Faith-Based Organizations, Community Development, and Room of Manoeuvre in Gang Controlled Neighbourhoods in San Salvador

This paper aims to contribute to our understanding of the feasibility and effectiveness of development interventions in situations of chronic violence, paying particular attention to the capacity and room of manoeuvre of intervening organisations in contexts where illicit groups have built up a power position. It analyses the interventions of two faith-based NGOs (FBOs) that aim to reduce violence and promote community development in selected gangcontrolled neighbourhoods of the Metropolitan Area of San Salvador (AMSS). Based on a literature study and data from fieldwork in several municipalities of the AMSS, it focuses on the ways the organisations navigate in extremely complex contexts. It is argued that the factors that contribute to the capacity of the FBOs to work in gang controlled neighbourhoods include their evangelical identity, the acceptance by and independence from gang and government, their longer-term engagement in selected neighbourhoods, and the combination of social interventions that have a local impact (education, healthcare) with the promotion of moral values. The paper also discusses some of the dilemmas and limitations of these kinds of approaches.


Introduction
Latin American cities continue to display relatively high levels of violence, and in 2013, 42 of the fifty most violent cities in the world were in Latin America (Chioda 2017: xi). A growing number of studies discusses the possibilities to mitigate chronic violence in urban contexts where armed non-state actors have taken control Muggah & De Boer 2019;Salahub et al. 2019). While there are interesting examples of violence reduction in Latin American cities, such as Medellin, and Rio de Janeiro (Muggah & De Boer 2019), there are no quick fixes. The complexity and tenacity of the drivers of chronic violence -a situation where high levels of physical violence persist over a longer period 1 -produce long-term challenges and dilemmas for policy makers (Adams 2017: xvi).
The dilemmas of responding to chronic violence are felt in El Salvador, a country with consistently high levels of violence over the past century or so (Walter 2016). After the civil war came to an end in 1992, street gangs -'pandillas' or 'maras' in Spanish -started to grow and developed a strong capacity and willingness to use violence. While the phenomenon of violence is complex and should not simply be attributed to specific actors such as gangs (Moser and Winton 2002), the policies of the Salvadoran governments have largely concentrated on gang violence. The government's focus on repressive approaches has been questioned by many authors, as it made scapegoats of gangs, and inadvertently contributed to the transformation and adaptation of gangs, not to its elimination (Cruz 2011;Wolf 2017). Efforts to dialogue with gangs -in particular during a gang truce in 2012 -delivered impressive results in terms of violence reduction, but were at the same time criticised on legal and ethical grounds (van der Borgh & Savenije 2019). A rash of social and preventive programs (such as sports programs) for 'youth at risk' has been implemented in the poorest and marginalised

Gangs in El Salvador
In the 1990s, when El Salvador's civil war had come to an end, the street gang phenomenon quickly started to grow. Gangs already existed before and during the war, and most new gang members were born before or during the war. Only to a limited extend did gangs absorb former soldiers or guerrilla members. The growth of the gangs was spurred by the deportation of illegal Salvadoran youth from the US with criminal records, who often had experience in American gang structures (Arana 2005). As a result, local maras that had operated relatively autonomous increasingly identified with and became cliques of the large gang structures of MS13 or 18 th Street Gang (Barrio 18, which later split into two factions). Over time, gangs became a powerful armed actor in many urban, marginalised neighbourhoods (Arana 2005;Savenije 2009). In El Salvador and in the neighbouring countries Honduras and Guatemala, the phenomenon is particularly violent, with local cliques of opposing gangs fighting each other, and using coercion and violence vis-à-vis residents, government officials and entrepreneurs (van der Borgh 2020).
There has been ample discussion about the causes of the gang phenomenon. At the 'macro level', the failure of the state to build a legitimate presence in marginalized areas and the unsuccessful post-settlement security reforms (in particular the police reform) played an important role (Cruz 2011). In addition, a number of risk factors explains the proliferation of street gangs. These include social exclusion and lacking economic opportunities, a culture of violence, and community and family disorganisation (Cruz 2007;Savenije 2009). Although gangs are not necessarily a direct or rational response to social and economic marginalization, gangs meet certain needs of youth, such as a social identity, protection, and some income. At the same time, these groups pose serious challenges to the security and livelihood strategies of residents, particularly the lower socio-economic strata (ICG 2017).
