0 NIWETU Endline Evaluation Final Report August 2020 1 KENYA NIWAJIBU WETU “NIWETU” NIWETU Endline Evaluation Final Report August 2020 1. TABLE OF CONTENTS Findings and Outcomes ....................................................................................................... 9 Lessons Learned and Recommendations ......................................................................... 11 2. Research Overview ................................................................................................................ 13 2.1 Limitations ............................................................................................................... 13 3. Findings and Lessons learned ................................................................................................ 14 3.1 OUTCOMES: (Goal) Capabilities to identify and respond to VE threats .......... 14 3.1.1 Effectiveness of systems-based CVE approaches ....................................................................... 14 3.1.2 Complexity-aware planning enhanced adaptive management ................................................. 15 3.1.3 Community-Government trust enhanced through dialogue ................................................... 16 3.2 LESSONS LEARNED AND RECOMMENDATIONS: (Goal) Capabilities to identify and respond to VE threats ................................................................................... 17 3.2.1 ‘Whole-of-Society’ approach key to CVE longevity .................................................................. 17 3.2.2 Pivot in VEO recruitment tactics requires CVE adaptation .................................................... 18 3.2.3 County-level CVE devolution, resourcing and clarity is patchy ............................................. 18 3.2.4 Evolving USG/GOK strategic priorities impact sustainability.................................................. 19 3.3 OUTCOMES: (Obj1) Community mobilization to address violent extremism 20 3.3.1 Knowledge and skills ......................................................................................................................... 20 3.3.2 Enhanced understanding of VE and CVE strategies ................................................................... 20 3.3.3 Openness to CVE dialogue expanded .......................................................................................... 21 3.3.4 Ability to identify and counter VE narratives enhanced ........................................................... 22 3.3.5 Inter-religious/ethnic tolerance boosted ...................................................................................... 23 3.3.6 Women perceived as critical to CVE response ......................................................................... 23 3.3.7 Networks ............................................................................................................................................. 24 3.3.8 Community-led strategies and initiatives ..................................................................................... 27 3.4 LESSONS LEARNED AND RECOMMENDATIONS: (Obj1) Community mobilization to address violent extremism ..................................................................... 29 3.4.1 Communities should opt for vigilance over suspicion .............................................................. 29 3.4.2 CVE activists must be protected .................................................................................................... 29 3.4.3 Greater input needed from ‘grass-roots’ CSOs and commercial actors ............................. 30 3.4.4 Restrictive gender norms remain a challenge ............................................................................. 30 2 3.5 OUTCOMES: (Obj 2) Government responsiveness to VE .................................. 31 3.5.1 Government CVE capacity .............................................................................................................. 31 3.6 LESSONS LEARNED: (Obj2) Government responsiveness to VE ..................... 33 3.6.1 Managing expectations in an evolving context ............................................................................ 34 3.6.2 CEFs effective in ‘hotspots’ with political buy-in ........................................................................ 34 3.7 Cross-cutting outcomes and lessons learned ........................................................ 35 3.7.1 Gender prioritized across portfolio .............................................................................................. 35 3.7.2 Legacy partner CVE capacity significantly strengthened ........................................................... 35 3.7.3 The journey to self-reliance requires sustained donor backing ............................................. 36 4. Conclusions and recommendations ...................................................................................... 37 5. ANNEX I – Methodology ...................................................................................................... 39 Research Approach ............................................................................................................ 39 Analytical Framework ........................................................................................................... 41 Sampling ............................................................................................................................... 45 Scoring Approach ................................................................................................................. 45 6. Annex II – Garissa County Report: Garissa County Endline Overview ............................. 51 7. Key Findings ........................................................................................................................... 52 7.1 Section 1 – The VE Context ................................................................................... 52 7.1.1 Summary of Findings ......................................................................................................................... 52 7.1.2 Causes of Insecurity .......................................................................................................................... 52 7.1.3 Disappearances ................................................................................................................................... 53 7.1.4 VE and Recruitment .......................................................................................................................... 53 7.1.5 Gang and Criminal Activity .............................................................................................................. 54 7.1.6 Safety and Security ............................................................................................................................. 55 7.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced ... 55 7.3 Section III – Skills and Knowledge ......................................................................... 57 7.3.1 Summary of Findings ......................................................................................................................... 57 7.3.2 Skills and Knowledge Index Scores ............................................................................................... 57 7.3.3 Participation in Training ................................................................................................................... 58 7.3.4 Improving Training Programs .......................................................................................................... 60 7.3.5 Sharing Lessons Learned .................................................................................................................. 61 7.4 Section IV – Community Networks ....................................................................... 63 7.4.1 Summary of Findings ......................................................................................................................... 63 7.4.2 Community Networks Index Score .............................................................................................. 63 7.4.3 Collaboration between Actors ....................................................................................................... 63 7.4.4 Conflict between Organizations ..................................................................................................... 65 7.5 Section V – Community-Driven Strategies & Initiatives ..................................... 66 7.5.1 Community-Driven Strategies and Initiatives Index Score ...................................................... 66 7.5.2 Participation in NIWETU-Sponsored Activities ......................................................................... 66 7.5.3 Community-Driven Strategies ........................................................................................................ 66 7.5.4 Effectiveness of CVE Programs ....................................................................................................... 67 7.5.5 Level of Engagement .......................................................................................................................... 69 7.5.6 Summary of Findings ......................................................................................................................... 69 7.6 Section VI – Objective 2: Improved Government Capacity ................................ 70 7.7 Section VII – Government Responsiveness to CVE ............................................. 72 7.7.1 Government Responsiveness Index Score .................................................................................. 72 3 7.7.2 Government Understanding of VE ................................................................................................ 72 7.7.3 Willingness to Engage in Programming ......................................................................................... 73 7.7.4 Engagement with Government Counterparts ............................................................................ 74 7.7.5 Summary of Findings ......................................................................................................................... 75 8. Annex iii – Isiolo County Report: Isiolo County Endline Overview ................................... 76 9. Key Findings ........................................................................................................................... 77 9.1 Section 1 – The VE Context ................................................................................... 77 9.1.1 Summary of Findings ......................................................................................................................... 77 9.1.2 Causes of Insecurity .......................................................................................................................... 77 9.1.3 Disappearances ................................................................................................................................... 78 9.1.4 VE and Recruitment .......................................................................................................................... 78 9.1.5 Gang and Criminal Activity .............................................................................................................. 79 9.1.6 Safety and Security ............................................................................................................................. 80 9.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced ... 81 9.3 Section III – Skills and Knowledge ......................................................................... 82 9.3.1 Summary of Findings ......................................................................................................................... 82 9.3.2 Skills and Knowledge Index Scores ............................................................................................... 82 9.3.3 Participation in Training ................................................................................................................... 83 9.3.4 Improving Training Programs .......................................................................................................... 85 9.3.5 Sharing Lessons Learned .................................................................................................................. 87 9.4 Section IV – Community Networks ....................................................................... 88 9.4.1 Summary of Findings ......................................................................................................................... 88 9.4.2 Community Networks Index Score .............................................................................................. 88 9.4.3 Collaboration Between Actors ...................................................................................................... 89 9.5 Section V – Community-Driven Strategies and Initiatives .................................. 93 9.5.1 Summary of Findings ......................................................................................................................... 93 9.5.2 Community-Driven Strategies and Initiatives Index Score ...................................................... 93 9.5.3 Participation in NIWETU-Sponsored Activities ......................................................................... 94 9.5.4 Community-Driven Strategies ........................................................................................................ 94 9.5.5 Effectiveness of CVE Programs ....................................................................................................... 94 9.5.6 Level of Engagement .......................................................................................................................... 96 9.6 Section VI – Objective 2: Improved Government Capacity ................................ 97 9.7 ......................................................................................................................................... 97 9.8 Section VII – Government Responsiveness to CVE ............................................. 98 9.8.1 Summary of Findings ......................................................................................................................... 98 9.8.2 Government Responsiveness Index Score .................................................................................. 98 9.8.3 Government Understanding of VE ................................................................................................ 99 9.8.4 Willingness to Engage in Programming ......................................................................................... 99 9.8.5 Engagement with Government Counterparts ......................................................................... 100 10. Annex IV – Mandera County Report: Mandera County Endline Overview ................ 103 11. Key Findings ..................................................................................................................... 104 11.1 Section 1 – The VE Context ................................................................................. 104 11.1.1 Summary of Findings .................................................................................................................. 104 11.1.2 Causes of Insecurity................................................................................................................... 104 11.1.3 Disappearances ........................................................................................................................... 106 11.1.4 VE and Recruitment ................................................................................................................... 107 11.1.5 Gang and Criminal Activity ...................................................................................................... 108 11.1.6 Safety and Security ..................................................................................................................... 109 4 11.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced . 110 11.3 Section III – Skills and Knowledge ....................................................................... 111 11.3.1 Summary of Findings .................................................................................................................. 111 11.3.2 Skills and Knowledge Index Scores ........................................................................................ 112 11.3.3 Participation in Training ............................................................................................................ 112 11.3.4 Improving Training Programs .................................................................................................. 114 11.3.5 Sharing Lessons Learned .......................................................................................................... 115 11.4 Section IV – Community Networks ..................................................................... 115 11.4.1 Summary of Findings .................................................................................................................. 116 11.4.2 Community Networks Index Score ...................................................................................... 116 11.4.3 Collaboration between Actors ............................................................................................... 116 11.4.4 Conflict Between Organizations ............................................................................................. 118 11.5 Section V – Community-Driven Strategies and Initiatives ................................ 120 11.5.1 Summary of Findings .................................................................................................................. 120 11.5.2 Community-Driven Strategies and Initiatives Index Score............................................... 120 11.5.3 Participation in NIWETU-Sponsored Activities ................................................................. 121 11.5.4 Community-Driven Strategies................................................................................................. 121 11.5.5 Effectiveness of CVE Programs ............................................................................................... 122 11.5.6 Level of Engagement .................................................................................................................. 123 11.6 Section VI – Objective 2: Improved Government Capacity .............................. 124 11.7 Section VII – Government Responsiveness to CVE ........................................... 125 11.7.1 Summary of Findings .................................................................................................................. 125 11.7.2 Government Responsiveness Index Score ........................................................................... 125 11.7.3 Government Understanding .................................................................................................... 126 11.7.4 Willingness to Engage in Programing ..................................................................................... 127 11.7.5 Engagement with Government Counterparts ..................................................................... 129 12. ................................................................................................................................................ 130 13. Annex V – Nairobi County Report: Nairobi County Endline Overview ....................... 131 14. Key Findings ..................................................................................................................... 132 14.1 Section 1 – The VE Context ................................................................................. 132 14.1.1 Summary of Findings .................................................................................................................. 132 14.1.2 Causes of Insecurity................................................................................................................... 132 14.1.3 Disappearances ........................................................................................................................... 133 14.1.4 VE and Recruitment ................................................................................................................... 133 14.1.5 Gang and Criminal Activity ...................................................................................................... 134 14.1.6 Safety and Security ..................................................................................................................... 136 14.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced . 136 14.3 Section III – Skills and Knowledge ....................................................................... 138 14.3.1 Summary of Findings .................................................................................................................. 138 14.3.2 Skills and Knowledge Index Scores ........................................................................................ 138 14.3.3 Participation in Training ............................................................................................................ 138 14.3.4 Improving Training Programs .................................................................................................. 141 14.3.5 Sharing Lessons Learned .......................................................................................................... 143 14.4 Section IV – Community Networks ..................................................................... 144 14.4.1 Summary of Findings .................................................................................................................. 144 14.4.2 Community Networks Index Score ...................................................................................... 144 14.4.3 Collaboration Between Actors ............................................................................................... 144 14.4.4 Conflict Between Organizations ............................................................................................. 147 5 14.5 Section V – Community-Driven Strategies and Initiatives ................................ 149 14.5.1 Community-Driven Strategies and Initiatives Index Score............................................... 149 14.5.2 Participation in NIWETU-Sponsored Activities ................................................................. 149 14.5.3 Community-Driven Strategies................................................................................................. 149 14.5.4 Effectiveness of CVE Programs ............................................................................................... 150 14.5.5 Level of Engagement .................................................................................................................. 152 14.5.6 Summary of Findings .................................................................................................................. 152 14.6 Section VI – Objective 2: Improved Government Capacity .............................. 154 14.7 Section VII - Government Responsiveness to CVE ............................................ 155 14.7.1 Government Responsiveness Index Score ........................................................................... 155 14.7.2 Government Understanding of VE ......................................................................................... 155 14.7.3 Willingness to Engage in Programming ................................................................................. 156 14.7.4 Engagement with Government Counterparts ..................................................................... 157 14.7.5 Summary of Findings .................................................................................................................. 158 15. Annex VI – Wajir County Report: Wajir County Endline Overview ............................. 159 16. Key Findings ..................................................................................................................... 160 16.1 Section 1 – The VE Context ................................................................................. 160 16.1.1 Summary of Findings .................................................................................................................. 160 16.1.2 Causes of Insecurity................................................................................................................... 160 16.1.3 Disappearances ........................................................................................................................... 161 16.1.4 VE and Recruitment ................................................................................................................... 161 16.1.5 Gang and Criminal Activity ...................................................................................................... 163 16.1.6 Safety and Security ..................................................................................................................... 163 16.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced . 165 16.3 Section III – Skills and Knowledge ....................................................................... 167 16.3.1 Summary of Findings .................................................................................................................. 167 16.3.2 Skills and Knowledge Index Scores ........................................................................................ 167 16.3.3 Participation in Training ............................................................................................................ 167 16.3.4 Improving Training Programs .................................................................................................. 169 16.3.5 Sharing Lessons Learned .......................................................................................................... 171 16.4 Section IV – Community Networks ..................................................................... 172 16.4.1 Summary of Findings .................................................................................................................. 172 16.4.2 Community Networks Index Score ...................................................................................... 172 16.4.3 Collaboration Between Actors ............................................................................................... 172 16.4.4 Conflict Between Organizations ............................................................................................. 174 16.5 Section V – Community-Driven Strategies and Initiatives ................................ 175 16.5.1 Summary of Findings .................................................................................................................. 175 16.5.2 Community-Driven Strategies and Initiatives Index Score............................................... 175 16.5.3 Participation in NIWETU-Sponsored activities................................................................... 176 16.5.4 Community-Driven Strategies................................................................................................. 176 16.5.5 Effectiveness of CVE Programs ............................................................................................... 176 16.5.6 Level of Engagement .................................................................................................................. 177 16.6 Section VI – Objective 2: Improved Government Capacity .............................. 179 16.7 Section VII – Government Responsiveness to CVE ........................................... 180 16.7.1 Summary of Findings .................................................................................................................. 180 16.7.2 Government Responsiveness Index Score ........................................................................... 180 16.7.3 Government Understanding of VE ......................................................................................... 181 16.7.4 Willingness to Engage in Programming ................................................................................. 181 16.7.5 Engagement with Government Counterparts ..................................................................... 183 6 17. Annex VII – Case Study: Champions For Change .......................................................... 185 17.1 Overview of Champions for Change .................................................................... 185 17.2 Who are the Champions? ..................................................................................... 186 17.3 What are the main achievements of the C4C initiative?................................... 186 17.4 Lessons Learned .................................................................................................... 190 18. Annex VIII – Case Study: County Action Plans .............................................................. 192 18.1 Engaging Government .......................................................................................... 192 1.1.1 NIWETU-NCTC collaboration in support of the Rapid-CAP development 193 18.2 Engaging CSOs ...................................................................................................... 195 1.1.2 CEFs Coordinate Response to Increase in AS Attacks ........................................... 195 18.3 Engaging Influencers ............................................................................................. 196 18.4 Engaging Communities ......................................................................................... 197 18.5 Creating Linkages Between Actors ..................................................................... 198 18.6 Lessons Learned .................................................................................................... 198 19. Annex IX – case Study: Kenya School of Government .................................................. 200 19.1 Establishment of the Institute .............................................................................. 200 19.2 Key Achievements ................................................................................................. 202 19.3 Long-Term Vision .................................................................................................. 204 20. Annex X – Case Study: NIWETU Engagement in Mandera County ............................ 206 20.1 Mandera in 2017 .................................................................................................... 206 20.2 Working Through a Systems-Based Approach .................................................. 206 20.2.1 Civil Society Organizations ...................................................................................................... 207 20.2.2 Government ................................................................................................................................ 208 20.2.3 Influencers .................................................................................................................................... 209 20.2.4 Communities ............................................................................................................................... 210 20.3 Mandera in 2020 .................................................................................................... 211 20.4 Lessons Learned .................................................................................................... 212 21. Annex XI – VE Index Scores ............................................................................................ 214 22. Annex XII - Q3-2019 - Quarterly Research Report: Comparative Analysis of Community and Government Responses in the Aftermath of the Westgate and 14 Riverside Attacks ......................................................................................................................................... 217 22.1 Key Findings ........................................................................................................... 217 22.1.1 A Brief History of Terrorism in Kenya ................................................................................. 217 22.1.2 National Government & Security Sector Response .......................................................... 218 22.1.3 Westgate attack response and the aftermath ..................................................................... 218 22.1.4 14 Riverside attack response and the aftermath ................................................................ 220 22.1.5 Summary analysis of GoK’s response .................................................................................... 222 22.2 Community Response ........................................................................................... 222 22.2.1 Community mobilization following the Westgate attack ................................................. 223 22.2.2 Community mobilization following the 14 Riverside attack ............................................ 224 22.2.3 Comparative analysis of community mobilization .............................................................. 225 7 23. Annex XIII- Q2-2020 - Quarterly Research Report: Understanding Gender-related outcomes of NIWETU grants, and a review of NIWETU’s Gender Plan ................................. 227 23.1 Key Findings ........................................................................................................... 227 23.1.1 Gender-Related Outcomes ..................................................................................................... 227 23.1.2 Additional Findings ..................................................................................................................... 233 23.2 Recommendations from Component 1............................................................... 234 24. Annex IV- Q3-2018 - Quarterly Research Report: The County Action Plan Formation Process: Isiolo, Garissa and Mombasa ....................................................................................... 235 24.1 Overview of the Research ..................................................................................... 235 24.1.1 Research Context and Objectives................................................................................. 235 24.1.2 Areas of Inquiry ...................................................................................................................... 236 24.1.3 Research Methods ................................................................................................................. 237 25. Future CAPs: Ten Essential Ingredients .......................................................................... 238 26. Stakeholder Perspectives and Analysis ........................................................................... 241 26.1 NCTC ..................................................................................................................... 241 26.2 National Government in the Counties ................................................................ 242 26.3 Security Agencies .................................................................................................. 244 26.4 County Governments ............................................................................................ 244 26.5 Lead Partners ......................................................................................................... 246 26.6 Local CSOs ............................................................................................................. 248 26.7 Wider Communities ............................................................................................. 249 8 ACRONYMS ACC Assistant County Commissioner C/VE Countering Violent Extremism C4C NIWETU Champions for Change CAP County CVE Action Plans CAP Cycle Complexity-aware Planning Cycle CASSS Central Asia Support for Stable Societies CC County Commissioner CEF CVE Engagement Forum CIDP County Integrated Development Plan CSO Civil Society Organization DCC Deputy County Commissioner GOK Government of Kenya IPL Isiolo Peace Link J2SR Journey to Self-Reliance KACPEN Kamukunji Community Peace Network KMS Knowledge Management System KSG Kenyan School of Government MCA Members of the County Assembly NCTC National Counter Terrorism Center NIWETU USAID Kenya Niwajibu Wetu NPR National Police Reservist NSCVE National Strategy for Countering Violent Extremism OCS Officer in Charge of Station POTA Prevention of Terrorism Act PREG Partnership for Resilience and Economic Growth RCAP Rapid County Action Plan SCORES Sustainability of Community on Radicalization and Empowerment Solutions SMI Security Management Institute SUPKEM Supreme Council of Kenyan Muslims USAID United States Agency for International Development VEO Violent Extremist Organization WPDA Wajir Peace and Development Agency WWFP Wajir Women for Peace 9 EXECUTIVE SUMMARY This endline assessment report is the culmination of a longitudinal study of USAID Kenya NiWajibu Wetu (NIWETU) interventions over the span of four years from 2016 to 2020, aimed at improving CVE capabilities to identify and respond to VE threats. The study was conducted by Wasafiri, NIWETU’s research and learning partner, as part of its broader portfolio of research, analysis, and monitoring work across the activity. The methodology utilized tracking of qualitative and quantitative indices, using a mixed methods approach at key intervals, including baseline, midline, and endline assessments. The assessments focused primarily on the results of activity interventions in NIWETU’s target counties of Garissa, Isiolo, Mandera, Nairobi, and Wajir. Data gathered included targeted county-level micro￾surveys, tracking of sentinel indicators, and qualitative interviews with key stakeholders. Additionally, the study incorporated key findings from concurrent research activities, including thematic and geographic case studies, outcome harvesting assessments, and quarterly research reports. In light of the findings and lessons learned, it is evident that NIWETU had significant impacts across its primary objectives and corresponding results areas. Its contribution to CVE in Kenya manifested in numerous ways, particularly with regard to building resilience to VE among its key government, community, and civil society stakeholders. Within an often-challenging context, the team piloted a variety of innovative and novel approaches that established a model for the community of practice, and for future interventions of this nature. A summary of these observations is presented below. Findings and Outcomes The findings are categorized according to the activity’s two primary objectives: 1) Enhancing community mobilization to address VE, and 2) Improving government responsiveness to VE. The observations are then structured according to the intended results areas within each objective.  Effectiveness of systems-based approaches to CVE – NIWETU pioneered a unique approach to CVE that leveraged a multi-pronged strategy to build cooperation and synergies to tackle VE threats. The approach benefited from its holistic partnerships, creating a culture of awareness and action that extended beyond the activity. Key achievements to which this process contributed included institutional reform, driven by a facilitated and inclusive County Action Plan (CAP) development strategy, which brought together a diverse range of national and county-level CVE actors around a unified goal. Further, the strategy spurred a surge in collective CVE dialogue and the emergence of multi-stakeholder CVE networks. In certain instances, there was evidence of the activity’s contribution to improved responses to VE attacks, such as following the 14 Riverside attack in 2019 and the increase in VE incidents in the North East in early 2020.  Complexity-aware planning enhanced adaptive management – The Complexity Aware Planning (CAP) Cycle facilitated a constant process of learning and adjustment. The approach built on USAID’s portfolio of complexity-aware tools, including most significant change, outcome harvesting, sentinel indicators, and contribution analysis. The CAP cycle contributed to enhanced targeting of VE ‘hotspots’ at the outset of the activity, based on context mapping of key actors, challenges, and potential resiliencies. Ongoing action-research activities also facilitated the formulation of a deliberate, activity-level gender strategy and streamlined improvements to the rapid rollout of CAPs across the country. Periodic learning reports enabled real-time course correction during implementation and incremental improvements to delivery modalities over the life of the activity.  Enhanced understanding of CVE through dialogue – The vast majority of participants of NIWETU-supported workshops and training said they gained significant benefits through enhanced CVE knowledge and skills. Key participants included government representatives, civil society, community members, and influencers. A critical outcome was the breaking down of perceived 10 barriers and stigmas associated with VE, enabling more open and constructive dialogue between stakeholders, and providing psychosocial support to communities affected by trauma. Increased CVE awareness and skills had cascading impacts and contributed to the empowerment of peacebuilding ‘ambassadors’ and Champions for Change (C4C). Furthermore, the interventions led to an enhanced ability for previously vulnerable communities to identify and counter VE narratives, mitigating many VEO ‘pull’ factors.  