February 2013 This publication was produced at the request of the United States Agency for International Development. It was prepared independently by International Business & Technical Consultants, Inc. (IBTCI). EVALUATION Peace in East and Central Africa (PEACE II) Program Final Evaluation Report PEACE IN EAST AND CENTRAL AFRICA (PEACE II) PROGRAM FINAL EVALUATION REPORT February 9, 2013 Prepared by: Robert Grossman-Vermaas, Principal, Monitoring & Evaluation, IBTCI Adam Reisman, Senior Manager, Monitoring & Evaluation, IBTCI This publication was produced for review by the United States Agency for International Development. It was prepared under Contract Number AID-RAN-I-00-09-00016, Task Order Number AID-623-TO-12-00002 by International Business & Technical Consultants, Inc. (IBTCI) DISCLAIMER The authors’ views expressed in this publication do not necessarily reflect the views of the United States Agency for International Development or the United States Government. TABLE OF CONTENTS List of Acronyms ................................................................................................................................................................. i Executive Summary ............................................................................................................................................................ ii Introduction and Purpose ................................................................................................................................................. 1 Project Background ............................................................................................................................................................ 3 Evaluation Methods & Limitations .................................................................................................................................. 5 Monitoring & Evaluation: Conclusions, Supporting Findings and Recommendations ........................................ 10 Cumulative Impact: Perceptions of Security and Cross-border Interaction in the PEACE II Border Areas ................................................................................................................................................................................... 14 Peace Dividends Theory of Change: Conclusions, Supporting Findings and Recommendations ...................... 20 Trauma Healing & Social Reconciliation Theory of Change: Conclusions, Supporting Findings and Recommendations ............................................................................................................................................................ 28 Local Collaborative Peace System & Sector Response Theory of Change: Conclusions, Supporting Findings and Recommendations .................................................................................................................................... 37 Strategic Conclusions and Recommendations ............................................................................................................. 46 i LIST OF ACRONYMS AMISOM The African Union Mission in Somalia CCP Community Contracting Process CEWARN Conflict Early Warning and Response Mechanism CEWERU Conflict Early Warning and Response Unit CPU Community Policing Unit CWG Community Working Group DC District Chief DfID U.K. Department for International Development EA Enumeration Area F2F Face-to-Face Interview FGD Focus Group Discussion GOK Government of Kenya HATI Horn of Africa Training Institute HH Household IBTCI International Business & Technical Consultants, Inc. IR Intermediate Result KDF Kenya Defense Forces KII Key Informant Interview LAPSSET Lamu Port-Southern Sudan-Ethiopia Transport LCPS/SR Local Collaborative Peace System/Sector Response M&E Monitoring and Evaluation PEACE II Peace in East and Central Africa II PLA Participatory Learning Approach PMP Performance Management Plan RCMG Regional Conflict Management and Governance Office RSA Research Solutions Africa SOW Statement of Work ToC Theory of Change TFG The Transitional Federal Government TH/SR Trauma Healing/Social Reconciliation TIS Transition Initiatives for Somalia TPM Third Party Monitoring USAID/EA U.S. Agency for International Development/East Africa ii EXECUTIVE SUMMARY Evaluation Purpose and Evaluation Themes International Business and Technical Consultants, Inc. (IBTCI) was contracted to perform the final performance evaluation of USAID/East Africa’s (USAID/EA) Peace in East and Central Africa II (PEACE II) program. The five-year PEACE II program was USAID/EA’s primary conflict mitigation activity in terms of financial investment and profile, and operated from 2007-2012. The Cooperative Agreement (CoAg) was awarded to PACT in September 30, 2007 and the program ended on December 31, 2012, and was managed by the USAID Regional Conflict Management and Governance Office (RCMG). PEACE II program activities were active in multiple peace corridors along the Kenya-Somalia border and in one corridor on the Kenya-Uganda border. In the context of PEACE II, peace corridors are considered discrete geographic zones with notable historical cross-border activity, as well as evidence of cross-border or inter-clan conflict. Most peace corridors consist of two townships, one on either side of the targeted border. This is an evidence-based evaluation that investigates the effectiveness of PEACE II in 1) building community conflict prevention, mitigation, and response capacity and 2) improving security in targeted communities. The evaluation also supported USAID/EA to learn which theories of change (ToCs) proved valid, which ones did not, and, where monitoring and evaluation (M&E) data was not sufficient, to make a validity determination. The ToCs – hypothetical statements supported by critical assumptions about the communities to which they refer – can be summarized as follows: 1. Peace Dividend Theory of Change: Through co-managing and sharing tangible development projects, cross-border communities will develop strategic relationships and long-lasting peace networks. These relationships will provide the basis for a joint response when faced with violent conflict in the future. 2. Trauma Healing and Social Reconciliation Theory of Change: Increased stakeholder understanding of broken relationships and trauma caused by conflict will increase stakeholder resilience and leadership in conflict transformation when faced with future violent conflict. 3. Local Collaborative Peace System and Sector Response Theory of Change: Through increasing capacity of local organizations (e.g. peace committees, sector response units) to respond to conflict, a critical mass of peace actors is formed to proactively deal with conflict when it emerges. The evaluation will be used to inform the next generation of USAID/EA conflict mitigation activities. It builds upon previous monitoring and evaluation data and reports already captured by the PEACE II M&E Plan and M&E implementation strategy, as well as previous internal assessments and annual reports, to provide a more contextual, evidence-based analysis of program outcomes and impacts achieved, and to provide insight into the validity of the program’s ToCs. Because the program’s objectives deviated throughout the five-year period of performance, there are some inconsistencies in the data, but these inconsistencies have been mitigated through the evaluation team’s use of freshly collected data and analytical findings. The evaluation was structured around several key questions that USAID/EA was particularly interested in having answered. These questions were grouped according to the following themes:  What were the program impacts on security, cross-border interactions and community life in the border areas? Were there unintended impacts? Can these impacts be attributed to PEACE II’s activities?  Was there evidence of individual and community attitudinal or behavior change over time? Can these changes be attributed to PEACE II?  What were the community perceptions of PEACE II outputs and outcomes within the border areas?  Were the ToCs valid? iii  What lessons learned emerged from the evaluation findings and conclusions, and what recommendations can be based upon these findings and conclusions? PEACE II Background The PEACE II program built on the foundations of PEACE I, which was based primarily in the Mandera area of Kenya, and focused on the institutionalization of multi-layered networks of intergovernmental, governmental, non-governmental and community-based organizations, and other representatives of civil society to manage and respond to cross-border conflict. PEACE II sought to expand upon several PEACE I constructs (e.g. peace corridors and groups, such as peace committees and sector-based units such as the Mandera District Livestock Marketing Council), but its key expansion goal was to create an ongoing local presence, not just in Mandera but throughout the border area, to implement its programs. PEACE II established offices on the ground with teams supporting and mentoring community groups, combining this support with provision of grants to these groups. Over its five-year life, the program has evolved programmatically and geographically to adapt to environmental and operational constraints, including the cessation of its programming in Ethiopia due to the passage of new civil society legislation that restricted NGOs from working on democracy, human rights, gender and peacebuilding issues; the presence of Al Shabaab and closure of the border by Kenya; and military engagements between Kenya Defence Forces (KDF), the African Union Mission to Somalia (AMISOM), the Transitional Federal Government of Somalia (TFG) and militia groups, including Al Shabaab. Evaluation Design, Methods and Limitations The evaluation of the PEACE II program employed a mixed-methods approach designed to collect data to inform each of the evaluation questions. To support the data collection, IBTCI sub-contracted Nairobi￾based Research Solutions Africa (RSA) and Somalia-based Horn of Africa Training Institute (HATI). Four main methods and one corollary method were used: 1. Key Informant Interviews (KIIs): The team conducted interviews with more than 105 purposively￾selected key individuals representing the targeted areas of the Kenya-Somalia and Kenya-Uganda border areas (see site collection map below). These included community elders, representatives from women’s and youth groups, religious leaders, business community representatives, members of cross￾border working groups, peace committee representatives, and Government of Kenya officials, including senior police staff. 2. Focus Group Discussions (FGDs): More than 35 FGDs in the targeted border areas (see site collection map below) were conducted with community representatives, including local chiefs (in the town of Garissa), men, women, youth, community elders, inter-faith religious elders, and members of the media, including those representing Reuters, Standard Media, Nation Media Group, Risal FM, and Salama FM. 3. Comprehensive Face-to-Face (F2F) Survey: The team designed and conducted an in-depth F2F survey through which it queried 587 heads of households, and 79 Caretakers in the targeted areas of the Kenya-Somalia and Kenya-Uganda border areas, regarding their perceptions of security in and around their communities, as well as of peace dividend projects, trauma healing/social reconciliation, and sector response unit activities in their area. The household respondents were randomly selected, and represented a combined 4,000 household members. 1 1 The F2F demographic data suggests that while 587 household heads were interviewed, these heads represented all members of their respective households, or over 4,000 men, women and children. iv 4. Desk Study: The team conducted an extensive review of PEACE II program and additional relevant materials, drawing on a range of strategic, programmatic, and performance documents provided by USAID and PACT, as well as those of relevant conflict programs focused on cross-border, nomadic, and pastoralist populations, and of Mission-recommended and reliable non-USAID data sources. 5. Corollary Method - CEWARN Data Analysis: This corollary method was undertaken through the collection and analysis of significant activities data provided to the evaluation team by the Conflict Early Warning and Response Mechanism (CEWARN) office in Ethiopia, and the triangulation of the subsequent findings with those to emerge from the four above methods. This additional method allowed the evaluation team to incorporate objective, third-party data on violent clashes, livestock incidents, raids, etc., as data to augment, refute or corroborate findings from the other methods. Several factors related to security in the targeted sites affected the collection of data, the number of survey respondents interviewed, and the timing of when the data was made available for analysis. Specifically, conflict or tensions in and around Mandera, Belet Hawa, Damassa Lokiriama, Ausqurun and DaresSalaam delayed the team from undertaking its data collection as planned until a later date, required changes in the originally planned site visit order, and created logistical issues for transporting the data back to Nairobi and D.C. for analysis. This resulted in data being delivered for analysis roughly 10 days later than anticipated. Separately, it should also be noted that appropriate and sufficient data did not exist to allow the team to make evidence-based judgments on program impact, and the evaluation team encountered some challenges in establishing correlative linkages between PEACE II activities and community-level outcomes and impacts in part due to the insufficiency of the existing PEACE II M&E data, or due to the existing data being either anecdotal or outputs-based. Kenya Team Sites Visited Somali Team Sites Visited Additional Interviews & Focus Groups Mandera Belet Hawa ElWak ElWak Rhamu Takaba Banissa Wajir Diff Diff Liboi Dobley Hulugho Kolbio Amuma Waldena Garissa Dolo Garbahare Bardhera Afmadu Gherille Gherille Konton DaresSalam Lafey Damassa Badhaadhe Dadaab Habaswein Daadjabula Wadajir Luq Bura Fafi Hosingo Kokani Khorof Harar Suftu Burgabi Kuda Iresteno Shantabaq Khorof Harar Awsqurun SOMALIA KENYA PEACE II Evaluation – Kenya-Somalia Border Map: Field Team Data Collection Sites v Selected Conclusions and Recommendations: Based on the evidence, the PEACE II projects and activities were largely successful as mechanisms to facilitate peacebuilding, trauma healing, social reconciliation, peace networking and cross-border interaction and collaboration between communities. They successfully leveraged existing community peace mechanisms such as the Peace Committees, elders, women’s and youth groups, traditional and culture-specific trauma healing and training mechanisms, and response networks to build additional and enhanced capacities for change in the corridors. The following are selected evidence-based conclusions and recommendations, with the Supporting Findings and Data in the full version of the final report. The selected conclusions and recommendations are categorized by ToC. Conclusion on Cumulative Impact – Perceptions of Security and Cross-border interaction in the border areas The PEACE II evaluation was innovative in that it sought not only to evaluate the performance of the PEACE II program along the Kenya-Somalia and Kenya-Uganda border areas, but also to measure community members’ perceptions of safety and security over time to ascertain with a strong degree of scientific reliability whether or not people felt that their environments changed over time. This was done using proxy measures, which were then used to guide the development of questions for a behavior change survey that was fielded in the border areas as a discovery exercise to measure impact. Overall, there were significant changes in survey respondents’ perceptions of security and cross￾border interactions from 2007-2012, with perceptions of security and of interactions generally improving. There were no marked shifts in perceptions of market life, education, farming and herding, these being subject to market, herding and trade vagaries, as well as natural occurrences such as weather. Perceptions of the ability to move from place to place are, and have been, poor. While the improvement in perceptions of security and cross-border interactions has been in some cases as high as 20-30% toward it being more positive today than five years ago, this shift may be due to several factors that have changed or have been perceived as changing over time, including enhanced infrastructure, such as roads; the enhanced presence of KDF and Somali government forces; a greater capacity to address natural resources demands, supplies, and shortages; and enhanced capacity within the communities themselves as they work with themselves, other communities and/or donors. Peace Dividend Theory of Change Conclusions The majority of Household (HH) respondents in the border areas are aware of the presence of peace dividends, but have not necessarily been engaged in the process. Not surprisingly, all of the KII respondents were aware of the presence of peace dividends and registered an overwhelmingly positive opinion of their impact on the communities. There were some exceptions to this trend, most notably from a former PACT program manager and conceptual developer for PEACE II, who highlighted that the ToC itself was flawed as it was based on the theory that conflict can be mitigated by addressing immediate, albeit joint, needs, rather than the fundamental sources of the conflict itself. According to the HH data, a vast majority of respondents stated that they were aware of peace dividends and were using the facilities, but their perceptions of peace dividend impacts and of the processes themselves are inconclusive. The majority of HH respondents in the PEACE II intervention areas were aware of the presence of peace dividends, but not necessarily that they were PEACE II peace dividend projects (and therefore that they were meant to facilitate cross-border interaction, peacebuilding and reconciliation). There is evidence that HHs in the communities were aware of the existence of new projects and that they had positive perceptions of these projects, but also that these HHs were unaware of the provenance of these projects. More alarmingly, the caretakers, or those whom have been elected by the communities to care for projects once implemented, are also unclear as to the provenance of these projects. This makes attribution vi very challenging. The Community Contracting Process (CCP) is a comprehensive process that has not only mentored communities on procurement and contracting mechanisms, but has mitigated procurement cronyism. But there is evidence that the CCP was not always sufficiently participatory. Access to peace dividends has strengthened the ability of partners to conduct other PEACE II activities. There is evidence of there being in place symbiotic, networked, relationships between the peace dividends and trauma healing, social reconciliation and local peace networks activities. There is a correlation between the existence of peace dividend projects and perceptions of security. But the correlation is inconclusive and warrants follow-on analysis. There is a correlation between the presence of PEACE II and perceptions of the future. The team included in the HH survey a question about peoples’ perceptions of the future, as related to the existence of new projects, and found a positive association between the two. There were also challenges associated with the peace dividends, which resulted in unintended outcomes and impacts. These include a number of non-use cases (although low); instances of inconsistent or insufficient community involvement in the peace dividend process; issues related to access to the projects once completed; and there being no verifiable proof of peace dividend processes actually being successful or unsuccessful, leading the team to some inconclusive findings and conclusions. According to KIIs with PEACE II staff and the partners, and FGDs with community women, men and youth groups, the peace dividend ToC is valid. But verifiable evidence of this is still inconclusive because of missing or deficient M&E tools and practices. Recommendations: 1. The peace dividend process was creative, evolutionary, comprehensive and participatory and its elements should be retained for follow-on programs, and even expanded to comparable programs such as Transition Initiatives for Somalia (TIS). 2. The Participatory Learning Approach (PLA) process and CCP are innovative component steps of the peace dividend process, and USAID should consider using these processes in other donor-funded, community-implemented programs, where transparency and accountability are key. 3. The peace dividend ToC is valid, and the peace dividend projects were perceived as positive by partners and beneficiary organizations. This may be in part due to the fact that this ToC was the only one with truly tangible, measurable – physical – outputs. However, the impact of PEACE II peace dividend projects was difficult to ascertain given the lack of baseline data, the lack of periodic third-party evaluations during the period of performance, the lack of monitoring data to verify the existence and continued use of the projects; and the anonymous nature of the projects. As documentation of these efforts is critical for correlation, let alone attribution, to USAID, definitive conclusions are impossible. The team recommends a rigorous M&E component be included in any follow-on program’s Scope of Work (SOW) to ensure rigorous and verifiable outcomes and impacts to USAID. In addition, the team recommends that an independent, third-party, M&E program be considered to provide ongoing support to USAID to monitor, evaluate – and prove – program outputs, outcomes and impacts. 4. FGD and survey findings showed evidence that peace dividend projects were not always equitably accessible once built, and that the processes in which they are designed and developed was sometimes exclusionary. The team recommends follow-on research to determine the contextual causes for this, as they will affect the successful implementation of a follow-on program. vii Trauma Healing and Social Reconciliation (TH/SR) Theory of Change Conclusions Despite having experienced more intense trauma, Somali border communities are cautiously optimistic about the future and show a greater willingness than Kenyan border communities to forgive and reconcile with those who perpetrated trauma against them. The fact that all of the surveyed Somali sites experienced trauma at far higher rates than their Kenyan corridor counterparts, and that these same sites (and the northern Somali sites in particular) appear to show greater empathy toward those who had experienced trauma at their hands and willingness to accept their group’s role in causing that trauma represents a positive outcome – though not an outright causal linkage – in areas in which PEACE II conducted trauma healing and reconciliation trainings. However, the reasons for these differences in trauma and empathy levels in the survey responses are also likely tied to broader external developments affecting most if not all of the Somali border area. Some individuals with trauma are not benefitting from or aware of trauma healing training. It should be noted that the team was not able to assess the role of trauma and social healing activities in conflict management/transformation from as large a percentage of respondents as desired, given the relatively low numbers who were aware of the existence of trauma healing trainings (approximately 20%). While this could be the product of other factors (e.g., PEACE II had not targeted these communities for trauma healing to begin with owing to greater needs for the training elsewhere), the fact that so many were unaware of the trainings across all of the survey sites, and that so many stated they had experienced trauma, potentially points to a greater need for trauma healing trainings in additional border communities. There are overall indications of positive trauma-healing trends within the border communities, but not all of these trends are consistent. Based on the team’s findings regarding the effects of the trainings from the FGDs, KIIs and survey results, all demographic groups across peace corridors are increasingly willing to empathize, forgive and work with traditional antagonists, and to accept that they had a role in causing trauma to others. More specifically, there is greater positive change within the Somali communities regarding such feelings about trauma and those who have caused it, and regarding the possibilities of reconciliation. Many respondents attributed positive changes to PEACE II activities, including its trauma healing activities. Sizable minorities within the border communities – an average of 33% of those surveyed – retained unforgiving perceptions. Based on an increase in the number of individuals in specific sites who strongly disagree with statements showing empathy or a willingness to forgive antagonistic groups, and that this increase was seen across multiple age groups and geographic areas even as overall numbers for these groups/areas were trending toward greater empathy, there appear to be sizable hardcore minorities that have not been offered trauma or social reconciliation training, are unwilling to take it if offered, or are taking it but are not absorbing its lessons. This 33% figure is an average of the responses of all HH survey participants demonstrating disagreement or strong disagreement with survey statements advocating empathy toward those who caused trauma to their communities. Notably, this average includes a 3% increase in those strongly disagreeing. In specific border environments, TH/SR activities have limited effectiveness. There are settings in which it has been more difficult to implement or maintain progress on reconciliation processes. The validity of the TH/SR ToC could not be validated. This theory of change could not be validated due to the lack of metrics and monitoring conducted. viii Recommendations: 1. As noted above, the fact that a significant number of household and caretaker respondents had experienced trauma suggests that the trainings would be beneficial within these communities. A needs assessment of communities along the border could present a more accurate picture of where such activities would be most warranted. Awareness of such trainings also could be increased in the form of more aggressive promotion by PEACE II-funded groups (e.g., women, religious leaders, youth) and through available media channels (e.g., STAR FM station). 2. There is a need to track trauma healers more closely and identify those who are able to spot participants most resistant to trauma healing (e.g., those most in pain) and create a transformation in their thinking. This could create opportunities to reach specific, more recalcitrant groups, such as the significant minorities described in the TH/SR findings section who remain solid in their strong disagreement with statements advocating empathy and responsibility with regard to trauma. This could also create a better chance of having antagonists present together during a peace process who are genuinely interested in reaching out and finding solutions with the other side (i.e., as opposed to those who are just engaging to “buy time” or strengthen their positions before the next round of attacks). 