pdf Download PDF

Enrique Desmond Arias - Faith in Our Neighbors: Networks and Social Order in Three Brazilian Favelas - Latin American Politics & Society 46:1 Latin American Politics & Society 46.1 (2004) 1-38

Faith in Our Neighbors:
Networks and Social Order in Three Brazilian Favelas

Enrique Desmond Arias

After nearly 20 years of democratization, residents of Rio's favelas suffer high levels of civil and human rights abuse at the hands of both police and drug traffickers. The government is generally unable to guarantee the political order necessary to protect the rights of residents in these communities. Existing theories of democratization and advocacy networks offer little to explain how the types of endemic violence that affect poor neighborhoods in the developing world can be brought under control. Based on more than two years of participant observation and interviews in Rio de Janeiro, this article examines how democratic order can be extended to favelas. It argues that networks can link favela residents to organizations in civil society, and state actors can play a critical role in reducing violence and establishing democratic order.

The military government that ruled Brazil from 1964 to 1985 committed extensive human rights violations. Despite a laudable transition to democratic rule, human rights abuse and violence against the populace has significantly increased over the past decade-and-a-half (Pereira 2000, 217; Paoli and Telles 1998, 64-65). This is nowhere clearer than in Rio de Janeiro's favelas (shantytowns), where years of neglect and broken promises have caused the Brazilian state to appear to lose control of these communities to gangs of highly organized drug traffickers who enforce order, provide social services, and adjudicate disputes. In October 1994, President Itamar Franco responded to the growing outcry about public safety by sending the Brazilian military into Rio de Janeiro's favelas to ensure order. This operation resulted in substantial bloodshed (Human Rights Watch Americas 1997; Leeds 1996, 75). Dramatic clashes between state forces and drug traffickers are still the norm.

Draconian state violence, however, has failed to bring order to Rio's favelas and has actually had the reverse effect of reinforcing criminal legitimacy, as residents suffer police abuse and lose faith in the state (Garotinho et al. 1998, 134-42). Drug traffickers maintain prominent roles in most poor communities, and their conflicts with police continue to make residents' lives difficult (Jornal do Brazil 1996; Zaluar 1998, 218-20). The recent histories of Peru and Colombia indicate that high levels of violence can be destabilizing to democratic governments [End Page 1] (Correa Sutil 1999, 266). How can democratic order, which is necessary to guarantee civil and human rights, be extended to Rio's favelas?

This paper recounts the struggles of the residents of three Rio favelas, Vigário Geral, Tubarão, and Santa Ana, to protect their rights and live their lives. 1 The empirical cases are presented in the historical context of Rio's poor communities. Existing theories of democratization are used to build an integrated model for establishing democratic order in the context of ongoing social violence. The model shows how social networks can help translate protest and governmental reform efforts into concrete political change when traditional strategies for political transformation are ineffectual.

Detailed local-level analysis offers a complex picture of the microlevel politics that play such an important role in controlling local conflict. The evidence shows that local civic groups and social movements, networking with state actors, play a critical role in stimulating police reform and controlling violence.

The History of Favelas in Rio de Janeiro

The Brazilian state has had a complex and contentious relationship with Rio's favelas. Settled illegally by ex-slaves and new urban migrants over the course of the last hundred years, favelas have received only sporadic improvements in basic urban services. State intervention, for the most part, has been limited to policing, repression, and clientelist vote seeking.

As a result, local leadership plays an important role in favelas' internal governance. During the late 1950s and early 1960s, residents' associations (associações de moradores, AMs) were formed in many Rio favelas, usually at the behest of outsiders trying to deliver resources to residents (Resende de Carvalho et al. 1998, 20-22; Perlman 1976, 39-40). In the 20 years of authoritarian rule that began in 1964, AMs worked to protect favelas from removal and provided government services in large areas of the city (Santos 1995, 158-64; Perlman 1976, 23-57). With the return of democracy in the 1980s, clientelist politicians again used contacts with AMs to gain access to votes. These efforts resulted in improved urban services in favelas, but weakened the AMs and their statewide interest group as the government met many favelas' most pressing infrastructural needs (Jornal do Brasil 1996; Zaluar 1985, 1998, 218-20; Gay 1994, 31, 40-41; Leeds 1996, 74).

In the 1980s, drug traffickers began to employ many favela residents in their operations, providing needed assistance to the poor and fortifying their political leadership. Politicians, seeing the AMs' growing fragility, worked more directly with traffickers to secure votes. During this period, AMs began acting as intermediaries between traffickers, residents, [End Page 2] and state officials. The consolidation of drug gangs' political power has provoked significant violence in Rio.

The relationship between Brazilian favelas and the state contributes to ongoing conflict (Pinheiro 1997, 270-71). Upper-level politicians, police, and bureaucrats often do not have effective control over lower-level state representatives, and as a result, police take bribes and engage in brutal repressive operations. Many bureaucrats, police, and politicians take kickbacks or otherwise work with traffickers to accomplish personal objectives. These actions strengthen drug traffickers, who, as a result, operate with relative freedom inside favelas. Traffickers provide economic aid, maintain order, and resolve conflicts in efforts to build their legitimacy (Junqueira and Rodrigues 1992).

Narcotics trafficking broke the already fragile political links between government and the poor (Resende de Carvalho 1998, 31). Today, for example, residents refuse to help the police but offer assistance to the criminals with whom they live on a day-to-day basis. Often tied to traffickers, AMs criticize police violence but publicly say little about criminal activity.

Rio's police provide almost no help in solving these problems. Formed in the nineteenth century mainly to repress popular groups, the police have historically maintained abusive relations with the poor, and today are known to be extremely corrupt (Holloway 1993, 288-91; Leeds 1996, 64). Rio has two major police forces: the military police (policia militar, PM) and the civil police (policia civil, PC). The PM wears uniforms and conducts street patrols. The PC wears civilian clothes and conducts investigations. The state governor commands both forces under the Secretaria de Segurança Pública. Though corruption is pervasive in the PM, the PC is generally considered to be more corrupt (Leeds 1996, 63-66). In both organizations, these problems result in poor administration and deployment of resources. Corrupt police fail to enforce the law; they inform traffickers of police activities and set up police operations in places where they will not interfere with trafficking.

The interaction between society and the state in Rio's favelas raises serious questions about local-level democratization. In poor and peripheral areas of major cities and some parts of the countryside, the Brazilian state does not exercise an effective monopoly on the means of organized violence (Pinheiro 1998, 13-14; Pereira 2000, 234-35). The fragmented and diffuse nature of the Brazilian state, characterized by corruption, official impunity, and lack of hierarchical controls, both emerges from and engenders a divided society. Corrupt police strengthen drug traffickers by taking bribes that allow the traffickers to operate openly in the favelas (Pessoa 2002, 5; Leeds 1996, 64-65). Honest police and politicians do not trust AMs or residents because of many AMs' relationships with traffickers. [End Page 3]

As Martha Huggins has argued in a broader context, public security has become privatized as independent agents have taken control of different elements of the state's repressive apparatus and as heavier weapons have come into civilian hands (Huggins 1998, 202-4). With many police beholden to the highest bidder, neither the state nor the federal government can guarantee human rights. The poor, unable to hire their own security guards, have to rely on criminals for protection. Corrupt state officials work with locally empowered delinquents to enrich themselves and win votes; criminals engage in conflicts with one another; out of fear, the population calls for increasing police repression; and violence spins out of control (Pinheiro 1997, 264-70; Caldeira 1996, 208-9).

Theoretical Building Blocks

These conditions are by no means unique to Rio de Janeiro. The persistence of localized authoritarian enclaves characterized by state impunity and criminality poses serious challenges to democratic consolidation in many countries. Neoliberal reforms and a difficult transition from authoritarian rule have limited poor peoples' access to the state and have exacerbated conditions of violence (Yashar 1996, 99; Chalmers et al. 1997, 552-53; Fernandes 1994, 104-7). Crises of confidence in democratic regimes' abilities to control everyday violence and corruption have resulted in the emergence of semiauthoritarian democracies in many parts of Latin America (Correa Sutil 1999, 266). How does a democratizing state guarantee order in crime-ridden neighborhoods without resorting to draconian force?

Institutional reform plays an important role in any effort to control continuing social violence. Guillermo O'Donnell notes that democratic governance depends on states' efforts to extend the rule of law and basic protections to excluded groups (O'Donnell 1999, 322-23, 325). Under the prevailing conditions in Latin America, many scholars argue, political elites must reform judicial institutions in order to treat citizens more equally and to hold police accountable under civilian laws (Méndez 1999, 221-26; Zaverucha 1999, 72; Correa Sutil 1999, 255-71). To end ongoing abuse of poor and minority populations, furthermore, these same leaders must also reform the police by "inculcating" democratic values and ending corruption (Pereira 2000, 234-35). This will help not only in limiting police violence but also in decreasing criminal violence as police, responding to reform efforts, begin to effectively enforce the law.

Ending state corruption, however, is only part of the battle. Government efforts politically to incorporate the poor also play an important role in extending the rule of law. Latin American governments must also worry about the role of criminals in providing informal security and [End Page 4] services to groups at the margins of the economic system. Some scholars argue that by offering basic welfare guarantees and by recognizing irregular urban land tenure, states can more effectively intergrate these portions of the population into the democratic system and reduce their dependence on illegal actors (Becker 1999, 144-46; De Soto 2001, 19). 2

While these reforms make some sense from a theoretical perspective, there little reason to believe that governments will implement these suggestions. Why, for example, would government agents initiate reforms if the principal beneficiaries of those changes have little presence or representation in the state? Even if the government actually does begin reforms, how are those policies implemented when the responsible state agents' interests may run counter to the policy? (For a discussion of this problem, see Migdal 1994, 16-17.)

The literature on civil society and on social movements provides some answers. Writers on social movements argue that mobilization and protest is one avenue for political change in exclusionary democracies. Historically excluded groups in Brazil have demanded political change through movements. 3 By taking to the streets, excluded populations can put pressure on the government to alter policies and behaviors (Cardoso 1992, 291). Writers on this subject have begun to investigate the complex role of movements not just in pressing the state to change policy but also in reforming authoritarian and repressive practices in society (Alvarez et al. 1998, 18-19).

While the study of social movements goes a long way toward explaining how marginal groups can engender political change, the approach is more limited in explaining the role of civil society in ongoing political activity. Social movements tend to operate in mobilization cycles and deactivate after they achieve a few major objectives, factionalization emerges, groups are coopted, or activists tire (Tarrow 1998, 147-50). In these periods between marches and public clashes, civic institutions play an important role as a bridge between citizens and the state. Representative groups work with state officials to help implement policies by mediating relations between government and the population (Evans 1995, 50, 72, 228; Kurtz 1999, 295; Wang 1999, 232, 237-46; Migdal 1994). These ongoing contacts help build public accountability and ensure effective governance.