Emergence of sustainable CVE networks – A variety of formal and informal networks emerged as a result of NIWETU activities, including youth groups, women’s groups, boda boda collectives, and C4Cs. Additionally, the activity sparked the resurgence of CVE ‘clubs’ in schools, ‘self-help’ groups for men and boys, and small-business, entrepreneurial collectives. There was evidence that many of these groups developed organically and were sustainable beyond the life of the activity. Stakeholders indicated these groups were instrumental in promoting more open CVE conversations, cohering communities around collective action, and advocating for social justice within VE hotspots.  Formulation of targeted CVE action plans – NIWETU facilitated the creation of concrete CVE strategies in several forms, including the CAPs and C4C action plans, among others. CAPs rolled out in target counties led, in most cases, to strengthened linkages between national and county-level CVE actors, in addition to the formation of County CVE Engagement Forums (CEFs). While resourcing of CAPs remains a challenge, they established a clear roadmap for further interventions in key areas. C4C action plans represented a highly localized approach to individual action, that led to the creation of a number of CVE-focused community-based organizations (CBOs), in addition to a national C4C coalition with longer-term ambitions outside of their activity-related commitments.  Collective government responses to CVE improved – CVE knowledge and skills among county-level government actors improved significantly over the life of the activity. Resultantly, there was evidence that overall responsiveness of county officials to community concerns and national policy imperatives improved in some areas. The degree to which capacity was enhanced, however, was not consistent across all target counties and was heavily influenced by pre-existing relationships between communities and local authorities and the individual personalities present in county offices. Examples of collective action included the creation of the Mandera CVE Champions, modelled on NIWETU’s C4Cs and initiatives by the office of the Isiolo County Commissioner to open up additional dialogues with local youth.  Provisions for longer-term capacity building and research established – Part of NIWETU’s targeted assistance to national government actors included interagency capacity building and CVE training. The activity facilitated the creation of the Security Management Institute (SMI) within the Kenya School of Government (KSG) as a cross-cutting, sustainable means of providing for longer-term capacity building initiatives. The SMI developed an executive professional and diploma training course, a training and facilitation manual, a five-year strategic plan (2020- 2025), and conducted a CVE training of trainers (ToT) for 25 government staff. The institute also developed a range of case studies and policy briefs, based on ongoing research, including a review of policy gaps in responding to VE in Kenya, and a critical examination of gaps in the implementation of existing policies. 11 Lessons Learned and Recommendations The lessons learned over the course of implementation related to process considerations, such as the methods and modalities of activity delivery, in addition to contextual shifts, such as the evolving nature of CVE in Kenya. The recommendations highlight how best practice and some of NIWETU’s more innovative systems can be built on for future activities of this nature.  ‘Whole-of-Society’ approach key to CVE longevity – NIWETU demonstrated the importance of nesting CVE within national and sub-national governance portfolios. Members of NIWETU’s Steering Committee emphasized the importance of integrating CVE with wider development initiatives to harmonize activities, destigmatize the sector and provide for greater sustainability. This approach is particularly important in light of the often-fragile relationships and divergent priorities that exist within communities. In counties where political will to address CVE was relatively low, NIWETU boosted the role of civil society to fill these gaps, and vice versa where the impetus for action shifted. Future CVE initiatives should also place greater importance on conducting rolling political economy analyses (PEAs) in order to better understand underlying power dynamics and the incentives of key individuals and institutions to work for or against policies that improve the CVE landscape. These PEAs would allow a CVE program to strategize more effectively to mainstream CVE into relevant government and policy structures that will have varying degrees of political appetite for addressing the structural drivers of VE.  Evolving VEO tactics require further adaptation – While a notable reduction in recruitment activities was observed in NIWETU target counties, there was evidence that VEOs had adapted their tactics in terms of geographic focus, messaging modalities, and targeting of particular groups. Stakeholders suspected that VEOs were increasingly targeting in more peripheral areas where they had not encountered as much resistance, and where knowledge of VE threats was comparatively less. There was also evidence that VEOs were increasingly targeting younger individuals due to their perceived vulnerability to VE messaging. Countering these narratives requires more creative engagement through online and offline platforms to build resilience among vulnerable youth.  County-level CVE resourcing crucial to effectiveness – Uncertainty regarding responsibilities for resourcing and funding the action plans remained one of the primary obstacles to their effective implementation. This uncertainty stems in part from the lack of devolved security responsibilities; as security remains a national-level mandate, county governments are still not properly incentivized to resource CVE initiatives in meaningful ways. And yet while some county governments pledged significant amounts for CAPs, the legislation did not exist to allow reallocation of funds from the five-year County Integrated Development Plan (CIDP). County officials noted that Members of the County Assembly (MCAs) are crucial to include in any fiscal integration and/or reallocation process. Furthermore, a knowledge management system is crucial to enabling county governments to input regular updates on the status of CAP activities, track expenditures, and prevent duplication of efforts.  USG/GOK policy clarity critical for journey to CVE self-reliance – USAID support to CVE in Kenya faces uncertainty, and national- and community-level stakeholders perceive a lack of clarity and vision on the part of USAID regarding the nature of its support to the government to undertake such activities. Given the relatively recent emergence of the CVE sector in Kenya, it requires sustained support and cooperation in order to achieve longevity in the current context. To maintain the positive momentum driven by NIWETU and its partners, clear and coordinated leadership is required from national and international counterparts. Such efforts should build on 12 the CAPs and collaborative engagements facilitated by NCTC nearer the end of the activity to foster more effective programming. 13 2. RESEARCH OVERVIEW This evaluation was carried out in NIWETU’s core counties of intervention (Garissa, Isiolo, Mandera, Nairobi1, and Wajir) and utilized a mixed-methods approach including stakeholder quantitative surveys, household quantitative surveys, and key informant interviews. The evaluation also gathered data against the sentinel indicators identified during the midline evaluation to evaluate and understand the wider context at the objective level. The findings from this evaluation are driven by qualitative and quantitative data, pre-existing reports written under NIWETU’s CAP Cycle Framework, and case studies that were developed for this evaluation. Details on the research methodology and sampling are offered in Annex I. 2.1 Limitations  This research was broken out into two phases, with the second phase of research scheduled to begin in mid-March 2020, just after the Covid-19 outbreak hit Kenya. A decision was taken (in advance of the government banning travel into and out of Nairobi) to carry out phase two of fieldwork remotely via telephone interviews. As a result, the depth of information gathered from phase two of interviews was somewhat limited, as attention spans for phone interviews are often shorter than in-person conversations. The availability of certain county government and security sector officials was also limited, as priorities were in responding to the Covid-19 pandemic.  This evaluation was not designed to assess impact or ascribe attribution. As such, this evaluation is not able to ascribe attribution specifically to NIWETU, but instead identifies ways in which NIWETU contributed to changes at the county and national levels.  The sampling approach for the stakeholder surveys was purposive and designed to reach certain stakeholder groups with some degree of knowledge or relation to CVE and VE at the county level. As a result, findings from the stakeholder survey (presented in depth in the county reports) cannot be extrapolated to the wider population. Respondents were, however, identified using the same approach across the baseline, midline, and endline to enable longitudinal tracking of findings among a stratified sample.  Household survey data presented in the county reports comes from 40 interviews per county (and from two neighborhoods within the main city in each county), and therefore does not represent a statistically significant sample. As such, findings from the household survey data cannot be extrapolated even to the village level.  Research took place in the urban centers of each county. While key informants shared perspectives that touched on the wider context across the county, perspectives gathered from the household and stakeholder surveys are limited to the town centers and cannot be extrapolated to more rural areas. Findings that relate to the county context are highlighted accordingly. 1 This endline evaluation covers NIWETU’s work in Kangemi and Kibra. While NIWETU also works in Kamukunji sub￾county, a separate evaluation framework covers NIWETU’s work there. 14 3. FINDINGS AND LESSONS LEARNED This section, as noted in the methodology, highlights the primary observations gleaned over the course of activity implementation. The observations are drawn from highlights captured through Wasafiri’s research and evaluation activities and NIWETU’s internal reporting, in addition to a variety of proxy and secondary sources relevant to the broader context in which the activity operated. From these sources, the findings and lessons learned seek to build on the baseline and midline, while identifying the most important intended and unintended outcomes and impacts of NIWETU programming. It also puts forward recommendations for future activities of a similar nature. The sections are structured according to NIWETU’s updated logic model, leading with goal/impact-level observations, followed by a discussion of findings within each of the two primary objectives and corresponding results areas. There are, inevitably, overlapping elements of each intervention stream, which the report attempts to categorize according to the most relevant results area. Cross-cutting observations are explored at the end, culminating in a series of conclusions and recommendations. 3.1 OUTCOMES: (Goal) Capabilities to identify and respond to VE threats NIWETU’s Senior Management Team (SMT) revised its goal/impact statement in 2019 from ‘Reduce VE among at-risk individuals and communities’ to ‘Improved CVE capabilities to identify and respond to VE threats.’ The adjusted goal more accurately reflected the activity’s role in building CVE capacity among key government and community stakeholders in Kenya, while acknowledging the numerous external variables contributing to overall levels of VE in target areas. The subsequent findings, therefore, relate to the cumulative, longer-term effects of activity interventions in support of its two primary objectives, namely: 1) Enhancing community mobilization to address VE, and 2) Improving government responsiveness to VE. These interventions contributed to a number of higher-level impacts, including enhanced understanding of CVE frameworks, strengthened intragovernmental CVE linkages, improved community-government dialogue, and incremental institutional reform through the development of CAPs. These achievements are unpacked in greater detail below, followed by overarching lessons learned related to NIWETU’s goal. 3.1.1 Effectiveness of systems-based CVE approaches NIWETU pioneered a unique approach to CVE that leveraged a multi-pronged strategy to build cooperation and synergies to tackle VE threats. This strategy relied on sustained and proactive engagement with key stakeholders, including government counterparts, civil society, communities, and key influencers across its areas of focus. The systems-based approach benefited from its holistic partnerships, creating a culture of awareness and action that extended beyond the activity. Some evidence of this impact included:  Institutional strengthening driven by CAP process – CAP formulation was a relatively recent phenomenon that tested the coordination efforts of national and county-level actors. In several instances, it exposed cracks in the governmental architecture and suffered due to resourcing constraints. On the whole, however, it facilitated dialogue and engagement on CVE that led to constructive and inclusive responses. Enhanced collaboration and initiative in counties, such as Isiolo and Mandera, led to the emergence of well-coordinated and respected CEFs, and improved relations between county and national-level entities, such as the National Counter Terrorism Centre (NCTC), Anti-Terror Police Unit (ATPU), National Police Service (NPS), and County Commissioners.2 2 See Isiolo and Mandera County Endline Reports 15  Reduction in VE recruitment – Stakeholders perceived a decrease in rates of youth recruitment by VEOs, which was cross-verified by endline surveys. In several areas, government officials indicated that joint government-community action had been taken, as a result of increased awareness from NIWETU-supported activities and in collaboration with community leaders, either to remove or replace imams considered to be promoting radicalization among youth.3  Surge in collective CVE dialogue – Discussions of CVE expanded outside traditional forums and actors to include informal youth networks, women’s groups, and boda boda collectives, among other groups. Findings from the endline evaluation suggested that knowledge of concepts, terminology, and potential resiliencies had become part of the fabric of communities and government counterparts in target counties.  Emergence of multi-stakeholder networks – NIWETU’s broad approach to stakeholder engagement catalyzed the sustained linkages between diverse actors. CEF meetings led to the formation of downstream CSO partnerships; Champions for Change (C4C) networks facilitated cross-sectoral responses; and the Security Management Institute (SMI) strengthened interdepartmental relationships through capacity building initiatives.4  Improved responses to attacks – While such initiatives fall outside of NIWETU’s immediate remit, there was evidence that CVE activities were indirectly linked to increased speed and coordination of response to recent VE attacks. This was evident when comparing the response to the Westgate attack versus the 14 Riverside attacks and more recent volatility in the North East. Improved relationships between county-level leaders and national agencies were credited with reducing the number of indiscriminate arrests and sparking more coherent community interventions in ‘hotspots’ following the 14 Riverside attacks. Similarly, the rapid formation of response committees in Wajir and Garissa, following attacks in early 2020, precipitated better resourcing of National Police Reservists (NPRs) and teacher recruitment practices.5 The above findings demonstrate the necessity and effectiveness of holistic, systems-based approaches for CVE, in addition to broader areas to be prioritized for further engagement. Further examples of such outcomes are elaborated under subsequent sections as they pertain to specific objectives and results areas. 3.1.2 Complexity-aware planning enhanced adaptive management The Complexity Aware Planning (CAP) Cycle Framework was promoted as an approach to implementation, based on a constant process of learning and adjustment. The consequent evolutionary nature of the activity relied upon the incorporation of adaptive learning capacities within its decision￾making processes. The CAP Cycle’s approach to monitoring and research was designed to complement NIWETU’s performance monitoring systems. The approach built on USAID’s portfolio 3 See Kamukunji and Majengo Outcome Harvesting assessments. 4 See C4C and KSG Case Studies. 5 See ‘Comparative Analysis of Westgate vs 14 Riverside Attacks’ assessment and December 2019 Monthly Learning Report in Annex XII J2SR for CAPs NIWETU trained county staff, national government officers, and community members using its CVE facilitation guide to enhance their participation in the CAP process. The national and county governments committed to co-chairing the process and establishing CVE specific mandates across their departments. The activity also supported civil society actors and key community influencers, to advocate their elected representatives to introduce county-level legislation to increase funding for CVE activities prioritised through the CAP process. It is expected that the passage of these bills in 2020 will greatly improve CVE resourcing across target counties 16 of complexity-aware tools, including most significant change, outcome harvesting, sentinel indicators, and contribution analysis. There were a number of notable instances where the CAP Cycle and resultant learning contributed to course correction and a refinement of activity strategy. Some examples included:  Enhanced targeting of VE hotspots – Context mapping of key actors, challenges, and opportunities during the activity’s inception phase laid the groundwork for the identification of NIWETU target areas, and better tailored interventions in light of key findings. The mapping also provided a baseline of VE knowledge and skills within each location, to improve the relevance and utility of CVE workshops and trainings.  Development of explicit gender strategies – Periodic reviews of gender dynamics at the activity-level and in target communities ultimately paved the way for the development of a deliberate gender strategy that sought to prioritize and integrate gender-based design throughout interventions. The effectiveness of this process is elaborated further under lessons learned in subsequent sections.6  Improvements to CAP processes – Early research during the CAP formulation process uncovered a range of lessons and recommendations to inform the refinement of engagement strategies in target counties, in addition to the Rapid County Action Plans (RCAPS) extended to the rest of the country following the 14 Riverside Drive attacks. These recommendations were presented to NCTC during RCAP rollout, which included streamlining the pillar development process to ensure stakeholders had a clear understanding of the localized areas of focus.7 The CAP cycle included baseline, midline, and endline assessments to track programmatic progress, analyze pathways to observed outcomes, and provide formative feedback to activity stakeholders. Additionally, the CAP Cycle involved quarterly research reports, outcome harvesting sessions, county￾level assessments, and periodic case studies across a range of data sets and thematic areas prioritized iteratively by the SMT. 3.1.3 Community-Government trust enhanced through dialogue There were frequent examples across target counties where CVE interventions boosted levels of trust and engagement between stakeholders, particularly local residents and security actors. Improved mutual respect and confidence between police and communities was seen as crucial to the effectiveness of localized CVE strategies, involving open channels of communication and joint actions to build resilience. Observed changes in levels of trust and cooperation were attributed to NIWETU￾led efforts to create safe spaces for open dialogue and interaction between security actors, community leaders, and youth. Some key examples included:  Improved community-security relations in Wajir County – The Head of the Peace, Cohesion, and Integration Commission confirmed that engagement between community and security personnel had improved significantly. Youth participating in county-level dialogue sessions also noted that their relationship with the Deputy County Commissioner (DCC) was much better than it had been previously, enabling them to coordinate on security-related issues.  Enhanced government-community engagement in Isiolo – A local Champion for Change said the primary focus of the County Engagement Forums (CEF) had been building trust between local residents and police. Security officers involved in CAP activities, led by the County Commissioner’s office, trained 72 female security officers across different agencies, and participated in regular football matches with youth as part of NIWETU-supported activities. In 6 See ‘Understanding gender-related outcomes of NIWETU grants’ 2020 Quarterly Research Report in Annex XIII 7 See ‘CAP Formulation Process’ 2018 Quarterly Research Report in Annex IV 17 Garbatulla, a sub-county of Isiolo, the Chief explained that his community was more open to sharing security-related information with local authorities, with the knowledge they would be protected, regardless of the outcome.8  Barriers to effective coordination overcome in Majengo – Youth from the area explained that they had been reluctant to apply for registered identification cards because the cards were associated with systematic harassment and marginalization. However, due to diligent outreach efforts by their Chief and observed changes in attitudes among local authorities, youth reported that there had been a spike in registration for IDs. A community leader pointed to the proactive behaviors of the Officer in Charge of Station (OCS), in neighboring Shauri Moyo, whom they said had become more trusted by local residents because she did not accept bribes, was responsive to phone calls, and attended dialogue sessions organized by the Kamukunji Community Peace Network (KACPEN).9 The above examples demonstrate the importance of a systems-based approach to building shared CVE visions and commitments among a multiplicity of key stakeholders, to improve trust and cooperation at the community-level. These case studies further emphasize the critical role individual personalities played in driving forward constructive engagement. Such outcomes were, however, not observed uniformly across NIWETU target areas, and modest gains in community cohesion can be quite fragile and vulnerable to a variety of external factors. These variations and causal effects are discussed in more depth under lessons learned. 3.2 LESSONS LEARNED AND RECOMMENDATIONS: (Goal) Capabilities to identify and respond to VE threats 3.2.1 ‘Whole-of-Society’ approach key to CVE longevity The preceding sections highlight the results of NIWETU’s efforts to cohere relevant government and community stakeholders around collective implementation of the pillars of the National Strategy for Countering Violent Extremism (NSCVE) and CAPs. While these efforts drove significant progress toward key objectives, some partners felt that CVE strategies should be better harmonized and integrated with regional development agendas. Despite considerable gains in intergovernmental and civil society coordination, CVE in Kenya is often siloed as a separate sector, thus insulating it from broader coordination enjoyed by more traditional development interventions.  CVE must be nested within broader governance portfolio – NIWETU demonstrated the importance of nesting CVE within national and sub-national governance portfolios. In some instances, CEFs were perceived as being overly ‘securitized’ and exclusive of civil society, such as when the CEF was first launched in Isiolo County in 2018. However, in Mandera County, the local government was able to integrate a significant portion of its CVE strategy into its County Integrated Development Plan (CIDP), enabling greater resourcing of activities.  Integration of CVE mitigates stigmas – The Prevention of Terrorism Act (POTA), for instance, stipulates that organizations working on CVE must report their activities to NCTC before, during, and after implementation, and allow NCTC representatives to conduct random spot checks. Some CSOs said they halted their CVE activities from July 2019 onward, due to the perceived invasive nature of these requirements. Members of NIWETU’s Steering Committee therefore emphasized the importance of integrating CVE with wider development initiatives to 8 See ‘Key Actions for CAP Implementation’ 2020 Quarterly Research Report 9 See Majengo and Kamukunji Outcome Harvesting Reports 18 harmonize activities, destigmatize the sector, and provide for greater sustainability, in addition to continuing coordination with national government counterparts. 3.2.2 Pivot in VEO recruitment tactics requires CVE adaptation While stakeholders observed a notable reduction in recruitment activities in NIWETU target counties, there was evidence that VEOs had adapted their tactics, in terms of geographic focus, messaging modalities, and targeting of particular groups. These shifts were seen in NIWETU’s later research activities conducted closer to the end of the activity, indicating they occurred gradually over the course of implementation, which included:  Targeting more peripheral – During outcome harvesting in Kamukunji, a religious leader said that many in his CVE network suspected that VEOs were increasingly targeting other areas, such as Nyeri, where they had not encountered as much resistance and where knowledge of VE threats was comparatively less. In such cases, amplifying CVE messaging through media with great reach could fill gaps in peripheral areas.  VE messaging increasingly online – Stakeholders reported that recruitment tactics had shifted from largely in-person engagement to social media and messaging apps, such as WhatsApp and Telegram. As such, officials noted that VE narratives had become harder to isolate and track, providing VEOs with an additional layer of secrecy under which to conduct recruitment activities. VEOs are also increasingly targeting younger individuals, with respondents saying that children age eleven and younger are now being influenced, due to their perceived vulnerability VE messaging. Countering these narratives requires more creative engagement through online platforms, to build resilience among vulnerable youth. 3.2.3 County-level CVE devolution, resourcing and clarity is patchy The NSCVE delegates the bulk of the responsibility for funding and implementation of CAPs to county￾level governments. The 14 Riverside Drive attack, in January 2019, prompted President Uhuru Kenyatta to announce the rapid expansion of CAPs across all 47 counties. In response, NCTC requested rapid assistance from USAID and the British High Commission (BHC), resulting in NIWETU’s direct support to the process. NIWETU had to swiftly scale-up its CAP-support operations across many counties where VE was not broadly perceived as an issue at the time. The process, consequently, encountered several hurdles, including a lack of understanding among county officials of CAP requirements, uncertainty about whom was responsible for implementation, and how it would be funded.  More CVE responsibilities should be devolved to the counties – The POTA vests responsibility for CVE and NSCVE implementation, including formulation of CAPs, with NCTC. There are, however, different interpretations of how CVE functions should be divided between NCTC, at the national level, and county governments. NCTC reportedly believes that education and prevention are the responsibility of county governments, but this is not confirmed in writing or in legislation, which contributes to confusion, as security functions as a whole remain a national government function. Given the recently expanded mandate of NCTC, county-level officials, particularly those involved in the RCAPs, are often unaware of it. The lack of clarity with regard to roles, combined with the absence of legislated funding, leaves many county governments in a quandary, where they are not incentivized to be proactive to take ownership of CAP implementation. As NIWETU demonstrated, in such instances it is important to identify potential frictions and opportunities and work to mitigate tensions, while building consensus through an inclusive process.  County level pillar formulation required localization – Over the course of CAP expansion, they became increasing localized. Initial, coastal CAPs simply adopted the national pillars, partially to due to a lack of understanding of how they translated into county-level agendas. The NIWETU- 19 supported process complemented the national pillars with three to four county-level pillars, but there was some disagreement over prioritization. In Isiolo and Garissa counties, for instance, NCTC initially wanted to focus on the pillars in the national General CAP Guidelines, however county stakeholders saw them as less pertinent to local issues. Upon the advent of the RCAPs, NCTC began encouraging counties to simply prioritize two-to-three pillars, whether they be national-level or county-specific. Overall, the process resulted in more localized and streamlined CAPs, which generated valuable lessons learned for future processes of this nature.  Resourcing of CAPs critical for implementation – Uncertainty regarding responsibilities for resourcing and funding the action plans remained one of the primary obstacles to their effective implementation. While county governments pledged significant amounts for CAPs, the legislation did not exist to allow reallocation of funds from the five-year County Integrated Development Plan (CIDP). The CIDP in Mandera County reportedly earmarked funds specifically for CVE initiatives, which allowed the County Commissioner to allocate a portion to CAP implementation. County officials noted that MCAs are crucial to include in any fiscal integration and/or reallocation process. Future initiatives should consider securing national government commitments to fill county-level funding gaps prior to the necessary legislation being passed, as a prerequisite for support to CAP processes.  Centralized knowledge management key to effective monitoring – A KMS is crucial to enabling county governments to input regular updates on the status of CAP activities, track expenditures, and prevent duplication of efforts. NIWETU advocated to relevant actors for the creation and widespread adoption of a CAP-level KMS, and NCTC now has a platform in development. The KMS is reportedly being piloted in Nyeri and Isiolo, with plans to expand to other counties in the near future. 3.2.4 Evolving USG/GOK strategic priorities impact sustainability In recent years, USAID has adopted a global policy framework referred to as the ‘Journey to Self￾Reliance’ (J2SR), which aims to ‘end the need for foreign assistance by fostering self-reliance’. As part of this framework, USAID missions have been developing country-level J2SR ‘road-maps’ that provide an overview of how their strategy seeks to promote these principles within national contexts. Broad indicators were developed to track areas of national ‘commitment’ and ‘capacity’ to guide aid considerations. USAID-funded activities are, consequently, expected to align their goals with country road maps as part of a more aligned and coherent strategy.  USG prioritization important for regional CVE longevity – Funding uncertainty made it more difficult for NIWETU to complete the planned CVE activities with local partners in the final stages of the activity, and risks USAID’s reputation and regard as a serious partner on security issues with local Kenyan partners, including the Government of Kenya. There was also a lack of clarity on the part of USAID regarding longer-term CVE interventions in Kenya and the nature of its support to the government to undertake such activities.  Greater donor support needed to support transition – Isiolo’s CEF, for instance, served as a conduit for donor interest and funding which was also heavily reliant on funding from external donors. Respondents reported donor support to CEF meetings has ‘dried up’, meaning that since July 2019 to date, there have been just two meetings. This highlights the limited sustainability of CAP processes without donor funding. 20 3.3 OUTCOMES: (Obj1) Community mobilization to address violent extremism Objective 1 aimed to enhance community mobilization to address violent extremism in NIWETU target counties. Types of activities in support of this objective included sub-grants to key partners to conduct community-based CVE initiatives, the creation of formal and informal CVE networks, and strategic communications on CVE-related themes. The following sections highlight important outcomes observed under each results area, followed by objective-wide lessons learned. 3.3.1 Knowledge and skills Efforts to improve CVE knowledge and skills among key stakeholders represented a foundational activity component. The strategy was multi-pronged and involved partner-led knowledge and awareness workshops, in addition to skills training on dispute resolution, conflict mediation, and how to effectively manage CVE networks. This component was not intended to be standalone, rather it was woven throughout implementation streams fed into networking and community-led CVE strategies. 3.3.2 Enhanced understanding of VE and CVE strategies While the issues surrounding VE in affected communities in Kenya are not new, research activities at the outset of the activity found considerable obstacles preventing open discussion of VE in target areas. These obstacles included fear among residents of talking about local problems stemming from radicalization, due to perceptions they would face enhanced suspicion and scrutiny from VEOs and local authorities. Inter-religious tensions and fractures along ethnic lines further dissuaded citizens from engaging in meaningful dialogue about the underlying drivers of instability in their communities. The activity, consequently, was tasked with breaking down perceived barriers and creating safe spaces in which to facilitate discussions about CVE concepts. Building a foundation of CVE knowledge and skills among community and government stakeholders was seen by NIWETU as interwoven and crucial to the effectiveness of the adjoining programmatic pillars. Various assessments over the life of the activity measured the degree to which stakeholders felt they had acquired new CVE knowledge and skills, the perceived utility of the information, and whether they were able to apply the skills in their day-to-day lives. The charts show an increase in participation in CVE training over the past 12 months, with the exception of Wajir County where there has been a slight decline. The utility of these trainings, however, have declined in all counties except for Mandera, where there was an increase in perceptions of the relevance of these trainings. It is possible that the decline is a result of NIWETU’s sustained 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4 Garissa Isiolo Mandera Nairobi Wajir Average Utility (1-4) Utility of Training Content Baseline Midline Endline 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Garissa Isiolo Mandera Nairobi Wajir Percentage of Respondents Respondents that Participated in CVE Trainings Baseline Midline Endline 21 effort in these counties over the past four years, and that sensitization and awareness raising has been effective, but there is now a hunger and a need to shift strategies to continue building off of the gains already made at the county level.  Majority gained new CVE knowledge and skills through training – The majority of endline respondents said they had gained new knowledge and skills as a result of their participation in NIWETU-supported workshops. The main topics and themes where stakeholders demonstrated changes in their understanding included: the role of communities and security actors in CVE, the dynamics of radicalization, the importance of inter-religious tolerance, and the legal frameworks governing acts of terrorism.  Utility of skills diverged along gender lines – In terms of gender dynamics, men and women reported similar levels of skill usage, while slightly more women report using new skills ‘very often’, whereas men said they used them ‘a few times’. This slight disparity in utility between men and women was considered to be influenced by a number of variables, most notably differences in socio-cultural norms, making women more open to reconciliation and dialogue than men in some target communities.  Increased CVE knowledge had cascading impacts – Respondents frequently alluded to the role of NIWETU-supported CVE workshops as empowering community members to become peacebuilding ‘ambassadors’ and ‘influencers’ among their wider social circles. This broader catalytic effect was evidenced through the organic emergence of informal youth networks to share CVE related knowledge, the creation of women’s self-help and entrepreneurial groups, and the identification of Champions for Change (C4C) candidates. The reach and impact of these groups is discussed in greater depth in subsequent sections.  Champions for Change significantly improved understanding of VE issues – The C4C case study pointed to a greater understanding of the drivers of extremism, what makes people in their communities at risk, and the narratives that are used to recruit and radicalize. In addition to community champions, who participated in the yearlong C4C activity, NIWETU also trained approximately 30 government champions, such as Regional Commissioners (RCs), DCCs, and CCs, including former CCs from Mandera, Wajir, Garissa, and Isiolo counties. A community champion from Mandera said his engagement with NIWETU-supported events helped demonstrate the variety of roles residents play in maintaining security in their communities, which helped to inform the design of the one-year action plan. There was also evidence, among government stakeholders, of enhanced knowledge of VE and CVE strategies. The levels of knowledge and interest, however, varied considerably across institutions and target counties, depending on the extent of political will, the perceived negative impacts of VE, and specific personalities involved. Some CCs, such as in Isiolo, demonstrated particular interest in CVE efforts and thus mobilized their subordinates, police, and other security actors to engage with NIWETU-supported workshops. Their training had a catalytic affect by progressing CVE awareness in new counties where they were subsequently transferred. In other locations, such as in Kamukunji in Nairobi, heavy-handed security crackdowns meant that police were less proactively involved in partner-led workshops and demonstrated lower levels of knowledge of community CVE strategies. There was, consequently, a strong correlation between the level of political backing for CVE from the County Commissioners’ offices and the level of participation in knowledge and skills training of lower-level officials. Specific examples and lessons learned are discussed in subsequent sections. 3.3.3 Openness to CVE dialogue expanded As found at the goal-level, interventions under this objective made significant strides in helping stakeholders open up about VE issues they were facing in their communities. This marked a notable 22 step change from the fear and stigmas surrounding discussion of these issues in many areas prior to outreach by activity partners. This willingness to engage on the underlying dynamics of VE extended beyond community residents, to local authorities and security actors who participated in NIWETU￾supported workshops.  Local government more willing to engage in CVE in some areas – The Wajir Peace and Development Agency (WPDA), for example, said that prior to its trainings, heads of local government departments typically shied away from dealing with security-related issues, because they felt these tasks were only the responsibility of police and specific security actors. Over time, however, government stakeholders voiced their understanding that effective CVE requires a community-wide response with the support of local authorities at all levels.  Youth assume role as ‘peace ambassadors’ – Youth in Wajir said the knowledge and skills initiatives had increased the number of ‘peacebuilders’ in their area. The Wajir County Cohesion and Integration’s Department Peace Officer stated that there was a realization that VE was usually an issue from within, and not necessarily brought from the outside. Similarly, youth focus group discussion respondents in Garissa said that NIWETU-supported workshops had helped them become ‘peace ambassadors’ and ‘Champions for Change’ in their communities.  