3. In interviews with representatives of cross-border working groups, examples were provided of receiving additional support from PEACE II in the form of trauma healing training, and subsequently being able to use such training in the management of their peace dividend projects, and even to assist community members with trauma. However, it is not clear from this research what role Community Working Groups (CWGs) are supposed to play in trauma healing, as their representatives also spoke of not being able to travel to – or have people brought to – training locations due to a lack of financial resources. Local Collaborative Peace System & Sector Response Theory of Change Conclusions There are examples of linkages between the existence of PEACE II programming and the extent to which this programming had a relative impact on conflict mitigation and response capacity-building. However, the data do not (yet) suggest a broader (e.g., regional) trend in their impact on cross-border conflict-mitigation and peacebuilding efforts. Instead, they suggest modest progress toward that goal, with the establishment of new or the nurturing of heretofore unsupported groups that are now able to conduct limited, independent, activities in their communities. The work of a smaller number of groups in key areas (e.g., Mandera) have actually been able to influence or guide community peacebuilding efforts (within the ‘Local Collaborative Peace System’ section, see Conclusions 2 and 4, along with their respective sets of supporting findings). The majority of communities’ respondents were aware of and working with conflict resolution groups, but it is unclear if these were PEACE II groups. PEACE II sought to expand PEACE I’s capacity building efforts with local groups and the work of such groups with additional communities. As PEACE II ends, the data demonstrates some progress toward achieving this end, with more than half aware of such groups, and an overwhelming percentage of these reporting that their communities have worked with them and been helped by this support. PEACE II-supported sector response units have played effective roles in collaborative peacebuilding activities within their communities, but they have not yet evolved into the regional, cohesive, mutually-supporting network envisioned under the program. The targeted groups under PEACE II’s sector response strategy, i.e., women, youth, elders, religious leaders, business leaders, etc., have contributed to peacebuilding efforts along the border and to empowering at-risk groups – i.e., women and youth – as part of this process. However, while their intervention in a number of cases highlights the effective work they can perform, the data are not conclusive beyond individual and community-based examples. Peacebuilding trainings could be more effective with individual job training-based components. The benefit of adding job- or vocational-based elements to trauma healing, reconciliation and other interventions ix would be as an incentive to validate and reinforce peacebuilding trainings by the group. One could call these peace dividend projects for the individual. While the majority of partners’ work will continue after PEACE II in at least limited ways, long-term sustainability of some groups is questionable without additional institutional capacity building. The skills in which community representatives have been trained under PEACE II have helped to create a grassroots cadre of proficient practitioners in specific corridors able to conduct peacebuilding awareness and trauma healing exercises, but less mature groups will require additional building of their institutional capacity and mentoring if these groups and their work are to be sustained. The PEACE II “small grants” approach has provided uniform support to a diversified set of grassroots organizations representing key demographic segments of border-area communities, but has also proven problematic for these same groups in expanding or simply maintaining their current level of work. PEACE II’s ‘small-group’ strategy keeps any one of its sector response units from becoming better funded or resourced than another, but also keeps groups from realizing greater potential, or possibly in some cases from surviving. Sector-based conflict response activities have reinforced – and been reinforced by – activities under its TH/SR and Peace Dividend theories of change. Sector response units’ conflict response actions have intersected and been supported by other PEACE II activities, demonstrating a complementary relationship between the program’s theories of change, particularly between Local Collaborative Peace Systems/Sector Response and Trauma Healing/Social Reconciliation. Validity of Local Collaborative Peace System and Sector Response ToC is unclear at present. Evaluators found inconclusive evidence to validate this theory of change, at least at this point in the long￾term implementation of peacebuilding efforts along the border. Recommendations: 1. The PEACE II ‘small-grants’ approach should be reconsidered under any future PEACE program. The team suggests allowing larger grants to be provided to partner organizations that have proven their effectiveness thus far and have presented practical plans for expansion of their work. Directing additional resources to specific groups may engender jealousies from other clans and ethnic groups, but if such groups cannot yet stand on their own once PEACE II ends, then the whole point of helping to establish or strengthen such groups in the first place is moot. Such groups may also be able to secure funding from non-USAID sources, which may generate the same jealousies PEACE II was seeking to avoid anyway; in such a scenario, a successor program would also lose the ability to guide such groups as part of its overall ‘peace network’ strategy. 2. A livelihoods component should be considered for a future PEACE program. The consensus view among FGD and KII participants was that their trauma-healing, reconciliation and other training and mentoring efforts would resonate more strongly with targeted audiences, particularly at-risk youth, if complemented by a work-related component. There is evidence within USAID’s own literature that such tandem efforts yield results under the appropriate post-conflict conditions, and the successor program to PEACE II should consider its inclusion, building on activities such as the ones being implemented by the Somali Support Group and El Wak Youth for Peace. 3. Within a future PEACE program, implementers should create knowledge-sharing services and networks to leverage other sources of support for sector response groups. There have been piecemeal efforts – and, at least in the case of the Mandera Mediation Council, successful ones – to harness support from non-USG sources. The MMC, for example, leveraged the support of the business community in the form of KSH 50,000 and free rent (through the beginning of 2013) on their office space in Mandera. x Strategic Conclusions and Recommendations: Based on the findings and conclusions in the previous sections, below are selected strategic conclusions and recommendations relating to the USAID/EA and PEACE II operating environment that can help to inform future USAID programming in the region. These higher-level conclusions and recommendations evolved through the course of this evaluation, and in particular during focused meetings with USAID/EA staff and PEACE II beneficiaries, as it appeared that there were very important extant, environmental and strategic issues that not only affected the PEACE II program and its impacts, but that could affect the design of a successor program. More specifically they augment the evaluation questions and their responses with conclusions and recommendations that are forward-thinking, holistic and address the “wicked problems” associated with conflict mitigation in complex environments. Finally, the conclusions evolved primarily from interviews with the KII participants and in particular from the very frank discussions with partners, government staffs, and beneficiaries, crossed with the evaluation survey and FGD data. They represent focal areas of current and future interest and concern. Oil and gas exploration in North West Kenya will impact the Kenya-Uganda border area: Recommendation: A stronger outreach to and coordination with the Government of Kenya (GOK) is recommended. Any successor program to PEACE II will need to leverage USAID’s relationship with the GOK to ensure that there is consistency in approaches in mitigating the social and economic impacts of oil exploration in the Karamoja and in mitigating resultant conflicts between groups and communities that already possess grievances with other groups or communities. This will include using consistent approaches, such as using PLA and other participatory approaches, to assess potential impact. USAID and the GOK should ensure that communities are mentored to understand and participate in this new sector, including by developing livelihoods programs and training programs for youth in the oil and gas or services sectors. The Lamu Port-Southern Sudan-Ethiopia Transport (LAPSSET) Corridor will impact the Kenya￾Somalia border areas: Recommendation: USAID should consider incorporating conflict assessment, needs analyses, and impact assessment methods to support the design of any successor program to PEACE II. Successor programs to PEACE II will be faced with needing to understand and mitigate the possible negative social, economic, environmental impacts of the LAPSSET Corridor on border communities in the North Eastern Province, including social upheaval, mass economic migrations, economic disparity and rivalry, competition for limited natural resources, and the impacts on already distressed and violent towns such as Garissa and Ijara, as they continue to become key regional centers for the border communities. A conflict assessment such as one guided by the Conflict Assessment Framework (CAF) 2.0 or Interagency Conflict Assessment Framework (ICAF) is recommended as an initial step in the implementation of a successor program to PEACE II. In addition, successor programs to PEACE II will need to better emphasize the linkage between peace (and stability) and livelihoods programs. In addition to peace dividend projects, cross-border interaction, social reconciliation and trauma healing components as key precursors to peace and stability, successor programs to PEACE II should continue to ensure that the identification of and implementation of such activities in this region are designed to mitigate the potential economic and social impacts of LAPSSET. The shifting nature of conflict on the Kenya-Somalia border and in the “Little Mogadishu” of Garissa has affected stability in the Kenya-Somalia border area: Recommendation: A cross-border community needs assessment and conflict assessment should be considered prior to future programming in the Somali Corridor. As the findings and conclusions above demonstrate, there is strong evidence of positive impacts on communities based on the presence of PEACE II activities. However, an overwhelming majority of KII and FGD participants responded that in addition to cross-border conflict, they are equally impacted by acts of terrorism or by armed clashes between government xi and militia forces in or near their communities. This will undoubtedly have an effect on individuals’ personal perceptions of trauma, reconciliation, and peace in their communities. It will also have an as yet undetermined impact on the lasting effectiveness of PEACE II peace dividends and capacity-building activities. To better prepare for follow-on programming, the evaluation team suggests that tools such as CAF 2.0 or ICAF be used to help guide a comprehensive conflict assessment to inform programming of a successor to PEACE II. In addition to this, it is suggested that USAID consider conducting a robust and participatory needs assessment to better ascertain communities’ priorities relative to peace and security within the context of the new threat of terrorism. There is evidence of conflict “fatigue” and a shift in worldviews in PEACE II communities: Recommendation: A cross-border community needs assessment and/or conflict assessment should be considered prior to future programming in the Somali Corridor and USAID should reconsider worldviews as contextual foundations for the design of successor programs. There are likely several contextual, environmental and historical reasons for the shift in worldviews, and it is recommended that these be explored fully during a comprehensive conflict assessment. For example, one of the reasons for the shift in worldviews to have come out from KIIs and FGDs with elders, youth and women is that there is a general perception that there has been a relative shift in safety and security between 2007 and 2012 from good to bad in Kenya, and from bad to good on the Somalia side of the border. This is particularly evident in Garissa town, Mandera, and Hulugho, where HH survey, FGD and KII respondents expressed disappointment in their worldviews, whereas in Belet Hawa and Dobley there were expressions of optimism. The resiliency and adaptability of local communities is key to future stability: Recommendation: Build the concept of resilience into PEACE II principles and its successor programs. As per guidance in the new USAID policy on resilience, Building Resilience to Recurrent Crisis: USAID Policy and Program Guidance, it is recommended that future PEACE programs in the region frame their objectives, Intermediate Results (IRs) and sub-IRs (or at least a subset thereof) in terms of having an impact on the resiliency of the intervention communities to withstand or adapt to stressors and shocks, and, therefore, that they also establish M&E systems capable of measuring this impact. Page 1 of 54 INTRODUCTION AND PURPOSE International Business & Technical Consultants, Inc. (IBTCI) was contracted to perform the final performance evaluation of USAID/East Africa’s (USAID/EA) Peace in East and Central Africa II (PEACE II) program. The five-year PEACE II program was USAID/EA’s primary conflict mitigation activity in terms of financial investment and profile, and operated from 2007-2012. The Cooperative Agreement (CoAg) was awarded to PACT in September 30, 2007 and the program ended on December 31, 2012, and was managed by the Regional Conflict Management and Governance Office (RCMG). PEACE II program activities were active in multiple peace corridors on the Kenya-Somalia border and in one corridor on the Kenya-Uganda border. This is an evidence-based evaluation that investigates the effectiveness of PEACE II in 1) building community conflict prevention, mitigation, and response capacity and 2) improving security in targeted communities. The evaluation also supported USAID/EA to learn which theories of change (ToCs) proved valid, which ones did not, and where monitoring and evaluation (M&E) data was not sufficient to make a validity determination. The ToCs are hypothetical statements supported by critical assumptions about the communities to which they refer, and can be summarized as follows: 1. Peace Dividend Theory of Change: Through co-managing and sharing tangible development projects, cross-border communities will develop strategic relationships and long-lasting peace networks. These relationships will provide the basis for a joint response when faced with violent conflict in the future. 2. Trauma Healing and Social Reconciliation Theory of Change: Increased stakeholder understanding of broken relationships and trauma caused by conflict will increase stakeholder resilience and leadership in conflict transformation when faced with future violent conflict. 3. Local Collaborative Peace System and Sector Response Theory of Change: Through increasing capacity of local organizations (e.g. peace committees, sector response units) to respond to conflict, a critical mass of peace actors is formed to proactively deal with conflict when it emerges. The evaluation will be used to inform the next generation of USAID/EA conflict mitigation activities. It builds upon previous monitoring and evaluation data and reports already captured by the PEACE II M&E Plan and M&E implementation strategy, as well as previous internal assessments and annual reports, to provide a more contextual, evidence-based, analysis of program outcomes and impacts achieved, and to provide insight into the validity of the program’s ToCs. Because the PEACE II program’s objectives deviated throughout the five-year period of performance, there are some inconsistencies in the data, but these inconsistencies have been mitigated through the evaluation team’s use of fresh data collects and analytical findings. The evaluation was structured around several key questions that USAID/EA was particularly interested in having answered. The complete list of the key questions is in Annex B, and is grouped according to the following themes:  What were the program impacts on security, cross-border interactions and community life in the respective PEACE II intervention communities? Were there unintended impacts? Can these impacts be attributed to PEACE II’s activities?  Was there evidence of individual and community attitudinal or behavior change over time? Can these changes be attributed to PEACE II?  What were the community perceptions of the PEACE II outputs and outcomes within the respective PEACE II intervention communities?  Were the ToCs valid?  What lessons learned emerged from the evaluation findings and conclusions, and what recommendations can be based upon these findings and conclusions? Page 2 of 54 The evaluation report is structured as follows: it commences with a background to the PEACE II program, followed by a section in which the team describes the methods used to collect and analyze the data to support the report’s findings, conclusions and recommendations. This section will be followed by an evaluation of the existing PEACE II M&E tools and techniques, and their appropriateness to inform the evaluation questions. This section will be followed by five discrete, yet inextricably intertwined analytical sections that report on 1) overall perceptions of security in the PEACE II Kenya-Somalia and Kenya-Uganda border area communities; 2) the conclusions, supporting findings and recommendations for each of the three ToCs; and, 3) strategic conclusions and supporting findings that may help to inform USAID/EA future programming in the region. Page 3 of 54 PROJECT BACKGROUND The Peace in East and Central Africa Phase II (PEACE II) program was as a five-year, regional program launched in October 2007 by USAID’s East Africa Regional Mission and implemented by PACT as a conflict mitigation and transformation intervention focusing on key border areas between Kenya, Somalia, and Ethiopia and Uganda divided between the “Somali Cluster” in the east and the “Karamoja Cluster” in the northwest. The program’s two primary objectives were to: 1) Strengthen cross-border security through local community security initiatives; and 2) Contribute to cross-border peace committees’ ability to prevent, mitigate, and respond to conflict in focus border areas. Tensions and conflict along the Kenya-Somali border have long been fueled by inter-clan competition and violence over natural resources and political influence, as well an ongoing lack of investment in education, health and infrastructure, the presence of large cohorts of unemployed youth and weak rule of law. The porous border, which has allowed for the free flow of militias and small arms, along with the presence of Al Shabaab and the Kenya Defence Forces (KDF) (which was fully integrated into the African Union Mission in Somalia (AMISOM) in 2012) to dislodge Al Shabaab from their strongholds) has further exacerbated this situation. The Kenyan-Ugandan border faces similar resource and poverty issues, coupled with cross-border cattle rustling and recurring droughts. PEACE II represented part of USAID/East Africa’s ongoing work to address these issues by working with cross-border communities and local, national and regional conflict management networks in the Horn of Africa to improve the capacity of local communities and intergovernmental organizations to manage conflict in the region. PEACE I to PEACE II: The PEACE II program built on the foundations of PEACE I, which was based primarily in the Mandera area of Kenya, and focused on the institutionalization of multi-layered networks of intergovernmental, governmental, non-governmental and community-based organizations, and other representatives of civil society to manage and respond to cross-border conflict. PEACE II sought to expand upon several PEACE I constructs (e.g., peace corridors) and groups (e.g., peace committees and sector-based units such as the Mandera District Livestock Marketing Council), but its key expansion goal was to create an ongoing local presence not just in Mandera but throughout the border area to implement its programs. PEACE II established offices on the ground with teams supporting and mentoring community groups, combining this support with provision of grants to these groups. A related goal was to create systemized processes on the ground, with the ability to replicate them across multiple border locations. The Participatory Learning Approach (PLA) and the Community Contracting Process (CCP), described in greater depth in the “Peace Dividends Findings, Conclusions and Recommendations” section of this report, were examples of such processes. Program Evolution: Over its five-year life, the program has evolved programmatically and geographically to adapt to changing environmental and operational constraints. In its first year, for example, the program shifted emphasis to programming in the Kenya-Somalia border area (rather than in both the Kenya-Somalia and Kenya-Uganda border areas) and gave a greater role to civil society partners building the capacity of local communities, with support from PEACE II social mobilization grants. Operationally, all three individuals who have served as Chiefs of Party for the program, as well as a majority of partners working with PEACE II at that point, agreed that USAID funding to the program was intermittent, delayed and/or truncated during the first year, and that this resulted in programmatic inertia during that period. During year two, the program also launched many of the major components that defined its work from then until its end: peace dividends, trauma healing, sector response strengthening, and support to the Conflict Early Warning and Response Mechanism (CEWARN) and the country-based Conflict Early Warning Response Unit (CEWERU) system. However, the program was constrained operationally by developments along the Kenyan-Somali border, which became a more challenging work environment after Al Shabaab fully took control of the Somalia side of the border and the Kenyan government officially closed the border. Additionally, Ethiopia enacted and implemented new civil society legislation – called the “Charities and Page 4 of 54 Societies proclamation” – that restricted NGOs from working on democracy, human rights, gender and peacebuilding issues, which eventually compelled PACT with USAID agreement to cease PEACE II’s Ethiopia programming. The succeeding years of the program saw ongoing implementation of peace dividends under the management of newly-established Community Working Groups (CWGs) in the peace corridors established along the border (by the end of the program, there were 10 such corridors along the Somali-Kenya border and one along the Kenya-Uganda border). However, the security context at the border was subject to more active military engagements between the Transitional Federal Government of Somalia (TFG), with AMISOM and the Kenyan Defence Forces (under the AMISOM banner), and Al Shabaab forces. Additionally, Al Shabaab attacks inside Kenya increased and the operating environment remained unstable. Theories of Change: The program was built around three theories of change (ToCs): Peace Dividends, Trauma Healing & Social Reconciliation (TH/SR), and Local Collaborative Peace Systems & Sector Response These ToCs are described in greater detail in their respective sections below, and their themes are captured here:  Peace Dividends – Under the Peace Dividend ToC, cross-border communities working within “peace corridors” along the border, collaboratively identified, developed, implemented and managed much￾needed infrastructure projects. Most peace corridors consist of two townships, one on either side of the targeted border. The intention, through projects such as new school wings, health dispensaries, bore holes, and meeting halls, was that these communities, through their cross-border working groups, would develop strategic, and inter-dependent, relationships that would lead to more sustainable, collaborative networks.  Trauma Healing/Social Reconciliation – Under the Trauma Healing and Social Reconciliation ToC, PEACE II trauma activities focused on increasing individuals’ understanding of cycles of violence and trauma, including giving them a context and language for articulating their grief and anger, with the aim of creating empathy among both perpetrators and victims, which would lead to forgiveness among the relevant parties. SR activities provided a process, through negotiation and implementation of local agreements, for reconciliation between community groups.  Local Collaborative Peace System and Sector Response – Activities under the Local Collaborative Peace System and Sector Response (LCPS/SR) ToC were designed to empower local organizations representing major sectors of society and help them respond more effectively to conflict. More specifically, these trainings and other capacity–building activities were intended to form a viable grassroots network of “first responders” when conflict breaks out, providing fora and outreach for parties to the conflict in which to engage, supply trauma healing and social reconciliation to their, neighboring and cross-border communities, and to function as effective “bridges” between communities and state-run institutions. Page 5 of 54 EVALUATION METHODS & LIMITATIONS I. Team Methods PEACE II’s objectives are based on three theories of change. It is these theories of change (ToCs) – Peace Dividend, Trauma Healing and Social Reconciliation, and Local Collaborative Peace System and Sector Response – that were at the center of this evaluation methodology, and indeed drove its development. Given the difficulties in proving complete and defensible causality between PEACE II’s activities and the desired goals within these theories of change, the team focused on a more testable approach that drew logical connections between 1) the theories of change, 2) the Mission’s Key Questions, i.e., those questions the Mission wanted to have answered through the course of the evaluation, and 3) the team’s practical, operational and proxy questions that it used in its data collection instruments to gather relevant data for analysis. The team then developed a “codebook”, or cross-walk document to ensure that each ToC and each evaluation question was logically and sufficiently addressed by 1) more than one proxy question, and 2) more than one evaluation method. Data Collection Methods & Implementation: The evaluation of the PEACE II program employed a mixed-methods approach designed to collect data to inform each of the evaluation questions. The four main data collection methods utilized were key informant interviews (KIIs), focus group discussions (FGDs), a comprehensive, face-to-face (F2F) behavior-change survey, and an extensive document review. This approach allowed for the verification of the findings through triangulation. To support the data collection, IBTCI sub-contracted Nairobi-based Research Solution Africa (RSA) and Somalia-based Horn of Africa Training Institute (HATI). A more in-depth description of the four individual methods – as well as a fifth corollary method employed – follows below: 1. Key Informant Interviews (KIIs): The team conducted purposive interviews with 105 key individuals representing the targeted the Kenya-Somalia and Kenya-Uganda border area PEACE II intervention communities. These included PEACE II staff, community elders, representatives from women’s and youth groups, religious leaders, business community representatives, members of cross￾border working groups, peace committee representatives, CEWARN heads, and Government of Kenya (GOK) officials, including the head of the National Steering Committee, Administrative Police and Provincial Police senior staff. 