Youth, gangs and violence: Analysing the social and spatial mobility of young people in Guatemala City

Abstract

Youth gangs have become an increasingly significant, and controversial, social institution in low-income communities in many cities in Central America, and yet the local-level impacts of this phenomenon, especially on young people, remain under-explored. Drawing on research with young people in Guatemala City, this paper explores the multiple barriers to the social and spatial mobility of both gang and non-gang members resulting from gang violence. It also examines how young men and women cope with violence, and, given the severe impacts of gang activity on young people, highlights the pressing need for social alternatives to gangs.

Keywords:

• Youth
• violence
• social and spatial mobility
• Guatemala

Introduction

Almost everyone has their weapons, and when you go out into the street you never know if you're going to return, what's going to happen to you, when they're going to assault you. You don't have security—we don't live in peace […] There have only been deaths.

(Maria, aged 20, Guatemala City)

Despite the seemingly interminable academic and non-academic concern over ‘youth violence’, rarely does it stretch to consider the effects of violence on young people, and the extent to which they themselves are victimised. It is certainly true that in Guatemala, as noted in the context of the UK by Pain (2003, p. 152), ‘the issue of young people as victims is submerged by the increasing criminalisation of youth’. In analysing the impact of youth gang activity on young people in Guatemala City, this paper attempts to give some insight into a phenomenon in which young people are at once the perpetrators and principal victims of community violence, and to go some way towards revealing the nature of everyday risk for young people in this context. In particular, although youth gangs have become an increasingly significant (if in some ways negative) social institution in low-income communities in urban Guatemala, and throughout much of Central America, little research has examined the ramifications of gang activity for the communities in which they operate (but see Rodgers, 2003; Moser and McIlwaine, 2004). In essence, this paper attempts both to broaden the existing literature on youth gangs in Latin America through considering the changes in social landscapes brought about by this phenomenon, while also contributing to the emerging body of geographical literature on the socio-spatial construction of fear of (violent) crime among young people.

The now substantial literature on the geographies of childhood and youth has established not only that there certainly is a place for children and youth in geography (cf. James, 1990), but that ‘research in children's geographies is now beginning to reach critical mass’ (Holloway and Valentine, 2000; see also Skelton and Valentine, 1998; Matthews and Limb, 1999; Aitken, 2001). The central tenets of this stream of research need not be rehearsed here, but particularly welcome for the purposes of this paper have been the problematisation of ‘moral panics’ surrounding youth (for example Lucas, 1998), and the broader aim of deepening understandings of the various ‘lifeworlds’ of children and young people in both the North (see for example, Katz, 1993; Valentine, 1996; Pain, 2001, 2003; Nayak, 2003) and increasingly, the South (for example Katz, 1993; Beazley, 2000; Punch, 2000; Young, 2003). More specifically, there has been increasing attention paid to the particular experiences and responses of young people in relation to fear of crime1 in the North (see Pain, 2001, 2003; also Nayak, 2003), but this is yet to be accompanied by similar work in the South (but see Bartlett et al., 1999). Especially important, and again largely neglected, is the fact that community violence carried out by young people has important repercussions for young people not involved in its perpetration. Yet, the relationships between and within these two apparently distinct groups of young people also need to be explored.

Furthermore, the face of ‘youth’ in Guatemala is portrayed as distinctly male and disaffected, with little regard for the lives and experiences of young women, and indeed the diverse experiences of young men, in such an environment. In addition, existing work on gangs in Latin America has scarcely considered the spatial dynamics of violence beyond the construction of identity through gang territories (for example Riaño-Alcalá, 1991; Rocha, 2000b). There are, therefore, both connections and barriers among young people, and the wider community, to be considered. It is useful to consider youth violence in terms of exclusion, since different layers of (at times conflicting) socio-spatial exclusions and inclusions were found to be integral not only to the functioning of youth gangs themselves, but also to their wider effects (see also Pain, 2001). In turn, while it has been suggested that wider processes of social exclusion can themselves be short-circuited at a local level (Gore et al., 1995), it is also important to recognise that local phenomena (with, of course, their own macro-level influences) may also produce acute exclusions at this level.

Methodology2

The paper is based on research which focused on the well-being of young people living in two low-income, marginal settlements in Guatemala City, both of which were established in the mid-1980s: El Progreso, a settlement situated to the south west of Guatemala City, and Ciudad Sueños, a more geographically isolated settlement to the west of the city.3 Marginal areas of the city such as these are populated by close to one million people, where there is a profound lack of basic services, only limited educational and health services, few formal sector employment opportunities, and very high levels of delinquency and social violence (Poitevín et al., 2000). Research was carried out with young people in these communities between October 2000 and June 2001, using a broadly qualitative approach. A total of 250 young people participated in the research (127 women, 123 men), aged between 9 and 23, with the majority in their mid-teens. They were accessed both through local schools and youth organisations, and subsequently via snowballing. Research was carried out mainly in these schools and organisations, and occasionally in participants' homes. Some discussion groups were held in the street, although this was generally not possible for safety reasons. Since the research was not explicitly focused on either violence or youth gangs, sampling was not determined by gang membership, although a small number of participants were themselves ex-gang members, and others had friends currently involved in gangs. It is possible that some participants were involved with gangs, but chose not to divulge this. More specific research which had been planned with active gang members had to be abandoned owing to events towards the end of the research period.

