2012/Belize/Engagement Report

Bureau of Conflict and Stabilization Operations (CSO)
U.S. Department of State

Table of Contents

Belize Engagement Evaluation Report: Executive Summary

I. Introduction
II. Engagement Background
III. Evaluation Background
IV. Baseline for Longer-Term Outcomes and Impact
V. Findings

  1. Mediation
  2. Community Dialogues
  3. Sustainability and Networking
  4. Environment & Transferability

V. Recommendations


  • Methodology
  • Mediation Data
  • Baseline Data for Longer-Term Outcomes and Impact

Engagement planning documents available from CSO:

  • Engagement Strategy
  • Results Frame
  • Theory of Change Map and Narrative


Belize Engagement Evaluation Report: Executive Summary

CSO piloted a project May-July 2012 to establish mediators, mediation trainers, and community dialogues in South Belize City, at the request of the government organization RESTORE Belize. The pilot pursued three objectives: 1) Increase local capacity to reduce gang violence; 2) Strengthen community resilience by linking civil society and leaders; 3) Develop CSO tools for reducing conflict in the context of gang warfare. Longer term, the goal is to reduce gang presence and thus vulnerability to Transnational Criminal Organizations (TCOs). This report measures early outcomes one month after the engagement, sets a baseline for further results, and recommends actions to achieve impact and sustainability.

Key Findings

1. Mediation is having significant effects, is socializing well, and shows promise for expansion. Mediators and disputants rate mediation as “very strongly effective” in general and “effective” with gangs. Nearly all non-gang mediations, and most gang mediations, produce lasting agreements. Most agreements prevent violence, especially in gang mediation. Mediators are unanimously “very strongly interested” in continuing, and 83 percent of disputants are interested in becoming mediators.

2. Community dialogue is resonating but making slow progress. The dialogue in St. Martin’s community has empowered dialogue participants but has not affected the broader community. However, plans to collect data on the community and elect officers should lead to activities with broader impact in coming months. No other communities have established community dialogues, but broad interest and some preliminary action among mediators indicate the concept has gained traction.

3. Indications are mostly positive for sustained mediation and community dialogue. Sustainability requires broad local ownership, and multiple institutions developed interest in mediation, forming a nascent network for mediation and community dialogue. Trainers plan to train staff and peer mediators at several institutions. Inter-institutional challenges currently limit mediator availability, but this is being addressed. RESTORE Belize is strongly dedicated. Though it lacks dedicated staff to manage the program and funds to expand it, it is seeking funding from private and public sources. The prime minister’s office is confident about solving funding and inter-institutional challenges.

4. This context was particularly favorable. This context had significant advantages, including a strong local partner and strong political will from the prime minister. Also, the Belizean gangs are small, poorly organized, and not yet dominated by TCOs, though engaged in drug trade with them. Other contexts are likely to be more difficult, and efforts elsewhere should be tailored to these factors.

Key Challenges with Recommendations

1. Unresolved inter-institutional issues could undermine the program. RESTORE Belize must complete MoU’s to ensure that agencies support the mediation program. The prime minister’s office should create incentives for directors to support employees mediating as part of their jobs.

2. Additional resources are necessary to sustain & expand the program. RESTORE Belize should create dedicated staff positions for this program. CSO and U.S. Embassy Belmopan should advocate for public and private resources for RESTORE Belize and the Belizean Association of Mediators.

3. Community dialogues have not yet reached beyond direct participants or expanded to new communities. RESTORE Belize should clarify roles, goals and timelines with relevant partners. RESTORE Belize should create a forum to promote community dialogue, share ideas among facilitators, and expand training, especially for potential facilitators native to target communities.

4. There is more need for gang mediation than the Conscious Youth Development Program (CYDP) can address. RESTORE Belize should develop more gang mediators, preferably civilian residents of gang-ridden neighborhoods, and encourage CYDP to train non-CYDP mediators in gang mediation.

I. Introduction

CSO’s Belize engagement in mid-2012 sought to capitalize on the window of opportunity created by the September 2011gang truce to catalyze developments that can prevent the occurrence of the “balloon effect,” in which counter-narcotics efforts elsewhere in Latin America have pushed narco-trafficking operations into new locations, prevent transnational criminal organizations (TCOs) from exploiting local gangs, boost community resilience, and reduce violence in Belize City. The Belize engagement tied into CSO’s larger North Central America strategy and regional goals.

Although it is too early to be able to assess the macro-level impact of CSO’s work in Belize, this evaluation provides timely feedback on the early outcomes of CSO’s efforts. This analysis suggests that the engagement’s near-term theory of change – that the creation of a sustainable mediation and community dialogue system in Belize City will resonate and grow, build a sense of hope and empowerment, and reduce violence at the micro level – appears to be accurate. The mediation component is generally on track and reaching gang members and the neighborhoods from which they emerge, although implementation of the community dialogue component has thus far been slower than anticipated. This report identifies ways to support the success of CSO’s investment in Belize in the future, highlights specific areas of attention moving forward, and provides insights on the potential applications of this pilot program’s approach to other engagements.

II. Engagement Background

The CSO engagement in Belize in May-July 2012 was a pilot project that sought to establish mediators, mediation trainers and community dialogues in South Belize City that could jointly reduce gang violence and presence and promote community resilience, at the request of the Government of Belize. Within an overarching, regional long-term goal to enable local communities to build resilience against TCO activity via local gangs, the Belize engagement was designed to achieve three primary objectives:

1. Increase local capacities to reduce violence and homicide rates.

2. Strengthen community resilience by linking neighborhood civil society organizations and local leaders across sectors.

3. Develop and strengthen CSO tools for conflict prevention and mitigation in the context of gang warfare that can be applied elsewhere.

CSO worked closely with RESTORE[1] Belize, an organization established by the Prime Minister to address the gang issue, which is committed to coordinating and sustaining these activities. Other partners include the Conscious Youth Development Program (CYDP), an organization under the Ministry of Human Development, Social Transformation and Poverty Alleviation that specializes in gang intervention, and the Center for Community Resource Development (CCRD), a community development NGO in the St. Martin’s neighborhood. The engagement also partnered with the Belize Central Prison, Youth Cadets, Youth Hostel, Community Rehabilitation Development (CRD), and three high schools. Details of the engagement design are provided in the Annex.

The engagement team delivered three mediation courses which trained 36 mediators and conducted 72 co-mediations with the trainees. It increased the number of gang mediators from 2 to 11. The CSO team also trained 10 new mediation instructors to help the program sustainably expand. They helped teach the third mediation class and also a peer mediation class at a high school, training 29 students. In addition to the mediation activities, the CSO team worked with CCRD to establish a community dialogue in the St. Martin’s neighborhood of Belize City. The team also helped establish new relationships between key Belizean institutions, including all the partners mentioned above, thereby laying the foundation for a new, cross-cutting network for mediation and locally-driven change. Finally, it worked with RESTORE Belize to build its capacity to manage and expand the program.

The overall vision behind CSO’s Belize engagement is that the creation of a sustainable, locally-owned system comprising mediation and community dialogue capacities in Belize City will reduce and prevent gang violence, membership and activity, by addressing underlying security, psychosocial and economic drivers. Each demonstration of success at resolving conflict, preventing violence, building relationships and community, solving community problems and promoting development will contribute to shifting the local narrative to one of empowerment and hope in the medium term. Combined with building Belizean training capacity, developing the network among mediators and community dialogue facilitators, and strengthening institutional capacity to manage the project, the interest generated by demonstrated success will result in growth of mediation and community dialogue activities and capacities, broadening and deepening participation and progress. In the short term, effects will be concentrated on the dialogue participants, but through accumulated and secondary effects as the project matures, medium-term macro effects on broader communities should become apparent in lowered violence and improved attitudes. In the long run, in combination with other efforts, this should help to foster economic development. The lowered violence and increased social and economic wellbeing of Belizean youth will reduce gang membership and influence in Belize City. This will ultimately increase communities’ resilience to TCOs because they will be less able to act via local gangs, youth will be less vulnerable to direct recruitment by TCOs, and communities will be able to react more effectively to TCO attempts to establish a presence or operate locally.