2. Focus Group Discussions (FGDs): More than 35 FGDs in the targeted border areas (see site collection map in Annex F) were conducted with community representatives, including local chiefs (in the town of Garissa), men, women, youth, community elders, inter-faith religious elders, and members of the media, including those representing Reuters, Standard Media, Nation Media Group, Risal FM, and Salama FM. 3. Comprehensive Face-to-Face (F2F) Survey: The team designed and conducted an in-depth F2F survey in which it queried 587 household (HH) respondents and 79 caretakers2 - i.e., those responsible for the functioning or management of a community-based infrastructure project – from within the targeted areas of the Kenya-Somalia and Kenya-Uganda border areas - regarding their perceptions of security in and around their communities, as well as of peace dividend projects, trauma healing/social reconciliation, and sector response unit activities in their area. The HH respondents were randomly selected, and represented a combined 4,000 household members. The sample site selection was linked to PEACE II sites, and the size was determined based in part on 2 The team initially targeted 600 HH respondents and 79 caretakers, but local surveyors were unable to reach 13 of the HH individuals in the targeted locations, owing to security-based constraints on the survey team (see ‘Limitations’ below); surveyors also stated they were only able to find 79 of the caretakers within the visited site locations). Page 6 of 54 census data (to the extent this was available; if it was not available, the team extrapolated based on estimates); demographic/population concentration; methodological constraints such as assumptions of data accessibility and validity; and, operational constraints such as time and resource availability, and security and access restrictions. 4. Desk Study: The team conducted an extensive review of PEACE II program and additional relevant materials, drawing on a range of strategic, programmatic, and performance documents provided by USAID and PACT, as well as those of relevant conflict programs focused on cross-border, nomadic, and pastoralist populations, and of Mission-recommended and reliable non-USAID data sources. 5. Corollary Method - CEWARN Data Analysis: This corollary method was undertaken through the collection and analysis of significant activities data provided to the evaluation team by the CEWARN office in Ethiopia, and the triangulation of the subsequent findings with those to emerge from the four above methods. This additional method allowed the evaluation team to incorporate objective, third-party data on violent clashes, livestock incidents, raids, etc., as data to augment, refute or corroborate findings from the other methods. Survey Site Selection – The identification and selection of the final survey sites was based on the results of discussions between the team’s local partner, RSA, as well as with PEACE II management. Through this selection process, it became apparent that the intensity of the program implementation in the various sites was not uniform, with some areas receiving more program attention based on the frequency with which, and for how long, community conflicts occurred. The overall sampling approach was informed by this reality on the ground, so that the various sites were aggregated into four main clusters: 1. Sites with the greatest level of program activities: a. Kenya: Mandera, El Wak, Liboi; b. Somalia: Belet Hawa, Dobley 2. Sites with moderate levels of program activities: a. Kenya: Gherille, Hulugho; b. Somalia: Gherille, Kolbio 3. Sites with limited program activities: a. Kenya: Diff, Khorof Harar, Lafey, Lokiriama (PEACE II Kenya-Uganda intervention site); b. Somalia: Diff, Ausqurun, Damassa; c. Uganda: Nakiloro (Kenya-Uganda PEACE II intervention site) 4. Sites with very limited program activities: a. Kenya: Konton, Amuma; b. Somalia: DaresSalaam, Waldena. In deciding the household target samples per study site, the degree of program activities in the target sites was a significant factor, and the survey purposefully allocated 45% of the respondents to sites with the most activities, 25% to those with moderate activities, 20% to those with limited activities and 10% to those with very limited activities. The 600 target household respondents were assigned a total of 550 for the PEACE II￾targeted Kenya-Somalia border area, (300 in Kenya and 250 in Somalia) and the remaining 50 for the PEACE II-targeted Kenya-Uganda border area. The selected sites and survey respondents from each are provided in the chart below: Page 7 of 54 Survey Samples by Site Kenya Somalia Household Caretaker Household Caretaker Mandera 45 8 Belet Hawa 38 7 El Wak 45 8 El Wak 37 6 Gherille 38 5 Gherille 31 5 Liboi 45 8 Dobley 37 6 Hulugho 37 5 Kolbio 31 5 Diff 20 4 Diff 17 3 Khorof Harar 20 4 Ausquran 17 3 Lafey 20 4 Damassa 16 3 Konton 15 2 DarEsSalaam 13 2 Amuma 15 2 Waldena 13 2 Sub-Total 300 50 Total 250 42 Kenya Uganda Household Caretaker Household Caretaker Lokiriama 25 4 Nakiloro 25 4 Sub-Total 25 4 25 4 Total 325 54 275 46 It should be noted that the survey was designed to analyze response rate and type to form simple statistics, such as mean, median and mode, and more importantly, generalization. It was a purposive survey; the sample consisted of people within only those areas in which PEACE II was active. During the planning stages of the design, the team proposed a more quasi-experimental design, with control and treatment respondent groups, but this approach was not selected. Given the restricted access to certain areas; the lack of government information on population size and concentration; the lack of a baseline or comparative sample; and the small sample sizes in some of the sites, the survey should not be considered statistically significant. Also, the raw data did not allow for stringent P-value tests, T-tests or regressions. These forms of standard statistical tests could not to be consistently performed with all data, although the team did conduct rudimentary one-sample location tests.3 3 Given the small sample size in some of the field locations, the survey did not lend itself to formal statistical reliability, but the comprehensiveness of the combined qualitative and quantitative collection methods enabled the evaluation team to perform evaluative data analysis for response reliability, validation, corroboration and refutation. Note that statistical reliability is dependent in part on there being criteria for repeatability and replicability and while it was an integral and scientifically valid method for this evaluation, true statistical reliability can only be observed when analyzing the data from subsequent surveys. Moreover, the survey was designed to provide quantitative data from HH and Caretaker respondents and as there was no baseline, acted both to provide an indicative data set, and as a proxy baseline for follow-on analyses. Page 8 of 54 Selection of Respondents  Quantitative (i.e. survey) Respondents – For survey respondents, an enumeration area (EA) was first defined by the general area of each respective project. The target households were those presumed to be the immediate/nearest beneficiaries of the project by virtue of being within the said target area. The boundaries of the area were defined by the village guide and/or local administration (chief, assistant chief or village elders), and verified by the field supervisor or team leader as being operationally realistic. The verification was undertaken primarily by assessing how far the given households were from the project in question. Once the general area for a project was defined as above, the target households were then identified and selected using systematic random sampling, through the “left-hand-rule” walk pattern. This survey method drew in part on lot quality assurance (LQA) sampling methods, as well as more generalized stratified sampling methods where smaller numbers of respondents are required.  Qualitative Respondents – For the KIIs and FGDs, the team selected individuals by two processes: 1) the team purposively selected KIIs based on program knowledge and experience of the sector or the region, and 2) whether or not the individuals represented active partner organizations in the corridors. Selection was also based on the type of organization represented, the availability and willingness of the respondents to be interviewed, as well as how long an official had been with the organization (at least six months). Individuals were selected from among three key groups: o Implementing Organizations (IOs) – These were groups active in the targeted area of interest that must have received some grant or funds from PEACE II to manage or implement given community project(s) in the area. o Local Beneficiary Organizations (LBOs) – These were groups made up of members from the targeted local communities, and whose offices and officials were based in the community. These representatives should not have received any donation/grants from any external source connected to PEACE II to manage any community project on behalf of the community. They were not be among the organizations that PEACE II liaised with during the implementation of the given program projects in the target area of interest in the survey. Examples included youth and women’s groups. o Local Beneficiary Community Members – These individuals participated in the study through the FGDs. The participants were required to be persons residing in the general area of the targeted project(s), and were categorized into men’s, women’s and youth FGDs. During the processing phase, the qualitative and quantitative data sets were disaggregated by site, clan, age, and gender. Logistics: In the field, local guides were used in each of the data collection sites to facilitate the identification of target respondents (and other stakeholders such as area chiefs, district chiefs (DCs), etc.), enhance survey teams’ reception into the community and, especially at the household level, provide valuable information regarding the best route plans from one enumeration area to the next. Local guides also identified appropriate FGD venues and helped to invite participants to the sessions. In areas where it was advisable and recommended to move with police escort, the teams liaised with the DC or area chief to hire police officers. This was mainly implemented in the Lower Karamoja border area However, in some sites, such as Mandera, no police escort was recommended, as officers were presumed to be in collusion with the government, or targeted by Al Shabaab or their sympathizers. Thus their presence would have threatened the teams’ safety. In these cases, only the services of the local guides were utilized. Page 9 of 54 II. Limitations Several factors related to security in the targeted sites affected the collection of data, the number of survey respondents interviewed, and the timing of when the data was made available for analysis. Conflict or tensions in and around Mandera, Belet Hawa, Damassa, Lokiriama, Ausqurun and DaresSalaam delayed the teams from undertaking until a later date, required changes to the planned site visit order, and created logistical issues for transporting the data back to Nairobi and Washington for analysis. Specific incidences were reported by field team members in Mandera, Belet Hawa, Kolbio and, in the Kenya￾Uganda border area, at Lokiriama. In Mandera there was an inter-clan fight during the Mandera team’s first day in the field, forcing the team to stop the day’s data collection activities, and based on the advice of the DC, to move to the next sample point - El Wak (K) - until the security situation in Mandera improved. The team left for El Wak (K) the following day and worked the Mandera site last. Al Shabaab attacked Belet Hawa when the HATI team was about to leave the area for Mandera, delaying their departure and forcing the team to take cover. In Kolbio, there were several gun shots heard while the team was in the field, forcing them to detour to the nearest police camp for safety. The fieldwork activities in Lokiriama occurred at a time when Kenyan government authorities were conducting an intensive security operation in the area to weed out persons suspected of having participated in the killing of over 40 police officers in Baragoi area in November 2012. Additionally, the data collection team could not access Ausqurun and DaresSalaam due to the presence of Al Shabaab and their sympathizers. The HATI team, which was working these two sites, was able to make alternative arrangements with the area chief to have additional HH members identified and invited these HH members to survey interviews in El Wak (K). Unfortunately, this resulted in fewer survey interviews than anticipated - only 7 out of 17 household respondents in Ausqurun, and 7 out of 13 from DaresSalaam, respectively, made it to El Wak (K) for the interviews. The team also encountered some technical limitations which are fully described in Annex C and in the M&E section below, and included methodological constraints such as the lack of baseline data, or indeed the lack of a counterfactual or a geographical comparative group with which it could compare the PEACE II findings to determine relative impact. The evaluation team attempted to address these technical limitations (and in particular the lack of baseline indicator or perception data), by designing a rigorous methodology and data collection plan that included qualitative KIIs and FGDs that generated perceptions of contextual and attitudinal change over time, as well as a comprehensive retrospective, or recall, survey that generated perceptions of environmental and attitudinal change over time. Page 10 of 54 MONITORING & EVALUATION: CONCLUSIONS, SUPPORTING FINDINGS AND RECOMMENDATIONS This section provides the evaluation team’s M&E Conclusions, Findings and Recommendations. It is presented as an independent analysis of the PEACE II program’s M&E capabilities, but also as a narrative precursor – or qualification - to the analysis that follows. As will be stated below, the evaluation team encountered some challenges in establishing correlative linkages between PEACE II activities and community-level outcomes and impacts in part due to the existing PEACE II M&E data being either solely anecdotal or solely outputs-based. That said, the existing PEACE II data provided a rich, interwoven, oral and contextual background to PEACE II activities from 2007-2012, reflecting the very human nature of the program’s objectives and impacts. Through KIIs with the partners, especially, the team was able to uncover a nuanced and at times emotional source of information to support its findings. This is also rich information not accessible in standard evaluation methods such as surveys. Conclusion 1: Appropriate and sufficient data was not available to allow the team to make evidence￾based judgments on program impact. As the most recent Chief of Party of the PEACE II program stated in his KII, “How do you measure the absence of violence?” This philosophical question is a challenge for evaluation teams: how does one measure the absence of an event such as conflict? Equally challenging is how one measures attitudinal or behavior-change over time in a conflict environment, and then to prove that this change was, at least in part, a direct and attributable result of a USG-funded program’s many activities. To prove the impact of a program based on the absence of an event –in this case conflict - is surely the larger of the two challenges; to prove attribution is workable, but only if the proper M&E tools, data and analyses are in place at the outset. The evaluation team did have access to a very rich PEACE II outputs data set, but this was insufficient for a full evaluation. This evaluation of the PEACE II program would have benefitted from having access to a clear, logical, RCMG M&E logframe, or results framework, clearly linked to its implementer’s logframe, or results framework (see Annex J). Ideally, the PEACE II framework, and its M&E indicators, would be nested within, and feed into, the RCMG results framework to provide USAID/EA and USAID/Washington with periodic measures of success (or failure) and to illustrate accountability and transparent reporting to USG. In both cases this logical structure of “how” to track indicators as they link causally in the hierarchy of a results framework to USAID/EA goals and objectives (or even to the USAID/Kenya’s Country Development Cooperation Strategy (CDCS)) was lacking, making any retrospective analysis of progress difficult. Lacking too was the logical structure of how USAID, RCMG and PEACE II could effectively measure progress over time, according to pre-defined, and relatively fixed, objectives, Intermediate Results (IRs), sub-IRs, indicators and data collection techniques, i.e., the operationalization of M&E plans for PEACE II. Supporting Findings: Because there were challenges associated with the RCMG Performance Management Plan (PMP) and the PEACE II M&E plan, these findings are separated into Mission and PEACE II findings. Mission: A thorough examination of the RCMG PMP and results framework, and the PEACE II M&E Plan, resulted in the team realizing that the documents were sometimes two years out of date or in draft form, and therefore were not aligned. Linkages between RCMG and PEACE II PMPs and results frameworks were not completely clear, and it was very difficult to ascertain progress according to these documents. The RCMG PMP was lacking linkages to higher level documents, such as a strategic plan, or Mission strategy document. It was also lacking tangible outcome indicators that could be used to indicate progress toward the achievement of its IR and objectives, and more importantly, how, and how many of, these indicators could be tracked by partners such as PEACE II. The evaluation team reviewed the latest version of the RCMG “USAID/EA Conflict Management and Mitigation Program Narrative” and found little, if any, M&E program correlative or connecting linkages to partners such as PEACE II, again making the tracking of progress toward desired impacts difficult, if not impossible. This is not to say that RCMG and PEACE II Page 11 of 54 programming was not linked, or indeed in-line with Mission, or USAID, strategy, but rather that the evaluative components to “prove” the linkages between PEACE II outcomes and RCMG desired outcomes was not evident. PEACE II: Shifting priorities and goals resulted in a failure to accurately measure toward RCMG and PEACE II objectives and results (and impacts). PEACE II benchmarks changed “over and over”, according to the PEACE II M&E officer again making any reliable measure of progress toward desired outcomes and impacts unrealistic. This description is not intended to criticize the means in which the PEACE II M&E system was implemented; rather, it is intended to draw attention to specific findings that lead to conclusions and actionable recommendations for USAID and its subsequent programs. The measure of impacts was made even more challenging as PEACE II had, through the course of its program, abandoned its impact indicators. These were innovative “scorecard” indicators that while difficult to operationalize and then analyze, would have been valuable sources of information for PEACE II and for the evaluation team. Instead PEACE II relied almost exclusively on reporting output and “F” indicator data, which when analyzed is relatively weak, and unfortunately cannot provide a sufficient evaluation of contextual progress in these environments. Within the PEACE II M&E Plan there was an over-reliance on output data and standard “F” indicators that: a) do not measure outcomes and impacts; b) do not effectively indicate behavior change and resiliency in conflict environments. The monthly M&E reports from partners consisted almost exclusively of output data (# of people involved in training, # of activities, etc.), making an evaluation of behavioral change progression very challenging. Finally, this monthly reporting to RCMG was entirely self-reported from the partners, a process that is open to bias, and although some individual reports from partners were very comprehensive, most did not consist of uniformly gathered metrics to allow USAID to envisage progress over time. Conclusion 2: Appropriate and sufficient data was not fully available to allow the team to make evidence-based judgments on changes in security for targeted communities and the relationship between those changes and PEACE II activities. Supporting Findings for Conclusion 2: The above conclusion was derived by the evaluation team after conducting a thorough review of the PEACE II M&E documentation, and finding that there was good and reliable outputs data that could be used for the evaluation to confirm program achievements. However, the evaluation team also found that this data was not sufficient and additional data collection tools and techniques were required. These tools and techniques were subsequently developed by the evaluation team and included a coded KII instrument; a coded FGD instrument; and coded quantitative HH and Caretaker survey instruments that would be used to analyze behavior change over time in the border areas. Recommendations: 1. There should be a clearer logic and a more demonstrable hierarchical flow between RCMG’s PMP and program PMPs, such as that of PEACE II. In addition, both RCMG’s and the program’s M&E Plans should provide more logical and defensible linkages between desired results and actual, real outputs, outcomes and impacts, including who is responsible for what result/outcome and impact, and what organization or individual is responsible for the indicator data collection that helps to determine whether or not the outcome or impact has been achieved. This recommendation is suggested for action for both RCMG’s and its partner programs. 2. Independent M&E and Third Party Monitoring (TPM) are critical to objectively measuring progress. PEACE II was a nuanced behavioral change program that operated in two very complex geographical, ethnic, clan/tribal, resource-driven, environments. Given the complexity of the program and of the environments themselves, the team recommends that RCMG consider tasking objective, third-party organizations with conducting verifications, and monitoring and evaluation activities of its programs, and their subsequent outputs, outcomes and impacts. This recommendation is suggested so as to reduce bias Page 12 of 54 and eliminate self-reporting on critical indicators; mitigate the reliance on anecdotal evidence; and, add scientific rigor to the M&E process. As part of this independent M&E program, the team recommends adding a TPM capability to specifically a) verify the 1st tier peace dividend projects, and their existence, maintenance, and use; and b) verify 2nd and 3rd tier activities such a trauma healing trainings and peacebuilding and reconciliation activities for, for example, participation levels, numbers of participants, numbers and type of outputs and number and types of outcomes. It is common for 1st tier verifications to be conducted by TPMs, e.g., Kenyan, Ugandan or Somali organizations, trained by the Mission’s selected independent M&E provider, and to be equipped with GPS recorders so that their verifications reports can be downloaded to a GIS-enabled clearinghouse capable of analyzing these data to provide the Mission with instant, accurate, verifiable, reporting (see Annex L for an example of a TPM verification report). Finally, it is recommended that USAID consider conducting periodic program performance evaluations through its independent M&E provider. These evaluations would provide USAID with regular rigorous progress checks on its programs, rather than having it wait until a final evaluation that may provide findings too late for corrective action. 3. A behavior change baseline survey, with subsequent time-phased follow-on surveys, is important to measuring behavior change and impact over time. It has been stressed throughout this evaluation that a baseline data set would have provided the evaluation team with a stronger “starting point” for its evaluation of progress, and of impact. In the absence of this, the evaluation team developed a proxy baseline with its HH and Caretaker surveys. As a work-around option, this was scientifically rigorous and programmatically satisfactory. That said, the team recommends that USAID consider commissioning a baseline survey to yield “first-instance data” necessary for any follow-on to PEACE II that wishes to measure behavior change over time, and to consider implementing follow-on surveys to establish trend￾lines indicating progress. The selection of respondents for this baseline will be based on those communities selected for intervention i.e., there are “eligibility requirements” for intervention, and these requirements can also shape a baseline survey. Eligibility requirements include: 1) history of inter- or intra-community conflict; 2) the socio-economic status of residents, 3) willingness of the community to engage in peacebuilding initiatives, 4) clan balance, and 5) other characteristics. Communities not selected for a program intervention will not meet one, some, or all of the implementation criteria and will not be part of the survey sample. By extension, some intervention and non-intervention communities may be significantly different from each other along the four criteria and thus may not be compared using an outcome measure because the difference in outcomes may be due to the criteria and not the intervention. An intervention area may have lower conflict rates because it may have had significantly more willingness in the community to engage in security and development initiatives to begin with. Conversely, conflict rates in non-intervention areas are higher because of the absence of this willingness. It is these types of criteria that the team recommends a future M&E program consider in designing a baseline survey sample for future programs, and to maintain this sample in repeat, time-phased, follow-on surveys to establish trend. 4. More innovative tools for measuring perceptions of progress are needed, and these include multi￾media/video interviews of partners and beneficiaries; GIS enabled data analysis tools; sense maker tools; rapid appraisal techniques, and participatory techniques such as “expert opinion”. GIS enabled clearinghouses can act both as an archive and as an indicator data analysis engine. Other innovative participatory techniques include “expert opinion” sessions that derive perceptions of change in complex environments from multi-staged Q&A exercises with informants like stakeholders, funders, beneficiaries, partners, etc. These sessions combine the Delphi-method and nominal group processes that when conducted in multiple stages yield rich, multi-faceted, responses about change that can be linked to the Page 13 of 54 individual or to their affiliation, e.g., community, clan, tribe, etc.4 Finally, there are models that better allow USAID to evaluate trauma, peacebuilding, social reconciliation and Community Policing Unit (CPU) training programs by applying methods that examine the effectiveness of capacity-building over time, using pre/post tests and panel data analysis. The Kirkpatrick model can be used to evaluate training programs at every phase: – Reaction or what was felt during training – Learning or knowledge gained from training – Behavior change in the community after training – Results or application of peacebuilding training The Kirkpatrick model can also provide a framework to evaluate evidence that a program will be sustainable and will have lasting effects, particularly in the areas of behavior change and results, and therefore inform future, successor programs. Any follow-on or successor program can develop a baseline questionnaire that will collect data relevant to the four levels of the Kirkpatrick model. Illustrative questions that can be in the module include: – Reaction: What were your initial impressions towards PEACE II? – Learning: What additional knowledge in community policing have you gained with PEACE II? (This can also be evaluated quantitatively, e.g. pre/post, with available data.) – Behavior Change: How has PEACE II affected the local environment? – Results: How is PEACE II knowledge being passed on to those who have not experienced it? 4 The type of session recommended draws from the Delphi method in that it is a multi-phased, participatory yet structured, session, intended as an interactive forecasting approach. The session relies on participation from a panel of academic experts, leaders, community leaders, or functional experts (UN, USAID, DfID, etc.) In these sessions, experts respond to written and/or oral questionnaires in three rounds, first individually on paper, then in structured groups, and then finally in a plenary session. After each round, a facilitator provides an anonymous summary of the participants’ opinions from the previous rounds, as well as the reasons they provided for their responses. This structure allows for experts to revise their answers in light of the replies derived during the individual, group and plenary phases. At the end of the final phase, the session is halted and there is an achievement of consensus, and the median or average scores of the final, plenary, round determine the “answers” to the questions. Page 14 of 54 CUMULATIVE IMPACT: PERCEPTIONS OF SECURITY AND CROSS￾BORDER INTERACTION IN THE PEACE II BORDER AREAS I. Introduction The PEACE II evaluation was innovative in that it sought not only to evaluate the performance of the PEACE II program in the Kenya-Somalia and Kenya-Uganda border areas, but, as per the evaluation questions, to also measure community members’ perceptions of safety and security over time to ascertain with a degree of statistical reliability whether or not people felt that their environments changed over time. This was done using proxy measures, which were then used to guide the development of questions for a F2F behavior change survey that was fielded in the border areas as a discovery exercise to measure impact. This survey and its findings do not intend to take the place of an experimental impact method using control and treatment groups, or even quasi-experimental evaluation methods using doubles difference or other research methods to measure impact over time. What it does purport to do is illustrate relative environmental change over time based on respondents’ perceptions of safety and security between 2007 and 2012. Being cognizant not to assert any direct causal relationship between the existence of PEACE II programming and respondents’ perceptions of change, the evaluation team does state that some of the below illustrative findings can be correlated to the existence of PEACE II activities, and thus that PEACE II programs can be seen as having a relative impact on some communities and their capacity to manage conflict. In other sections of this report, the authors draw more specific correlations based on these data and findings to the existence of PEACE II programming in the border areas and their impact on the communities. See, for example the “Local Collaborative Peace System & Sector Response Theory of Change: Conclusions, Supporting Findings and Recommendations” section. II. Findings & Conclusions In this section, the evalution team responded to the following key questions from USAID/EA:  What evidence is there that cross border peace building interventions have strengthened community relationships?  