It is widely recognised that providing a range of alternative modes of expression is particularly important in research with young people (see also Morrow and Richards, 1996; Johnson et al., 1998; Thomas and O'Kane, 1998; Ansell, 2001; Young and Barrett, 2001). Similarly, the recent ‘participation revolution’ in development—aiming to prioritise local knowledge and promote genuine participation in the research process—has increased the use of ‘alternative’ methods in this context, albeit not always appropriately (see below). With this trend comes an increasing recognition of the benefits of combining these methodological approaches in research with youth in the South (see for example Beazley, 2000; Young and Barrett, 2001). Underlying this paper, therefore, is the understanding of young people as different to adults, yet not inferior; diverse amongst themselves, but marginalised as a group; enmeshed in wider socio-cultural structures, yet possessing their own understandings of the world; and above all as eminently capable. In particular, it is argued that incorporating Participatory Appraisal techniques (see Chambers, 1994a, 1994b, 1994c) increased the flexibility and versatility of the framework, and it is especially well-suited to work with young people. One significant advantage of using participatory tools is that they allow the focus of the research to be immediately and visibly shaped by the priorities of participants (see below).

However, given evident problems with the overblown ‘empowerment’ claims of Participatory Appraisal (see for example Crawley, 1998; Guijt and Shah, 1998; Cornwall, 2000; Pain and Francis, 2003), I find it more constructive to use the method without the label, since genuinely ‘participatory’ research involves more than mere innovation at the data collection stage. In addition, I do not see the key principles of PA4 (respect of local knowledges, the respectful behaviour and attitudes of the outsider, accessibility and adaptability) as exclusive to this approach, nor as constituting ‘participation’. Unless it is the basis of direct action, resulting from sustained ‘community’ participation (which is itself a rather problematic idea), PA cannot claim to be anything but a preferable alternative, or complement, to more traditional methods. Surely, however, there is a self-evident difference between ‘participatory appraisal’, and the more action-based ‘participatory learning and action’, or ‘participatory action research’. Users of these different approaches need to be explicit about the aims and purpose of their methodological approach. In the current context, expectations of this approach are limited to it being an inclusive, appropriate and flexible method of data collection. Perhaps the best way to avoid confusion over the meaning of ‘participation’ would be to avoid using it as an explicit label unless the entire research process is genuinely participatory. I would also suggest that it is best used to inform more in-depth research, rather than to replace it altogether.

The research employed a primarily qualitative methodological framework, which included 61 diagram-based discussion groups, in which various (PA) tools, such as causal flow diagrams, maps, pie charts, timelines, and drawings, were used by young people to explore and discuss problems and ideas. The role of the researcher in this context is as facilitator/participant. Discussion groups were conducted with small groups of young people, averaging 4, to first, gain some understanding of the broad range of issues affecting their lives and the community more generally, and second, to explore the emerging issues as prioritised by participants. Where possible, the groups comprised those of similar ages, although this was not always feasible. Most groups contained both young women and men, but some were single sex. It was also used with some individuals as an alternative to interviews, when they expressed this preference.

In addition, 20 young people were given cameras—either individually or in small groups—and asked to take photographs of the positive and negative features of the community, and of the places and people which were important to them. They were later asked to comment on the photos they had taken to explain why they had taken the picture, and were all given copies of their photographs. The pictures, comments and associated discussions provided valuable insights into the daily lives and experiences of young people, which usefully complemented those gained through other methods (see also Morrow and Richards, 1996; Bolton et al., 2001; Young and Barrett, 2001).

More in-depth qualitative information was obtained through 39 in-depth interviews with young women and men aged between 14 and 21, the majority of whom had already participated in one or more of the discussion groups. In addition, a short questionnaire survey was completed after the discussion groups to elicit basic information about the participants, together with 47 daily routine surveys. This framework allowed young people some choice over how they participated, and enabled data triangulation.

Living with everyday violence: The research context

Following the official end of the 36-year armed conflict in 1996, violence in Guatemala has transformed from political, to more widespread social and economic violence. As such, the ‘norm’ of violence has been transformed and maintained, while at the same time the media continues to sensationalise this violence (De Orrellana, 1997; Smutt and Miranda, 1998; Torres-Rivas, 1999; Garbarino et al., 2002). The upsurge in social (and often economic) violence is commonly blamed on the proliferation of youth gangs, or ‘maras’, which have become an important concern in Guatemala, particularly in urban areas, and throughout the region (see, for example, AVANCSO, 1996; Aboutanos, 1997; Cruz, 1997; De Cesare, 1997; FLACSO, n.d.; Ramos, n.d.; Smutt and Miranda, 1998; Rodgers, 1999a). At a local level, therefore, gang violence is at once both normal and scandalous, which has significant psychological implications for the communities affected by it. As Clauss-Ehlers and Lopez Levi (2002) suggest in the context of the US, young people may come to expect violence in order to protect themselves from it; regularly experiencing or hearing about violence is traumatising, and yet leads to desensitisation. As a result, ‘community violence among youth becomes viewed as something routine rather than an unacceptable exception’ (ibid., p. 270). It has also been argued that young people who are faced with seemingly uncontrollable violence may develop a sense of fatalism and short-termism, through which violence is further perpetuated (see NNFR, n.d.). In the case of Guatemala, also significant is the potential loss of hope as a result of the continuation—indeed, in many ways worsening—of violence since the end of the armed conflict (see Moser and McIlwaine, 2004). Gibson et al.'s (1991)5 findings from Alexandra, South Africa are in many ways comparable:

