19817 August 1999 Latin America and Caribbean Region Sustainable Development Working Paper No. 4 Urban Peace Program Series Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey August 1999 Dennis Rodgers The World Bank Latin America and Caribbean Regional Office Environmentally and Socially Sustainable Development SMU LCR Sustainable Development Working Paper No. 4 Urban Peace Program Series Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey August 1999 Dennis Rodgers The World Bank Latin America and Caribbean Region Environmentally and Socially Sustainable Development SMU The opinions expressed in this document are attributable to the authors, and do not necessarily reflect the position or official policies of the World Bank. This working paper series is produced by the Environmentally and Socially Sustainable Development Sector Management Unit (John Redwood, Acting Director) of the Latin America and Caribbean Regional Office. Additional copies may be obtained either from the authors, or from LCSES Projects Assistant Peter Brandriss (pbrandriss@worldbank.org, or tel. 202-473-9379). Cover photos (clockwise from upper left): Sao Paulo, Brazil-Francis Dobbs Trujillo, Peru-Peter Brandriss Tegucigalpa, Honduras-Peter Brandriss Guatemala City, Guatemala-Cathy Mcilwaine Guatemala City, Guatemala-Cathy Mcllwaine Background: La Paz, Bolivia-Curt Carnemark Contents Foreword ......................................v 1. Introduction .....................................1 2. Conceptualizing Youth Gangs: An Overview ......................................1 3. Latin American and Caribbean Youth Gangs and Violence: A Country Review ......................................4 Argentina 5 Brazil 6 Chile 7 Colombia 7 Costa Rica 8 Cuba 8 El Salvador 9 Guatemala 10 Jamnaica 1 0 Mexico 11 Peru 12 Puerto Rico 12 4. Nicaragua: A Detailed Case Study ..................................... 13 5. Youth Gangs and Violence in Latin America and the Caribbean: Elements of an Analytical Framework ................................... 15 Poverty 15 Particularism 16 Community Links 16 Drugs 17 Migration 18 6. Gang Policy Interventions and Approaches .............. ..................... 18 Youth Gang Suppression Program 18 Community Organization Program 19 Social Intervention Programs 20 Opportunity-Providing Programs 20 7. Conclusions ................................... 22 References .................................... 25 Boxes 1. Basic Issues for Youth Gangs in Latin America and the Caribbean 15 2. Summary of Policy Approaches to Youth Gangs 18 iii Foreword This document is part of a series of papers produced by the Urban Peace Program of the Latin America and Caribbean Region's Environmentally and Socially Sustainable Development Sector Management Unit (LCSES). The Urban Peace Program is funded jointly by the World Bank and the Swedish International Development Authority (SIDA). The program focuses on the dynamics of violence in Latin America and the Caribbean, its effects on poor communities, and the development of appropriate multisectoral strategies for violence reduction that would in turn help promote peace and development. Violence has emerged as a significant economic, social welfare, health, and governance issue throughout the region. It is important not only in countries experiencing political unrest, such as Colombia and Peru, but also in war-to-peace transitional societies, such as Guatemala, El Salvador, and Nicaragua, where levels of crime and violence remain high. Crime and violence erode physical, human, natural, and social capital, undermine the investment climate, and deplete the state's capacity to govern. Previously regarded as an issue of criminal pathology or human rights, violence is now recognized as a macroeconomic problem. These papers synthesize information generated by one stage of activities of the Urban Peace Program. In turn, they are a contribution to the growing information infrastructure of the World Bank's Knowledge Management System in the area of Social Development. The papers are published through the LCR Sustainable Development Working Paper series produced by LCSES. The series seeks to share the results of analytical and operational work, present preliminary findings, and describe "best practices" with regard to major sustainable development issues confronting the region. The findings, interpretations, and conclusions expressed in these papers are entirely those of the authors and should not be attributed to the World Bank, members of its Board of Executive Directors, or the countries they represent. John Redwood Department Director, Acting Environmentally and Socially Sustainable Development SMU Latin America and the Caribbean Region The World Bank v 1. INTRODUCTION Violence has been a chronic feature of Latin American and Caribbean societies. It takes many forms and has a wide range of social, political, and economic consequences. Historically, its most obvious manifestation in the region has perhaps been in the political arena. But since the beginning of the 1990s the nature of violence has changed significantly in much of Latin America and the Caribbean, and "now the most visible forms of violence stem not from ideological conflicts over the nature of the political system but from delinquency and crime" (Caldeira 1996, p. 199). Violence in the region has not lessened, however, and it arguably remains the "social pandemic" of Latin America in the late twentieth century (Organizacion Panamericana de Salud 1996). Youth gangs are among the main features of the new landscape of violence in Latin America and the Caribbean. Whatever the name they are know by-pandillas, maras, bandas, galeras, quadrilhas, barras, chapulines-they are ubiquitous throughout the region, and are recognized as a serious social development problem by all levels of society. While it is difficult to quantify the proportion of criminal and delinquent activity directly attributable to youth gangs, there can be no doubt that they account for a significant, although variable, share of actual and perceived violence. In Nicaragua, for example, almost half of all crimes and delinquent acts are attributable to youth gangs, which are regularly cited in opinion polls as the most likely perpetrators of violence in urban areas (Rodgers 1999). Understanding youth gangs and their violence is therefore crucial to the design and implementation of almost any social development project dealing with violence in contemporary Latin America and the Caribbean. But although youth gangs are an immediately visible manifestation of violence in the region, they have been the subject of relatively few studies, many of which often present a distorted or partial view of the phenomenon. This paper surveys the literature and identifies possible elements of a conceptual framework through which to analyze Latin American and Caribbean youth gangs and their violence. 2. CONCEPTUALIZING YOUTH GANGS: AN OVERVIEW The conventional definition of a gang has not been substantially improved since Frederick Thrasher's classic 1927 proposition: A gang is an interstitial group, originally formed spontaneously, and then integrated through conflict. It is characterized by the following types of behavior: meeting face to face, milling, movement through space as a unit, conflict, and planning. The result of this collective behavior is the development of tradition, unreflective internal structure, esprit de corps, solidarity, morale, group awareness, and attachment to a local territory. (Thrasher 1927, p. 57.) This description covers many of the basic structural and institutional features of gangs. But gangs are also extremely heterogeneous social institutions. They can have different motivations and activities-some have political overtones, others are concemed mainly with the accumulation of wealth, and still others revolve around identity and social status. Furthermore, gangs can vary considerably both within and across cultures. In the United States, for example, Latino or cholo gangs are very different from Black gangs. Similarly, gangs in post-communist Russia are distinct from those in the French urbanbanlieues. 1 2 LCR Sustainable Development Working Paper No. 4 Although some gangs have both adult and nonadult members, the notion of a youth gang implies that the gang members are all youths. Defining youth is problematic, however, because cultural boundaries that separate youth from adult are highly variable and arbitrary. In North America adulthood is legally considered to start at 18; this is not the case in much of Latin America. Furthermore, in North America the category of youth is ambiguous, often relating to an behavioral patterns and activities rather than age. While many youth gangs do solely involve people under the age of 18, others frequently have members as old as 25 who are still perceived as youths. Generally, the sociocultural category of youth is more an ascribed social role than a physical state. Consequently, it is important to take a broad view of what constitutes a youth when conceptualizing youth gangs. Youth gangs are only one of a range of youth formations, that are not always easily distinguishable. Youth congregate in more or less recognized peer groups and engage in collective behavior; this is a universal aspect of the youth life cycle, when young people learn to socialize and interact with their physical and social environment through the group, which provides definite referential parameters and behavioral codes. Sports and recreational clubs, friendship networks, and youth gangs are all examples of juvenile peer groups. Although conceptually distinct, many of these groups are interrelated and overlapping. While some of these groups differ substantially from gangs, others, for example a football team or a close-knit group of neighborhood friends, share many of the characteristics that Thrasher (1927) attributed to gangs-meeting face to face, milling, displaying solidarity andesprit de corps, and developing group awareness and attachment to a local territory. Inherent association with illegal and violent activities is what generally distinguishes youth gangs from other juvenile peer groups (Miller 1982; Vigil 1988). Of course, such behavior is not the exclusive preserve of youth gangs and their members. Other youth groups can be involved in such activities, although usually on a lesser scale than youth gangs. Juvenile delinquency is more than a youth gang phenomenon, even if there is evidence to suggest that being a member of a gang increases the likelihood of delinquent behavior (Curry and Spergel 1988; Fagan 1990). But what differentiates the illegal and violent behavior of youth gangs from that of other delinquent groups and individuals is that it is considered by wider society to be inherent to the youth gang. Although other youth groups can be violent or engage in criminal enterprise, this behavior is generally not perceived as normative; it might be seen as induced by alcohol or juvenile immaturity or as a form of temporary rebellion against authority. Youth gangs, by contrast, are seen as violent and criminal by nature. Even if many instances of illegal and violent behavior are attributable to individual gang members or small groups, society associates them with gang membership, rather than individual delinquency (Cohen 1990). The degree of criminal and violent activity is not the issue. The Brazilian quadrilhas are deeply implicated in drug dealing and regularly display extremely violent behavior, including murder. By contrast, the Guatemalan maras are little more than neighborhood friendship networks that revolve around more or less legitimate pursuits-such as playing sports, listening to music, drinking, or consuming drugs-and engage only in low-level violence, such as street brawls, muggings, and pickpocketing. Yet both are considered to be youth gangs by society because they are inherently associated with illegal and violent behavior. As these two examples suggest, the types of illegal and violent activities of youth gangs can vary considerably. Most gangs have regular conflicts with other youth gangs because of territorial encroachment or arguments between rival gang members. These conflicts can remain limited to fist Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 3 fights or include weaponry ranging from sticks and stones to knives, automatic firearms, mortars, and