What evidence is there that targeted communities have, or have not, changed the way they manage conflict?  How has the security environment in targeted communities changed? What drove those changes?  What relationship did PEACE II-supported local actors and PEACE II activities have to changes in community security? What can be attributed to PEACE II partners and/or interventions? It should be noted that the evaluation team also addressed these questions in other sections of the report. Conclusions: Overall, there is evidence of significant changes in respondents’ perceptions of security and cross-border interactions from 2007-2012. Perceptions of security and interactions generally rose, and there is evidence that in some cases community relationships have changed or even strengthened. There is also evidence that in some communities people have changed the ways in which they manage conflict, and these mechanisms are more fully described in the ToC sections of the report. That said, there is no evidence of marked shifts in peoples’ perceptions of market life, education, farming and herding, these being subject to market, herding and trade vagaries, as well as natural occurrences such as weather. Perceptions of the ability to move from place to place, i.e., freedom of movement, are, and have been, poor. With respect to attribution, while the rise of perceptions of security and cross-border interactions has been in some cases as high as 20-30% toward it being more positive today than five years ago, this shift may in part be due to PEACE II interventions; or, it may be due to several other factors have changed or have been perceived as changing over time in the sites, such as enhanced infrastructure, such as roads; Page 15 of 54 enhanced presence of KDF, AMISOM and TFG forces; a greater capacity to address natural resources demands, supplies, and shortages; and enhanced capacity within the communities to work with themselves, other communities and/or donors. Supporting Findings (one year and five year patterns): One-Year Patterns: According to the HH and Caretaker surveys, in general respondents along the Kenya￾Somalia border feel more safe and secure than they did one year ago (see figure 1 below), but there are some distinct outliers. In the survey of 587 respondents, most felt more safe and secure in their communites than one year ago. There were some positive outliers. For example, in Konton and El Wak (K), 100% of respondents felt more safe and secure today than one year ago. Interestingly, only 57% of respondents in El Wak (S) felt more safe and secure today than one year ago, but 86% of respondents in DaresSalaam felt more safe and secure today than one year ago. This is particularly interesting as it illustrates that there is consistency in responses in the Konton-DaresSalaam corridor, but not in the El Wak-El Wak corridor. In Khorof Harar, 90% of respondents felt more safe and secure today than one year ago, whereas in Ausquran, 75% of respondents felt more safe and secure today than one year ago, showing some consistency in this corridor. The negative outliers are in the Kenya-Somalia border area communities such as Mandera, where 67% do not feel more safe and secure today than one year ago. In Gherille (S), a resounding 94% responded that they did not feel more safe and secure that one year ago. Dobley had similar findings with 92% of respondents stating that they did not feel more safe and secure than they did one year ago. In Kolboi, Diff (K) and Waldena, respondents also had startling results, with 87%, 88% and 85% respectively stating that they did not feel more safe and secure than they did one year ago. In contrast to Mandera, 66% of respondents in Belet Hawa felt more safe and secure today than one year ago, while ony 33% did not. Given the nature of the PEACE II programs in this corridor and the close geographical promximity between the two sites, these findings are anomalous, and yet may be explained given the November inter-clan clashes in Mandera which coincided with the evaluation team’s data colletion for this site. There are also contexual explanations for this anomaly that are explained in further detail below in the Strategic Conlusions section. There is additional evidence of divergent percetions of safety and security within corridors. In contrast to Gherille (S), 55% of respondents in Gherille (K) felt more safe and secure today than one year ago. In contrast to Dobley, 67% of respondents in Liboi felt more safe and secure today than one year ago, illustrating a divergence of perceptions in corridors with recent PEACE II activities. Similarly, there were Figure 1. Page 16 of 54 divergences in other corridors. In contrast to Kolboi, 63% of respondents in Hulugho felt more safe and secure today than one year ago. In contrast to Diff (K), 65% of respondents in Diff (S) felt more safe and secure today than one year. In contrast to Waldena, 67% of respondents in Amuma felt more safe and secure today than one year ago. The precise causal reasons for these divergences is beyond the scope of this evaluation. However, as noted in the peace dividends ToC section below, the authors suggest that there is a correlative relationship between the presence of peace dividend activities and peoples’ perceptions of safety and security. A particularly intriguing finding is in the Kenya-Uganda border area, where there was a sizeable divergence between respondents in Lokiriama and Nakiloro when asked if they feel more safe and secure than they did one year ago. In Lokiriama only 40% of respondents felt more safe and secure today than one year ago, whearas in Nakiloro, 72% of respondents felt more safe and secure today than one year ago. In the FGDs, KIIs and F2F surveys, there are data leading to concusions that suggest that certain respondents on the Kenya side of the Kenya-Uganda border area have a more pessimistic view of security than those on the Uganda side (see figure 1 at the bottom of the previous page). There is another illustrative set of findings when the data is further disaggregated. When respondents were asked if they feel more safe and secure than they did one year ago, female and male respondents responded slightly differently, with 57% of men responding “yes”, and 51% of women responding “no” (see figure 2, on the top left of this page). When the data was disaggregated by age, there is an interesting demographic trend: 65% of youths aged 16-25 stated that they did not feel more safe and secure today than one year ago. This negative response rate changes with the next age groups. 53% of adults aged 26-35 stated that they felt more safe and secure today than one year ago; 54% of adults aged 36-50 stated that they felt more safe and secure today than one year ago; and, 59% of adults aged 51 and older stated that they felt more safe and secure today than one year ago (see figure 3, bottom right). This implies that youths and women in the border areas feel less safe and secure in their communites than do adult males. Five-Year Patterns: In addition to asking HH and Caretaker respondents general questions about perceptions of safety and Figure 2. Figure 3. Page 17 of 54 security in their communities today compared to one year ago, the evaluation team asked the respondents about their perceptions of daily life today compared to five years ago according to seven factors:  Security  Market Life  The Ability to Move from Place to Place, or “freedom of movement”  Farming  Herding  Cross-border interactions with other neighboring communities  Education By analyzing the responses based on these criteria, the evaluation team sought to better understand overall attitudinal or behavior change over time. In addition, the team triangulated the F2F findings with those from the document review; the KIIs; and the FGDs, where similar or comparable questions were developed and then answered. In the absence of a pre-existing survey baseline, the evaluation team adopted this method, which while not statistically significant, is scientifically rigorous. Should there be a follow￾on cross-border program comparable in scope and proposed impacts to PEACE II, the F2F questionnaires can help to inform the development of a baseline survey for such a program. When the evaluation team analyzed the demographic data sets from the F2F survey in more detail, the evaluation team identified significant findings and conclusions relating to security and cross-border interaction. Data associated with questions in the F2F survey on topics such as perceptions of market life, farming, herding and education yielded findings that suggest that there was little, if any, change in peoples’ perceptions over time. As will be further described below, the data sets associated with the question of how respondents felt about their ability to move from place to place now and five years ago yielded generally inconclusive findings and did not allow the evaluation team to come to any definitive conclusions. In all cases, respondents claimed that there was little difference in their perceptions of the ability to move from place to place, i.e., between communities; to market; between family; to their herds; to a peace dividend project, etc., today than there was five years ago. More interestingly, the question was presented as a Likert scale question, with five possible responses: Very Good; Good; Neither Good nor Bad; Bad; or, Very Bad, and in all cases there was a relatively equitable distribution of responses between all possible responses. The findings below illustrate changes in perception in each of the corridors, in both border areas, using data from questions that asked the respondents how they felt about key issues today and five years ago (see Annex E for the full question sets and the survey instruments). Figure 4. Page 18 of 54 When asked their perceptions of security now and five years ago, an overwhelming 78% of respondents felt that the security situation was very good or good today (see figure 4, middle left). That said, 1% of respondents, and 9% of respondents felt that the security situation in their community was very bad, or bad today, for a total of 10% of overall respondents feeling the security situation was negative. 13% and 22% of the population felt that the security situation in their communities was very good or good respectively five years ago, for a total of 35%. 17% and 24% felt it was bad or very bad for a total of 41% feeling it was negative overall. Overall findings are that there is a perception that the security situation is significantly better today than it was five years ago, which is consistent with the team’s analysis of CEWARN data for the Kenya-Somalia border area (Kenya side).5 When analyzed, these data suggest that there was a downward trend in armed clashes along the border from 2007-2012 (see figure 5, top left)6 . That said, there was a sharp and very distinct rise in organized raids during the period, and an exponential rise in the number of human casualties reported. This leads to several conclusions (see figure 6 below). First, although the CEWARN data only reflects the Kenya side of the Kenya-Somalia border area, there is a clear rise in human casualties along the Kenya-Somalia border (although the sites are not specified). Second, although there appears to be a decline in “armed clashes”, there was a rise in “organized raids”, i.e., inter-clan or inter-community conflicts. Third, and most revealing, despite the rise in violent raids and in human casualties in the Kenya-Somalia border area (Kenya side), the F2F data indicates that people in the border communities in this area have a positive perception of security. The obverse argument is that when one analyzes the outlier data from Mandera, Hulugho and other Kenya borders sites, these are cases in the Kenya-Somalia border area (Kenya side) where negative perceptions of security correlate to the rise in the “organized raids” and in human casualties as reflected in the CEWARN data. In the Kenya-Uganda (Karamoja) border area two positive trends to emerge from the analyzed CEWARN data. In both livestock lost and human deaths, there was a significant downward trend from 2007-2012, save for a spike in deaths on the Kenya side in 2011 (see figures 7 and 8, on the middle of the following page). This is consistent with HH and 5 The evaluation team was granted access to raw CEWARN data for the Kenya-Somalia border area (Kenya side). Data from the Somalia side of the border area was immature and not available for this report. 6 Fourth quarter data (July-December, 2012) was not available at the time of the writing of this report. Figure 5. Figure 6. Page 19 of 54 FGD responses from Kenya-Somalia border area, but it should be noted that this actual decline does not correlate to the negative perceptions of security for respondents from Lokiriama on the Kenya side. An analysis of this anomaly is further discussed the Strategic Conclusions section below. In terms of peoples’ perceptions of cross-border interactions, 33% of the surveyed population felt these to be very good or good five years ago, while 56% of respondents felt that cross-border interactions were very good or good today. Overall, cross-border interactions are better today than they were five years ago. Again, this rise in community respondents’ perceptions of cross-border interactions corralates both to the overall decline in violent incidents as per the CEWARN data, and to overall perceptions of security in the border areas (outliers not included), and one can draw correlations between the presence of PEACE II programming in the border areas and this trend, but the team wishes to stress that there is at present no scientifically rigorous causal linkage between the presence of PEACE II activities in the border areas and these trends (see figure 9, at the bottom of this page). Figure 9. Figure 7. Figure 8. Page 20 of 54 PEACE DIVIDENDS THEORY OF CHANGE: CONCLUSIONS, SUPPORTING FINDINGS AND RECOMMENDATIONS I. Introduction From the outset in 2007, the PEACE II program developed the concept of “peace corridors” to focus its cross-border conflict mitigation work to specific geographical corridors. These peace corridors are discrete geographic zones with notable historical cross-border activity, as well as evidence of cross-border or inter￾clan conflict. Most peace corridors consist of two townships, one on either side of the targeted border. The peace dividend ToC assumes that through co-developing, co-managing and co-maintaining much-needed infrastructure projects, the communities within the peace corridors will develop strategic, and inter￾dependent, relationships that would lead to more sustainable, collaborative networks. These relationships would then act as joint frameworks for responding to conflict, leading to a more lasting and strengthened cross-border security. There were 42 peace dividend projects in the corridors, for a total of 123 completed structures including schools, classrooms, maternity hospitals, dispensaries, water pans, water reticulation works, wells, etc. (see Annex K for the complete list of USAID-funded peace dividends). The PEACE II program developed a comprehensive, transparent, and accountable, process for planning and delivering peace dividends in the corridors, which included the following steps: receiving the initial request for the peace dividend from the communities; conducting consultative meetings with the communities; the selection process of local partners; the training of these partners in the PLA approach; the conduct of the actual PLA process itself as a guide to educating the community members in the collaborative processes involved in the prioritization and budgeting of the infrastructure project; the conduct of the Community Action Planning process which guides the proposal and eventual implementation plan for the project; the development of a community contracting manual; undertaking the community contracting process to identify and contract for the construction of the project; and, the construction of the project itself. There is also a step in the process for monitoring and evaluating the peace dividend, once operationalized, but the team did not see any evidence of PACT conducting this follow-on activity verifying use or maintenance of a specific peace dividend. The evaluation team noted several key outcomes and impacts associated with the peace dividend ToC, and the resultant processes for its realization:  Peace dividend projects enhanced or strengthened cross-border interactions  There is a significant awareness of peace dividends among the border area communities  The CCP is a comprehensive process that has not only mentored communities on procurement and contracting mechanism, but has helped to mitigate procurement corruption  According to all of the KIIs (PACT staff, implementing partners, and beneficiaries) and the majority of FGDs, peace dividends help to bridge communities in conflict and acts as “triggers” for collaboration, healing, and social reconciliation, but the HH survey responses are less conclusive There were also challenges associated with the peace dividends, which resulted in unintended outcomes and impacts. These include a low number of non-use cases; instances of inconsistent or insufficient community involvement in the peace dividend process, and issues related to access to the projects once completed; and, there being no verifiable proof of peace dividend processes actually being successful or unsuccessful, leading the team to some inconclusive findings and conclusions. Page 21 of 54 II. Findings and Conclusions In this section, the evalution team responded to the following key questions from USAID/EA:  What have been the key impacts, intended or unintended, of peace dividends on cross-border community relations?  What evidence of improved relations and coordination between cross-border communities attributable to peace dividends can be documented?  Is there evidence that peace dividend projects impacted security in the immediate cross-border area? If so, what was the impact?  How did the community contracting process influence the success or failure of the peace dividend process? How did it affect relationships within and between cross-border communities?  Are there lessons to be learned regarding peace dividend activities?  Did the peace dividend theory of change prove valid? If it did in some locations, but not others, to what was the difference attributable? Conclusion 1: The majority of HH respondents in the border areas are aware of the presence of peace dividends, but have not necessarily been engaged in the process. Not surprisingly, all of the KII respondents were aware of the presence of peace dividends, and registered an overwhelmingly positive opinion of their impact on the communities. There were some exceptions to this trend, most notably from a former PACT program manager and conceptual developer for PEACE II, who highlighted that the ToC itself was flawed as it was based on the theory that conflict can be mitigated by addressing immediate - albeit joint - needs rather than the fundamental sources of the conflict itself.7 According to the HH data, a vast majority of respondents stated that they were aware of peace dividends, and were using the facilities, but their perceptions of peace dividend impacts and of the processes themselves are inconclusive. Supporting Findings: 74% of HHs surveyed, or 432 individuals, were aware of new projects being started in their communities recently. Of the FGDs, all except for the Belet Hawa youth were aware that there were new projects in their communities. It should be noted, however, that HH respondents’ awareness of new projects does not indicate awareness of a PEACE II project. The data from the HH survey suggests that there have been several new projects in the communities, only some of which were PEACE II peace dividend projects. As with the other ToCs, attribution of outcomes and impacts directly to PEACE II is therefore unclear. Of these 432 people, 55 were actively engaged in some capacity, with 25% acting as committee members; 2% as PACT/PEACE II community mobilizers; 9% as “other” mobilizers; and the vast majority as labor support to the project. Mandera town had the highest proportion of those engaged, at 14.5%, and El Wak (K), Diff (K), Ausquran, Lafey, Damassa, Waldena, each have no engagement at all. In the Kenya￾Uganda border area, Nakiloro had 9 survey respondents say that they were involved in the peace dividend process, a number that heavily eclipses the proportion within many of the Kenya-Somalia border area site responses. Conclusion 2: The majority of HH respondents in the border areas were aware of the presence of peace dividends, but not necessarily that they were PEACE II peace dividend projects (and therefore that they were meant to facilitate cross-border interaction, peacebuilding and reconciliation). There is evidence that HHs in the communities were aware of the existence of new projects and that they had positive perceptions of these projects, but also that these HHs were unaware as to the provenance of these 7 KII with Simon Richards (11/1/2012) Page 22 of 54 projects. More alarmingly, the Caretakers, or those whom have been elected by the communities to care for projects once implemented, are also unclear of the provenance of projects. This makes attribution very challenging. Supporting Findings: As the successive PEACE II Chiefs of Party and the PEACE II program manager have confirmed, PEACE II peace dividend projects were intentionally not marked or branded.8 This was partly to mitigate possible retaliation or targeting of project staff or beneficiaries, and, relatedly, to avoid any association between the projects and USG. It was also designed so as to facilitate greater ownership of the projects from the communities. Of course, in a region where there are several other donors, this strategy also made it difficult to ascertain attribution for PEACE II. For example, HH and FGD respondents often did not refer to PEACE II projects at all. When asked about new projects, the Mandera men’s FGD alluded to a much-lauded U.K. Department for International Development (DfID)-funded social hall in Belet Hawa and the mediation center in Mandera. In Dobley, the men’s FGD respondents stated that they were very pleased with both recent classrooms and a community hall, but neither of these were PEACE II peace dividend projects. The Liboi, Kolbio and Hulugho women’s FGDs stated the positive impact that recent classroom, school and social hall projects had on their communities – projects implemented by CDF, AFREC and CDF respectively. Conclusion 3: The Community Contracting Process (CCP) is a comprehensive process that has not only mentored communities on procurement and contracting mechanisms, but has in some cases mitigated procurement cronyism. But, there is evidence that the CCP was not always sufficiently participatory. Supporting Findings: KIIs with the PEACE II Project Engineer, and other PEACE II staff, as well as with partners, indicated that the CCP was a robust, fair and transparent process.9 The process has been standardized in a manual for use by other programs in war-torn countries.10 This sense of fair practice and transparency was also evident in the communities. For example, in an interview with Abdinoor Hussein of the El Wak Youth for Peace, he provided evidence that the CCP for the youth center was not only fair and transparent, but that as a consequence of this fairness had resulted in a contract award to the only person in the community who had vociferously complained that the CCP was corrupt. This was a common sentiment among KIIs with CCP participants in projects along the Kenya-Somalia border. Interestingly, however, of the 74% of HH respondents who were aware of recent peace dividend projects in their communities, only 26% felt that the process for selecting the contractors for the project, e.g., using the CCP, had sufficient participation from the community, and 38% said that it did not. 35% did not know. This suggests that community members are not fully supportive of the CCP, or that they feel it is insufficient, or that the CCP may have affected cross-border communities negatively. However, these findings are inconclusive: first, there is a very high percentage of respondents stating that they “don’t know”; second, given that there are respondents in several FGDs who are not certain of the organization(s) responsible for the recent projects in their communities, one cannot be certain that the 38% of HH respondents who feel that the process for selecting the contractors for the projects are referring to the PEACE II CCP, or another donor’s mechanism for contracting. In 9 of the 30 FGDs conducted by the evaluation team, respondents claimed that their communities were involved in some capacity in the contracting and/or peace dividend process. However, in some cases such as in Hulugho especially, there were accusations of “unfair practices” being used. In Hulugho, both the men’s 8 KIIs with Angi Yoder-Maina (10/29/2012); Nikolai Hutchinson (11/2/2012); Jebiwot Sumbeiywo (11/2/2012); Wendy Marshall (11/14/2012) 9 KII with Isaac Njunguna (11/2/2012) 10 PACT/ACT!, Community Contracting in War-Torn Countries Handbook, 2011 Page 23 of 54 and the women’s FGD respondents stated that the selection and building of the projects was done so using “unfair” practices. As stated above, however, this sentiment could be referring to other USAID programs such as Transition Initiatives for Somalia (TIS), DfID or other donor-funded peace dividend projects. Conclusion 4: Access to peace dividends has strengthened