But how can these societal efforts to create reform work in extremely violent communities? Under dictatorships, civic actors often became targets of government repression. Today, social leaders working to control violence can become the targets of corrupt officials or criminals (Leeds 1996, 69-73). Furthermore, if some state officials are more closely attached to violent social actors than to other officials concerned with the rule of law (and vice versa), it makes little sense to operationalize politics along a single state-society dimension. [End Page 5]

The model of a network suggests an alternative, flexible framework to explain local-level political organizing. It transcends conventional understanding of state-society relations by allowing for participation by nontraditional political actors, such as international organizations and the media; it explains how social groups can remain active even under the threat of violence; and it offers a complex model of the state that reflects the internal divisions common in Latin American politics today.

Networks are horizontal, informal organizations based on connections between actors with similar interests (Chalmers et al. 1997, 567-68; Keck and Sikkink 1998, 8). Member groups use connections to help build coalitions that transcend state-society and domestic-international boundaries and exploit the unique resources and skills of different member groups to accomplish common goals. Networks thereby provide a description of how different organizations relate to one another. Network analysis helps to explain the specific dynamics of how politically interested actors build cross-institutional connections and exploit those connections for political gain (Keck and Sikkink 1998, 10-38).

Through a network, groups with similar objectives can share labor and risk among themselves. Networks help member groups to accomplish complex objectives under difficult circumstances by promoting specialization among members, linking like-minded groups, and allowing those that want to change political conditions in violent places to share risks in such a way that it is more difficult to stop efforts to promote change (from a business perspective see Kanter 1991, 68; Uzzi 1997, 70-71; Powell 1994, 303-4).

Contemporary scholarship on social networks suggests that this framework can provide channels for communication, collective action, and representation under conditions in which traditional forms of political activism are ineffectual. One way networks do this is by bringing together a set of functionally and spatially diffuse actors. For example, significant evidence shows that transnational networks that bring together nongovernmental and intergovernmental organizations and governments have had success in controlling human rights abuse in places where authoritarian regimes made domestic political opposition difficult or impossible. By deflecting attention from particular actors exposed to violence and using alternative channels of communication and representation, network connections helped to achieve political change under difficult circumstances, such as the brutal Latin American dictatorships of the 1960s and 1970s (Keck and Sikkink 1998, 75-120; Risse and Sikkink 1999, 17-35).

This analysis also applies to local-level networks dealing with social violence. In bringing multiple groups together, networks raise the cost to violent actors of silencing individual groups. Networks allow more protected actors to take riskier roles in controlling violence and more [End Page 6] exposed actors to take less public but equally important roles in those efforts.

The activities of networks also make state leaders aware of community problems so that they can change existing policies or discipline negligent officials. Local groups have access to important information about police behavior and state activities that outsiders may not have. By gathering data and disseminating it through contacts in the media and civil society, local groups can help bring information to, and put pressure on, state officials. As state leaders become more aware of the activities of criminals and local-level bureaucrats, they can more effectively monitor and implement policies.

A third role of networks is to facilitate efforts to build norms of social and political behavior. Through ongoing social activities, civic groups can help construct local norms that may constrain violent actors (on movements and society see Alvarez et al. 1998, 18; Putnam 1993, 171-73). Networks, furthermore, allow groups with different and often contradictory strategies to work together to achieve similar objectives (on coordination between network members see Sikkink 1996; for an example of these tensions see Hochstetler 2000, 169). A group engaged in open protest thereby can operate in conjunction with a group that puts more weight on negotiating accommodations with the government.

Through informal networks, different local organizations can interact with outside NGOs, residents, and the state to seek outside assistance, develop innovative solutions to problems, and change state policy. As a result of these factors, network links restructure the relationship between state and society to extend the rule of law to favelas.

While the model developed in this paper draws most distinctly from network theory, it also reflects the insights of the other literature discussed here. As the data from the three cases show, a coherent state organization, capable of making reforms and ensuring that lower-level officials follow the law, is critical in controlling local-level violence. Networks help empower state actors concerned with the rule of law to undertake these vital reforms. Dense, tightly linked social networks, indeed, play an important role in the proper functioning of any democratic state (Putnam 1993, 181-89).

Networks also contain various types of social actors, including social movements. Spontaneous and organized social protest can provide an important impetus in forming social networks and causing state reform. Protest is a way of strengthening latent social contacts and activating groups around new issues. By distributing risks, networks can help to sustain state reform and social mobilization in the face of high levels of violence.

Civil society, social movements, and networks, however, are not always positive or politically progressive institutions. Significant writing [End Page 7] on Brazil, Latin America, and the developing world has suggested that many civic institutions and networks engage in undemocratic behavior and promote violence. Leigh Payne has shown that some civic actors are ambivalent about democratic government (Payne 2000, 3-5). Orin Starn has pointed out that the rondas campesinas in Peru, while reflecting residents' concerns, are not necessarily either democratic or progressive, and often engage in violent behavior (Starn 1999). Mark Duffield, Mary Kaldor, and Phil Williams each suggest that criminal networks linked to states and other important entitites promote international conflicts and crime (Duffield 2001, 13-15; Kaldor 1999, 91-111; Williams 2001, 73-77).

In the case of Rio's favelas, Robert Gay and others have pointed out that many AMs have links with drug traffickers (Gay 1994, 12, 97-98; Alvito 2001, 153-54; Leeds 1996, 70-73). This study argues that such links undermine efforts to control violence. These groups will often try to chip away at efforts to promote the rule of law, and they pose serious challenges to progressive political networks. Ultimately, networks themselves have biases, and member groups do not always make effective political decisions.

In Rio's favelas, where criminal groups are strongly connected to some civic leaders and government agents, networks provide the most effective way to extend democratic governance and protect citizens' rights. Only by building connections between residents and organizations outside the favelas can groups with resources to help reduce violence effectively penetrate neighborhoods dominated by criminal organizations. The specific, flexible structures of networks allow groups concerned with controlling conflict to work together to build the confidence of fearful and isolated populations, to undertake significant political reforms, and to decrease residents' reliance on traffickers.

The Role of Social Networks in Controlling Violence

What types of activities help protect the citizenship rights of favela residents? This section examines three Rio favelas that experienced conflict in the 1990s. While they initially responded to particular crises in similar ways, the internal political structure of each favela and its links with outside groups had important effects on political outcomes. In communities where residents effectively networked together and made contacts with outsiders concerned with human rights, homicide and violence levels declined. In favelas where criminals dominated local politics through connections to community leaders and police, violence remained high. The data in this section show how the internal and external political dynamics of these favelas affected levels of violence after tragic events. [End Page 8]

The first community, Santa Ana, is dominated by drug traffickers and has a small number of internal organizations that maintain limited relations with outside groups and the state. During one three-month period in 1997, 28 people were murdered in and around the favela. Despite one NGO's efforts, little has changed here, because a powerful criminal organization has coopted much of the local leadership. The second community, Vigário Geral, has a higher number of internal organizations with links to outside actors. These groups succeeded in bringing violence under control for several years in the mid- and late 1990s. During a nine-month period in 1997 and 1998, residents reported only one murder.

The third community, Tubarão, has both a powerful drug-trafficking presence and a number of active civic organizations. Between August 1998 and May 1999, police and traffickers murdered approximately 30 residents. After the police killed five residents in 2000, however, police working with local groups set up an innovative community policing program that contributed to a rapid drop in violence and the elimination of local police homicides in the period between September 2000 and August 2001.

Santa Ana: Persistent Violence and Network Failure

Caught in the all-too-common crossfire between the police and drug traffickers, Santa Ana is known as one of Rio's most violent areas. The community is full of signs of conflict, ranging from damaged walls and satellite dishes to scars on the bodies of young men and tired faces of residents who have difficulty sleeping through the nightly gunfights. During one three-month stretch in 1997 there were 28 gunfights, 27 murders, and 14 occasions when residents denounced police abuse ranging from unwarranted searches to threats, beatings, and killings (Joselino 1997).

With about four thousand residents, Santa Ana sits on a hill just west of downtown Rio that it shares with several other favelas (Josias and Antônio 1996; Josias and Manoel 1997). Access to Santa Ana is available from a city-maintained road. About three-quarters of the way up the hill, Santa Ana gives way to a larger favela that contains a small police post, a school, and some commerce.

Santa Ana's residents build their homes with brick and cement. Some own cars, and the neighborhood skyline shows an ample number of satellite dishes. Many private businesses operate in the community; yet the favela has few paved roads and a number of open sewers. Traffickers and their allies control most of the public space. The central building is the headquarters of the AM, which is run by the father of a [End Page 9] trafficker. It contains a large, semi-enclosed room used for public functions, an office, and storage space. To obtain votes, politicians have funneled money through the AM and the traffickers to build a covered dance floor, to which the girlfriend of a powerful trafficker holds the keys (Joselino 2002).

The only other semipublic buildings are a chapel and daycare center, or crèche, maintained by the Catholic Church and a social club maintained by an NGO. Two small stages on the main street provide space for mounting loudspeakers and for addicts to take cocaine. Just up the main street from the AM is the traffickers' boca de fumo (a term used for point of sale and drug headquarters). Drug buyers, residents trying to get financial assistance, and police demanding bribes all frequent the boca (Josias and unknown woman, 1997).

Most of the violence that affects Santa Ana results from ongoing low-level conflicts between drug traffickers and police. Both overzealous law enforcement and police extortion contribute to this violence. In the winter months of 1997, the police released several traffickers after receiving payoffs (Joselino 2001).

Traffickers who maintain connections with police and politicians dominate Santa Ana's politics. To build their legitimacy, these criminals provide social and economic assistance to residents. To increase the security of their trafficking operations, drug dealers pay bribes to police stationed nearby. The AM acts as an intermediary between traffickers, residents, and state officials. Josias, the AM president, has helped traffickers by providing them access to the AM for social events, mediating relations with politicians, and helping with communications during police operations.

The Catholic-run crèche also provides important local services. In 1997 the crèche leader agreed to help the traffickers throw a community party. A social club also operates in Santa Ana; it works with an outside NGO to provide guidance to at-risk adolescents. The club is the only local organization that eschews contact with traffickers. Its leaders, however, are virtually marginalized because of their affiliation with outsiders and, probably, their efforts to control violence.

Although the Catholic Church provides some services to residents, its leaders put little time into addressing questions of violence. Many residents worship in Evangelical Protestant churches. Some of the AM leaders are affiliated with one Evangelical church that holds services in the AM headquarters. AM leaders actively marginalize social club leaders from the community, complaining about their activities and critiquing their ties to outsiders.