grenades. Similarly, gang larceny can range from vandalism, pickpocketing, mugging, and petty theft to drug trafficking, murder, and rape. The relationship of youth gangs to their local community is generally portrayed negatively. The criminal and violent activities of gangs often affect all indiscriminately, including local communities. However, in many cases the gang and the community develop a working relationship based on mutual aid and respect. Youth gangs refrain from harming community inhabitants and damaging community property, sometimes even going so far as to act as informal vigilante groups for the community, which then tend not to denounce gang members to the police. This is not surprising, since the youth gang members are themselves inhabitants of the community and embedded in wider kinship and friendship networks. But it is also obviously much more difficult for a youth gang to operate in a hostile context, and it makes strategic sense for a gang to make such alliances. More broadly, youth gangs are an urban phenomenon, possibly because a critical mass of youth is necessary for a gang to emerge (Fischer 1975). The socioeconomic background of gang members is generally, a low-income one. Youth gang structure is extremely variable. The size of a gang, which might range from as few as 8 members to more than 100 members, is not important. What is important is recognition of the group as a distinct social aggregation by the wider society. Some gangs display clear leadership, well-developed lines of authority, and other organizational features, while many are less cohesive. Generally, however, most gangs have some form of initiation process-a feature that they share with other social institutions, such as university fraternity clubs-to accede to membership. Initiation rites can vary considerably. In Nicaragua, aspiringpandilla members have to prove their worth in the face of danger, participate in some form of theft, and demonstrate that they share the same basic values as other gang members. Aspiring gang members in Panama have to kill a taxi driver (Rodgers 1999). In many ways, such rites of passage are social mechanisms to test whether an individual will be an adequate gang member, and whether the gang will be able to count on him or her unreservedly. Gang membership can often be categorized as either "core" or "peripheral" (Klein 1995). Just as with any social group, some (core) members are more involved in collective group activities than others (peripheral members). Generally, such a distinction does not affect gang membership, although some youth gangs do not permit selective involvement in gang activities. The overwhelming majority of youth gang members are male. A 1975 survey of U.S. youth gangs found up to 90 percent of gang members to be male (Covey, Menard, and Franzese 1992, p. 16). Most female gang members tend to be little more than sexually exploited appendages to the gang, although there are exceptions. Some Guatemalan maras have female leaders. Furthermore, all-female youth gangs exist in the United States, Mexico, and Nicaragua, although they are comparatively rare (Cummings 1994; Giordano 1978; Harris 1988; Rodgers 1999). The age of youth gang members tends to range from 7 to 25 years old. However, the modal age range of gang members is generally between 13 and 19, and the age range for many gangs is often much more restricted. For example, the gang members of the Costa Rican chapulines are all between 12 and 16 years old (Revista Centroamericana 1994). Youth gang membership is a finite social role, for obvious physiological reasons. Of course, depending on the degree of violence which a youth gang engages in, death can also be a limiting factor to the length of youth gang membership. 4 LCR Sustainable Development Working Paper No. 4 Individuals rarely remain associated with the gang beyond the normal age range. Members "mature out" of the gang, usually when they reach the upper limit of the age range. Ex-gang members typically integrate into mainstream society, taking a job or starting a family (sometimes, a cause of "maturing out"). But a significant share of youth gang members will turn to full-time criminality, either as members of an adult gang or individually. Many of theories have been put forward to explain the formation of youth gangs. Perhaps the most influential is the "social ecology" theory that gangs are the result of "social disorganization" of the poor urban areas in which they proliferate (Shaw and McKay 1942; Whyte 1943). Youth gangs are conceived as partial replacements for crucial social institutions such as the family, school, or labor market, which the poverty and administrative breakdown characterizing slums and inner cities have weakened or rendered dysfunctional. Other explanations have conceptualized gangs as examples of lower-class subculture (Cohen 1955; Miller 1958), forms of resistance to limited or "blocked" opportunities (Cloward and Ohlin 1960; Moore 1978), male maturation process and identity creation (Bloch and Niederhoffer 1958; Katz 1988), economic enterprise (Sanchez Jankowski 1991), or in terms of gang members' supposedly deviant or sociopathic personality traits (Yablonsky 1959). No single explanation has established itself as orthodoxy, however, and generally there is little consensus on the causes, circumstances, and processes of gang formation. In summary, a youth gang can be any discernible, self-formed group of socially recognized youths who: C Consider themselves to be a distinct group - Are generally perceived as a specific social aggregation by the widernongang population Are bound together by shared values Are collectively associated with illegal and violent behavior, both in practice and in general perception. Most gang members are male, fall between the ages of 7 and 25, and come from impoverished or socially excluded urban backgrounds. At the same time, however, youth gangs are heterogeneous in their structure, criminal and violent activities, and links to the local community. 3. LATIN AMERICAN AND CARIBBEAN YOUTH GANGS AND VIOLENCE: A COUNTRY REVIEW The literature on Latin American and Caribbean youth gangs is neither extensive nor comprehensive. The studies reviewed in this section cover: Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, El Salvador, Guatemala, Jamaica, Mexico, Nicaragua, Peru, and Puerto Rico. Youth gangs also reportedly exist in Belize, Honduras, Panama, Uruguay, and Venezuela, but detailed information about the gang situation in these countries is not available. Because many studies of youth gangs in Latin America and the Caribbean are unique, little intercountry comparison is possible. Argentina, Brazil, and Peru are notable exceptions. Much of the information presented in this paper has been obtained indirectly from more general studies of criminality, youth culture, or urban slums. Consequently, the country overviews are uneven in their coverage. The data are overwhelmingly qualitative, which also makes systematic comparison difficult. Although some of the information is recent, much of it is relatively dated, which since youth gangs are extremely volatile social institutions. Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 5 Argentina DeFleur's 1970 study of youth delinquency in the slums of C6rdoba remains one of the most extensive examinations of Latin American youth gangs. Drawing on court records and extensive interviews with 47 incarcerated, self-professed gang members, she found the barras, as the youth gangs were called, to be relatively small, with between five and eight (exclusively male) members averaging 16 years old. Organizational structures varied from gang to gang; some had strong, stable leaders who held the group together, but most (over three-quarters of the barnas identified) were more egalitarian and fluid (although even those with no recognizable leader or internal hierarchy tended to have relatively stable membership). Although ease of entry varied from gang to gang, there seemed to be no formal initiation rites for aspiring gang members. Many C6rdoban barra activities were nonviolent, such as talking, playing football (soccer), corner socializing, going to the movies, smoking cigarettes, and consuming alcohol. Drug use was almost unknown, perhaps reflecting the era rather than any conscious decision not to use drugs. Theft and robbery were also major occupations of gang members. Although armed assaults and robberies were not unknown, most of this criminal activity remained at the level of petty delinquency-such as pickpocketing, shoplifting, purse or watch snatching, breaking into cars, and generally stealing anything possible. The study provides little in the way of a wider C6rdoban context, but does compare C6rdoba juvenile court statistics with amalgamated U.S. juvenile court statistics for the same period. C6rdoban delinquents seem to have been involved more in larceny and violent crime than their North American counterparts, who specialized in robbery, burglary, and car theft. There seemed to be an economic motivation to barra criminal activities, for the youth almost invariably sold whatever they stole. Although there were more or less institutionalized "fences"-mainly local small business owners-the gang members frequently sold their stolen wares at bargain prices to people in the neighborhood, which implies a symbiotic relationship with the community. Other majorbarra delinquent activities, such as fighting with gangs from other neighborhoods, were significantly less important than theft and robbery and were linked to territoriality and traditional rivalries between neighborhoods. DeFleur argues that the underlying reason for emergence of youth gangs in Cordoba's lower-class neighborhoods was the poverty, exclusion, and "social disorganization" of these slums, which prevented the systematic and effective socialization of their inhabitants into mainstream Argentine society. Kuasfiosky and Szulik (1996) developed a similar argument to explain the formation of barras de la esquina (corner gangs) in contemporary Buenos Aires, although they emphasize the symbolic aspects of the social exclusion of Argentine youth more than their material conditions. The Buenos Airesbarra they describe is significantly different from those DeFleur studied in C6rdoba. It consists of 30-40 mainly male members-with a core of some 15 individuals-between 18 and 23 years old. Gang members are characterized by an extremely apathetic and nihilist outlook and display little solidarity either with each other or with the community. Szulik and Kuasfiosky (1997) add that the gang members tend to be unemployed, have low levels of education, and have spent substantial time in prison. The main barra activity is to gather regularly, particularly at night, in a specific location to play football, listen to rock music, drink alcohol, and, especially, consume drugs (usually marijuana). Petty delinquency is also widespread and the barra often clashes violently (and sometimes fatally) with police, although more for antiauthority reasons rather than territorial instincts, according toSzulik and Kuasfiosky. However, Auyero's 1992 study of Buenos Aires barras implies a sense of territoriality: he 6 LCR Sustainable Development Working Paper No. 4 describes how barras have names linked to their locality, such as Los de Itati or Los Locos del Cuarto Camino, both named after the streets where they tend to congregate. Fights between barras are relatively rare, and the police are considered to be the group's primary enemy. Brazil Youth gangs have a long history in Brazil-the nineteenth-century maltas in Rio de Janeiro are among the earliest recorded examples-and are a common feature of the social landscape of contemporary Brazil. Ranging from the relatively tame galeras cariocas to the more violent drug-trafficking quadrilhas, the Brazilian youth gangs are among the most diverse and violent in Latin America. Violence has been increasing in Brazil since the early 1980s, particularly in urban areas. In 1982, for example, Rio de Janeiro and New York shared the same homicide rate of 23 per 100,000 population. By 1989, New York's had declined to 17 per 100,000 while Rio's had exploded to 57 per 100,000, and has continued to increase (Zaluar 1994, p. 242). During the same period, the participation of minors in violent crime also increased, and by the mid-1990s