At the time of this evaluation, the results of CSO’s Belize engagement are in the early outcome stage. It will be months or years before the long-term impact of CSO’s work can be expected to show in macro indicators such as more hopeful and empowered community narratives, lowered overall violence, and reduced gang membership/activity. However, early outcomes can already be seen among individuals in the current attitudes of participants in mediation and community dialogue, in effective problem solving and violence prevention through mediation, and in emerging actions taken by local stakeholders to sustain and promote the use and growth of mediation and community dialogues.

III. Evaluation Background

This evaluation of the Belize engagement was conducted one month after the end of the engagement by two staff members from the CSO Office of Learning and Training. Engagement team members and members of the larger Operations Team were consulted on the design of the evaluation, but were not involved in collecting data or conducting the analysis. The two-person data-collection team had five days to collect data in Belize. The evaluators interviewed 45 individuals, including mediators, trainers, disputants, community dialogue participants and administrators, Embassy staff, and personnel from relevant Belizean institutions.[2]

This evaluation examines CSO’s pilot effort at developing mediation and community dialogue capacities in order to catalytically increase community resiliency against a strategic threat which preys on disempowered communities. This evaluation focuses on outcomes and impact of the project, rather than its implementation and outputs. It has six main purposes:

1. Measure initial outcomes

2. Establish a baseline for longer-term outcomes and impact

3. Provide a methodology for future monitoring and evaluation by host-country partners

4. Assess challenges moving forward in scaling the project up to achieve the desired impact, how local partners can meet those challenges and needs for further assistance

5. Assess sustainability

6. Inform future engagements

IV. Baseline for Longer-Term Outcomes and Impact

In the longer term, this project is aimed at effects on more than simply the direct participants in mediation or community dialogue. It is intended to catalyze a process of broader social change that will reduce the presence, violence and influence of gangs and increase resiliency against TCOs that would operate via local gangs. We do not expect to see significant effects at this level at this time, but we have measured a baseline on some indicators that can be used to show effects at this level in the future.

Our choice of variables is based on three main factors leading people to join or cooperate with gangs. One is safety – joining a gang may be seen as necessary for protection against that gang and others. Another is economic – in the high unemployment of South Belize City, gangs provide income through drug trafficking and protection rackets. A third is psychosocial – for youth with little direction, broken families, no hope and low self-esteem, gangs provide a sense of group identity, pride and routine. Several individuals interviewed in the evaluation process noted that gangs provide a pseudo-family structure for children and youths lacking a social support system in their lives. These factors reflect our theory of change. First, mediation and community dialogues reduce violence, thereby addressing the safety factor. Secondarily, this creates a cascading effect by providing a more favorable environment for other actors to spur economic development. This can in turn reduce the economic incentive for joining and cooperating with gangs. Both mediation and community dialogues address the psychosocial reasons for gang affiliation and activity by connecting and empowering people, rebuilding relationships, and developing problem-solving and communication skills – all of which provides positive substitutes for the psychosocial role played by gangs. These same developments can also help young people find satisfactory employment, further undermining the economic driver of gang affiliation.

Drawing on these elements of the theory of change, the evaluation team measured indicators of the following concepts: gang violence; feeling safe; gang affiliation; attractive job opportunities; hope; and empowerment. Ideally, this baseline would draw on survey data and systematically collected statistics such as police data on crime and gangs, but we have not yet been able to access these (except for an estimate of gang membership). As a second-best substitute, we asked a “convenience sample” of Belizeans – the mediators, disputants and administrators we were interviewing for other purposes – to think beyond their individual attitudes and estimate broader social attitudes and other social-level variables, based on their experiences as part of the society. Their responses were captured both qualitatively and quantitatively (as counts or on 1-5 scales).

Respondents estimate on average that gangs account for “almost all” violence in South Belize City (4.9 on a 1-5 scale). They estimate that people feel closest to “not very safe” (2.3) on the streets, though youth feel “very safe” (4.7) in school. The average respondent thinks it is between “moderately important” and “important” (3.4) for youth to be in a gang to feel safe. They rate the pressure to join a gang as “strong” (4.0) and estimate “about half” (2.8) of youth are in gangs. (CYPD estimates 3,000 gang members, which is about 20% of the number of males aged 10-39 in all of Belize City – or perhaps as high as 40% of those in Southside.) Job opportunities for young people that are better than being in a gang are considered “rare” (1.8). Youth are thought to be only “fairly hopeful” (2.8) about their futures.


Date: 10/09/2014 Description: Bar chart on Importance for young people to be in a gang to feel safe. Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

Respondents are “fairly confident” (3.0) about the ability of youth to work together to improve the community. A detailed discussion of the baseline data is in the appendix.

We also assessed the level of TCO presence and activity in South Belize City, by consulting the best-informed people we could access – CYDP, prison officials, and relevant officials at the U.S. Embassy. Indications are that TCOs are not present in South Belize City, and Belizean gangs remain structurally independent. However, local sources state that Belizean gangs are heavily involved in trafficking drugs for TCOs for consumption in third countries. Embassy sources believe these drugs are a relatively small volume meant for local consumption. TCO presence and influence will be monitored by periodically interviewing a panel of informed local and Embassy sources.

The baseline data currently acquired will serve as a minimum quality of measurement if more systematic data cannot be achieved. The metrics used in this evaluation were reviewed with RESTORE Belize, which committed to conducting regular six-month measurements of both these social-level indicators and the individual-level indicators discussed below. RESTORE Belize agreed to explore whether an existing periodic survey has suitable indicators, or whether they can add questions to such a survey. One possibility may be cooperation with the Department of Human Development on its Community Action for Public Safety (CAPS) Survey, which will next be measured in October on a six-month cycle. RESTORE Belize will also investigate the potential for cooperation with the University of Belize, and pursue use of police statistics.

V. Findings

1. Mediation


Mediation Outputs:

• 36 mediators trained

• Gang mediators increased from 2 to 11

• 72 co-mediations completed

• 81% of co-mediations prevented violence

• 10 mediation trainers trained

• 29 peer mediators trained in one high school

The CSO engagement trained a total of 36 mediators in three courses, which was one course more than planned. Some of these had experience with gangs and agreed to become gang mediators, raising the number of gang mediators in Belize from 2 to 11. These mediators work principally in CYDP or the prison. The other 25 mediators (69%) are primarily focused on mediation in schools and youth facilities, which can help prevent gang affiliation and activity. Each mediator was to complete three co-mediations with a mentor, and almost all of the first two courses did this with the CSO trainers. All 72 of these co-mediations came to an agreement, including 58 which prevented violence where violence was threatened or previously occurred. The CSO team also delivered a course which trained 10 Belizean mediation trainers. Most of these newly trained trainers helped teach the third mediation course. The students from that course were to be mentored in co-mediations by the Belizean trainers, but this generally has not happened yet, for reasons discussed below. The other trainers gained experience by helping teach a peer mediation course at a high school, which trained 29 students and established a peer mediation program. This program is intended as a model to be replicated in high schools throughout Belize City.