Indeed, street gangs have institutional characteristics as they have developed 'relatively permanent arrangements of behaviors, roles, norms, and values that structure aspects of human activity in patterned ways' (Eisner 2009: 48). They have a strong capacity to reproduce and to adapt themselves to changing contexts. In this regard, the massive incarceration of gang members that started around 2003 led to extortion schemes that are often controlled by imprisoned gang leaders (Gomez Hecht 2013). Street gangs are quite different from other Latin American non state actors, such as the Mexican cartels, who are involved in much more lucrative criminal activities. Most gangs are involved in criminal activities, such as extortion of businesses and residents (van der Borgh 2019). However, most gang members do not get rich from their involvement in gangs but remain poor (ICG 2017). 4

Dealing with Chronic Violence and Room for Maneuver in Micro-level Armed Regimes
How to deal with chronic violence is complex and can become a highly politicized and contested topic. A distinction can be made between responses that focus on the actors held responsible for violence (such as gangs), and the responses focusing on the prevention of chronic violence and insecurity. The first type of policy usually focuses on the most visible armed actors that have built up power positions, and can variously include repression, reintegration and/or negotiation. In many cases, these policies leave out violence used by other actors, such as the police and the military, and violence taking place in other spheres of life (the household). Efforts to include these sectors, or to use 'softer approaches' (such as dialogue with gangs, or effort to reintegrate) can meet strong resistance and repressive and aggressive (illegal) responses to perpetrators of violence can be presented as the only way out. 5 The second type of policy response goes beyond the focus on ' armed actors' and seeks to deal with the causes and reproduction of chronic violence in cities or communities. This preventive or social response can take into account actors using violence, but more often, the focus is on the broader group of residents affected by chronic violence.
The need for preventive and social approaches is broadly recognized. Social and preventive interventions come in different forms, including criminal justice, public health, urban renewal, citizen security, conflict transformation and interventions directed at the 'rebuilding' of social capital (Moser 2004: 12). While many preventive interventions focus on individuals (youth at risk) or violent spaces (the community level), others focus on the national level, such as the security sector (Muggah & de Boer 2019). The choice for a preventive approach is always informed by assumptions about societal change, either implicitly or explicitly ('theories of change'), and the assumptions can differ. On the one hand, it is argued that a focus on the most violent spaces and groups is the most effective strategy to deal with chronic violence (Abt & Winfield 2016;Chioda 2017). On the other hand, a focus on coping and resilience by disadvantaged communities has been criticized for neglecting the broader political context. 6 For example, Reid-Henry and Sending (2014: 436) criticize the "humanitarization" of urban violence, where 'technical policy language […] serves to downplay rather than elevate the political and structural stakes involved'. The two arguments are, however, not necessarily at odds. Recognizing the structural and complex causes of violence (Adams 2017: xvi) takes a more pragmatic approach and stresses the need to strengthen human and social resilience where and when strategic opportunities exist. 7 However, this is easier said than done. In places where non-state armed actors have built up strong power positions, the space of intervening agencies can be quite limited (Cano 2012: 48). This raises questions about the strategies used to identify or forge space to work in the most difficult environments. The question how intervening organizations use or create 'strategic opportunities' has only received scant attention in the academic literature. NGOs face difficult questions about how to relate to local non-state armed actors and government agencies in what Arias (2017) has called the 'micro level armed regimes'. The space to intervene largely depends on the actions and policies of other actors, most notably the non-state armed actors and the state itself. Non-state actors can prohibit or try to control the entrance of NGOs, but state action can also have an impact on the work of NGOs. In this regard, Machado (2016) shows how the 'pacification' schemes in Rio de Janeiro led to a reconfiguration of the relations between state organisations and NGOs, which affected and destabilised the local informal order. This is where the question of the legitimacy of intervening agencies and local orders comes in. Abst & Winship (2016, 19) argue that legitimacy is an important element of effectiveness in preventive programs, pointing at positive feedback loops between formal control (police) and informal social control (of the community). Their assumption is that a new local social contract between the state and marginalized local communities should be forged. However, as will be discussed in the next section, for FBOs working in gang controlled neighborhoods in San Salvador, the endorsement of gangs to work in these neighborhoods is of primary importance. And, since gangs oppose military and police, FBOs must keep a distance from these entities. Hence, the need to be accepted by local gangs implies that FBOs will be reluctant to team up with state law enforcement agencies.