the significance of hope for allowing people to cope with the most adverse of circumstances cannot be underestimated. It represents the capacity to hold on to the possibility of something good even in the face of overwhelming badness. In Alexandra, the years of continuous violence may have undermined this usually very resilient capacity which has also been injured by the abrupt raising of expectations of peace, followed by the increase in violence.

Indeed, one 17-year-old young woman in Guatemala City noted that: ‘There will be more people involved in gangs, more deaths. This is going to increase here. There's not much hope’. Thus, the psychological effects of everyday violence, and the often associated lack of hope affect the experiences and perceptions of young people, and influence the choices that are available to them in their daily lives, all of which further complicates their roles as both victims and perpetrators of violence.6

The nature of youth gangs in Guatemala

There is little consensus, or reliable data, on the proportion of crime actually attributable to youth gangs in Latin America, but perhaps more significant than actual rates is the fact that that the problem is widely perceived and portrayed as one of the most serious facing Guatemala today, and certainly has a profound impact on many communities. Indeed, of all problems noted by young people participating in the current research, the problem of the maras and the associated violence and insecurity was noted most frequently, and the majority of participants also considered this to be the most severe of the many problems confronting them (see also Moser and McIlwaine, 2004). It is important to note, however, that not all violent crime can be attributed to youth gangs, and that gangs themselves are not uniformly criminal (see Moser and Winton, 2002). In many ways youth gangs are the manifestation of young people's need to feel part of a group in situations of multiple exclusion and the absence of alternatives, yet this seemingly innocuous need can have grave consequences both for themselves and others. Participants identified a range of different types of activities which gang members were involved in, many of which were violence-related (see Table I). It is both the economic crime gangs often engage in, and the social violence of territorial conflicts which have a strong impact on the spatiality of communities.

Table 1.  List of mara activities and outcomes reported by participants

Mara activities	Outcomes
Fight each other	Fear
Kill	Insecurity
Hurt people	You can't go out when they fight
Sexual abuse, rape	Death
Have tattoos	Lack of communication
Shoot	Enemies with parents
Take drugs	Family breakdown
Vandalise	Lack of trust
Violence in school
Rivalry
Violence in the community
Rop people and houses
Drug dealing
Source: discussion groups

A predominantly (although certainly not exclusively) male phenomenon, maras are generally associated with low-income urban areas, due to high levels of inequality and the lack of social, economic and cultural alternatives available in this context. Data on the numbers of gangs or maras and their members can only be approximate. Although the maras ‘18’ and ‘Salvatrucha’ (or MS 13) operate throughout Central America, the areas where gangs operate will usually have a number of smaller neighbourhood gangs, which may or may not be part of one of the larger gangs. The Prensa Libre (cited in UNDP, 1998) identified a total of 53 different maras operating in 12 different zones of Guatemala City in 1997, although this is a conservative figure. A different source put the figure for Guatemala City at 330 in 1995 (PRODEN, 1996). Moreover, it has almost certainly increased significantly in subsequent years. While gang activity is most prolific in marginal urban areas of Guatemala City, gangs also operate in 10 departments outside the capital (Rodríguez and de León, 2000; see also Moser and McIlwaine, 2001).

Although not always used, the widespread possession and use of firearms (not least as a consequence of the recent armed conflict) heightens the fatality of territorial conflict between gangs, and also increases the risk of injury or death for those not involved in the disputes. Indeed, the most common cause of death among 15–24-year-old Guatemalans in 1997 was firearms (Poitevín et al., 2000). Far from being gratuitous, however, territorial violence is key to maintaining the identity of individual gangs. Group identity is based on separation both from other gangs and the rest of the community, with unity rooted in difference from others, yet commonality amongst the group. It is on the basis of this strong group identity, and the (often intimidating) image this projects, that gangs exercise power, with identity rooted in space as symbolic of cultural difference.

The upside of the structure of gangs for members themselves is the large amount of social capital they inhere, in terms of the mutual trust and solidarity associated with the social networks in place. ‘Belonging’ to a gang offers support in practical and emotional terms, and is an attractive remedy to a lack of these resources in other spheres, especially the home (which is itself the result of wider social and economic problems) (see also FLACSO, n.d.; Hearn, 1994; SELA, 1995; Rocha, 2000b; Moser and McIlwaine, 2001; Ramos n.d.;). As one interviewee, Verónica,7 aged 17, suggested,

if your parents don't support you, what are you going to do? If you find a, a friend who is … a drug addict, or who is involved in bad things, you go around with them too—you feel that you have their trust, and not your parents.