Because of their connections, local organizations linked to traffickers have only limited contacts with society outside the favela and do not work to diversify those contacts. Camila, the director of the crèche, [End Page 10] noted that once she had participated in a group that wanted to put a better public face on Santa Ana by throwing some events and bringing in the media to show that Santa Ana had positive activities for children and adults. Traffickers, however, vetoed her efforts for fear of attracting too much attention. AM leaders have contacts with some state officials, but these are either weak, low-level bureaucrats with whom AM leaders file paperwork, or corrupt politicians who have worked with traffickers.

Crisis and Response

On July 19, 1998, a PM group conducted an operation in Santa Ana to shake down traffickers. As police entered the area, a group of criminals fled up the hill. The police opened fire, striking a day laborer in the head as he was drinking a soda. Angry residents quickly came out and surrounded the police; the police fired bullets in the air but failed to disperse the crowd. Fearing that more violence would break out, one resident, an Evangelical hardware store owner, helped get the body into the ambulance and persuaded the police to withdraw as the vehicle left. The victim died before reaching the hospital.

After the police retreated, residents prepared to riot. The director of the social club, however, a resident and activist trained both by NGOs and in Catholic comunidades eclesiásticos de base, persuaded residents not to riot and organized a group to march peacefully down the hill and block the street below. A major newspaper carried the story. At the residents' behest, the traffickers moved their boca far from the favela's entrance.

For a short time, violence decreased dramatically, as both police and traffickers avoided public confrontations (Sandrão 1998; Camila 1998). Eventually, however, without a strong institutional framework to continue pressuring the state, violence returned to high levels. This outcome differed from that in a nearby favela where, when police murdered a resident, other residents rioted and burned a police booth. Had it not been for the leader of the social club, with her extensive activist training, Santa Ana residents might also have rioted.

Press reports about the protest contributed to pressure on the police to reduce killings. Residents noted that after newspapers reported on killings, violence tended to decline for a short time. This was the case when stray bullets killed several people in Santa Ana and neighboring communities in March 1997 and when the protest occurred in 1998.

This account illustrates how residents' discontent and public attention, when effectively channeled, constrain both police and trafficker behavior. Both sets of actors made efforts to reduce the potential fallout of their conflicts. This story also shows how effective collective action can bring positive press attention. By organizing and walking to the street, residents attracted media interest and forced the police to change [End Page 11] their behavior over the short term. We also see here how connected and well-trained local leadership can play an important role in defusing a violent situation. The director of the social club helped refocus residents' energies away from attacking the police and toward protest.

Historically, favela residents often march or engage in spontaneous violence when upset with the government. What is interesting and different about this case, from the perspective of networks, is that while the community was prepared to riot, a resident with extensive outside training and contacts persuaded other residents to adopt a political posture that was more likely to garner positive press attention. As a result of her contacts with the outside and significant leadership training, the social club leader had the knowledge and skills to persuade residents to peacefully protest rather than riot.

Without a dense network concerned with human rights, however, locals could not sustain efforts to reduce violence. With only the social club directly committed to controlling conflict, a local critical mass did not exist to maintain regular pressure on police or criminals. Leaders of the social club engage in no collaborative activities with either the AM or the crèche. The social club has no internal support for its activities to control violence and, as a result, directly and regularly reaches only about 40 children and adolescents. While the organizers of the social club have leadership training, their limited support in Santa Ana prevents them from having a long-term impact on local violence. The AM, which is closely allied with the traffickers, maintains contacts with state officials, but usually these officials are corrupt, and have little interest in controlling violence. The AM is openly critical of the contacts that the social club maintains with outside groups.

None of the groups in the community, moreover, have regular contact with the press. As a result, the press visits only during periods of high violence, and residents have little influence with reporters, who often depend on the police for information. Ultimately, the focus of local organizing around the AM and traffickers prevents the community from building the types of organizations, programs, and contacts needed to control violence. At the same time, evidence shows that when residents have contact with outsiders, such as with the social club leaders, they have some success in reducing bloodshed.

Vigario Geral: Local Organizing and Violence Control

The Parque Proletário de Vigário Geral sits on the northern edge of Rio. The community is bordered on the north and east by the São João de Meriti River, on the west by a wall that runs next to a railroad, and on the south by its neighbor (and longtime enemy) Parada de Lucas, another [End Page 12] shantytown. In 1997 Vigário's approximately ten thousand residents could reach the favela either by a dangerous road that passed through Parada or by foot over a small bridge that crossed the railroad tracks.

Crisis and Response

On the August night in 1993 after Brazil's national soccer team won a qualifying game for the next World Cup, a group of Rio de Janeiro's PM invaded Vigário and killed 21 residents—some of whom were in the streets quietly celebrating Brazil's victory—in alleged retaliation for the murder of several police officers the night before (Leeds 1996, 65-66). The police were involved in a corruption scheme that allowed the local traffickers to operate on the condition that they regularly pay bribes. Tired of paying kickbacks, the traffickers attacked the police. The day after the massacre, the brutal killings made headlines throughout Brazil (Ventura 1994, 66-68).

In the days and weeks that followed, a group of well-educated young residents and survivors organized a movement called Mocovige (Movimento Comunitário de Vigário Geral) to protest against violence. Mocovige quickly made links to outside NGOs and exploited these alliances to gain and maintain media attention. With the advice of an outside activist, Mocovige leaders organized a difficult but symbolically important 40-kilometer march from the Candelária Church in downtown Rio (where police had murdered street children several months earlier) to Vigário. This event, a very creative reinterpretation of traditional favela protest marches, received extensive press covereage and put local activism on the map.

Within months of the march, Mocovige, with the help of the prominent NGO Viva Rio and others, reorganized into an NGO called the Casa da Paz (CdP). Later, two other NGOs, the Médicos Sem Fronteiras (MSF, the European Doctors Without Borders) and the Fundação Afro-Reggae (FAR), set up projects with the CdP's help. Since 1993, a number of other groups have operated in the favela, including a work cooperative, an internationally supported crèche, a local community management group, and two environmental groups.

Solutions to Violence in Vigário

In the wake of the massacre, high levels of violence continued as a result of ongoing police retaliation. Between 1993 and 1996, the police had no permanent presence in Vigário and would enter the community to conduct actions against drug traffickers who operated openly in their absence. When the police did enter, shootouts ensued (Pedro and Wesley 1997; Almeida and unknown resident 1997). Both police and [End Page 13] traffickers, in such an unstable situation, believed that nearly all residents posed threats to them (Ferraz 1997b; Charles 1997b; Evanildo 1998). Stories abound of police violence ranging from unjustified detention to beatings and killings (Charles 1997a; Mateus 1998; Daniel and Joana 1998). During operations, police regularly conducted unwarranted searches and, in one case, mistakenly dropped grenades on a home they believed traffickers occupied (Charles 1997a; Mateus 1998; Jorginho and Eric 1998; Katarina 1998; Carlinha 1998). Traffickers, for their part, killed residents they thought had informed on them (Miguel 1997; Evanildo 1998).

This changed in October 1996, when Marcello Alencar, then governor of Rio, ordered police permanently to occupy Vigário. The PM presence rapidly reduced violence and suffocated trafficking. Addicts could no longer freely walk into the favela and openly buy and consume drugs; and traffickers could no longer overtly move around. Major traffickers left the favela, and more minor dealers took control of day-to-day operations (Unknown resident 1997). The police thus had little to gain from vigorously pursing the remaining traffickers, so shootouts decreased (Pedro and Wesley 1997; Almeida and two unknown residents 1997; Charles and outside filmmaker 1997). The reduction in the volume of drug sales also reduced the interest of outside traffickers in Vigário. As a result, between mid-1997 and early 1998, residents confirmed only one murder. Traffickers, who previously had patrolled the community with large weapons, did not appear armed in the streets, and residents consistently indicated that the community had dramatically improved (Almeida and women residents 1997; Cynthia 1997a; Luis 1997; Daniel 1998).

The state's decision to occupy Vigário was based largely on the high levels of sustained media attention in the years after the massacre. The massacre alone, however, does not explain ongoing media attention or policy changes after 1993. Many homicides occur in Rio each year; only a few of them attract public attention, and none retained attention as Vigario's did.

Unlike Santa Ana, Vigário benefited from well-connected, locally active NGOs, which skillfully attracted public attention to the situation. Only through the activities of networked NGOs were media regularly attracted to the community. The CdP kept the massacre trials in the headlines by engaging in such innovative activities as taking coffins to the courthouse during the trial of one of the police accused in the massacre, holding a candlelight vigil outside the courthouse, and informing the press when one of the prosecutors in the case toured the favela (1997). As in the march from Candelária, residents and outsiders often worked together to stage these activities.

The FAR's work with adolescents and children also gradually drew more public attention to the social and cultural life of Vigário. While the [End Page 14] CdP protested, moreover, the FAR avoided direct confrontation with the state. Vigário benefited from this combination of activities. According to one activist, a watershed moment occurred in mid-1996 when a resident unveiled a sculpture of a family made of bullet casings that local children had collected in the favela and exchanged at the CdP for candy. The work attracted both national and international attention, appearing in U.S. newspapers and on CNN. A week later, the police occupied Vigário. The active and innovative work of NGOs kept media and state attention on the favela by exploiting connections between residents and outsiders (Charles 1997c; CdP activists 1998).

While the police maintained low levels of crime, NGOs played an important role in reducing police violence. For one thing, the simple presence of NGOs limited police abuse. For example, the director of the MSF post is affiliated with an important INGO; in Brazil's hierarchical and unequal political system, he would have more credibility when testifying if he witnessed an act of brutality than would most residents. The MSF director, furthermore, has his own contacts in government. When the Merití River flooded in January 1998, he quickly brought public health officials to Vigário (Jaime 1997; Jaime and city officials, 1998a, 1998b). In a case of serious police violence, he could have drawn on those same contacts to bring officials to the favela more quickly than local leaders could.

NGOs also educate residents politically and provide venues to denounce rights violators. The work of the FAR is exemplary in this area. Through classes and conversations, participants learn about their rights in a democratic society. The FAR also has an agreement with another NGO to support and finance lawyers to denounce human rights abuses on a case-by-case basis (Roberto 1997b). If levels of violence increased, a structure would be in place that would allow residents to bring complaints quickly against abusive PMs (Visit of NGO workers 1997; Cynthia 1997b).