Theory of Change

Mediators can affect gangs in South Belize City in two main ways: intervention and prevention. The intervention approach in the engagement sought to develop additional mediators able to focus on gang mediation. Gang mediators attempt to directly reduce gang violence between gang members, or at a more advanced level, between gangs at the leadership or organizational level. Gang truces, however, tend to be unstable unless underlying factors feeding gang affiliation and activity are addressed. Thus, it is necessary to also preventatively address the underlying drivers of gang affiliation and activity. Mediation helps prevent gang affiliation and activity by solving problems nonviolently before disputants turn to gangs or relatives in gangs to help, and by teaching youth communication and problem-solving skills which build better relationships and reduce the psychosocial and economic drivers of gang affiliation.

Mediation cases and resources are expected to grow. Demonstrating effectiveness at solving problems and preventing violence will generate interest in using mediation, interest in becoming mediators, and interest of institution administrators in hosting mediators and mediation training. This, combined with training trainers who sustainably offer good-quality training, will cause mediators and peer mediators to grow in numbers and to be present in more institutions and communities. RESTORE Belize will organize and support training and ensure quality is maintained.

Mediations that are not gang-related (and thus able to build a record of success more easily) can build a reputation and capacity that may ultimately lead to gang-related mediation. Mediation which is not focused on gangs but deals with youth may lead to gang mediation cases because they may raise gang issues, provide contact with gang members, and/or demonstrate the effectiveness of mediation to associated gang members. Gang mediation is a more difficult specialty that is facilitated by having knowledge about working with gangs as well as mediation. Some mediators without this knowledge may initially be uncomfortable with gang mediation, but this is likely to improve if provided opportunities to gain gang experience and knowledge. Also, exposing trainees to gang mediation in their co-mediations will make it more likely that they become gang mediators by giving them some experience and increasing their comfort with gang mediation.

Early Outcomes

The findings in this section are based on four datasets, as further detailed in the annex. We interviewed 15 of the 36 mediators with our mediator questionnaire. On a few of these questions, we only have 10-13 responses. Three additional mediators were interviewed at the prison, but using a different questionnaire – insights from these interviews are included below in separate observations as appropriate. We also interviewed ten disputants with another questionnaire, but not all of them were asked all questions, so the number of respondents varies between six and nine. Finally, we interviewed eight of the eleven mediator trainers with a questionnaire tailored for them.

Overall, mediation resonates well in Belize, and mediation cases and resources are likely to expand.

Mediators are continuing to mediate after CSO’s departure.

This is evidence that the Belizean mediators are taking ownership of the mediation process and suggests that the skill resonates in their environment. Mediators continued to mediate in the month since the engagement ended, although at a slow rate overall of 1.9 cases per mediator, across our 15 respondents.[3] There is broad agreement that the primary impediment has been the vacation season, because schools are not in session and thus not providing cases, in addition to mediators being away on vacation. The number of cases is expected to increase now that school has started – one school already referred four cases in its first week. The eight mediation trainers interviewed all expect to complete co-mediations with the third cohort of trainees soon now that vacation season is over. (Another impediment remains however, in the inter-institutional challenges discussed under sustainability below.)

Mediators are unanimously “very strongly interested” in continuing to mediate.[4]

This demonstrates strong local buy-in among mediators to mediation as a useful activity in the local context, and signals that the mediation rate will increase. Mediators also are generally happy with the training they received and many expressed interest in taking advanced, specialized or refresher courses that might become available, providing further evidence of the mediators’ enthusiasm about mediation.

Gang mediation is continuing at a strong pace, though only at CYDP and the prison.

The six CYDP mediators, who deal almost exclusively with gang mediations, have been exceptionally active, accounting for two-thirds of the 28 mediations in the past month among the 15 mediators interviewed. Their caseload was not reduced by the school vacation, nor were they affected by the logistical and interagency impediments. Their activity level is reason to hope that the other mediators, given their interest level, will pick up the pace once these impediments are removed. About half of the 30 cases per month at the prison are gang-related.[5] Other mediators, however, are generally not engaging gangs or gang issues after their co-mediations (some of which were intentionally gang-related).

Date: 10/09/2014 Description: Bar chart on Belizeans find mediation effective - slightly less so with gangs. Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

Mediation is generally effective, but gang mediation is more challenging and slightly less effective.

Mediation without known connection to gangs has consistently led to an agreement. These agreements are still holding 1-2 months later, to the knowledge of our respondents on this question – including six disputants[6] and four non-CYDP mediators. Most gang mediations result in agreements, though slightly less often: 17% of CYDP cases in the past month did not. [7] These agreements are less stable: one or two required re-negotiation, and a disputant in another case was shot by a third party. Mediation is generally successful at preventing violence when a dispute involves violence or threats of violence, but again, it is somewhat less successful when gangs are involved. Violence was an issue in approximately 80% of non-prison cases in the past month. Most of these were CYPD gang-related cases, which all involved violence or threats, while only 40% of the non-CYPD cases did. The non-CYPD mediators brought 100% of these cases to an agreement to prevent violence, while only 56% of the gang-related cases of CYPD resulted in agreements to prevent violence. Gang cases are more complicated[8] because third parties from either side often take violent actions against one of the parties to the agreement, and sometimes this is at the behest of one of the signatories. If the whole gang is not party to the agreement, it is vulnerable. Coming to an agreement or even starting a mediation is also complicated by recent violence among gangs, which can put them “beyond talking” for some time. Overall, 21 respondents (9 disputants and 12 mediators) rated mediation between “effective” and “very effective” on average, (4.5 on a 1-5 scale).[9] Sixteen respondents (6 disputants and 10 mediators) rated gang-related mediation closer to “fairly effective” at 3.3 on average.

There is growing interest in learning to mediate and establishing mediation programs.

The data are consistent with the theory that successful mediation breeds interest. All six disputants asked[10] are willing to use mediation again. Five of seven disputants spread the word about mediation’s effectiveness to select people. Sixteen of the 28 new cases that the mediators in our sample have handled in the month since CSO departed were requests from people who had heard that they are now mediators. This shows that word is spreading about mediation, and generating interest. Five of six disputants asked are interested in learning to mediate, which is good news for plans to expand peer mediation programs. Importantly, at least three respondents with interest in continuing to use mediation and/or becoming mediators had gang history, supporting the idea that it can appeal to gang members. Additionally, administrators at the prison, several schools and other youth facilities expressed interest in mediation training for both staff and peer mediators.

Mediation trainers are dedicated and preparing to expand the number of mediators.

When asked to rate their interest in training mediators, the eight trainers we asked responded with a 4 or 5 on a 1-5 scale, either “strongly” or “very strongly” interested. When asked if they had scheduled any additional mediation training courses, only one clearly had, although five others are in various stages of preparation. One trainer at CCRD had just taught mediation as part of a community leadership course, and is about to begin discussing a peer mediation course with high schools. Two other trainers, at Youth Cadet and Gwen Lizaragga High School, are looking at dates to conduct peer mediation training and mediation classes for staff this fall. Two trainers at RESTORE Belize expressed intentions to organize a full 40-hour mediation course and a peer mediation course before the end of the year. The trainer with the Community Police intends to teach a course for Community Police, but does not yet have plans. The two CYDP trainers have no current plans to teach mediation, although one has been teaching shorter conflict resolution courses and has two more planned. The high level of interest among trainers, administrators and former disputants suggests training plans will be implemented, as long as logistical and resource challenges can be overcome.

Co-mediations were delayed by logistical challenges, but are likely to continue now.

None of the eight trainers interviewed had mentored students in co-mediations yet, except the two at CYDP, who have nearly completed them. Other trainers cited difficulty coordinating schedules during the vacation period and the paucity of cases outside the distant prison while school was not in session. (Note that youth being on the streets instead of in school probably increases gang-related cases for CYDP). Most trainers plan to be in contact shortly with their trainees now that summer is over and school has begun. Since trainer interest is strong and logistics have eased, co-mediations are now likely to pick up pace.