the ability of partners to conduct other PEACE II activities. There is evidence of there being in place symbiotic, networked, relationships between the peace dividends and trauma healing, social reconciliation and local peace networks activities. Supporting Findings: There is a universal perception among respondents - PEACE II staff, partners and beneficiaries alike – that there was a correlation between communities having access to peace dividend projects, and these same communities being able to further other PEACE II activities such as trauma healing and social reconciliation. When asked whether the existence of these peace dividend projects changed their lives and acted as change mechanisms themselves, the majority of FGD respondents states “yes”, except for, coincidentally, the Belet Hawa men’s FGD and the Mandera men’s FGD, both of which stated a resounding “no” that the peace dividend projects had no major impact and did not help to facilitate other activities. There is also evidence that the physical structures themselves served, and continue to serve, dual purposes, being used for several cross-border initiatives. For example, in El Wak (K) the youth center that was built also acts as a resource center for the community and for trauma training activities. KIIs with PEACE II staff and partners also clearly identified the linkages between communities having access to peace dividends, and being able to further other PEACE II activities such as trauma healing and social reconciliation. For example, in an interview with a former PEACE II program officer, he stated that peace dividends themselves were “trauma-healing and social reconciliation mechanisms,” emphasizing that wherever there was a peace dividend project, there was a TH/SR mechanism in place, all as nodes in the collaborative peace system.11 In an interview with the most recent and final PEACE II Chief of Party, he stressed the concept that peace dividends projects were not about “giving people things”, but about facilitating change in the communities, and that this change was incremental and part of a holistic process of peacebuilding and reconciliation.12 Conclusion 5: Peace dividend projects can provide a catalyst for additional cross-border community communication & coordination Supporting Findings: In all of the PEACE II staff and partner KII responses, there was concurrence that the presence of peace dividend projects facilitated or enhanced additional cross-community coordination activities. The FGDs were also overwhelmingly positive in their perceptions that peace dividend projects had 11 KII with Ahmed Sheikh (10/31/2012) 12 KII with Nikolai Hutchinson (11/2/2012) Figure 10. Page 24 of 54 facilitated cross-border interactions, especially in cases where the peace dividends created a natural inter￾dependency between the communities, and those that were co-managed through the Cross-Border Working Groups. According to the HH respondents, here too there is an overwhelming perception that interactions with neighboring communities have changed, and that there is a correlation between the existence of a project and cross-community interactions. As figure 10 (see previous page) indicates, in those communities where there was a new peace dividend project, 62% of HHs felt that there has also been a change or a significant change in how they interact with cross-border communities. Conclusion 6: There is a correlation between the existence of peace dividend projects and perceptions of security. But the correlation is inconclusive, and warrants follow-on analysis. Supporting Findings: As mentioned above, the majority of survey and FGD respondents had positive perceptions of the security situation in their communities, but there were some outliers. The evaluation team also received generally positive responses from the partner KIIs.13 However, when analyzing the correlation between the existence of a peace dividend project and perceptions of security, the team received mixed responses. Of the 432 border area site respondents who stated that they were aware of new projects in their communities, 77% felt that the security situation in their community was very good or good, compared with 78% of the 587 overall HH respondents who felt that the security situation in their community was also generally very good or good. This implies that the existence of a new project had no or very little impact on their perception of security (and that 155 people were either unaware of any projects in their community or were unsure). 57% of those aware of new projects perceived cross-border interaction as very good or good, which is only a small majority. Of the 432 who responded that they were aware of new projects in their communities, 34% felt that the security situation in their community was very good or good five years ago, compared with 33.8% of the overall respondents who felt that the security situation in their community was very good or good now. Conversely, for those 155 people who responded that they were not aware of any new projects in their communities, 80% felt that the security situation in their community was very good or good, compared with 78% of overall respondents who felt that the security situation in their community was very good or good now, again implying that the existence of new projects in the border areas had no or little impact of peoples’ perceptions of security from 2007-2012. Conclusion 7: There is a correlation between the presence of PEACE II and perceptions of the future. As a primary question to determine stability or progress, in most HH surveys in fragile or conflict environments, the team included in the HH survey a question about peoples’ perceptions of the future, and then cross-tabbed these findings to responses relating to the existence of new projects. Supporting Findings: During KIIs with partners and beneficiaries, the majority of respondents stressed that the peace dividend projects with which 13 It should be noted that to avoid bias, in the KIIs with PEACE II staff the team did not ask for the staff to draw correlations between the existence of a PEACE II project and their perceptions of security. Figure 11. Page 25 of 54 they were most familiar had a very positive impact on their lives and the lives of those in their communities. There was also demonstrated evidence of pride, ownership, responsibility, and optimism about the future. There were some outliers, and these are discussed above and in the Conclusions section of the report below. Similarly, the team asked respondents about their perceptions of the future, and the majority of respondents answered with a positive response. When these responses were analyzed alongside data that confirms the existence of new projects, 87% of respondents felt positive about the future. This is a remarkably high proportion, and suggests that there is a correlation between the existence of new projects and peoples’ perceptions of the future. That said, it should be noted that there is no indication that the projects to which they HHs are referring in these data are PEACE II projects, or that these projects are associated with larger, more holistic, peacebuilding or conflict mitigation programs. The projects could be traditional development infrastructure projects. Conclusion 8: There were challenges associated with the peace dividends projects, resulting in unintended outcomes and impacts. These challenges form the basis for lessons learned. As mentioned above, despite the successful implementation of the peace dividends ToC, there were some associated challenges. Supporting Findings: According to KIIs with PEACE II staff, and in particular the PEACE II project engineer, there were a small number of non-use cases of peace dividend projects.14 It should be noted that this is a very small number and is therefore not indicative of poor planning or mismanagement. In two cases, non-use was due to project location, e.g., the Mandera Market and the Belet Hawa Market. For the Belet Hawa Market, non-use was also due in part to insecurity in the area, and partly due to it being planned too far from the community it was intended to serve. There is also evidence of a small number of non-use cases due to a lack of qualified staff to man them once completed, e.g., the Konton dispensary and the Morato dispensary. In these cases, it should be noted that the identification of qualified staff to man the projects was not within the mandate of the PEACE II program specifically, although it was in part the responsibility of the community to ensure that their projects were sustainable. This responsibility could have been better trained during the PLA process or during consultative processes, including PEACE II training on community advocacy with the government to ensure reliable and sustainable staffing for these projects, as well as continued government support and buy-in once the projects have been implemented. Finally, there were two cases of non-use due to other, exogenous, circumstances, e.g., the Dobley School is no longer being used after a nearby water pan dried up. According to the interview with the PEACE II project engineer, there were some cases of peace dividend projects being completed without site visits, leading to unintended outcomes such as, for example, damage caused by projects being implemented on a floodplain. As mentioned above, according to the HH survey only a small number of respondents in the communities claimed to have been involved in the peace dividend process, and this is a contradiction to the assertions of 14 KII with Isaac Njunguna (11/2/2012) Figure 12. Page 26 of 54 PEACE II staff and partners. In addition, some HH survey and FGD respondents claimed that access to peace dividend projects was exclusionary, and not universal, once the projects were developed and in place. This was particularly the case in Hulugho and Liboi. In addition, when asked the question: “did you feel that the process for developing and maintaining this project involved sufficient participation from your community?,” the Gherille women’s; El Wak (K) men’s; El Wak (K) women’s and Hulugho women’s FGDs all answered “no”. Of course, ensuring universal access to projects was also not within the mandate of the PEACE II program, but it is worth noting these examples, as well as the prospect that there may be underlying causes for this exclusion, e.g., inter-clan related or otherwise, that can be explored in follow-on USAID programs. As mentioned in the M&E section above, the most significant challenge associated with the peace dividend ToC is that there was no monitoring or verifications program in place within the PEACE II program to ascertain with some degree of accuracy, accountability and rigor that: the peace dividend projects were actually being built on time and on schedule; the CPP was conducted according to the guidelines developed by PEACE II; the PLA and other participatory methods were being used effectively and produced actionable outcomes; any evidence of graft or inconsistencies in the CPP were flagged and that any subsequent resolutions were documented; the project was is use after being built, and that it was being used in the manner in which it was intended to be used; some degree of verifiable, documented, follow-up be in place to demonstrate PEACE II adherence to “value for money” principles for USAID. Conclusion 9: According to KIIs with PEACE II staff and the partners, and FGDs with community women, men and youth groups, the peace dividend ToC is valid. But verifiable evidence of this is still inconclusive because of missing or deficient M&E tools and practices. Supporting findings: According to the KIIs with the partners, peace dividend projects provided key resources to communities traditionally in conflict, and provided tangible evidence of mutual benefit when both communities buy into the notion that peace dividends are part of the collaborative peacebuilding process. According to the KIIs with the partners, the projects also had identifiably positive impacts on individuals and communities, and cross-border interactions. The findings from the FGDs and the survey are less conclusive, and suggest instead that peace dividends may be welcomed by communities, but may have little or no impact on peoples’ perceptions of security. The challenge for the evaluation team was that there was no baseline data from which to measure attitudinal, let alone behavioral, change in the border area communities from 2007-2012. For much of the evaluation, evidence of positive impact was provided by the partners or the PEACE II staff, using anecdotal evidence, and often relying on memory. FGD and survey data augmented these data sources, and provided data for key findings related to the peace dividend ToC, but these data also revealed some inconclusive findings that will require additional analyses to more definitively, and rigorously, show impact. Moreover, because of the lack of, or inconsistent use of, M&E principles and tools, the team was unable to draw completely definitive correlations between the existence of peace dividend projects and peoples’ perceptions of security in the border areas. Finally, based on the evidence it did receive, and the analysis conducted thus far, the team suspects that the peace dividend ToC is a fully valid one, but this is a presumption that cannot be fully supported given the lack of baseline data, monitoring reports, observer reports and verifications reports that would have provided substantial evidence of project use and impact over time. III.Recommendations: 1. The peace dividend process was creative, evolutionary, comprehensive and participatory and its elements should be retained for follow-on programs, and even expanded to comparable programs such as TIS. 2. The PLA process and CCP are innovative component steps of the peace dividend process, and USAID should consider using these processes in other donor-funded, community-implemented programs, where transparency and accountability are key. Page 27 of 54 3. The peace dividend ToC and the projects themselves were perceived as positive by partners and beneficiary organizations, and this may be in part due to it being the only ToC with truly tangible outputs and outcomes. However, the impact of PEACE II peace dividend projects was difficult to ascertain given the lack of baseline data, the lack of periodic, third-party, evaluations being conducted during the period of performance, the lack of monitoring data to verify the existence and continued use of the projects; and, the anonymous nature of the projects. As documentation of these efforts is critical for correlation, let alone attribution, linking program impacts causally and exclusively to USAID is impossible. The team recommends a rigorous M&E component be included in any follow-on program’s SOW to ensure rigorous and verifiable outcomes and impacts to USAID. In addition, the team recommends that an independent, third-party, M&E program be considered to provide ongoing support to USAID to monitor, evaluate - and prove - program outputs, outcomes and impacts. 4. FGD and survey findings showed evidence that peace dividend projects were not always equitably accessible once built, and that the processes in which they are designed and developed was sometimes exclusionary. The team recommends follow-on research to determine the contextual causes for this, as they will affect the successful implementation of a follow-on program. Page 28 of 54 TRAUMA HEALING & SOCIAL RECONCILIATION THEORY OF CHANGE: CONCLUSIONS, SUPPORTING FINDINGS AND RECOMMENDATIONS I. Introduction Under the Trauma Healing and Social Reconciliation Theory of Change, PEACE II trauma activities focused on increasing individuals’ understanding of cycles of violence and trauma, including giving them a context and language for articulating their grief and anger, with the aim of creating empathy among both perpetrators and victims; SR activities provided a process, through negotiation and implementation of local agreements, for reconciliation between community groups. Findings demonstrate some evidence of achievement of short￾term outputs related to these aims, as well as establishment of the foundation for long-term impacts. In particular, the team identified individual- and community-based examples and trends demonstrating:  Forgiveness of (or willingness to forgive) past grievances, or the willingness to ask for forgiveness and accept compensation;  Understanding and application of TH & SR principles by PEACE II-supported groups; and  Empowerment of key groups, such as women, through leveraging of their training to address past trauma through non-traditional bodies. However, the challenge under this ToC was in identifying – and attributing to PEACE II trauma healing and social reconciliation activities – broader trends demonstrating empathy on a more widespread scale both geographically and demographically, as well as comprehensive reconciliation efforts leading to long-term solutions to persistent regional disputes and normalization of relations among involved communities. II. Findings & Conclusions In this section, the evalution team responded to the following key questions from USAID/EA:  What have been the key impacts, intended or unintended, of PEACE II trauma and social healing activities?  What evidence of changes in behavior/attitudes with regards to people’s views of conflict and their personal response attributable to PEACE II activities can be documented?  What are community perceptions vis-à-vis the role of trauma and social healing activities in conflict management/transformation? What examples do community members cite to support their views?  Are there lessons to be learned regarding the trauma and social healing activities?  Did the trauma healing theory of change prove valid? If it did in some locations, but not others, to what was the difference attributable? Conclusion 1: Despite having experienced more intense trauma, Somali border communities are cautiously optimistic about the future and show a greater willingness than Kenyan border communities to forgive and reconcile with those who perpetrated trauma against them. A similar trend was not evident along the Kenyan-Ugandan border. The fact that all of the surveyed Somali sites experienced trauma at far higher rates than their Kenyan corridor counterparts, and that these same sites (and the northern Somali sites in particular) appear to show greater empathy toward those who had experienced trauma at their hands and willingness to accept their group’s role in causing that trauma represents a positive outcome – though not an outright causal linkage – in areas in which PEACE II conducted trauma healing and reconciliation trainings. However, the reasons for these differences in trauma and empathy levels in the Page 29 of 54 survey responses are also likely tied to broader external developments affecting most if not all of the Somali border area. Such developments would likely include incremental progress toward more democratic governance in Somalia, including the shift during the summer and fall of 2012 from transitional governance to the newly formed parliament and selection of the new president and cabinet, as well as the expulsion by AMISOM and the KDF of Al Shabaab from Kismayo, their last major urban stronghold within the country, as well as large sections of the Somali-Kenya border. Such events were also cited within the Somali FGDs and KIIs as reasons for cautious optimism. Supporting Findings: Of the two-thirds (390) of HH respondents and nearly three-quarters (57) of caretakers who said they had experienced trauma, both stated that it had come in the form of the death of a family member or friend (27% HH; 28% caretakers). While this was the most common answer, 16% of households simply reported experiencing “fear,” while 14% of households and 18% of caretakers said they were subject to conflict between clans, and 6% and 9%, respectively, said they had been forcefully evicted or displaced.15 Men and women, as well as those 16-25, 26-35 and 36-50 had experienced trauma at similar levels – between 66-68% of them answered affirmatively. Less than 10% of interviewees had ever taken vengeful or other actions against members of other communities as a result of this trauma. Nearly 58% of respondents said they had dealt with their trauma experiences through prayer, while 14% had internal discussions with family, and 18% said they had joined support groups. Across nearly all of the peace corridors, the Somali site respondents consistently reported experiencing trauma at a significantly higher level than their Kenyan counterparts. This was most notable in the Diff Corridor (25% of Kenyans versus 94% of Somalis), Khorof Harar-Ausqurun (20% versus 100%), Konton￾DaresSalaam (20% versus 100%), although differences of close to 50% were common in the majority of other corridors. Interestingly, only 53% of those interviewed from Mandera stated that they had undergone trauma (97% in Belet Hawa said they had), suggesting differing interpretations of what rises to that level of shock. Other evaluation data supported this finding, with focus groups on the Somali side of each peace corridor more consistently and comprehensively speaking of trauma within their communities than their Kenyan corridor counterparts, with particular emphasis on occupation of their community by Al Shabaab and/or militias, or fighting among these groups or against the KDF. With regard to Mandera, for example, the male FGD participants said they had not experienced any trauma within the last five years, while Belet Hawa male FGD participants stated that although the overall level of security now was “not bad,” there had been more conflicts and violence in the Belet-Hawa District and the broader Gedo region over the last five years, including fighting between clans, cross border communities and federal government and Al Shabaab militants. Dobley men spoke of the two militia groups who fought for control of their town and injured or displaced several community members. They stated that “in dealing with such issues, we just pray to God to keep us safe as we wait for the war to end and see who has won so that we know which rules and regulations to follow.” In Gherille-Somalia among both men and women, people were traumatized by the many killings from the fighting between the KDF and Al Shabaab forces, which included KDF planes strafing the area with gunfire and forced them to “live in the bush.” The Kolbio Women’s FGD stated that militias had “ruled their town” for almost two years, and that they were being forced to support Al Shabaab in their fighting with other militias. Within the Kenya-Uganda border area, survey respondents in both Lokiriama and Nakiloro experienced 15 Nearly three-quarters of the surveyed caretakers (57) said they had experienced trauma, with nearly the same percentage as households (28%) also reporting the death of a family member or friend. 18% of caretakers said they too were subject to clan conflict, while 9% said they had been evicted against their will. Page 30 of 54 trauma at the same levels (50%), with both demonstrating a high degree of willingness to forgive those who had caused them trauma (84% among Nakiloro respondents and 97% among those in Lokiriama agreed or strongly agreed with statements advocating forgiveness and payment of reparations). On both sides of the border, the loss of livestock played a significant role as a source of their trauma in several ways: First, the theft of animals and killing associated with this theft fueled cycles of retribution to avenge these losses, according to the Lokiriama women’s FGD; second, the loss of these animals deprived them of a means of livelihood and essential sources of food, which led to “cancerous poverty,” the Nakiloro men’s FGD reported. Conclusion 2: Individuals with trauma are not benefitting from or aware of trauma healing training. It should be noted that the team was not able to assess the role of trauma and social healing activities in conflict management/transformation from as large a percentage of respondents as desired, given the relatively low numbers who were aware of the existence of trauma healing trainings (approximately 20%). While this could be the product of other factors (e.g., PEACE II had not targeted these communities for trauma healing to begin with, owing to greater needs for the training elsewhere), the fact that so many were unaware of the trainings across all of the survey sites, and that so many stated they had experienced trauma, potentially points to a greater need for trauma healing trainings in additional border communities. Supporting Findings: Only some 20% of those surveyed were aware of any recently held events, discussions, or trainings designed to help individuals deal with trauma, with the remainder either not aware or unsure,16 although of those who were aware, nearly 85% had participated in such trainings themselves (i.e., about 17% of the total survey sample of 587), with 79% having participated once in the past year, 16% twice, and fewer than 5% having done so more times than this.17 Of those who had participated, more than half (54%) described the event, discussion or training as one focusing on peacebuilding; a much smaller percentage (7.5%) described it as focusing on trauma healing. The most participants came from Gherille￾Somalia and Belet Hawa, while some of the lowest participant numbers came from their opposite communities in Gherille-Kenya and Mandera. This trend was also apparent among the FGD participants: In these same corridors – i.e., among the youth and women’s groups in Mandera and all groups in Gherille￾Kenya – and within other corridors – i.e., among the youth and women in Liboi and all groups in Hulugho, no FGD respondent who had experienced trauma had participated in any events or discussions to help them cope with their feelings; in contrast, all of the FGDs representing their counterpart cross-border communities – i.e., Belet Hawa, Gherille-Somalia, Dobley, and Kolbio – reported having participated in such events/discussions. Approximately 20% of those surveyed had participated in a training or event enabling the discussion or resolution of past or current disputes over natural resources, trade/migration routes, and other issues. Of this group, 62.7% had taken it once, 28% twice. The men’s FGDs in both Gherille-Somalia and Dobley both said their communities had participated in peace and reconciliation meetings in Malindi, and that they had been organized and led by PACT, DAI, (for the Dobley FGD only) TIS, and others. All of the communities from lower Juba were involved in their meetings, according to the Gherille-Somalia group. In contrast, no members from either the Liboi Women, or Gherille-Kenya men or women’s FGDs reported participating in such trainings. Conclusion 3: There are overall indications of positive trauma-healing trends within the border communities, but not all of these trends are consistent. Based on the team’s findings regarding the 16 A slightly higher percentage of caretakers – 41% – had been aware of these events. 