In this regard, it is important to note that government entities, including those that are supposed to benefit residents -like healthcare and education -are not necessarily accepted by gangs. The vast majority of the staff working for local governments and NGOs that were interviewed during a study in 2017 and 2018 reported that they need to deal with gangs, either indirectly, for instance, through local neighbourhood associations that are in touch with street gangs, or directly by negotiating access to certain neighbourhoods (van der Borgh & Abello Colak 2018: 6). Thus, community interventions of NGOs and government agencies involved in social service provisioning have to take into account the local gang, and often need the permission of gangs to enter the neighbourhood and to implement services.
As mentioned before, religious organisations tend to be respected by gangs and as a result their possibilities to intervene in gang controlled neighbourhoods are better. But this is not to say that churches and FBOs have a ' carte blanche' to do whatever they want. The degree of legitimacy of FBOs vis-à-vis gangs is fragile, and the volatile local and national context requires constant manoeuvring and interaction with gangs.

Community work in Los Abetes and El Valle
This section discusses how two FBOs developed their interventions in two different gang controlled neighbourhoods in the AMSS. The focus is on the room to maneuver of organizations that seek to address a number of problems of local residents, with the consent of the local gang. After a brief introduction of the methodology used in this study, the characteristics of local gang control are discussed. The section moves on with a discussion of the navigation strategies of the FBOs and their contribution to community development.

Methodology
The research is based on data from fieldwork in several municipalities of the metropolitan area of San Salvador (AMSS) and in a number of provincial towns in the period between 2014 and 2019. In addition, a literature study into interventions in gang-controlled neighbourhoods was conducted. The research project dealt with urban (in)security, and during the research, a large number of local government officials, church leaders and NGO staff were interviewed. As a result, the research also generated a considerable amount of data about different social and preventive programs and projects implemented in gang-controlled neighbourhoods.
From this scan, it became clear that there is a wide variety of interventions in gang-controlled neighbourhoods and 11 cases were identified. The majority of the interventions of government agencies, NGOs and international programs focus on youth at risk. There were a few smaller initiatives of local churches that focused on gang members themselves. Most of the preventive programmes involve sport and cultural activities and educational (after school) projects. Two organisations are evangelical NGOs that intervene in a selected neighbourhood, over a longer period. These are the ones selected for this study.
It was agreed with directors and staff of the selected FBOs that the names of the organisation, the neighbourhood and interviewees will not be disclosed. Both FBOs work in municipalities that belong to the ones with a very high gang presence. FBO A works in a neighbourhood that I call Los Abetes, and FBO B in a neighbourhood that I call El Valle. Interviews were conducted with staff of FBO A during research visits in 2017, 2018 and 2019. In addition, short field visits were made, and conversations with local residents and a gang leader were conducted. The experience of one other FBO (FBO B) was recorded in 2018 with the director of the FBO. This interview took place after a first round of interviews in and about the work of FBO A in Los Abetes.
The two FBOs work in neighborhoods that were founded before the civil war and belong to the poorest areas of the AMSS. Los Abetes consists of small one or two-story houses connected through a labyrinth of small alleyways, which are not accessible to vehicles. The neighborhood can be reached through a number of entrances that are monitored by youth related to the gang. The main sources of subsistence of the residents are in the informal sector, such as informal street trade and garbage collection. Just like other marginalized neighborhoods of San Salvador, the problems of social exclusion and violence are strongly intertwined (Savenije & Andrade Eekhof 2003). Domestic violence is and remains an important phenomenon, and with the emergence of gangs, the already existing situation of poverty and local insecurity became more complex.
Although FBOs A and B didn't cooperate and the directors didn't seem to know about each other's experiences in the two neighbourhoods, there were striking similarities between their experiences. This points to a ' certain logic' of these kinds of longer-term FBO interventions in gang-controlled neighbourhoods. While the discussion of these cases provides insights into the dynamics of the interventions, the findings are preliminary. More research is needed to gain a deeper understanding of the possibilities and limitations of these kinds of interventions. Additional interviews with local stakeholders (gang members, local residents, representatives of local community organizations, the municipality) can provide more detail about how the interventions played out and were perceived. However, the interviews with staff of the FBOs were of high quality. Staff had obtained intimate knowledge of the developments in the neighborhoods and were willing to reflect on their experiences over the past 10 years or so.