Challenges for Achieving Broader Outcomes and Impact

Not all mediators are providing required information to RESTORE Belize.

RESTORE Belize is trying to establish a system for referring cases to mediators, which requires each mediator to submit a profile. Only a third of the 36 trained have submitted profiles so far, but this may pick up now that the vacation season is over. RESTORE Belize also asks the mediators to submit a data sheet on each mediation. This is important for detecting and fixing problems, ensuring mediators are getting sufficient practice, and tracking data that can aid fundraising by showing results. Overall cooperation has not been strong at 46% of the 13 mediators asked if they have sent data. However, this rises to 86% if excluding CYDP mediators – who claim the problem is simply being too busy, not that they do not have the information or are reluctant to share it. This might mean that compliance will drop amongst other mediators as they increase their caseloads, but the problem might simply be a need to build it into the organizational culture at CYDP. RESTORE Belize may need to create an incentive system or otherwise facilitate the collection of these data. Creating a stronger sense of community among mediators, including those at CYDP, especially through regular meetings/social events (and also correspondence, a website, newsletters, etc.) could help raise compliance. Such venues could be used to ask for these data and also to increase interest by showing how they are used.

More gang mediators are needed, which requires focused efforts beyond current plans.

Even though the engagement increased the number of gang mediators 450%, the ratio of gang members to gang mediators is still nearly 300 to one, and there is more gang mediation to be done than CYDP can do on its own. While prevention work is at least as important as direct intervention, this situation suggests that a further increase of gang mediators is still in order. Given the enthusiasm for general mediation, it is likely to grow fairly easily, but gang mediation is more difficult and will require additional focused efforts to expand. Contrary to one element of the theory of change, the increase of cases in schools is unlikely to lead to exposure to gang mediation because respondents tell us that gang issues are kept out of the schools. Since gangs are unlikely to take cases to court, gaining court-referred cases may also do little to reach gangs. Extending mediator presence in gang-ridden neighborhoods, through community centers, churches, etc. – and in conjunction with community dialogues – seems more likely to achieve this. In particular, recruiting and training civilians who live in gang-ridden neighborhoods would provide a helpful community-rooted alternative to the government officials at CYDP, who are mostly police officers. Specialized gang mediation mentoring/training by CYDP for external mediators would be helpful to building their skills. Taking time for this would ultimately help CYDP lighten its workload and build stronger community connections.

Trainer enthusiasm and creativity needs to be balanced with maintaining quality.

The enthusiasm for training among trainers, former disputants and administrators bodes well for the expansion of the number of mediators. However, maintaining quality training is a challenge, as trainers are tempted to shorten training to fit into the busy schedules of their trainees and themselves. RESTORE Belize wants to centrally organize training and control access to the training slides in order to maintain standards. The danger is that too much control will dampen trainer enthusiasm and prevent flexibility and innovation that could result in more people being trained. A balanced, flexible and creative approach is needed to capitalize on and nurture trainer enthusiasm and innovation while maintaining essential standards. An additional challenge discussed under sustainability below is providing funding for training.

2. Community Dialogues


Community Dialogue Outputs:

• 1 community dialogue initiated in the St. Martin’s community

• CSO partnered with & built on the preparatory work of a local NGO

• Mediators used the community dialogue model to establish the Belizean Association of Mediators

The community dialogue concept was introduced to mediators as part of a mediation package, but there was no separate, focused course on the subject. The small size of the engagement team plus the decision to conduct a third mediation course at the request of local partners limited the amount of mentoring provided for community dialogue. The CSO team planned to help start two-three community dialogues, but after talking with local partners, the team determined that significant work was needed to prepare a community for dialogue in this difficult environment. A local NGO, CCRD, had been working with the St. Martin’s community for 18 months, building the relationships, hope, confidence and skills such as cooperation and teamwork necessary for the community to work together on development. The team decided to start the first community dialogue in this community. It proved infeasible to devote the time and resources necessary to establish community dialogues in other communities. The mediators applied the community dialogue concept among themselves to create the Belizean Association of Mediators (BAM), but this is different from the envisioned application to empower disadvantaged communities.

Theory of Change

Community dialogues help to identify and address important action items for the community. They accordingly identify and mobilize community leaders, relevant resources in the community, and external resources from government, business and civil society. They build confidence of community members and leaders that the problems can be solved and promote transfer of relevant knowledge and skills. To grow, community dialogues need to show success in improving their communities in some way, so that they attract participation and resources, feed a sense of ownership and confidence in the community dialogue process, and inspire similar dialogues in neighboring communities. The accumulation of visible cases of success at bringing the community together to solve community problems eventually leads to changing the dominant community narrative in South Belize City from “there’s nothing I can do” to “we can solve this.” Community dialogues can contribute to the intended impact of this engagement by helping to address the safety, psychosocial and economic factors underlying gang affiliation and activity, thereby reducing gang violence and presence, and ultimately creating resilience against TCOs.

Another aspect of the project design is that community dialogues could work in concert with mediation, taking advantage of the atmosphere provided by the reduced violence that gang mediation is intended to produce, and conversely making gang mediation easier by contributing to a less violent and more hopeful environment. They could also build on the improved relationships, communication skills and problem solving skills promoted by mediation, scaling them up to the community level. Community dialogues could furthermore identify cases for mediation and connect them with community mediators.

Early Outcomes

Overall, community dialogue is making slower progress than hoped, but shows some signs of resonating with mediators and communities, and is continuing to develop.

St. Martin’s Community Dialogue

Data regarding this community dialogue principally comes from two detailed interviews, with the lead facilitator from CCRD and with the CCRD Director. The ratings given below come from these interviews. We also observed one session of the dialogue and briefly spoke with two participants and another facilitator. Future data collection should be broader and more systematic with participants.

The community dialogue is building hope and empowerment among dialogue participants, but has not yet had concrete impact on the broader community.

The participants interviewed expressed enthusiasm for the process and said it was building hope among participants. They think their one activity so far – designing and fielding a survey of skills in the community - is important, and they feel that they are on the path to accomplishing something. We asked CCRD staff to estimate how hopeful community dialogue participants are about their futures, and they gave an average rating between “fairly hopeful” and “hopeful” (3.5 on our 1-5 scale). Asked to rate this for a month ago, they averaged between “slightly hopeful” and “hopeful” (2.8), suggesting a modest improvement. Based on similar questions about the broader community, the CCRD staff responses suggest that the participants started out more hopeful than the community and had a notably stronger improvement, while there was very little improvement in the community (as one would expect at this point). Participants’ confidence in their ability to work together to improve their community also was on average higher than the level of confidence in the broader community and has increased in the past month, while there has been little or no improvement in confidence among members of the broader community (all according to CCRD estimates). Asked to rate the effectiveness of the dialogue in improving the community, the CCRD facilitator rated it 4 of 5 (“effective”), saying it is affecting community attitudes by word of mouth even though no programs targeting the broader community have started yet. In contrast, the CCRD program director stated that the community dialogue is only affecting the participants at this point, although she expects that it will eventually affect the community.

Dialogues are still led by CCRD staff, rather than community members.

To be sustainable, community members need to take ownership of the dialogue and lead it themselves, but CCRD has not yet been able to build their confidence, skills or ownership sufficiently for this to happen.[11] The CCRD facilitator noted that they are trying to transition the group to self-facilitation, by involving community members in aspects of running the meetings, like taking notes, and that this should accelerate in coming weeks, once the participants select officers for the association they are creating. The idea to conduct the community survey came from the participants, and they designed it, showing a certain degree of local ownership.

There is potential for broader effects to blossom in coming months.

Once the survey results are complete (several members have turned in results, but most have not and the deadline was extended until late September), the community dialogue will use them to design activities affecting the broader community. The selection of officers may help them move forward on implementation more efficiently. When CCRD staff were asked about their confidence that the dialogue will continue to develop and empower the people of St. Martin’s to improve their community, CCRD staff averaged 4 of 5, or “confident.”