Communities take initiative in leading CVE forums – A teacher from an Isiolo girls’ school who participated in a workshop organized by the Supreme Council of Kenyan Muslims (SUPKEM), noted that it was not only facilitators who led the discussions, but impromptu talks were given by other community members who were interested in CVE and played an active role in coordinating activities. The subsequent sections provide examples of key knowledge areas that emerged as a result of these initiatives. 3.3.4 Ability to identify and counter VE narratives enhanced A common theme observed across stakeholders was a significantly improved ability to identify and counter VE narratives that were circulated within target communities. These narratives included messages aimed at radicalizing youth, extremist religious doctrines, and recruitment propaganda promoting the economic incentives of participating in VE activities. Changes in capacity to spot and respond to these messages were seen across communities, including youth, teachers, religious leaders, and boda boda networks.  VE narratives countered in educational settings – A teacher who participated in a SUPKEM￾organized workshop, said that school administrators in Isiolo were better equipped to identify dangerous messages that were being disseminated by religious leaders who led different classes in their schools. As a result, they were able to collectively address the problematic messages and prevent certain guest lecturers, who shared this propaganda, from being invited back.  Moderate ideological interpretations more common – The interpretation of the meaning of Islamic jihad was a topic that emerged frequently in conversations with Kamukunji youth. In an outcome harvesting session with girls from the area, they recounted a program set up by a prominent imam at a local mosque, explicitly aimed at recruiting youth to become mujahideen to fight in Somalia. They explained that this program had been shut down by joint government￾community action and they were now aware that the conflict in Somalia was not a true jihad and therefore did not deserve support.  VEO incentives less of a ‘pull’ factor – Youth participants in an outcome harvesting session in Kamukunji said that, as a result of the Jichanue workshops organized by the Kamukunji Community Peace Network (KACPEN), they now understand that the alleged opportunities offered by VEOs operating in Somalia are a ‘scam’. One outcome harvesting participant explained that all of his friends now understood that there areare no legitimate financial gains to be offered by VEOs. He also pointed out that some individuals who had shared their experiences during the 23 Kumekucha workshops suggested that life in Majengo is better than what they might face in Somalia.  C4Cs contribute to counter narratives – C4Cs were examples of the cascading impacts of efforts to counter VE narratives in affected communities. One Champion from Kibra Constituency in Nairobi said that he began to use his platform as a religious leader to disseminate important messages through weekly sermons and discuss VE issues in more intimate settings with the members of his mosque. He underscored the importance of countering false religious narratives with correct interpretations of religion to push back against VEOs. 3.3.5 Inter-religious/ethnic tolerance boosted In addition to countering VE narratives, community stakeholders exhibited increased tolerance for other religions and a willingness to interact with individuals from different religious backgrounds, whom they would have otherwise avoided prior to participating in NIWETU-supported initiatives. This change in levels of tolerance was seen by activity staff and community leaders as critical to mitigating tensions and boosting social cohesion in areas with historic rifts between religious groups. In particular, there was an acknowledgment that VE issues were not exclusive to Muslim communities and required a collective response from residents of all religious backgrounds, in order to build resilience.  VE seen less as a ‘Muslim’ issue – Local residents in Isiolo confirmed that previously VE had been perceived as only a ‘Muslim issue,’ leading Christians to blame Muslims for a broad range of issues that arose in the county, including some which were not directly linked to VE. They said, however, that perceptions had shifted to viewing VE as a community-wide problem that did not exclusively affect one religious group, noting that a variety of factors make certain areas more vulnerable to VE recruitment. These observed improvements in religious tolerance were attributed, in part, to workshops delivered by the NIWETU-supported Isiolo Peace Link (IPL).  Greater inter-religious collaboration – Respondents said they had observed more constructive interaction between Christians and Muslims. They said they learned, as a result of their participation in workshops delivered by NIWETU-supported partners, that a variety of communal dynamics played into VE issues. Such changes in understanding was said to have contributed to a reduction in ethnic-based violence and ‘tribalism’ in Kibra. Similarly, religious leaders who participated in a NIWETU-supported trauma healing initiative said they used the relationships established during the sessions to organize inter-faith forums between Christians and Muslims. The assessment noted, however, that significant fractures persist along ethnic and religious lines that are deeply embedded within mixed faith communities and will likely continue to contribute to tensions. 3.3.6 Women perceived as critical to CVE response Across target areas, stakeholders point toward negative gender dynamics as a key contributor to communal vulnerability to VE. NIWETU-supported research found that this vulnerability was linked to a discriminatory view of the role of women in CVE that promoted segregation and exclusion. From the outset, NIWETU programming was designed to focus explicitly on issues of gender by ensuring activities were inclusive and geared toward breaking down gender barriers in target communities. There was a further focus on emphasizing the critical role women play in effective CVE response, especially in building resistance to VEO recruitment activities.  Women more open to discussing issues of radicalization – Participants from the Pastoralist Girls Initiative activity in Garissa said that women had become more comfortable sharing information and talking about early signs of radicalization in their families. In Garissa and Wajir, women noted feeling more comfortable reporting to other women and expressing their concerns to local female leaders. Women in Garissa said they felt more comfortable sharing 24 security-related information with their local chief, but when they were not comfortable going directly to the chief, they shared the information indirectly through other women’s representatives. They noted that the most trusted networks for reporting are the Nyumba Kumi and peace committee members.  County-level CVE forums increasingly gender-balanced – Many stakeholders attributed this, in part, to the creation of a gender-focused pillar within CAPs, which they felt helped institutionalize the importance of women in CVE responses. A grant to the Isiolo County Commissioner’s office, for instance, to build the CVE capacity of female security officers and representatives, played an important role in outreach and knowledge dissemination throughout the community. The Rights Organization for Advocacy and Development (ROAD), noted that the cultural and security context in Garissa is known to be very male dominated. ROAD therefore proactively engaged peace committee heads to emphasize the importance of including women in meetings and establishing separate meetings for women and men where cultural barriers persisted.  Single mothers better equipped to talk about VE issues with family – This was true in Majengo, where stakeholders talked about the challenges faced by single mothers in addressing issues of radicalization. Women participating in CVE knowledge and skills workshops delivered by NIWETU partners said the lessons had helped them discuss difficult issues with their children and other family members in more constructive ways. They further noted that, while family dynamics were still changing, the levels of trust and respect between them had improved significantly. Overall, support for improved knowledge and skills, through a gender-sensitive lens among stakeholders, created the foundations for gender-inclusive CVE networks that emerged over the course of implementation. These networks, which included a variety of women’s groups, are discussed in more detail in the next section. 3.3.7 Networks NIWETU’s interpretation of ‘networks’ was broad and thus encompassed formal and informal entities ranging from more established peace committees to ad hoc youth groups in target counties. Some networks NIWETU supported existed prior to the activity and others newly emerged as a direct result of its interventions. This section looks at the most relevant networks within this sphere and their relationship to wider CVE responses. 3.3.7.1 Diverse linkages created across CVE actors Activities implemented by NIWETU partners contributed to a variety of newly formed or strengthened linkages across CVE stakeholders. There was also evidence that many of these connections would likely not have occurred had it not been for forums convened through the activity. Respondents most frequently reported new engagements with youth groups, women’s groups, religious leaders, chiefs and government officials. Examples include:  The assistant chief of Arbaqeramso in Wajir East said that the activity connected him to new actors like teachers, religious leaders, and county-government officials who he would not have otherwise been in touch with.  In Majengo, seven influential local chiefs created a CVE-focused WhatsApp group following their participation in Kumekucha sessions to support each other, particularly when it came to dealing with issues in their respective communities. The chiefs said, in outcome harvesting sessions, that the regular interaction had helped them form closer bonds that extended beyond their official, work-related roles.  A chief from Garbatulla said that boda boda networks were now seen as an important source of information for authorities about security-related incidents occurring in their areas. 25  Boda boda drivers in Isiolo said they had created spaces where they could openly discuss CVE issues, usually around their stands during working hours. The charts provide a snapshot of sentinel indicators regarding community perceptions of trends in the activity of peace committees and Nyumba Kumi between midline and endline assessments.10 Encouragingly, there are notable upticks in the activity of these entities in most target counties, demonstrating greater awareness of their presence and actions among stakeholders. 3.3.7.2 Champions for Change connect CVE activists NIWETU’s C4C initiative aimed to leverage existing CVE networks to identify individuals seen as positive influencers. In its C4C case study, Wasafiri found that the relative effectiveness of the Champions networks was due to robust selection criteria. Champions were vetted according to a) their understanding of CVE issues, either through work in the sector or through related work in peace, security and/or conflict resolution; and b) their reputation in their communities and how they were perceived by community leaders and local authorities. In all, the activity supported 52 C4Cs identified in its core counties of Garissa, Isiolo, Mandera, Nairobi, and Wajir.  NIWETU helped C4Cs establish CBOs – As part of the activity, several Champions registered community-based organizations (CBOs) to enable them to compete for funding and develop a track record of CVE initiatives. A Kibra-based Champion began collaborating on CVE initiatives with BAWAKI, a local CSO, that brought together Christian and Muslim religious leaders. Afterward, he started independently raising awareness about the risks of VE and the impact on youth in particular. Similarly, Salim Juma created the Kamukunji Community Peace Network (KACPEN) that worked to spread CVE messages to Majengo residents. Salim was awarded the state commendation award by President Uhuru Kenyatta in 2018 for his fight against extreme violence.  Champions boost CVE messages through media – Media outlets often rely on NIWETU’s C4Cs for analysis of VE trends in the region. Following an attack in early 2020 on a school in Saretho, Garissa County C4C Sheikh Hassan appeared on local television to encourage collective action to resist the influence of VEOs. Voice of America Somali also interviewed Champions from Garissa and Mandera about CVE in Kenya. Another C4C is the manager of Wajir Community Radio, a highly popular station in the region, where she champions CVE-related radio content drawn from key stakeholders in her network. 10 Certain ‘sentinel’ indicators were identified iteratively during implementation and therefore data for some were not collected as part of the baseline assessment. 0 1 2 3 4 5 6 7 8 Garissa Isiolo Mandera Nairobi Waji r Average Agreement with Statement (1-10) "I feel that peace committees are active" Midline Endline 0 1 2 3 4 5 6 7 8 Garissa Isiolo Mandera Nairobi Waji r Average Agreement with Statement (1-10) "I feel that Nyumba Kumi are active" Midline Endline 26  C4Cs build national consortium – NIWETU held a final C4C networking event in March 2020, where the Champions agreed to form a national consortium of five of their CBOs to promote learning and collaboration and improve their ability to coordinate CVE initiatives. Additionally, they established a C4C WhatsApp group to share topics for debate, confirm security events, and ask CVE￾related questions. There were numerous other examples of CVE Champions that emerged, that were not linked to the C4C, but took particular initiative in driving forward CVE strategies. A prominent example was Abdia Mohamud of IPL, who was commended for being diplomatic, personable and resourceful, not only as an entry point to the community, but also in helping activity teams understand the context. She developed a strong social network within and outside of Isiolo, was well connected to security actors, and nurtured women’s leadership in CVE. Similarly, the former Isiolo County Commissioner, John Ondego, proactively Championed CVE work and is now participating in KSG CVE workshops as CC for Kitui. 3.3.7.3 Informal youth networks strengthened NIWETU supported research activities found that youth were likely the most influential demographic in building community-level resilience to VE in target counties. The activity, consequently, took deliberate steps to engage with youth through a variety of platforms. This resulted in the establishment and strengthening of several types of formal and informal youth CVE networks. These networks formed around sports and livelihood activities, educational institutions, and social media.  Self-help groups for men and boys formed – Following the trauma healing initiatives as part of the Kumekucha activities, youth in Majengo created a discussion group for local boys called Maendeleo ya Wanaume and formed a WhatsApp group as an informal, self-help network for survivors of trauma to keep in touch and support those still struggling with related issues.  CVE ‘clubs’ created in schools – In Wajir, youth set up membership programs, in a local secondary school, on peace and security for youth to share their perspectives on CVE issues. In Isiolo, youth set up ‘peace clubs’ in schools, following their involvement in a NIWETU-supported workshop organized by SUPKEM. In Isiolo and Mandera, school-based Amani Clubs played a transformative role in CVE initiatives as sustainable student and faculty-led forums for engagement, integrating CVE into official club guidelines.  Inter-community youth networks formed – Youth in Kamukunji established CVE-related networks that spanned multiple villages, such as Katanga and Mashimoni, to raise awareness of CVE issues and address emerging challenges in their communities. These bases are not a new phenomenon, but their utilization as forums to engage on CVE-related issues was considered a relatively recent development, linked to NIWETU-supported activities. 3.3.7.4 Women’s groups play critical CVE role NIWETU-supported research research found that a number of formal and informal women’s groups appeared in NIWETU target areas over the course of implementation. Some were found to have been a direct result of activity interventions, while others grew organically in response to extant needs in communities. Additionally, there was evidence that these groups were considered effective CVE mechanisms and were well regarded among local residents. Examples include: J2SR for C4Cs The activity positioned C4Cs for sustainable commitments by developing a leadership training package and facilitation guide and coaching them on how to independently seek other sources of funding for their action plans across VE hotspots. This included support to the creation of local CBOs to submit funding proposals, and the formation of a national C4C consortium to lobby national government actors for support. An additional, unintended outcome was the commitment of the Mandera County government to launch its own CVE Champions program, following support from NIWETU C4Cs, with community members and government representatives across the county. 27  Women’s self-help groups formed – Following the Kumekucha sessions, women in Majengo said they formed ‘self-help’ groups, which they referred to as ‘table-banking’, to provide informal psychosocial support, and sometimes financial assistance through lending programs. Some such networks existed prior to the activity, but many were said to have sprung up as a result of the dialogue forums.  Small business collectives established – The CVE initiatives, in Kamukunji, which occurred following the trauma healingc sessions, helped to further formalize these women’s networks. Examples included small business collectives and other entrepreneurial initiatives.  Women advocate for social justice – In Wajir, women involved in NIWETU-supported activities organized by Wajir Women for Peace (WWFP), said the knowledge and skills gained had empowered them to challenge unreasonable crackdowns by security actors, by amplifying women’s voices through better coordination. Teachers in Isiolo who participated in a workshop organized by SUPKEM said they gained significant confidence in their abilities to discuss VE issues and advocate to local authorities on behalf of their community. Stakeholders in Garissa also said women’s groups had become collective support networks and forums through which to challenge injustice. 3.3.8 Community-led strategies and initiatives The sequencing of interventions under the first activity objective followed a trajectory from the enhancement of CVE knowledge and skills, to the formation of networks, culminating in the formulation of community-led CVE strategies. While the evolution of this process was understandably non-linear, the report structure observes this progression. 3.3.8.1 Psychosocial support builds community resilience The expansion of trauma healing services by NIWETU in affected communities represented a first-of￾its-kind initiative at this scale. Intervention logic was built on the assumption that family reconciliation and open discussion of traumatic events, especially between parents and at-risk youth, is crucial to building resilience to VE narratives and recruitment activities that target vulnerable demographics. The success of the Kumekucha initiatives was evidenced by:  Support groups continue organically – Informal ‘self-help’ groups were continued, following activity completion, throughout target areas, often facilitated by community leaders. Additionally, these sessions unpacked in more depth, the push-pull factors of VE that residents struggled with on a regular basis.  Local authorities more open to community concerns – Government officials, in particular, self-reported this change, saying it helped them engage with their constituency with more patience and understanding.  Better intra-household relations – Participants often reported that the sessions contributed to more harmonious household environments and a reduction in domestic violence. Parents and their children reported improvements in overall communication, particularly regarding issues of VE, societal influences, and potential risks.  Reduction in public disputes – Women who participated in trauma healing sessions said they observed a reduction in conflicts at community waterpoints where they often congregate to fetch water and wash clothes. They remarked that, where previously communal differences would spark frequent arguments, they now used it as an opportunity to share knowledge and lessons more widely from the psychosocial initiatives. This resulted in an increase in female membership to informal ‘self-help’ groups created following the completion of the Kumekucha activities. 28 3.3.8.2 C4C action plans establish CVE roadmap While preceding sections touch on initiatives taken by NIWETU’s Champions for Change as part of their action plans, there were a number of instances where the plans cascaded knock-on activities in their broader communities. The ability of C4Cs to catalyze this collective action was indicative of their already established positions as influencers and mobilisers in their respective counties and sub￾counties. Some examples documented in Wasafiri case studies include:  The establishment of CVE committees in schools – NIWETU supported the creation of a committee in Garissa Township, comprised of teachers, Parent Teacher Association members, Boards of Management members, Ministry of Education officials and some student leaders, representing longer-term community commitments to integrating CVE concepts into the local education system, to build sustainable resilience among students and faculty.  Recurring CVE forums – Champions in Mandera East sub-county created a regular CVE forum involving elders, religious leaders, security officers, and county government representatives, in order to learn about community perceptions and cultural beliefs affecting VE, to demonstrate the importance and effectiveness of ‘softer’ approaches to policing and community security.  Community art exhibitions – Kamukunji-based Champions in Nairobi channeled different CVE messages through informal, public art exhibitions, which they said were cost-effective and successful in attracting youth for further discussions. One champion observed that such approaches have an added benefit of removing the expectation of compensation or reimbursement from local residents for participation. 3.3.8.3 CAPs catalyze civil society collaboration A positive impact observed in many target counties was the ripple effect the CAP process had on connecting local CSOs for further CVE collaboration, outside of the specific action plans themselves. NIWETU worked closely with NCTC in theCAP formation process in Garissa, Isiolo, Mandera and Wajir counties, while supporting local partners to lead strategies within each county. A critical factor influencing the likelihood and effectiveness of observed ripple effects, was the level of inclusiveness attributed to CAPs, especially the degree to which civil society was perceived to be consulted in a representative, ongoing manner. CAP development timelines often followed very different trajectories, and NIWETU worked with the Malaika Foundation, which led the drafting of the CAPs in each county, to support these trajectories. Some outcomes included:  Inclusive CAPs spurred greater initiative – In Isiolo, participants commended the inclusive and collaborative CAP process and environment that helped cascade positive outcomes of the sessions even further. Stakeholders said the consultations included a diverse range of stakeholders including interfaith groups, the Takuma Youth Group, the Council of Imams, Nomadic Women for Sustainable Development (NOWSUD) and Sustainability of Community on Radicalization and Empowerment Solutions (SCORES), among others. They noted, from a practical standpoint, that the creation and effective management of a CAP WhatsApp group was key to promoting communication and notification of relevant events across numerous groups and actors.  CAPs facilitate CVE civil society mapping – 350 participants, including youth from local schools, senior government and military officials, women, human rights advocates, and CSO representatives, attended the CAP launch in Wajir. WPDA, NIWETU’s local partner in Wajir, which led the formation process, conducted a thorough CSO and stakeholder mapping across the county in order to identify and invite attendees for the session. The proactive engagement of civil society contributed to greater effectiveness of its resulting CEF, which mobilized a 12-member committee to focus on hotspots in Wajir East and Tarbaj. While the majority of target counties have seen encouraging progress in CVE initiatives due to the proactive coordination of CSOs mentioned above, the relative effectiveness of CEFs is heavily reliant on constructive and supportive government involvement. Key considerations regarding levels of 29 political will to implement CVE responses and corresponding limitations are discussed further under lessons learned for Objective 2. 3.4 LESSONS LEARNED AND RECOMMENDATIONS: (Obj1) Community mobilization to address violent extremism The following lessons learned were derived from, and build on, processes related to the key outcomes discussed in the previous section. Some of these findings were perhaps unintended or unanticipated results of activity activities that could not be avoided but are worth noting for wider community of practice implementing activities of a similar nature. Overarching conclusions and recommendations are discussed at the end of this report. 3.4.1 Communities should opt for vigilance over suspicion Enhanced community vigilance about VE activity is typically considered a positive outcome of programming. In some cases, however, a byproduct of enhanced vigilance was xenophobia and increased suspicion of outsiders. There were several instances identified where stakeholders reported negative consequences of heightened sensitivities to perceived VE threats:  Migrant newcomers should be safeguarded – Local residents in Wajir County acknowledged a strong link between increased CVE vigilance and unwarranted levels of suspicion of ‘newcomers’ to their communities. This was said to have contributed to an overall improvement in security, but also subjected many innocent residents to unnecessary scrutiny. Care must be taken to safeguard those who are new to a community but pose no legitimate threat.  Local authorities must not betray community trust – In Kamukunji, one community leader said that the Assistant County Commissioner (ACC) encouraged ex-combatants to report to his office if they wished to reintegrate, but that these individuals sometimes were disappeared or killed shortly thereafter. Authorities must be persuaded to demonstrate good will in their communities by protecting returnees and ex-combatants from violence. Previous Wasafiri reports noted that a high degree of suspicion of outsiders or newcomers is not always a good thing. It risks alienating and side-lining visitors or new residents who may not be familiar with the customs and norms of a new location, and the need to identify themselves or register with relevant groups or individuals. Coupled with the security sector’s often heavy-handed approach, any potential false accusations against new entrants into a community could put them in undue jeopardy. 3.4.2 CVE activists must be protected The nature of CVE work often entails higher levels of visibility of key stakeholders and organizations involved in outreach and engagement activities. This heightened visibility will sometimes lead them to be targeted by militant groups, or those seeking to deter wider CVE interventions from occurring in their areas. While activities typically take deliberate steps to enhance duty of care and protect at-risk stakeholders, there are instances when enhanced vulnerability to targeting as a result of participation in CVE activities is unavoidable.  Activists require sheltering support networks – A SUPKEM representative noted that VEOs watched their activities closely and presented a security threat both to the organization and participants of their workshops. They said they often worry that individuals who reside in areas where CVE events are held could subsequently be exposed to attacks. Meanwhile, CSO representatives in Kibra said that, although VE is less of a threat in comparison to other hotspots, participants have received threatening text messages when talking about CVE in the community. In such cases, activists should be advised on how to insulate themselves from violence and call on local groups for help if needed. 30  Mechanisms needed for anonymous reporting – NIWETU progress reports note that young people in VE-affected areas, such as Mandera East, Mandera North, and Lafey, are frightened of VEOs and thus hesitant to name the group or openly discuss CVE initiatives. They further noted that VEOs have spies within communities who can report individuals back to sleeper cells and deter teachers from speaking out about VE in their classrooms. CVE implementers must therefore be acutely aware of these risks and accordingly take steps to integrate duty of care and conflict￾sensitivity measures throughout activity design. 3.4.3 Greater input needed from ‘grass-roots’ CSOs and commercial actors The CAP development process highlighted tensions between CSOs in some target counties, due to perceptions that some organizations had greater presence in certain areas and were therefore more representative of local interests than others. This distinction was also acknowledged by local authorities who were resistant to larger, external organizations which they suspected of seeking influence and resources over other, better-placed CSOs. While most CAP processes involved a mixture of organization types, there were clear capacity gaps in some counties that required the facilitation of external entities, which was sometimes resented.  CEFs should prioritize local organizations – A government official from Garissa County explained that it was often challenging to manage civil society expectations, as many entities are ‘briefcase’ organizations that have no physical, local presence and are involved primarily to receive monetary stipends. A CSO representative familiar with the process in Garissa said, when a CSO meeting is convened, there sometimes were more than one hundred organizations in attendance, many of whom have no tangible links to the community. In such cases, CSOs with demonstrable presence in target areas should be prioritized.  CVE outreach should move beyond urban centers – In Isiolo County, stakeholders observed that that only a few CSOs have the capacity to implement CAP pillars outside of the main town. This dovetails with concerns around ‘elite-capture’ when conducting community outreach for CVE initiatives. Stakeholders in Isiolo County said that many important stakeholders outside of peri-urban areas, such as pastoralists and grazing communities, were not involved in CAP development due to access issues. Future activities should take deliberate steps to engage rural and marginalized communities, such as NIWETU C4Cs in Garbatulla, who held activities at local watering points.  Commercial and private sector stakeholders should be better engaged – Stakeholders in some areas observed a lack of participation by shopkeepers and private business owners in CVE activities. In Kamukunji, for example, a local resident emphasized the importance of including merchants from the Gikomba Market in Majengo, who are directly impacted by VE issues that jeopardize their safety, productivity, and broader economic interests. CVE initiatives should find ways of increasing their voice in community workshops, as an untapped area of knowledge and resources. 3.4.4 Restrictive gender norms remain a challenge While this report provides a number of examples where significant progress occurred in the promotion of greater gender balance in CVE response, entrenched cultural paradigms remain a primary barrier to the broader engagement of women across activities. NIWETU took deliberate steps to integrate gender consideration across its portfolio, however, more actors in the CVE sphere should make this a priority in order to tackle broader marginalization.  CVE practitioners should build on NIWETU’s gender strategy – Domestic conflicts will likely continue to be an issue, as women from male-headed households said they were often forbidden from joining different groups. NIWETU stakeholders, such as women participating in activities in Wajir, reiterated that to engage in peacebuilding in their communities, they must fight 31 stereotypes and create a space for themselves to be involved. NIWETU’s gender strategy took deliberate steps to create safe spaces where women could feel comfortable engaging on CVE issues, even in extremely conservative cultural contexts.  Sociocultural activities should encourage gender-inclusive participation – Some forms of cultural discrimination are more challenging to overcome, for instance, women still face limitations in their ability to actively participate in certain sports tournaments. While women may have been able to attend a football match, they are not allowed to play as a part of a team. In such cases, implementers must find creative ways to overcome these forms of segregation to promote more inclusive programming. 3.5 OUTCOMES: (Obj 2) Government responsiveness to VE Objective 2 focused on promoting government responsiveness to VE. At the intervention-level, this included direct support to national and county governments in the form of technical advisory services, capacity building workshops, and logistical resourcing. While NIWETU’s conceptual framework only listed one results area under this objective – ‘Increased Government CVE Capacity’ – it encompassed a broad range of activities, which are examined in greater detail below. 3.5.1 Government CVE capacity To be effective, government CVE capacity building efforts must take into consideration the multidimensional power dynamics within government agencies and between national and county-level institutions. Given that national government CVE policy is still somewhat in an embryonic phase, activities often operate in a grey area, managing horizontal and vertical relationships where clear CVE responsibilities are not yet defined. Within this environment, NIWETU made considerable progress in the following areas. 3.5.1.1 County-level knowledge and responsiveness enhanced CVE knowledge and skills among county-level government actors improved significantly over the life of the activity. Resultantly, there was evidence that overall responsiveness of county officials to community concerns and national policy imperatives improved in some areas. The degree to which capacity was enhanced, however, was not consistent across all target counties and was heavily influenced by multiple factors, including whether counties were viewed as VE ‘hotspots’, pre-existing relationships between communities and local authorities, and the individual personalities present in county offices.  Community-government relations strengthened in Isiolo – This was largely due to proactive engagement of county officials in NIWETU-supported CVE workshops, and special initiative taken by individuals, such as the Deputy County Commissioner to respond to concerns 0 0.5 1 1.5 2 2.5 3 3.5 4 Garissa Isiolo Mandera Nairobi Wajir Average Ranking (1-4) "How well does the county government understand VE?" Baseline Midline Endline 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 Garissa Isiolo Mandera Nairobi Waji r Average Ranking (1-4) "How well does the national government understand VE?" Baseline Midl ine Endline 32 of local residents. Respondents acknowledged that collaboration between CVE actors and youth improved during 2019, because of better and safer channels of reporting about returnees and instances of radicalization. Activities that facilitated more constructive interaction between youth and police, such as football matches and local dialogues, created a positive environment.  NIWETU workshops bolster government responsiveness in Wajir – The senior chief of Wajir-Bor reported that government officials had begun sharing important CVE skills and awareness tactics that they had gained from NIWETU-supported workshops. He also explained that he organized a joint dialogue activity, as a result of his participation in CVE trainings, that included religious leaders, youth, women’s groups and government officials. The Tarbaj sub-county administrator in Wajir County also said that he had used the CVE training manual to train ward administrators in Tarbaj on key issues.  Improved government-youth relationships led to greater collaboration on CVE in Kamukunji – As noted in previous sections, there were examples in Nairobi constituencies of Kamukunji and Majengo where the respective OCS and DCC had taken initiative to build constructive relationships with youth, leading to greater collaboration on CVE issues. In some cases, youth reported that the officials had provided their personal mobile numbers and encouraged them to get in contact if there were urgent needs. The increase in CVE responsiveness among government officials was, however, negatively impacted by the high rates of turnover among local authorities and security actors, which is discussed in more detail under ‘Lessons learned’. 3.5.1.2 Intergovernmental CAP coordination enhanced The CAP development process in many target counties was quite challenging due to the often-unclear roles and responsibilities between county and national government stakeholders. There is evidence that the CAP process did, however, lead to greater intergovernmental coordination, since CEFs were typically co-led by representatives from both county and national government, ensuring that both levels had an equal role to play. While county officials’ ongoing participation was still lacking in some counties, communication and coordination with national government counterparts nonetheless improved.  Rapid CAP expansion improves NCTC-County coordination – There were examples where the RCAP expansion process improved relations between the NCTC and counties that were historically seen as low priority for CVE programming. In target counties, stakeholders said the national government played a proactive and effective role in the CAP formation process. The effectiveness of intergovernmental coordination at the county-level was heavily influenced by the level of buy-in from county officials.  NIWETU bolstered CSO CAP ownership where government gaps existed – The absence of the Governor in the Isiolo CAP development process was particularly noticeable, because of the community’s keen interest in its implementation. The active role of the County Commissioner was, consequently, credited with ensuring that the CAP document was successfully finalized.  National-county relations strengthened in some areas – County and national government representatives in Mandera played an active role in CAP development and implementation. In Wajir, relationships between county government and some CSOs were strengthened through the CAP process, which created a platform for further engagement. Likewise, in Garissa, county government and national government were reported to both be committed to the process since it is considered a hotspot for VE attacks. 33 3.5.1.3 Security Management Institute established Part of NIWETU’s targeted assistance to national government actors included interagency capacity building and CVE training. The activity explored various avenues for delivering this support and ultimately identified the creation of the Security Management Institute (SMI) within the Kenya School of Government (KSG), a parastatal entity to offer advisory, consultancy, research and training services to the public sector. NIWETU saw this as a cross-cutting, sustainable means of providing for longer￾term capacity building initiatives.  NIWETU navigates agency mandates to foster cooperation – NIWETU initially had to navigate the potentially overlapping roles of the NCTC’s capacity building mandate and that of the KSG. This took considerable time and required an MoU to minimize potential areas of duplication between KSG and NCTC, which resulted in delays to the implementation plan. The MoU, however, ultimately provided for greater NCTC buy-in to the SMI capacity building curriculum and resulted in more direct coordination between the two entities.  Interdepartmental CVE capacity strengthened through new curriculum – Following the MoU, SMI developed a training curriculum for a certificate-level course, targeting government civil servants and senior-level managers in the public sector, across the judiciary, prisons service, police, intelligence, and other security sector stakeholders. A pilot training took place over a five-day period in February 2020, with participants from the Ministry of Interior, the Anti-Terror Police Unit, National Intelligence Service, and National Police Service, amongst others. This training took place as part of the NIWETU grant activity with SMI, but equally were strategically integrated into its five-year strategic planning, as government officials will be mandated to attend either the executive course, or the non-executive course, depending on cadre. Implementation of all of SMI’s activities will depend on the implementation of this five-year strategic plan, which includes resource streams from government and also donor support. These government actors also contributed through research validation workshops and the dissemination of relevant policy briefs.  