There is, however, a conflict between attaining emotional (and perhaps economic) well-being on one hand, and facing constant physical danger on the other. Young people involved with gangs were met with the threat of injury or death on a daily basis, with the clustering of territories in small areas placing further limits on their personal freedom. Indeed, one young woman, aged 16, spoke of serious problems of rivalry in her block alone. It was not safe, she said, for the gang members there to go to work, or to do anything with a set routine, as those from the other gangs would know when and where to get you. To belong to a gang, therefore, can mean surrendering a ‘normal’ life for one which is largely contained within the structure of the gang. This highlights the often neglected fact that young people involved in the perpetration of violence are also victims on many levels. As Pain (2003) argues, offending and victimisation are both connected in the lives of marginalised youth, both in terms of being the direct victims of crime (and/or violence), and also in a more abstract sense of exclusion and lack of appropriate opportunities or care. Nowhere is this more visible than in the case of youth gangs.

Fear, spatial mobility and social networks

If gang members are themselves subject to fairly strict social and spatial boundaries, how does gang activity impact on the socio-spatial freedom of young people not directly involved? Broadly, social contact beyond the home is contingent on spatial mobility, and use of space may thus expose the temporal and spatial extent of social networks. Overall, young people appeared to spend alarmingly little time outside school or the home. The daily routine survey, completed by 47 young people aged between 11 and 17, revealed that an average of only 0.4 hours per day was spent outside the home either with friends or playing sport, with very little variation according to gender. In addition, only 17 of the 47 respondents reported spending any time away from the home socially, and even surveys referring to the weekend showed an average of only 0.5 hours spent on this type of activity per day. The majority, then, spent their free-time at home, either spending time with their family, or in solitary activities such as reading, drawing, or watching television, which was also reflected in interviews.

The reason for participants' general aversion to public life in both communities was, perhaps unsurprisingly found to be the danger associated with general delinquency, and with the maras in terms of assaults and territorial fighting. In particular, the (at times) violent behaviour of maras has major repercussions for the use of space of other young people in the communities: the omnipresence of gangs leads to their perceived or actual dominance over public space, including recreational areas (such as basketball courts, football pitches and children's play areas), and the construction of this space as dangerous for the rest of the community. Particularly significant is that—contrary to findings elsewhere (Rodgers, 1999b, on Nicaragua; Smutt and Miranda, 1998, on El Salvador)—participants did not see gangs as protectors of their neighbourhood, but rather defenders of their own gang. This was itself related to the fact that the communities were said to contain a number of smaller gangs, rather than a strictly identifiable ‘neighbourhood’ gang.8

Accordingly, a striking aspect of the community maps drawn by participants, and also the photographs taken by them, was the danger associated with recreational areas. All the maps identified these sites as dangerous at night, and in only a quarter of cases were they shown as safe during the day (see for example, Figure 1). As one young man, aged 15 commented, ‘they arrive to play ball, and end up being killed’, while two young women, aged 14 and 15, drew someone being shot in the ‘children's park’ in their map. If even the supposedly recreational areas are perceived as dangerous by the very people they are supposed to serve, then it stands to reason that violence was consistently identified as the main factor influencing young people's spatial mobility. In addition to these specific sites, in many cases danger was seen to be more generalised, and associated with ‘the street’ as a whole (see Figure 1). This was related to the relatively small area covered by gang territories, which made violence less predictable and hence, less avoidable. Given the danger often associated with public space, it is perhaps hardly surprising that so many appeared to choose to stay at home (see also Kaztman, 1997, on Uruguay; Moser and McIlwaine, 2004, on Guatemala and Colombia; see below on coping strategies).

Figure 1. Local map of safety and danger, Guatemala City: drawn by two young women and one young man, aged 14–15.

There was therefore a lack of safe places to meet, compounding general fear of public space both by young people themselves and their parents. As noted above, the appropriation of public space is vital for gangs to attain some degree of spatial and cultural domination within the community, which has the (not unintentional) effect of alienating other young people and community members. As Carlos, aged 21, commented,

these young people take a space to go, to annoy and get drunk, to smoke marijuana as some of them who go there do, to have sex with the girls from their gang, and all that—they have a space which is supposedly for recreation, but it's not. They're always there.

Furthermore, some participants spoke of the limits placed on their freedom to move beyond as well as within the communities. With armed attacks on buses an all too common occurrence, some young people like Esteban, aged 16, felt limited as to where they could go: ‘you can't go up there [on the bus], because maybe if you're in a bus, they get on to rob you, or if not start to assault people’. Thus, connections beyond the community are also inhibited by fear of violence. Overall, therefore, violence had an unsurprisingly negative effect on the development and maintenance of some young people's social networks, restricting the extent to which they participated in social activities outside the home (see also Smith, 1987; Moser and McIlwaine, 2004). While the effects of gang violence were certainly not limited to young people, it will be seen that the social immediacy of young people involved in gangs bore specific implications for young people.

Paradoxically, it is perhaps the gang members themselves who had the strongest sense of ‘place’, albeit highly specific, and yet who were also subject to significant levels of community level exclusion as a result. This exclusion operated both in terms of the territorial restrictions resulting from gang rivalry (see below), and also their social separation from the rest of the community.