The dialogue does not yet directly engage gangs, but CCRD plans to engage gangs and believes the process can reduce gangs and violence in the future.

Although gang members have not yet been brought into the community dialogue process, CCRD staff who were trained as mediators through the CSO engagement plan to begin mediating with gangs soon. They expect this to build relationships with gang members while also improving their problem-solving and communicative skills, such that it becomes feasible to productively draw them into the community dialogue. When asked how relevant the current community dialogue and associated activities are to reducing gang membership and influence, the CCRD facilitator rated it a 3 of 5 now (via prevention rather than intervention), and projected 3.5 six months from now. Asked if it has been effective in reducing violence, the facilitator rated it 2.5 of 5 now, and projected 3.5 six months from now. The CCRD Director believes that the community dialogue could be a good, non-threatening process for reaching gangs and for reducing violence in the future, but is not yet having such effects.

Slow progress is partially due to infrequent meetings with irregular attendance.

The dialogue meets only every other week to avoid competing with another community meeting CCRD organizes, which slows progress. Participation in the community dialogue over the past two months has varied: there is a small, core group of individuals who regularly attend, and others who are irregular, making continuity of discussion and progress difficult. The group included some young people early on, but they developed scheduling conflicts with their jobs. Men are starting to attend more regularly, but it is still mostly women. There have not yet been intentional efforts to publicize the dialogue, but one CCRD facilitator has described community dialogue while participating in a local talk radio show, and several people called in about this, demonstrating potential for more publicity to create interest.

Other Community Dialogues

BAM meetings were suspended after three meetings with dwindling attendance.

Attendance at meetings dwindled from ten individuals in the first meeting to only a few attendees in the two subsequent meetings. The group consequently suspended meetings until mediators completed the 240 hours of mediation each mediator committed to complete within a year, “to see who is serious.” However, the unanimous “very strong” interest we found among mediators is at odds with interpreting the poor attendance as lack of commitment. The same problems that inhibited mediation in the past month – vacation season and logistics – are likely the culprits behind low attendance at BAM meetings.

Two mediators are actively pursuing establishment of community dialogues in other communities, and 40% of the 15 mediators asked have intentions to do so.

Date: 10/09/2014 Description: Pie chart on Mediators intending to start new community dialogues. Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

This suggests that the community dialogue concept has gained traction with mediators. The proportion of mediators intending to start community dialogues rises to 67% if excluding the six CYDP mediators, who say they are too busy, but would be willing to participate if someone else organizes them. Although CCRD does not intend to start a new dialogue in another community within the next year, they may expand to cover other zones of St. Martin’s, and they plan to start a dialogue between youth and police in St. Martin’s. They said there is interest from the Yarborough neighborhood. RESTORE Belize is interested in supporting this, although the Yarborough Community Police mediators were disinterested.

Challenges for Achieving Broader Outcomes and Impact

Better support is needed for increasing the efficacy and expansion of community dialogues.

A redoubled effort by RESTORE Belize in advising CCRD might help find ways of speeding up the development of the community dialogue. The CCRD Director expressed uncertainty regarding the concrete goals and timeline of the community dialogue, and clarifying this might help. She also expressed concern about ambiguity on the division of labor between CCRD and RESTORE Belize, especially in creation of dialogues in other communities, where she feels CCRD cannot currently play a leading role, but could advise. As other mediators move forward with establishing community dialogues elsewhere, it would be helpful for them to have a forum to share experiences and ideas, but none currently exists. Such a forum would also be helpful for promoting community dialogue and motivating others to get involved. RESTORE Belize is best placed currently to take the lead on this, and BAM might play a role if revived.

It is unclear how much preparatory work is needed to prepare communities for new dialogues.

CCRD’s experience is that communities may need preparatory work in issues such as effective collaboration and cooperation before a community dialogue can be established. They used service delivery events which bring the community together, such as family days, to help build these attitudes and skills. Such programs built trusting relationships that allowed CCRD to invite community members to other activities, ultimately including the community dialogue. The CCRD Director also emphasized the importance of “community organizing” involving informal conversations to sell the idea of community action. Much of the preparatory work involved gathering information, making connections and building relationships between CCRD and the community. For an organization or community leader that is already well-established in a community, less preparation of this kind might be necessary. Consequently, it is unclear that the full 18 months CCRD spent preparing would be necessary for all communities. The CCRD facilitator said it could be done in as little as six months, though the Director stated 18 are necessary. Asked how to judge the time needed to prepare a community for community dialogue, the CCRD facilitator suggested a set of community indicators including participation rates, number of ideas for programs offered by the community, interest level in development, and resources available, including organizations, funding and people.

The economic drivers of gang affiliation and activity are critical to address, but require more resources than community dialogue alone.

Respondents gave at least equal weight to economic factors as they did safety or psychosocial factors in driving gang affiliation and activity. Recall that respondents rated the availability of jobs that are more attractive than gangs as “rare.” But addressing the economics of South Belize City is very difficult, as the experience of CCRD – which has tried several different means of stimulating economic development in St. Martin’s – attests. It is thus critical that additional resources specialized in economic development be targeted to neighborhoods with community dialogues to capitalize on the progress they make in improving security and psychosocial factors, as well as whatever progress they may make economically. The economics driving gang affiliation and activity are strong, and failing to address this is likely to endanger the strategic success of the overall effort. We discussed this challenge with the CEO of the Prime Minister’s Office, who agreed that this is critical, though very challenging, and agreed to consider how economic development resources could be coordinated with community dialogues.

Mediators are deprived of support BAM could be offering.

Leaving BAM suspended until mediators finish 240 hours could take a year, depriving mediators support it could provide such as sharing experiences, facilitating training, public education and advocacy on behalf of mediators – support that could aid mediators in finishing those hours. If the prime reason for poor attendance is logistical rather than lack of interest, then reviving BAM in the near term might produce better attendance now that vacation season is over. Attendance could be facilitated through creative use of communication technologies such as virtual meetings via the Internet. The stronger sense of community and opportunities for communication BAM might create could help improve compliance with information requests, as noted above.

3. Sustainability and Networking

To promote effectiveness, local ownership and sustainability, CSO worked with multiple local partners on this engagement. The closest relationship was with RESTORE Belize, which requested the program and was a partner from the earliest stages in the design and implementation of the project. CSO helped develop a nascent network of private and public actors for mediation and community dialogue through discussions with relevant institutions and common training. RESTORE Belize is at the heart of the network, with functions of assigning cases/venues to mediators, coordinating training, maintaining standards and monitoring progress. Other network members provide mediation cases and/or mediators, or facilitate community dialogue, including high schools and youth institutions, Community Police, CYDP, CCRD, Belize Central Prison, the Family Court, and the Community Rehabilitation Department. CSO worked closely with RESTORE Belize at the end of the engagement and remotely afterwards to develop a system for managing mediation referrals and training, and the readiness of RESTORE Belize to manage it. CSO also helped RESTORE Belize (partly via this evaluation) to develop metrics for monitoring progress and effects of the project in the future.

Theory of Change

Date: 10/09/2014 Description: Flowchart on the network for mediation and community dialogue. Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

Partnering closely with local partners from the start in design and implementation of the project will foster local ownership and sustainability. Success of the efforts inspired by the project, especially success in reducing gang affiliation, activity and violence, will enhance the buy-in and continued support of RESTORE Belize, the Government of Belize and other local actors. These entities will therefore provide important support, including funding, that facilitates continued mediation, community dialogues, and mediation training.