SMI facilitates development of indigenous CVE research capacity, knowledge sharing and dissemination – The policy briefs developed by the SMI staff, with support from consultants hired through SMI, included a review of policy gaps in responding to VE in Kenya, and a critical examination of gaps in the implementation of existing policies. Additionally, it published two case studies, focusing on returnee policy and capacity-building gaps in government. The SMI also hosted NCTC’s and the Tony Blair Institute’s two-day workshop on CVE in the public sector in early March 2020, providing an early indication of longer-term sustainability. 3.6 LESSONS LEARNED: (Obj2) Government responsiveness to VE Lessons learned for Objective 2 are intended to inform activity design and implementation processes for future activities of a similar nature. J2SR for GOK Capacity Building Following the facilitation of an MoU between the NCTC and the KSG, both entities committed their own resources toward collaborative public sector CVE capacity building through the SMI. NIWETU provided capacity support to SMI staff that enabled them to conduct their own CVE policy gap analysis of the public sector and develop a curriculum to fill some of these gaps. KSG also mainstreamed CVE concepts throughout its leadership and management courses that are delivered to high-ranking public servants, including its Strategic Leadership Development Program that is provided to incoming county commissioners. 34 3.6.1 Managing expectations in an evolving context Improvements in government responsiveness and policing tactics were not consistent across government stakeholders involved in NIWETU-supported CVE initiatives in target counties. As noted earlier, a number of variables, internal and external, contributed to observed disparities. Internal variables (i.e. those within the activity’s sphere of control) included the comparative quality of the interventions, the capacity of delivery partners, and the level of resource allocation in specific geographies. External variables (i.e. those outside of the activity’s sphere of control) included pre￾existing levels of instability, political buy-in of government counterparts, and community openness to CVE interventions.  Communities must adapt when authorities are unresponsive – Stakeholders noted cross￾government policies that encouraged security actors to handle VE-related cases in a ‘friendlier’ way, to avoid a further deterioration of community-security relations. However, in the Dadaab sub-county of Garissa, a C4C noted that county officials often fail to show up to meetings or activities to which they have been invited and did not respond to requests for logistical support to continue carrying out CVE activities. Another Champion similarly pointed to difficulties in coordinating with county officials in Garbatulla, suspecting that they let CVE issues rest with national government counterparts. In such cases, NIWETU worked with stakeholders to find broader mechanisms for resourcing and engagement in addition to government support.  Resilience must be built to overcome fragile government-community relations – In Wajir County, CVE stakeholders observed greater participation of security actors in activities but did not see a link to improved policing behavior. The Wajir County Director of the Peace, Cohesion and Integration Commission, noted that security forces had, “changed their attitudes, but not their actions.” Additionally, there were examples where specific security incidents harmed relations between communities and security actors who were otherwise moving in a positive trajectory. This was true in the case of an extrajudicial killing on January 16th, 2020, by police of Hemedi Majini, a 19-year-old resident of Kamukunji. Within this context, NIWETUT encouraged communities to view CVE through a longer-term lens, understanding that stakeholder relations will continue to be volatile. 3.6.2 CEFs effective in ‘hotspots’ with political buy-in There is a strong correlation between the effectiveness of CAP/CEF processes and the perception of whether a county or sub-county was considered to be a VE ‘hotspot’. There was a substantially greater likelihood of CAP processes enjoying broad government support and resourcing in areas where VEOs had a greater presence and incidents were more frequent. 0 1 2 3 4 5 6 7 8 9 Garissa Isiolo Mandera Nairobi Wajir Average Agreement with Statement (1-10) "The government made effot to participate in dialogue" Midline Endline 0 1 2 3 4 5 6 7 Garissa Isiolo Mandera Nairobi Wajir Average Agreement with Statement (1-10) "The county government made effort to improve peace" Midline Endline 35 Previous Wasafiri CAP studies recommended that future CAP processes, especially those not currently regarded as VE ‘hotspots’, should find ways to build political buy-in towards CVE through other means. Crucially, they noted that NCTC and other partners should reframe discussions around VE from focusing primarily on VE attacks to VE recruitment. Stakeholders suggested that the 14 Riverside attack be referenced as an example, where a large number of the perpetrators originated from non-hotspot counties. 3.7 Cross-cutting outcomes and lessons learned The endline assessment explored a number of outcomes that cut across NIWETU’s programming streams, such as programming modalities and thematic considerations, which contribute to good practice and should inform future approaches among the community of practice. These outcomes include the effectiveness of the activity’s complexity aware planning cycle, the results of gender mainstreaming, and CVE capacity building for legacy partners. 3.7.1 Gender prioritized across portfolio NIWETU’s gender strategy served as a deliberate mechanism to promote the integration of gender considerations throughout NIWETU’s activity design and delivery processes, in addition to being emphasized through partners at the activity level.  NIWETU gender strategy serves as model for future CVE interventions – The activity took deliberate steps to focus on gender through its programming, including issuing calls for proposals for CVE activities specifically addressing gender issues. The modalities used in implementing these initiatives included convening separate meetings for women (as had been done in Isiolo), and separate group sessions for women during activities (as had been done in Garissa), and targeting women’s peer networks, where women feel safer opening up.  Women encouraged to lead and facilitate initiatives – Within individual activities, CSO partners reached out to other women in the community about the best approaches to design and implementation, including the use of gender scorecards. Partners also encouraged women to lead on activity implementation, through facilitating a training session using the gender-conscious CVE facilitation guide, or direct management of grants. Similarly, open consultation spaces, or ‘field desks’ were created where women and youth could come to share issues of harassment by the security forces, which would then be taken up by relevant authorities.  Awareness activities lead to more equitable parenting – Some NIWETU supported activities, such as the Isiolo Women of Faith grant and a component of an IPL activity, specifically focused on reaching families affected by VE, through training, counseling and strengthening peer support. Respondents believed the activity had a ripple effect on the broader community, where both men and women took a more active role in raising their children. While significant social and cultural barriers to equitable gender participation persist, such as those described in previous sections, the inclusive approach adopted by NIWETU represents a strong and replicable framework for future activities of this nature. 3.7.2 Legacy partner CVE capacity significantly strengthened Stakeholders across NIWETU’s portfolio highlighted the strong CVE capacity of its longer-term CSO partners in target communities. While feedback varied somewhat according to partner, there was general consensus that the activity’s proactive and collaborative management approach had contributed to significant gains in CVE knowledge and skills among its legacy implementers. There was further evidence that partners had cultivated constructive and sustainable relationships with key CVE actors, including national and county-level authorities, community leaders, and downstream civil society representatives, including: 36  The Malaika Foundation effectively facilitated CAP processes – The organization received positive feedback forfor its technical advisory and broader support during the formation of the CAPs, following the NSCVE and General CAP Guidelines. Respondents across all counties mentioned them as an important partner in the formation and implementation of respective CAPs. NIWETU consequently invested in Malaika as the lead training provider for the development of the nation-wide CVE toolkit for Kenya. Similarly, there was agreement that IPL had done a very good job of convening the appropriate stakeholders and ensuring the momentum of the process, with several noting that IPL had actually initiated a CAP formation process directly.  Partner capacity evolved significantly during implementation – For instance, ROAD in Garissa, was not noted as being as much of a driving force during the CAP process, while SUPKEM initially was cited as having more influence in many of the sessions, according to research respondents. ROAD took a more active role as the process unfolded, due to more consistent guidance and feedback from participants. According to NIWETU-supported research, in the early stages of CAP formulation in Mombasa, HAKI Africa was perceived negatively by many stakeholders, due to poor relationships with government counterparts. They were, however, ultimately considered instrumental in the process after having taken concrete steps to improve theirtheir rapport with local officials. Stakeholders expect that such improvements in partner capacity will contribute to their journey to self-reliance and an established cadre of experienced, grass-roots CVE implementers in Kenya going forward. 3.7.3 The journey to self-reliance requires sustained donor backing NIWETU alignedT its J2SR strategy with the USAID Kenya roadmap, in addition to specific objectives for each intervention. These objectives were intended to promote the sustainability and longer-term national ownership of key activity activities. Key strategy elements, such as delegating CAPs to county governments, empowering C4Cs to develop independent action plans, and supporting the creation of the Security Management Institute, represent deliberate steps to feed into broader J2SR aims. Processes requiring sustained support include:  Expansion of CVE coordination – NCTC has held donor coordination meetings for some time, but recently (at time of drafting) initiated meetings to improve coordination and cohesion between CVE implementing partners. This included the development of a matrix highlighting where each IP was working, which local partners they are collaborating with, and how their activities are supporting CAP pillar implementation. This matrix will help both NCTC and the IPs identify current gaps in CVE programming. NCTC encouraged similar matrix efforts at the county level to support CAP implementation.  Engage MCAs in CVE resourcing – County-level officials pointed to the need for greater involvement of MCAs in the development and implementation of their county’s CAP to promote greater sustainability. They said this is key given that MCAs are responsible for approving or rejecting Annual Development Plan and Country Fiscal Strategy PaperPaper provisions for CAP funding. They also have constituencies in their home wards and are thus able to reach out to communities and gain their support and participation in CVE activities. In counties where the more J2SR for CVE CSOs NIWETU engaged Hedayah, an internationally recognised CVE think tank, in partnership with the Malaika Foundation, to create a nationally relevant, locally tailored CVE toolkit. The toolkit built on a needs assessment to inform the development of tools appropriate to the Kenyan context. The Malaika Foundation committed to develop its own training curriculum, independent of NIWETU, to improve the CVE capacity of civil society. This helped them develop CVE specific theories of change and implementation plans. The toolkit is expected to provide a capacity building roadmap for partners that will extend beyond the activity. 37 senior county government leadership is less active or interested in CAP implementation, approaching the MCAs for support is a good alternative.  Deepening collaboration across CVE implementers – Closer to the end of the activity, NIWETU noted in a quarterly progress report that it deepened relationships with other USG￾funded activities in Kenya through Partnership for Resilience and Economic Growth (PREG) meetings and informal information sharing and collaboration. In addition, NIWETU has led coordination meetings on a semi-regular basis across implementers in the CVE space in Kenya in order to align activities and share emerging lessons. NIWETU also continued to share its emerging good practices with USAID CVE activities in other regions. For example, NIWETU shared its CVE Facilitation Guide with USAID’s Central Asia Support for Stable Societies (CASSS), which will adapt the guide for the Central Asia CVE context Additionally, NIWETU held a virtual cross￾program learning event with staff from Tuko Pamoja, a USAID-supported CVE activity based in Mozambique. Staff from Tuko Pamoja and DAI’s Center for Secure and Stable States (CS3) requested the session to learn from NIWETU’s experience as they embark on a series of CVE activities in the coming months. Much of Tuko Pamoja’s upcoming portfolio mirrors NIWETU’s work in Kenya, including supporting a cadre of C4Cs, developing strategic communications materials, and working closely with regional government to implement context-specific CVE activities. Following the session, Tuko Pamoja staff requested one-on-one conversations with members of the NIWETU team to learn more about NIWETU’s successes and challenges. 4. CONCLUSIONS AND RECOMMENDATIONS In light of the findings, lessons learned, and recommendations presented above, it is evident that NIWETU had significant impact across its primary objectives and corresponding results areas. Its contribution to CVE in Kenya manifested in numerous ways, particularly with regard to building resilience to VE among its key government, community, and civil society stakeholders. Within an often￾challenging context, the activity team piloted a variety of innovative and novel approaches that established a model for the community of practice, and for future interventions of this nature. The immediate results can be seen in better, more tailored responses to VE at the national and county￾levels, particularly among its target communities. This synthesis document presented the higher-level observations and findings gleaned over the course of NIWETU’s implementation, through a large volume of research and evaluation activities. It is not, however, an exhaustive synthesis and the annexed county reports and case studies should be consulted for a deeper understanding of specific outcomes and programming examples. A summary of conclusions and overarching recommendations are presented below:  Systems-based approaches and adaptive management critical to building sustainable CVE networks and capacity – The study found that the systems-based approach employed by NIWETU was instrumental in strengthening capabilities and connections with multiple institutions and key stakeholders working on CVE issues at local, county and national levels. The ‘whole-of-society’ principle in engaging with government, civil society, community members, and key influencers concurrently, to promote coordination and collaboration on VE issues, was crucial to delivering the results achieved by the activity. Similarly, the adaptive management process embodied in the activity’s complexity aware planning process was found to be critical in enabling course-correction in response to rapidly evolving contexts in target areas. The complexity-aware planning process further demonstrated the catalytic impacts of tailoring interventions, according to a robust thematic and technical evidence base.  Enhanced VE knowledge and skills foundation for further dialogue – NIWETU fostered significant increases in knowledge and skills among government, civil society, and 38 community stakeholders as a result of its programming. This enhanced CVE capacity provided the starting point for improvements in psychosocial support among populations affected by trauma, in addition to opening the space for greater dialogue about VE issues. This study found that dialogue between key actors gained momentum over the life of the activity and will likely continue to build as a result of strengthened relationships. Future interventions should leverage this boosted CVE capacity and support protracted dialogue activities to identify more effective responses going forward.  CVE networks underpin resilience and longevity – A number of CVE networks were either created or strengthened by NIWETU, including youth groups, women’s groups, boda boda collectives, and the Champions for Change. There were indications that these networks were sustainable from independent initiatives taken by their participants to respond to VE in their communities. The networks contributed directly to community-led CVE actions that, by nature, were more holistic and inclusive than siloed interventions without broad participation from a diverse cross-section of society. Future activities should build on these latent and more formalizedz groups, to drive locally tailored solutions to VE throughout affected communities. Similarly, implementers should seek to build wider and stronger linkages between key civil society, government, and community stakeholders to tackle greater challenges.  Action plans are an effective entry point for future initiatives – A variety of action plans were developed as a result of NIWETU activities. These include the CAPs and RCAPs which led to the formation of active CEFs in most target areas, and concrete implementation roadmaps for CVE at the local level. Additionally, Champions for Change developed personal action plans as part of their engagement with the activity, leading to the creation of local CBOs and a national-level coalition. In many cases these plans extended beyond the end of NIWETU and provide a concrete entry point for future interventions to leverage and build upon.  J2SR for CVE requires sustained coordination and vision – Given the relatively recent emergence of the CVE sector in Kenya, it requires sustained support and cooperation in order to achieve longevity in the current context. The opaqueness of the USG longer-term strategy for CVE in the region, along with unpredictabilityunpredictability of government resourcing of CVE at the national and county-levels, mean that key stakeholders are unsure how initiatives will be supported going forward. To maintain the positive momentum driven by NIWETU and its partners, clear and coordinated leadership is required from national and international counterparts. Such efforts should build on the collaborative engagements facilitated by NCTC nearer the end of the activity to foster more effective programming. 39 5. ANNEX I – METHODOLOGY Research Approach This research utilized a combination of quantitative and qualitative research tools and analysis of secondary data, including reports produced through NIWETU’s CAP Cycle Framework. The stakeholder survey was a replication of the baseline and midline surveys that aimed to gather up-to￾date information and identify changes in the local VE and CVE context over the past year, as they relate to NIWETU’s conceptual framework. The midline evaluation’s household survey that gathered information on sentinel indicators (discussed in further detail below) was also replicated for this endline. Qualitative tools were designed to gather more detailed and in-depth information to provide greater insight, context, and nuance to the findings, as well as gather data specifically for the case studies. The methods that this research employed are listed below, and will be expanded upon further in the following sections: 1. Case studies 2. Stakeholder surveys (for baseline/midline/endline) 3. Sentinel indicators 1. Case studies Case studies of some of NIWETU’s key achievements were written for this endline evaluation. They focused on key outcomes related to NIWETU’s activities in the following areas: (1) Champions for Change, (2) County Action Plan development, (3), Kenya School of Government, and (4) NIWETU’s intervention in Mandera County. Information for the case studies was gathered via key informant interviews at the county level, as well as through interviews with NIWETU team members and other Nairobi-based stakeholders. Findings from previous quarterly research reports written under NIWETU’s CAP Cycle Framework also fed into the case study analysis. The case studies are annexed in this report, and findings fed into the key outcomes and lessons learned presented in the main body of the report. 2. Stakeholder surveys (for baseline, midline and endline) This questionnaire was a replication of the stakeholder surveys employed during the baseline and midline evaluations. Wasafiri refined some questions from the baseline evaluation during the midline in order to clarify language, remove questions that did not yield relevant information, and add additional questions to better assess respondents’ engagement over the reporting periods. No changes were made to this tool for the endline evaluation. Findings from the stakeholder survey comprise the main source of data in the county reports. The county reports are annexed in this report, and findings fed into the key outcomes and lessons learned presented in the main body of the report. 3. Sentinel indicators This research utilized sentinel indicators to understand how change can happen at the objective level. The indicators were designed during the midline evaluation in order to assess the wider context that affects change at the objective level. Specifically, the indicators looked at other factors in the context that help to assess communities’ tendency to mobilize around peace and security issues through existing structures, to understand whether similar or the same structures could be utilized to support mobilization around CVE, which is NIWETU Objective 1. They also look at government responsiveness to VE through the presence of legislation and funding for CVE initiatives, and 40 perspectives on the relationship between government (including the security sector) and communities to support analysis around the Objective 2 level. The research approach used household surveys to gather data against sentinel indicators, and to derive the scores for each sentinel indicator. The household survey utilized a meta-perceptions tool to gather individual perspectives and assess wider social norms. Assessing social norms gives a better indication of behavioral or perspective change that can be expected. Perceptions of social norms essentially guide individuals’ own perceptions and behavior, because of perceived risks from deviating from that social norm. In this context, the survey gathered meta-perceptions to construct a clearer picture of the relevance of peace and security structures, and attitudes toward government in the research counties.11 The survey also collected individual perceptions via key informant interviews to identify improving or worsening trends: if the social norm is more positive than the individuals’ views, it indicates potentially positive momentum, as people will tend to act or think according to the norm. If the social norm is more negative than the individuals’ views, it indicates worsening perspectives on the issue in question. Scores derived from the household survey data are presented on a scale of zero to ten; scores closer to ten indicate agreement with the statement, and scores closer to zero indicate disagreement with the statement. Further details on meta-perceptions and the development of the household survey are offered in Annex II. Wasafiri developed two sentinel indicators that align with each of NIWETU’s objectives. Each indicator includes sub-indicators to more clearly specify what will be measured within each sentinel indicator. Objective 1: Community mobilization to address violent extremism enhanced Sentinel Indicator 1: Level of community mobilization around peace and security issues 1. Degree to which peace committees are active and responsive 2. Degree to which communities actively engage in community barazas 3. Degree to which Nyumba Kumi initiatives are active 4. Type of community networks that exist 5. Presence of other networks, organizations, or actors that work on peace and security within the community. Objective 2: Government responsiveness to violent extremism improved Sentinel Indicator 2: Degree to which the government engages with and supports the community on issues of peace and security 1. Level of government and community participation in community-government dialogues 2. Presence and size of line items for CVE activities in county budgets 3. Presence of legislation (national and county level) to support the implementation of County Action Plans 4. Level of trust in the security sector. Each sentinel indicator was given a score based on an aggregation of the findings from each sub￾indicator. Scores for each sub-indicator were determined through the findings from relevant questions in the quantitative surveys or led solely by the document review (for questions on budgets and legislation). Once questions are developed for each indicator, clear scoring criteria will be articulated. An average of the sub-indicator scores within each sentinel indicator will determine the overall score of that particular sentinel indicator. 11 The meta-perceptions approach asks individuals to first think about their own perspective on a question, and then ask how their peers would answer the same question. Individual perceptions are derived from the questions asking about their own perspective, and meta-perceptions come from answers to questions about peers’ views. These perspectives are averaged to identify individual trends and social norms. 41 Analytical Framework Wasafiri developed an analytical framework to guide the analysis at the goal, objective and result area levels for this report. The analytical framework summarizes the key questions that this evaluation sought to answer, and the data sources that fed into the analysis. 42 Results Areas Lines of Inquiry Key Questions Data Streams/Evaluation Products Additional Tools to Fill Gaps in Knowledge How relevant has NIWETU's strategy been in coherence with national policy agendas? How has project strategy informed national objectives? How effective has coordination been with national and sub-national actors? What have been the key challenges and opportunities? To what extent is NIWETU's approach replicable, scalable, and sustainable? How is its approach being received and expanded? How is NIWETU perceived within the broader CVE architecture in Kenya? What are observations from senior stakeholders? To what extent did NIWETU's approach adapt to the evolving VE context in target areas? How did this manifest in program processes? How well were research and learning processes structured to enable iterative programming? What were levels of dissemination and uptake? To what extent was the complexity-aware planning cycle appropriate for the context? How did it influence learning and decision-making? How did program intervention logic evolve over the course of implementation? How did this impact ToCs and evaluation methods? How has overall capacity to mitigate VE threats changed during implemenation? What factors contributed to these changes? What are the general perceptions of security and VE threats in target communities? How does pereception differ from reality? How do perceptions of CVE capacity differ between target and non-target areas in similar locations? What contribution can be inferred? How effective have sentinel indicators been in tracking overall VE threats and stability? What lessons have been learned for forumulation? Have there been changes in the methods, tools, and frequency of communication about VE issues? What role has ICT played? What changes have been observed in levels of national vs sub-national dialogue on VE issues? How has this affected community actions? Which agencies/groups coordinate most effectively? To what extent has this been leveraged? What has been learned? Have there been any negative results of info-sharing, such as rumors and hatespeech? How has this been facilitated and/or mitigated? Have there been any changes in actions by government and security agencies operating in target areas? Examples? Are there actions implemented jointly by communities and the sub-county administration? How has this changed during implementation? How have levels of VE knowledge and skills changed among security actors? Where are there still gaps? Have there been any changes in local authorities' perceptions and awarnes of VE challenges and community capacity in target areas? Why? Increased Government Capacity Effectiveness of inter-agency CVE coordination and info-sharing CVE knowledge and effectiveness of County-level administrators and security providers •Research: CAP development and implementation •Case Study: Support to the Kenya School of Government (KSG) •Case Study: Support to the formation of CAPs/RCAPs •OH: Skills and Knowledge - Majengo •OH: Kamukunji Interventions •Case Study: Support to the formation of CAPs/RCAPs •Sentinel Indicators: Government responsiveness •Validation/triangulation of case study findings •KIIs w/ county government representatives on the CEF •KIIs w/ national CVE reps •Desk review •KIIs w/ MOI, NCTC •KIIs w/ CVE implementers in Kenya •KIIs w/ Steering Committee •Desk review •KIIs w/ MOI, NCTC •KIIs w/ County Administrators •KIIs w/ Steering Committee •Validation/triangulation of case study findings •NIWETU semi-annual reports •Government strategic plans (if available) •Steering Committee feedback •Case Study: Systems strengthening in Mandera County •OH: Gender and inclusion •OH: Skills and Knowledge - Majengo •Case Study: Champions for Change (C4C) •Case Study: Systems strengthening in Mandera County Effectiveness of systems-based CVE approaches Contribution to greater VE resilience in target communities Improved CVE Capabilities to Identify and Respond to VE Threats •Complexity-aware Planning Cycle •NIWETU MEL documents •NIWETU quarterly reports •Steering Committee feedback Effectiveness of adaptive and complexity-aware approaches •Analysis of NIWETU CAP Cycle •Review quarterly reports •KIIs w/ SMT + decision makers 43 Results Areas Lines of Inquiry Key Questions Data Streams/Evaluation Products Additional Tools to Fill Gaps in Knowledge How relevant has NIWETU's strategy been in coherence with national policy agendas? How has project strategy informed national objectives? How effective has coordination been with national and sub-national actors? What have been the key challenges and opportunities? To what extent is NIWETU's approach replicable, scalable, and sustainable? How is its approach being received and expanded? How is NIWETU perceived within the broader CVE architecture in Kenya? What are observations from senior stakeholders? To what extent did NIWETU's approach adapt to the evolving VE context in target areas? How did this manifest in program processes? How well were research and learning processes structured to enable iterative programming? What were levels of dissemination and uptake? To what extent was the complexity-aware planning cycle appropriate for the context? How did it influence learning and decision-making? How did program intervention logic evolve over the course of implementation? How did this impact ToCs and evaluation methods? How has overall capacity to mitigate VE threats changed during implemenation? What factors contributed to these changes? What are the general perceptions of security and VE threats in target communities? How does pereception differ from reality? How do perceptions of CVE capacity differ between target and non-target areas in similar locations? What contribution can be inferred? How effective have sentinel indicators been in tracking overall VE threats and stability? What lessons have been learned for forumulation? Have there been changes in the methods, tools, and frequency of communication about VE issues? What role has ICT played? What changes have been observed in levels of national vs sub-national dialogue on VE issues? How has this affected community actions? Which agencies/groups coordinate most effectively? To what extent has this been leveraged? What has been learned? Have there been any negative results of info-sharing, such as rumors and hatespeech? How has this been facilitated and/or mitigated? Have there been any changes in actions by government and security agencies operating in target areas? Examples? Are there actions implemented jointly by communities and the sub-county administration? How has this changed during implementation? How have levels of VE knowledge and skills changed among security actors? Where are there still gaps? Have there been any changes in local authorities' perceptions and awarnes of VE challenges and community capacity in target areas? Why? Increased Government Capacity Effectiveness of inter-agency CVE coordination and info-sharing CVE knowledge and effectiveness of County-level administrators and security providers •Research: CAP development and implementation •Case Study: Support to the Kenya School of Government (KSG) •Case Study: Support to the formation of CAPs/RCAPs •OH: Skills and Knowledge - Majengo •OH: Kamukunji Interventions •Case Study: Support to the formation of CAPs/RCAPs •Sentinel Indicators: Government responsiveness •Validation/triangulation of case study findings •KIIs w/ county government representatives on the CEF •KIIs w/ national CVE reps •Desk review •KIIs w/ MOI, NCTC •KIIs w/ CVE implementers in Kenya •KIIs w/ Steering Committee •Desk review •KIIs w/ MOI, NCTC •KIIs w/ County Administrators •KIIs w/ Steering Committee •Validation/triangulation of case study findings •NIWETU semi-annual reports •Government strategic plans (if available) •Steering Committee feedback •Case Study: Systems strengthening in Mandera County •OH: Gender and inclusion •OH: Skills and Knowledge - Majengo •Case Study: Champions for Change (C4C) •Case Study: Systems strengthening in Mandera County Effectiveness of systems-based CVE approaches Contribution to greater VE resilience in target communities Improved CVE Capabilities to Identify and Respond to VE Threats •Complexity-aware Planning Cycle •NIWETU MEL documents •NIWETU quarterly reports •Steering Committee feedback Effectiveness of adaptive and complexity-aware approaches •Analysis of NIWETU CAP Cycle •Review quarterly reports •KIIs w/ SMT + decision makers 44 What formal and informal networks emerged? How do communication and organization processes differ? How were they leveraged? Are there indications of changes in levels of trust between different com munity groups? What effect has this had on coordination? How has NIWETU approached support to established and newly created CVE networks differently? How did effectiveness vary? How is membership defined formally/informally? To what extent are marginalized groups represented? How are existing CVE networks regarded by key stakeholders in target com munities? What is their perceived level of influence? What role do CVE networks plan in obtaining, fact-checking, and disseminating information regarding VE activity in com munities? To what extent are CVE networks contributing to the forumulation of CAPs in target areas? How do they shape the consultation process? What actions have CVE networks taken? What is the effectiveness of peace committees? Which actors do they typically engage? Have changes been observed in the level of local government support for com munity action plans? If so, what exa mples are there? How has the level of local authorities engagement with, and perceptions of, trauma healing changed? Do they find it useful? Why? How are local authorities involved in harmonizing action plans across target areas? How are these linked to national VE objectives? Have there been any changes in local authorities' levels of trust in community stakeholders as a result of coordination? If so, how? How involved have communities been in developing local VE action plans? How has this changed over the course of the project? To what extent have partners been successful in engaging and including marginalized groups? Are there unreached groups? Why? To what extent to CAPs in target communities align with broader national / sub-national VE priorities? How do they differ? Which partner activities were most sucessful? Which were the least successful? What lessons were learned from this experience? How has the level of knowledge of VE issues changed among target com munities? How do levels of kno wledge vary by stakeholder? To what extent have NIWETU strategic communications activities contributed to icreased levels of VE awareness in target areas? How have communities' abilities to identify, fact-check, and disseminate credible sources of VE information changed? What gaps in kno wledge and undersatnding of VE issues still exist? Which gaps are most critical to building community resilience? What links can be identified between increased understanding of VE issues and changes in CVE skills and behaviors? How willing are communities to discuss issues of VE, crime, and security? How has this changed since project inception? How have the nature and scope of CAPs changed in target areas? What contribution have NIWETU activities had in observed changes? How have the kno wledge, influence, and practices of com munity CVE champions changed over the course of the project? Examples? Improved CVE Skills and Knowledge Knowledge and awarenes off CVE concepts and issues •Research: CAP development and implementation •Case Study: Champions for Change (C4C) •Case Study: Systems strengthening in Mandera County •FGDs w/ community •FGDs w/ CSOs •KIIs w/ religious leaders •Validation/triangulation of case study findings Level and effectiveness of CVE skills and behaviors •Case Study: Champions for Change (C4C) •Case Study: Systems strengthening in Mandera County •OH: Gender and inclusion •OH: Skills and Knowledge - Majengo •FGDs w/ community •FGDs w/ CSOs •KIIs w/ religious leaders •Validation/triangulation of case study findings Community-led CVE Strategies and Initiatives Existence and formulation of community-led CVE strategies •Case Study: Support to the formation of CAPs/RCAPs •Case Study: Champions for Change (C4C) •Research: CAP development and implementation •FGDs w/ community •FGDs w/ CSOs •KIIs w/ religious leaders •Validation/triangulation of case study findings Implementation and effectiveness of CVE initiatives •OH: Gender and inclusion •OH: Skills and Knowledge - Majengo •OH: Kamukunji Interventions •Sentinel Indicators: Community mobilization •FGDs w/ community •FGDs w/ CSOs •KIIs w/ religious leaders •Validation/triangulation of case study findings Existence and formation of CVE networks Effectiveness and sustainability of CVE networks CVE Networks Established •FGDs w/ community •FGDs w/ CSOs •FGDs w/ peace committees •KIIs w/ religious leaders •KIIs w/ grantees •FGDs w/ community •FGDs w/ CSOs •KIIs w/ religious leaders •Validation/triangulation of case study and prior research findings •Case Study: Support to the formation of CAPs/RCAPs •OH: Skills and Knowledge - Majengo •OH: Kamukunji Interventions •Sentinel Indicators: Community mobilization •Case Study: Support to the formation of CAPs/RCAPs •OH: Skills and Knowledge - Majengo •OH: Kamukunji Interventions •Sentinel Indicators: Community mobilization 45 Sampling The following section outlines the sample sizes for stakeholder survey interviews in each county. Stakeholder Survey Garissa Isiolo Mandera Nairobi Wajir Total Community-based organization 7 5 8 6 11 37 Community leader 4 5 7 8 4 28 County administration 7 5 6 0 4 22 Educational professionals 5 7 5 1 6 24 Media 4 5 3 3 7 22 Private sector 8 4 4 7 2 25 Security sector 2 7 5 3 10 27 Women’s groups 5 6 6 10 1 28 Youth representatives 6 6 6 5 4 27 Total 48 50 50 43 49 240 Household surveys reached 40 individuals in each of the five counties for a total of 200 respondents. Gender breakdown of the survey respondents by category is as follows: Garissa Isiolo Mandera Nairobi Wajir Female 15 18 12 23 21 Male 25 22 28 17 19 The majority of respondents were also Muslim, and respondents held varying levels of schooling. Respondents all came from urban villages within the main town in each county. No household surveys were conducted in rural areas of the county. The sample was created utilizing the random walk method. Researchers started from a central starting point, or landmark, in the village (a school, watering hole, mosque/church, for example) and walked in different directions. They stopped at every third house and asked to carry out an interview with the individual who answered the door, provided they were over the age of 18.12 If no one of appropriate age was home, the researchers went to the next door to conduct the survey. Result Area Scoring Criteria The below section discusses the scoring approach for each result area. The questions that factor into the scoring and the scale for each section will be outlined below. Scoring Approach Scores for each section were created by taking an average of the key selected questions. The scale for each section was based on a four- or five-point scale. Answers that demonstrated positive findings were given a higher score and answer that demonstrated negative answers were given a lower score. On each scale, the breakdown is as follows:  Four-point scale: o 1-2: negative answers o 3-4: positive answers 12 As noted in the limitations section, in some of the more traditional areas, the elder of the household would be summoned for the survey, slightly skewing the age range of respondents upward. 