The NGOs' facility at attracting the media severely limits police violence. As in Santa Ana, and as Marcos Alvito has documented in Acarí, in most favelas the police dominate crime reports because reporters have few contacts with residents and are often afraid to venture alone into the favelas (Alvito 1998, 188). 4 In Vigário this is not the case. Reporters would often show up alone or in groups and always without a police escort. Journalists have developed relationships with local groups, and report on events using local perspectives. Under the gaze of a "connected" press, the police must respect the rights of residents (Visit of British Vice Foreign Minister 1997; CdP leadership and staff 1997; Jorginho 1997b; Charles 1997a).

The social activities of the NGOs and the AM also decrease the role of traffickers in providing assistance to residents. For example, the traffickers [End Page 15] historically had thrown a Christmas party for the favela's children, where they would distribute presents and food. With the decline of trafficking in 1996, it was apparent that no one would throw a similar party in 1997. Worried that residents would become nostalgic, the CdP gathered funding from outside groups to host a party (Ferraz 1997b; Cynthia and other women 1997; Cynthia and Tânia 1997; Jorginho 1997a; Cynthia and Felipe 1997; Daniel 1998).

This same pattern can also be seen with the distribution of water in the favela. Until the mid-1990s, the traffickers piped in the water and employed a plumber to ensure distribution. By 1997, without trafficker assistance, water pressure and quality had declined (Almeida, Cynthia, and Lorivaldo 1997; Almeida, Cynthia, and unknown resident 1997; Almeida and PROSANEAR administrator 1997). In response, the AM secured government help to improve water service (Almeida 1997; Almeida and PROSANEAR administrator 1997; Almeida and social scientist, 1997). The government was probably more receptive to the AM's demands because of both the attention the killings had already focused on the community and the ways that extended activism in the community had built links between local leaders and state officials.

The FAR made efforts to draw children and adolescents out of crime through cultural and educational work. FAR leaders would work to persuade adolescents at the edges of trafficking to participate in the group's activities. Slowly the FAR would envelop them in group social life and offer them jobs. Once a participant had become sufficiently immersed, FAR leaders would ask the youth to leave trafficking (Pedro and Wesley 1997; Roberto 1997a). The CdP, through technical classes and social programs, has offered young people alternatives in building their future (Ferraz 1997a).

The MSF played an important role in delivering medical and public health services to residents. In the context of Rio's favelas, this was especially important; one of the ways traffickers traditionally supported residents was to provide them with money to seek medical care. While the MSF pulled out of the community in 1999, it helped create the Community Management Movement (Movimento de Gestão Comunitário, MOGEC). Today this group provides medical and public health services to residents, researches local problems, and lobbies the government. By offering services traffickers had provided, network members decrease residents' support for criminals.

Religious groups also have a prominent role in Vigário, partly because many of the victims in the massacre were Evangelicals. A Protestant group from another neighborhood provided the funds for Mocovige to purchase the property that became the CdP. On occasion, Evangelical groups came through the community to provide medical services to residents. The Catholic Church has considerably less projection [End Page 16] in Vigário than it does in Santa Ana. While a chapel and a congregation are there, the church plays no political role in the community and church leaders do not attend local meetings.

These efforts are supported by contacts that each of the locally based groups maintains with outside organizations. Without backing from Viva Rio, Amnesty International, Evangelicals, and other groups, the CdP would not exist. The MSF, of course, drew on the international MSF's network of support. The FAR drew funding from the Ford Foundation, Cirque de Soleil, the British Council, and the EU. Both established a local presence through contacts with the CdP. One result of the presence of active NGOs and, in particular, the activities of the MSF was that a city administrator regularly attended local activists' meetings to respond to questions about the role of municipal government in the community.

The combination of the implementation of a constructive policing policy and the development of social programs reduced local dependence on traffickers. While the police presence increased the risks of being involved in trafficking, the NGOs increased the rewards for not being involved and restricted the activities of the police by giving voice to the community's needs.

The structure of the Vigário network differs considerably from that of Santa Ana. First, Vigário contains a large number of locally active groups connected to each other and to the outside. Each of these groups engages in extensive local activities and has relatively little contact with criminal elements. Because none of these groups are seriously compromised to traffickers, all are able to maintain and develop extensive outside networks. Whereas in Santa Ana most local civic groups are connected to traffickers and have limited outside contacts, in Vigário the combination of dense local interlinking and extensive outside contacts provided a basis for local program development, funding, and political activism.

In the pattern of their external contacts, the groups in Vigário have more in common with the social club in Santa Ana than they do with the other organizations in that community. Trafficking and some complications in the network, however, created difficulties in Vigário.

Continuing Violence

While groups active in Vigário succeeded in attracting public attention and controlling police, the network also encountered some problems. In general, local groups worked well together. The AM provided security at FAR events, and the leaders of the FAR, the AM, and the MSF sat on the board of MOGEC. This was not the case with the CdP, however. Until early 1998, the president of the CdP was Caio Ferraz, a university graduate and former resident who had taken a significant leadership role in Mocovige. Although Ferraz won Brazil's first human rights award, [End Page 17] many residents and activists did not like his aggressive, outward-looking leadership and the fame that he acquired (Daniel 1998).

In 1995, police officers whom Ferraz had publicly accused of taking bribes threatened him and forced him into exile. In his absence, an executive director was appointed to run the CdP. Errors made by the first executive director led to a significant budget deficit and resulted in the appointment of a new one (Carlinha 1998; Evanildo 1998; Lourenço 1998). The new administrator, a Protestant activist who had worked in several other favelas before coming to Vigário, arrived with an obligation to cover the budget deficit. His efforts to do that prevented the CdP from paying its employees. The new executive director also made residents unhappy by bringing in outside workers and not building close local contacts.

These tensions came to a head on January 20, 1998, when a group of community residents, leaders, and CdP workers gathered in the CdP for a conference call with Ferraz. By the end of the phonecall, Ferraz and the executive director both had agreed to step down. In the next several weeks, one of Rio's major newspapers published a prominent story on the conflict, as different groups wrangled over the CdP's leadership. The CdP failed to reorganize itself and significantly decreased its operations (Commission to Reorganize CdP 1998). Organized public pressure on issues of violence and the trials of those accused in the massacre decreased drastically. In November 1998, with little fanfare or comment, the Supremo Tribunal de Justiça, Brazil's highest court, acquitted 11 of the police accused in the massacre of all charges. Courts also released two police officers who had earlier been convicted. By mid-2000, Onda Azul, a growing NGO that recycles plastic bottles to make furniture, had taken over the CdP building (Sneed 2000).

Linkages in the Vigário network contributed to events that brought down Ferraz and ultimately the CdP. On the most basic level, dissatisfaction with the CdP stemmed from popular dissatisfaction with Ferraz. Despite his accomplishments, many residents thought he was smug and egotistical. A number of Mocovige activists said that he had assumed leadership of an organization that should have no leader. This created an alliance of Mocovige activists and massacre survivors against Ferraz. Mismanagement of funds under the first executive director worsened dissatisfaction with Ferraz. With that deficit, the CdP could not pay its employees. Local dissatisfaction with the work of the new executive director helped to build an alliance between disgruntled residents and unpaid workers. Prior contacts between the new executive director and a powerful trafficker from another favela in which he had previously worked worsened an already tense situation. Some people whispered that Vigário's traffickers wanted the second director out because of this relationship. He agreed to leave, but explained to residents why his relationship with the other trafficker should not have created a problem. [End Page 18]

This evidence reveals three important points about network organizing. First, not all networks or network components operate effectively all the time. Some network member group leaders make poor choices that cause their organizations to fall apart or that separate them from their constituencies. Second, members of networks do not always agree with each other and may compete among themselves. Under some circumstances, this may cause member groups to undermine each other. In this case, members of the Vigário network became upset with the CdP and allied against its leadership in the hope of reforming it. While this might have had the positive effect of improving the CdP and its relationship to the rest of the network, it had instead the negative effect of significantly reducing the CdP's activity, and also reducing pressure on the government to control violence.

Much of this outcome was caused by a breakdown of open communication among different groups in Vigário. The discussions that led to the breakdown of the CdP occurred behind closed doors. In the runup to the conference call when Ferraz resigned, one resident pulled this researcher aside and secretively told me about the efforts to remove the CdP leadership. Activists called in outside reporters and others only when Ferraz had already agreed to step down.

A final point about network organizing is that aggressive actors, such as criminals, may use connections into a network to undermine organizations antagonistic to their interests. In this case, traffickers used contacts with residents effectively to eliminate the CdP.

Although the policies of NGOs and the state have had great success in reducing violence in Vigário, some problems continue. Residents still resent the police in many ways (Luis 1997; Roger 1997; Miguel 1997). Traffickers continue to operate in the community, and many residents and outsiders remain extremely tense about their firepower. With the end of the CdP, violence began to increase again as pressure on the government declined and police corruption rose. In short, a decrease in outside connections and problems with internal connections resulted in a moderate increase in violence in Vigário in 1998 and 1999 (Eric 1998; Carlinha 1998). Organized efforts to control violence continued nevertheless, and levels of bloodshed did not rise dramatically immediately after the CdP's breakdown. The presence of other groups kept pressure on the state to minimize violence until at least 2000. Thus the existence of a network in Vigário allowed local activist efforts to survive the destruction of the CdP.

Tubarão: Top-Down Network Formation

Located in the wealthy Zona Sul, the Morro do Tubarão (Tubarão Hill) is privileged with spectacular views of the city's beaches. Two favelas [End Page 19] sit on Tubarão Hill: Tubarão proper and Ceuzinho. Combined they have roughly 15,000 residents. The favela of Tubarão has a large migrant population from the Northeast, while Ceuzinho's residents are predominantly locally born. Both favelas have a lively cultural life, with an active samba school, weekly forros (a popular northeastern dance), and hip-hop balls funded by local traffickers (Carlos 1998; Elizete 1998; Baile Funk 1998; Oscar 1999).

Tubarão and Ceuzinho have experienced tensions, which, in the 1980s and 1990s, culminated in open gang warfare between the two communities. In recent years, however, relations have calmed as traffickers from Tubarão have taken control of both hills (Bernardo and unknown resident of Ceuzinho 1998; Anderson 1998; Sister Elena 1998; Elizete 1998). Access by wealthy buyers to Tubarão's bocas, however, creates heavy drug sales and, eventually, intense conflict among rival drug gangs for control of local points of sale. Caught between the sea and the sky, the Morro do Tubarão stands out as an example of the contradictions of wealth and poverty that define Brazil (Sara and Bartolomé 1998; Sara 1998; Denise 1998a; Vilma 1998; MC Big 1998; Carolina 1998).