Gang control
Los Abetes and El Valle already had a certain reputation of being a hotbed of criminal groups before the civil war broke out, and the gang phenomenon quickly grew after the civil war. The presence of the local gang in Los Abetes goes back to the 1980s and 1990s, and many families in the neighborhood have members in the gang (many of them in prison). In the period 2017 and 2018, NGO staff working in los Abetes reported that the local gang had weakened considerably as a result of the extremely repressive approach taken by the government. This was also acknowledged by a local gang leader, Nino (a man in his early 30s) who recognized that life with the gang had become extremely hard. 'You have to flee for the police, and to take revenge on the other gang.' FBO staff reports that despite changes in the strength of the local gang 'gang life and control continue'.
The staff of FBO A acknowledged that for gang members that are tired of the gang, the only way to leave is to join an evangelical church. It was also stressed that it is easier to take this step for youth close to the gang, but not yet a member of the gang. FBO B, working in El Valle, explained that several gang members had joined the local church and had been allowed to leave the gang. But not all gang members want to leave. An FBO worker in Los Abetes explained that she tells gang members that God can offer them a way out, 'but they said in response that this is no longer possible, that they have lived their life, they can die, they know they are bad.' Interestingly, older (veteran) gang members often don't want their kids to follow their example, and this is why FBO A is allowed to work with their and other kids in the neighbourhood. Even though many gang leaders want a different future for their children, recruitment doesn't stop. Young children continue to be interested in the gang, and it seems that gangs have no problem in recruiting new kids (of other families than their own). In the words of one FBO worker, 'it is a nasty chain'. 8 A staff member of FBO A explained that the gang controls the local neighborhood: 'They are the law'. This is not to say that the gang actually governs the local neighborhood in the sense that they take care of local residents; but rather that they are the local organization that has the armed power to enforce its will. Over time the relationship of the gang with the neighborhood has changed. Nino, who had joined the gang at a young age, argued that gang life had become more structured over time. In the old days, they lived the crazy life, 9 robbing wherever they wanted and using violence against different people, including local residents. A local FBO worker remembers that in the first decade of the century, a number of older leaders were arrested that had done 'terrible things'. By the end of the decade -around 2010 -a set of internal rules came into existence that started to govern the behavior of local gang members and their relations with local residents.
Nino explained a number of the rules that govern local gang life. Among other things, he mentioned that it is not allowed to use violence vis-à-vis local residents or to harass girls, and gang members are not allowed to drink alcohol (marihuana is allowed). In addition, FBO staff presumed that local residents are not extorted by the gang, but that shopkeepers probably had to pay. The local gang structure functions in close relation with imprisoned leaders. One staff member told how he sometimes is connected with someone in prison, a chief. 'They let you know that they know what you do and where you work. They make it clear that they rule. You have to be very careful.' Local residents have to obey the code of the street (to hear, see and remain silent), and residents that are not trusted by the gang can be expulsed from the neighborhood. When people want to sell their house, the gang must be informed and can object against selling. There were also stories of the gang confiscating houses, using the house themselves or selling. Also, the gang decides where new houses can be constructed. Despite the control of the gang, its legitimacy tends to be low, according to an FBO worker.
On the other hand, there is a high degree of co-existence. People have lived together and know the gang, and local residents feel more or less at ease with the situation in the community. Instead, the police and military that enter the neighbourhood during invasions are feared, especially by non-gang youth who are harassed by the police. Over the past years, several youths have been hit by bullets of the police. One FBO staff member told about the example of a 14-year-old boy who was not a member of the gang, but worked for and 'flirted with' the gang who was seriously injured. 10 A worker of the day-care centre told that young children often tremble when they see police. 11

Working in gang controlled territories
FBO A started to work in 2004 in Los Abetes, and FBO B in 2009 in El Valle. In both cases, the directors of the FBOs, who were also pastor, contacted local community representatives and (through them) local gang leaders. The director of FBO A, who had no church presence in the neighborhood and was not known by the local gang leaders before 2004, had a four-hour conversation with four leaders. The reason for the local gang leaders to accept the presence of the FBOs was that they wanted a different life for their children. This was van der Borgh: Faith-Based Organizations, Community Development, and Room of Manoeuvre in Gang Controlled Neighbourhoods in San Salvador 149 the reason for them to endorse the work of the FBO and to give permission to use a local community center. However, gang leaders in Los Abetes made it clear that they were in charge of the neighborhood and that they would closely monitor the activities of the FBO. Thus, during the first two years, they frequently entered the local meetings of the FBO, which was experienced as rather intimidating. 'They entered our meetings, sat down, didn't talk, and looked us in the eyes'. Reportedly, after a couple of years, this stopped.