46  Five-point scale: o 1-2: negative answers o 3: neutral/no change o 4-5: positive answers Answers from respondents in each county were averaged to obtain the average score for each selected question. Each average was then averaged into an overall score for each result area to demonstrate the current status of each area. It is expected that these scores will change as they continue to be tracked over time, though adjustments might be minimal. The coding scheme used breaks down as follows: Four-point scale: 1.00-1.50 1.51-2.50 2.51-3.50 3.51-4.00 Five-point scale: 1.00-1.50 1.51-2.50 2.51-3.50 3.51-4.50 4.51-5.00 questions that factor into the scoring for each section, and their corresponding values, are as follows: 1. County VE Context Extremism and recruitment to join extremist groups is less of a problem now than it was 12 months ago. Strongly agree 5 Somewhat agree 4 No change 3 Somewhat disagree 2 Strongly disagree 1 Organized crime and gang activity is less of a problem now than it was 12 months ago. Strongly agree 5 Somewhat agree 4 No change 3 Somewhat disagree 2 Strongly disagree 1 Community members feel safer now than they did 12 months ago. Strongly agree 5 Somewhat agree 4 No change 3 Somewhat disagree 2 Strongly disagree 1 47 Over the past 12 months, the number of terrorist attacks in this county has decreased. Strongly agree 5 Somewhat agree 4 No change 3 Somewhat disagree 2 Strongly disagree 1 Over the past 12 months, the number of terrorist attacks throughout Kenya has decreased. Strongly agree 5 Somewhat agree 4 No change 3 Somewhat disagree 2 Strongly disagree 1 2. Skills and Knowledge Have you participated in any trainings on preventing violent extremism or understanding the causes of violent extremism in the last 12 months? Yes 4 No 1 How much, if at all, do you think these trainings have helped your understanding of violent extremism and how to prevent it? A lot 4 A little 3 Not very much 2 Not at all 1 How much, if at all, do you use the information you have learned in these trainings? A lot 4 A little 3 Not very much 2 Not at all 1 Have you shared the information you have learned with others in the community? Yes 4 No 1 3. Community Networks Do you work with any other organizations on CVE? Yes 5 No 1 48 Do you collaborate more, less or the same amount with these actors than you did 12 months ago? A lot more 5 A little more 4 The same amount 3 A little less 2 A lot less 1 How often are there conflicts or disagreements between organizations working on CVE? Very often 1 Somewhat often 2 Occasionally 3 Rarely 4 Never 5 Overall, do you think improved networks between individuals and organizations working to prevent violent extremism has been very helpful, somewhat helpful, or not at all helpful? Very helpful 5 Somewhat helpful 3 Not at all helpful 1 4. Community-led Strategies Have local organizations worked with other stakeholders to develop programs on CVE in the last 12 months? Yes 5 No 1 How effective do you think these methods are in teaching others about violent extremism, and countering violent extremism? Are community-based organizations more or less active in working to prevent violent extremism than 12 months ago? Much more active 5 Only a bit more active 4 About the same amount 3 A little less active 2 Much less active 1 Very effective 5 Somewhat effective 4 Sort of effective 3 Not very effective 2 Not at all effective 1 49 To what extent do community members engage with these organizations? Engage a lot 5 Engage a little 3 Do not engage 1 5. Government Responsiveness to CVE How well do you think the county government understands the problem of violent extremism in this community? Very well 4 Somewhat well 3 Not very well 2 Not at all 1 How well do you think the national government understands the problem of violent extremism in this community? Very well 4 Somewhat well 3 Not very well 2 Not at all 1 How willing do you think county government is to work on programs to prevent violent extremism? Very willing 4 Somewhat willing 3 Not very willing 2 Not at all willing 1 How willing do you think national government is to work on programs to prevent violent extremism? Very willing 4 Somewhat willing 3 Not very willing 2 Not at all willing 1 How easy do you think it is to work with national government counterparts (e.g. Assistant Chief, Chief, Deputy County Commissioner, or County Commissioner) Very easy 4 Sort of easy 3 Sort of difficult 2 Very difficult 1 50 51 6. ANNEX II – GARISSA COUNTY REPORT: GARISSA COUNTY ENDLINE OVERVIEW Data collection took place in Garissa Town over 12 days in February to March 2020, carried out by two researchers. NIWETU facilitated introductory meetings to key stakeholders to explain the goals of the research and to facilitate further introductions to relevant actors. The research team conducted both qualitative and quantitative research with key stakeholders in Garissa, as well as the wider community. The breakdown of respondents in the stakeholder questionnaire is as follows: Category of respondent Number of respondents County administration 7 Youth representatives 6 Educational professionals 5 Women’s groups 5 Community-based organization 7 Private sector 8 Security sector 2 Community / Religious leader 4 Media 4 Total 49 Household surveys were conducted with 40 individuals around Garissa town selected through the random walk method. Survey respondents were equally split between men and women. Ten of those respondents have had no schooling at all, while 14 have at least a college diploma. Key informant interviews were carried out with eight individuals. 52 7. KEY FINDINGS 7.1 Section 1 – The VE Context 7.1.1 Summary of Findings Perceptions of the VE context in Garissa are somewhat negative, particularly in comparison to findings at the time of both baseline and midline. This is mostly in respect to the frequency of extremist activities where perceptions are largely influenced by the current context. The study was conducted at the peak of terror attacks in the county. Disappearances continue to occur frequently, as youth who disappear are believed to either have joined AS or been arrested or killed by security agents when suspected to have engaged with VE groups. Gang activity and organized crime is still not as big a threat in Garissa County, although there is a mention of active youth gangs known as ‘’Squad Chafu.’’ Although there seems to be an agreement that crimes such as murder and rape are reported from time to time, there still is not sufficient evidence to suggest that such crimes are organized or are a result of gang activities. As was indicated in the midline, Garissa continues to face threats from inter￾clan conflicts. 7.1.2 Causes of Insecurity Predictably, nearly three-quarters of respondents noted that the threat posed by violent extremist groups was one of the most significant security threats in Garissa. Other significant threats mentioned were land disputes, youth disappearances and high levels of unemployment and inter-ethnic tension were reported by 31% respectively. Security threats Baseline Midline Endline Violent extremist groups 92% 73% 88% Land disputes 26% 18% 43% Youth disappearances 21% 24% 37% Inter-ethnic tension 57% 36% 31% High levels of unemployment 64% 62% 31% Radicalization 4% 16% 27% Drug abuse - 40% 25% High levels of poverty - 24% 10% Lack of education - 6% 14% Local politics 36% 44% 8% Theft 9% 15% 8% Criminal gangs 2% 9% 4% Inter-religious tension 2% 0% 0% National politics 2% 15% 0% Other 26% 0% 0% Key informants all acknowledged the significant threat posed by violent extremist organizations in Garissa County. This threat was more prevalent between January and March 2020, where Garissa and neighboring counties witnessed a spike in terror attacks. There was an estimate of ten AS-related attacks in Garissa during this period, ranging from abductions, torching homes, attacking key government and security installations and police stations, and killings that mostly targeted non-local teachers and security officers. A key informant notes that this was an ‘’awakening call for almost everyone in the CEF and those working on CVE in the region.” Further fueling this growing sense of insecurity was a press release from AS leadership that called for a fatwa on all non-locals living in the North-Eastern region, which caused widespread fear and led to an exodus of teachers causing an educational crisis. Still, local communities are targeted by AS activities as well. In one incident in Saretho village, AS assailants killed four children in a nearby primary school. In another incident, assailants torched homes, blaming area residents for working with security agents, particularly NPR. 53 As it was expressed during the midline, informants noted that the community fears discussing VE or reporting suspicious activity openly, out of fear of harassment and intimidation by the security sector, and also by VEOs themselves. One informant from a local CBO stated: “Need to improve confidentiality of the information that the community shares with the security as there are several cases where some community members shared intel on security threats and their identity were leaked to the VEOs and later become a target to AS. The government need to clean their inhouse as there are several moles within the security sector who work with the VEOs.” 7.1.3 Disappearances Respondents with at least some degree of knowledge on the rate of disappearances noted that this phenomenon occurs with at least some frequency. Only a women’s group representative, during the midline, indicated that youth disappearances never occur. During the endline, several respondents (31%) noted that such disappearances occur very regularly, indicating an increase in comparison to both baseline and midline. Only 10% of respondents, a journalist, two members of a women group and one youth representative, believed that young people did not disappear in the county. How often do youth in Garissa County disappear? Baseline Midline Endline Very regularly 21% 13% 31% Somewhat regularly 23% 22% 16% Rarely 32% 24% 27% Never 17% 1% 10% Do not know 6% 40% 16% As in the midline, most respondents indicated that youth who disappear either go to Somalia or disappear at the hands of the security agents. Close to half of the respondents indicated that youth go missing regularly. Over half of these respondents stated that those who disappear are believed to have either unlawful groups like Al-Shabaab or Islamic State (ISIS), or are taken by the security agencies (specifically the ATPU) if suspected of being members or sympathizers of terror groups. Others did not know where youth who disappear go. “Nowhere to be found, they have been taken out of their homes by unknown people.” – Educational Institution Representative. A female youth representative was quoted saying that “young people join Al-Shabaab for employment”, while another noted, “Vulnerability index of youth at the sub-county level has increased due to lack of services like education and health.” This may indicate that, the perception that youth are driven by economic reasons to join such groups is still prevalent, despite increased awareness programs that demystify this. A county administrator stated that the youth are recruited into extremism, especially in areas where the government and security actors are less present. 7.1.4 VE and Recruitment Respondents tended to believe that the VE context concerning recruitment in Garissa has improved over the past 12 months. 43% tend to agree that it is now comparatively less of a problem. Notably, also is the finding that 8% of the respondents did not have an opinion on the context compared to nearly 28% during the midline. 54 Extremism and recruitment to join extremist groups is less of a problem now than it was 12 months ago. Midline Endline Strongly agree 22% 10% Somewhat agree 7% 33% No change in context 9% 12% Somewhat disagree 15% 10% Strongly disagree 19% 27% Do not know 28% 8% Still, more respondents held negative views on this issue. Some felt that the tensions due to the escalation in attacks between December 2019 and March 2020 were an indication that recruitment is still taking place. Nonetheless, the number of those who strongly believed extremism and recruitment was more of a problem now than it was 12 months ago was slightly less than the midline, indicating that the seemingly escalating VE activities at the time did not overshadow their perspective on the context when reflecting over a longer period. A media representative quotes that: “The number of attacks has increased so I think recruitment has increased. The government seems less concerned.” Another youth stated that recruitment is still happening, because there have been reports in the community. He quotes: “We still hear stories on recruitment still happening in the community.” When asked why they believe that recruitment remains prevalent, respondents mentioned that the government measures to curb recruitment have not been very effective. Perhaps this perceived ineffectiveness of the government’s approach to VE can be tied to earlier complains about the extra￾judicial killings and exposure of government informants in the community that may have strained community-security relations. 7.1.5 Gang and Criminal Activity Likewise, respondents’ views on the frequency of organized crime and gang activity pointed to a somewhat improving context. More than half of those who responded to this question agree that gang and organized criminal activities have reduced compared to 12 months ago. Some credited this to the enhanced awareness on the dangers of crime, vigilance, and cooperation on issues of security. A media representative highlighted that: “There have been increased efforts from the government to deal with crimes.” Other reasons given for the perception of decreasing gang activity and organized crime included the increase in educational opportunities that have improved youth’s technical and vocational skills and knowledge to strengthen their capabilities in self-employment, giving hope for secure and better livelihoods. This includes vocational training and free primary and secondary education, by both county and national governments. Organized crime and gang activity are less of a problem now than it was 12 months ago. Midline Endline Strongly agree 20% 18% Somewhat agree 16% 35% No change 13% 16% Somewhat disagree 6% 14% 55 Strongly disagree 33% 10% Do not know 13% 6% Those disagreeing with the statement pointed to an increasing rate of crime. A security actor noted that they receive reports of criminal activities from time to time but did not confirm what type of crime he was referring to; whether it is isolated incidents, or activities related to gangs. A youth representative noted that crimes reported are mostly related to murder and rape cases, and often occur during the night. While some respondents simply denied that Garissa faces challenges of gang activity, a CBO representative indicated there are examples of well-known gangs operating in the county. He quotes: “Youths are still engaging in crime e.g. it’s unsafe for Boda Bodas to pass at Bula Mzuri due to "Squad Chafu" which is a group youth involved in criminal activities.” 7.1.6 Safety and Security Perceptions on the context of security at the national level regarding terrorism were slightly more optimistic. 51% of surveyed respondents disagree that the number of terror attacks in Kenya has reduced, a 9% decline from the midline. Still, over 30% of the respondents believe that the attacks have reduced in the last 12 months, while 16% believe that the situation has not changed from what it was a year ago. Over the past 12 months, the number of terrorist attacks in Kenya has decreased. Midline Endline Strongly agree 13% 4% Somewhat agree 20% 27% No change 4% 16% Somewhat disagree 7% 18% Strongly disagree 53% 33% Do not know 4% 2% Views in respect to the county were much more pessimistic. 69% of respondents mentioned that terror attacks in Garissa County had increased compared to 12 months ago. This again could have been influenced by the increased attacks in the county in the period when the respondents were interviewed. Only 14% somewhat agreed that the number of attacks had reduced. Others noted that there has been no change. 7.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced The table below compares scores for the sentinel indicators from the midline and endline evaluations on indicators that track communities’ willingness to mobilize in support of peace and security objectives at the county level. While NIWETU does not work to support these indicators, we assume that communities’ willingness to engage with these structures, or their trust in these structures’ ability and relevance, will affect their perspectives on the wider community’s willingness to mobilize around CVE. Individual scores reflect respondents’ personal views on the corresponding statement, while the ‘meta’ scores, or meta perceptions scores, reflect how respondents believe their peers would answer. Research has shown that respondents are more likely to act in accordance with what they think are their peers’ views, or the social norm. This means that when meta scores are higher than the individual scores, views in general tend to be more positive about a particular statement, and vice versa. 56 These findings come from a sample of 40 individuals in two neighborhoods in Garissa and should therefore not be considered statistically significant or representative of a wider sample of Garissa residents. Figures presented below are out of a ten-point scale13. Degree to which peace committees are active and responsive. Midline Endline I feel that the peace committees are active and easy to access in my area. Indiv. 5.50 5.71 Meta 5.95 4.75 I feel that the peace committees have an adequate understanding of the issues in my community. Indiv. 4.33 5.04 Meta 4.77 4.65 I feel that peace committees effectively respond to issues in my community Indiv. 5.20 5.74 Meta 5.08 4.73 Overall score 5.14 5.10 Degree to which communities actively engage in barazas. Midline Endline I frequently attend community barazas. Indiv. 3.67 3.32 Meta 4.84 3.48 When I go to barazas, I feel like my opinion is considered and my concerns are responded to. Indiv. 2.72 5.60 Meta 3.30 4.05 I always feel positive after participating in barazas. Indiv. 4.82 5.29 Meta 4.98 4.15 Overall score 4.06 4.32 Degree to which Nyumba Kumi initiatives are active. Midline Endline Nyumba Kumi initiatives are active in my community. Indiv. 4.16 4.99 Meta 4.44 6.02 I feel that Nyumba Kumi initiatives are positively addressing the issues that concern me. Indiv. 3.07 4.76 Meta 5.97 4.28 I have confidence in Nyumba Kumi initiatives and trust that it will help improve peace in my area. Indiv. 5.70 4.83 Meta 5.12 4.26 Overall score 4.74 4.86 Presence of other organizations/actors that work on peace/security. Midline Endline There are numerous other networks and organizations making a positive impact on peace and security in my community. Indiv. 3.56 3.79 Meta 4.18 4.25 Overall score 3.87 4.02 Perspectives on community security structures in Garissa township remain rather low, and across the board, scores hardly varied between midline and endline, with slight declines or improvements in perceptions. Views on the relevance of peace committees were most positive, though still somewhat neutral, and no notable changes were identified in perceptions from midline to endline. Community members’ views on the effectiveness of Nyumba Kumi initiatives and barazas were also somewhat negative, with most respondents at least somewhat disagreeing that these are effective structures for addressing issues of concern. Respondents also generally disagreed with the statement that they attend barazas frequently, and also believed that their peers would respond in the same way. 13 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 57 This suggests that barazas are not effective or relevant platforms for disseminating important messages to the community in urban settings, and government and security actors should find other avenues. Respondents were also still somewhat negative in their views on the prevalence of networks and organizations working to make a positive impact on peace and security in the community. There was a slight improvement from the midline, but this finding suggests that those working on these issues in Garissa are not all that visible to the wider community. 7.3 Section III – Skills and Knowledge The following section looks at the skills and knowledge that exist within the stakeholder community on VE and CVE related activities. The primary focus is on the different types of training that exist on these topics, as well as the degree to which respondents have participated in such activities already, and what, if anything, they are doing with the skills and knowledge thereafter. 7.3.1 Summary of Findings Participation in CVE skills and knowledge training activities is somewhat widespread, but less so than at the time of the midline. Respondents feel that the training programs should be more inclusive by particularly focusing on youth, security actors and elders, in order to target the most vulnerable and increase the appreciation of community security relations in the county. Further to this, recommendations included increasing the number of participants and enhancing the outreach to focus of community members at the grassroots. This could be done by diversifying the channels of communication to include creative media programs and simplifying CVE messaging to audio-visual approaches and local languages. This was also seen as a way of sustaining ownership and dissemination of CVE knowledge and skills among communities in the urban, rural, and border towns that are prone to VE. 7.3.2 Skills and Knowledge Index Scores The following table shows the average scores on a four-point scale14 from several key questions related to skills and knowledge. All scores are presented side-by-side to highlight changes over the period. Findings from the midline demonstrate a slight increase in baseline scores on the level of skills and knowledge amongst stakeholders in Garissa. This, however, declines slightly at the endline. There was an increase in the number of participants at the training programs at the endline compared to the baseline and midline. The degree to which the training programs have improved understanding of VE and its use thereof has declined slightly at the endline, but the frequency with which the knowledge acquired is shared has been positively maintained from the midline to the endline. All scores are still overwhelmingly high, meaning that even though there has been a decline, sentiments toward the relevance and effectiveness of training are still largely positive. Questions Baseline Score Midline Score Endline Score Have you participated in any training on CVE and understanding the causes of VE in the last 12 months? 2.85 2.72 3.00 How much, if at all, do you think these training have helped you understand VE and how to prevent it? 3.79 3.81 3.73 How much, if at all, do you use the information you have learned from these trainings? 3.83 3.84 3.58 Have you shared the information you have learned with others in the community? 3.36 4.00 4.00 14 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 58 Overall average score 3.46 3.79 3.58 7.3.3 Participation in Training Nearly two-thirds of respondents participated in trainings on countering violent extremism during the midline but dropped to a quarter at the endline. At the endline, representatives from community-based organizations and county administration categories interviewed had not attended trainings that year. In addition, women’s group, media, community/religious and youth representatives participated the least. However, participation by these categories was higher during the midline. Participation in trainings by Category Midline Endline Participated Did not participate Participated Did not participate CBO 4 1 0 7 Community/Religious leader 3 4 1 3 County administration 4 4 0 7 Educational professional 5 2 3 1 Media 2 2 1 3 Private sector 2 3 4 4 Security sector 2 2 0 2 Women’s representative 5 1 1 4 Youth representative 4 4 1 5 Total 31 (57%) 23 (43%) 12 (25%) 36 (75%) Overall, 50% of the respondents indicated having participated in NIWETU-sponsored trainings at the endline. Most of the participants were able to name the organizations that led the training which included: ROAD International; Women Kind; SUPKEM – Garissa; Malaika Foundation; Pastoralist Girl’s Initiative (PGI); and C4C training sessions. Non-NIWETU partners mentioned, such as CODI, Life and Peace Institute (LPI), Kesho Alliance, Action Aid Kenya, Waris Women Group and the county government, facilitated sessions. As in the endline and midline, most commonly, respondents participated in trainings on understanding violent extremism. The interest in understanding violent extremism has remained steady throughout the activity and could be a reflection of the need to keep up with the evolving trends of recruitment and VE dynamics. Other themes frequently mentioned included how to prevent or counter violent extremism, why people join violent extremist groups, and how to work with the government to prevent and counter violent extremism. As in the midline, least common was participation in monitoring and evaluation training. Other themes mentioned were understanding the role of youth in countering violent extremism. Types of Training Baseline Midline Endline Understanding violent extremism 62% 84% 80% How to prevent or counter violent extremism 52% 68% 63% Why people join violent extremist groups 59% 74% 40% How to work with the government to prevent or counter violent extremism 45% 45% 33% Conflict resolution 45% 26% 33% How to report extremist activities 31% 19% 28% Training others on preventing or countering violent extremism 24% 36% 28% How to talk to others about violent extremism 38% 32% 25% 59 Leadership skills 35% 39% 23% How to work with other organizations to prevent or counter violent extremism 31% 42% 23% How to lead interfaith discussions on violent extremism 7% 39% 13% Advocacy skills 7% 10% 18% How to identify recruiters and recruitment methods 10% 7% 15% Monitoring and evaluating 4% 7% 8% Other 38% 0% 8% Identifying signs of recruitment 17% - - Respondents have diverse views on the kind of topics that each would consider helpful. 59% of those who participated in trainings believed that their new skills would be relevant in their work. This statement was echoed across all categories. Many respondents pointed out that the opportunity to attend the training helped them to meet new colleagues working on similar issues or strengthened their existing knowledge on CVE. All survey respondents stated that the trainings have helped to improve participants’ understanding of the subject matter. Likewise, all but three respondents, including a media representative, county administrator and private sector representative, stated that they use the skills and knowledge acquired. A religious leader noted: “I managed to disengage some people who wanted to go and join as who were radicalized by sharing the skills, I learned with them and referred them to available avenues.” To that end, survey respondents articulated a better understanding of VE compared to the midline. Some respondents suggested that VE is the use of violence to pursue certain goals or progress certain ideologies, while others referred specifically to religious ideology or extremism. Others were less categorical about VE being associated exclusively with religious extremism. A women’s group representative defined violent extremism broadly as “any violation against human rights”, while several respondents highlight the use of fear to intimidate others, without necessarily mentioning any specific violent extremism group. While some respondents had a much clearer understanding of CVE, others viewed it within the counter-terrorism lens. A community-based organization stated that CVE is a combination of “measures to prevent vulnerable individuals or groups from getting radicalized” while a youth representative defined it as “Ways of strengthening the community against radicalization and religious conflict.” A county administrator, on the other hand, defined CVE as “one forming groups to deal with VE’’, as a media representative noted that it is “Kupambana dhidi ya Ugaidi” (dealing with terror issues). Survey respondents also discussed the various ways in which they have utilized the new skills and knowledge acquired from these activities: 1. Training others on VE and CVE 2. Sharing information gained with a wider audience, including madrassa students, family and friends and network of individuals 3. Working to deradicalize youth 4. Becoming more cautious of one’s surroundings 5. Strengthening peace advocacy campaigns 6. Mentoring and supporting youth 7. Sharing what was gleaned from the training on social media platforms. 60 7.3.4 Improving Training Programs There were several proposals made by the respondents on how the trainings can be improved for them to be more effective. Respondents recommended that to be more effective, CVE trainings in general should be more inclusive and consider the diversity and numbers of those targeted, with a particular focus on the youth who are most vulnerable. One media representative emphasized that ordinary community members, even at the village level, should be able to access some of the knowledge shared in training forums. A county administrator stated that the trainings should also include more elders at the village level. This, perhaps, is an acknowledgement of the degree of influence such elders have among communities in Garissa, especially in the more remote areas, to enhance community ownership. “Training should be taken to the grassroots – The programs should be owned by the community, not government or other stakeholders.” – Women’s Group Representative Likewise, including the religious leaders into CVE trainings and activities is vital, so they can preach the skills learned in mosques to reach the larger community. To reach out to these groups, respondents recommended more training for trainers and leaders who have local relevance, such as Nyumba Kumi agents and local government authorities. This, they noted, would be important in improving information sharing and disseminating CVE content at the lower levels of the society. A religious leader stated that: “The trainings should focus on improving community – security relations. They should take CVE narratives to the village especially to the sub-counties at the border.” NIWETU completed community-security relationship building work in border areas, including Dadaab, Fafi, Hulugho, and Ijara and, since data collection was done within Garissa Township, respondents might not be aware. Other ideas on how to improve outreach were also shared, with respondents proposing the use of media channels, such as local FM stations, TV shows and animation to simplify the message, but also to appeal to more youths. These materials, a community leader adds, could be translated to local languages. One respondent from the private sector was in favor of more follow-up training and monitoring to ensure that those trained can incorporate the skills and knowledge into their daily work, thus to impact the spaces they occupy and also determine the gaps in knowledge and skills that are most useful. Respondents also pointed to specific future training needs. The most useful type of training identified during the endline was about understanding VE, views that were similar at the midline. Other themes proposed include preventing or countering violent extremism, why people join extremist groups, conflict resolution skills, and training others on preventing or countering violent extremism. Least relevant training identified over the two periods was on monitoring and evaluation. What types of trainings would be most useful? Baseline Midline Endline Understanding violent extremism 62% 74% 73% How to prevent or counter violent extremism 38% 65% 40% Why people join violent extremist groups 68% 65% 33% Conflict resolution 23% 44% 28% 61 Training others on preventing or countering violent extremism 45% 56% 25% How to report extremist activities 17% 37% 18% How to work with the government to prevent or counter violent extremism 19% 44% 15% How to work with other organizations to prevent or counter violent extremism 34% 50% 15% Leadership skills 28% 41% 13% Advocacy skills 11% 30% 13% How to talk to others about violent extremism 32% 46% 13% How to identify recruiters and recruitment methods 2% 15% 10% How to lead interfaith discussions on violent extremism 15% 44% 8% Monitoring and evaluating 2% 6% 3% 7.3.5 Sharing Lessons Learned As was highlighted during the midline, sharing lessons learned through trainings with a wider audience and network of individuals was one of the most common ways in which respondents used their new skills and knowledge. Most respondents reported discussing the issues with their family members, colleagues, community leaders and other community members, which is a significant improvement from both the baseline and midline. There was also an increase in sharing information among colleagues within and outside of their organizations, as well as in mosques and churches during the midline, but this decreased during the endline. This perhaps is a reflection of the reduced number CVE activities that provided opportunities for interaction and networking. The increased sharing with less formal networks, such as friends and family, as earlier mentioned points to the sustainability of CVE messaging spread outside formal spaces and also points to the relevance of the training content at the household level. Perhaps indicative of somewhat poor relationships with county officials, only two respondents – a private sector and CBO representative – noted sharing what they learned with colleagues in government. It was noted that members of some women’s groups had shared the information and skills learned with people from their mosques or church and business partners. It is worth noting that there has been consistency in the culture of information sharing of knowledge and skills acquired by participants and stakeholders, identified from the baseline to the endline, perhaps pointing to an emphasis on sharing lessons, more platforms for doing so, or less fear of engaging in CVE conversations outside formalized settings to feel confident sharing with peers, family and the wider community. With whom have you shared your new skills and knowledge? Baseline Midline Endline Family members 21% 77% 55% Community members 34% 63% 50% Friends/school mates 4% 23% 45% Community leaders 30% 30% 35% Colleagues in my organization 40% 50% 18% People from my mosque/church 32% 37% 18% Colleagues at other organizations 26% 10% 18% Colleagues in the local government 19% 0% 10% Other 23% 0% 5% 62 Business partners 21% 20% 5% I have not shared the information or skills with anyone 0% 0% 3% 63 7.4 Section IV – Community Networks The following section assesses the existence and strength of community-based networks working to counter extremism in Garissa. This section addresses the community-driven CVE strategies that are underway in Garissa. An analysis of the finding will be followed by a summary. 7.4.1 Summary of Findings Overall, although it seems that less collaboration is happening now, the value from CVE networks in Garissa has increased and more actors are willing to address conflicts that arise among them when working together. This finding signals the need for continued positive and productive engagement between key stakeholder groups in Garissa. The establishment of the CEF was said to provide a platform to develop collective strategies in addressing issues of VE. As such, it provides an opportunity to develop more elaborate strategies that incentivize collaboration between stakeholders, rather than competing for recognition if they were not working through a platform designed to promote cooperation. 7.4.2 Community Networks Index Score The table below highlights the average scores of five questions (on a five-point scale15) related to establishing community networks to counter violent extremism. The scores indicate that perceptions are not as positive about collaboration between stakeholders on issues of CVE in the last 12 months. Respondents were also less positive about the frequency of conflict and disagreement between actors, but still overwhelmingly suggested that these issues arise infrequently. On a positive note, respondents largely agreed that improved networks between actors working on CVE have been helpful, showing an increase from baseline and midline scores. Questions Baseline Score Midline Score Endline Score Do you work with any other organizations on CVE? 4.15 3.55 3.37 Do you collaborate more, less, or the same amount with these actors than you did 12 months ago? 3.41 4.32 4.06 How often are there conflicts or disagreements between different actors working on CVE? 3.85 3.62 3.15 Overall, do you think improved networks between organizations working on CVE has been helpful? 4.41 4.40 4.59 Overall average score 3.78 3.81 3.79 7.4.3 Collaboration between Actors Over 67% of the respondents reported working with other actors on CVE related issues. The most involved organizations that the actors work with are youth organizations. Women’s groups and religious leaders were also frequently engaged in CVE activities. Eleven respondents mentioned having worked with security actors too. Least engaged included self-help groups, private sector representatives and teachers. A CBO representative mentioned having worked with similar CBOs in the community. Reasons why respondents preferred working with specific types of stakeholders include: 15 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 64  Community-based organizations and peers were accessible and approachable, either because they are actors the respondents work with in programs, are conversant with CVE and therefore are interested in participating in discussions on this topic.  Community, youth and religious leaders are respectable and have large networks in areas where the respondent may not. This means that working with such groups is essential in disseminating messages, mobilizing and getting buy-in from the community and other stakeholders.  Security actors and county administrators are seen to have the mandate on issues of security – a media representative noted that his choice of working with security actors is because they are directed by law to work on issues that relate to security. Perhaps the journalist likely feels more comfortable reporting the information without verifying further, since they can refer back to an authority figure as the source, and their fears of being targeted are also less, since that security official chose to provide the information.  A CBO representative noted that his choice to work with other stakeholders was enshrined in the grant agreement with NIWETU, while a community leader notes that working with others is a commitment to inclusivity.  An actor’s ability to disseminate messages, because they have a large constituency or following and are reputable.  