With the exception of religious groups, few outside organizations operate in Tubarão. Those that do are either tightly connected to Bernardo, the AM president, or to the traffickers with whom he has worked. Bernardo uses his connections with other groups to limit possible challenges to his authority and to maintain tranquil conditions for traffickers. Ceuzinho's political structure is much more divided, with a number of interlinked groups operating separately, with different connections to the state. Despite this more complex structure, Bernardo and the traffickers occasionally interfere in politics in Ceuzinho. As a result, collective action in Ceuzinho has been rare over the past 15 years.

Tubarão is a very violent favela. Traffickers and (often corrupt) police frequently engage in shootouts. In the nine months I regularly worked in these favelas during 1998 and 1999, more than 20 shootouts occurred. During this time also, traffickers and police killed at least 10 residents. One resident reported that in 1999 alone, 30 were murdered in the two favelas (Cedric 2001). Residents indicated that violence had actually occurred at much higher rates in earlier years, because of gang warfare.

Overall, the favela has benefited from contacts with the government that have led to extensive housing and infrastructural improvements. The favela has also received significant assistance from NGOs and religious organizations (Bernardo 1998; Denise 1998b). This assistance, however, has been largely clientelistic and funneled directly through the AM. This has reinforced the importance of AM leaders, who have been linked to traffickers since the 1980s. [End Page 20]

Crisis and Response

On May 14, 2000, elements of Rio's PM went into Ceuzinho and murdered five residents. While police claimed they had killed traffickers, residents disagreed and rioted in the streets, breaking windows with rocks and cooking gas canisters. In the aftermath of the riots, with extensive news coverage on everyone's mind, the state government promised to improve local policing and began a series of discussions with angry residents (Amalia 2001). Despite these meetings, however, the government did little in the immediate aftermath of the murders. PM Major João Antunes, however, began to work with Viva Rio to develop a community policing program. One Viva Rio leader knew about Operation Cease Fire, a police program centered on harm reduction that had successfully controlled violence in the Dorchester section of Boston, and arranged for Antunes to travel there (Samuel 2001).

On his return, Major Antunes began the collaborative process of setting up a police program based on the Boston experiment. It involved building the confidence of residents to denounce police abuse and, at the same time, establishing an effective police presence to prevent serious crimes. Initially, Antunes tried to work with Bernardo of the AM, but encountered significant resistance. He then made contact with Jorge, the president of Ceuzinho's AM. Long an important local figure, Jorge had many contacts with outside political and civic leaders and little sympathy for the drug traffickers who, years earlier, had briefly driven him out of the favela. Jorge readily gave the police access and helped them build local contacts. Eventually, the traffickers again forced Jorge out, but not before the police had set up the program and established strong ties to some residents. A former Ceuzinho AM president who had once worked closely with Jorge said,

The police in the community control violence and there are few shootouts. That is good. It wouldn't have happened without Jorge, who was willing to come with his chest out and take a lead in bringing the police into the community and to walk up and down the hill with the major to help start the program. This was a huge risk that almost cost Jorge his life, but he got the police in there, and it made a difference. (Alexandre 2002)

On September 22, 2000, the Rio PM began a social occupation of Tubarão with the support of Viva Rio and the Ceuzinho AM. A team of one hundred police took control of key points in the favela and adopted a nonconfrontational crime-control strategy based on monitoring access to the hill and slowly suffocating trafficking. PM officials worked with Viva Rio to set up a police training program focused on increasing professionalism and improving service to residents (Police group 2001). Antunes told arriving police that they should see themselves as serving [End Page 21] the needs of residents, and quickly removed police who did not follow orders. This strategy reduced conflict and resulted in a decline in reported homicides.

In conjunction with the change in policing, Viva Rio and the state government began a series of social programs designed to build local confidence in the police and reduce dependence on traffickers. The State of Rio invested in high school completion courses for three hundred at-risk teenagers. The government paid students in these courses a stipend. Former members of the Ceuzinho AM and other respected residents took the lead in these programs to help build up the confidence of residents (former AM president in Ceuzinho 2001).

Viva Rio provided legal assistance to residents, helped another NGO to set up a volleyball program, organized a second high school equivalency class, organized competing local religious groups to work together collectively to address neighborhood problems, and set up other programs to help build local capacity and meet residents' needs (Viva Rio 2001). Viva Rio, the federal government, and an important local foundation worked with UNICEF, the United Nations Children's Fund, to open a large sports program designed to serve 1,500 children from the hill and surrounding neighborhoods (Sports project, 2001). Finally, Viva Rio organized a community forum to give local leaders a space to meet with each other, police leaders, and NGO workers to discuss problems. These activities helped groups in Tubarão to engage with one another.

Solutions to Violence

The efforts of police, local leaders, and the outside groups had a significant affect on Tubarão. Residents and police reported a considerable decline in shootouts. One resident who worked with Viva Rio said, "Things have gotten a lot better in the community. There are no more shootouts. There is no violence" (Cedric 2001).

Although residents remained wary, relations with the police thawed. Residents reserved special complements for Major Antunes, who they thought was a good man working in a bad institution. One NGO worker said, "many people say that they don't like the police but say that Major Antunes is okay" (Guilherme 2001). One resident deeply linked to trafficking, both personally and through family connections, and who had once served time in prison, sought out Antunes for help when she had no work. She said that while she could not trust most of the police, she felt that she could trust him (Elizete 2001). Another resident, an Evangelical, said that while she did not trust police in general, she did think the major was a little bit better than the others (Carolina 2001).

Growing confidence in the police hierarchy allowed residents to denounce police behavior to officers. As a result of residents' assistance, [End Page 22] Major Antunes removed 40 of the 100 police assigned to him for disciplinary reasons (Antunes 2001). A resident noted,

the work of Viva Rio and the new policing program has changed people's vision of the police. Now they [the police] have a way to talk to the community. The police walk in the community, residents have access to the police; they talk to them. [When there were problems with other police]...the major went to the battalion to fix things. He gave his cell phone number to [community] leaders. Residents didn't trust the new program at first, but after they [the 40 police accused of extortion] were removed, it gave a new viability to their work. (Cedric 2001)

Major Antunes's commitment to this program and his ability to negotiate the complex relationships between the state, the favela, and the NGOs allowed this program to move forward. Through his contacts with NGOs, Antunes developed strategies for effective police reforms. By maintaining those contacts, he enlisted the help of Viva Rio to advance the project and to obtain outside funding to realize the social component of the police occupation. Through his connections with the PM, Antunes secured the resources necessary to set up and maintain the project. A populist state government interested in developing innovative public security programs facilitated this. Antunes and other high-ranking police also maintained extensive contacts with the press to promote international awareness of the project and obtain additional support.

The work of Viva Rio was important for several reasons. First, Viva Rio helped Antunes develop his program by putting him in touch with Operation Cease Fire. Later, the NGO monitored violence, trained police, and offered advice to both residents and state representatives. Viva Rio also helped legitimate the social aspects of the project; the participation of a civic organization showed residents that this project had support beyond the state.

Viva Rio also brought significant funds to the social and educational sides of the project, and served as a liaison between residents and the government. Through the funding, Viva Rio helped to meet some local needs, gave small stipends to at-risk teens in equivalency courses, and generally decreased residents' dependence on traffickers. Finally, Viva Rio helped build local capacity to pressure police and traffickers to control violence. It did this by building a community leadership council to discuss problems, establishing a network of religious organizations to develop new programs to help meet residents' needs, and setting up a training program to help professionalize the police. As one resident put it, "The community is growing and learning. The [leadership] council is proof that the community is uniting and making demands on the government" (Cedric 2001). [End Page 23]

Jorge, Ceuzinho's AM president, represented the important role of local leaders in that he helped to bring in the police and set up social programs after Bernardo refused to help. Jorge walked around the favela with Antunes and other outsiders, introducing them to residents and community organizations. Through these introductions, outsiders built a strong network of contacts that proved invaluable in implementing the program. Having community leadership was essential to these programs because the target audience was underserved, at-risk teenagers close to drug traffickers, who had reason to distrust outside authorities.

Of course, some actors resisted the program. Because of his ties to the police, Jorge was forced out by the traffickers, and Bernardo took over the Ceuzinho AM. Bernardo tried to limit the role of the police by calling for the removal of a number of police posts in the favela. Some police took bribes and not only allowed drug dealing but also provided drug traffickers with information about police operations. Viva Rio also had some internal divisions that led to poor decisionmaking and the hiring of outsiders to work in the community when residents could easily have done the job. As one former Ceuzinho AM president who was involved with a sports program said, "[a foundation] gave [Viva Rio] a lot of money [to start the UNICEF sports program] and they used it to hire outsiders." He went on, "They pay them R$2000 or R$3000 a month and they run things. Residents were only given jobs as janitors" (Alexandre 2002). 5 Over time, this decreased Viva Rio's community support, much as a similar decision by the CdP executive director to hire outsiders decreased his support in Vigário.

Despite all these problems, the contacts that had already been established in the favela helped maintain the program. Community leaders gave legitimacy to the police presence by publicly speaking out in favor of it even after Bernardo took over the Ceuzinho AM. The vice president of the Ceuzinho AM publicly said at a leadership council meeting,

The police are in the hill every day. If the police go down the hill, the community will pay. They [the residents] need someone to go to directly to make sure they are satisfied. If they get rid of these police, other police will go up and invade the community. (Tubarão leadership council 2001)

Although Bernardo immediately and publicly admonished him, other residents at the meeting endorsed what the vice president said. Statements such as this help to reinforce and develop community support for positive police actions.

Viva Rio helped to resolve some of its own problems in working with the community by hiring Eduardo, an activist from a nearby favela, to act as its intermediary with local leadership; and by employing Cedric, a local Baptist pastor, to work on a number of its projects. With [End Page 24] their own strong local contacts and history of working with leaders in these communities, Eduardo and Cedric helped immeasurably in dealing with difficulties in Tubarão. When Bernardo would resist the program, they could advise him that he was hurting their work as favela residents. This would be politically untenable for Bernardo as a local leader committed to defending local interests.

By establishing a strong network of local contacts, efforts to maintain the new policing program succeeded. Residents who previously may not have had the courage to speak out publicly about these issues now were connected to other residents and outsiders who shared their concerns. Residents supported each other in efforts to improve policing despite traffickers' attempts to limit those activities.

Without the support of such a variety of actors, the intervention would not have had significant success. If, for example, Bernardo provided the only link between Tubarão and the outside, the police could not have set up the program. Efforts to set up, maintain, and develop the project encountered such resistance that Jorge was forced out of the favela; yet the complex, interlocking network of government agents, civic leaders, and residents made this intervention succeed.