Interestingly, both FBO A and B reported that there was some resistance against their work from local residents. They thought that it was not ethical to work with the gang. However, both FBOs made it clear that they did not work with the gang itself, but with all local families, particularly with mothers and children in the neighborhood. FBO B also reported that it had undertaken some work in the field of reconciliation and had facilitated dialogue between former gang members and residents that had suffered under gang violence. In this regard, the fact that the FBOs had a religious identity and combined spiritual and social work not only provided it with some legitimacy vis-à-vis the gang but also generated confidence among local residents. The staff of both FBOs stressed that even though gangs are explicitly non-religious, most gang members have grown up in Christian families and with Christian values and -as one FBO staff member said -'they know the word of God'. The church is probably the only institution that still has some respect. It should be stressed, however, that while the religious identity is very important, it is possible for nonreligious NGOs to work in gang-controlled neighborhoods. The work of a non-religious NGO that promoted local youth leadership and human rights in a different part of town explained that its work was endorsed by local gangs most of the time. 12 Ongoing contact with gang leaders and the provision of information about the projects of the FBOs and the staff working in the neighborhood is key to secure the continuing support of gangs. As local gang leadership tends to change (when leaders are imprisoned or killed), support can be problematic. New gang leaders may distrust the organization or are simply not interested in its work. This happened in one of the other neighborhoods where FBO A worked. In this place, where gang presence was relatively recent, a new gang leader opposed the work of the FBO, leading to the decision of the local promotor of FBO A to leave. In this regard, the presence of a more stable local gang leadership (veterans) and the support from imprisoned leaders with contacts in the neighborhood seems to be an important factor in guaranteeing the long-term presence of FBOs. In El Valle, the permission to work in the community was given by four local leaders, but by 2018 three of them had passed away and one lived in the US. In los Abetes, imprisoned gang members supported the interventions of FBO A and had intervened (by telephone) when a problem about the use of a local accommodation had emerged.
Regarding the relationship with the local gang, the strategy of FBO staff seems to be a combination of acceptance of the local power of the gang, a distance to the local gang, and communication when necessary. All staff emphasised that they had to abide by the rules of the local gang, 'they are in charge'. FBO staff also indicated to be extremely careful and respectful towards gangs. This implies that staff of the FBO has to abide by the unwritten code to 'see, hear and remain silent'. For example, when I asked local church leaders in a different neighbourhood of AMSS to conduct life histories with local residents, this immediately raised concerns, and the answer was 'we have to ask them' (the gang). In a conversation with two staff members of FBO A, they explained, 'we know a lot of things, but cannot talk about it'.
Nevertheless, problems can and do emerge. A staff member reported that when the gang felt insulted and threatened when an article was published in a national newspaper of the work of one of the FBOs that was not appreciated by the gang. The FBO decided to apologize publicly. The example shows that trust is fragile and that the FBO should not disclose information about the gang and should respect the presence and rules of the local gang.
FBO A and B focus on preventive action for mothers and for youth that is not involved in the gang. However, both FBOs also have the capacity to dialogue and work with gang members. In El Valle, FBO B had conversations with local gang members about the need to lower the homicides, to stop local recruitment and even requested the gang to hand in weapons. A number of local gang members left the gang, and local recruitment was reduced, but other gang members expressed they would never abandon the gang, nor hand in weapons. The director of FBO B also recognized that transformation of gang members is very difficult. He noted that he did not necessarily believe gang members when they told him that 'they don't extort, don't kill, and that others are blaming them', but it was nevertheless possible to make agreements with them. Importantly, despite this, the director sustained that the work of the FBO had gained the respect of the gang.
A similar pattern exists in Los Abetes. Over the years that FBO A worked in Los Abetes, the local leadership was replaced a couple of times when members were arrested or killed. A local staff member says that she respects the local gang, but also tells them that what they do is not correct. In addition, local gang members have repeatedly insisted that the FBO should not leave. After seven years of FBO work with children, local gang leaders approached FBO A, expressing their appreciation and respect for the work of the FBO and asking if they were also willing to work with them. This led to an income-generating project with local gang members (see the next section).