Women’s group representatives working with youth organizations, security sector actors, county or national government actors, note that their choice is because these actors are the most affected and they need to link with government and security. What organizations do you work with? Baseline Midline Endline Religious leaders 48% 54% 52% Youth organizations 74% 54% 49% Community leaders 45% 43% 42% Security sector actors 62% 69% 39% Women’s groups 50% 34% 27% At-risk youth - 11% 21% County or national government actors - 49% 21% Teachers 6% 14% 18% Other community members 29% 31% 12% Business owners 21% 11% 9% Self-help groups 29% 3% 6% Other 36% 6% 3% Although there was an increase in collaboration between actors and specifically with security actors, women’s groups, and religious leaders at the midline, the rate at which actors are collaborating decreased slightly during the endline. This perhaps is also as a reflection of the reduction in CVE activities and opportunities for engagement. Less than half of the survey respondents confirmed that they collaborate at least a little more than they did before. Most of the respondents who noted that they collaborate much more explained that this was a reaction to an increased number of VE attacks. Additionally, owing to a recent attack on teachers and security personnel, a county administrator felt that it was necessary to bring different actors on board to engage the community collectively. A youth representative pointed to increased appreciation for the fact that they serve the same problem, while a women’s group representative echoed that this collaboration has helped build trust, 65 especially with security actors. A youth representative said his engagement in CVE activities, especially training, improved his capacity to collaborate and engage other actors over the last 12 months. Do you collaborate more, less, or the same amount now as you did 12 months ago? Midline Endline A lot more 54% 36% A little more 27% 36% The same amount 6% 24% A little less 9% 3% Do not know 3% 0% The formation of the CEF, which is perceived by key informants to be quite inclusive, provides a platform for CVE actors and community leaders to engage. The forum meets monthly and quarterly and allows for sharing and collective learning among actors. This has also enhanced networking among individuals active on issues of CVE. A key informant noted: “The CEF is the most appropriate forum for coordination, mobilization and collaboration of the CVE work in the county but it needs to be strengthened in terms of effectiveness.” The value is not just exclusive to members. A key informant noted that, although they are not officially members of the forum, they benefit from the forum’s engagements. Referring to the CEF, he states that: “They do sometimes make follow-ups on us and we sometimes engage them in our activity analyses, spot-checking, sharing our deliverables, progress we have made on the ground and our strategies for activity implementations.” 7.4.4 Conflict between Organizations At the endline, more than half of the respondents noted that conflict occurs rarely or does not occur at all. As opposed to the midline where 55%, claimed to not know the frequency of disagreements between organizations, only 20% did not know. This perhaps could indicate improved interaction and awareness on the relationship dynamics among actors. How often do disagreements occur between organizations working on CVE? Midline Endline Very often 4% 16% Somewhat often 7% 8% Rarely 22% 29% Never 13% 27% Do not know 55% 20% Highlighting the source of conflict among CVE actors in general, respondents point to disagreement over ownership of activities, choice of target participants, and duplication of activities. A women’s group representative notes that stakeholders at times “fight for power like which actor to work on which activity and where to focus,” while a community leader further stated that such competition exists because “sometimes they have the same activity with the same participants”. Others, according to a youth representative, are confronted by other actors and community members for being perceived to be politicizing or taking advantage the CVE and other development programs, to mobilize political or financial capital to benefit themselves. 66 Occasionally, disagreement over differing opinions on strategies and approaches to VE and CVE do occur. An example noted was conflict over who took credit or ownership of CVE programs and as a youth representative states, “opinions on how to report VE suspects”. Over the last 12 months, conflicts and disagreements that are presented to the CEF, or involve members, are resolved via a dialogue at CEF forums. Other forums, where stakeholders meet even outside of the CEF, encourage participants to promote a safe space that accommodates divergent views and opinions. All but one, a representative from the media, believe that community networks have been somewhat useful. 7.5 Section V – Community-Driven Strategies & Initiatives This section addresses the community-driven CVE strategies that are underway in Garissa. An analysis of the finding will be followed by a summary. 7.5.1 Community-Driven Strategies and Initiatives Index Score The table below presents average scores from four key questions on a five-point scale16, about the effectiveness of community-driven initiatives to counter violent extremism. Scores indicate a decrease in the rate at which organizations work with other stakeholders to develop CVE programs in the last 12 months. However, the extent to which such organizations engage communities on CVE activities has increased, even though the perception of the effectiveness of the programs seems lower than it was during the baseline and midline. It is worth noting that, as described below, less than half of these respondents had participated in NIWETU activities. Survey findings also indicate that CBOs are as active as they were during the midline which was much higher than the baseline period. Questions Baseline Score Midline Score Endline Score Have local organizations worked with other stakeholders to develop programs on CVE in the last 12 months? 4.09 4.57 4.04 How effective do you think these programs are at teaching others about VE and CVE? 3.91 4.59 3.73 Are CBOs more or less active now than they were 12 months ago in implementing such programs? 2.87 3.52 3.51 To what extent do community members engage with these organizations? 3.39 4.05 4.13 Overall average score 3.57 4.18 3.85 7.5.2 Participation in NIWETU-Sponsored Activities Participation of respondents in NIWETU-supported CVE initiatives was limited to 19 respondents or 39% of the total. Those involved participated mostly in activities led by SUPKEM, WomanKind Kenya, CODI, ROAD International, and SUPKEM. One mentioned having participated in C4C activities. Others mentioned Ijara Women for Peace, Kenya Red Cross, Life and Peace Institute, Kesho Alliance, the county government, or did not remember the name of the organization running their activity. SUPKEM, WomanKind Kenya and ROAD have been partners of NIWETU. 7.5.3 Community-Driven Strategies As it was during the midline, the majority of respondents, or 80%, acknowledge that local organizations have been able to develop CVE programs and activities in the last 12 months. Only 14% respondents, 16 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 67 a security actor, community leader and a private sector and women’s group representative, disagreed; while 6% did not have an opinion on this. Respondents mentioned that informal meetings in the community were more common in the last 12 months, while formal meetings with religious leaders and security actors were active, but a little less than during the midline. This is perhaps indicative of the growing emphasis on community engagement on CVE, whose impact or effectiveness may not be easily measured like formal meetings would. CVE Activities Occurring in Garissa Baseline MidlineMidline Endline Informal conversations in the community 40% 45% 80% Meetings with religious leaders 38% 81% 53% Formal meetings with community leaders or members 36% 64% 51% Training sessions 17% 40% 47% Meetings with the security sector 32% 64% 47% Meetings with government authorities (local or national) 28% 47% 29% Meetings with other community organizations 47% 30% 27% Sports activities 11% 15% 29% Mentorship programs in schools 6% 15% 16% Advocacy campaigns 4% 9% 20% Discussions in schools 4% 9% 2% Roleplaying 4% 4% 6% Do not know - 4% 10% There are no activities on this topic taking place 4% 2% 2% Other 11% 0% 0% Activities such as roleplay, discussion in schools and school mentorship programs were cited as least common. It is worth noting that schools have been closed in Garissa because of insecurity, meaning that reaching teachers for this research was difficult and respondents may not be fully aware of what has happened in schools. One key informant noted that there have been interfaith dialogues, women’s empowerment programs for CVE and leadership programs targeting youth, including those active in the peace and CVE space. 7.5.4 Effectiveness of CVE Programs 74% of respondents believe that CVE programs in Garissa are at least somewhat effective, while only 14% disagreed that these methods are effective. 10%, however, did not know whether they were effective or not. 40% indicated that meetings with religious leaders and formal meetings with community leaders were the most effective. This could mean that, even though meetings with community leaders were not mentioned by any responding as being active over the last 12 months, they still consider them effective. Besides, there was an increased appreciation in the value of CVE training sessions than during the midline. Perhaps due to the increase in training programs that demonstrated the value through content and networking as highlighted in previous sections. The least effective initiatives were roleplaying and discussions in schools, as reported by 3% respectively. 68 Most effective CVE initiatives Baseline Midline Endline Meetings with religious leaders 38% 63% 40% Formal meetings with community leaders or members 17% 52% 38% Training sessions 21% 19% 38% Meetings with the security sector 23% 48% 28% Meetings with other community organizations 28% 19% 25% Sports activities 24% 13% 20% Advocacy campaigns 9% 8% 15% Informal conversations in the community 30% 44% 13% Meetings with government authorities (local or national) 28% 28% 13% Mentorship programs in schools 2% 8% 13% Discussions in schools 6% 13% 3% Role playing - 2% 3% Other 6% 0% 0% Some reasons were given on why respondents evaluated the initiatives in this way:  They help to disseminate messages to a wider audience.  They are inclusive and bring the right people together who are informed, have authority and are respected.  They help other stakeholders understand community issues.  They enhance networking and build relationships especially between security actors and community members.  They help inform policymakers.  The security actors were considered default stakeholders because they have the authority and machinery to deal with issues of security especially VE. A county government representative noted that religious leaders are instrumental actors in the county for they are very knowledgeable on religious teachings and help others interpret the teaching of the Quran accurately in understanding VE messaging or designing counter-narratives. A youth leader stated that: “Although religious leaders are respected by everyone and it is community leaders who are listened to by the community.” The leadership of such an initiative is seen as important in the effectiveness of community engagement. A CBO representative notes that initiatives led by locals who have a good understanding of local dynamics make it easier for stakeholders to understand the perspective of the community and create more ownership of the process. A private sector actor who views such activities as somewhat effective and is more skewed to trainings, stated that: “It is the easy way to reach to people like me as we don't get time mostly and it is the training sessions that are easy for us to attend to CVE issues and community issues in general.” Still, some respondents who were a bit concerned about the efficacy of such initiatives noted that, despite the efforts, many vulnerable people are not reached adequately through some the activities, especially those living in remote areas. 69 7.5.5 Level of Engagement As it was established during the midline, a large majority, 94%, of respondents noted that community members engage with CBOs working on peace and security issues at least a little bit, pointing to the fact that they are well-embedded and well-established entities in Garissa town. This provides a strong case for strengthening CVE community engagement programs by leveraging the constituencies of religious and community leaders. To what extent do community members engage with these organizations? Midline Endline Engage a lot 53% 33% Engage a little 47% 60% Do not engage at all 0% 2% Do not know 0% 4% As it was during the midline, faith-based organizations were viewed by over half of the respondents as the most effective form of community-based organization, pointing to the high regard placed on religious leaders among the communities in Garissa County. 51% of respondents also believe youth organizations to be the most effective in these activities, as they can work with and reach the members of the community deemed to be the most vulnerable, while women’s groups are viewed as important perhaps due to their commitment in community service and strong social networks. As was in the midline, the least effective are cultural organizations, according to survey respondents. 7.5.6 Summary of Findings The findings presented above point to a largely improved context concerning community-driven activities and strategies in Garissa County. Notable has been the increased informal engagements in the community and formal meetings with the security actors. Still, on effectiveness, meeting with religious and community leaders is viewed as the most effective strategy, even when such activities seem to have reduced significantly over the last 12 months. Respondents acknowledge that local leadership is key to strengthening CVE strategies for local leaders to understand the context, have a level of influence in the community and can sustain such activities after activities are completed. This is an improvement from the midline, where informal and non-donor funded activities were not identified, and recommendations made included working with groups and local organizations to identify ways to develop and implement CVE strategies in ways that encourage sustainability and self￾reliance. 70 7.6 Section VI – Objective 2: Improved Government Capacity The table below compares scores for sentinel indicators from the midline and endline evaluations related to Objective 2. Scores for the meta perceptions indicators are again on a ten-point scale17 and reflect the same sampling dynamics as the Objective 1 questions above. Scores for the statements regarding legislation and budgets were identified through a review of the CAPs and County Integrated Development Plans (CIDPs) which spell out the county budgets. Level of government and community participation in community￾government dialogues Midline Endline When I express my views about violent extremism to members of the government, I feel like my opinion is considered and my concerns and ideas are responded to. Indiv. 3.75 3.04 Meta 8.51 2.83 The government has made a strong effort at participating in community dialogue over the past two years. Indiv. 8.06 7.95 Meta 5.38 7.45 I feel like I have been adequately included in the county government’s response to peace and security. Indiv. 5.32 5.23 Meta 3.25 3.32 Overall Score 5.71 4.97 Community perceptions of the national and county government on issues relating to peace and security Midline Endline The national government has made a great effort at improving the peace in my area over the last two years. Indiv. 6.02 2.49 Meta 5.86 2.48 The county government has made a great effort at improving the peace in my area over the last two years. Indiv. 4.70 3.65 Meta 4.72 3.09 The security sector has made a great effort at improving the peace in my area over the last two years. Indiv. 7.12 2.55 Meta 7.26 2.28 My community has become more willing to cooperate with the security sector over the past two years. Indiv. 8.06 3.51 Meta 8.39 3.18 Overall Score 6.52 2.90 Midline Endline Presence of line items for CVE activities in county budgets 0 0 Presence of legislation to support implementation of the CAPs 0 0 The scores presented above demonstrate a rather significant decline in perceptions about community￾government and community-security relations. Respondents, on average, were neutral in their views on the level of community and government participation in joint dialogues, while at the midline, they were somewhat positive. During the midline, respondents believed their peers would feel their concerns on VE would be responded to, even while as individuals they mostly disagreed. This suggested that individuals are likely to act in line with what they think their peers experience, even if their personal opinions are different. By the endline, however, respondents felt that their peers would be equally negative in their perceptions on government’s responsiveness to their concerns about VE, 17 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 71 suggesting that there is a growing divide between communities and government officials, even while they tended to agree that the government has made an effort at participating in community dialogues over the last two years. Views also declined significantly in perceptions on the role of county and national government in addressing VE related matters. While views of government tended to be rather negative across the board, respondents were particularly negative in their assessments of the national government and security sector. Likely because these arms of government have been mandated with addressing security matters – while county government remains minimally involved in these issues – the increase in VE-related attacks, and the resulting closure of schools are likely blamed on national government level officials. As suggested, these findings are not surprising, given the uptick in VE activity that has significantly negatively impacted Garissa County in the early part of 2020. Respondents are likely responding to a deteriorating security context and placing the blame on government and security sector officials. The current insecurity in the region likely colors respondents’ perceptions of the steps the government has taken; if communities are not able to experience safety and security, they likely perceive all the government’s efforts to be insufficient and are therefore more likely to express negative views. It is possible that, had this research been conducted prior to the increase in attacks in the county, views would have been more positive. 72 7.7 Section VII – Government Responsiveness to CVE The following section presents findings on the responsiveness of county and national level government to address VE. An in-depth analysis is described followed by a summary of the findings. 7.7.1 Government Responsiveness Index Score The below questions are scored on a four-point scale 18 to understand overall government responsiveness to VE at both county and national levels. The findings demonstrate an overall increase in the perceptions of the government responsiveness to VE in Garissa County from the baseline, even though there has been a decline since the midline, reflective as well of household survey responses on government’s engagement with communities on matters of security and stability. Most notable is the increased pessimism on the respondents’ perceptions of the willingness of the county government to work on VE-related issues. Questions Baseline Score Midline Score Endline How well do you think the county government understands the problem of VE in this community? 2.28 3.15 3.04 How well do you think the national government understands the problem of VE in this community? 3.67 3.67 3.44 How willing do you think county government is to work on issues of VE? 1.82 4.67 2.61 How willing do you think the national government is to work on issues of VE? 3.47 3.73 3.08 How easy do you think it is to work with national government counterparts? 3.40 3.47 2.96 Overall average score 2.93 3.74 3.03 7.7.2 Government Understanding of VE The majority of survey respondents still believe the county government understands the problem of VE at least somewhat well, while only 10% suggested that county officials have a poor understanding of the subject matter. All those who indicated that the county government understands the subject very well cut across different categories, as opposed to the midline where CBOs were more optimistic. This perhaps points to the diversity of actors engaging the county government probably due to the establishment of a more inclusive engagement forum such as the CEF. How well does the county government understand VE? Baseline Midline Endline Very well 26% 46% 40% Somewhat well 17% 33% 35% Not very well 15% 13% 15% Not at all 40% 9% 10% Don’t know 2% 0% 0% The existence of the CEF co-chaired by the county government continues to offer a platform for government representatives to learn more about security issues from the community and key security county actors, even though security may not be a devolved function. This means that the county government has more opportunities to understand the community’s perceptions on violent extremism and is better updated on the trends in the county. 18 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 73 At the national government level, survey respondents’ views were even more positive, with all but four respondents believing that the national government has a good understanding of VE issues in the county. A key informant explains that the national government has a mandate on security issues, resources, and access to information because of the strong network of across the county. 7.7.3 Willingness to Engage in Programming Half of respondents seem to believe that the county government is at least somewhat willing to work on issues of CVE, a sizeable number did not. These views are slightly less positive as compared to the midline. How willing are county government is to work on CVE? Baseline Midline Endline Very willing 15% 33% 33% Somewhat willing 11% 35% 18% Not very willing 11% 17% 31% Not at all willing 58% 13% 14% Do not know 6% 2% 4% Key informants are split on the county government’s commitment to CVE processes in Garissa. There is credit given to the county’s involvement in the CAP process and outside of the county’s CEF engagements. A key informant, who is not a member of the CEF, mentions having worked with the following department under the county: “Department of youth gender affairs, county security and operation and the county religious affairs.” Still, according to a key informant, the perceptions of the county’s unwillingness to work on CVE issues was symbolized by the fact that the governor chooses to delegate his function to other county officials all the time. Those delegated to coordinate the forum at times have not been committed to CVE work in the past and are not generally aware of what is happening in the sector. Another key informant notes that another reason was the fact that the county government had not provided financial support to the CEF and activities of its members as much as they were expected to. “We closely work with the deputy county secretary in terms of security who is the head of peace and security for the county government. There are challenges in engaging with the county government as they don’t show up when invited in our activities and in our updating meetings with other CVE actors which we conduct after every activity in our work-plan.” – Key Informant “We also requested the county government to support us in terms of logistics to continue in our programs as a champion and we have not yet heard from them.” A key informant from the county government, on the other hand, notes that they are supportive of the CVE process but are selective on who they chose to work with. He notes that “very few organizations are active on CVE in Garissa” and further states that the role of the county government continues to be limited by the existing legal frameworks for security. This being an ongoing challenge, a key informant noted that there are plans to strengthen the county laws to guide the county government engagement on CVE, including funding activities. “County government are not that much involved in security issues since security is under the docket of the national government. It is not a devolved function and we don’t have the legislative right to do anything about it.” – Key Informant 74 “Needs for the legislative role and responsibility-sharing of CVE between the county the national government as the county always push everything to the national government as far as it concerns CVE.” – Key Informant As was observed during the endline, views on the national government’s willingness to engage on issues of CVE remain positive. 82% of survey respondents believe that the government is at least somewhat willing. However, those who perceive the national government as very willing to work on CVE issues has decreased significantly. Only three respondents mentioned that there are some difficulties when working with the national government, while most respondents identified chiefs and assistant chiefs as the easier national government officials to engage on issues of CVE. Still, a key informant notes that the national government has not provided financial support to CVE activities in line with the Garissa CAP. A frequent and unpredictable transfer of county commissioners was also viewed as a challenge, for a new official took time to build rapport with and catch up on progress made, especially on the CAP and community engagement. A key informant noted: “Though there are challenges of transfer of the county commissioner at any time and the change of county government stakeholders at every five years elections which is not giving grantee to the current team who are working on the CAP implementation otherwise the CAP implementation seems promising for now and we have high hopes that it will continue for years to come.” How willing are national authorities to work on CVE? Baseline Midline Endline Very willing 78% 75% 31% Somewhat willing 13% 24% 51% Not very willing 3% 2% 14% Not at all willing 3% 0% 4% Do not know 0% 0% 0% 7.7.4 Engagement with Government Counterparts The frequency of engagement with both county and national government counterparts appeared to have reduced compared to the midline, although engagement with the national government is still reported to be stronger than engagement with the county government. Consequently, a higher number than that observed during the midline, 29% and 22%, rarely engage with county and national government officials, respectively. The number of respondents who mentioned that they engage the national government daily has reduced too. This could be as a result of reduced CVE activities, or other engagements or some reflection of some of the challenges highlighted above. Frequency of engagement with national government counterparts Baseline Midline Endline On a daily basis 43% 37% 14% Once a week 4% 2% 6% A few times a week 0% 6% 8% Once a month 2% 6% 8% A few times a month 11% 6% 14% Occasionally 23% 24% 27% Rarely 15% 19% 22% Do not know 2% 2% 0% Unsurprisingly, survey respondents pointed to significant challenges in dealing with both levels of government. In describing the challenges faced engaging the county government, respondents commonly noted that the county does not involve them at all, fails to honor scheduled meetings when 75 invited, and often imposes programs without consulting the stakeholders. Only five respondents acknowledged not having any problems engaging the county government. Although respondents reported having engaged the national government much more than the county government, there were some challenges reported. 39% of respondents stated the national government’s prioritization of law enforcement and arrests was a strain on their relationship with CVE stakeholders. Other problems faced by respondents were that the national government did not involve them, try to talk to them or do not seem to value their contributions. Only ten respondents had no issues engaging the national government. Frequency of engagement with county government counterparts Baseline Midline Endline On a daily basis 11% 26% 12% Once a week 0% 2% 2% A few times a week 0% 9% 8% Once a month 0% 7% 4% A few times a month 0% 7% 16% Occasionally 53% 22% 25% Rarely 21% 22% 29% Do not know 0% 6% 4% The high degree of frustration expressed by respondents with both levels of government perhaps indicates that the largely positive findings highlighted earlier in this section are perhaps slightly positively biased. As repeatedly noted, respondents demonstrated some degree of hesitation when speaking about VE and CVE related matters, out of a heightened sense of fear and suspicion. As such, these findings should be considered in context. 7.7.5 Summary of Findings Although the perception of the county and national governments’ understanding of VE issues has been steadily positive, there is a growing frustration that both levels of government are not demonstrating a strong commitment to working on CVE issues in the county. Further to this, engagement with government counterparts has reduced among stakeholders. Most feel less valued or involved by both governments in shaping CVE programs. The establishment of the CEF is seen as instrumental in strengthening these relationships but required both governments to demonstrate political will by not only investing in the forum, but also by taking leadership responsibilities in coordinating key community stakeholders. The legal framework for security continues to define the relationship between both governments and is seen as a source of ambiguity in defining clearer roles among actors in the county. 76 8. ANNEX III – ISIOLO COUNTY REPORT: ISIOLO COUNTY ENDLINE OVERVIEW Data collection took place in Isiolo Town over a seven-day period, carried out by two researchers. NIWETU facilitated introductory meetings to key stakeholders to explain the goals of the research and to facilitate further introductions to relevant actors. The research team conducted quantitative surveys with key stakeholders in Isiolo, as well as the wider community. The breakdown of respondents in the stakeholder questionnaire is as follows: Number of respondents Category of respondent Number of respondents Security sector 7 Educational professionals 7 Women’s groups 6 Youth representatives 6 Community-based organization 5 Community leader/Religious leader 5 County administration 5 Media 5 Private sector 4 Total 50 Household surveys were conducted with 40 individuals around Isiolo town, selected through the random walk method. Thirteen respondents were female and 27 were male; 27 respondents were Muslim and the remaining respondents Christian. Respondents come from a range of occupations, though the largest proportion, 30%, were unemployed. All but one had at least some schooling, while 55% had received a college diploma or higher. Key informant interviews were carried out with ten individuals, representing the following categories: Category of respondent Number of respondents Community-based organization 3 Community leader 3 Security sector 1 County government representative 1 Youth representative 2 Total 8 77 9. KEY FINDINGS 9.1 Section 1 – The VE Context The following section presents findings about the broader VE and security context in Isiolo County, including the ‘VE index’ score, in comparison with scores from the baseline. 9.1.1 Summary of Findings Positive perceptions of the VE context in Isiolo have slightly declined, with findings showing relatively mixed views amongst key stakeholders. Increased awareness on CVE has brought more attention to the risks posed by VE actors in a way that communities did not see before. The 14 Riverside attack and its connection to Isiolo is still fresh in people’s minds and seen as evidence that VE remains a problem, even though no attacks have happened in the county. While there have been recent cases of recruitment and youth disappearances reported, respondents generally believe that the frequency with which these phenomena occur has decreased. As findings presented in subsequent sections in this report will suggest, greater information sharing, cooperation, and knowledge of VE issues and recruitment methods may result in a heightened sense of vigilance and awareness of extremist activity in Isiolo. 9.1.2 Causes of Insecurity A variety of security threats, which affect residents, exist in Isiolo County. While the scores presented above indicate a slightly worsening VE context, respondents still do not believe extremism to be the biggest threat to safety and security in the county. Still, VE is now elevated to a major threat, alongside unemployment, theft, and inter-ethnic conflicts that are prevalent in Isiolo County. Worth noting is the rise of national politics as an even bigger threat to security, with 56% of respondents stating as much, compared to the midline phase where not a single respondent highlighted this as an issue of concern. The level of threat posed by drug abuse has dramatically increased, from 34% believing it to be a problem in the midline evaluation, compared to 82% at the endline. Key informants, however, were quick to point out that many of these issues are interlinked with extremism and extremist recruitment; unemployed youth are particularly vulnerable to recruitment into extremist groups, as are drug users. The escalating levels of drug abuse could be an indication of an increase in other factors that make individuals at risk, such as unemployment. Security threats Baseline Midline Endline Drug abuse19 - 38% 82% National politics - 0% 56% Youth disappearances 0% 0% 54% Theft 36% 34% 40% Inter-ethnic tension 70% 32% 32% Violent extremist groups 19% 26% 32% Lack of education - 2% 28% Criminal gangs 11% 8% 18% High levels of unemployment 28% 12% Land disputes 64% 20% 10% High levels of poverty - 8% 8% 19 Response options for this question were expanded for the midline and endline evaluation to enable easier analysis. Options are marked with a dash during the baseline in cases where these options were not available. 78 Inter-religious tension 4% 0% 6% Radicalization 17% 20% 4% Local politics 34% 2% 2% Other 30%* 0% 0% * Cattle rustling, illicit arms trade, drought, drug abuse It is worth noting that the data collection process was conducted at the height of a polarizing nationwide political campaign dubbed ‘Building Bridges Initiative’, commonly referred to as BBI. BBI was the result of a reconciliation between contesting parties after the 2017 election. It was designed as a national guide to resolve current and longstanding grievances that were caused by political extremism and poor governance. BBI proposals sought to uplift the economy, address issues of corruption at the county level, strengthen institutions and advance inclusion – all uniting issues at the national level, but implementing the recommendations required a referendum and public awareness campaigns. However, politicians who viewed this as a tool to advance their political agenda rallied for or against it. The two competing groups used public awareness campaigns as platforms to amalgamate political capital in readiness for the next general elections. In the first quarter of this year, political rallies were hosted by both camps, across counties including Isiolo and neighboring towns. The support for or against BBI was viewed by political, religious and community leaders as a way of positioning themselves to lobby for their interests and that of their respective constituencies. This situation could have escalated political tensions in Isiolo County, where political leaders have been accused of stirring ethnic conflicts to gain political mileage in past elections. Political debates and conversations around this were resounded in all media outlets, at formal and informal gatherings in the markets, bus stops and social occasions. This could be the reason why national politics were mentioned frequently by respondents between January and March 2020. 9.1.3 Disappearances Youth disappearances in Isiolo are reported to be a major problem. 56% identified youth disappearances as a major security threat, representing a sharp increase over the year. This could be attributed to some cases of recruitment that were reported in December 2019 and, perhaps more so, to the increased reporting of such cases to relevant authorities. Respondents noted that the channels of communication between security actors and community members had improved a lot in the last year, making it easier for communities to know when a young person disappears, because there is less stigma or risk in reporting and talking about it. This perhaps explains why respondents were much more aware of or willing to talk about disappearances compared to previous years. Although disappearances were not identified as a major security threat in Isiolo County during the baseline and midline, respondents did acknowledge that they did occur with some frequency. Five respondents highlighted recent reports of disappearances, one key informant noted that: “We only hear that there is no recruitment and radicalization, but there are still young people joining the extremist groups. There are young people who have left for AS a few months ago and must have been recruited in the community.” Most respondents believe that there have been intensified efforts from the authorities and community leaders to enhance vigilance, including strengthening the Nyumba Kumi and community policing initiatives, and intensifying security patrols to respond to any reports. 9.1.4 VE and Recruitment 62% of respondents tended to agree that extremist activity and recruitment into extremist groups was more of a problem now than it was 12 months ago. A CBO representative noted that, although 79 security measures had been increased after the 14 Riverside attack last year, there has been a laxity by the community and security actors in past few months. He noted that: “Last time recruitment went down after government intervention, but now it has once again evolved after efforts slowed down.” While 26% of respondents disagreed with the above statement, 12% believe there has been no change in the context. While perceptions were generally positive in this regard, respondents were more positive in their opinion regarding this statement than at the time of midline. There seems to be more awareness whenever a case of recruitment or disappearance has occurred; most who referred to recent cases pointed to the same timeframe, stating that this happened over the past few months. Those who believe recruitment is less of a problem say so for a number of reasons. First, some believe that the government, overall, is more alert; more youths are deterred by the stories told by returnees, and there is better coordination and sensitization on issues related to VE. Storytelling drawn from returnee testimonials seems to have been significant in dissuading youths from joining VE groups. Several respondents, including a youth leader, note that there is a better understanding among the young people that extremism is not a way out of unemployment or idleness, and that no one has found a better life from joining such groups. They note that CVE and counter terrorism activities were heightened after the 14 Riverside attack but seem to have lost momentum in the recent months. There is an acknowledgement that community members may not have the information to determine the level of recruitment, and that they are not equipped to rule it out as being a significant problem. When determining the threat of VE in Isiolo, two respondents referred to the 14 Riverside attack as the main indicator that VE is a threat to the community. 9.1.5 Gang and Criminal Activity 64% of the survey respondents believe that crime and gang activity is less of a problem now than it was one year ago, while only 28% of respondents disagree. A majority of those who agreed with this statement were security actors, followed by private sector and county administration representatives. However, of this percentage, only five respondents strongly agreed that crime and gang activity is less of a problem. Those who disagreed cut across all sectors. While views on organized crime and gangs in Isiolo are generally positive, this finding points to a very minimal change in the perception compared to midline and baseline. Organized crime and gang activity are less of a problem now than it was 12 months ago. Strongly agree 14% Somewhat agree 50% No change 8% Somewhat disagree 22% Strongly disagree 6% Widespread police presence, such as patrols and better cooperation between the police and communities (to be discussed in subsequent sections in this report), is credited with contributing to the reduction in criminal activity. A representative of a women’s group in Isiolo town pointed to NPRthe Kenya Police Reserve (KPR) deployment as a reason for the reduced crime within the town. In contrast, a religious leader who disagrees with the statement, noted that although there seemed to be an increase in activities by security actors, and some policemen are insincere because they continue to take bribes from those reported to be engaged in crime. Those who saw no change simply say that the levels of crime or gang activity remain inconsistent and it varies year around. 80 Still, nearly one third of respondents still believe crime in Isiolo is a problem and, although various respondents highlighted drug use as a cause of insecurity, they also noted that drug-related crime is often petty and sporadic. A respondent further noted that organized gang activity is not a threat within Isiolo town due to increased government surveillance; while one respondent pointed out that cattle rustling and livestock theft continue to plague more remote areas, specifically mentioning the Samburu highway, which is a hotspot for stock raids. 