While it may seem remarkable that gangs have been able to exert such levels of control and influence over the local population, the police offered little resistance to gang activity. As Berta, aged 14, noted: ‘now instead of the groups of maras being afraid of the police, the police are afraid of them’. Indeed, the police were widely perceived as both inefficient and untrustworthy, and were (together with maras) frequently identified in discussion groups as the most negative institution in both communities (see also Moser and McIlwaine, 2004). This lack of efficacy was connected first, to a severe lack of resources, and second, to the endemic corruption and weakness of state institutions (see Winton, 2004a). Community resistance, while found to exist in other areas of the city, was not said to operate cohesively in either of the research communities.9

Gender, fear and violence

Emerging from the research were interesting gender differences in the type of fear and violence experienced by young women and men. While there was no discernible difference between the levels of violence experienced or feared by young men and women, the nature of fear and violence did vary according to gender, as did the spatial and temporal occurrence of this violence.

Together with the danger associated with getting caught up in gang conflict, young people also reported being the specific target of verbal and physical abuse. Young men in particular often noted how they would be threatened or hit by some gang members. Although young men were reported to be vulnerable to physical (and, less often, sexual) abuse in the home, and to male rape outside the home, they were more likely to discuss violence in terms of public violence. Indeed, it appeared that young men could less easily evade the violence of the streets (see Goodey, 1997, on the UK; see below on coping strategies). As Sandra, aged 16 noted: ‘for women it's not so hard—more for the men because they see them going around and hit them and everything, and sometimes they're not even involved in the maras’. This is notably different from research in other contexts, which highlights perceptions of public space as particularly dangerous for women (for example Day, 2001, on the US).

While young women also reported feeling afraid of what gang members may do, their experiences tended to centre around intimidation and verbal abuse, and the fear of sexual abuse, rather than physical violence. Thus, although young women often discussed public violence in terms of delinquency, they were also concerned about specifically gendered violence, including rape, and more general harassment (see also Koskela, 1999; Moser and McIlwaine, 2001). Olga, aged 16, for example, described how ‘they see that [a woman] is walking alone, and what the men do is start to annoy her for fun’. Much research in the North has suggested that such incidences of sexual harassment in public places serve to remind women of the threat of gendered sexual violence, and thus ‘evoke fear of more severe sexual attack through routinely creating a state of insecurity and unease among women’ (Pain, 1991, p. 421; see also Valentine, 1992; Smith and Torstensson, 1997; Koskela, 1999; on examples from the UK and Europe). Since young women were much more likely to discuss domestic violence than young men (see also Moser and McIlwaine, 2001), domestic violence may therefore be indicative of the ‘hidden victimisation’ of women (Smith and Torstensson, 1997) which, it has been argued (again in the context of the North) ‘may affect the broader spatial experiences and choices of women affected by or threatened by it’ (Pain, 1991, p. 417; also Koskela, 1999). Thus, if young women perceive themselves as vulnerable to gendered violence both in the home and in ‘public’, as the findings suggest, perceived danger in public and private space becomes mutually constitutive.

In terms of gender differences in the temporal experience of fear, it was more common for young men to talk about danger during the day than young women. It is telling that of the seven community maps on which recreational areas were marked dangerous during the day, five had been drawn by young men. It would appear that fear was more temporally constant for young men than for young women, with women more likely to fear attacks in public specifically at night. For one interviewee, Brenda, aged 14, for example, it was necessary to return home before six in the evening for fear of what may happen to her. As is now widely accepted, it can be seen that space (and hence fear in, and of, space) is socially constructed, and is therefore conditioned by the multiple social, and especially gendered, power relations it inheres.

Coping with everyday violence: Avoidance, compliance and engagement

So how did young people deal with risk on a daily basis? Young people's strategies depended to some extent on whether the problem of violence was deemed to be localised (and hence to some extent avoidable) or more pervasive (and less avoidable). It is useful to divide these strategies into three main categories: avoidance, compliance and engagement, within which there were often differences according to gender and age.

Avoidance is a logical, but highly isolating response to risk, which was most common among young women, and younger age groups (early teens and below) in general (see also Moser and McIlwaine, 2004). For these young people public space(s) had become synonymous with unmanageable danger, which was avoidable only through extreme distancing from public life. While in some cases this was the choice of parents who wanted to protect their children from both the violence and the negative influence of ‘the street’ (see below), it was often the choice of young people themselves. Carmen, aged 16, for example, explained the reason she did not spend time in the street: ‘it's because I don't want to, because I'm afraid of going out in the street because of everything that happens’. Such strategies therefore concerned tangible danger of which young people themselves has ample experience, both first and second hand, and which continually affirmed their beliefs (see also Pain, 2003). Other participants reported more localised avoidance of certain places that were particularly associated with territorial fighting, such as particular streets, places such as basketball courts (especially after dark), or larger areas with a reputation of gang conflict.