The network for mediation and community dialogue will help mediators support each other and improve the quality, sustainability and complementary interaction of mediation and community dialogue. With RESTORE Belize in a central role, the network will facilitate coordinated referral of cases to mediators, the growth of mediation cases, ongoing training, maintenance of standards, development of additional mediators, and development of additional community dialogues. As the most relevant government entity, and with its support by the Prime Minister, RESTORE Belize is well-placed to play the coordinating and fundraising role in this network.

Early Outcomes

RESTORE Belize staff are optimistic about the future societal impact of the program and demonstrate local ownership.

Since RESTORE Belize staff are driving the program, their ownership is critical to sustainability. When three key staff members were asked about their confidence that the program can ultimately reduce gang presence, they averaged 3.7 on our 1-5 scale, closer to “confident” than “fairly confident.” They emphasized the importance of reaching into schools and families for prevention, expanding the number of mediators, mediators working together across agencies, and changing the environment, not just individuals. Asked about their confidence that the program can make people feel more hopeful and empowered, they averaged 4.0, “confident.” They discussed the role of community dialogues in helping people find solutions and see change, and the importance of community members owning the process. These responses indicate substantial buy-in and understanding regarding the theory connecting program activities to goals, which is important for effectiveness, scaling up and sustainability.

RESTORE Belize is committed to six-month measurements of indicators.

As noted in the earlier discussion of the social-level baseline, RESTORE Belize is committed to ongoing measurement of indicators at all levels on a six-month cycle. This reflects ownership and will promote effectiveness by providing feedback on how well various aspects are achieving desired effects. RESTORE Belize and the Prime Minister’s CEO see data demonstrating effectiveness as essential to fundraising, and hence connect this to sustainability – but also recognize the importance of ensuring objective data collection by seeking the assistance of an independent party such as the university. The Prime Minister’s CEO stated she would inform departments that the data need to show results in six months, to help motivate cooperation.

Challenges for Achieving Broader Outcomes and Impact

The nascent network for mediation and community dialogue needs further development.

CSO’s work has connected key public and private institutions and individuals at multiple levels of Belizean society, in a nascent network for mediation and community dialogue which may improve the effectiveness, growth and sustainability of the program. However, its potential is currently limited by the failure of some mediators to provide profiles and/or mediation case data to RESTORE Belize, as discussed above, and by the inter-institutional issues discussed below. The suspension of BAM is also not consistent with the development of this network. When we asked mediators and trainers about their level of consultation via this network, we found virtually none so far other than within CYDP – but this might increase now that vacation season is over. As noted above, more interaction between CYDP and other community-based mediators could help develop more gang mediation capacity and ease the CYDP workload, but this is not currently happening. As also noted above, the network is not currently providing a forum for sharing experiences among those attempting to establish community dialogues. RESTORE Belize and other members of the network need to find ways to consolidate it and enhance its functioning.

RESTORE Belize lacks sufficient staff and funding to consolidate and expand the program.

RESTORE Belize currently lacks the funds to hire a dedicated staff member to drive the program - who understands the overarching vision for lasting peace and can coordinate many actors to get there - as well as a full-time scheduler for mediations and related meetings. They would ideally like to fund both these dedicated positions, but if they can add a dedicated staff member for at least one of these, they will be better able to cover the other one with existing resources. RESTORE Belize also needs financial resources for conducting trainings and meetings and for transportation. When asked about their confidence in continued funding for all parts of the program, the three RESTORE Belize staff we interviewed averaged 2.3, closer to “not very confident” than “fairly confident.”[12] They have made several additional proposals since CSO left, primarily with the private sector. Some grants have already been rejected, and others are still pending. However, Audrey Wallace, Central Executive Officer of the Office of the Prime Minister, expressed “high confidence” that they will find external funding to continue the program, based on donor enthusiasm in a recent donor coordination group meeting. Even if external funding is not secured, she is confident that her government will continue to provide enough funding for the mediation and community dialogue efforts to continue at a basic level. She also committed to finding funding for a brochure on mediation that could help build the caseload and attract new mediators.

Inter-institutional issues limit mediator availability.

One major impediment to the solidification of the mediation network has been inter-institutional. Most mediators work for government agencies, whose directors do not necessarily support them operating as mediators as part of their regular work hours or provide logistical support such as transportation[13] to mediation venues. RESTORE Belize is working on Memoranda of Understanding (MoU) with these departments to create scheduled times when these mediators will be available and resolve other inter-institutional issues. Additionally, the Prime Minister’s Office is working to persuade directors to more strongly support use of their mediators, expressing confidence in its ability to do so. RESTORE Belize staff noted that they lack sufficient technical capacity to do necessary strategic planning and design a structure to coordinate mediators across agencies. Both RESTORE Belize and the Prime Minister’s Office requested additional presence of a CSO advisor for this purpose, noting that the “directors listen more if the experts are here.”

If active, BAM could be an important civic partner for RESTORE Belize.

Given RESTORE Belize’s limited resources, BAM could be a valuable civic partner that could ease some of the workload if it were active. Furthermore, BAM can serve as a backup organization for coordinating mediation activity and training should RESTORE Belize ever shift priorities or lose political support. Since this would require developing organizational capacity, an independent reputation and sustainable private sources of funding, it might take years for BAM to be able to securely play this role. However, BAM is not currently functioning (due to an apparent misperception that most mediators are not committed, as described above), and lacks funding from either private or international sources. It is important that the association be reinvigorated and developed.

4. Environment & Transferability

The Belizean context at the time of CSO’s engagement had significant advantages, including a strong local partner and strong political backing from the Prime Minister, and the success of CSO’s efforts was greatly facilitated by their ownership and support of the effort. Such a confluence of facilitating features might not exist in other contexts, and this should be taken into account when determining the resource needs and feasibility of a similar engagement elsewhere. Decisions on whether and how to transfer this approach will be better informed if they occur after the next wave of data on this engagement is analyzed, which is scheduled for collection in late February 2013, and will allow comparison of macro-level variables against the baseline.

Key Observations

This environment had a confluence of significant advantages that might not exist elsewhere.

The conditions in Belize are unusually favorable in a number of ways. CSO has many strong, dedicated and connected local partners, including the Prime Minister’s Office, RESTORE Belize, CYDP, CCRD, CRD, the prison, two youth facilities and several schools, and a recognized leadership role for RESTORE Belize with the highest political backing. Gangs are small, not very well organized, independent from TCOs (although dealing with them), not as violent as TCOs and not active in the schools. The culture has a strong respect for written contracts, which increases the efficacy of mediation agreements and enhances the likelihood that agreements will be maintained, at least among signatories. In addition, Belize is an English-speaking country, which eases the choice of CSO personnel and materials. In other contexts, gang influence and characteristics, language barriers, community characteristics, and pre-existing local institutions are among potential issues that will need to be taken into consideration when determining the feasibility, potential success, and design of similar engagements elsewhere.

The individualistic culture in Belize City slowed community dialogue development.

Conversely, several respondents pointed out that Belize has a very individualistic culture, and it may be easier to establish community dialogues more rapidly in more collaborative cultures. (However, it may also be more difficult for outsiders to be trusted in more collaborative cultures.)

This was an undersized engagement even for Belize.

Even given the favorable environment, the CSO team lacked sufficient time and personnel to devote the necessary attention to each component of the engagement. This suggests that more time and personnel should be provided to pursue an engagement of this scope elsewhere, especially if the environment is less permissive. Specifically, other engagements including community dialogue should have at least one dedicated staff member for at least a month to focus on training and mentoring for community dialogues, and a specific course focusing on community dialogues should be developed. Resources for mediation training and mentoring are minimally 3 people for 3 weeks per course. Additionally, time and personnel should be included for the following, each of which could involve a dedicated staff member for weeks to a month:

  • Engagement scoping to find partners, tailor training and apply local culture to role plays
  • Co-teaching with new trainers on all segments of each kind of training
  • Development of metrics and measurement of baselines for outcomes and impact
  • Aiding institutionalization and development of strategy and funding for expansion and sustainability.