9.1.6 Safety and Security 90% of respondents agree that residents in Isiolo feel safer now than they did 12 months ago. No respondent disagrees with this statement, while only 10% believe there has been no change; the respondents come from the media, education institution, a CBO and youth representative. The score for this question represents only an increase from the baseline and midline score, perhaps due to the heightened security and community engagements over the past few months, which have given a sense of confidence in the security ecosystem in Isiolo County. Community members feel safer now than they did 12 months ago. % of respondents who agree Baseline 79% Midline 78% Endline 90% There was broad agreement that the frequency of attacks and incidents in Isiolo had declined in comparison to the previous 12-month period, in which a number of incidents took place in the lead up to the attack on the 14 Riverside complex (one of the attackers had connections to Isiolo). While there have been no extremist attacks in the county, recruitment remains a threat. Respondents do not necessarily perceive VE activity (including attacks or other VE related incidents) to be a significant threat, but they still acknowledge that recruitment is ongoing. At the national level, 48% of respondents believe that the number of attacks countrywide has increased, which is 12% less in perceptions shared in the midline. This could be attributed to the rise in sporadic extremist attacks along the coastal and north eastern region, and threats to Kenya by AS between January and March 2020 that heighten security activities across the country. Still, 20% of respondents agreed that the number of attacks has decreased, while 12% saw no change, pulling the average score for this question much lower compared to both baseline and midline scores. Those disagreeing cut across all categories, though only one county administrator agreed. At the time of data collection, February through March 2020, there was a spike of isolated violent extremist attacks in the North and Coastal areas of Kenya, which could explain the perception of respondents on the threat of terrorism at the national level although the community does feel safer now. Over the past 12 months, the number of terrorist attacks throughout Kenya has decreased. % of respondents who agree Baseline 98% Midline 60% Endline 48% 81 9.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced The table below compares scores for the sentinel indicators from the midline and endline evaluations on indicators that track communities’ willingness to mobilize in support of peace and security objectives at the county level. While NIWETU does not work to support these indicators, we assume that communities’ willingness to engage with these structures, or their trust in these structures’ ability and relevance, will affect their perspectives on the wider community’s willingness to mobilize around CVE. Individual scores reflect respondents’ personal views on the corresponding statement, while the ‘meta’ scores, or meta perceptions scores, reflect how respondents believe their peers would answer. Research has shown that respondents are more likely to act in accordance with what they think are their peers’ views, or the social norm. This means that when meta scores are higher than the individual scores, views in general tend to be more positive about a particular statement, and vice versa. These findings come from a sample of 40 individuals in two neighborhoods in Isiolo Town and should therefore not be considered statistically significant or representative of a wider sample of Isiolo residents. Figures presented below are out of a ten-point scale where 1 represents strongly disagree and 10 represents strongly agree. Degree to which peace committees are active and responsive. Midline Endline I feel that the peace committees are active and easy to access in my area. Indiv. 4.49 5.06 Meta 3.72 5.94 I feel that the peace committees have an adequate understanding of the issues in my community. Indiv. 4.21 5.27 Meta 3.89 6.28 I feel that peace committees effectively respond to issues in my community Indiv. 3.62 3.08 Meta 3.34 6.24 Overall score 3.87 5.31 Degree to which communities actively engage in barazas. Midline Endline I frequently attend community barazas. Indiv. 3.92 4.45 Meta 4.32 5.31 When I go to barazas, I feel like my opinion is considered and my concerns are responded to. Indiv. 3.64 4.32 Meta 4.17 5.37 I always feel positive after participating in barazas. Indiv. 5.80 4.66 Meta 6.02 5.64 Overall score 4.65 4.96 Degree to which Nyumba Kumi initiatives are active. Midline Endline Nyumba Kumi initiatives are active in my community. Indiv. 3.34 4.62 Meta 4.08 6.02 I feel that Nyumba Kumi initiatives are positively addressing the issues that concern me. Indiv. 3.97 5.41 Meta 4.71 4.28 I have confidence in Nyumba Kumi initiatives and trust that it will help improve peace in my area. Indiv. 5.61 4.67 Meta 7.11 4.26 Overall score 4.80 4.88 Presence of other organizations/actors that work on peace/security. Midline Endline There are numerous other networks and organizations making a positive impact on peace and security in my community. Indiv. 3.56 3.48 Meta 4.18 4.25 Overall score 3.87 3.87 The findings in this table suggest, in general, perceptions of structures that exist at the county level to address peace and security issues have improved from the midline evaluation one year ago. During the midline, perspectives on peace committees tended to be rather negative, while the endline shows 82 more neutral perspectives on the relevance and effectiveness of these structures. Individuals’ views on the effectiveness of peace committees decreased slightly from the midline, but their thoughts on their peers’ perceptions increased dramatically, suggesting that, as a whole, the community still finds peace committees to be structures that effectively respond to the issues. Both individual and peer perceptions of barazas also increased marginally, though views still remain rather neutral on the effectiveness of barazas. Likewise, overall views on Nyumba Kumi improved slightly, though there was a noticeable decline in the degree of confidence that respondents (and their peers) have in Nyumba Kumi. The decline in peer perceptions on this question was significant; while at the midline, respondents believed their peers were quite confident in Nyumba Kumi’s ability to improve peace (which suggests that respondents themselves would likely harbor the same degree of confidence); at the endline, respondents believe their peers would answer that question more negatively. Overall, however, views on Nyumba Kumi still remain neutral. It could be expected, however, that allowing more time following the conclusion of the grant to the Isiolo County Commissioner’s office, which trained female Nyumba Kumi representatives (in addition to other female security stakeholders), before measuring these perceptions, could result in an even further improvement in views of Nyumba Kumi’s effectiveness. Once residents begin to interact more with Nyumba Kumi representatives related to CVE matters, we could see a further improvement in their views. 9.3 Section III – Skills and Knowledge The following section looks at the skills and knowledge that exist within the stakeholder community on VE and CVE related activities. The primary focus is on analyzing the different types of training that exist on these topics, as well as the degree to which respondents have participated in such activities already, and what, if anything, they are doing with the skills and knowledge thereafter. 9.3.1 Summary of Findings Many respondents have participated in trainings on CVE in Isiolo township, including NIWETU￾sponsored trainings, most of which were run by Isiolo Peace Link. All participants got at least some utility out of the trainings, stating that they shared important messages and lessons about extremism and methods to counter it. Respondents acknowledged efforts made to reach more of the at-risk in the community, especially youth and affected families, but noted that there needs to be more focus on similar groups in rural areas in the county. Recognizing the need to disseminate the key messages from the trainings received, respondents across the board indicated sharing what they have learned with others in their personal or professional networks. 9.3.2 Skills and Knowledge Index Scores The following table shows the average scores on a four-point scale20 from a number of key questions related to skills and knowledge. Baseline and midline scores are presented side-by-side to highlight changes over the past three years and demonstrate a slight increase in the score for those who have participated in CVE skills and knowledge trainings, but minimal change in perceptions of the utility of these trainings. 20 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 83 Questions Baseline Score Midline Score Endline Score Have you participated in any training on CVE and understanding the causes of VE in the last 12 months? 2.72 2.92 3.20 How much, if at all, do you think these trainings have helped you understand VE and how to prevent it? 3.81 3.88 3.50 How much, if at all, do you use the information you have learned from these trainings? 3.81 3.84 3.50 Have you shared the information you have learned with others in the community? 4.00 4.00 4.00 Overall average score 3.58 3.66 3.55 9.3.3 Participation in Training 78% of all survey respondents have participated in trainings on countering violent extremism. Those not participating in trainings represent individuals in the community/religious leader, educational institution, media, private sector and security sector categories. County administration and youth, women’s groups, and CBO representatives were most present in trainings on CVE issues, while those from the private sector and media participated in these activities the least. The researchers noted confusion in what is meant by training in the midline phase, as some respondents interpreted consultation meetings as trainings. Category Did not participate Participated Women’s group 0 6 Youth 0 6 Community-based organization 0 5 County administration 0 5 Security sector 2 5 Educational institution 3 4 Community/Religious leader 1 4 Media 2 3 Private sector 3 1 Grand Total 11 (22%) 39 (78%) Of those who participated in training, all but one was involved in NIWETU-sponsored training. Twenty-seven mentioned participating in the Isiolo Peace Link sponsored activity; nine participated in SCORES; six in the training via the County Commissioner’s office; and two in the Isiolo Women of Faith trainings. One participant, one of the champions from a women’s group, participated in a training conducted by the Rural Agency for Community Development and Assistance (RACIDA); while one respondent from the private sector could not remember the organization that led the training. 80% of all respondents noted that they participated in a wide range of trainings. 71% participated in trainings on understanding violent extremism, while 64% trained in understanding why people join VEOs. Over half participated in trainings on ways of countering violent extremism. None of the survey respondents has ever participated in monitoring and evaluation training. There was a steady increase in the number of those who participated in trainings for trainers, how to report extremist activities, and working with thegovernment to prevent or counter violent extremism, as compared to midline. A significant number of respondents noted that they have participated in other types of trainings as well. Most engaged in training on the role of different community stakeholders, such as women, parents and local CSOs, in countering violent extremism. Others said they participated in trainings on recognizing early warning signs of radicalization, the impact of violent extremism, and on providing opportunities for youth. One respondent mentioned participating in a session that focused on 84 understanding the Isiolo County Action Plan (CAP) and how to implement some of its pillars. This could have been part of the dissemination activities under the ICAP. Types of Training Baseline Midline Endline Understanding violent extremism 45% 50% 71% Why people join violent extremist groups 38% 56% 64% How to prevent or counter violent extremism 47% 41% 62% Other 23% 53% 38% How to identify recruiters and recruitment methods 34% 6% 36% How to talk to others about violent extremism 30% 16% 27% How to report extremist activities 39% 9% 22% Training others on preventing or countering violent extremism 17% 0% 22% How to work with the government to prevent or counter violent extremism 17% 6% 16% How to lead interfaith discussions on violent extremism 32% 6% 16% How to work with other organizations to prevent or counter violent extremism 17% 13% 11% Advocacy skills 6% 0% 9% Identifying signs of recruitment 31% - - Leadership skills 6% 0% 7% Conflict resolution 11% 0% 7% Monitoring and evaluation 4% 0% 0% Most respondents recognized that trainings on understanding violent extremism, why people join extremist organizations, and how to prevent violent extremism, have been the most helpful. Key informants, however, emphasized that, despite the increased number of those trained, the distribution is largely within Isiolo town and especially among those who are well-networked with community leaders. The choice of venue for such training and engagement is also restrictive, as costs for hotels and conference halls limit the number of participants, and such events rarely reach those most at-risk, or most in need in more remote and rural localities such as Merti and Khina. This perception could have been shaped by the fact that interviews were conducted within the town and the respondents might not have been aware of CVE activities conducted by NIWETU and partners in remote areas like Garbatulla and Merti, among others. 78% of those who participated in training programs stated that the training helped them carry out their regular work, while over half of them mentioned that they gained new knowledge that helped them strengthen their existing skills in this area. In the midline, respondents stated that the trainings have been particularly helpful for the security sector, teaching them to handle VE related issues “more humanely” and in better coordination with the community. They further noted that, for these trainings to continue to be effective and helpful, they must be offered continuously and should target diverse community leaders within and outside of the town centers. Two thirds of participants acknowledged that the training had helped them understand violent extremism a lot more, and all agreed that they have used what they learned through these trainings at least a bit, demonstrating that the topics and issues covered have been beneficial to participants in the sessions, in strengthening skills and knowledge on these key issues. Respondents pointed to a number of important ways in which they have used the skills and knowledge gained: 85  Youth and CSO representatives report using lessons from the training to talk to peers in both formal and informal spaces, such as football clubs, churches, entertainment spots and social media platforms, specifically WhatsApp and Facebook.  Media representatives note designing radio and TV programs, informed by some of the learnings and stories shared during the training. Other journalists have created spaces for youth and invited ‘experts’ on C/VE, including CSO representatives and security agents, to talk about these issues on radio talk shows. A media personality stated, “as a TV presenter I have invited youth to express their feelings thus increasing their self-worth. You see most of them are idle and unemployment is the main push factor.”  Some participants acknowledged that the information acquired, and interactions made during the training, have helped them construct better counter-narratives and strengthen their capacity as facilitators at the community level. The most notable has been dispelling the falsehoods about the perceived benefits of recruitment among youths.  County administrators and community leaders mention use of training in strengthening community security structures, for example engaging affected families, and improving messaging during community barazas. A women’s leader reported using the knowledge to counsel and connect affected families to get help. In general, most respondents report sharing what they have learned with others, particularly with the youth viewed as most at-risk, and using both formal and informal spaces available to them to disseminate key messages on CVE. Because of their engagement and interactions on issues of CVE, some have noted gaining more confidence from affected families, especially parents. 9.3.4 Improving Training Programs While respondents broadly agreed that the trainings in which they participated have been effective in improving their skills and knowledge around CVE and gaining a wider understanding of VE issues, survey respondents and key informants alike outlined a number of ways in which training programs could be more effective. Most recommendations for improving training focused on the need to reach different and a wider network of individuals. Respondents stressed that many trainings target specific – and according to key informants, often the same – individuals or stakeholder groups, and fail to reach the most at-risk or those closest to the at-risk. As such, they made the following recommendations: Enhanced outreach. Participants proposed that the training be held in towns like Merti, Khina, Garbatulla and Sericho, where at-risk groups are not knowledgeable on CVE and more isolated than those in Isiolo town. This could mean decentralizing points of convening from hotels within town, to public meetings and other community spaces in more rural areas: elders’ meeting sheds, maskanis, DSTV entertainment joints and football fields, which are common in both urban and rural town centers. Such spaces may not be favorable for women in rural areas, but other venues such as school, Koran center, churches and homes that may be hosting chama meetings could be considered. “Yes. The community believed the government was their number one enemy. The CVE meetings/forums/activities have helped, but we still need more to be done. These activities have mainly 86 taken place and target people within the municipality - the three wards – Bulla Pesa, Wabera and Burat. Other people in the rural areas also need this information, activities need to be taken to wards in place like Garbatulla, Merti and Oldonyiro”. – Key Informant from CSO Category They also noted the need to ensure that there is more ethnic inclusivity, representing at least all the major communities (known as the ‘Big 5’ – Turkana, Somali, Meru, Borana, Gabbra). One of the participants noted that it is important to: “Ensure that every community is represented. Somalis are mostly not invited.” Youth inclusion. A few recommended the use of social media platforms to reach out to younger people, both in and out of school, but to do so with more interactive messaging and languages that speak to the youth across different backgrounds and cultures. For example, translating some of the concepts into local languages and using storytelling to communicate efficiently, especially in more rural areas. A respondent noted that there should be more youth facilitators and youth-friendly facilitation methods that allowed for more interactions, peer learning and experience sharing. One example highlighted was the use of returnees as a strategy that could help young people make decisions when faced with a dilemma and are not open to talk about it. This is because most young people will understand and be more accepting of narratives that they find most relatable, because those presenting them have gone through it. In addition, a youth participant recommended, in return, that such a program would support returnees to reintegrate. In contrast, one participant from the media category cautioned that, before engaging returnees, there should be a discussion with authorities on how to identify genuine returnees lest such training sessions idolize returnees who have come back with ulterior motives, or act as points for initial contacts by recruiters. With drug abuse, unemployment and low levels of education being one of the biggest challenges among the youth, three participants recommended that CVE training include a component of youth empowerment, or be integrated within other youth engagement programs, such as entrepreneurship and livelihood skills training. A security actor echoed sentiments shared in the midline and requested that future trainings be increased and be made more consistent; while one other participant further emphasized that the trainings which have been happening are one-to-two-day sessions over a few months in a year, with little engagement in between, which does not allow for continuous learning. A participant recommended that there be more TOTs to build additional capacity of local facilitators who understand local context and language across different regions for sustainability. Respondents also indicated the specific trainings they believe would be most useful. The findings demonstrate a sustained interest in more trainings on the topic, but a slight reduction from the midline. Over 60% of survey respondents pointed to the need to better understand why people join VEOs and understand the nature of VE. A significant number of respondents suggested trainings on how to prevent or counter VE, training others on such activities, how to work with others, and how to talk to others about extremism are needed. What types of trainings would be most useful? Baseline Midline Endline How to prevent or counter violent extremism 46.8% 56% 66% Why people join violent extremist groups 27.7% 78% 64% Understanding violent extremism 31.9% 80% 60% How to talk to others about violent extremism 46.8% 52% 38% How to identify recruiters and recruitment methods 27.7% 50% 32% How to report extremist activities 25.5% 50% 30% 87 Training others on preventing or countering violent extremism 34% 54% 22% How to work with other organizations to prevent or counter violent extremism 27.7% 54% 20% How to work with the government to prevent or counter violent extremism 27.7% 48% 16% Advocacy skills 14.9% 26% 12% Identifying signs of recruitment 29.8% 0% 0% Leadership skills 10.6% 28% 10% Other 34% 32% 8% Conflict resolution 25.5% 30% 8% How to lead interfaith discussions on violent extremism 27.7% 46% 6% Monitoring and evaluation 10.6% 12% 2% Other training topics noted include:  Developing media content, production policy and guidelines on reporting on CVE  Digital communication, especially targeting chiefs, security actors and other local leaders  Impact of VE on other communities outside Isiolo town  Understanding the ICAP and how the different pillars could be implemented  How to strengthen economic opportunities for youth  Working with and engaging returnees  Entrepreneurship and livelihood skills (confirmed across the board by key informants)  Countering religious narratives and applying religion in psychosocial counselling  Addressing drug and substance abuse (seen as a key driver of vulnerability)  How to implement the CAP pillars across the county. 9.3.5 Sharing Lessons Learned As discussed above, most respondents reported the importance of sharing what they have learned with others in their networks, and highlighted that the ways in which they have used their newly￾acquired skills and knowledge, and the way for trainings to be more effective, is to share what is gained with a wider group of people, including the at-risk and other stakeholders and individuals encountering these issues frequently. As such, all respondents who participated in trainings reported sharing what they learned with others. Over 50% indicated sharing what they learned with less formal but close networks, such as friends and schoolmates, while 32% shared with family members. More than half of respondents noted disseminating what they learned to the wider community, perhaps through barazas or other community dialogues which respondents referenced, and 43% mentioned colleagues at their organization. No respondents mentioned sharing with local government officials. Twelve of those respondents indicating that they shared the lessons learned with other community members, noted that they discussed these issues specifically with youths, affected families, and a returnee seen as more vulnerable. They also specifically mentioned sharing what they learned with self-help group members, madrassa students, neighbors and a customer. Throughout the baseline, midline and endline, sharing of lessons learnt between respondents and both local government administration and business partners remains relatively low. Perhaps this could be the limited or structured interaction with county administration officials and minimal engagement of the private sector actors respectively. From the table, respondents tend to share information much more with those they have frequent interactions with. 88 With whom have you shared your new skills and knowledge? Baseline Midline Endline Community members 44.7% 50% 61% Friends/school mates 2.1% 68.8% 50% Colleagues in my organization 55.3% 65.6% 43% Family members 10.6% 21.9% 32% Other 21.3% 6.3% 32% People from my mosque/church 23.4% 18.8% 30% Community leaders 36.3% 25.0% 27% Colleagues at other organizations 27.7% 34.4% 9% Business partners 4.3% 0% 2% I have not shared the information or skills with anyone 0% 0% 2% Colleagues in the local government 29.8% 6.3% 0% 9.4 Section IV – Community Networks The following section assesses the existence and strength of community-based networks working to counter extremism in Isiolo. A brief summary of survey responses, followed by an analysis of the findings, is presented below. 9.4.1 Summary of Findings Networks have noticeably strengthened in Isiolo over the past twelve months. With the county’s CEF Forum, organized through NIWETU, actors working on the key issues related to extremism and countering violent extremism have started working together more closely, especially with youth and women organizations. Unlike the midline, there has been a significant decrease in the frequency of conflicts between organizations, but a few issues have remained, especially around duplication of efforts. This improvement is credited to better coordination through the County Engagement Forum, and increased networking during CVE activities, strengthening relationships among actors. When looking specifically at the benefits that result from stronger networks, a number of key themes emerge in the endline. First, information sharing and knowledge of the VE threat is improving; second, there is greater skills transfer across actors; and third, more actors are engaging in the crucial field of countering violent extremism. 9.4.2 Community Networks Index Score The table below highlights the average scores of five questions (on a five-point scale21) related to establishing community networks to counter violent extremism. On average scores have slightly decreased compared to the midline, but indications are that more organizations are collaborating even much more than before. The scores also indicate perceptions of slightly less conflicts and disagreements between stakeholders, though 30% of respondents stated that they did not know, which could mean they did not feel comfortable sharing, or they were simply not well informed enough to be able to express an opinion. A majority still place value in the value of networks and relations created while working together on CVE issues. Questions Baseline Score Midline Score Endline Score Do you work with any other organizations on CVE? 4.14 4.84 4.30 21 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 89 Do you collaborate more, less, or the same amount with these actors than you did 12 months ago? 3.29 3.88 4.30 How often are there conflicts or disagreements between different actors working on CVE?22 4.24 4.21 4.45 Overall, do you think improved networks between organizations working on CVE has been helpful? 3.68 4.57 3.78 Overall average score 3.84 4.38 4.21 9.4.3 Collaboration Between Actors All respondents acknowledged working with others to at least some extent on CVE programming. Only one respondent from the media category noted that they collaborate with others in CVE matters with less frequency, noting that he has received fewer invitations from other CVE actors in the last 12 months. Which organizations do you work with? Baseline Midline Endline Youth organizations 40% 72% 56% At-risk youth - 25% 51% Security sector actors 57% 34% 44% Women’s groups 32% 54% 42% Religious leaders 53% 40% 40% Community leaders 57% 30% 33% County or national government actors - 22% 26% Teachers 17% 16% 14% Other 17% 8% 14% Other community members 40% 26% 9% Self-help groups 4% - 7% Business owners - 7% Over half of respondents mentioned working with youth organizations, and youth at-risk. Women’s groups, security actors, religious and community leaders were identified as categories of actors with which respondents collaborate. More respondents noted having worked with security actors than in the midline, which reaffirms testaments of increased cooperation between security and community and/or CSOs in Isiolo. With the ICAP launch, more spaces traditionally occupied by security actors were opened up to CSOs and community leaders, and more efforts were placed in improving existing community security processes, such as community policing. Other actors mentioned included specific non-governmental organizations, such as Isiolo Peace Link, Caritas, NOWSUD, and Coffey (through the REINVENT program). Other categories mentioned included media, students, affected mothers and community health volunteers. Several respondents acknowledged that collaborations between CVE actors and the youth have increased in the last year, because young people and youth leaders are more receptive to CVE conversations than before. This is attributed to the strengthened relationship between youth and security actors and availability of safe spaces with less stigma for youth at-risks and affected families to channels concerns through networks, knowledgeable and trusted community leaders – spaces that are more accountable to the society, such as CEF. Another reason given for this improvement is that CVE messaging has been more effective in demonstrating the negative impact of engaging in VE. 22 Higher scores for this question indicate fewer conflicts while lower scores indicate more conflicts. 90 “There are more youth and women being invited to CVE events, while also there are more actors engaging in CVE activities, the County Engagement Forum engages more stakeholders who are involved in the county action plan.” – CBO Representative A security sector actor quoted that she: “Built more confidence among the youth who have also accepted that they can work with the police.” Improved interaction between police and youth is helping CVE actors strengthen their understanding of VE trends and tactics in the county. Being the most targeted group, youth are believed to come across information on VE activities that security agencies and stakeholders may not access to. A representative from the media notes that this is a big step in understanding and solving the problem: “Most of the information comes from the youth, e.g. drugs, real stories of crossing, attempts, etc. so we tend to focus on them.” A youth leader from Tullu Roba credits his increased collaboration with youths in the community to the fact that he has been more exposed to knowledge on opportunities for engaging the youth on CVE issues. There were several respondents who acknowledged the value in interpersonal relationships built across various participants in NIWETU-supported and other CVE activities. The meetings and workshops have not only facilitated frequent interactions that have helped actors identify synergies but have also established friendships and trust. A community leader noted that: “There have been friendships created between the government and the religious and community leaders in the community. The religious leaders have can now speak freely without fear of being arrested by the police.” This trusted link has also reduced the stigma of affected families and returnees, and some of the participants note that their improved networks and understanding of CVE issues has helped then bridge the gap between such groups and other CVE actors. “To me I am in between affected parents and other people, if they can't see me, they are not ready to go to some of these forums and they fear being victimized, but when we are all around there is a feeling of safety.” – Participant from the Women’s Group Category In implementing ICAP pillars such as education, respondents noted that collaboration with the relevant ministries and authorities, such as the Ministry of Education Science and Technology, was a requirement. The more they expanded their programs to target schools and students, the more engagement is maintained with school administrators, relevant CSOs and both national and county governments. Additionally, coordination of CVE activities by CEF-facilitated collaborations helps to reduce duplication. Frequency of collaboration in comparison to 12 months ago Baseline Midline Endline A lot more 33% 31% 58% A little more 30% 41% 28% The same amount 17% 17% 2% A little less 0% 10% 2% A lot less 0% 2% 9% Do not know 20% 0% 1% 91 Respondents noted increased collaboration between actors over the past year, with just over 50% noting that they collaborate a lot much more now than they did 12 months ago. The five respondents who reported having collaborated less in the last 12 months, said that this was a result of a decline in CVE programs in Isiolo, due to decreased funding and high turnover of individuals, especially youth, to other social development programs. A respondent from the media category quoted that: “Activities where in high gear last year, but this year there has been a reduction.’’ While some related less funding to reduced collaboration among CVE actors, one respondent noted that the limited funding for CVE activities was motivating organizations to collaborate with others in order to leverage on each other’s resources. A CBO representative adds that: “There are more opportunities to collaborate in the community, because more organizations are implementing CVE activities.” In general, respondents noted that collaboration has dramatically improved, perhaps because of increased spaces for CVE actors in Isiolo, and that has improved the trust and common reference point for such activities in the ICAP. As such, relationships with the government are strengthening, as the community feels more secure and open to sharing information on issues related to VE, while those at-risk or affected have found more confidence in seeking help. Findings related to government capacity will be addressed in a later section of this report. Despite the increased interaction and collaboration, the frequency of conflicts between organizations has decreased tremendously, perhaps due to better and more institutionalized coordination led by the CEF. Over 70% agreed that improved networks between individuals or organizations working on violent extremism is very helpful, with only two respondents noting that it has not been useful to them at all, but both agree there has been more coordination in the last 12 months and new ideas on CVE have been designed as a result. Specifically, they pointed to a number of effects of greater coordination and collaboration. Almost two thirds agreed that more actors are learning about violent extremism, and more than one third noted that different groups are supporting each other and their activities. Only one respondent, a private sector representative, stated that he has not noticed any real changes as a result of enhanced collaboration. Other changes noted by a respondent from the CBO category was reduced duplication of activities. Effects of more collaboration Baseline Midline Endline More actors are learning about violent extremism 23% 55% 64% There is more coordination between different actors 26% 16% 48% We are learning new skills from each other 23% 14% 44% More actors are working to prevent or counter violent extremism 26% 33% 38% We are sharing our successes and lessons learned 26% 4% 34% Different groups support each other and their activities 45% 39% 32% New ideas on how to prevent violent extremism are forming 38% 14% 18% Other 21% 6% 2% Do not know 11% 6% 2% There are no real changes that I have noticed 2% 2% 2% 92 93 9.5 Section V – Community-Driven Strategies and Initiatives This section addresses the community-driven CVE strategies that are underway in Isiolo. A summary of the survey findings will be followed by a deeper analysis of the findings. 9.5.1 Summary of Findings The findings presented above demonstrate a somewhat mixed view on the degree to which community-driven strategies and initiatives are taking place in Isiolo. While a number of formal and informal activities have been identified by survey respondents and key informants, the distribution of these activities is uneven. Being a transit route for VE organizations, remote communities that are also at-risk of recruitment have not engaged on CVE as widely as those in Isiolo town. In addition, while some respondents believe that the threat of extremism in Isiolo is still evident, CVE efforts in the community seem to have decreased as compared to the midline, perhaps due to NIWETU’s wind down over the last few months. The community must continue to engage on these issues and remain vigilant. 9.5.2 Community-Driven Strategies and Initiatives Index Score The table below presents average scores from four key questions on a five-point scale23 about the effectiveness of community-driven initiatives to counter violent extremism. The findings from the endline evaluation demonstrate a slight decrease from the time of midline. While respondents agree that local organizations are working with various stakeholders to develop CVE programs and that these programs are effective in teaching about CVE, there was a slight decline in scores with regard to the degree to which CBOs are active now as compared to 12 months ago. Still, there was an increase in the extent to which community members engage with these organizations. These findings are somewhat reflective of some respondent sentiments expressed above, indicating that fewer initiatives are happening in the CVE space and could be as a result of a decline in funding for CVE, as donors are pushing for more sustainable and locally resourced approaches, including working through existing community security structures. Although, it was expected that, in the aftermath of the 14 Riverside attack, Isiolo could see a spike in engagement around these issues – that is yet to be reflected. However, it could also mean that CVE approaches are more integrated informally into other processes that are not necessarily activity driven, such as engagements at the household level. Fewer activities from CBOs could also be explained by the increased partnerships across stakeholders, meaning that similar activities are merged. Questions Baseline Score Midline Score Endline Have local organizations worked with other stakeholders to develop programs on CVE in the last 12 months? 4.56 4.83 4.25 How effective do you think these programs are at teaching others about VE and CVE? 4.33 4.38 3.96 Are CBOs more or less active now than they were 12 months ago in implementing such programs? 4.19 3.96 3.44 To what extent do community members engage with these organizations? 4.68 3.72 4.43 Overall average score 4.44 4.22 4.02 23 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 94 9.5.3 Participation in NIWETU-Sponsored Activities Participation of respondents in NIWETU-supported CVE initiatives was limited to 14 respondents, or 28% of the total. Those involved include respondents across all categories, with a concentration of CBO representatives, education institutions, security sector actors, community/religious leaders and women’s groups. 50% of those involved mentioned participating in IPL-sponsored activities while others mentioned activities implemented by Isiolo County Commissioner’s Office, RACIDA, SCORES, and Isiolo Women of Faith (IWOF). 