For some, however—particularly young men—the problem was deemed too pervasive to avoid altogether, as one young man, aged 16 noted: ‘you see them like that on every corner—it's hard to avoid them’. So young people were often faced with the problem of how to deal with potential confrontations. Many used various degrees of compliance to manage these situations, perhaps presenting themselves as being of little interest (‘tell them I'm nobody’, as one young man suggested), and to submit if they were confronted (‘give them what they want, otherwise they hit you’, as noted by another young man).

Yet, beyond these strategies of compliance, others—particularly older age groups and young men—described the constant struggle of maintaining enough distance from the maras not to get drawn in, but to know enough to be able to deal with them (see also Walklate, 1998, on the UK). Indeed, one interviewee, Efrain, aged 18, noted how fear of violence from gang members induced some level of forced interaction with them to ensure their own personal safety. After all, he remarked, ‘they're not going to harm themselves’. Indeed, it has been suggested that some young people who feel threatened by violence in public may join a gang in order to be protected, and thus that the security offered by gangs counters individuals perceptions of their vulnerability (see Rodgers, 2001; Garbarino et al., 2002). As Carmen, aged 16, noted ‘[in the gang] you're protected … and this is why young people are joining the gangs because they feel more than you—more than the others’. Group-based violence is therefore sometimes best countered with a collective solution, in the sense that groups offer more protection than would individual strategies (see also Pain, 2001).10

These engagement strategies suggest that a logical retreat from public life as a protective measure may prove counterproductive, since fear may be magnified through a lack of understanding. Isabel, aged 16, for example, described the various types and cycles of violence typical of different parts of the community, which she said had to be understood in order to deal with the violence on a daily basis. For example, gang violence was more severe in some sectors than others, and occurred more often at certain times of day, while some specific areas were more vulnerable to delinquency. Violence, she suggested, followed loose patterns, and so awareness of the specific risks of violence in different areas was useful in predicting where and when it might occur. As such, parental or self-restriction of young people's use of space (through using avoidance as a strategy) may actually heighten their feelings of insecurity and vulnerability, since it would reduce their ability to ‘read’ public space in terms of the nuances of violence, and hence render them less able to manage it. There is, however, a fine line to tread between this engagement and becoming (or being seen to be) involved.

Indeed, there was a tension visible in the perceptions of many young people between what were seen to be ‘a risk’ and ‘at risk’ young people (see Pain, 2003), or comparably those seen to be ‘good’ or ‘bad’ (see Rocha, 2000a). Yet, within this the potential co-existence of these categories was often reinforced. For example, one 16-year-old young man commented that his main problem was the maras ‘who hit us—maybe until they kill us, and get us into drugs’. This reflects concerns not only of being a victim of violence, but also of getting involved with so-called ‘malas juntas’ (the wrong crowd). What may seem to be distinct boundaries between different groups of young people can easily be crossed, such that young people may become victims on many different levels, as noted above. Yet, maintaining these barriers was sometimes considered an important way of presenting themselves as separate from what were seen to be ‘bad’ young people, such that there were seen by some to be clear divisions between different groups of young people in both communities.

Violence and ‘Community’

But how does this spatial behaviour relate to broader community relations? Broadly, violence associated with gangs affects not only the daily lives of their members, but also the community as whole (see Moser and Holland, 1997; McIlwaine and Moser, 2001, on Jamaica; Moser and McIlwaine, 2004). While some participants did report a sense of ‘community’ where they lived, it was also found that what unity did exist was not universal. Indeed, it was common for participants to report a sense of community only in their school, block or amongst a particular set of neighbours or interest group. Unity was therefore restricted to groups which were brought together either through repeated interaction, with unity therefore spatially restricted, or shared interests, with associated social capital thus ‘pocketed’ and often exclusionary. As David, aged 17 commented, for example: ‘the Catholics are like a community, the Catholic community, and the evangelical community, and the maras in their community’. While this pocketing was also connected to factors other than gang violence, not least to the fact that both communities were established relatively recently, and therefore lacked historically established social ties, it will be seen that it had been exacerbated as a result of violence.

It is significant in this sense that participants who did report feeling ‘part’ of the community did so because of belonging to local organisations who tried to reach beyond these divisions. One youth organisation in particular included the ‘strengthening of community integration’ in its mission, and endeavoured to involve as much of the community as possible in its various groups and activities. It ran various long-term family–community integration and development initiatives, together with an annual festival which aimed to bring young people, and other community members, together (see Winton, 2004b).

For most young people, however, a general lack of unity was reflected by the often discussed divisions between different groups of friends, and indeed gangs, as a result of rivalry and associated distrust. Yet, this is not to suggest that there were always strict divisions between gang and non-gang members. In the case of coping strategies outlined above, young people used communication with gang members as a way of reducing their own risk of being targeted, and other young people talked in more positive terms of friendships across the ‘boundaries’ of gang membership. Yet, this brought with it its own risks, mainly that other gangs may mistakenly see them as targets, and also the risk that they would become stigmatised by the wider community (see below). As Maria, aged 20, noted: ‘I talk to someone who's in a gang, and they confuse me for them—this is the consequence for young people who are not in gangs. That they confuse you’. For many young people in these communities, therefore, dealing with gangs required a balance between associating with them for the sake of protection, and distancing oneself enough to avoid first, being mistaken for a gang member, and second, becoming involved.