The efforts of the government and Viva Rio in Tubarão were part of a broader public security policy undertaken by the administrations of Governors Anthony Garotinho and Benedita da Silva between 1998 and 2002. These governors were elected on a reform platform written by Luis Eduardo Soares, a sociologist and Viva Rio collaborator. The platform, and government policies during this period, focused on building links between favelas, civil society, and the state to decrease the power of traffickers in these communities and to control violence. Antunes actually had been Soares's aide while he was in government. Soares himself was also, no doubt, aware of the activities of Viva Rio in Vigário in the early and mid-1990s. Any overall decrease in the growth of violence in Rio during the Garotinho and da Silva administrations is most probably attributable to a broad set of policy recommendations that involved increasing police transparency and strengthening ties between poor communities and the police. The success of the efforts to control violence in Tubarão reflects those policies.

Postscript on Tubarão

Bernardo died suddenly in a car accident in mid-2002. His death seems to have made some communications between police and traffickers difficult, and his replacements were unable to smooth relations between police and traffickers (Guilherme 2003). This difficulty, combined with the increasing distance between Viva Rio and the community and the transfer of Major Antunes from the favela, seems to have weakened the [End Page 25] network. By July 2003, one contact in the community and an outside activist reported increased levels of conflict in the favela (Elizete 2003a, b; Guilerme 2003).

An Analysis of the Factors that Promote Network Success

The events in these three communities offer a number of insights into the basis for political action in highly violent neighborhoods in Latin America. Both police reform and social action play a critical role in any effort to control violence in favelas. In all three communities, the government initiated police reform only after significant local protests. In the two successful cases, a significant amount of time passed between the peak of crisis, when protests occurred, and policy change. Only as a result of ongoing contacts with the government, outside groups, the media, and residents did change occur. Even with reform, furthermore, many problems persisted. What structures of state-local engagement made these reforms successful over the long term? The answer lies in the activities of networks.

In both communities where police reform offered some palpable relief to residents, complex networks evolved that brought numerous local groups together with actors from civil society and the state. In both Tubarão and Vigário, government, civic, and local actors all had important roles in helping to change local conditions. Each group had a functional role in controlling local violence by drawing on its own unique resources to restrain police and traffickers. Government leaders ordered police reforms and provided funds to supplant traffickers' aid to residents. Civic leaders provided funds and helped mediate the relationship between the government and the community. Residents provided information about police violence necessary for reforms, and provided government and civic leaders access to the community. Only through the concerted activities of groups and individuals committed to change could organizations committed to controlling violence overcome the resistance of traffickers and their allies.

In Santa Ana, by contrast, a lack of significant outside connections limited the impact of local groups' protests. With most local actors linked to criminals, outsiders had little access to or contact with the favela. The activities of the social club offer some hope; the one group effectively connected to outsiders and not compromised with the traffickers played an important role in organizing the protest against police violence.

Government agents have access to certain types of political resources that other groups do not, and play a fundamental role in building citizenship and controlling violence (Velho 1996, 22-23). The Brazilian state, however, is often very divided and corrupt. 6 As a result, state [End Page 26] actors concerned with political change and decreasing violence need to manipulate links both inside and outside the state to promote policy outcomes that help them achieve their goals. In Tubarão, Major Antunes had a critical role in the reforms that occurred. Using his contacts in the state and in civil society, Antunes built support for an innovative new police program that effectively controlled homicides. Antunes used his position of power in the police to promote reform and informed residents of his efforts. In Vigário, the orders of Governor Alencar forced the police to change their behavior. In both Vigário and Tubarão, changes in government behavior created changes in levels of violence.

Civic actors also help in efforts to transform local politics. In each of the three cases, outside actors provide financing, training, and political and media connections that help in effecting policy change. By providing money and training, outside civic groups help build local institutional capacity, build programs to decrease dependence on traffickers, and maintain pressure on state actors. Networks facilitate this process by accelerating the learning process as groups become more densely connected; networks also provide an efficient system for the transfer of resources (see especially Podolny and Page 1998, 64-66; Uzzi 1996, 677-80, 682; Putnam 1993, 174; Powell 1994, 303-4).

Through political and media connections, groups in civil society put actors in favelas in touch with outside groups to which they might not previously have had access. The FAR in Vigário, for example, regularly brought in political dignitaries and reporters and maintained routine contact with lawyers whom residents and activists occasionally contacted. Isolated poor people can use these connections to the state to put new types of pressure on political leaders, as human rights groups did throughout Latin America in the 1970s and 1980s (Sikkink 1996, 70-71; Chalmers et al. 1997, 580-82).

By putting direct pressure on the state and increasing public awareness of violence, the media contribute to this process. At the same time, the media, when connected to the wrong actors, can also create justifications for further police violence. The media themselves form parts of these networks, but only rarely develop intimate links with particular favelas. Essentially, groups with experience dealing with the media have some idea of how to control what the media publish; but it is rare that groups with this experience are active in or linked to favelas. Often residents believe that they are in so much danger that they will not talk to the press. Thus it is difficult for activists to get the type of reporting that can pressure the state, over the long term, to control violence.

Distrust and violence also make it difficult for the media to gain access to favelas. The media therefore depend on contacts with possibly corrupt, violent, and complicit police to report on crime. Only when favela residents develop connections with groups that know how to manage the [End Page 27] media can residents get the type of attention, and the media get the type of access and data, that will help control violence over the long term.

The evidence from all three favelas shows that the actions of both state actors and civil society groups depend on their coordination with groups inside the affected communities. In closed, violent places, it is difficult for well-intentioned outsiders to gain access without the help of trusted locals. The evidence here supports the theory that networks help to transfer legitimacy between different groups through introductions and peer surveillance (Powell 1994, 303-4; Putnam 1993, 171-72; Podolny and Page 1998, 64-65). In Vigário, numerous outsiders gained access through politically active locals who introduced them to residents. These groups then used their outside connections to develop programs to help residents and to monitor the police. Tubarão followed a similar pattern. The new policing program and Viva Rio gained access through the numerous leaders in Ceuzinho who favored change. The program, as many observers noted, could have been set up only because of Viva Rio and the government's contacts with Jorge, the Ceuzinho president.

Poor, isolated communities also tend to have strong protective ties among residents that make it difficult for outsiders to obtain information (Granovetter 1983, 212-13; Espinoza 1992, 52). 7 Local groups in Vigário, Tubarão, and Santa Ana were long aware of the details of violence in their favelas. Only through links with outsiders, however, could residents get their information safely into the hands of institutions with the connections necessary to get the information to those with the means to effect change. In Vigário these groups brought attention to police violence through the media and through contacts with high-level state officials. In Tubarão, groups worked through Major Antunes and Viva Rio by making these authoritative actors aware of police abuse.

By linking multiple actors, networks also provide member groups protection from violence and help to ensure that efforts to bring about political change will continue even in circumstances where criminals successfully intimidate a number of activists. When groups operate together, the cost to violent actors of threatening or coopting member groups increases. If violent actors eliminate only one leader, it may not deactivate the network as a whole. This can be seen in both Vigário and Tubarão. The network as a whole continued to operate, if at a slightly lower level, despite the departure of Caio Ferraz in Vigário and Jorge in Tubarão and Ceuzinho. This stands in sharp contrast to Santa Ana, where one civic organization, the AM, ran the favela.

The different groups in Vigário illustrate how members of networks collectively can pursue different political strategies that a single group could not effectively pursue. Thus, while the CdP adopted a more confrontational attitude toward the state, the FAR and MSF worked more [End Page 28] collaboratively with state actors. Ultimately, networks succeed in changing policy because their decentralization and specialization produces a concert of diverse voices and perspectives focused on achieving similar changes in a specific area. Politicians, instead of confronting isolated local leaders, face groups that not only represent but demonstrate, through their independent actions, local agreement about the need for change. Different groups say and do different things and do not strive, as social movements often do, to produce a single voice with which to pressure state actors; but they nonetheless coordinate their activities and speak to the same local issues. In a place characterized by very high levels of violence, this decentralization, specialization, and polyvocality are essential in creating political change. The consistent activities of multiple groups also help to build new local norms that discourage residents from turning to traffickers and empower them to call on the government to solve local problems.

Of course, networks and civic groups have some negative aspects. Not all groups active in poor communities are politically progressive or concerned with human rights. Police, traffickers, and local leaders worked together to limit reform in Santa Ana and provided significant resistance in Tubarão. With access to violence, these groups pose significant threats to networks whose members are concerned with controlling conflict. At times, these violent groups may try to use connections in a network to undermine groups they are unhappy with, as in Vigário when traffickers felt threatened by the connections between the new executive director and a trafficker in another favela. Because these groups pose the threat of violence, however, only networks offer a significant hope for change, because they can share risks and more easily resist cooptation than other types of organizations can. This resilience can be seen in both the Vigário and Tubarão networks; despite demobilization efforts, traffickers could not end network activities. The flexibility of networks helps them deal with challenges that criminals pose and allows them to work to reduce violence.

Criminals and their associates are not the only challenge to groups concerned with controlling conflict. As evidence from both Vigário and Tubarão shows, poor political decisions on the part of network group leaders can result in reduced network activity, local dissatisfaction, and increased violence. Furthermore, evidence from Vigário suggests that networks' connections can even bring about the deactivation of some network members. The CdP stopped functioning because a coalition of activists quickly came together to undermine its leadership. The evidence suggests that when communication breaks down and groups do not effectively distribute or manage funds, problems can arise in the network.

In general, organizations that are more densely organized and maintain more effective open connections among member groups will tend [End Page 29] to avoid these problems. As connections become more distant and groups communicate less effectively on a formal level, it is easier for outside groups to have negative influence on networks. One of the problems in the CdP was that, by moving to the United States, Caio Ferraz cut himself off from his contacts in Brazil. At the time the CdP fell apart, both Ferraz and the new executive director had poor links to groups living and working in Vigário.

The real challenge that activists working in networks will confront is how to structure their networks and their communications so as to limit the impact of bad political decisions and self-destructive internal coalitions. Overcoming these problems probably means that group leaders must make constant efforts to maintain contacts with diverse groups inside and outside the community.

This all leaves one important question unanswered: where do these networks come from? Evidence in this paper suggests that many communities, even those with powerful trafficking organizations, contain latent activist groups with connections to the outside. Under some circumstances, such as Santa Ana, some groups with limited connections to the outside can be forced to limit their political activism; but these groups exist nevertheless. The question is why and how they activate. The data here suggest that at certain critical junctures, when violence becomes unbearable or when certain widely held community norms are suddenly violated by state agents, the community will rise up. This process can be seen in Santa Ana, Tubarão, and Vigário. While sometimes this might develop into short-lived social movements, in some cases groups will activate that have effective links with the outside and the ability to develop new connections. Where groups can develop connections and use them to mobilize external support, violence can be brought under control; this was the case in Tubarão and Vigário. The emergence of the network, however, can follow either a top-down pattern, as in Tubarão, where outsiders initiated changes and then activated groups in the favela; or a bottom-up pattern, as in Vigário, where mobilization began in the community and made links to the outside.