17 Of those caretakers who were aware of such events, 81% had participated in these trainings, with 53.8% having done so once within the last year, 26.9% twice, and a combined 19.1% more than this. Page 31 of 54 effects of the trainings from the FGDs, KIIs and survey results, all demographic groups across peace corridors are increasingly willing to empathize, forgive and work with traditional antagonists, and to accept that they had a role in causing trauma to others. More specifically, there is greater positive change within the Somali communities regarding such feelings about trauma and those who have caused it, and regarding the possibilities of reconciliation. Respondents attributed positive changes to PEACE II activities, including its trauma healing activities. However, it should also be noted that these changes could also be attributable in part to the positive developments described above that are taking place in Somalia, especially since the degree of change among the Somali communities was higher than among the Kenyan communities. Credit to PEACE II for positive changes in feelings of empathy and forgiveness, must also be viewed as part and parcel of the broader changes the program has sought to effect through the combined activities under all of its theories of change. The ability to effectively deliver these trauma healing and social reconciliation trainings to communities most in need, for example, has been strongly tied to PEACE II’s efforts to build the capacity of sector response units and related groups, which have been trained and provided resources to conduct these trainings (and which fall under the ‘Local Collaborative Peace System & Sector Response’ ToC explored under the next section). It therefore remains difficult, other than through examples, to attribute clear changes in people’s views and behavior with regard to trauma and conflict solely to PEACE II’s trauma healing activities. Supporting Findings: Among those responding to the team’s survey, a majority who had taken the trauma￾healing training – 79% – said it had changed their views positively, with the greatest number explaining that it had improved harmony among their and other communities. A slightly smaller number (71%) said it had altered their behavior in positive ways as well.18 Qualitatively, however, respondents were more measured in describing how the trainings had changed their views and behaviors. Some in the Belet-Hawa men’s group, for example, said they could now forgive their enemies, but others said they would still take action in order to make them feel the same pain they had caused; the Belet-Hawa youth group’s response was similarly mixed, with two of them saying they would forgive those who perpetrated violence or conflict against them, three of them saying they would react and take action against these individuals, and the rest of the group saying that they would organize seminars for the “perpetrators” and train them on the importance of peace. The Gherille-Somalia men’s FGD stated that the relevant training conducted by PACT Kenya had been good for the community, but that it had only been held once and that more were needed. The El Wak-Somalia men noted that their training led to the creation of social network between the communities and eradicated bad attitudes both within the community and in the cross-border communities. The Kolbio men stated that their discussions had played a key role toward changing the behavior of those affected by trauma, and by doing so, had helped them to develop trauma coping mechanisms. FGD, KII and survey respondents were separately asked about their views after having participated in social￾reconciliation trainings. 81% stated that the training had helped them to build trust or solve problems with traditional antagonists. Qualitatively, respondents were more consistently positive about the benefits they had taken away from the trainings. The Belet Hawa Women’s FGD, for example, emphasized the importance of such trainings, stating that “these mechanisms act as agents of change,” and had changed the views and perceptions of the communities on peace. Beyond the training, 22% had contributed their own time or money to hold a reconciliation event themselves. Within the Dobley men’s FGD, for example, respondents noted that members of their community had interceded in the conflict between the militia then controlling their town and those constructing the town’s new maternity ward, one of the PEACE II peace dividend 18 Among caretakers, 100% said the training changed their views positively, while 88% said it altered their behavior in a positive way. Page 32 of 54 projects being implemented within the Liboi-Dobley Corridor. Their participation contributed to the resolution of the dispute over the new project and facilitated its completion. While interview and FGD responses articulated general support and appreciation for the SR process, there were few concrete examples of impact, and some were dubious about the effectiveness of its processes because the degree of “buy in” from participants was unclear. “The problem is that much of social reconciliation – in the form of peace processes – can be superficial,” said PEACE II’s Chief of Party. “It’s never clear whether or not the two sides have a stake in the process and are ready to reconcile or are just ‘regrouping’ and planning on renewing the conflict in the near future.” In terms of changes in levels of empathy, all of the Somali peace corridor sites showed at least some change toward greater support over the last year for empathy statements, either through decreases in disagreeing views, increases in agreeing perspectives, or usually some combination of both. FGDs with those in and around the Somali sites also frequently demonstrated a more comprehensive understanding of the principles and or the benefits they accrued, in some cases in the absence of any participation in trauma trainings by their peace corridor partner sites across the border (e.g., Gherille-Somalia versus Gherille-Kenya, according to both the men’s and women’s FGDs from both sites). The aforementioned differences in responses on the Somali side were underscored by several examples among the Somali FGDs – Dobley men, Gherille men – of cautious optimism expressed regarding the new Somali government and its ability to stabilize security along the border in the future. Conclusion 4: Sizable minorities within the border communities – an average of 33% of surveyed populations – retain unforgiving perceptions.19 Based on an increase in the number of individuals in specific sites who strongly disagree20 with statements showing empathy or a willingness to forgive antagonistic groups, and the fact that this increase was seen across multiple age groups and geographic areas even as overall numbers for these groups/areas were trending toward greater empathy, there appear to be sizable hardcore minorities that have not been offered trauma or social reconciliation training, are unwilling to take it if offered, or are taking it but are not absorbing its lessons (see figure 13, top of the following page). Alternatively, it appeared harder for this minority to agree with statements requiring pro-active gestures (e.g., asking for forgiveness) or admissions of culpability regarding their own group members’ roles in damaging the whole group. 19 The 33% figure is an average of all household individuals across all survey sites disagreeing or strongly disagreeing with the statements posed in two survey questions (C13 and C14), the first assessing the individual’s willingness to forgive past antagonists, the second the willingness to make reparations. 20 The average number of those strongly disagreeing in their responses to these two questions rose from 14 to 16%. Page 33 of 54 Supporting Findings: Respondents were asked a series of questions designed to assess how their views on trauma had changed over a one-year period, both from the perspective of those who had experienced or been peripherally touched by trauma, and of those who had caused trauma to others. Responses among all demographic groups demonstrated increasing willingness to empathize, forgive and work with traditional antagonists, and to accept their own role in causing trauma to others. However, there were notable exceptions to this among certain groups and in specific sites along the Kenyan-Somali and Kenyan-Ugandan borders. Among both men and women and across a majority of corridor sites, for example, significantly more strongly agreed with the statement that they would feel no sympathy if they saw a member of the other group suffer now than one year ago, even as the overall number of those disagreeing strongly rose and those “agreeing” with the statement fell. On the Kenyan side and part of the Somali side of the border, the percentage of those strongly disagreeing or agreeing with statements advocating asking for forgiveness from or making reparations to the “other” group, or admitting that the actions of a few in their group damaged the majority, increased considerably (e.g., in some cases rising 10-15%, and increasing to 30-35% of the total sample population for that site), or stayed the same compared to one year ago. The only exceptions to this were in Belet Hawa, El Wak-Somalia, Damassa and DaresSalaam, which showed only increases in support for these statements; Ausqurun had only very small increases in disagreeing (i.e., with the statement on few causing damages to many in the group). Together, these communities constitute the northern half of the Somali-side PEACE II peace dividend sites. This same set of questions was broken down by each survey community visited, revealing several significant trends within particular corridors:  Mandera – Belet Hawa: There was little correlation in the findings between the two communities. In Mandera, there were significant increases from one year ago in those “strongly disagreeing” with statements demonstrating empathy and their own responsibility with regard to trauma; more or less no changes in levels of those “agreeing” from a year ago; and significant decreases in those “strongly agreeing” with these statements. In Belet Hawa, the opposite was true, with the number of those Figure 13. Page 34 of 54 disagreeing and disagreeing going down considerably and a corresponding increase taking place in those agreeing/strongly agreeing. There was also strong support regarding hope about the future, which jumped from 10% to 71% of the sample. This is supported by FGD findings from Belet Hawa, where the women’s group spoke positively about security in the area compared to in previous years, as well as about cross-border interactions (due to social interaction such as women’s peace committees) and market life (due to interaction through trade).  Liboi-Dobley – Liboi showed increases in levels of empathy and responsibility, although respondents were slightly less hopeful about the future; in contrast, Dobley showed noticeable increases in those strongly disagreeing across a range of these questions, with 18% to 43% of the sample holding these views.  Hulugho-Kolbio – Both Hulugho and Kolbio showed noticeable spikes in those strongly disagreeing with empathy statements; this was particularly true in Kolbio, where those strongly disagreeing with statements advocating asking for forgiveness from the “other” group and demonstrating willingness to make reparations increased to 35 and 38% of the sample, respectively.  Diff (Kenya-Somalia) – Diff Kenya showed increasing support for empathy and responsibility, while Diff Somalia demonstrated a significant increase in those strongly disagreeing with these statements, especially with regard to making reparations to the other group (those strongly disagreeing jumped from 12 to 53%).  Khorof Harar-Ausqurun – Khorof Harar demonstrated notable jumps in those strongly disagreeing with empathy and responsibility statements, and no respondents showing any agreement with the statement that a few members damaged their whole group. Ausqurun respondents were overall in agreement or strong agreement with these statements.  Lokiriama-Nakiloro – Both communities showed a significant decrease in those agreeing that they were “hopeful about the future” (e.g., Lokiriama by 50%). While Lokiriama also showed significant drops in people disagreeing and strongly disagreeing with the survey’s empathy and responsibility statements, there were no corresponding upticks in those agreeing or strongly agreeing with these statements, leaving the largest chunk of their responses as “not sure.” In Nakiloro, there was slightly more support for these statements, but the majority of respondents was also not sure. Conclusion 5: In specific border environments, trauma healing & social reconciliation activities have limited effectiveness. There are settings in which it has been more difficult to implement or maintain progress on reconciliation processes. Supporting Findings: Within the Dadaab refugee camp in the Garissa-Lower Juba area, PEACE II￾supported groups such as the Garissa Mediation Council and Garissa Women for Peace have been engaging with camp-based youth and women stakeholders in peacebuilding, networking and empowerment activities. This has been viewed as an important area in which to provide project support, as there is a broad perception that Dadaab has become a recruitment ground for forces in Somalia, including Al Shabaab.21 However, the sheer scale of the refugee population in Dadaab’s six camps (approximately 430,000), its accompanying humanitarian issues, and a “volatile and unpredictable” security situation,22 when combined with limited resources and access to the camps by PEACE II groups, and the lack of peace dividend projects to serve as 21 PEACE II Quarterly Report (April 1, 2012 – June 31 2012). Pg. 9. 22 UNHCR Situation Report - Refugee Camps in North Eastern Province, Alinjugur & Dadaab Sub-Offices (July 17-31, 2012) Page 35 of 54 an “anchor” for other PEACE II activities, has made conducting trauma and reconciliation work there challenging at best, according to interviews with local partners.23 It should be noted that such environments can be temporary in nature, and may exert a transitory – albeit sizeable – negative effect on individual perceptions regarding their ability to empathize with or forgive past antagonists. Drawing from the team’s individual corridor findings, Mandera demonstrated a significant increase in opposition to statements of empathy and forgiveness, while Belet Hawa demonstrated nearly the opposite change. This disparity can likely be attributed, on the Mandera side, to the timing of the survey, which was administered just after the November 2012 violence that engulfed the town. On the Belet Hawa side, the findings are at least in part due to the recent developments in Somalia mentioned above, and the guarded hope this has generated. In the Kenya-Uganda border area, findings based on the empathy/forgiveness questions revealed the Corridor’s ambivalence and uncertainty about the future, but within the FGD and survey findings, many attributed this uncertainty to recent violence related to cattle rustling, and generally expressed optimism regarding the future based on assistance they were receiving that lessened competition or conflict over pastureland for cattle, migratory routes and water points. Conclusion 6: The validity of the Trauma Healing and Social Reconciliation Theory of Change could not be validated due to the lack of metrics and monitoring conducted. Supporting Findings: The Trauma Healing and Social Reconciliation Theory of Change states that increased stakeholder understanding of broken relationships and trauma caused by conflict will increase stakeholder resilience and leadership in conflict transformation when faced with future violent conflict. The team found inconclusive evidence to suggest that this ToC had been proven valid as part of the PEACE II program, owing in part to the lack of baselines, targets and other metrics established to assess increases in stakeholder understanding of broken relationships and conflict-based trauma, and to measure increases in resilience and leadership among stakeholders in a conflict transformation context. Additionally, monitoring of those who had participated in trauma healing, social reconciliation and other trainings was inconsistent. Much of this is also linked to the fact that the sector response units and related groups conducting these trainings had insufficient PEACE II funding to allow them to track the post-training progress of those who had participated in their training events. III. Recommendations 1. Awareness of TH trainings/activities needs to be strengthened. As noted above, the fact that a significant number of household and caretaker respondents had experienced trauma suggests that the trainings would be beneficial within these communities. A needs assessment of communities along the border could present a more accurate picture of where such activities would be most warranted. Awareness of such trainings also could be increased in the form of more aggressive ‘advertising’ by PEACE II-funded groups (e.g., women, religious leaders, and youth) and through available media channels (e.g., STAR FM station). 2. Trauma-Healing-to-Social-Reconciliation transition monitoring needs to be strengthened. Need to track trauma healers more closely and identify those who are able to spot participants most resistant to trauma healing (e.g., those most in pain) and create a transformation in their thinking. This could create opportunities to reach specific, more recalcitrant groups, such as the significant minorities described in the TH/SR findings section who remain solid in their strong disagreement with statements advocating 23 Such groups also noted the importance of addressing immediate (e.g., humanitarian) needs as paramount to stabilization, and at times a necessary precursor to peacebuilding. Page 36 of 54 empathy and responsibility with regard to trauma. This could also create a better chance of having antagonists at the “table” during a peace process who are genuinely interested in reaching out and finding solutions with the other side (i.e., as opposed to those who are just engaging to “buy time” or strengthen their positions before the next round of attacks). 3. Continue strengthening and clarifying post-peace dividend roles of CWGs. In interviews with representatives of cross-border working groups, examples were provided of receiving additional support from PEACE II in the form of trauma healing training, and subsequently being able to use such training in the management of their peace dividend projects, and even to assist community members with trauma.24 However, it is not clear from this research what role CWGs are supposed to play in trauma healing, as their representatives also spoke of not being able to travel to – or have people brought to – training locations due to a lack of financial resources. 24 KII with Safia Bare, Hulugho Cross-Border Working Group (11/10/2012); KII with Abdinoor Haji, Diff Cross￾Border Working Group (11/7/2012). Page 37 of 54 LOCAL COLLABORATIVE PEACE SYSTEM & SECTOR RESPONSE THEORY OF CHANGE: CONCLUSIONS, SUPPORTING FINDINGS AND RECOMMENDATIONS I. Introduction Sector response units were created or supported to increase the capacity of local organizations representing key demographic sectors to respond to conflict; more specifically, such groups were intended to form a viable grassroots network of “first responders” when conflict breaks out, providing fora and outreach for parties to the conflict in which to engage, supplying trauma healing and social reconciliation to their neighboring and cross-border communities, and to function as effective “bridges” between communities and state-run institutions, and in some cases counter-balances to these institutions at the local level, such as peace committees. A number of examples and trends identified by the team demonstrate that efforts under the Local Collaborative Peace System & Sector Response Theory of Change have led to:  Easing of tensions with police and Kenyan defense forces;  Increased cross-border relations among key groups;  Empowerment of women’s and youth-based groups; and  Ability of sector response units to counter-balance ineffective peace committees.25 PEACE II was effective at building localized community conflict prevention, mitigation and response capacity. While there are examples of the work that PEACE II-supported groups have been doing on and around the targeted border areas that indicate ties to the program’s capacity-building activities, the data do not (yet) suggest a broader (e.g., regional) trend in their impact on cross-border conflict-mitigation and peacebuilding efforts. Instead, they suggest modest progress toward that goal, with the establishment of new or the nurturing of heretofore unsupported groups that are at this point able to conduct limited activities in their communities, and the work of a smaller number of groups in key areas (e.g., Mandera) actually able to influence or guide community peacebuilding efforts. Among this smaller set, the team did find examples in which PEACE II capacity-building efforts in certain communities, and among certain groups (particularly religious leaders and youth), were tied to new or stronger responses to conflict and processes designed to mitigate the spread of violence (see Conclusions 2 and 4, along with their respective sets of supporting findings, below). This was the case in Mandera, for example. However, it should be noted that even among this smaller set, data demonstrating a direct, causal link between this increased capacity and a prevention or earlier end to violence was not evident. II. Findings & Conclusions  What have been the key impacts, intended or unintended, of PEACE II local peace organization capacity building? 25 Although PEACE II provided grant support to specific district peace committees in certain instances and they were viewed as a necessary component of the Peacebuilding System, there was general concern among PEACE II implementers that the DPCs had become extensions of the state, permanent homes for aspiring politicians, and vehicles for inter-clan warfare rather than a genuine bridge between the state and the communities as originally developed. For example, the Garissa DPC was viewed as having become an inter-clan platform for a land dispute between competing local clans; in Mandera, divisions of administrative boundaries along clan-based lines created additional sub￾municipalities, each with its own DPCs advancing that particular clan’s interests. Page 38 of 54  What have been the key impacts, intended or unintended, of PEACE II sector response activities, e.g. with women, youth, religious leaders, and media groups?  Is there evidence that stakeholders are proactively responding to conflict when it emerges?  Can conflict response actions be attributable to PEACE II sector response capacity building? Can such actions be attributed to other aspects of the PEACE II program?  Are there lessons to be learned regarding local peace organization capacity building and sector response activities?  Did the sector response theory of change prove valid? If it did in some locations, but not others, to what was the difference attributable? Conclusion 1: The majority of communities’ respondents were aware of and working with conflict resolution groups, but it is unclear if these were PEACE II groups. PEACE II sought to expand PEACE I’s capacity building efforts with local groups and the work of such groups with additional communities. As PEACE II ends, the data demonstrates some progress toward achieving this end, with more than half aware of such groups, and an overwhelming percentage of these reporting that their communities have worked with them and been helped by this support. Of these, nearly 50% said they were familiar or working with PEACE II-type sector-response groups (i.e., women’s and youth groups, CPUs), which represents a potentially robust presence for PEACE II groups within these communities. However, there are caveats to these findings below based on questions of attribution to PEACE II. Supporting Findings: A minimal but key criterion for assessing the degree of penetration of the peace system and sector response groups within the targeted border communities was their level of awareness of the work of such groups. Overall, 57% of HH respondents were aware of their work. Familiarity was relatively close to this percentage among both men and women (60/55%), and among three out of the four age groups, with the 16-25, 26-35, 36-50 groups showing 57-60% awareness (the 50-and-above age group’s awareness stood at 48%). However, awareness was not consistent across the peace corridor sites. At five sites (Liboi, Gherille-Kenya, Damassa, Ausqurun, and Amuma), more respondents were not aware of groups whose role it was to respond to conflict than were aware. The highest percentage of respondents stating that they were familiar with such groups was found in Lafey, Konton, Dobley, Belet Hawa, and Lokiriama. Of those who were aware of the existence of these groups, respondents’ were by far more aware of the existence of peace committees (44%) than other types of groups, although sizeable minorities were also aware of women’s groups (20%) and youth groups (19%). Among those who were aware of such groups, 86% stated that these bodies had worked with their communities. All but a few of the sites surveyed reported similarly high numbers among site respondents; at nine of the sites, 100% of the interviewees stated such groups had worked with them, with only four sites (in the Lafey and Damassa Corridor, El Wak-Kenya, and Lokiriama) under 70%. Respondents overwhelmingly (48%) stated that this work had come in the form of these groups conducting peace meetings. Other examples of peace groups’ work in respondents’ communities included the conducting of general trainings/seminars (11%), the raising of peace awareness (10%), the initiation of peace committees, and mobilization of the community (both 5%). The overwhelming number of respondents (98%) indicated that working with such groups had helped their community, with the type of help dependent on the peace corridor in question. In the Nakiloro-Lokiriama area of Karamoja, for example, such groups had assisted in reducing fighting over pastureland for cattle, migratory routes and water points, lessening cattle rustling, and in enabling mediation for the return of stolen livestock. However, it should be noted that far more of these respondents - 44% - said they were familiar or working with traditional, Kenyan-government peace committees. Moreover, although the three types of sector-based Page 39 of 54 groups described by respondents were the types of groups being supported by the program, it was unclear from the data if the three were specifically PEACE II-supported groups. Conclusion 2: PEACE II-supported sector response units have played effective roles in collaborative peacebuilding activities within their communities, but they have not yet evolved into the regional, cohesive, mutually-supporting network envisioned under the program. The targeted groups under PEACE II’s sector response strategy, e.g., women, youth, elders, religious leaders, etc., have contributed to peacebuilding efforts along the border, and to empowering at-risk groups – i.e., women and youth – as part of this process. However, while their intervention in a number of cases highlights the effective work they can perform, the data aren’t conclusive beyond individual- and community-based examples. It is also unclear in many of these cases how much of their positive roles within these examples can be attributed specifically to PEACE II support,26 since a number of such groups existed before the implementation of PEACE II (and in some cases, before PEACE I), have several sources of support, and have built up knowledge and expertise independent of this program. Supporting Findings: The following analyses of the work of several key groups that were supported at least in part by PEACE II demonstrates groups able to train, advocate and collaborate on behalf of the sectors and communities and solid progress toward establishing the foundation for the aforementioned network, assuming support continues under PEACE III or from a similar program in the future. Women’s Groups: Women’s groups have made important contributions to cross-border peacebuilding efforts with PEACE II support and within the context of PEACE II’s LCPS/SR theory of change, yet also face ongoing challenges for stepping outside of “traditional” roles from conservative segments of the border communities. Nevertheless, 1 in 5 survey respondents were aware of or cognizant of their communities working with such women’s conflict-resolution groups, and the team’s findings include a number of positive demonstrations of their roles, which fall into three main categories:  Training: Women’s groups who have been trained in trauma healing, social reconciliation, leadership and communication skills and other relevant themes, have in turn provided training to other women’s groups and thousands of additional women on both sides of the border. Focus group and interview findings most prominently indicate that these trainings have given them a new awareness of trauma issues affecting them, a “language” for discussing these issues, and fora for progressing toward understanding and forgiveness with traditional adversaries. Culturally, they also have been able to meet with local women where men would not be.  