As a result, there were often said to be limits to these relationships, as noted by Ana, aged 13: ‘you can have friendships with gang members, but not very … not very involved ones’. Aside from being confused as gang members, some young people voiced concerns about becoming involved in gang activity themselves, such that maintaining distance from them removed any temptation or pressure to take part. On balance, as David, aged 17 commented: ‘it's not very convenient to have friendships like that’. It would seem that despite some connections between these two groups of young people, there remained overarching differences between them. For example, Karen, aged 17, noted that there were ‘two types of youth culture—those who are involved in church groups [like herself] and those who are involved in maras’, despite also saying that she had friends who were gang members.

Notwithstanding these (not unproblematic) social connections, a general lack of unity and mutual concern was rooted in a pervasive lack of trust and communication as a result of everyday violence (see also McIlwaine and Moser, 2001; Moser and McIlwaine, 2004). Fundamentally, gang violence was found to decrease security, increase fear, and therefore reduce the propensity of people to trust each other, with the result that horizontal networks of communication and co-operation were destroyed (see Table I). As Efrain, aged 18, suggested, someone could be a friend one minute, but ‘have pistol to your head’ the next. In this context, it becomes more normal not to co-operate, and thus the ‘virtuous circle’ of trust and participation so celebrated by Putnam (1993) becomes a vicious circle, whereby a lack of trust breeds further distrust (see Brehm and Rahn, 1997; Bartlett et al., 1999; Narayan, 1999). In the words of Mario, aged 16, ‘because of the violence—there isn't any unity between neighbours, between peers […] It's rare … it's rare to see a person who—who you trust’. It could be argued, therefore, that the seemingly contradictory terms ‘community’ and ‘violence’, become strongly associated in the minds and experiences of many community members (see Clauss-Ehlers and Lopez Levi, 2002). ‘Community’, in this sense, ceases to be something positive associated with (as suggested by participants) ideals of order, unity and trust, but in its broadest sense is associated with violence, and with the lack of order, unity and trust brought about by public (or indeed, community) violence. In such a context, ‘community’ is commonly positive only with reference to certain ‘pocketed’ groups, as noted above.

Fear of youth violence is also a significant factor in the general perceptions that older members of the community have of young people. Just as participants reported an overall lack of trust in the wider community, only two participants who discussed this in interviews felt that young people in general were trusted by other members of the community, while five interviewees suggested that it depended on the persons concerned. Common to all these accounts was the importance of communication in gaining the trust of older members of the community. Trust therefore depended on young people proving themselves to be an exception to the rule of distrust in order to gain respect. As such, the majority of participants reported an acute lack of trust of young people by older members of the community, with young people excluded and stigmatised by association. The source of this intergenerational conflict was largely seen to be rooted in the negative image of youth portrayed by members of maras, which had led to young people being labelled as potential or actual criminals. Brenda, aged 14, noted that ‘my mum doesn't trust youth now who hang around the streets […] because of the danger. Now you have to go around watching your back because if you don't, you get killed’. This was indeed supported by the majority of adults interviewed. One woman, aged 45, for example, described young people simply as ‘very … very badly behaved’ (see also Walklate, 1998). Figure 2—a site photographed by two groups of young people independently—illustrates the visibility of gang culture, and the effect this has on the perceptions of older community members.

Figure 2. Wall of mara graffiti, Guatemala City: photographed by two groups independently.

Many participants saw this treatment of youth to be unjust, as one interviewee, Ingrid, aged 15, suggested:

There's not a lot of trust because right now the community … doesn't … doesn't support us, because they tell us that maybe the community is being ruined by youth. But it's only some young people—it's not all of us.

However, it is interesting that some young people reinforced the stereotypes themselves. For example, Maria, aged 20, suggested that young people were very aggressive, and involved in gangs, and Jorge, aged 16, saw the majority of youth going around annoying people for fun. It is notable that, as in these two cases, young people sometimes separated themselves from the ‘youth’ of popular imagination (see also Tucker, 2003, on the UK). In addition, some young people not involved in gangs felt excluded as a result of the attention which is focused on youth (as) gangs. As one interviewee, Olga, aged 16, noted, ‘I don't [feel part of the community] because they don't take us into account. The truth is that they take more notice of the gang members’.

A lack of support from older generations is compounded when even those young people trying to help others involved in gangs are automatically assumed to be getting involved themselves, as one interviewee, Mayra, aged 16, complained, ‘what you try to do is to support—to help the others, and what they think is that you're also involved in the maras’. Both the problem of gangs, and the prejudice and discrimination arising from it combine to reduce the contact young people have both with each other, and with other community members, and thus reduce the benefits associated with a range of social contacts.

In summary, for the majority of young people a sense of ‘community’ was present only in isolated circumstances. This was seen to be connected to a lack of communication between different groups of people, especially between generations. Furthermore, young people were often thought to be excluded from the community as a result of the negative image of the gangs, and similarly, young people not involved in gangs were commonly set up in opposition to those who were.