two of every three homicides involved youths (Zaluar 1997, p.24). Among the main contributors to these trends, the quadrilhas are primarily drug-trafficking gangs of male youths between 13 and 25 years old. They are usually led by the eldest gang member, known as the cabeva (head). Gang members, or quadrilheiros, are initially attracted to the gang bymachismo and status, but later by a more materialist ethos of enriching oneself, similar to Colombian gang members. Quadrilhas are territorial, insofar as they physically stake out their drug markets and will defend them violently against other drug-trafficking gangs and criminals. Quadrilha wars tend to be murderous, and it is rare for a gang member to live beyond the age of 25. During 13 years spent studying a low-income bairro (neighborhood) in Rio de Janeiro, Zaluar (1994) recorded the deaths of 722 youths between 13 and 25 years old, about twice the number of quadrilheiros operating in that bairro at any one time. An unspecified number of casualties from the local population, all too often caught in the cross fire, can be added to the tally. As a result, the relationship between quadrilhas and the neighborhoods in which they operate is ambiguous. Quadrilhas act as a vigilante group for the community against petty criminals, and quadrilha cabecas often distribute goods in the community in which their gang operates. On the other hand, they may behave violently toward the local population, and show little identification with the community. Zaluar maintains that quadrilhas are not much more than illegal economic corporations of individuals brought together by self-interest. They see the community simply as a conglomeration of potential buyers of their wares. Keeping petty criminals at bay thus makes economic sense, because criminals would reduce the buying power of "clients" in the community, and distributing goods is considered good advertising. The other main type of youth gang in Brazil is the galeras cariocas. The se gangs have developed from samba and carnival schools. They strongly associate with their neighborhood, and generally take on the name of their bairro or favela (shantytown), whereas the quadrilhas tend to adopt the name of their chiefs. In stark contrast to the quadrilhas, the galeras cariocas. are centered mainly on pleasurable pursuits such as listening or dancing to funk music, consuming drugs, or drinking-although they will often fight wars with other galeras over territorial encroachment. They are nevertheless much less violent than their drug-trafficking counterparts, as these conflicts tend to be fought with stones and knives, rather than automatic firearms. These wars result in deaths and tend to mark the culmination of Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 7 semi-ritualized conflicts rather than their beginning, as is often the case withquadrilhas. Furthermore, galeras often act as informal vigilante groups for their communities. Galeras are more fluid than quadrilhas, without a strong leader, and are much more a reflection of youth culture than the drug- trafficking gangs. Chile Valenzuela's 1984 study of Chilean youth is based on over 600 interviews with individuals between 15 and 24 years old in three low-income poblaciones (settlements) of Greater Santiago de Chile. The study examines youth collective action in the context of the social dislocation and economic crisis that characterized Chile during the early 1980s after the second global oil crisis and the restructuring of the economy undertaken by the Pinochet regime. Youth gangs, orpandillas as they are known in Chile, are mentioned as one formn of youth mobilization. Valenzuela argues that the crisis of the early 1980s precipitated the exclusion of huge swaths of the urban lower-class youth. Work and education declined as mechanisms of integration into mainstream society as youth unemployment and underemployment reached 60 percent by 1983 and school attendance levels fell sharply. Valenzuela details a number of forms of youth collective mobilization in the resulting context of general apathy and frustration, including involvement in religious groups, such as Christian-based communities, and anti-Pinochet political activism, both organized and unorganized. However, these two forms of social action were not as common as widespread drug and alcohol consumption, and violent delinquency, both of which usually occurred in a gang context. These pandillas were characterized by strong leadership and displayed great internal solidarity. Gang members spent substantial time together, listening to rock music and playing pinball, and were violently aggressive toward passersby, destructive of public property, and often engaged in criminal activities ranging from armed assaults to holdups. Youth gangs in contemporary urban Chile seem to be less violent. Their activities are limited to football and occasional petty delinquency. Pandillas are perceived to be an inevitable minor nuisance associated with adolescence. Colombia Youth gang violence is difficult to analyze because it is so generalized. There were over 250,000 violent deaths recorded in Colombia between 1985 and 1996, which amounts to a homicide rate of 70 per 100,000 people-making it one of the most lethal countries in the world (Martin 1996-97, p. 181). Youth crime, not necessarily involving gangs, is responsible for most of this violence. Furthermore, youth violence in Colombia inevitably involves the ubiquitous drug trade. Although youth gangs predate the drug cartels, which formed in the mid- I 970s, the gangs and cartels have developed a strong symbiotic relationship, contributing to the proliferation of both. By 1990 there were no fewer than 120 youth gangs in Medellin, Colombia's drug capital, mostly located in the poor northeastern part of the city. They involved approximately 3,000 youth, averaging 16 years old (Salazar 1990). The first cartel-connected youth gangs emerged toward the end of the 1970s. They consisted of unemployed male adolescents from poor neighborhoods (although some middle-class youth gangs exist) who were recruited as sicarios (paid killers) for confrontations between rival drug groups. The initial motivation of sicarios was to make money "to have a good time, live life to the full, and help the family" (Salazar 1990, p. 110). However, the development of a "culture of violence" in Colombia over the past two decades has changed sicario attitudes and made them violent for violence's sake. As the 8 LCR Sustainable Development Working Paper No. 4 state attempted to bring these gangs under control in the 1980s, members began assassinating policemen, judges, and even politicians, often in spectacular suicide attacks. Death became an expectation, almost an aim. Contract killing is only one side of being a sicario. When not "on a job," sicarios congregate like conventional youth gangs. Strictly hierarchical and centered around a strong leader who holds the group together and acts with absolute authority, these gangs are in many neighborhoods the main form of socialization for children, who join at around 12 or 13 years old. Extensively involved in robbery, theft, drug consumption (marijuana, bazuco, cocaine), and assaults, and constantly at war with rival gangs, the gangs provide ideal sicario training for the pelados, as the younger kids are known. Their relationship with the local community varies. While in some neighborhoods they protect the local population, in others they are committing crimes and the local inhabitants are a primary target(Human Rights Watch/Americas 1994). In Bogota, where drug-related gangs are less predominant, neighborhoods nevertheless remain fraught with violence as a result of the numerous gangs of gamines (street children), who assault and rob indiscriminately. The existence of these children has given rise to local paramilitary vigilante groups that have the self-attributed mission of "social cleansing." The Centre for Research and Popular Education (CINEP) recorded 124 cases of gamines being killed by such paramilitary groups in Bogota between 1988 and 1993 (Human Rights Watch/Americas 1994, p. 3). Costa Rica Although Costa Rica has a significantly smaller juvenile delinquency problem than its Central American neighbors, youth gangs are increasingly influential actors in the growing national economy of violence. Since the beginning of the 1990s they have begun to appear in urban areas, particularly in San Jose. In 1993 the judicial police arrested and tortured six members of a gang called Los Chapulines, eventually killing one. The episode profoundly affected public opinion, so much that Costa Rican youth gangs are now known generically as chapulines. They tend to have up to 45 members, generally between 12 and 16 years old. By 1994 there were at least six of these gangs operating in San Jose (Revista Centroamericana 1994, p. 7). By 1997, however, they had proliferated and converted parts of the city into no-go areas. Although not as violent as their counterparts in El Salvador, Guatemala, or Nicaragua, Costa Rican youth gangs are increasingly criminally active, particularly in mugging and petty theft. The Costa Rican media often claim that these gangs consist of Nicaraguan immigrants, but there is little evidence to support this assertion (except that there are approximately 500,000 Nicaraguans in Costa Rica, which has a national population of 3.5 million). Cuba Not surprisingly, there is little in data available on youth gangs and violence in Cuba. Salas's 1979 study of social control and deviance in Cuba in the late 1960s and early 1970s provides some material, although dated. Part of his study focused on juvenile delinquency, which at the time was a rapidly growing phenomenon in Cuba, particularly in urban areas. In 1971, 50 percent of all crimes committed in Cuba involved a minor. The peak ages of delinquency were between 15 and 17 years of age (51 percent of all juvenile offenders brought to court), although a significant number of 12 to 14 year-olds were also Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 9 involved (33 percent of juvenile offenders). Most were involved in petty theft and some public property destruction. Many of the delinquent youth operated in gangs with characteristics similar to those of contemporary U.S.gangs. They had names such as The Zids, Los Chicos Now, Los Chicos Melenudos, Los Chicos del Crucifijo, Los del Palo, Los del Tercer Mundo, and Los Sicodelicos. Youth gangs were often linked to adult criminal gangs. Interestingly, studies of delinquency prior to the 1959Castrist revolution show similar trends to the late 1960s and early1970s, although the author provides few comparative references. This implies a temporal continuity in the existence of youth gangs. Although not directly relevant to Cuba, Page's 1997 study of Cuban-American youth gangs in Miami is perhaps a useful comparative source for the contemporary youth gang situation in Cuba. Using interview data collected in 1978-80 in the Little Havana area of Miami, Page looks at the rise and fall of two Cuban-American youth gangs called the Vulcans and the Jutes. Drawing on a "social disorganization" approach, he argues that when Cuban exiles initially arrived in Miami after the 1959 revolution, they experienced difficulties in integrating-including family disruption, ethnic friction, struggles with a new language, discrimination, and poverty-which resulted in the formation of youth gangs as a form of self-defense. However, by 1978 these gangs were disappearing as a result of the growing economic power, prosperity, and integration of the Cuban exile community and the increasing mainstream employment opportunities for Cuban-American youth. Although the context is different, Page's analysis resonates with present-day Cuba in transition, which-if anecdotal evidence from tourists is to be believed-is afflicted by rapidly rising crime and delinquency. El Salvador Violence in El Salvador has actually increased since the end of the 12-year civil war in 1992. With a homicide rate of 117 per 100,000 people, El Salvador is now the most violent country in Latin America, according to the United Nations (Ortiz 1997). The maras dominate the landscape of violence-gangs of youths and young adults that have staked out their turf in all but the smallest settlements of the country-although they remain concentrated in San