This article has looked at three cases of political change in violent neighborhoods in Rio de Janeiro. The data presented here show that while both institutional reform and popular mobilization play critical roles in addressing problems in those communities, networks provide essential impetus to support reforms and stimulate long-term mobilization.

Active organizational contacts help local groups transcend their own community to bring in resources, support important programs, and pressure government agents to undertake reforms. Network contacts promote [End Page 30] learning and communication, thereby helping to create more effective and longer-term popular mobilizations and to compel state leaders to control crime. These efforts promote local administrative reforms that result in controls on police violence. As police corruption diminishes, drug-trafficking violence will tend to subside, as police enforce the law and, as in Tubarão, residents begin cooperating with police. By operating in networks, groups concerned with reducing violence can limit the effect of attacks against member groups and continue efforts to control conflict.

Evidence presented in this paper also suggests, nevertheless, that networks struggling to reduce violence may experience significant internal and external challenges from groups that would prefer things to continue as they are. The drug traffickers who operate in these communities, allied with local organizations, can resist some efforts to control violence and can use network resources to prevent groups concerned with controlling violence from building the local connections necessary to reduce conflict. Criminals can also use internal conflict and the specific structure of networks to undermine network cohesion. To succeed, therefore, network members must be aware of the problems they can confront and must work to keep communication channels open. Only through mutual cross-institutional support can networks overcome challenges from criminals.

The stories told here suggest the need for further research in small case studies to understand the dynamics of local-level organizing in Latin America today. Under conditions of ongoing social violence, only close, detailed local studies can provide the types of insights necessary to understand the important localized political dynamics that will be essential in furthering the process of democratic consolidation in the region.

Enrique Desmond Arias is an assistant professor of government at John Jay College of Criminal Justice, CUNY, and a fellow at the Bildner Center for Western Hemisphere Studies at the CUNY Graduate Center. His current research concerns the roles of criminal gangs, human rights, and local-level governance in Rio de Janeiro.


I would like to thank Leigh Payne, Michael Schatzberg, Richard Merelman, Aaron Seeskin, Beth Dougherty, and four anonymous reviewers for invaluable advice and comments on different drafts of this article. Funding for the research discussed in this piece was provided by the Institute for the Study of World Politics; a Fulbright Scholarship; a Tinker Grant from the Latin American and Iberian Studies Program at the University of Wisconsin, Madison; a MacArthur Fellowship; a Scott Kloeck-Jensen International Internship Grant from the Global Studies Program at the University of Wisconsin, Madison; and a grant from Oberlin College.

1. Santa Ana and Tubarão are pseudonyms used to protect informants' confidentiality. The real name of Vigário Geral is used because a secondary literature already exists on this favela.

2. On the role of the state in extending guarantees to the poor, see Reis 1996, 122, 133-34. [End Page 31]

3. On mobilization for gender rights in São Paulo, see Alvarez 1989, 225-28; on labor mobilization see Keck 1989, 255-72. For examples from Rio's favelas, see Gay 1994, 20, 25-26; for suburban Rio see Mainwaring 1989, 168-69, 173-82.

4. Acari is a set of favelas and a public apartment complex in the municipality of Rio, many miles northwest of downtown.

5. At existing exchange rates, this would equal between US$800 and US$1,200. Typically, favela residents with jobs as janitors in the center would probably make around US$100 a month.

6. On the structure of the Brazilian state, see Evans 1995, 60-66. On shifting structures of the state and the role of networks in state decisionmaking, see Chalmers et al. 1997, 556-57.

7. On types of network ties, see also Putnam 2000, 22-23.


The research for this project was conducted in July and August 1996, April 1997 to May 1999, June and August 2001, and June 2002. The data presented here were obtained through extensive participant observation and selective interviews with local leaders, residents, and activists. With the exception of one public figure discussed in the project, all names used are pseudonyms. All interviews were conducted in Rio de Janeiro.

Almeida, President of Vigário AM, and unknown resident. 1997. December 5.

Almeida and administrator of PROSANEAR state program to improve water service. 1997. November 24.

Almeida and two unknown residents. 1997. June 28.

Almeida and two unknown women residents. 1997. June 18.

Almeida and social scientists working on Favela-Bairro city-run urbanization program. 1998. January 29.

Almeida, Cynthia, and Lorivaldo. 1997. Vice President of AM 1997-mid 1998; receptionist of AM; resident. October 2.

Almeida, Cynthia, and unknown resident. 1997. November 17.

Alexandre. 2002. Ex-President of Ceuzinho AM. June 13.

Amalia. 2001. Viva Rio lawyer working in Tubarão. July 3.

Anderson. 1998. Head of local government center, Ceuzinho. October 21.

Bernardo. 1998. President, Tubarão AM. November 24.

Bernardo and unknown resident of Ceuzinho. 1998. October 6.

Camila. 1998. Director of crèche, longtime resident, Santa Ana. September 9.

Carlinha. 1998. Massacre survivor, CdP activist, resident of Vigário. January 30.

Carlos. 1998. AM worker, longtime resident, Ceuzinho. Conversation, August 13.

Carolina. 1998. Evangelical Christian, resident of Tubarão. December 17.

----. 2001.

Cedric. 2001. Baptist pastor, liaison with Viva Rio, resident of Ceuzinho. July 10.

Charles. 1997a. Artist, activist, resident of Vigário. October 15.

----. 1997b. October 17.

----. 1997c. October 20. [End Page 32]

Charles and outside filmmaker. 1997. November 18.

Cynthia. 1997a. Former resident, AM receptionist, Vigário. September 16.

----. 1997b. October 7.

Cynthia and Felipe. 1997. Community sanitation worker, resident, Vigário. October 30.

Cynthia and Tânia. 1997. Vigário resident. October 9.

Cynthia and other women at the Vigário AM. 1997. October 7.

Daniel. 1998. Activist, Vigário resident. January 7.

Daniel and Joana. 1998. Daniel's mother, Vigário resident. January 7.

Denise. 1998a. Employee of government center, resident, Ceuzinho. November 25.

----. 1998b. December 3.

Elizete. 1998. Ceuzinho resident. November 12.

----. 1999. January 6.

----. 2001. July 4.

----. 2003a. July 15.

----. 2003b. July 18.

Eric. 1998. Associate Director, CdP, Vigário; nonresident. January 15.

Evanildo. 1998. Van driver, activist, Vigário resident. January 10.

Ferraz, Caio. 1997a. Activist, president of Vigário CdP, social scientist, former resident. August 27.

----. 1997b. Comments in CdP video. Watched October 1.

Guilherme. 2001. Viva Rio Organizer. June 28.

----. 2003. July 15.

Jaime. 1997. Director, Vigário MSF; nonresident. May 28.

Jaime and Vigário city officials. 1998a. January 9.

----. 1998b. January 29.

Jorginho. 1997a. Executive Director, Vigário CdP; nonresident. October 13.

----. 1997b. October 14.

Jorginho and Eric. 1998, January 15.

Joselino. 1997. Evangelical, hardware store owner, longtime Santa Ana resident. July 3.

----. 2001. July 22.

----. 2002. June 5.

Josias and Antônio. 1996. President and Officer, Santa Ana AM. July 26.

Josias and Manoel. 1997. Vice President, Santa Ana AM. May 13.

Josias and unkown woman. 1997. Santa Ana resident. June 26.

Katarina. 1998. Mãe de Santo (Candomblé religious leader), Vigário resident. January 29.

Lourenço. 1998. Vigário CdP activist, resident of nearby community. January 30.

Luis. 1997. Vigário community business leader, former AM president. October 9.

Mateus. 1998. Nonresident teacher, community activist, Vigário. January 7.

MC Big. 1998. Radio disc jockey, Tubarão resident. December 16.

Miguel. 1997. Former plumber, Vigário AM member. November 25.

Oscar. 1999. Occasional resident, Tubarão. Conversation. May 8.

Pedro and Wesley. 1997. Members of FAR band, administrators of FAR Cultural Center in Vigário, Vigário residents. May 28.

Roberto. 1997a. Vigário FAR director, activist, nonresident. October 10.

----. 1997b. October 15. [End Page 33]

Roger. 1997. Vigário resident. Conversation. October 17.

Samuel. 2001. Coordinator for Human Rights and Public Security, Viva Rio. June 28.

Sandrão. 1998. Social club director, longtime resident, Santa Ana. September 9.

Sara. 1998. Doctor at local health clinic, Tubarão. December 1.

Sara and Bartolomé. 1998. Epidemiologist, local health clinic, Tubarão. November 24.

Sister Elena. 1998. Foreign nun living in Tubarão. November 25.

Unknown Resident, Vigário. 1997. July 2.

Vilma. 1998. Evangelical Christian, longtime resident, Tubarão. December 6.

Author Observations

Baile Funk. 1998. Tubarão, April 16.

CdP activists. 1998. Meeting at CdP. January 30.

CdP leadership and staff. 1997. Meeting. September 30.

Commission to Reorganize CdP. 1998. Meetings. January 30, February 8.

Police group. 2001. Meeting at Viva Rio, July 18.

Sports project. 2001. Observations and conversations at opening of sports facility funded by UNICEF. August 11.

Tubarão leadership council. 2001. Meeting. July 7.

Visit of British Vice Foreign Minister. 1997. Vigário, September 18.

Visit of NGO workers. 1997. Vigário, October 2.

Visit of prosecutor in massacre case. 1997. Vigário, September 19.

Viva Rio. 2001. Meeting. July 17.


Alvarez, Sonia. 1989. Politicizing Gender and Engendering Democracy. In Stepan 1989. 205-51.

Alvarez, Sonia, Evelina Dagnino, and Arturo Escobar. 1998. Introduction: The Cultural and the Political in Latin American Social Movements. In Alvarez et al. 1998. 1-32.

Alvarez, Sonia, Evelina Dagnino, and Arturo Escobar, eds. 1998. Cultures of Politics/Politics of Cultures: Re-visioning Latin American Social Movements. Boulder: Westview Press.

Alvito, Marcos. 1998. Um bicho-de-sete-cabeças. In Um século de favela, ed. Alba Zaluar and Alvito. Rio de Janeiro: Editora Fundação Getúlio Vargas. 181-208.

----. 2001. As cores de Acarí: uma favela carioca. Rio de Janeiro: Editora Fundação Getúlio Vargas.

Amnesty International. 1998. Urgent Action Appeal. <www.ai.org> December 11.

Antunes, João. 2001. Major, Rio military police; commander of police program in Tubarão. Statement at Tubarão leadership council meeting. July 7.