Advocacy: Women’s groups are empowering female stakeholders within their own and neighboring communities, confronting perpetrators of trauma against them in public fora who are taking those accused of rape and abuse to court, as in the case of Wajir Women for Peace (WWFP), and Mandera Women for Peace (MWFP), with the latter currently pursuing five rape cases that are pending in Mandera, and helping to secure a 14-year sentence for a convicted rapist. WWFP asserts that incidences of rape and conflict associated with such incidences have gone down, although this could not be independently confirmed, as many rape cases go unreported.  Mediators: There were several examples in which women’s groups have worked directly with members of the police and defense forces. The El Wak Somalia Women’s’ FGD, for example, noted that their ‘women’s’ committee’ was “now known by police” and able to establish an arrangement 26 For example, MWFP’s role in bringing rape cases to court has been supported by both PEACE II and United Nations Development Program (UNDP) trainings, the latter involving educating women on their legal rights and training them as paralegals. Page 40 of 54 with them in which those accidentally arrested while crossing the border to utilize the peace￾dividend-supported school on the Kenya side were to be released without being charged. Additionally, the presence of these groups – and the longevity of several of them27 - demonstrates an ongoing and increasingly successful effort to shift attitudes among husbands and community elders28 toward acceptance and support of their participation in peacebuilding activities – not an insignificant development in this religiously conservative area. However, their work is still not without risk or cost – members of Belet Hawa Women for Peace, for example, have been tortured, injured and killed while attempting to resolve conflict in their town.29 There was also near-universal acknowledgement among women within the FGDs and KIIs of the ongoing fear among women regarding participation in such trainings and local opposition from elders and religious leaders (and from Al Shabaab in particular on the Somali side of the border). Youth Groups: As noted under the first conclusion above, nearly 1 in 5 survey respondents were aware of, or aware of their communities working with, youth-based conflict-resolution groups. PEACE II supported – or supported the establishment of –CPUs, predominantly youth-based grassroots groups working to reduce crime at the local level by building relationships and sharing information with the police. The Mandera CPU played a leading role in this effort, both in working with local authorizes – e.g., the Mandera District Security Intelligence Committee – to deter criminal activity in the Mandera area, and in establishing and supporting other CPUs in Garissa, Liboi, and Hulugho over the last 2-3 years of the PEACE II program. CPUs have become an important component of the collaborative peace network PEACE II has sought to build along the border, but that significance has been based more on making youth stakeholders in the security of their communities and having police authorities validate that role than on actually lowering crime. There are only isolated examples of the latter30, and aside from the Mandera CPU, a more mature organization with deeper linkages with local authorities, the other border CPUs are still in the nascent stages of their development. Additionally, PEACE II staff acknowledged that the CPUs have had to contend with ongoing distrust from specific community members, who sometimes view them as “spy networks,” and a lack of ongoing formal coordination with either the Administrative or Provincial Policing Units, according to police officials interviewed by the team. Religious Leaders: Groups composed of religious leaders have the potential to exert a major influence on peacebuilding activities, and in specific cases within the PEACE II context, they have done so. Given their central role as teachers, advisors, and mediators of disputes within their communities, they have a natural role within the peace system. PEACE II support helped them to leverage this role outside of the mosque by mediating as religious but neutral arbiters in fighting between clans. The Mandera Mediation Council, for example, was formed out of the 2009 fighting between the Garre and Murulle clans, and the mediation they provided that facilitated resolution of the conflict; the PEACE II training they have received since – in mediation, arbitration, negotiation and trauma healing – has enabled them to continue and enhance their mediation role, and to intervene early and rapidly, in successive conflicts up to and including the November 2012 fighting in Mandera that ensued while the evaluation team was in country (this mediation, including for the most recent fighting, included “rapid response” interventions in which meetings were convened with conflicting clans or parties and dialogue enabled, leading to incremental agreements or “next steps” declarations that have helped contribute to resolution of these crises or at least allowed for development of action plans to be implemented once the conflicts “cooled” sufficiently; however, no data was identified by 27 Wajir WFP & Mandera WFP were both started in 2000. 28 KII with Maryan Noor, Mandera Women for Peace (11/3/2012) 29 Belet Hawa Women’s FGD (11/2012) 30 The Liboi CPU worked with the police to lift a 6PM curfew in the town that kept its hospital closed at night and on weekends due to area insecurity. The curfew was lifted and the CPU now helps provide monitoring and security. Page 41 of 54 the team demonstrating a causal link between these interventions and an earlier end to fighting or fewer deaths/casualties). The Mandera Council has also partnered with youth groups to administer counseling for at-risk youth, as well as with women’s and other key sector units to reach these other key groups through trainings and sector￾focused fora.31 Additionally, it also has smoothed the way for other groups to participate in this system, with the Council supporting the inclusion of Mandera Women for Peace and other groups in training and peace negotiations. “The respect we are given as religious leaders means that we can act as an umbrella for these other groups,” explained the Council’s Coordinator.32 Cross-Border Working Groups: These working groups have been formed under PEACE II during the participatory process of developing peace dividend projects (explored in detail under the ‘Peace Dividend’ section of this report). Their relationship is built in incremental steps around the identification, selection, planning, implementation and management of these projects, and has resulted within a number of peace corridors in the formation of collaborative, pro-active, ongoing cross-border relationships that have continued well beyond the launch of their peace dividend projects. They were also found to have leveraged the training and cooperation built through development and implementation of these projects to enable additional cross-border cooperation and support for their respective communities. For example, the Diff Cross-Border Working Group has worked with the staff at the area’s new maternity ward to assist women who are about to deliver with the costs of childbirth.33 Inter-Network Support: There was limited success demonstrated in the findings among sector response units supporting development of other groups in the broader peace system. However, the ability of the religious leaders to help form or support other mediation councils – or for these groups to establish themselves – has been limited, although there were a few examples of this. The Mandera Mediation Council and Mandera Community Policing Unit, for example, traveled to Garissa to help intervene in a land-based dispute between the Abdalla and Abduwak sub-clans in 2010. As part of their intervention and outreach within the community to mediate disputes, they helped to mobilize and mentor local groups, including the Garissa Mediation Council. With the exception of this, however, the Mandera Council has thus far not been successful in creating a planned religious network spanning the Northeast Province. Additionally, the planned establishment of the Wajir Mediation Council was put on hold after funding was required for local school construction. Conclusion 3: Peacebuilding trainings could be more effective with individual job training-based components. The benefit of adding job- or vocational-based elements to trauma healing, reconciliation and other interventions would be as an incentive to validate and reinforce peacebuilding trainings by the group. One could call these peace dividend projects for the individual. Supporting Findings: Focus group findings, particularly youth FGDs, demonstrated a strong preference for job-based training to accompany or follow from their participation in trauma healing or social reconciliation trainings. “The [trauma-healing] trainings helped us,” said one El Wak Somalia youth FGD participant. “But [job] training and jobs would help us more.” Several also pointed out that pastoralists could particularly benefit from such training, as many had lost their traditional means of livelihoods, not to mention family members, through drought, famine, conflict, or a combination of these. Several PEACE II-supported groups have built economic components into their programming. Combining PEACE II training and an economic 31 KII with Mohamed Abdinoor, Mandera Mediation Council (11/7/2012) 32 Ibid. 33 KII with Abninoor Haji, Diff Cross-Border Working Group (11/7/2012) Page 42 of 54 component, for example, the Somali Supportive Group (SSG) has established income-generating activities for jobless and stranded youths. Specifically, they have set up training and work opportunities with hairdressing sites and salons. A separate economic training – basics of setting up a small business – was offered by El Wak Kenya Youth for Peace, while the El Wak peace dividend project, a youth center, was stocked with 4 computers and 5 sewing machines, although PEACE II support for that training ended when the grant money ran out.34 Conclusion 4: While majority of partners’ work will continue after PEACE II in at least limited ways, long-term sustainability of some groups is questionable without additional institutional capacity building. The skills in which community representatives have been trained under PEACE II have helped to create a grassroots cadre of proficient practitioners in specific corridors able to conduct peacebuilding awareness and trauma healing exercises, but less mature groups will require additional building of their institutional capacity and mentoring if these groups and their work are to be sustained. Supporting Findings: Institutionally, the team found during its interviews with the program’s implementing partners that the majority stated they could continue their work without the program’s support, albeit at a less comprehensive level in terms of the number of interventions they could conduct and the geographic scope of those activities. All were actively seeking, had recently obtained, or possessed pre-PEACE II funding through other donors, NGOs, the private sector, the GOK, or some combination of these. Generally, more established, mature groups were to be found in and around Mandera and Wajir. This was based in no small part on PEACE II having been built on PEACE I and other earlier interventions in these areas, which the SOW for this evaluation notes resulted in a more advanced “community peace infrastructure” there. The groups created from these earlier efforts have evolved operationally, and frequently include diversified sources of funding, office space, and paid staff. The Belet Hawa Women for Peace group, for example, receives non-PEACE II support from the UN, OXFAM, and local NGOs. “PEACE II took our existing work and made it easier, but it wasn’t at the core of what we’re doing,” said Lullu Mohamed of the woman’s group, explaining that they were able to conduct additional conflict workshops through the program’s support.35 The Mandera-based District Livestock Marketing Council and Mandera Mediation Council both have received support from the town’s business community (the Council, which has also received support from the Council of Elders, noted that local businesses turned it down the first time they were approached regarding donations; they did start donating, however, after their businesses were burned or robbed). While PEACE II’s capacity-building efforts among local organizations have provided or augmented the technical skills to become viable participants in networking and peacebuilding activities, a majority of interviewed partners, particularly more recently-formed ones and those with support from fewer – or just one – source, were critical of the program for having given them the technical abilities but not the operational skills – e.g., fund-raising, proposal writing – to diversify their sources of support. “If we don’t have a means of sustaining the empowerment under PEACE II after the program ends, it’s going to disappear,” representatives of Garissa Women for Peace and Ijara Women for Peace told evaluators. Several also referenced support services they had previously been able to provide but were no longer able to since they had stopped receiving PEACE II support, such as the ability to provide paid transportation to bring participants to training sites, or to transport themselves to such sites.36 34 KII with Abdinoor Hussein, El Wak Kenya Youth for Peace (10/31/2012) 35 Other PEACE II partners, especially women’s groups, echoed this sentiment, noting that the means to transport these practitioners to other crisis or conflict-prone areas, including across the border, had been correlated directly with PEACE II resources. 36 KIIs with Mandera Women for Peace (11/3/2012), Garissa Women For Peace (11/10/2012) Page 43 of 54 Conclusion 5: The PEACE II “small grants” approach has provided uniform support to a diversified set of grassroots organizations representing key demographic segments of border-area communities, but has also proven problematic for these same groups in expanding or simply maintaining their current level of work. PEACE II’s ‘small-group’ strategy keeps any one of its sector response units from becoming better funded or resourced than another, but also keeps groups from realizing greater potential, or possibly in some cases from surviving. Supporting Findings: PEACE II adopted a strategy of support to its sector response units in which small grants are provided to these organizations to ensure that all stay involved and are supported at a uniform level. The deliberate avoidance of providing larger grants has been based on PEACE II implementers’ belief that this can lead – and in several instances has already led – to the perception or reality of favoring one particular sector or clan over the others, stirring up clan antagonism, and the “commercialization” of peace work.37 They cite CEPAR/Wajir Peace University as an example of a regionally-based organization that was given a broader mandate and corresponding resources to help provide training and education services throughout the Northeast Province, but did not work as a PEACE II intervention. They found after some time had elapsed that its stakeholders were too narrowly focused on the interests of Wajir to the exclusion of Mandera, Garissa and other areas along the border, and that these other areas were soon demanding the establishment of universities in their districts as well. The small-grants approach does increase available resources to enable support of additional groups, and does help to temper any perceptions, or the actuality, of more resources being directed to a particular area or clan. However, this approach also has meant intermittent funding for these groups – e.g., a five-month grant, then a six-month wait until they win the next one – as well as ongoing staff turnover, with people with PEACE II￾supported training now possessing marketable skills leaving for other jobs, in part because network groups can’t afford to match higher salary offers.38 Group representatives also note that this approach has meant that while they have the resources to provide an initial round of trainings to individuals, they have been unable to follow up with these same sets of individuals to reinforce and augment what has been taught already, or even to track their post-training progress.39 Conclusion 6: Sector-based conflict response activities have reinforced – and been reinforced by – activities under its Trauma Healing/Social Reconciliation and Peace Dividend theories of change. Sector response units conflict response actions have intersected and been supported by other PEACE II activities, demonstrating a complementary relationship between the program’s theories of change, particularly between Local Collaborative Peace Systems/Sector Response and Trauma Healing/Social Reconciliation. Supporting Findings: The sector response units have participated in trauma-healing and social￾reconciliation trainings and are now able to conduct similar trainings themselves for others, or at the very least, understand the concepts and “language” involved in such activities. Additionally, peace dividend projects gave a number of such groups, literally, a base of operations for conducting networking and training activities; several of them have closely tied themselves to peace dividends projects that align with their interests and/or have been able to serve as a means for them to better carry out their work, thus improving their ability to reach their targeted audiences. For example, the El Wak Youth for Peace consistently uses the town’s Youth Resource Center – a peace dividend project – as a meeting space for trauma, empowerment and other activities. The Hulugho Cross-Border Working Group has similarly utilized its peace-dividend 37 KII with PEACE II Chief of Party (11/2/2012) 38 KIIs with Mandera Women for Peace (11/3/2012), Garissa Women For Peace (11/10/2012) 39 Ibid. Page 44 of 54 project, a community center, as a source of income by renting it out to groups conducting seminars and workshops, but this was a DfID-funded – not a USAID PEACE II-funded – project. Conclusion 7: Validity of Local Collaborative Peace System and Sector Response Theory of Change is unclear at present. Evaluators found inconclusive evidence to validate this ToC, at least at this point in the long-term implementation of peacebuilding efforts along the border. Supporting Findings: The LCPS/SR Theory of Change asserts that by increasing capacity of local organizations to respond to conflict, a critical mass of peace actors is formed to proactively deal with conflict when it emerges. The team was unable to corroborate this ToC due, to a significant degree, to the lack of a baseline, targets, and indicator information to describe what would constitute “critical mass” both for individual sector groups seeking to meet this criteria, and for the network as a whole in terms of how proactively it respond to conflict.” Additionally, evaluators would suggest clarifying “increasing capacity” to include increases in institutional capacity, which PEACE II programming could have assisted more, in addition to technical and leadership capacity, which PEACE II helped increase, in some cases significantly, across a range of sectors and locations. One area in which it could be argued that this ToC demonstrated greater validity is Mandera, where a local network of groups was able to marshal diverse resources, create fora for rapid engagement of relevant parties to the conflict, and address conflict issues as proactively as they can be addressed within the setting for the conflict and sometimes when the fighting is still taking place. Again, however, it should be noted that given other and previous sources of support, only part of the validation of this theory within Mandera can be attributed to PEACE II. III. Recommendations 1. The PEACE II ‘small-grants’ approach should be reconsidered under a future PEACE program. The team would suggest allowing larger grants to be provided to partner organizations that have proven their effectiveness thus far and have presented practical plans for expansion of their work. Such groups would have to meet minimal operational and programmatic criteria to demonstrate that they not only possess sufficient potential for sustainability after PEACE II ends, but are also proving effective, even on a small scale, in their interventions with targeted communities. Grants could be launched as smaller-scale pilot initiatives in the form of, for example, highly-targeted training to build a group’s operational or fund-raising capacity, so that the group’s progress could be tracked and informed decisions on continuing funding made. Directing additional resources to specific groups may engender jealousies from other clans and ethnic groups, but if such groups aren’t given sufficient resources to allow them at least the potential to survive and continue conducting interventions once PEACE II ends, then the whole point of helping to establish or strengthen such groups in the first place is moot. Such groups may also be able to secure funding from non-USAID sources, which may generate the same jealousies PEACE II was seeking to avoid anyway; in such a scenario, a successor program to PEACE II would also lose the ability to guide such groups as part of its overall ‘peace network’ strategy. It is logical to want to have all demographic groups and geographic areas represented within the network, but not providing (additional) funds to organizations that could potentially have a greater positive impact than others essentially penalizes these groups and limits the potential of the entire network. 2. A livelihoods component should be considered for a future PEACE program. The consensus view among FGD and KII participants was that their trauma-healing, reconciliation and other training and mentoring efforts would resonate more strongly with targeted audiences, particularly at-risk youth, if Page 45 of 54 complemented by a work-related component. There is evidence within USAID’s own literature40 that such tandem efforts yield results under the appropriate, post-conflict conditions, and the successor program to PEACE II should consider its inclusion, building on activities such as the ones noted above being implemented by the Somali Support Group and El Wak Youth for Peace. Such efforts would be designed to strengthening the resiliency and “traction” of associated peace-building activities, and could be done on minimal budgets when paired with existing infrastructure supported under a future PEACE program – e.g., the use of the El Wak youth center, which was stocked with small numbers of computers and sewing machines – or through a similar arrangement with a non-USAID program funded by another donor or the Kenyan government. These small-scale livelihood activities would also be directed toward particularly high-risk groups, such as youth and women. 3. Within a future PEACE program, implementers should create knowledge-sharing services and networks to leverage other sources of support for sector response groups. There have been piecemeal efforts – and at least in the case of the Mandera Mediation Council, successful ones – to harness support from non-USG sources. The MMC, for example, leveraged the support of the business community in the form of KSH 50,000 and free rent (through the beginning of 2013) on their office space in Mandera. 40 See “Livelihoods and Conflict Toolkit,” USAID Office of Conflict Management and Mitigation (http://transition.usaid.gov/our_work/cross￾cutting_programs/conflict/publications/docs/CMM_Livelihoods_and_Conflict_Dec_2005.pdf) Page 46 of 54 STRATEGIC CONCLUSIONS AND RECOMMENDATIONS To augment the findings and conclusions in the previous sections, below are selected strategic conclusions and recommendations relating to the USAID/EA and PEACE II operating environment that can further help to inform future USAID programming in the region. These higher-level conclusions and recommendations evolved through the course of this evaluation, and in particular during focused meetings with USAID/EA staff and PEACE II beneficiaries, as it appeared that there were very important extant, environmental and strategic issues that not only affected the PEACE II program and its impacts, but could affect the design of a successor program. The higher-level conclusions also align well to the evaluation questions relating to lessons learned and impact. More specifically they augment the evaluation questions and their responses with conclusions and recommendations that are forward-thinking, holistic and address the “wicked problems” associated with conflict mitigation in complex environments. Finally, the conclusions evolved primarily from interviews with the KII participants, and in particular from the very frank discussions with partners, government staffs, and beneficiaries, crossed with the evaluation survey and FGD data. They represent focal areas of current and future interest and concern. It should be noted that while the authors do suggest that the below themes are critical to better and more comprehensively understand the contextual environment so as to shape future programs, there is no assumption that USAID/EA programming had intended to or planned to address these macro-level issues through the course of the PEACE II program. Also, there is no suggestion to USAID/EA to parse already limited resources to address these larger issues, which are likely out of operational focus. That said, the below Conclusions and Recommendations are meant to provide USAID with strong contextual, socio-political, thematic, and cultural areas of focus, which may benefit programming in the near future. Conclusion 1: Oil and gas exploration in North-West Kenya will impact the Kenya-Uganda border area Supporting Findings: Kenya and its regional neighbors have become hot-spots for oil and gas exploration. 41 Canada's Africa Oil and Tullow, its partner British exploration firm, have large and potentially lucrative oil and natural gas assets in East Africa, including those in areas in close proximity to the Kenya-Uganda border area, such as Ngamia and Twiga. In October 2012, Africa Oil and Tullow announced discovery of fresh oil deposits in its Twiga South-1 well, 30 km west of Ngamia.42 The commercial viability of many of the finds has yet to be ascertained, but these new developments increase hopes for Kenya to become a petroleum producer and exporter once the oil is found to be commercially viable. There is concern, however, that the rapid nature in which the deposits were found, combined with allegations of government collusion on land sales, suggests that this new discovery is a double-edged sword. This said, in order for this project to be sustainable and regionally representative, the Kenyan government will need to ensure full transparency in how the extraction will be done and how the revenues will be utilized. “I know Kenya has discovered oil, but the government should not at any time let it be the major driving sector of the economy. That is what clever countries have done,” according to Professor Banji Oyeyinka, United Nations Habitat Director in Nairobi.43 There are several triggers of conflict in the region. First, there is an overall lack of adequate infrastructure in the region. Indeed, the region is one of Kenya’s most neglected districts. Second, whenever there is a famine, chances are high that this region will be fully affected. According to Juliet Torome: 41 Standard, More oil deposits found on Turkana’s Ngamia 1 well, July 5, 2012. 42 Africa Oil Corporation, 2012 Third Quarter Report, December, 2012. http://www.africaoilcorp.com/s/Operations_Update.asp 43 Capital FM, Tullow Strikes More Oil in Turkana, October 31, 2012, http://www.capitalfm.co.ke/business/2012/10/17434/ Page 47 of 54 Gado, a renowned cartoonist for one of Kenya’s leading newspapers, summed it up best, depicting a jubilant Kenyan President Kibaki leading a pack of bureaucrats and dogs in suits to Turkana to announce to the people, “Rejoice! We have discovered oil!” A Turkana woman asks him, “And when will you discover water?” 