Salvador (Tuckman 1996). A recent study of maras reports that three-quarters of members are male, their average age is 18.5 years, and most come from single-parent families (Revista Centroamericana 1994). They are extremely violent-often simply for the sake of being violent-and are extensively involved in assaults, theft, killings, and warfare against rival gangs, generally for territorial reasons. As Elias Bolailos, director of the Ricaldone Institute in San Salvador, remarked in a recent interview, "the members of the maras are 'children of war,' having only seen or heard of war since they were five or so . . . now, aged 17, they are bringing to bear these experiences in a new context" (Revista Centroamericana 1994, p. 5). Certainly, many members of maras are ex-guerrilleros or army personnel who experienced the war and since 1992 have had few employment opportunities in a saturated labor market. In contradiction to Bolafnos's analysis, other countries, most notably Guatemala, have also had long wars but have not spawned youth gangs as violently destructive as the Salvadoran maras. Perhaps the most important factor in explaining the emergence of the maras is the high rate of return migration and deportation of illegal Salvadoran nationals from the United States back to El Salvador since the 1992 peace accords. The deportation rate of illegal Salvadoran immigrants from the United States is the highest in Central America. Many of these returning migrants and deportees were members of Latino gangs return with "the education they received on the mean streets of Los Angeles." (Tuckman 1996, p. 3). Salvadoran maras have adopted violent behavioral patterns typical of 10 LCR Sustainable Development Working Paper No. 4 U.S. youth gangs, including drive-by shootings, high-casualty gang warfare, and daylight armed assaults. (One of the most important youth gangs in San Salvador is called theMara 18, after 18th Street in Los Angeles.) As a consequence of return migration and deportation, most gang members are not from the lowest socioeconomic classes, but from the working or lower-middle class (Revista Centroamericana 1994). Guatemala As in El Salvador, youth gangs in Guatemala are called maras. However, the similarity between them ends there. Although there is also a "culture of violence" in Guatemala as a result of the 30-year civil war that ended in 1996, the Guatemalan gangs are much less violent and destructive than their Salvadoran namesakes. A 1996 study of the maras in the San Antonio de Padua barrio (neighborhood) in Guatemala City found that most of the Guatemalanmara delinquency involves low-level mugging of drunks, women, or children, and pickpocketing on crowded buses in small groups of two or three (Nufiez 1996). Some maras also charge a "toll" for crossing their territory. Firearms are rarely used, even in gang conflicts, although knives are commonplace. However, recent anecdotal reports suggest that this is changing, partly as a result of guerrilla demobilization. Maras have existed in Guatemala since the 1980s, mainly in urban areas. They generally consist of 30 to 40 youth, on average about 14 years old. About half the gang members go to school, according to a 1988 study (Levenson-Estrada 1988). The maras have an egalitarian, democratic structure, although they are divided into age groups. Surprisingly, the study found thatmaras are also relatively egalitarian in terms of gender. Many maras have female leaders (Revista Centroamericana 1994). Initiation into and exit from the maras is informal and voluntary, and membership is considered a finite part of a youth's life. Being a member of a mara is to be part of a collectivity where everything is shared and everybody helps each other. Members of maras not only share the profits of theft and mugging, and drink, consume drugs (mainly marijuana), and sniff glue together, but every mara also has its own history, codes, territory, language, conflicts, and alliances. A mara's territory can cover just a few blocks or several barrios and will be demarcated by gang- specific graffiti. Most neighborhoods have several youth gangs. Interestingly, maras in the same neighborhood often fight each other to defend their territory. The 1996 survey of barrio San Antonio de Padua notes that there were six maras-the UVA-26 (Uni6n de Vagos Asociados de la 26 calle- United Association of Delinquents of 26th Street), Noruega, 22, San Juan, Alf, and los Burgueses, also known as BK-which all fought each other with stones and knives, often in alliance with maras from the surrounding neighborhoods. There are also "ethnic maras," consisting of individuals from a single indigenous group. These are probably a result of the internal displacement indigenous people to Guatemala City during the civil war. Jamaica Jamaican youth gangs tend to be overshadowed in the research literature by the more notorious, violent international drug-dealing posses and yardies (Gunst 1995; Small 1995) that tend to have an adult membership. A 1997 World Bank study of urban poverty and violence in Jamaica noted that youth gangs remain an important part of the violence landscape, particularly in low-income communities (Moser and Holland 1997). These groups of males 12 to 15 years old are well-armed and often fight each other and the police violently, contributing to the high local homicide rate (twice that of the U.S. Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 11 rate in the late 1980s) (Moser and Holland 1997, p. 1). Their relationship to the local community seems to be negative; local inhabitants restrict their movements as much as possible as a result of the ever- present danger. The gangs seem to emerge as a result of a power vacuum in the community; inhabitants in one of the communities surveyed by the World Bank study claimed that when there had been a strong leader in their neighborhood during the early 1980s the gang problem had been much less serious than it is today. Increased unemployment and poverty have also affected the development of gangs, although gang members are generally not the poorest of the poor, but often educated youths. Becoming a gang member is more about obtaining respect: "no gun, no girl" was a common expression reported in the study (Moser and Holland 1997, p. 15). Mexico In their 1968 cross-cultural study of delinquency and crime, Cavan and Cavan reported that youth gangs were largely absent in Mexico City and other urban centers, but did exist in rural areas of Mexico. However, by the early 1970s youth corner gangs were already present in the peripheral slums of Mexico City such as Ciudad Nezahualcoyotl (Velez-Ibafiez 1983). By the early 1980s Mexico City and most other sizable urban centers in the country were plagued by the notorious chavos bandas (Feixa 1995; Klein 1995, pp. 216-17). These early youth gangs, inspired by the British punk movement, were mainly interested in listening to music, talking slang, or consuming drugs. Rival groups occasionally fought each other, but in a relatively nonviolent manner; fatalities occurred rarely. Feixa (1995) theorized that gangs were attempts by marginalized youth to establish an alternative lifestyle in the absence of opportunities to integrate into mainstream culture. However, the 1990s have seen new developments in the Mexican youth gang culture. As Martinez (1997) points out, "Mexico ... is being transformed in this era of free trade and global culture: It is on the receiving end of a tremendous amount of cultural influence from the United States, for better and worse." Among the worst features of this cultural exchange has been the emergence of violent Los Angeles-style cholo Latino gangs. They have been a noticeable presence for several years in provincial cities and towns of southwestern and northem Mexico, which have the highest rates of immigration to the United States. Since 1995 these gangs developed rapidly in Mexico City and its surrounding slums and suburbs. The devaluation of the peso in December 1994 and the subsequent economic crisis are probably at the root of this trend, having led to increased migration to the United States from the capital city (Martinez 1997, p. M6). These new gangs, called clikas (cliques), have not only adopted the cholo style of dress (baggy trousers and bandannas), musical taste (hip-hop music and oldies), or slang (Spanish peppered with pochismos-Anglicisms and neologisms combining both languages). They have also adopted the behavioral patterns of their North American counterparts: graffiti tagging, gang warfare, drive-by shootings, and crack cocaine consumption (although not dealing). Some of the behavior comes from the influence of North American culture, but more importantly many gang members were once emigrants to the United States, who assimilated U.S. gang culture either directly or indirectly. Gang-related killings have become an almost daily occurrence in Mexico City and its slums as youth gangs proliferate unchecked. In Ciudad Nezahualcoyotl, one of Mexico City's largest slums, it is estimated that there is at least one clika of about a dozen male gang members in every one of the 85 12 LCR Sustainable Development Working Paper No. 4 neighborhoods (Martinez 1997). Furthermore, as Cummings's 1994 study of youth gangs in Chihuahua indicates, there are often parallel female groups as well, also extensively involved in violent gang fighting, though their the weapons tend to be limited to sticks, stones, and knives. Peru The decline in the activities of both the Sendero Luminoso and MRTA revolutionary movements notwithstanding, fear and violence remain primary characteristics of life in late-1990s Peru. However, it is now juvenile delinquents rather than guerrilleros who underpin the climate of insecurity (Reyna 1996). Following a decade of relative decline, Peruvian juvenile delinquency has risen dramatically since 1992, reaching unprecedented levels. This increase in youth crime has been marked by the appearance of youth gangs, or pandillas, especially in Lima (Canepa 1993). Most of the pandillas are based in the slums of Lima, particularly those north of the city (Ruiz Secada 1993), but there are also pandillas de pitucos (rich kids' gangs), which originate in wealthier neighborhoods (Portocarrero 1996). The overwhelming majority of youth gang members are male. They range between 10 and 24 years old, with the modal age range 15 to 17 (Villalobos 1993). The gang is hierarchically structured, with a strong leader, who is legitimized through his strength or violence. A 1987 study of a Lima gang claimed that the leader, or capo, is the gang member who commits the most dangerous assaults or robberies (Sanchez Le6n and de Mastro 1993, p. 143-45). Most youth gang larceny consists of snatching bags, chains, or watches, shoplifting, or mugging. There is also a certain amount of more sophisticated criminal activity, such as stealing automobile air filters or radios, but this is often directed by adults. Youth gangs are not a new phenomenon in Peru-many operated in the 1970s in Lima-but they are so numerous today that they have begun to overshadow the more professional criminal bandas (bands). Despite this overlap of delinquent activity, pandillas and bandas are very different. Bandas tend to be small groups of under a dozen, generally adult professional criminals who combine their skills to enrich themselves. The pandillas tend to be larger aggregations of 20 to 30 individuals from the same neighborhood. Furthermore, pandilla delinquency is more amateur and spontaneous than bandas criminality, and is neither spectacular nor motivated primarily by economic considerations. The more violent and better armed bandas do not tolerate serious competition, just as the numerous organized drug-trafficking rings that proliferate in Lima do not permitpandillas to get involved in anything to do with drugs other than consumption (mainly marijuana and cocaine). The pandillas provide youths with a sense of identity, often grounded in the local community in which they emerge, rather than a means of pulling themselves out of poverty through crime (Sanchez Le6n and de Mastro 1993). Puerto Rico A 1975 study of delinquent and nondelinquent youth in the slums of San Juan between 1966 and 1972 found that: The "West Side Story" syndrome, with youth gangs as a way of life and gang warfare an endemic problem, [was] almost completely absent. Juvenile crime [was] likely to be solo or dyadic. Rumbles, turf, gang identification symbols, and all the other aspects of group conflict so aptly described and carefully studied by several generations of United States criminologists Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 13 [were] foreign to San Juan, even though prevalent among the cousins of these same youths in New York. (Ferracuti, Dinitz, and Acosta de Brenes 1975, p. 126.) Gangs did exist, with members typically being between 11 and 17 years old. However, they were viewed as a collective means of coping or surviving in the low-income slums of the city, and centered on economic activity (some of it illegal, and generally involving low levels of violence). 