V. Recommendations

1. Mediation

Recommendation: Expand support systems for mediators.




• Ensure that trainers quickly complete co-mediations and continue mentoring.

• Quarterly meeting of mediators to share best practices and discuss growth.

• Provide mediators with pamphlets on mediation for distribution to institutions and public. Add website and newsletter on mediation.

Belizean Association of Mediators (BAM)

• Do not wait for completion of 240 hours requirement – get functioning now.

• Explore use of website, email and virtual meetings to increase participation.

• Help mediators share experiences, get training, educate public, advocate interests.


• Promote development of BAM with encouragement, advice and connections.

U.S. Embassy Belmopan

• Support development of BAM through existing programs as appropriate, and by advocating funding and support by Belizean and international sources.


Recommendation: Expand gang-related mediation by getting mediators deeper in communities.[14]




• Increase mediator presence in communities, via community organizations and community dialogues.

• Train more community activists (vice government employees) as mediators.

• Promote gang mediation training and mentoring by CYDP.


Recommendation: Promote quality training that meets the interests and needs of trainers, current mediators, new trainees, and institutions interested in hosting training.




• Creatively balance quality control with building on trainer enthusiasm/innovation.

• Create a plan for peer mediation training in every Belize City high school, including training for high school counselors on training peer mediators and establishing/managing peer mediation programs.

• Work with prison to create specialized mediation training for prison officials and a peer mediation program among prisoners.

• Provide refresher and advanced topic training for current mediators


2. Community Dialogues

Recommendation: Promote more rapid and effective development of community dialogues




• Create forum for community dialogue facilitators to share experiences and ideas, and to promote expansion. This forum could be hosted by or involve BAM.

• Take leadership role in promoting and advising community dialogues, with focus on impact. Clarify RESTORE Belize role versus other organizations involved.

• Clarify general goals and timelines for community dialogues.

• Develop and deliver training on community dialogue establishment & facilitation.

• Recruit and train potential facilitators who are rooted in target communities.

Prime Minister’s Office

• Organize targeted economic development advice and assistance to build on achievements of community dialogues and address economic drivers of gangs.


• Provide guidance on project management.


3. Sustainability

Recommendation: Resolve remaining inter-institutional issues




• Complete interagency MoUs

Prime Minister’s Office

• Help directors to see mediation in their interest and treat mediation time as regular work time, provide transportation, put mediation in job descriptions and performance evaluations. Consider providing small transportation budgets.

U.S. Embassy Belmopan

• Emphasize importance of effort with other ministries/directors

• Follow up with Prime Minister’s Office


Recommendation: Ensure adequate resources, information and planning for sustainable expansion adequate to achieve intended impact.




• Find a way to create dedicated staff positions, either through additional funding or by shifting assignments, and complete Scope of Work for dedicated positions.

• Measure indicators on Feb/Aug six-month cycle and schedule this now. Explore using or attaching to existing surveys for social-level indicators. If possible take social level baseline survey now. Partner with university.

• Incentivize and aid mediators to provide requested profiles and mediation data.


• Seek independent private and international funding to ensure sustainability.


• Continue to remotely advise RESTORE Belize on strategy, institutionalization, & expansion until they are confident.

• Continue to advise RESTORE Belize on M&E, but not conduct it for them (with the partial exception of facilitating embassy input on TCO presence and activity).

• Assist RESTORE Belize in finding funding by providing suggestions, connections and recommendations.

• Occasional short trips may be worthwhile, but do not provide TDY staff to manage the program for RESTORE Belize.

U.S. Embassy Belmopan

• Consider funding position, logistics, publicity – or helping to find funds.

• Monitor interaction of TCOs and local gangs


4. Environment & Transferability

Recommendation: Ensure adequate resources to complete all components of the engagement.




• Base decisions on whether and how to apply this approach on analysis of contextual factors affecting implementation and impact, including local partners, political support, gang structure, TCO engagement, violence, economics, communalism and conflict culture. Anticipate more difficulty elsewhere.

• Add specific course and dedicated officer for community dialogue development.

• Include time and resources for co-teaching with new trainers, institutionalization of concepts and procedures, monitoring & evaluation, and sustainability planning.




This evaluation of the Belize engagement was conducted by two staff members from the CSO Office of Learning and Training. Engagement team members and members of the larger Operations Team were consulted on the design of the evaluation and commented on this report, but were not involved in collecting data or conducting the analysis, nor did they control the content. The two-person data-collection team was allowed five days to collect data in Belize, in late August, 2012.

Because the longer-range outcomes and impacts would occur at a social level, beyond the level of individuals who directly participated in mediation or community dialogue, it would be ideal to use surveys or other data systematically gathered across South Belize City, such as police statistics. We are pursuing such options for future use, but generally have not been successful so far in securing appropriate data of this sort. As a backup strategy, we asked people we were interviewing anyway – mediators, disputants, school administrators, etc. – to estimate how people generally feel on subjects such as safety, hope and confidence.[15] This approach treats them as a panel of well-informed judges of how Belizeans think. This small “convenience sample” is not as reliable as a larger random or representative sample, but it is the best we could do with our resources.

We developed questionnaires for various types of respondents – disputants who are prisoners versus students, for instance - based on our evaluation questions. Some questions, especially at the social level, were essentially common across respondent types, but the wording was sometimes tweaked to better suit a particular respondent type. In particular, for some respondents it made sense to focus social-level questions on a particular environment familiar to the respondent, such as the prison or the St. Martin’s neighborhood. For others it was more appropriate to ask about South Belize City overall, because that is the scope of their work.

To ease summarizing information and comparing the current situation with measurements in the future, many questions were quantitative, such as counts of mediations, or quantifications of qualitative answers using ordinal 1-5 response scales such as:

1=Not Effective, 2=Slightly Effective, 3=Fairly Effective, 4=Effective, 5=Very Effective.

However, we collected open-ended qualitative comments on these questions, and also had some purely qualitative questions, to provide a richer understanding and capture unanticipated responses.

We met first with RESTORE Belize and got their feedback on these questionnaires, partially through using some of them to interview RESTORE Belize staff who were trained as mediators and mediation trainers. The questionnaires were revised based on this feedback, and also evolved as we experienced interviews with others. On our last day, we once again reviewed the questionnaires with RESTORE Belize, because they will use them for collecting future waves of data.

45 people were interviewed, including:

  • 18 of the 36 mediators
  • 9 of 11 mediation trainers
  • 3 peer mediators
  • 10 disputants (4 prisoners, 2 high school students, 2 Youth Cadets, 2 at Youth Hostel)
  • 4 people involved with community dialogue (2 CCRD staff, 2 participants)
  • 3 RESTORE Belize staff
  • Administrators at 5 relevant institutions (prison, Youth Cadets, CCRD, high schools)
  • 3 U.S. Embassy Belmopan staff
  • Several informal incidental interviews with additional mediators or activists.

Most interviews were of one person at a time, which arguably produces more frank responses. Several were group interviews due to miscommunications or time constraints, including the interview of prison officials/mediators, of CYDP mediators, and at the U.S. Embassy.