Community development in Los Abetes
As mentioned, FBO A focuses on the needs of women and children and places emphasis on educating the 'next generation'. This is realized by a combination of activities, including a day-care center, after-school programs for children, scholarships to attend high school ('bachillerato'), vocational training courses, home visits, food aid, and a local medical center. The educational projects are the backbone of the NGOs intervention in the neighborhood. The idea is that day-to-day contact with local youth, including the children of gang members, will lay the foundation for moral and social change in the community. The after-school program offers a combination of biblical readings, play, and homework support. The FBO stresses that its work has moral, spiritual, and practical dimensions. A part of the activities with children focuses on biblical readings and biblical principles, while a lot of attention is paid to the other activities. The staff of the organization expresses the mission both in lay terms ('We want to show them that there is a different way of life. They don't have to end up like their parents.') and in evangelical terms ('The challenge is to make them understand that God has a better future for them. We want to reach the hearts and minds of the people'.) In addition, a local medical center is run by the FBO, where residents from Los Abetes and residents from some neighboring neighborhoods receive medical assistance (visitors are asked to contribute one dollar per consult.) The medical center is supported with funds from international organizations, and is, just like the day-care center, owned by the FBO. In El Valle, there is no local health clinic ('health workers don't dare to enter the neighborhood'), and FBO B supports a local church member who has received basic training and has medical supplies in his home.
FBOs do have access to but do not necessarily control the real estate they use in the neighborhood. In Los Abetes, the permission of the gang to use local buildings is essential. This implies that the gang allows FBOs to use certain buildings or community centers, while access to the building is controlled by the gang. In other cases -such as the medical center and the day care center in Los Abetes -the buildings are owned and controlled by the FBO.
It should be stressed that government services like education and healthcare are not necessarily respected by the local gang. This was also the case in Los Abetes. While the clinic serves residents, the government had abandoned it years ago because of burglary and a lack of security of local staff. When FBO A reopened the medical center and contracted a medical doctor, it counted on the consent of the local gang. In the day to day relation with gang members, challenges do come up, however. Local gang members are not excluded from consults, but may prefer to send family members to the hospital for medicines or ask that medical staff visits them at home. It had also occurred that gang members hid in the hospital when the police and military intervened in the neighborhood. This may put the security of the medical staff and visitors and the very existence of the local medical center, at risk. The capacity of the FBO to take this up with local gang leaders is of key importance for the center to continue. Indeed, the continuation of a local medical center requires the capacity to enter into dialogue with local gang leaders when the need emerges.
Without the respect of the local gang and the continuation of the respect of the gang, FBO A would not be able to work in Los Abetes. Indeed, respectful treatment of gang members is key in this contact, but this does not imply that FBO staff endorses the gang as such. The strategy seems to be one of 'non-judgment'; gang life as such is not accepted, but not condemned either. Gang members are treated with respect, and it is stressed that 'God loves them' and that they can change. It should be stressed that the FBO can only do its work and, when necessary, enter into contact with the local gang, because the gang respects it. That respect is at least in part based on the tangible results that the FBO delivers, combined with its long-term presence and relationship with local community members.
The respect of the gang for the work of FBO A also became clear when gang members asked the FBO to start a vocational training project with them. This led to an initiative with gang members, where the gang constructed a local workplace, and the FBO assisted the gang with an instructor. The instructor focused on vocational training but started each day with a prayer, a lecture from the bible and a discussion about the meaning of these texts. The local gang owned the building; the instructor hired by the FBO did not have keys Community Development,and Room of Manoeuvre in Gang Controlled Neighbourhoods in San Salvador 151 to the property. The project continued for over a year, but it ended when the police raided the workplace. After that, the project continued with local adolescent youth that is not with the gang. The experience shows that there is a difference between working with children and youth that live in gang controlled areas (including those that are ' close to the gang') and youth and young adults that are already part of the gang. It is fair to say that vocational training and income generation projects are easier with gang members that have taken a distance from the gang (calmed down, or joined a church). Projects with active gang members easily run aground, as the logic of active gang membership (fast money, internal dynamics and culture of gang life, including revenge against other gangs) may be hard to combine with the demands of entrepreneurship.