4. NICARAGUA: A DETAILED CASE STUDY Nicaragua, despite the end of its civil war in 1990, has seen steadily rising levels of violence and criminality during this decade. Crime rates almost doubled between 1990 and 1995, to about 1,300 crimes per 100,000 people in 1996. Over 40 percent of all crimes are perpetrated in Managua. Violent crimes-especially armed robbery, assault, rape, homicide, and grievous bodily harm-have increased most, accounting for almost half of all reported crimes (Granera Sacasa and Cuarezma Teran 1997; El Nuevo Diario, January 4, 1997, p. 9). Although these trends are no doubt accurate, the figures are underestimations, in part reflecting the inefficiency of the police. An April 1997 Gallup survey found that one in six Nicaraguans had been assaulted in the past four months, a proportion that rose to one in four for the inhabitants of Managua, the capital city (La tribuna, May 2, 1997, p. SA). The most notable feature of the violence landscape in contemporary urban Nicaragua are the ubiquitous pandillas, which roam the streets of the city barrios, robbing, beating, terrorizing, and occasionally killing. Almost half of all crimes committed in Nicaragua is attributable to these youth gangs, which are cited in opinion polls as the most likely perpetrators of violence in urban areas. Pandillas can have between 10 and 120 members (pandilleros) ranging from 7 to 22 years of age. The overwhelming majority of gang members are male, with the notable exceptions of two all-female pandillas in the Managua barrio 19 de Junio and the outlying Ciudad Sandino. Members tend to all originate from the same location, generally a lower-class barrio or an asentamiento of the city. Pandillas tend to have a well-defined primary territory, that corresponds geographically to their neighborhood of origin. They can also have a secondary territory, an area outside their barrio of origin that they have taken from other pandillas. Large neighborhoods can have several gangs, that divide territory among themselves. Much pandilla violence is petty delinquency. Mugging, pickpocketing, or shoplifting constitute the major criminal activities of most gang members. However, many pandilleros are also extensively involved in rape, armed robbery, and assault. Clashes with the police are widespread and often bloody, as the pandillas are sometimes better armed than the authorities. Furthermore, territorial conflicts between pandillas frequently transform parts of Managua into quasi-war zones. Pandilleros fight each other to defend their territories in a semi-ritualized manner, initially with sticks and stones but later with knives, broken bottles, and eventually AK-47s, fragmentation grenades, and mortars, with dramatic consequences for both the pandilleros-on average between 3 and 5 percent of gang members are killed every year-and the local population. However, generally a pandilla does not harm or victimize its local barrio population. This behavior stems partly from the territorial logic of the pandilla; gang members generally consider themselves to 14 LCR Sustainable Development Working Paper No. 4 be the guardians of the community. Indeed, a pandilla often acts as an informal vigilante group for its barrio, confronting and attacking strangers. Gang members sometimes act as bodyguards for inhabitants when they run errands outside the neighborhood. In return, barrio residents tend not to call the police when the local pandilla is involved in violence, and although very critical of pandillas generally, they are more forgiving of their own youth gang. Thus, in a contemporary Nicaragua marked by uncertainty, confusion, and state disintegration, pandillas can be conceptualized as structuring institutions that provide the local barrio population with some security and constitute the only concrete form of community spirit in a context of widespread social atomization. The pandilla is also a structuring institution for its members. It is a collective group that is guided by definite referential frameworks, linked to considerations of territoriality, identity, and local values, but grounded in violence. To be a pandillero implies a sense of identification with the local community, expressed most visibly through violence. Although much of pandilla life revolves around violent activities, youth gang members are also involved in other activities. The pandillas are a forum for socialization and the creation of strong social networks that persist beyond the gang. Being apandillero is a finite social role; between the ages of 18 and 22 individuals leave the gang, most often because of the accidental pregnancy of one of their sexual partners. Gang members talk of needing to "become responsible," and their means of doing so is to assume a more normative adult behavioral pattern. Less violent pandillero activities include drug-mainly marijuana, glue, and florip6n (a local hallucinogenic)-and alcohol consumption. Drug dealing is generally not a pandilla activity, however. In terms of structure, pandillas are divided into age groups, generally 7 to 12 year-olds, 13 to 17 year- olds, and 18 to 22 year-olds, each of which specializes in different violent activities. The youngest group is a learning group; members are taught by members of the other two age groups how to street fight at the most basic level, for example. The middle group is mostly implicated in gang warfare and petty delinquency. Members of the eldest group tend to be involved mostly in "harder" criminal activities such as muggings, assaults, and robberies, while reducing their participation in gang warfare. These different age groups tend to operate separately, unless the barrio is under attack or there are advantages to "going onto the streets" as a large group. Pandillas can subdivide geographically within the primary territory and operate as separate units, although as a general rule, the subdivisions do not fight each other. Most pandillas have an identifiable leader, who is normally the individual considered to have the most violent and dangerous reputation. Each gang subdivision has such a leader, but one individual is commonly deferred to as the overall head of the gang. Pandillas are not a new phenomenon in Nicaragua; they have antecedents in the 1960s and 1970s. However, they were not as common then as today. They disappeared completely during the first half of the 1980s, but began reappearing slowly toward the end of the 1980s, probably as a result of the economic and social dislocation caused by the civil war. By 1988, it was estimated that there were 20pandillas operating in Managua. The end of the war in 1990 saw the demobilization of thousands of youths (the age of conscription in Nicaragua was 16) and an explosion in the number of pandillas. Gang membership provided these young ex-conscripts with a means of fitting into the new Nicaraguan social conditions. Although these demobilized youths "matured out" of the pandillas, gangs have continued to proliferate; there are now more than 100 pandillas in Managua alone, involving over 6,000 youths. This large number is likely related to the state of quasi-anarchy that characterizes contemporary Nicaragua as a result of continuing economic turmoil, social conflict, and political strife.Pandillas will probably continue to grow rapidly, as conditions seem unlikely to improve (Rodgers 1997; 1998; and 1999). Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 15 5. YOUTH GANGS AND VIOLENCE IN LATIN AMERICA AND THE CARIBBEAN: ELEMENTS OF AN ANALYTICAL FRAMEWORK It is obvious from the review of the literature above that there is a lack of information about youth gangs in Latin America and the Caribbean. Although the information is relatively extensive for some countries, most notably Brazil, Nicaragua, and Peru, for many, such as Colombia, Costa Rica, El Salvador, Guatemala, and Mexico, it is extremely poor. In a number of instances-Argentina, Cuba, and Puerto Rico-studies are also very dated, which is problematic considering that youth gangs can change rapidly. There are no data for over half the countries of the region, although in some cases- Chile and Jamaica, for example-it is possible to extract some information on youth gangs from broader studies of youth culture and violence. Further research is crucial to establish a proper understanding of youth gangs and their violence. Nevertheless, based on existing studies, it is possible to outline certain elements of an analytical framework with which to conceptualize youth gangs and their violent behavior in Latin America and the Caribbean. Five basic issues emerge that are especially pertinent for policy prescriptions and program design. They are summarized in Box I and discussed in greater detail below. Poverty Youth gangs in Latin America and the Caribbean generally emerge with in the context of urban poverty. There are some examples of middle-class gangs, most notably the pandillas de pitucos in Lima, Peru, but the vast majority of Latin American youth gangs are linked to the poorer barrios and Box 1. Basic Issues for Youth Gangs in Latin America and the Caribbean . Poverty. The majority of youth gangs in Latin America and the Caribbean tend to emerge in contexts of urban poverty, although there are some notable exceptions, such as the Peruvian pandillas de pitucos ("rich kids' gangs"), which originate in the wealthier neighborhoods of Lima. . Particularism. Youth gangs in Latin America and the Caribbean are significantly different from country to country in their organization, structure, logic, dynamics, and activities, including acts of violence. Local conditions affect both the formation of youth gangs and their behavioral patterns. * Community links. Some youth gangs have strong links to their local communities and do not direct their violence or their criminality toward the local community-for example Nicaraguan pandillas. But others have weak links and are indiscriminate in their illegal and violent activities-for example Colombian sicarios. . Drugs. Youth gangs involved in drugs trafficking (as opposed to drug consumption, which is almost a given among youth gang members), such as the Brazilian quadrilhas, tend to be more violent than those that are not. However, non-drug-trafficking youth gangs can also display extremely violent behavior-for example Salvadoran maras. * Migration. Post-war return migration and the increasing deportation of illegal emigrants from the United States have become important factors in the formation of youth gangs in certain Latin America and Caribbean countries during the late 1990s (El Salvador, Guatemala) and in the transformation of gang activities in others, generally involving more violent patterns of behavior (Mexico). 