To speed the scheduling of interviews in the short time allotted, the evaluation team relied on RESTORE Belize and one of the CSO trainers to use their relationships to help arrange interviews with mediators and at several institutions. This is arguably problematic with the mediators[16] because we interviewed only half of these, leaving potential for bias in the sampling towards the mediators with whom they have the strongest relationships – likely to be the most enthusiastic ones. Furthermore, the most enthusiastic mediators may have been more likely to cooperate even with random contacting – although a notably dissatisfied mediator arguably might have been eager to talk to us as well. This potential bias may have caused some over-estimation of the amount of mediator activity or enthusiasm. There is also an inadvertent slight bias in the overall sample of 18 mediators towards those who were trained first, and thus had more time to mediate: 9 from the 1st cohort, 4 from the 2nd, and 5 from the 3rd. However, since the three mediators at the prison are all from the 1st cohort, and were not interviewed with the mediator questionnaire, the summary data on mediators generally only have six in the first cohort, helping to ameliorate this problem (the three mediators at the prison are accounted for in the presentation below only in additional comments specific to them). Having interviewed so many as half of the mediators helps to reduce the probability of bias, and there were several who had little or no activity. It seems unlikely that this has dramatically skewed the results.

Disputants at the prison and youth facilities were selected to sample different disputes and different mediators as much as feasible. One of these disputes showed some indications of being gang-related, and a disputant in a different dispute disclosed she had gang history. However, we did not attempt to sample disputants from the gang mediations at CYDP. Thus, disputants from gang-related disputes are under-represented, which may have somewhat inflated estimates of mediation effectiveness, violence prevention and effects such as interest in learning to mediate. These effects are not expected to be large, based on the fairly small differences (one point in five, on average) in mediator opinions on effectiveness of mediations related to gangs versus those that are not.

Mediation Data

We constructed a mediator questionnaire focused on outcomes with which we interviewed 15 of the 36 mediators trained, generating the statistics in the table below. Because the mediators at CYDP, a gang intervention organization, were unusually active, and almost entirely active in gang mediation, statistics are also provided without the six CYDP mediators. (We also interviewed three mediators at the prison, but using a different questionnaire designed for prison officials – if we had comparable statistics from them, the proportion of gang-related mediations and violence-preventing mediations would rise).

Date: 10/09/2014 Description: Table chart on mediation data, showing all mediators interviewed and non-CYPD mediators. Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

The interview team used a set of questions to evaluate the immediate outcomes of mediation in addressing disputes and reducing violence, both in general and in relation to gangs. In addition to straightforward questions about effectiveness, we also asked if agreements were holding (1-2 months later) and add-on effects of creating interest in further use of mediation and learning to mediate.

Our primary source of information about mediation effectiveness at this point is the disputants who participated in mediations conducted by CSO-trained mediators. They provide a more objective point of view than the mediators rating their own performance. Since these mediations were co-mediations with a CSO trainer, they do not necessarily reflect the skill of the trainees alone, but the trainees did have a hand in these results. We interviewed ten disputants from nine disputes, including four prisoners (all males, apparently in their 20s or 30s) and two teenagers each from Youth Cadets (two boys), the Youth Hostel (two girls), and Wesley High School (two girls). Three revealed they had gang history, including two prisoners and one of the girls. None of the disputes were clearly gang-related, but one had potential for gang-related escalation. The table below shows averages across the disputants for whom we gathered data on each question (questions were not always asked of all respondents).

Date: 10/09/2014 Description: Table chart on 10 disputants (4 prisoners, 2 youth cadets, 2 youth hostel, 2 students). Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

The first two effectiveness questions were also asked of twelve mediators, including six gang mediators from CYDP, 3 prison mediators, two at RESTORE Belize and one at CCRD, the organization running the community dialogue in St. Martin’s. The table below combines their responses with those of the disputants discussed above. The results with these extra data points are very similar.

A slightly different version of these questions is also available, but with fewer respondents. Focusing on effectiveness at preventing violence, the same rating of 4.5 is reached for general mediation – halfway between “effective” and “very effective.” This indicator gives a surprisingly higher rating for gang-related mediation at preventing violence. Note, however, that there are only seven respondents for this aspect of the question, and six of these are from CYPD, which deals almost entirely with gangs.

Date: 10/09/2014 Description: Table chart on mixture of 10 disputants (4 in prison) and 12 mediators (6 CYDP, 3 prison, 2 RB, 1 CCRD). Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

The evaluation team interviewed nine of the eleven Belizean mediation trainers who were trained by the CSO engagement, and asked eight of these a set of questions on the quality of their training as trainers and the extent of their subsequent trainer activities and plans, as listed in the table below.

Date: 10/09/2014 Description: Table chart on mediation trainers trained. Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

Baseline Data for Longer-Term Outcomes and Impact

We measured these variables by asking a “convenience sample” using the disputants, mediators, and administrators we were interviewing for the other variables. These variables would ideally be measured more systematically throughout South Belize City, and we are pursuing options for this. Gang violence, affiliation and other measures of gang presence and influence may be available in police statistics which we are waiting to receive. Surveys using large representative or random samples would be ideal for the others. We discussed this with RESTORE Belize, and they will explore whether existing surveys have similar indicators or these indicators can be added on. In the meantime, data gathered through this small convenience sample serve as a better baseline than none.

The table below displays our baseline measurements. Below the table is a map of gang territory provided by CYDP; if a similar map is produced in the future, comparison of the amount of gang-controlled territory could provide another indicator.

Date: 10/09/2014 Description: Table chart on social outcome indicators. Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image

Graphic of CYDP Map of Gang Territory in Belize City, October 2010

Date: 10/09/2014 Description: CYDP Map of Gang Territory in Belize City, October 2010. Graphic from November 30, 2012 Belize Engagement Evaluation Report. - State Dept Image


[1] Re-Establish Security Through Outreach, Rehabilitation and Education

[2] For a detailed description of the evaluation methodology, see the Annex.

[3] Note: Most of these cases have been at CYDP. Among the nine non-CYDP mediators, the rate has been 1.1 cases per mediator in the past month.

[4] There are 13 respondents for this question.

[5] We do not compute mediations per month for the four prison-based mediators, because many of these mediations were handled by external mediators, and we did not collect case data specific to the prison-based mediators.

[6] Because mediators might not be objective judges of their own performance, the disputant data add credibility.

[7] The record is less clear in the prison. These three mediators first said “about half” of mediations result in agreements, but later say there are usually agreements. The first statement may have been intended to apply only to the gang-related cases, which they estimate to be about half their cases, meaning there are agreements in 75% of cases overall.

[8] CYDP mediators pointed out that mediating at schools is much easier than mediating with gangs. They feel that they were able to jump straight into gang mediation because of their prior experience working with gangs, and getting experience in easier cases first was not necessary for them. However, they expect that mastering mediation on easier cases first is necessary for those without gang intervention experience before trying to add the complications of mediating with gangs.

[9] The disputants polled rated it slightly higher, at 4.7. They are arguably the more objective source. Their rating for gang mediation was the same as the mediators, at 3.3.

[10] For these three questions asked of disputants, 2-3 disputants were prisoners and the rest were youth outside prison.

[11] The dialogue we happened to observe lacked an agenda and accomplished little despite being led by a CCRD facilitator, but we are told by CSO trainers that earlier dialogues were better facilitated.

[12] It is worth noting, though, that the person responsible for seeking funding gave the “fairly confident” rating, and one of the others noted she did not know much about the funding.

[13] While transportation needs are reduced by the opening of schools, which provides more mediation opportunities in town versus the prison outside of town, and by a “cluster” plan connecting mediators to nearby venues, inter-institutional cooperation in sharing limited transportation resources is still needed, and a challenge to be addressed in the MOUs.

[14] This recommendation is not meant to detract from the importance of preventive work in schools and elsewhere.

[15] We generally asked first how the respondent feels, and then how the respondent thinks others feel generally. This helped prime the pump conceptually for the more difficult broader question, and helped clarify that the second question is different that the respondent’s personal feelings.

[16] It is not problematic with other groups because the evaluation team decided whom to interview at the institution, or all or nearly all of the relevant people in a respondent type were interviewed.