and J. L. E. Miranda (1998), El fenómeno de las pandillas en El Salvador, UNICEF/FLACSO, San Salvador.

Thale, G. (2006), "Las pandillas juveniles centroamericanas y las respuestas de mano dura: sus efectos negativos en las reforma policial y el Estado de derecho", Quórum, no. 16, pp. 51–61.

Thrasher, F. (1927), The Gang, University of Chicago Press, Chicago.

Valenzuela Arce, J. M., A. Nateras Domínguez and R. Reguillo Cruz (coords.; 2007), Las maras: identidades juveniles al límite, UNAM/El Colegio de la Frontera Norte/Casa Juan Pablos, México.

Vigil, J. D. (2006), "A multiple marginality framework of gangs", in Egley, A. Jr, C. L. Maxson, J. Miller and M. W. Klein (eds.), The Modern Gang Reader, 3a. ed., Roxbury, Los Angeles, pp. 20–29.

Winton, A. (2004), "Young people's views on how to tackle gang violence in 'post–conflict' Guatemala", Environment and Urbanisation, vol. 16, no. 2, pp. 83–99.

Winton, A. (2011), "La violencia en Centroamérica: una historia de instituciones violentas", DESACATOS, núm. 37, pp. 111–124.

Wolf, S. (2010), "maras Transnacionales: origins and transformations of Central American Street Gangs", Latin American Research Review, vol. 45, no. 1, pp. 256–265.

YCARE (2007), Youth Justice in Action: Campaign Country Report: Honduras 2007, YCARE International, London.

Zilberg, E. (2004), "Fools Banished from the Kingdom: remapping geographies of gang violence between the Americas (Los Angeles and San Salvador)", American Quarterly, vol. 56, no. 3, pp. 759–779.

Herramientas del artículo

Envíe este artículo por correo electrónico (Inicie sesión)
Enviar un correo electrónico al autor/a (Inicie sesión)

Estadísticas del artículo

vistas descargas
anual histórico
vistas 9 752
descargas 3 768