16 LCR Sustainable Development Working Paper No. 4 favelas of the continent's cities. In many ways, poverty is more a context than a cause, since not all the slums of Latin America have youth gangs. Furthermore, although some gangs, such as the C6rdoban barras in late-1960s Argentina or the Puerto Rican youth gangs of the early 1970s, seem to have used crime as a form of economic enterprise, other gangs, such as the barras de la esquina of Buenos Aires or their Jamaican counterparts seem to be more linked to consumerist and identity considerations rather than their economic situation. Nevertheless, the material conditions from which youth gangs emerge are not to be discounted and provide the setting of any social development project that deals with Latin American and Caribbean youth gang violence. Particularism Particularism can be discussed at two levels in relation to youth gangs. At one level, youth gangs in the region display significant variation from country to country-and often within countries-in their organization, structure, logic, dynamics, and activities. At another level, and more importantly from a prescriptive perspective, this heterogeneity is partly linked to the specific national and local circumstances within which youth gangs develop and operate. For example, the degree to which youth gangs are violent varies considerably. While some youth gangs, such as the present-day Chilean pandillas or their Puerto Rican counterparts in the 1970s, are more or less nonviolent, others, including the Colombian sicarios and the Salvadoran maras, are extremely violent. This high level of violence is linked to country-specific factors. The political violence and civil strife characterizing contemporary Colombia, and the close association of the sicarios to the violent drug cartels, inevitably play a role in explaining the extreme violence of Colombian youth gangs. A culture of violence is also important in countries such as El Salvador, Guatemala, or Nicaragua, which have long histories of war and strife. However, no single such factor can be considered sufficiently explanative. Maras in Guatemala are much less violent than their Salvadoran counterparts, for example, despite over thirty years of civil war. From a policy perspective the particularism of youth gangs obviously has important implications. The specific local conditions within which youth gangs emerge and operate need to be taken into account, with programs varying according to different contexts and circumstances, as well as, of course, the heterogeneity of youth gangs themselves. Community Links Some gangs, such as Nicaraguan pandillas and the Brazilian galeras cariocas, have strong links with their local communities and do not direct either their violence or their criminality toward the local population. Both actually defend their local population from violent threats and maintain a form of social order. In retum, the community does not denounce them to the police, often aiding and abetting them instead. Other gangs, in contrast, have very weak links with their local communities and are indiscriminate in the scope of their illegal and violent activities-for example, the Colombiansicarios or the Brazilian quadrilhas. Although the climate of terror that the violence of these type of gangs induces can often impose a "law of silence" on the local population, the local victims of gang violence often appeal to the police and do not help the gangs. Certain forms of youth gang violence, such as gang warfare, would seem to inevitably have dramatic consequences for local populations, regardless of a gang's community links. However, closer Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 17 examination indicates that this is not always the case. Nicaraguan pandilla conflicts, for example, are highly organized and develop in a quasi-ritual manner, with a set pattern of escalating weaponry and priorities. The primary targets of gang warfare are members of the rival gang, and protection of the local population is paramount. While local population casualties do occur during the course of gang warfare in Nicaragua, the ritualized form it takes reduces the scope of harm and destruction and provides forewarning to the local community. The Colombian sicarios, by contrast, have no such restraint, and their wars tend to be indiscriminately devastating and bloody for the local population. These differences are related in the violent death rates in Colombia-70 per 100,000 people-and in Nicaragua officially 15 per 100,000 people, although the rate is probably at least three times higher (Rodgers 1999). Undoubtedly one of more important reasons for the difference in community links is the degree to which the state is present and can impose a form of social order. For example, contemporary Nicaragua is characterized by anarchy and social fragmentation, and the pandillas represent the only source of social organization beyond individualism. The same is probably true of the slums of Rio de Janeiro. However, the presence of both galeras cariocas and quadrilhas indicates that such circumstances alone are not enough to determine the nature of youth gang links to their local community. Other factors, such as specific local features and context, are also important to consider. The Nicaraguan pandillas, for example, draw on the country's ideologically charged past to justify their solidarity with their local communities, whether it be the obvious collectivism ofSandinismo or the sense of community the Contras developed in exile and opposition to the revolutionary regime (Rodgers 1999). But other factors can lead to a similar identification and attitude towards local community, such as ethnicity, for example. Certainly, this has been the case of some "ethnic" gangs in the United States, most notably in Chicago (Suttles 1968). The nature of youth gang community links has crucial policy implications. In a socially atomized society such as Nicaragua pro-community gangs such as the pandillas represent elements that can be actively integrated into social development projects to reduce the violence affecting communities. While the violent behavior of the Brazilian quadrilhas or the Colombian sicarios creates widespread fear and distrust in the local community, breaks down networks of solidarity and cooperation, and reduces physical mobility, pandilla violence is often directed at protecting the local population and providing a sense of security. In many ways pandilla violence can be conceptualized as a response to wider violence and chaos harnessed to further reduce the negative effects of these violent circumstances. Drugs Youth gangs, drugs, and violence are often linked. These links mainly take the form of consumption (marijuana, cocaine, and to a lesser extent glue-sniffing), which generally has little impact on the degree to which youth gangs, are violent (although it can influence individual acts of violence). However, there are exceptions; the Brazilian quadrilhas are drug-traffickers and the Colombian sicarios were closely associated with the Medellin drug cartels. Although, both gangs are among the most violent in Latin America and the Caribbean, whether their violence is a result of close association with the drug trade is open to debate. U.S. gang experience suggests that drugs could be a factor in gang violence (Covey and others 1992). Some non-drug dealing youth gangs, such as the Salvadoran maras, also display extremely violent behavior. 18 LCR Sustainable Development Working Paper No. 4 Migration International migration flows from the United States to the region, whether in the form of post-war return migration or deportation, seem to be increasingly a factor in youth gang formation or their transformation into more violent social entities. This is the case in El Salvador, which has one of the highest emigrant populations in the United States, and the highest return migration and deportation rates in Central America. Many returnees bring with them a violent gang culture that they assimilated in the United States, as gang members. Mexico has experienced a similar process; formnerly anodyne youth gangs have become extremely violent groups as result of the transference of U.S. gang culture to the Mexican setting by returning emigrants. The Nicaraguan media alluded to this potential consequence of international return migration in mid- 1997, before President Clinton signed an amnesty for illegal Nicaraguan migrants in November of that year (Rodgers 1999). This phenomenon should be taken into account in policy prescription and program design because it could affect a number of different countries and societies in Latin America and the Caribbean, including Cuba (post-Castro), Guatemala, Honduras, Jamaica, Panama, and Puerto Rico. 6. GANG POLICY INTERVENTIONS AND APPROACHES There is little information on gang policy interventions and program in Latin America and the Caribbean. To supplement what data there are on the region, this section draws on the U.S. gang experience. Gang policy interventions and approaches aimed at countering youth gangs and their violence can be divided into four basic categories (Box 2). Youth Gang Suppression Program The overwhelming majority of gang intervention program in the region tend to be suppressive. They involve a strict law enforcement approach to youth gangs and their violence, including the arrest, incarceration, and supervision of gang members to prevent and limit delinquent activities, for example. Such strategies are often perceived as the most effective means for law enforcement and justice agencies to respond to youth gang violence and frequently have the most favorable impact on public opinion. Box 2. Summary of Policy Approaches to Youth Gangs The first two approaches are preventative in nature: * Suppression program attempt to reduce youth gang activities throughpunitive means, usually involving police action. * Community organization projects aim to change the context withinwhich youth gangs operate by mobilizing and organizing gang-affected communities, thusreducing the scope for gang activities. The following two approaches are rehabilitative or corrective and seek to affect youth gangs and their members directly: * Social intervention projects include outreach and counseling program aimedat preventing youth gang delinquency and reducing gang membership through face-to- face contact with gang members, for example. * Opportunity-providing program seek to rechannel gang activities into more constructive pursuits and include job training, employment, and education program. Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 19 However, there is little to indicate that suppressive gang interventions are successful in consistently reducing the illegal and violent behavior of youth gangs or gang members. For example, in January 1997 the Nicaraguan national police began a wide-ranging suppressive campaign in the barrios and asentamientos of Managua known to be most affected by youth gang violence. Armed with automatic weapons, the police repeatedly conducted large-scale night raids in these gang-affected neighborhoods and arrested any youth they came across in the streets. At the same time, the standard fine for acts of vagrancy was doubled. Although initially successful in reducing pandilla activity, this decrease in youth gang violence was only temporary, lasting only a month and a half. The cost of mounting such raids meant that they could not be sustained within the constraints of the national police's budget. Furthermore, pandilleros began to plan strategically to avoid being caught, developing early-warning systems of lookouts and adapting their behavior to police night raids, for example. In many cases, violence in these neighborhoods worsened as a result of this suppressive campaign; police incursions and patrols gave rise to frequent armed clashes with youth gangs, increasing the risk of injury or death for neighborhood inhabitants (Rodgers 1998 and 1999). Community Organization Program Community organization programs can take a number of forms. All, however, take as a basic premise that youth gang violence is at least partly a result of social disorganization, and that if local-level social institutions or networks were created or strengthened within these communities, youth gangs would become less of a problem. Such a conceptual approach is also implicit within initiatives that aim to build social capital within violence-affected communities as a means of mitigating the effects of this violence (Moser and Holland 1997, pp. 41-43). Probably the best known U.S. community-organizing gang intervention program is the Chicago Area Project. Founded in the 1930s, the project operated until the mid-1970s. Its main thrust was to promote the active involvement of the community through the creation of autonomous local welfare groups, and provide them with the means to implement their own policies to reduce youth gang activity and violence (Empey 1982; Finestone 1976; Jensen and Rojek 1980; Kobrin 1969; Schlossman and Sedlak 1983). Another important initiative was the 1954-57 Boston Mid-City Project. Its primary objective was to inhibit or reduce the presence of youth gangs and delinquency in targeted communities, through a "total community philosophy." This philosophy involved two processes. First, local