Becker, David G. 1999. Latin America: Beyond "Democratic Consolidation." Journal of Democracy 10, 2: 138-51.

Caldeira, Teresa P. R. 1996. Crime and Individual Rights: Reframing the Question of Violence in Latin America. In Jelin and Hershberg 1996. 197-214. [End Page 34]

Cardoso, Ruth Corrêa Leite. 1992. Popular Movements in the Context of the Consolidation of Democracy in Brazil. In The Making of Social Movements in Latin America: Strategy, Identity, Democracy, ed. Arturo Escobar and Sonia Alvarez. Boulder: Westview Press. 291-302.

Chalmers, Douglas, Scott Martin, and Kerianne Piester. 1997. Associative Net works: New Structures of Representation for the Popular Sectors? In The New Inequality in Latin America: Rethinking Participation and Representation, ed. Chalmers, Katherine Hite, Martin, Piester, and Monique Segarra. Oxford: Oxford University Press. 543-82.

Chevigny, Paul. 1995. Edge of the Knife: Police Violence in the Americas. New York: New York University Press.

Correa Sutil, Jorge. 1999. Judicial Reforms in Latin America. In Méndez et al. 1999. 255-77.

De Soto, Hernando. 2001. The Constituency of Terror. New York Times, October 15: 19.

Downie, Andrew. 2000. Taking Boston's Lead, Police in Rio Lighten Up. Christian Science Monitor, November 8: 1.

Duffield, Mark. 2001. Global Governances and the New Wars: The Merging of Development and Security. New York: St. Martin's Press.

Economist. 2001. Nice Cops, Nasty Cops. September 8: 31-36.

Espinoza, Vincente. 1992. Networks of Informal Economy: Work and Community Among Santiago's Urban Poor. Ph.D. diss., University of Toronto.

Evans, Peter. 1995. Embedded Autonomy: States and Industrial Transformation. Princeton: Princeton University Press.

Fernandes, Rubem César. 1994. Private but Public: The Third Sector in Latin America. Washington, DC: Civicus Press.

Garotinho, Anthony, et al. 1998. Violência e criminalidade no Estado do Rio de Janeiro: diagnóstico e propostas para uma política democrática de segurança pública. Rio de Janeiro: Editora Hama.

Gay, Robert. 1994. Popular Democracy and Political Organization in Rio de Janeiro: A Tale of Two Favelas. Philadelphia: Temple Unviersity Press.

Granovetter, Mark. 1983. The Strength of Weak Ties: A Network Theory Revisited. Sociological Theory: 201-33.

Hochstetler, Kathryn. 2000. Democratizing Pressure from Below? Social Movements in the New Brazilian Democracy. In Democratic Brazil: Actors, Institutions, and Processes, ed. Peter R. Kingstone and Timothy J. Power. Pittsburgh: University of Pittsburgh Press. 167-84.

Holloway, Thomas. 1993. Policing Rio de Janeiro: Repression and Resistance in a Nineteenth-century City. Stanford: Stanford University Press.

Huggins, Martha. 1998. Political Policing: The United States and Latin America. Durham: Duke University Press.

Human Rights Watch Americas. 1997. Violencia x violencia: violações aos direitos humanos e a criminalidade no Rio de Janeiro. New York: Human Rights Watch Americas.

Jelin, Elizabeth, and Eric Hershberg, eds. 1996. Constructing Democracy: Human Rights, Citizenship, and Society in Latin America. Boulder: Westview Press.

Jornal do Brasil (Rio de Janeiro). 1996. Polícia critica candidatos que sobem favelas. August 12: 1, 3. [End Page 35]

Junqueira, Eliane, and José Augusto Rodrigues. 1992. Pasárgada revisitada. Sociologia: Problemas e Práticas 12: 9-18.

Kaldor, Mary. 1999. New and Old Wars: Organized Violence in a Global Era. Stanford: Stanford University Press.

Kanter, Rosabeth Moss. 1991. The Future of Bureaucracy and Hierarchy in Organizational Theory: A Report from the Field. In Theory for a Changing Society, ed. Pierre Bourdieu and John Coleman. Boulder: Westview Press. 63-87.

Keck, Margaret E. 1989. The New Unionism in the Brazilian Transition. In Stepan 1989. 252-96.

Keck, Margaret E., and Kathryn Sikkink. 1998. Activists Beyond Borders: Advocacy Networks in International Politics. Ithaca: Cornell University Press.

Kurtz, Marcus J. 1999. Free Markets and Democratic Consolidation in Chile: The National Politics of Rural Transformation. Politics and Society 27, 2: 275-301.

Leeds, Elizabeth. 1996. Cocaine and Parallel Polities on the Brazilian Urban Periphery: Constraints on Local Level Democratization. Latin American Research Review 31, 3: 47-84.

Mainwaring, Scott. 1989. Grassroots Popular Movements and the Struggle for Democracy: Nova Iguaçu. In Stepan 1989. 168-204.

Méndez, Juan E. 1999. Institutional Reform, Including Access to Justice. In Méndez et al. 1999. 19-24.

Méndez, Juan E., Guillermo O'Donnell, and Paulo Sérgio Pinheiro, eds. 1999. The (Un)Rule of Law and the Underprivileged in Latin America. Notre Dame: University of Notre Dame Press.

Migdal, Joel. 1994. The State in Society: An Approach to Struggles for Domination. In State Power and Social Forces: Domination and Transformation in the Third World, ed. Migdal, Atul Kohli, and Vivienne Shue. Cambridge: Cambridge University Press. 7-36.

O'Donnell, Guillermo. 1999. Polyarchies and the (Un)Rule of Law in Latin America: A Partial Conclusion. In Méndez et al. 1999. 303-38.

Paoli, Maria Cecilia, and Vera da Silva Telles. 1998. Social Rights: Conflicts and Negotiations in Contemporary Brazil. In Alvarez et al. 1998. 64-92.

Payne, Leigh A. 2000. Uncivil Movements: The Armed Right Wing and Democracy in Latin America. Baltimore: Johns Hopkins University Press.

Pereira, Anthony W. 2000. An Ugly Democracy: State Violence and the Rule of Law in Post-authoritarian Brazil. In Democratic Brazil: Actors, Institutions, and Processes, ed. Peter R. Kingstone and Timothy J. Power. Pittsburgh: University of Pittsburgh Press. 217-35.

Perlman, Janice. 1976. The Myth of Marginality: Urban Poverty and Politics in Rio de Janeiro. Berkeley: Campus Press.

Pessoa, Flávio. 2002. Negocios entre tráfico e polícia. O Globo (Rio de Janeiro), June 20: 15.

Pinheiro, Paulo Sérgio. 1997. State-Sponsored Violence in Brazil. In The New Inequality in Latin America: Rethinking Participation and Representation, ed. Douglas Chalmers, Katherine Hite, Scott Martin, Kerianne Piester, and Monique Segarra. Oxford: Oxford University Press. 261-80.

----. 1998. Polícia e consolidação democrática: o caso brasileiro. In São Paulo sem medo: um diagnóstico da violência urbana, ed. Pinheiro et al. Rio de Janeiro: Garamond. 13-17. [End Page 36]

Podolny, Joel M. and Karen L. Page. 1998. Network Forms of Organization. Annual Review of Sociology 24: 57-76.

Powell, Walter. 1994. Neither Market nor Hierarchy. In The Handbook of Economic Sociology, ed. Neil J. Smelser and Richard Swedberg. Princeton: Princeton University Press. 368-402.

Putnam, Robert D. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press.

----. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon and Schuster.

Reis, Fabio Wanderley. 1996. The State, the Market, and Democratic Citizenship. In Jelin and Hershberg 1996. 121-40.

Resende de Carvalho, Maria Alice, Zairo Borges Cheibub, Marcelo Baumann Burgos, and Marcelo Simas. 1998. Cultura política e cidadania: uma proposata de metodologia de avaliação do programa Favela-Bairro. Unpublished mss. Rio de Janeiro: Instituto Universitário de Pesquisas do Rio de Janeiro.

Risse, Thomas, and Kathryn Sikkink. 1999. The Socialization of International Human Rights Norms into Domestic Practices: Introduction. In The Power of Human Rights: International Norms and Domestic Change, ed. Risse, Stephen C. Ropp, and Sikkink. Cambridge: Cambridge University Press. 1-35.

Santos, Boaventura de Souza. 1995. Towards a New Common Sense: Law, Science and Politics in the Paradigmatic Transition. New York: Routledge.

Sikkink, Kathryn. 1996. The Latin American Human Rights Network. In Jelin and Hershberg 1996. 59-84.

Sneed, Paul. 2000. U.S. resident in Rio. Personal correspondence. August.

Starn, Orin. 1999. Nightwatch: The Politics of Protest in the Andes. Durham: Duke University Press.

Stepan, Alfred C., ed. Democratizing Brazil: Problems of Transition and Consolidation. 1989. Oxford: Oxford University Press.

Tarrow, Sydney. 1998. Power in Movement: Social Movements and Contentious Politics. Cambridge: Cambridge University Press.

Uzzi, Brian. 1996. The Sources and Consequences of Embeddedness for the Economic Performance of Organizations. American Sociological Review 61 (August): 674-98.

Velho, Gilberto. 1996. Violencia, reprocidade, e desigualdade: uma perspectiva antropológica. In Cidadania e violência, ed. Velho and Marcos Alvito. Rio de Janeiro: UFRJ Editora. 10-24.

Ventura, Zuenir. 1994. Cidade partida. Rio de Janeiro: Cidade das Letras.

Wang, Xu. 1999. Mutual Empowerment of State and Society: Its Nature, Conditions, Mechanisms, and Limits. Comparative Politics 31, 2: 231-49.

Williams, Phil. 2001. Organizing Transnational Crime. In Combating Transnational Crime: Concepts, Activities, and Responses, ed. Williams and Dimitri Vlassis. London: Frank Cass.

Yashar, Deborah. 1996. Indigenous Protest and Democracy in Latin America. In Constructing Democratic Governance: Latin America and the Caribbean in the 1990s, ed. Jorge I. Domínguez and Abraham F. Lowenthal. Baltimore: Johns Hopkins University Press. 87-122.

Zaluar, Alba. 1985. A maquina e a revolta. São Paulo: Editora Brasiliense. [End Page 37]

----. 1998. Crime, medo, e política. In Um século de favela, ed. Zaluar and Marcos Alvito. Rio de Janeiro: Editora Fundação Getúlio Vargas. 209-32.

Zaverucha, Jorge. 1999. Military Justice in the State of Pernambuco After the Brazilian Military Regime: An Authoritarian Legacy. Latin American Research Review 34, 2: 43-74.