Although the severely detrimental effects of gangs on all members of the community (especially young people) must not be made light of, neither should this phenomenon be treated as evidence of some pathologically violent trait of youth (see also Smutt and Miranda, 1998, on El Salvador). Moreover, the stigmatisation of gang members is major impediment to their rehabilitation. It is interesting in these terms that, despite the profound effects of youth gangs, participants overwhelmingly favoured comprehensive rehabilitative solutions over repression, pointing to (and empathising with) the profound causes of gang membership which repression does not address (see Winton, 2004b). While certainly not discounting the importance of security and social control, these views challenge the frequent media and government espousal of repressive security measures (see Winton, 2004b). Important to note, therefore, is that opinions vary as to what constitutes security, both in method and result.

Conclusion: violence and exclusion

Beyond the very real physical and psychological effects of violent victimisation, fear of violence leads to the spatial exclusion of community members, particularly young people, whereby assaults and territorial gang conflicts construct public space as dangerous and therefore restrict the extent to which young people feel positively connected to the community. Were gang territory to have been related more strongly to ‘community’ identity, rather than to sectors within this, other's fear of public space and the individuals/groups within it may have been significantly reduced. This relationship may also act in reverse, since greater social isolation, in the form of social exclusion and a weak position in the social network, serves to increase fear of crime (and violence), as noted above in terms of strategies of avoidance (Balvig, 1990, cited in Koskela, 1999). Thus, there is a (not unsurprising) mutually constitutive relationship between fear, spatial exclusion, a lack of social capital and social exclusion.

Together with affecting the amount of social interaction between young people and community members as a whole, public violence also significantly affects the nature of interaction, with interpersonal trust and communication often restricted to small, atomised groups. In addition, the very visible and negative image of ‘youth’ that the gangs portray can mean that trusting young people becomes considered dangerous. Thus, public violence restricts both the number and scope of young people's social networks. For gang members, however, this violence at once expresses and reinforces identity of their social group, which is itself based on explicit spatial boundaries. Yet, while the divisions between young people in this context may at first glance appear straightforward, and in this sense manageable, this research highlighted the difficulty with which young people negotiated the various influences, risks and relationships in their daily lives as youth ‘microgeographies’ (Matthews et al., 1998) become ever more complex and conflictive.

Some young people, it could be argued, find in gangs a sense of belonging which is lacking in the traditional social institutions of the family and community. Adopting this social identity at once offers them inclusion, while also exacerbating their wider exclusion. This rightly complicates the notion of ‘inclusion’ as the obvious remedy for exclusion. In addition, it could be argued that cultural difference may be incompatible with certain types of inclusion (see France, 1998), and indeed that the desire to ‘belong’ in broad terms of citizenship cannot be assumed as innate, especially in situations such as this where young people would gain very little in return. The question, then, of how to maintain the benefits of gang membership, while removing inherent violence and its multiple repercussions, becomes critical. It is also crucial to recognise, as highlighted by young people in this research, that the causes of gang membership are fundamentally structural, rather than individual. In this sense, therefore, increasing young people's access to economic, social and political opportunities is preventative in itself. Accordingly, demonising young people as deviant or innately violent does nothing to improve the situation, but simply represses any responsible and constructive responses which may take as their starting point the various needs of all young people, rather than the need to control them.

Acknowledgements

I gratefully acknowledge the ESRC studentship which financed this research, and I thank the research participants for sharing their time. I would like also to thank Nicola Ansell and Lorraine van Blerk for arranging the session on Children and Youth in Developing Areas at the RGS-IBG 2003 conference at which this paper was first presented. I am also grateful to the reviewers for their helpful comments, and Cathy McIlwaine for comments on an earlier draft.

Notes

1. There is obviously a distinction to be made between crime and violence, since not all crime is violent nor all violence criminal (see Moser and McIlwaine, 2004): the type of violence discussed here is specifically ‘public’ violence, be it overtly criminal or not.

2. It is beyond the scope of this paper to discuss the theoretical and ethical issues pertaining to research with youth, in the South. See for example Johnson et al. (1998), Morrow and Richards (1996), Young and Barrett (2001).

3. The names of the settlements have been changed to maintain confidentiality.

4. Originating from research in rural areas, the approach was first labelled Participatory Rural Appraisal (PRA). It has since been adapted for work in a range of contexts, and so is referred to here simply as Participatory Appraisal (PA).

5. Since this is an online document, no page numbers are available.

6. For further discussion of the psychological impacts of violence on children see for example Gibson et al. (1991), NNFR (n.d.).

7. Names of all participants have been changed to maintain confidentiality.

8. This has interesting implications for the particular meaning of ‘localism’ in this context (cf. Watt and Stenson, 1998).

9. It is worth noting that in the absence of adequate ‘legitimate’ state security and justice, this type of resistance at times verges on social cleansing, either through organised community vigilantism (in which state officials might themselves be involved) or sporadic assassinations (see Winton, 2004a).

10. To scale this up, is may be argued that youth gangs offer a collective solution to multiple exclusions, and may indeed be seen as a significant social movement (Moser and Winton, 2002), creating alternative routes to access key needs in the face of barriers to so many social, economic and political spheres (see also Rodgers, 2003).