citizen groups were encouraged and strengthened, and their direct and autonomous involvement in the anti-gang project was promoted. In particular, a major effort was made to obtain the cooperation of organizations and individuals with a direct relationship to youth gangs and their members, such as churches or schools. Second, trained social workers led these groups in change efforts with the youth gangs. Social workers acted as mediators with other community organizations, such as sports clubs, to facilitate the transfornation of youth gangs into a football or baseball club (Miller 1962). Neither project was successful in reducing youth gang activity and violence. In Chicago, local welfare groups tended to collapse very rapidly, and although the Boston Mid-City Project raised the awareness of the local population and promoted community empowerment, it had little success in positively reaching the youth gangs. Part of the reason for the failure of these two projects is that for a community organization to work, a certain prior level of social capital is necessary within the community. Although there was sufficient social capital in Boston, there was not in Chicago. One evaluation of the Chicago Area Project argued that the level of social disorganization in the targeted communities had 20 LCR Sustainable Development Working Paper No. 4 reached nonremedial proportions (Finestone 1976). In many gang-affected communities in Latin American and the Caribbean-as well as many non-gang-affected communities in the region- networks of trust and cooperation have totally collapsed and there is little community-level cohesion. In some cases, such as in urban Nicaragua, the local youth gangs represent the only form of community organization (Rodgers 1999). The failure to engage youth gangs directly is another reason for the disappointing project results. Even if gang members are linked to communities in other ways than through the gang, it is not enough to attempt to organize gang-affected communities (Lundman 1984). The nature of youth gang links with the community is also an important factor. In the Latin American and Caribbean context, some youth gangs, such as Brazilian quadrilhas or Colombian sicarios, have very weak links to the communities in which they operate; consequently, community organization is unlikely to have a significant effect. In the cases of the Nicaraguan pandillas or the Brazilian galeras cariocas, which have strong links to their local communities, community organization could have a positive effect. Perhaps one of the reasons for the disappearance of Nicaraguan youth gangs during the first half of the 1980s was the extensive local community organization networks that developed under the Sandinista regime. They began to unravel by the mid- 1980s, which coincides with the beginning of a new wave of pandillerismo. Social Intervention Programs Social intervention projects include outreach and counseling program that are aimed at preventing youth gang delinquency and reducing gang membership. These programs are predicated on the development of a rapport between social workers and youth gang members through repeated contact, often in areas that are familiar to gang members, such as schools. Part of the logic of such programs is to expose gang members to mainstream values and encourage them to adopt such values. Generally, however, these programs have limited success (Covey, Menard, and Franzese 1992), partly because the rationale for youth gangs is to be outside the mainstream, and because gang members are already exposed to mainstream values and culture, as demonstrated by the maturing out of gang members. Moreover, schools are often not considered to be "neutral" ground by gang members but rather are associated with authority. In a country like Nicaragua, for example, where few youth gang members go to school, and most of the country's youth have dropped out, such program have even less chance of success, according to a 1995 study (La Tribuna, June 26, 1997, p. 4A). A number of social intervention project evaluations have reported that an unexpected effect of such programs could be to increase gang cohesion, and thus actually help perpetrate the youth gang phenomenon (Klein 1967; Short and Strodtbeck 1965; Yablonsky 1959). Opportunity-Providing Programs Programs that focus on providing opportunities to gang members are probably the most effective type of youth gang intervention strategy. Underpinning this approach is the idea that gang members are marginalized or socially excluded, and thus lack opportunities to integrate into mainstream society. The advantage of such a strategy is that it tries to resolve some of the general causes of youth gang involvement and violence, which are often related to status, identity, and purpose. Both suppression and community organization are more focused on preventing the activities of youth gangs, while social intervention programs are generally passive. Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 21 One successful U.S. example of an opportunity-providing program is the 1967-69 Urban Leadership Training Program. It focused on co-opting youth gang leadership, by making individual gang leaders follow a study program designed to help them network and integrate into mainstream society. The aim was to reform delinquent youth into future leaders of their communities, and allow them to use their gang networks to establish a political base, thereby influencing other gang members. Although the initial project indicators seemed positive, funding ran out before the project could really progress; political intervention motivated partly by the need to guarantee gang leaders stable employment at the end of the study in exchange for their cooperation, caused the project's demise (Krisberg 1974). Within the Latin American and Caribbean context, a similar program seems to have been successful in Medellin, Colombia. Both state and private institutions in the city have offered long-term employment programs and social mobility opportunities to gang members, with the understanding that they would cease their gang involvement. In particular, diverse mechanisms have been developed through which youth gang leaders have been encouraged to involve themselves in local community politics and development, using gang structures as a means of action. In some cases, for example, rival youth gangs have been persuaded to sign peace pacts and work together on local social-development projects (Salazar 1994, p. 28). However, such programs have to be carefully implemented and monitored. The 1960s U.S. "Grants to Gangs" program, which allocated financial grants to gangs towards community development projects of their devising, is a spectacular example of the problems of a poorly controlled project. As the name of the project suggests, it involved making funds available to gangs to allow them to set up and manage community development programs of their choice. Some of the U.S. "supergangs," such as the Black P. Stone Nation and the Disciples, received substantial grants amounting to several hundreds of thousands of dollars, but with few positive results. The program was poorly supervised and monitored, and consequently there were numerous instances of fraud, corruption, and misappropriation. The project was also poorly conceived; in many cases gang leaders lacked the organizational and political skills to manage community development programs, and the project provided little technical training to gang members (Short 1976). Another frequent failing of opportunity-providing programs is their rigid and limited nature. Such projects must be as holistic as possible, providing not only funding but also education and training that take into account gang dynamics and local conditions. A Nicaraguan project run by the local nongovernmental organization HABITAT in Managua demonstrates such a holistic approach. It attempted to provide youth gang members with training, education, and financial aid. Individual gang members were contacted through prominent members of gang-affected communities and asked whether they wanted to participate in the project. In exchange for formally and publicly giving up the gang life-style, they received technical training and education, generally as carpenters or masons. At the end of their training, they were granted a loan to set up a small business. If a project beneficiary reverted to gang life, the loan arrangement was canceled and the individual was required to return the money. The project had limited success. There was a high dropout rate during training and education, and most participants failed to establish stable businesses after training. A preliminary project evaluation found two main reasons for these problems. First, by dealing with pandilleros as individuals rather than as part of a group, the program failed to recognize the collective nature of Nicaraguan pandillerismo and the youth gangs' association with their local communities, both of which are extremely important 22 LCR Sustainable Development Working Paper No. 4 motivations for the development of the gangs in the Nicaraguan context of social atomization and community collapse. Participating individuals felt that they had betrayed their community by being "egotistical" and thinking only of themselves. Furthermore, the project isolated individuals and did not provide a form of alternative group socialization. Second, in the context of the 60 percent combined unemployment and underemployment rate of the mid-1990s and economic stagnation, the low-level skills the project provided were insufficient for success, and so was the one-time small loal structure (interview with HABITAT director, Dr. Eva Altamirano, Managua, Nicaragua, March 14, 1997). In Nicaragua a successful opportunity-providing project would probably require the creation of a collective occupation for a gang within the local community. Such an occupation would need to benefit the whole community and provide useful skills to gang members once they "mature out." Community benefits are particularly important given the strong links that Nicaraguan pandillas have with their communities, and any successful opportunity-providing project in Nicaragua would need to respond to this aspect of pandillerismo. Furthermore, the positive nature of the links between the Nicaraguan pandillas and their local communities means that opportunity-providing projects can be articulated around activities that benefit the community more broadly than by simply eliminating gang violence, and thus can be included in a framework of larger social development initiatives. If carefully thought out, implemented, and monitored, opportunity-providing interventions probably stand the greatest chance of reducing youth gang violence, because they address some of the causes of youth gangs, rather than simply trying to hinder the gangs' existence, and they deal directly with the gangs. 7. CONCLUSIONS This literature survey makes apparent the crucial need for further research on youth gangs and violence in Latin America and the Caribbean. Information about the general youth gang situation in the region, as well as accurate and current information about the specific youth gang dynamics in each country of the region is necessary. Nevertheless, the data show that youth gangs in the region display considerable variation in their structure, logic, dynamics, and activities, including the degree to which they are violent. As described earlier, five distinct elements stand out from the available information on youth gangs: * Poverty-the majority of youth gangs emerge in the context of urban poverty. * Particularism-specific local conditions and contexts play an important role in shaping youth gangs and their violence. * Community links-youth gangs have varying links to their local communities; some gangs display very strong associations with their community, while others do not. The nature of the links a youth gang has with its local community affects the degree to which its violence is directed against the inhabitants of this community. * Drugs-there is weak evidence to suggest a link between the degree to which a youth gang is violent and its connection to drug-trafficking. * Migration--there is a connection between international migration from the United States, youth gang formation, and the degree to which a gang is violent. Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 23 Because of the paucity of data on youth gangs in the region an analysis of these elements is difficult, and their relationship to policy prescription and program design remains difficult to establish. The review of different types of youth gang policy interventions and approaches identified four kinds of initiatives. The first two-suppression and community organization-are preventive. The last two approaches-social intervention and opportunity-providing-are rehabilitative or corrective and seek to affect youth gangs and their members directly. Opportunity-providing projects need to carefully consider local conditions and gang-specific dynamics, and be implemented rigorously. The nature of a youth gang's links to the local community are important to consider in the planning of opportunity-providing projects. If the links are strong, youth gangs can be harnessed to initiatives that go beyond simply providing their members with opportunities and reducing the levels of violence in local communities, by creating stocks of social capital within the communities, empowering the inhabitants, and contributing to a wider social development process. References Auyero, J. 1992. "Juventud popular urbana y nuevo clima cultural: Una aproximaci6n." Nueva Sociedad 117(January-February): 131.4S. Bloch, H. A., and A. Niederhoffer. 1958. The Gang: A Study in Adolescent Behavior. New York: Philosophical Library. Caldeira, T. P. R. 1996. "Crime and Individual Rights: Reframing the Question of Violence in Latin America." In E. Jelin and E. Hershberg, eds., Constructing Democracy: Human Rights, Citizenship, and Society in Latin America Boulder, Colo.: Westview Press: 197-211. Canepa, M. A., ed. 1993. Esquinas, Rincones, Pasadizos: Bosquejos sobre Juventud Peruana. Lima: GP/Instituto Bartolome de las Casas. Cavan, R. S., and J. T. Cavan. 1968. Delinquency and Crime: Cross-Cultural Perspectives, Philadelphia, PA: Lippinscott. Cloward, R. A., and L. E. Ohlin. 1960. Delinquency and Opportunity: A Theory of Delinquent Gangs. New York: Free Press. Cohen, A. K. 1955. Delinquent Boys: The Culture of the Gang Glencoe, IL: Free Press. Cohen, A. K. 1990. "Foreword and overview.," in C. R. Huff ed. Gangs in America. Newbury Park, CA: Sage. Covey, H. C., S. Menard, and R. J. Franzese. 1992. Juvenile Gangs, Springfield, IL: Charles C. Thomas Publisher. Cummings, L. L. 1994. "Fighting by the Rules: Women Street-Fighting in Chihuahua, Mexico.," Sex Roles. 30, (3/4) (February): 189-198. Curry, G. D., and I. A. Spergel. 1988. "Gang Homicide, Delinquency and Community.," Criminotogy. 26, (3): 381-405. DeFleur, L. B. 1970. Delinquency in Argentina: A Study of C6rdoba 's Youth., Pullman, WA: Washington State University. Empey, L. T. 1982. American Delinquency: Its Meaning and Constructio. Homewood, IL: Dorsey. Erickson, M. L., and G. F. Jensen. 1977. "Delinquency Is Still Group Behavior! Toward Revitalizing the Group Premise in the Sociology of Deviance." Journal of Criminal Law and Criminology. 68 (2): 262-273. Fagan, J. 1990. "Social Processes of Delinquency and Drug Use Among Urban Gangs." in C. R. Huff (ed. Gangs in America. Newbury Park, CA: Sage. Ferracuti, F., S. Dinitz, and E. Acosta de Brenes. 1975. Delinquents and Nondelinquents in the Puerto Rican Slum Culture., Columbus, OH: Ohio State University Press. Feixa, C. (1995. "Tribus urbanas' y 'chavos banda': Las culturas juveniles en Catalufna y Mexico.` Nueva Antropologia. 14 (47) (March): 71-93. Finestone, H. 1976. Victims of Change: Juvenile Delinquency in American Society. Westport, CT: Greenwood. 25 26 LCR Sustainable Development Working Paper No. 4 Fischer, C. 1975. "Toward a Subcultural Theory of Urbanism." American Journal of Sociology. (80): 1319-1330. Giordano, P. C. 1978. "Girls, Guys, and Gangs: The Changing Context of Female Delinquency." Journal of Criminal Law and Criminology. (69): 126-132. Granera Sacasa, A., and S. J. Cuarezma Teran. 1997. "Evoluci6n del Delito en Nicaragua (1980-1995." El Nuevo Diario Managua: Editorial UCA. Gunst, L. 1995. Born fi' Dead: A Journey through the Jamaican Posse Underworld New York: H. Holt. Hagedorn, J. M. 1988. People and Folks: Gangs, Crime and the Underclass in a Rustbelt City. Chicago: Lakeview Press. Harris, M. G. 1988. Cholas: Latino Girls and Gang., New York: AMS Press. Human Rights Watch/Americas. 1994. Generation Under Fire: Children and Violence in Colombia. New York: Human Rights Watch. Isbester, K. 1996. "Understanding State Disintegration: The Case of Nicaragua." The Journal of Social, Political and Economic Studies. 21 (4) (Winter):455-476. Jensen, G. F., and D. G. Rojek. 1980. Delinquency, Lexington, MA: D.C. Heath. Katz, J. 1988. Seductions of Crime. New York, NY: Basic Books. Klein, M. W. 1967. A Structural Approach to Gang Intervention. The Lincoln Heights Project. San Diego: Youth Studies Center. Klein, M. W. 1995. The American Street Gang: Its Nature, Prevalence, and Control. New York: Oxford University Press. Kobrin, S. 1969. "The Chicago Area Project: A 25 Year Assessment," in R. S. Cavan (ed), Readings in Juvenile Delinquency. New York: Lippinscott. Krisberg, B. 1974. "The Politics of Delinquency Prevention: The Urban Leadership Training Program." Social Policy. (5): 53-56. Kuasfiosky, S., and D. Szulik. 1996. "Desde los Margenes de la Juventud." in M. Margulis (ed), La Juventud es Mas que una Palabra: Ensayos sobre Cultura y Juventud. Buenos Aires: Editorial Biblos. La Tribuna. Managua, Nicaragua. 2 May 1997, page 5A. La Tribuna. Managua, Nicaragua. 26 June, 1997, page 4A. Levenson-Estrada, D., with N. M. Figueroa, and M. Y. Maldonado Castillo. 1988. On Their Own: A Preliminary Study of Youth Gangs in Guatemala City. Ciudad de Guatemala: Asociaci6n para el Avance de las Ciencias Sociales (AVANCSO). Lundman, R. J. 1984. Prevention and Control of Juvenile Delinquency. New York: Oxford University Press. Martin, G. 1996-97. "Violences Strat6giques et Violences Desorganisees dans la Region de Uraba en Colombie." Cultures & Conflicts. (24/25) (Winter-Spring): 181-221. Martinez, R. 1997. "Mexico's 'Little Americas'." Los Angeles Times. 17 August, MI and M6. Youth Gangs and Violence in Latin America and the Caribbean: A Literature Survey 27 Miller, W. B. 1958. "Lower Class Culture as a Generating Milieu of Gang Delinquency." Journal of Social Issues. 14.3): 5-19. Miller, W.,B. (1962. "The Impact of a 'Total Community' Delinquency Control Project." Social Problems. (10): 168-191. Miller, W. B. 1982. Crime by Youth Gangs and Youth Groups in the United States. (report prepared for the National Youth Gang Survey), Washington, DC: Office of Juvenile Justice and Delinquency Prevention. Moore, J. 1978. Homeboys: Gangs, Drugs, and Prison in the Barrios of Los Angeles. Philadelphia: Temple University Press. Moser, C., and J. Holland. 1997. Urban Poverty and Violence in Jamaica. World Bank Latin American and Caribbean Studies, Washington, DC: The World Bank. Nuinez, J. C. 1996. De la Ciudad al Barrio: Redes y Tejidos Urbanos - Guatemala, El Salvador y Nicaragua. Ciudad de Guatemala: Universidad Rafael Landivar/PROFASR. OPS (Organizaci6n Panamericana de la Salud). 1996. La Violencia en las Americas: La Pandemia Social del Siglo XX. Comunicaci6n para la Salud No. 1 0,.Washington, DC: Organisacion Panamericana de Salud. Ortiz, F. 1997. "Peace Attained, Murder Now Stalks Central America." Reuters Feature release, 10: 12 PM ET, 24 December. Page, J. Bryan. 1997. "Vulcans and Jutes: Cuban Fraternities and Their Disappearance," Free Inquiry in Creative Sociology, 25 (1) (May): 65-73. Portocarrero, B. 1996. "El Lumpen Pituco: Bad boys." QueHacer. (104): 70-73. Revista Centroamericana. 1994. "Violencia Juvenil en Centroamerica." Revista Centroamericana (2):4-9. Reyna, C. 1996. "Las Subculturas de la Violencia: El Miedo de Lima." QueHacer. (104): 58-62. Rodgers, D. 1997. "Un Antrop6logo-pandillero en un Barrio de Managua." Envio. 16 (184) (July): 10- 16. Rodgers, D. 1998. "Chaos or Order?: Youth Gangs and Violence in Urban Nicaragua." presentation given to the LCSES Governance, Social Capital and Violence. seminar series. World Bank, Washington, D.C., 14 January. Rodgers, D. 1999. Living in the Shadow of Death: Violence, Pandillas and Social Disintegration in Contemporary Urban Nicaragua. Ph.D. dissertation, Department of Social Anthropology, University of Cambridge, UK. Ruiz Secada, R. E. 1993. "Imagen del Lider en Adolescentes de Sectores Urbano-marginales." in M. A. Canepa (ed. Esquinas, Rincones, Pasadizos: Bosquejos sobre Juventud Peruana. Lima: GP/Instituto Bartolome de las Casas. Salas, L. 1979. Social Control and Deviance in Cuba New York: Praeger Publishers. Salazar, A. 1990. Born to Die in Medellin. translated by N. Caistor, London: Latin America Bureau. Salazar, A. 1994. "Young Assassins of the Drug Trade." NACLA, XXVII (6) (May-June): 24-28. Sanchez Jankowski, M. 1991. Islands in the Street: Gangs and American Urban Society. Berkeley, CA: University of California Press. 28 LCR Sustainable Development Working Paper No. 4 Sanchez Le6n, A., and M. de Mastro. 1993. En el Juego de la Vida. Ser Delincuente en Lima. Lima: DESCO. Schlossman, S., and M. Sedlak. 1983. "The Chicago Area Project Revisited.," Crime and Delinquency. (29): 398-462. Shaw, C. R., and H. D. McKay. 1942. Juvenile Delinquency and Urban Areas. Chicago: University of Chicago Press. Short Jr., J. F. 1968. Gang Delinquency and Delinquent Subcultures. New York: Harper and Row. Short Jr., J. F. (ed). 1976. Delinquency, Crime, and Society. Chicago: University of Chicago Press. Short Jr., J. F., and F. Strodtbeck. 1965. Group Process and Gang Delinquency. Chicago: University of Chicago Press. Small, G. 1995. Ruthless: The Global Rise of the Yardies. London: Warner Books. Suttles, G. D. 1968. The Social Order of the Slum: Ethnicity and Territory in the Inner City. Chicago: University of Chicago Press. Szulik, D., and S. Kuasnfosky. 1997. "Los Extrafios de Pelo Largo: Vida Cotidiana y Consumos Culturales." in M. Margulis and others, La Cultura de la Noche: La Vida Nocturna de los J6venes en Buenos Aires. 2nd edition, Buenos Aires: Editorial Biblos. Thrasher, F. 1927. The Gang: A Study of 1,313 Gangs in Chicago, Chicago: University of Chicago Press. Tuckman, J. 1996. "Criminals take up soldiers' weapons." Financial Times, (27 September): 3. Valenzuela, E. 1984. La Rebeli6n de los Jovenes: Un Estudio sobre Anomia Social. Santiago de Chile: Ediciones Sur. Velez-Ibanez, C. G. 1983. Rituals of Marginality: Politics, Process, and Cultural Change in Urban Central Mexico 1969-1974. Berkeley and Los Angeles: University of California Press. Vigil, J. D. 1988. Barrio Gangs. Street Life and Identity in Southern California, Mexican American Monograph Number 12, Austin, TX: University of Texas Press. Villalobos R., S. 1993. "Un Acercamiento a la Agresi6n de los J6venes." in M. A. Canepa (ed), Esquinas, Rincones, Pasadizos: Bosquejos sobre Juventud Peruana. Lima: GP/lnstituto Bartolom6 de las Casas. Whyte, W. F. 1943. Street Corner Society: The Structure of an Italian Slum. Chicago: University of Chicago Press. Yablonsky, L. 1959. "The delinquent gang as a near-group." Social Problems. (7): 108-117. Zaluar, A. 1994. Condominio do Diabo. Rio de Janeiro: Editora Revan/UFRJ. Zaluar, A. 1997. "Gangues, Galeras e Quadrilhas: Globaliza,cao, Juventude e Violencia." in H. Vianna (ed), Galeras Cariocas: Territ6rios de Conflictos e Encontros Culturais. Rio de Janeiro: Editora UFRJ. LCR Sustainable Development Working Papers For back issues contact: Environmentally and Socially Sustainable Development Sector Management Unit (LCSES) Latin America and Caribbean Regional Office The World Bank 1818 H Street, N.W. Washington, DC 20433 U.S.A. Tel. (202) 473-9379 Fax. (202) 676-9373 E-mail: pbrandrissgworldbank.org No. Title Date Author 1 Environmental Guidelines for Social Funds December 1998 Douglas J. Graham Kenneth M. Green Karla McEvoy 2 A Conceptual Framework for Violence Reduction August 1999 Caroline Moser Elizabeth Shrader 3 Youth Violence in Latin America and the August 1999 Caroline Moser Caribbean: Costs, Causes, and Interventions Bernice van Bronkhorst 4 Youth Gangs and Violence in Latin America August 1999 Dennis Rodgers and the Caribbean: A Literature Survey 5 Violence and Social Capital: Proceedings of the August 1999 Caroline Moser LCSES Seminar Series, 1997-98 Sarah Lister