44 Third, according to Torome, in addition to famine, the region’s people have endured decades of raids by cattle rustlers from neighboring communities, and the discovery of oil presents Kenya with a rare opportunity to end many of the communities’ senses of marginalization. That said, a participatory dialogue on how best the oil exploration and extraction processes will evolve in these communities will need to start quickly, with the health and livelihoods of the pastoralist communities remaining paramount.45 Recommendation: 1. A stronger outreach to and coordination with the GOK is recommended. Any successor program to PEACE II will need to leverage USAID’s relationship with the GOK to ensure that there is consistency in approaches in mitigating the social and economic impacts of oil exploration in the Karamoja, and in mitigating resultant conflicts between groups and communities that already possess grievances with other groups or communities. This will include using consistent approaches such as using PLA and other participatory approaches to assess potential impact. USAID and GOK should ensure that communities are mentored to understand and participate in this new sector, to include developing livelihoods programs and training programs for youth in the oil and gas or services sectors. Conclusion 2: There is a persistent insecurity related to pastoralism Supporting Findings: Increasingly severe and always unpredictable droughts have forced many pastoralist communities in the North Eastern Province to migrate their herds more frequently. In the absence of sustainable alternative livelihoods programs or other government-sponsored economic and social diversification programs, pastoralist communities in Kenya, Somalia and Uganda have little recourse. Regions have suffered from economic neglect, not just in terms of infrastructure such as roads and water, but also in terms of core basics of human security such as protection. The absence of rule of law and of police in these regions also partly explains the prevalence of small arms, and of sporadic, but often deadly, clashes between communities. In the absence of viable alternative livelihoods, some individuals and communities have turned to banditry and armed robbery, making armed police escorts mandatory in many roads in northwest and western Kenya. According to IRIN/UNOCHA, another trigger of insecurity in regions such as Turkana is the so-called “proliferation of political boundaries”, i.e., what used to be a single district a few months ago, is now six. While the Kenyan government states that the new district structure under devolution will bring additional national government outreach and services to locals, critics argue that such boundaries boost conflict by instilling even more entrenchment, rivalry, ownership and incursion among communities that previously had regarded pastureland and water points as shared resources.46 44 Juliet Torome, Oil and Isolation, Project Syndicate (www.project-syndicate.org) http://www.koreatimes.co.kr/www/news/opinon/2012/11/160_111251.html 45 Ibid. 46 IRIN, In-Depth Another Kenya - The humanitarian cost of under-development KENYA The dangers of pastoralism Kenya Conflict Food Security Governance Refugees, 2012. Page 48 of 54 Recommendations: 1. Continued mentorship on peacebuilding in the communities is essential, but is only one component of effective, and sustainable, cross-border stabilization. This is, of course, at the very core of the PEACE II projects and activities -- to mentor, guide and above all facilitate cross￾border and cross-community interactions, healing, social reconciliation and overall collaboration between communities in conflict, and as this evaluation has revealed these triggers of insecurity are becoming increasingly deleterious to pastoralist communities. That said, there are clear indications from the KIIs and the FGDs that the community individuals themselves see community stabilization, and peace as a process within a complex system with peace dividends, trauma healing and social reconciliation, capacity-building and cross-border peace networking as key pre-requisites, but not exclusive pre-requisites, to peace. Alongside these are human security, food security, economic security, protection of individual and human rights, protection of persons and land, access to essential services, and good governance. Although beyond the mandate of PEACE II, and potentially of any successor to PEACE II, these requirements should be seen as integral to future peacebuilding programs in such communities. 2. Increased marginalization of small arms is recommended. Given the proliferation of small arms in the region, and an endemic, pre-existing inter-community rivalry in both border areas, the primary sources for conflict already exist and are entrenched. In both the Kenya-Somalia and Kenya-Uganda border areas, there is an overwhelming presence of small arms, which when mixed with the emotions of internecine rivalry and decades-long grievances, can lead to a perpetual state of conflict. Although mentioned widely from participants in the KIIs and FGDs, there is no mention in PACT PEACE II literature of the role that small arms plays in the stability of the border area communities and their potential impact on the overall effectiveness of cross-border interactions, trauma healing, and peacebuilding efforts. Conclusion 3: The Lamu Port-Southern Sudan-Ethiopia Transport (LAPSSET) Corridor will impact the border areas Supporting Findings: The $23 billion undertaking will connect Kenya's coastal Lamu region to South Sudan and Ethiopia with oil pipelines, railways, and super highways and to date is one of the biggest infrastructure projects in Africa. The LAPSSET Corridor route will have a significant impact on the communities through which it passes, including communities in the North Eastern Province. The route passes, from south to north, from Lamu port, through Ijara and then through Garissa northwest onward to South Sudan.47 According to a Kenyan government feasibility study, the economic benefit of the Corridor has been clearly identified by the Kenyan government, and includes enhancing Kenya’s position as a gateway and transport hub to the East African sub-region and the Great Lakes region; establishing a reliable sea access for the north and eastern parts of Kenya; the establishment of reliable sea access to Ethiopia and South Sudan; to stimulate economic activity in the north and eastern parts of Kenya; to facilitate trade regional economic integration and interconnectivity between countries in the region; and to improve the economic livelihoods of some 15 million people in Kenya. Proponents of the project also suggest that it will facilitate much enhanced trade and investment for Kenya with South Sudan and Ethiopia, boosting employment and, according to President Kibaki, increasing the economic prospects for some 167 million people. That said, there are concerns that the LAPSSET Corridor communities will face unprecedented social and economic impacts. The “discontent 47 Republic of Kenya, Ministry of Transport, 2 nd Transport and Economic Corridor – LAPSSET, October 8, 2011. http://kenyaembassy.com/pdfs/diaspora/Ministry%20of%20Transport￾2nd%20Transport%20and%20Economic%20Corridor-LAPSSET%20by%20P.S%20Dr%20Cyrus%20Njiru.pdf Page 49 of 54 stirred up by the project demonstrates once again how the imperatives of development and democracy can sometimes clash in transitional societies” such as those in Kenya.48 According to a risk analysis study by the International Livestock Research Institute (ILRI), “ninety percent of the LAPSSET corridor goes through arid and semi-arid lands, which are important areas for livestock production, tourism, biodiversity conservation, and cultural heritage. However, the [government-sponsored] feasibility study doesn’t give the full impact of the LAPSSET corridor on livestock and wildlife mobility or on water and pasture competition despite the fact that these issues already trigger deadly clashes between various communities.”49 Recommendations: 1. USAID should consider incorporating conflict assessment, needs analyses, and impact assessment methods to support the design of any successor to PEACE II. Successor programs to PEACE II will be faced with needing to understand and mitigate the possible negative social, economic, environmental impacts of the LAPSSET Corridor on border communities in the North Eastern Province, including social upheaval, mass economic migrations, economic disparity and rivalry, competition for limited natural resources, and the impacts on already distressed and violent towns such as Garissa and Ijara, as they continue to become key regional centers for the border communities. A conflict assessment such as one guided by CAF 2.0 or ICAF is recommended as an initial step in the implementation of a successor program to PEACE II. 2. Successor programs to PEACE II will need to better emphasize the linkage between peace (and stability) and livelihoods programs. In addition to peace dividends, cross-border interaction, social reconciliation and trauma healing components as key precursors to peace and stability, successor programs to PEACE II should continue to ensure that the identification and implementation of such activities in the region are designed to mitigate the potential economic and social impacts of LAPSSET. In addition to this, as the majority of KII respondents, and in particular women and youth respondents, stated once “triggered”, i.e., mentored and empowered through PEACE II activities to better respond to crises and more actively work with cross-border communities, there was a desire to become more engaged and enfranchised members of the community by leveraging groups such as the Women for Peace, Youth for Peace or the Somali Supportive Group. Such groups have been imperative for providing training programs; however, there is a requirement to build upon this success with livelihoods programs designed to train women and youth in the sectors associated with the LAPSSET project. Conclusion 4: The 2013 general election and devolution will likely impact the border areas Supporting Findings: It is hardly revelatory for the evaluation team to state that the 2013 general election will have an impact on all communities in Kenya, including those along the Kenya-Somalia and Kenya￾Uganda borders. However, given the complex inter-clan and inter-community relationships within the border areas, the potential political, social, and economic impacts may even be more acute. This year’s general election is scheduled for March 4, 2013 and will elect a new president and a host of leaders, including 48 Jeremy Flatteau, “A megaproject transforms Kenya, but not everyone is thrilled,” Foreign Policy, Web Edition, May 28, 2012. http://transitions.foreignpolicy.com/posts/2012/05/25/a_megaproject_transforms_kenya_but_not_everyone_is_thrill ed 49 LAPSSET needs better realignment to Vision 2030, Web Edition, May 8, 2012. http://peoplelivestockenvironment.wordpress.com/2012/05/08/lapsset-needs-better-realignment-to-vision-2030/ Page 50 of 54 governors and senators, for 47 counties established by the August 2010 constitution. The election will be the first under the transformative new constitution that Kenyans approved with a 67% majority in a referendum and was disseminated by President Kibaki. The constitution also devolves power and accountability to the country’s counties in a devolution program that is one of the most ambitious in the world. In addition, according to the World Bank, the political future in Kenya is fraught with tensions over pending prosecution by the International Criminal Court (ICC) of four key suspects of the 2007-2008 post-election violence, including politicians in the electoral race. Youth unemployment and regional imbalances, in particular those in the North Eastern Province, also pose political risk and insecurity especially in areas with high poverty levels such as along the Uganda and Somalia borders. Prior to and during the course of the evaluation, alliances forged by Kenya's main presidential contenders for elections in March were lining up a repeat of a largely ethnic-based contest for political power that exploded into bloodshed in the 2007 vote. According to Mzalendo Kibunjia, who heads a Kenyan national agency formed to reconcile tribes or clans after the 2007 violence, the source of any potential post-election violence is based squarely on tribal or clan structures: What do you expect? Our politics are about ethnicity. In Africa, democracy is about ethnic arithmetic not ideology.50 In addition to the potential impacts that the election will have on the border communities, any future USAID programs along the borders must also be cognizant of the attendant impacts of devolution. Devolution is a response to the enormous centralization of state power at the center and in the presidency, accentuated by the attrition of local government. For many people, the main contact with government has been with Provincial and District Commissioners and Chiefs, ultimately responsible to the President. On a more political level, this centralization of power, generally exercised by a small cadre of people around the President, was accused of marginalizing communities such as those along the Somalia border, and in regions that were perceived to be opposed to the regime. Devolution was designed to provide the districts with a much greater voice in and responsibility for the future of their communities, and would include greater accountability, transparency and lesser centralization from Nairobi. However, there are serious anxieties about devolution. Paradoxically, some people, as evidenced in the FGDs with elected representatives in Garissa County, are worried about too much powers being handed over to counties, while others that too little power is guaranteed.51 There are also fears of enhanced and increased inter-community clashes, especially in the wake of recent economic migrations within the North Eastern Province. Groups who have migrated into a county in recent times are fearful even of eviction, which is certainly relevant to the sizeable Somali migrant population in Garissa town52 Recommendation: 1. USAID should consider leveraging existing conflict assessment and early warning mechanisms with GOK to prepare for potential impacts of the elections. In the case of those communities within Garissa County, a recommendation would be to capitalize on the work already underway by the National Steering Committee and PACT, and especially under the conflict mapping 50 Mail & Guardian, Kenya: Fears of Another Violent Election, December 26, 2012, http://mg.co.za/article/2012-12-26- kenya-fears-of-another-violent-election 51 World Bank, Making devolution a game changer:10 ways to help transition succeed, February, 2012. http://siteresources.worldbank.org/INTAFRICA/Resources/257994- 1335471959878/Making_Devolution_a_Game_Changer.pdf 52 Hon. Peter Kenneth, MGH, Assistant Minister of State for Planning, National Development and Vision 2030, Devolved Kenya and the Promise, July, 2012. www.planning.go.ke/index.php?option=com_docman Page 51 of 54 and CEWARN initiatives.53 Moreover, the authors recommend conflict mitigation work be maintained or even expanded in this region, focusing on issues such as youth at risk and alternative livelihood programs; the impact of refugees on existing communities; the role of elders, women and religious elders as mediators and conflict mitigators in the community; and on educating the communities themselves on what devolution means for them, and how best they can progress within the new post-election governance structures. The question in this case, and indeed, in all conclusions in this section is whether there is sufficient community training, capacity and even motivation to prepare for and mitigate election violence. Conclusion 5: The shifting nature of conflict on the Kenya-Somalia border and the “Little Mogadishu” of Garissa has affected stability along the Kenya-Somalia border area Supporting Findings: Based on several KIIs and FGDs with community and religious leaders in the Kenya￾Somalia border area, the very nature of the conflict in this corridor has changed from 2007 to 2012. Whereas there were, and still remains today, sources of inter-community conflict based on pastoralism and economic migration, competition for natural resources, inter-clan rivalry and historical grievances, communities are now attempting to address what they term a “faceless” source of conflict in terrorism, both from within their communities and from outside. In more than 70% of KIIs and FGDs conducted during the PEACE II evaluation, and in particular those held with elders, community leaders, women, and youth, people in border communities have the perception that they now live in fear of what some called the “faceless” enemy responsible for the rise in grenade attacks, bombings, attacks, and improvised explosive devices, often without notice or attribution. In these cases, the PEACE II (or even traditional) mechanisms for conflict mitigation and reconciliation are insufficient to deal with this evolving threat. For example, what emerged from a FGD with religious and elders in Garissa, and a follow-on KII with Sheikh Ali Gure, a prominent religious leader in Garissa county, religious leaders have always been the peacebuilders in the communities. He has since the early 1990s. However, one of the prerequisites of peacebuilding and of conflict mitigation is the ability to communicate with one’s counterpart. In cases of terrorism and the “faceless” attacks on communities, this prerequisite is absent. In the words of Sheikh Ali Gure, elders cannot play a role in peacebuilding if they cannot “see” those with whom they are meant to build peace. To be clear, inter-clan conflict remains problematic in the corridors; during the fieldwork there was also a major inter-clan conflict that erupted in Mandera, causing the deaths of several people. During the fieldwork for this evaluation there were several acts of terrorism or Al Shabaab activity reported in the Somali corridor, and specifically in Garissa town. For example, in November alone an AP post in Garissa was bombed, killing one policeman, and there were two separate incidents of AP and KDF troops being killed by unnamed assailants. There were also Al Shabaab incursions into Belet Hawa, El Wak (K), and Damassa. PEACE II has been successful at mentoring and guiding in the development and implementation of inter-clan mediation activities, and in particular through its mentorship and peacebuilding training with the Mediation Councils, Peace Committees, and CPUs, it has made significant strides with communities in preventing, mitigating and lessening the violent effects of inter-clan and inter-community conflict. But, as the nature of conflict evolves to include Al Shabaab activities along the Kenya-Somalia border and further into the interior, the PEACE II peacebuilding and mediation trainings may prove insufficient. According to the FGD with elected county councilors in Garissa, the conflict in the region is now centered on Al Shabaab, and its retreat from Kismayo into Kenya, and the rise in refugees along the border as a result of the ongoing military operations in southern 53 National Steering Committee, Towards Peaceful Elections and Beyond: County Consultations and Dialogue in Kenya, June￾November 2011. Page 52 of 54 Somalia. This source of conflict originates following the AMISOM, Ras Kamboni Brigade and TFG incursions into Kismayo in 2012, and the KDF inclusions into southern Somalia in 2011 and 2012, all of which resulted in Al Shabaab migrating its operations into Kenya, and specifically into areas where there is a Somali majority. Garissa town itself has become a conflict-prone area, gaining the moniker “little Mogadishu” from organizations such as Garissa Women for Peace. Recommendations: 1. A cross-border community needs assessment and conflict assessment should be considered prior to future programming in the Somali Corridor. As the findings and conclusions above demonstrate, there is strong evidence of positive impacts on communities based on the presence of PEACE II activities. However, an overwhelming majority of KII and FGD participants responded that, in addition to cross-border conflict, they are equally impacted by acts of terrorism or by armed clashes between government and militia forces in or near their communities. This will undoubtedly have an effect on individuals’ personal perceptions of trauma, reconciliation, and peace in their communities. It will also have an as yet undetermined impact on the lasting effectiveness of PEACE II peace dividends and capacity-building activities. To better prepare for follow-on programming, the evaluation team suggests that tools such as CAF 2.0 or ICAF be used to help guide a comprehensive conflict assessment to inform programming of a successor to PEACE II. In addition to this, it is suggested that USAID consider conducting a robust and participatory needs assessment to better ascertain communities’ priorities relative to peace and security within the context of the new threat of terrorism. 2. USAID should consider expanding the peace dividend ToC to account for the additional conflict “trigger” of terrorism on community peace and stability. The ToC, and subsequent implementation of it, should consider including the proposed and anticipated implications of terrorism on local communities, and the relative efficacy of peace dividends in the new, more complex environment of the border communities. Conclusion 6: There is evidence of conflict “fatigue” and a shift in worldviews in PEACE II communities Supporting Findings: There is evidence of conflict “fatigue” and a shift in worldviews in some of the PEACE II communities. It should be noted that there is no evidence of a correlation between this feeling, and the degree and level to which PEACE II activities in these communities was present. The KIIs and FGDs in both Lokiriama and Nakiloro in the Kenya-Uganda (Karamoja) border area, and in Hulugho, Mandera, and Garissa County in the Kenya-Somalia border area revealed that individuals and groups (e.g., youth, women and men) have strong feelings of despondence, indifference, weariness, resignation and apathy when asked about their perceptions of violence and conflict. As mentioned above, there is evidence that the majority of respondents in the HH and Caretaker surveys feel that the security situation is better now than one year ago, and then five years ago, but when these data are analyzed alongside the qualitative responses, there is evidence that is not contradictory to the survey, but rather conditional. For example, in both communities in the Kenya-Uganda (Karamoja) border area, KII and FGD respondents stated that incidents of inter-tribe conflict, banditry and cattle-rustling were bad, to be sure, but that they were also a normal, underlying, if even benign, part of daily life and that they had resigned themselves to this. Conflict was enmeshed in daily life, and therefore conflict mitigation would be challenging. In Liboi and Mandera, there is also evidence of resignation. For example, the men’s FGD in Mandera described its perceptions of safety Page 53 of 54 and security in their community as benign (the HH survey showed a very low perception of security), stating the “security is stable, except for the inter-clan conflicts.”54 Similarly, there is also evidence in the data that worldviews have changed since PEACE II was launched. Specifically, the data implies that in some communities there has been a considerable increase in peoples’ perceptions of the future, their community, and of the ability to conduct daily activities such as herding, going to market or moving from one place to another, and that the majority of these communities are in Somalia. Conversely, there has been a considerable decrease in peoples’ perceptions of the future, their community, and of the ability to conduct daily activities in communities in Kenya. For one example, in the Mandera-Belet Hawa corridor there is a much more positive worldview evidenced in Belet Hawa than in Mandera. This would be no more than an interesting fact if it were not for the case that the data collected in Belet Hawa was done so during a major Al Shabaab incursion and subsequent KDF/AMISOM expulsions in the town, which would imply the evaluation team receiving negative rather than positive responses. Based on the HH survey data and FGDs, in the Kenya-Uganda (Karamoja) border area there is a resounding difference in peoples’ worldviews between respondents in Lokiriama and Nakiloro, with those in Lokiriama having a more pessimistic and resigned view of conflict than their counterparts in Nakiloro. Recommendations: 1. A cross-border community needs assessment and/or conflict assessment should be considered prior to future programming in the Somali Corridor. As above, it is critical to note that there have been significant contextual changes since the start of PEACE II in 2007. As a result, USAID should consider analyzing the needs of the Kenya-Somalia and Kenya-Uganda border areas for future programming as being relevant to the contextual issues for 2012, and thus to prepare and implement a conflict assessment assuming that there has been a shift in peoples’ perceptions from 2007 to 2012. 2. Reconsider worldviews as contextual foundations for the design of successor programs. There are likely several contextual, environmental and historical reasons for the shift in worldviews, and it is recommended that these be explored fully during a comprehensive conflict assessment. For example, one of the reasons to have come out from KIIs and FGDs with elders, youth and women is that there is a general perception that there has been a relative shift in safety and security between 2007 and 2012 from good to bad in Kenya, and from bad to good on the Somalia side of the border. This is particularly evident in Garissa town, Mandera, and Hulugho, where HH survey, FGD and KII respondents expressed disappointment in their worldviews, whereas in Belet Hawa and Dobley there were expressions of optimism. Conclusion 7: The resiliency and adaptability of local communities is key to future stability Supporting Findings: Resilience and flexible adaptability was not a focus of PEACE II, but perhaps should have been. A key measure of progress in stabilization or peacebuilding programs is evidence of resilience in the presence of ongoing or periodic violent shocks to communities experiencing these shocks. Several sites in the PEACE II Kenya-Somalia and Kenya-Uganda (Karamoja) border areas have demonstrated resiliency as an adaptive mechanism to remain active in an environment of perpetual conflict. This is the case in particular for Belet Hawa, Liboi, El Wak (S), Mandera, Garissa, and Hulugho where evidence from KIIs and FGDs has shown that despite ongoing violent conflict, youth, women’s and elders groups have been able to respond to and rebound from shocks to their communities, adapting in part due to PEACE II mentoring and the 54 Note that the data collection for Mandera was during the inter-clan conflict in November, 2012. Page 54 of 54 implementation of PEACE II peace dividends, PLA, CCP, social reconciliation and cross-border interaction initiatives, to remain thriving, living, evolving and organic communities. It would have been beneficial too, to have had indicators in place to ease in establishing correlations between PEACE II initiatives and these demonstrations of resilience, if only to provide some additional accountability to the USG in an era when there is increased interest in proving that USAID programs have an impact in enhancing community resiliency in challenging environments.55 Recommendations: 1. Build the concept of resilience into PEACE II principles and its successor programs. As per guidance in the new USAID policy on resilience, Building Resilience to Recurrent Crisis: USAID Policy and Program Guidance, it is recommended that future PEACE programs in the region frame their objectives, IRs and sub-IRs (or at least a subset thereof) in terms of having an impact on the resiliency of the intervention communities to withstand or adapt to stressors and shocks, and then also to establish M&E systems capable of measuring this impact. 55 USAID, Building Resilience to Recurrent Crisis: USAID Policy and Program Guidance, December 2012.