Discussion
The programs of the FBOs discussed in this paper provide a mix of social work, moral support, and services. While the paper does not aim to evaluate the outcomes and impact of the interventions, it is fair to say that the FBOs contribute to the improvement of the condition of the most vulnerable groups in these neighborhoods (healthcare, child care) and the creation of a different moral order. The intervention contributes to local resilience and may lay the foundation for more secure and more peaceful communities. Starting these initiatives and maintaining them not only has a price tag but also takes time. Moreover, these processes will have a limited impact when they are not complemented by changes in the political and economic realms. In this regard, a key challenge remains the limited socio-economic perspectives for youth that have participated in the programs of the FBO. Despite examples of youth that have gone to university and found employment, this remains a serious concern, and the limitations of the job market are a challenge that clearly cannot be addressed by this type of intervention alone.
The experience described in this paper contains some lessons about the possibilities to work in micro level armed regimes. Firstly, it is fair to say that the fact that the FBOs referred to 'the word of God' to explain their work resonated with gang members who ' are choosing not to follow God in their actions' (Offutt 2019: 432). Thus, the evangelical identity of the NGOs facilitated the access of the organization to gang controlled neighborhoods. While it is a well-known fact that evangelical churches abound in gang-controlled neighborhoods and focus on personal conversion, this paper shows examples of FBOs that created spaces to address some of the drivers of chronic violence. In a context where local gangs often distrust external organizations and where government services have suffered under the situation of insecurity or withdrawn altogether, FBOs succeeded in (re)building these services and generating a degree of trust, protection and respect for these services.
Secondly, FBOs don't have a carte blanche to work. The local gangs watched the FBOs, and the work of FBO A in Los Abetos shows that trust between the gang and FBO had to be forged over time. This type of intervention is not able to break the local armed control of gangs (Cano 2012: 47). However, FBOs do have a certain room for manoeuver when transforming the gang phenomenon in the longer term. While the FBOs have to abide by the rules of the gang, the programs focus on and promote a different moral and identity -Christians, ' citizens', and entrepreneurs -that is essentially at odds with the gang identity. Interestingly, gang leaders endorse and may even respect programs that educate and socialize children and youth (often family members of gang members) to prevent them from following their example.
Thirdly, while the religious identity of the organizations was important in getting the chance to work in the neighborhoods, another important factor seems to be the fact that FBOs can work around the local web of relations that exists between gangs, government agencies, political parties, and local community organisations. While the FBOs abided by the rules of the gang and maintained contact with the local leaders, they also were keen to maintain their autonomy. This seems to be one of the most difficult parts of intervention in local level armed regimes, as gangs may try to influence, control, or infiltrate churches, local governments and community organizations. It seems that the longer-term presence of the FBOs and their tangible contributions to local services, plus transparency about their activities, generated trust of the gang. It is important to note that the FBOs did not coordinate with government agencies. In this regard, it was particularly important to stay away from contact with the national police, which is highly distrusted by gang members and residents. Especially so in the years that extra-judicial killings of gang members increased (PDDH 2019).
The NGOs intervene in policy areas that can be considered the responsibility of the government. However, for governmental organizations, this kind of work will be more difficult to implement. Governments are supposed to 'be in control' over a certain territory and a population, while for NGOs there is no need to live up to that image. Also, government agencies encounter serious problems to reach gang-controlled neighborhoods as they don't have a single presence, and as a result, the relations developed by different state agencies (such as the local government and the police) strongly differ. In the words of one FBO staff member: 'The state passes by, giving a few things, prosecuting some people, sometimes destroying things, sometimes trying to construct things. But they don't know what they are doing, and they didn't stay.' 13 While the experience of the FBOs will be hard to replicate by government agencies, experiences in other places show that coordination between local government agencies, community organizations and FBOs may lead to interesting synergies. In one municipality, the local government had successfully initiated cooperation with FBOs, linking educational (after school) and sports activities. 14 This municipality had a higher socio-economic standard, and gang activities were concentrated in selected neighborhoods only.
Lastly, the strategy of FBOs to gain a territorial presence (rather than 'territorial control') and to work around the gang, may inform some of the government policies in the future. It should not be forgotten that in the case of El Salvador, the access and room of manoeuver of government agencies offering services (such as schools and health centers) is often quite limited. In other words, the very presence of the government is contested and resisted. In such a context, 'being present' is an art in its own right, and something can be learned from the work of FBOs. This is not a plea to simply accept the gangs and their power but to think creatively about the possibilities to maneuver government agencies and their possibilities to deliver at the local level.