9.5.4 Community-Driven Strategies All but nine respondents, or 82%, agree that organizations are working with other stakeholders to develop CVE programs and strategies. Four individuals, including a community leader and an education professional, disagreed with the statement, while five respondents did not know whether or not organizations are working with other stakeholders to develop CVE programs and strategies than 12 months ago. Survey respondents identified a number of specific CVE-related activities occurring in Isiolo. Half of the respondents stated that sports activities are taking place. The most well-known CVE activities occurring include sports activities, informal conversations in the community, training sessions, meetings with the security sector and other community organizations and formal meetings with community leaders or members. Overall activities seem to have increased as compared to the midline, apart from meetings with government authorities (both national and local). There has been a notable increase in mentorship and discussions in schools related to CVE. CVE Activities Occurring in Isiolo Baseline Midline Endline Sports activities 28% 36% 50% Informal conversations in the community 30% 32% 46% Training sessions 23% 30% 46% Meetings with the security sector 43% 28% 40% Meetings with other community organizations 40% 16% 38% Formal meetings with community leaders or members 62% 30% 36% Meetings with religious leaders 62% 20% 34% Discussions in schools 32% 10% 30% Mentorship programs in schools 13% 14% 24% Meetings with government authorities (local or national) 32% 20% 18% Advocacy campaigns 30% 6% 6% Do not know 2% 4% Other 26% 10% 2% Role playing 4% 0% 2% There are no activities on this topic taking place 4% 0% 0% Chief’s baraza 2% 0% 0% 9.5.5 Effectiveness of CVE Programs 39% of respondents noted that trainings were the most effective approach, stating that they provided opportunities for sharing experiences, especially among youth at-risk, and CSOs are keen to exchange ideas to strengthen their activities. It is also worth noting, as earlier mentioned, that respondents did confuse training with formal meetings hosted on CVE. Therefore, the explanations given on the effectiveness of formal meetings would probably be linked with those of training sessions. A female security actor noted that cash incentives or allowances made training most effective because it encouraged people to turn up. Cash incentives help to motivate continuous engagements and facilitate follow-up. 95 36% of respondents found informal meetings effective. However, when asked to speak to their effectiveness, only three respondents acknowledged that such meetings were very effective, while mentioning that trainings, formal meetings with other CBOs, community/religious leaders and security actors, were more effective. This could be explained by the fact that outcomes from informal meetings are difficult to track and discussions often do not immediately result in practical decisions about community issues, as more formal meetings would. In qualifying why respondents considered formal meetings as very effective, most noted that such sessions often result in actionable plans, since they involved people with influence and authority in the community. They also credited the effectiveness to the ability to consolidate a diverse idea, not only from the community, but also from external actors. A respondent noted that the ability of those engaged in formal CVE meetings to facilitate the implementation of proposals and disseminate information to a wider constituency, is important. Notable examples highlighted are those that bring security actors and youth, or community leaders together, because they provide a safe space to share information and offer a certain level of assurance for protection. Sports activities for example were viewed as having offered an opportunity for security actors to increase positive interaction with youth and foster trust. ‘’Dialogue forums between the police and youth removed fear as well as the games, the two sides asked questions and shared more friendly moments.’’– Representative from the CBO Category However, respondents did acknowledge that the level of understanding about CVE is still low in most parts of the county, because formal CVE activities are mainly held in Isiolo town and dissemination to more remote areas is minimal. Respondents agree that informal meetings which could include both planned and unplanned social meetings such as women social groups, joints frequented by youth, meetings with friends and other community members, such as are a more efficient way to expand the reach to remote areas simply because they are less costly and can be adapted to suit the local languages and social setups in rural areas. This could also mean that community leaders do not have to travel all the way to town to participate in CVE activities, and that CVE programs are integrated in the existing community social platforms such as Parent Teacher Association (PTA) meetings in local schools, women’s self-help groups, elders meetings, joints were youth convene to watch football or meetings of association members of local traders, such as boda boda drivers. They also allow community members, especially those at risk, to listen in and share the information more informally without fear. It is also important to note that both formal and informal CVE engagements seemed to complement each other. An example given was that of sports activities between the youth and the police, that address issues discussed in formal spaces during sporting activities. Both parties were given an opportunity to have an organized dialogue after the games are over. “Football creates a bond between the youth and police and becomes a pathway through which the two groups can share information.” – Community Leader Most effective CVE initiatives Baseline Midline Endline Training sessions 15% 19% 40% Informal conversations in the community 6% 23% 35% Meetings with the security sector 21% 21% 31% Sports activities 23% 29% 25% Discussions in schools 13% 6% 19% Formal meetings with community leaders or members 34% 19% 15% Meetings with religious leaders 34% 13% 15% Meetings with other community organizations 13% 10% 15% Mentorship programs in schools 6% 10% 10% Other 15% 13% 6% 96 Meetings with government authorities (local or national) 21% 4% 4% Advocacy campaigns 17% 4% 2% Do not know 0% 2% 2% Role playing 6% 0% 0% Overall, the perception on the effectiveness of these strategies was pegged to the ability to:  Enhance collective decision making and experience sharing among diverse and influential stakeholders, with the ability to scale the discussions to their constituencies in the community, especially those at risk. This is seen to enhance trust and transparency in how the community and stakeholders engage on CVE.  Target and gain confidence of those most at risk, specifically the youth. Informal activities are more youth-friendly, because the methods give the youth an opportunity to share their knowledge with their peers.  Ensure inclusiveness and relevance of those engaged. Strained resources and the low levels of awareness on CVE in the community require leveraging well informed, trusted, but also influential, community leaders who can be a link to a larger constituency and safeguard the correctness of messaging. 9.5.6 Level of Engagement Although respondents point to a decline in the frequency of CVE activities in the county, 70% of respondents believe that CBOs are more active on issues related to CVE now than they were 12 months ago. This could mean that there are more organizations participating in shared platforms or networks but does not necessarily translate to an increase in activities being carried out by individual organizations. This finding represents a slight improvement from midline findings, in which about 64% of respondents believed that CBOs were more active on issues of CVE than they were 12 months before. Those disagreeing include community and religious leaders, education professionals, journalists, women and youth representatives, county administrators, and one security sector official. Respondents had a range of views on the types of actors that are most effective in implementing CVE activities. More than half of survey respondents acknowledged that youth and women were most effective, followed by faith-based organizations, cultural and arts groups, and self-help groups. While 56% of the respondents note that communities are engaging with these organizations much more than they did before, 39% believe that engagement is limited, while 2% do not believe they engage at all. Three respondents, a community leader, security actor and a private sector representative, had no opinion about this. To what extent do community members engage with these organizations? Midline Endline Engage a lot 40% 57% Engage a little 56% 39% Do not engage at all 4% 2% Do not know 0% 2% 97 9.6 Section VI – Objective 2: Improved Government Capacity The table below compares scores for sentinel indicators from the midline and endline evaluations related to Objective 2. Scores for the meta perceptions indicators are again on a ten-point scale24 and reflect the same sampling dynamics as the Objective 1 questions above. Scores for the statements regarding legislation and budgets were identified through a review of the CAPs and County Integrated Development Plans (CIDPs) which spell out the county budgets. Level of government and community participation in community-government dialogues Midline Endline When I express my views about violent extremism to members of the government, I feel like my opinion is considered and my concerns and ideas are responded to. Indiv. 3.07 6.84 Meta 7.11 5.49 The government has made a strong effort at participating in community dialogue over the past two years. Indiv. 7.95 5.35 Meta 4.61 2.59 I feel like I have been adequately included in the county government’s response to peace and security. Indiv. 4.98 3.12 Meta 3.62 6.52 Overall Score 5.22 4.99 9.7 Community perceptions of the national and county government on issues relating to peace and security Midline Endline The national government has made a great effort at improving the peace in my area over the last two years. Indiv. 5.44 5.72 Meta 5.38 4.51 The county government has made a great effort at improving the peace in my area over the last two years. Indiv. 3.47 5.13 Meta 3.44 6.04 The security sector has made a great effort at improving the peace in my area over the last two years. Indiv. 5.13 5.87 Meta 5.02 6.69 My community has become more willing to cooperate with the security sector over the past two years. Indiv. 6.44 7.56 Meta 6.42 7.54 Overall Score 5.09 6.13 Midline Endline Presence of line items for CVE activities in county budgets 1.0 1.0 Presence of legislation to support implementation of the CAPs 0 0 Views on levels of engagement in community-government dialogue declined slightly from the midline evaluation, with the most notable decline appearing in views that the government has made a strong effort at participating in community dialogue over the past two years. Respondents’ views on their peers’ perspectives to this question were rather negative, though their own beliefs appear to be neutral, still suggesting that there is a lack of faith in the government’s willingness to engage in community dialogues. The meta perception score increased most significantly on the statement about feelings of inclusion in the county government’s response to peace and security, suggesting that there is perhaps more engagement on that level. Findings from this research corroborate those sentiments, though there is evidence of a greater willingness to share information and work together on these issues; even if individuals themselves might not feel wholly comfortable engaging with county government on these issues, they at least believe their peers are, pointing to an improvement. Perspectives on the national and county government’s engagement on peace and security issues improved from the midline evaluation. Views were particularly positive on the community’s willingness to cooperate with the security sector, showing an improvement from an already positive midline score 24 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 98 for this statement. Findings from this research will again support this sentiment, as there have been noticeable improvements in information sharing and cooperation across actors on CVE related issues. The presence of line items for CVE in county budgets remains the same, expectedly. The CIDPs outline county spending priorities over a five-year period, and while adjustments and realignments can be made every year, intensive lobbying efforts are required. The slow pace of developing and enacting legislation, and perhaps the lack of buy-in from parliamentarians to support legislation on CVE at the national level, likely explains why there has been no change over the past year on the presence of CVE-related legislation. 9.8 Section VII – Government Responsiveness to CVE The following section presents findings on the responsiveness of county and national level government to address VE. Summary findings are presented, followed by a more in-depth analysis of the findings. 9.8.1 Summary of Findings Relationships with the government in Isiolo have generally improved, and so has the perception of the county government’s willingness to engage on issues of CVE. The ICAP process profiled the role of the county government with more clarity and, although frustrations remain, particularly with regard to the level of engagement of county government officials, relationships with national government counterparts has improved compared to the midline. This finding is perhaps reflective of the confusion that remains over the level of government mandated to work on security issues; while the CAPs are county-owned and county stakeholder-driven, county officials have limited ability and resources to proactively engage. While relationships with the national government representatives in Isiolo are improving, relationships with county officials appear stagnant. 9.8.2 Government Responsiveness Index Score The below questions are scored on a four-point scale 25 to understand overall government responsiveness to VE at both county and national levels. Overall, the findings demonstrate a slight improvement across the board in perceptions of both county and national government’s willingness to work on issues related to VE and their understanding of the issues. Perspectives of the county government’s involvement in these issues, however, still fall short of views on the national government, with respondents expressing generally more positive views toward the national government on these questions. This difference will be addressed in more detail in this section. Questions Baseline Score Midline Score Endline Score How well do you think the county government understands the problem of VE in this community? 1.79 2.56 2.59 How well do you think the national government understands the problem of VE in this community? 3.62 3.72 3.70 How willing do you think the county government is to work on issues of VE? 1.80 2.00 1.92 How willing do you think the national government is to work on issues of VE? 3.36 3.68 3.50 How easy do you think it is to work with national government counterparts? 3.20 3.43 3.71 Overall average score 2.75 3.08 3.09 25 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 99 It is also important to note that respondents – both in the survey and key informants – viewed the security sector as a proxy for the national government. This is likely due to the fact that the national government functions, including security, are coordinated by the County Commissioner at the county level. 9.8.3 Government Understanding of VE 67% of survey respondents acknowledge that the county administration’s understanding of the problem of VE has improved, while the remaining either do not know how well the county government understands this issue or holds more negative views on county officials’ grasp of the problem. Those expressing more negative views represent respondents across all categories. How well does the county government understand VE? Percentage Midline Percentage Endline Very well 14% 33% Somewhat well 36% 24% Not very well 36% 24% Not at all 10% 12% Do not know 4% 6% As the research found in the midline evaluation, key informants’ perspectives echoed these findings; some noted that confusion over which agency has the mandate for CVE programming still remains, particularly in the aftermath of the launch of the County Action Plan (CAP). They noted that county government officials still fail to show up to any CAP follow-up meetings, such as the County Engagement Forum sessions, and are not taking an active role in CVE activities in Isiolo. A few key informants stated that this is likely because security issues, including CVE, still fall under the mandate of the national government and the county government does not have the resources to engage in or support CVE programming. Still, scores for this indicator represent a marked improvement from the midline. National government, on the other hand, is seen to have a stronger grasp of VE issues, according to Isiolo respondents. 86% of respondents believe that the national government understands the problem of violent extremism very well, as five respondents note that the government understands somewhat well. Only two respondents, from education institutions and the private sector, strongly disagreed. 9.8.4 Willingness to Engage in Programming Although 56% of respondents are pessimistic in their views on the county government’s willingness to work on CVE-related issues, there is an increase in the confidence that the county has made efforts to engage more. 36% express optimism compared to 28% during the midline. Those believing county government is very willing are distributed across all categories. Respondents from the education institution, youth, community leaders and CBOs category, were the most pessimistic, as compared to those from the county administration and security sector. How willing are county government is to work on CVE? Percentage Midline Percentage Endline Very willing 4% 32% Somewhat willing 24% 4% Not very willing 30% 22% Not at all willing 32% 34% Do not know 10% 8% Respondents were much more confident in the national government’s willingness to work on issues of CVE. Over 90% of all respondents believe that the national government was committed to working 100 on CVE. Despite the somewhat negative views on the county government’s willingness to engage in CVE programming, perspectives on the national government were much more positive than views toward county government. Those who expressed more negative views on the national government represented the education, media, and women’s group categories. As was found in the midline, no respondent believed that the government was not at all willing to tackle these issues. How willing are national authorities to work on CVE? Percentage Midline Percentage Endline Very willing 70% 62% Somewhat willing 28% 28% Not very willing 2% 8% Not at all willing 0% 0% Do not know 0% 2% Key informants suggested that, because of the improved relations with the security sector, local residents have a more positive outlook on the national government and its willingness to productively work with communities on this issue. While they acknowledge that some tension remains, there was broad-based agreement that the relationship between Isiolo residents and the security sector is continuing to improve. Indicators key informants highlighted, included the enhanced reporting by the community and improved police-youth engagements. 9.8.5 Engagement with Government Counterparts Engagement with government counterparts occurs on a somewhat regular basis in Isiolo, but there are differences in the degree to which respondents engage with the county and with the national government. Generally speaking, engagement on issues related to CVE appeared to be more common with national government than county government. 34% of respondents noted that they engage with the national government on a daily basis, while 16% engaged the national government several times a month. While for about 48% of respondents, engagement with the county government is rare. Frequency of engagement with national government counterparts Baseline Midline Endline On a daily basis 14% 8% 34% Once a week 2% 4% 2% A few times a week 2% 12% 16% Once a month 7% 4% 6% A few times a month 8% 22% 6% Occasionally 51% 31% 20% Rarely 8% 23% 16% Despite perspectives on limited engagement with the county government, a key informant noted that there could be an assumption that the county is less interested, as it prioritizes its support for CVE on indirect interventions to address vulnerability to VE, such as bursaries and enhancing youth employment opportunities. For most respondents, this support, addressing drivers of VE, might not be easily considered as a CVE activity and is an intervention that may not necessarily lead to frequent interactions with CVE actors in the county. Key informants who mention having worked with the county government on CVE issues, report that county government participation is related to attending meetings when invited and does not necessarily involve demonstrating active leadership. One key informant from the CSOs categories acknowledged that they work with the county government, particularly a member of the County Assembly and municipal administrator, at the village level. 101 Frequency of engagement with County government counterparts Baseline Midline Endline On a daily basis 0% 4% 8% Once a week 0% - 4% A few times a week 0% 2% 8% Once a month 0% - 4% A few times a month 2.1% 4% 4% Occasionally 18.8% 22% 24% Rarely 72.9% 68% 48% A key informant in the national government, however, did note that in the initial stages of the ICAP formation, they worked with the county government who supported the data collection process and provided financial support in hosting large gatherings. Still, a key informant notes that the one of the greatest challenges in implementing the ICAP, has been the disinterest from the county government which is generally expected to take more leadership alongside the CC. On many occasions, the county government has failed to attend CVE-related meetings and, when they do, their participation is limited, and they do not seem to own the process. A key informant stated that there was a sense of disappointment among stakeholders about the lack of political will from the governor – even after promising to allocate a budget to support ICAP pillars, no substantive steps have been taken to see this through. The county government has blamed this on the fact that it does not yet have the right policies in place to allow for such allocation. On the other hand, respondents to a great degree acknowledge the ease in working with the national government on matters related to CVE. Respondents note that the county commissioner’s office, including those at the local level such as chiefs and sub-chiefs, are very responsive, approachable and easy to work with. One respondent, however, places a disclaimer that this does not apply to police, who are seen as part of the national government structure. Respondents highlighted some of the key challenges they faced in working with government counterparts. They noted that county officials do not involve them in their activities and do not value their contributions, with only 6% acknowledging having had no issues at all. 30% of respondents pointed to other challenges faced when working with county officials. Some noted that they are not accessible, especially those in the Governor’s office, are marred by corruption, politicize CVE, and are not willing to fund ideas presented by CVE actors in the community. Others noted that the county government seems to have a clashing mandate with the national government over security issues and their role is not very clear. Two respondents noted that the county government only chooses to engage in issues that benefit officials personally or help them build political capital. A member from the county government admitted that, sometimes working closely with the national government officials on issues of CVE, is at times labelled as having taken sides on either national or county politics. Challenges Midline Endline Other 46% 30% They do not value our contributions 0% 26% They do not involve us at all 18% 24% They do not try to talk to us 9% 18% They do not show up to meetings that we schedule 14% 8% They impose programming without consultation 0% 8% There are no problems working with the government 14% 6% Do not know 5% 4% They prioritize law enforcement and arrests 0% 0% 102 Although challenges working with the national government exist, 96% of respondents noted that it was somewhat easy to very easy working with national government representatives on issues of CVE. However, when asked about the challenges they face, only 6% acknowledge not having had difficulties engaging national government. Some felt that the national government does not involve them, try to reach out, or value their contributions. This finding should be considered in context, however. It is likely that, because of the national government’s limited presence in Isiolo, there are few government representatives and those who are present, struggle with competing priorities further limiting opportunities to work together. So when they do, these targeted engagements are likely to be more productive. Another explanation would be the fact that, although it is easy to access the lower levels such as chief and assistant chief, it is not necessarily the same case with the CC or his deputies. The engagement of the national government on CVE is also channeled through the CEF and relevant county security committees, reducing multiple consultations outside the forums. Those who mentioned other challenges spoke of the red tape involved in engaging national government, while two respondents from the education and youth categories felt that the CC’s office was not very receptive to the youth. No respondent mentioned engaging with national government representatives out of Nairobi. 103 10. ANNEX IV – MANDERA COUNTY REPORT: MANDERA COUNTY ENDLINE OVERVIEW Data collection took place in Mandera Town over 11 days in February and April 2020, carried out by two researchers. NIWETU facilitated introductory meetings to key stakeholders to explain the goals of the research and to facilitate further introductions to relevant actors. The research team conducted both qualitative and quantitative research with key stakeholders in Mandera, as well as the wider community. The breakdown of respondents in the stakeholder questionnaire is as follows: Category of respondent Number of respondents Youth representatives 6 County administration 6 Community-based organization 8 Educational professionals 5 Media 3 Private sector 4 Security sector 5 Women’s groups 6 Religious and community leaders 7 Total 58 Key informant interviews were carried out with eight individuals, including county government representatives, NIWETU staff, CSOs and influencers. Interviews were not secured with national government counterparts, because of the complexities of conducting remote fieldwork as a result of travel restrictions in response to the Covid-19 pandemic. 104 11. KEY FINDINGS 11.1 Section 1 – The VE Context The following section presents findings of the broader VE and security context in Mandera County. 11.1.1 Summary of Findings This research points to several different perspectives concerning the VE context in Mandera. While key informants and stakeholders tended to be largely supportive of an improved context to VE, there were concerns that the spillover effect from the conflicts in Somalia would continue to threaten ongoing security mechanisms if not resolved consultatively. AS is still active in Mandera and is changing tactics to focus more on remote areas and places that were not known to be hotspots, such as Olla and Banisa. Youth remain vulnerable, especially herders. While disappearances have reduced, complainst persist over the role of security officers in such disappearances. Threat from criminal gangs remains less of a concern than it was during the baseline and midline. 11.1.2 Causes of Insecurity As was with the baseline and midline, respondents still believe that violent extremist groups continue to be one of the biggest security threats faced by residents in Mandera, alongside high levels of unemployment, poverty and drug abuse. There was less concern about the threat of local or national politics, criminal gangs, or theft. Although violent extremist groups continue to manipulate inter￾religious tension among communities in Northern Kenya, including Mandera, no endline respondents identified inter-religious tensions as a threat to security – perhaps as a result of religious homogeneity of the Muslim community (which is also reflected in the composition of respondents). Other security threats mentioned include porous borders with Ethiopia and Somalia, marginalization of some community members, drought, lack of freedom of movement, lack of identity cards, disconnect between security actors and community, cross-border related insecurity, poor infrastructure, corruption and prostitution. Security threats in Mandera Percentage of Respondents Baseline Midline Endline Violent extremist groups 80% 81% 80% High levels of unemployment 72% 76% 66% Drug abuse - 81% 46% High levels of poverty - 64% 40% Lack of education - 43% 32% Inter-ethnic tension (Interclan) 86% 54% 28% Radicalization 12% 67% 26% Youth disappearances 70% 54% 22% Land disputes 58% 52% 14% Local politics 58% 48% 4% Theft 16% 22% 4% Criminal gangs 4% 12% 4% Other 2% 9% 2% National politics 10% 12% 2% Inter-religious tension 6% 9% 0% *Respondents interpreted “inter-ethnic” tensions as intra-ethnic/interclan. However, the data still captured interclan variables. Respondents highlighted the impact of the recent row between Kenya and Somalia as a significant threat to security in the county, particularly since the conflict has spilled over into Mandera town, 105 facilitated by the porous border. At the time of conducting the fieldwork, tensions were high in Mandera town over fears that the government of Somalia was planning to attack forces from the semiautonomous state of Jubaland, who were rumored to have sought refuge within the town. Fighting erupted at the Bula Hawa border town in Somalia right after the quantitative data collection was completed. Key informant interviews were conducted immediately after fighting ceased. “This Jubaland issue has truly affected us. During the recent fighting by SNA and the Jubaland militia that was housed in Kenya, the grenades struck some houses in Mandera town, cars were burnt, a stray bullet killed one person, and several people were injured. One young man with a bullet lodged on his body was referred to Nairobi for removal of the bullet but the medics there refused to attend to him, and he was taken to Awassa, Ethiopia where the bullet was removed. This is not a healthy environment for CVE programming and the community’s perception of the government could be affected as a result.” – Key Informant from the District Peace Committee As a result of the tensions and fighting, some residents had also moved away from the border to seek safety in other parts of Mandera town. This row has also been complicated by the escalation of an ongoing martime dispute between Kenya and Somalia. A key informant noted that, the perception by the community that the Kenyan government was covertly providing refuge to the fugitive Jubaland Minister, Abdirashid Hassan Abdinur, also known as ‘Janan,’ accused of human rights abuses, had rolled back gains made in engaging the communities in fighting VE. The diplomatic rift between Kenya and Somalia had not only created insecurity along the border but was also creating a conducive environment for Al-Shabaab activities in Kenya, as security forces appear to be distracted by the row between Kenya and Somalia. A key informant noted that the threat of insecurity from these political tensions had diverted CVE efforts by actors and local leaders. The perceived position of the GOK in providing refuge for Janan has also driven a wedge between the national and county government of Mandera representatives. “For the period tensions were, the CVE programs were halted and compromised by both the CSOs and security agencies as the focus went to the border tension. Kenya government supporting Jubaland forces and hosting them on Kenyan soil to scale up attacks in Bula Hawa was an upfront provocation that was opposed by the residents of Mandera.” – CSO Key Informant Although 64% of respondents note that terror attacks in Mandera have reduced, key informants fear that the tension and movement of Jubaland militants along the border in between February and March 2020 may have created a loophole for AS militants from Somalia to enter Kenya more easily. A key CBO informant noted that: ‘’There has been confusion over the identity of who between Jubaland forces and Al-Shabaab forces had been using routes outside the town to cross in and out. During the same time attacks occurred and people killed in the Kenyan side, but no one knows who was responsible for the attacks.” Some of the respondents were interviewed at a time that was marked by sporadic but isolated terror attacks across the North-Eastern region, mostly aimed at targeting non-local residents. A community leader stated that: “There was a time when this has reduced considerably, but since the beginning of 2020, we have seen an increase in terror activities. We still have incidences, e.g. today we had an attack on a bus.” There were concerns over the increased insecurity in remote areas and especially sub-counties profiled as VE hotspots. Authorities and other actors have had reservations about covering areas perceived vulnerable to VE such as Lafey leaving a loophole for recruitment. 106 “There are wrong perceptions created about some places like Lafey sub-county that it is not secure. This led to a situation where no one goes the place without an escort. This has made the place to be a very soft target for VE recruitments.” – Security Sector Representative 11.1.3 Disappearances Disappearances continue to be a problem in Mandera County, but there were mixed reactions from the respondents. Comparing baseline to endline, there has been a marked increase in respondents who believe that disappearances never happen, shifting from 0% at baseline to 36% of respondents at endline. Conversely, 18% of endline respondents believed that disappearances happen somewhat or very regularly, a significant decrease in percentage since baseline when more than 75% of respondents reported regular disappearances. Those who were most affirmative include a CBO representative, security actor, community leaders, and county and youth representatives. Overall, results suggest that there has been a consistent decline in the rate of disappearances, or at least the perception of the rate of those disappearances, in the county in the past three years. As was observed during the midline, a significant number of respondents were not able to or not willing to comment on the frequency with which this occurs, as compared to the baseline. It is possible that respondents are more confident that disappearances, especially those linked to government, are reducing, but that there is consistent hesitancy to speak about Al-Shabaab related disappearances for fear of reprisal. How often do youth in Mandera County disappear? Baseline Midline Endline Very regularly 34% 9% 10% Somewhat regularly 42% 24% 8% Rarely 14% 36% 25% Never 0% 9% 37% Do not know 10% 22% 20% A majority of respondents who stated that disappearances occur at least rarely agreed that youth are victims of extrajudicial killings (specifically mentioning ATPU and Kenya Defence Forces), while two respondents believe they join or are abducted by AS. A women’s group representative states that: “A lot of people disappeared in the hands of the government security. AS will just kill people and leave the body for the deceased family to bury their loved ones but the disappearance is only done by the government of Kenya which is worse than killing.” “There are still some youth who disappeared especially from Mandera north but not that much though we don’t know where they are going.” – Youth Representative A representative from the county government, who believes disappearances rarely occur, noted that it is not easy to determine the degree to which this happens, because AS has changed focus and now mostly targets herders who live in remote areas and are not easy to track because they are very nomadic. A key informant, however, notes that although extra-judicial cases and indiscriminate punishment of civilians has reduced: “Occasionally, there are cases of youth disappearances reported. The government needs to come clear and abide by the laws of the land to prosecute suspects in the court of law.” 107 One respondent, a security actor who believes disappearances were very regular, was unwilling to share her thoughts on where such youth go. This could also be an indicator of why many respondents seemed cautious in addressing this section compared to other counties reviewed. The respondent states: “It’s not clear and even if we know we cannot tell as we will become targeted.” As such, of those stating that disappearances happen with at least some frequency, ten respondents stated that those disappearing engage with unlawful groups, such as AS or ISIS. Still, however, just over half of survey respondents did not know the frequency with which recruitment into these groups occurs. 22% noted that recruitment happens at least every few months. In the recent past, AS has launched several abductions in the North-Eastern region, including Mandera. In January 2020, suspected militants raided a village in Fino, Lafey sub-county, Mandera County, and abducted five locals. The most targeted were medics, teachers, security officers, mechanics and community members believed to be more informed about security operations in the area, and who could divulge such intel to the militants. These abductions, however, were publicized and respondents may not have considered them to be disappearances, but VE incidents. Only one respondent, a representative of a community-based organization, who believes that disappearances are rare, quoted that: “Disappearances are at times due to abductions by AS. Some youth working for an NGO were abducted but they were later released.” Those who believe that disappearances never happen state that they have not heard of any disappearances recently, and therefore believe this to be less of a problem. It is worth noting that tracking trends on disappearances is difficult; they happen sporadically and are often not widely reported. As such, it is likely only people with specific knowledge of certain issues are inclined to say that disappearances are still occurring and could suggest that they not happen all that frequently, otherwise a higher proportion of respondents would acknowledge this trend. 11.1.4 VE and Recruitment As it was during the midline, survey respondents are in broad agreement that the VE context in Mandera has improved in the last 12 months, despite discussion around the continued threat posed by the tension at the government and community level, over the conflict between Jubaland forces and Somali Federal Government. Nearly 70% of respondents agreed that extremism and recruitment is less of a problem now than it was last year. Though responses from midline showed an even stronger agreement in the improvement of the VE context since baseline, the endline results are still very notable given the security context within Mandera was unstable during the time of interviews. Extremism and recruitment to join extremist groups is less of a problem now than 12 months ago. Baseline Midline Endline Strongly agree 16% 68% 36% Somewhat agree 38% 19% 32% Somewhat disagree 22% 7% 10% Strongly disagree 2% 0% 10% Do not know 22% 5% 12% Most survey respondents who agreed with the statement credited the reduction in VE activity and recruitment to increased awareness of VE in the community, especially among youth and parents. This, they noted, has changed the mindset of many who now view terrorism as a detractor to development for the county. Credit was given to increased networking between the community, local community- 108 led organizations, local religious leaders, and county-level government that strengthen CVE sensitization in mosques, schools and public forums. “People have opened up, we never used to hear people saying AS directly – they used other terms to describe them such as Ali Garoob, Arsenal but now they say it without fear.” – A CBO Representative As it was noted during the midline, these networks also provide better and more effective reporting mechanisms, that encourage locals to discuss issues of concern directly with community elders or chiefs, who then report up through the appropriate channels as needed. These methods are reported to be effective, because the individuals reporting can do so through other figures they trust, and they are therefore able to maintain some degree of anonymity. While some respondents view increased government surveillance and counter-terrorism measures as a contributor to the reduction of VE activities in Mandera, several respondents disagreed, noting that hardcore measures are marked by atrocities that have led many to join AS to avenge the extrajudicial killings or forced disappearances. A CBO representative, disagreeing with the statement that extremism and recruitment to join extremist groups is less of a problem in Mandera now than it was 12 months ago, also pointed to a change in recruitment tactics: “The number of youths joining these groups has gone down, but the influx of the insurgents is increasing and this time they are moving beyond the borders to the interior of the county. They have changed tact from targeting town people to herders and farmers who are more vulnerable.” As will be discussed later in this report, pieces of training are often limited to select individuals or stakeholders in Mandera town and Lafey, meaning the community at large, and residents in more remote areas, do not adequately benefit from these activities. Still, NIWETU’s support for the Mandera County Action Plan and the county engagement forum has strengthened multi-stakeholder networking and collaboration. This provides an avenue for strengthening local networks and strategies to expand outreach to more remote areas, resolving government and community tensions, and complementing other peace and security in the county. 11.1.5 Gang and Criminal Activity As it was during the midline, gang and organized crime are viewed as less of a problem in Mandera than 12 months ago, with nearly 70% of respondents at endline believing that the situation has improved, in comparison to nearly 50% of respondents at midline, and only 26% of respondents at baseline. Organized crime and gang activity are less of a problem now than it was 12 months ago. Baseline Midline Endline Strongly agree 20% 36% 36% Somewhat agree 6% 12% 32% Somewhat disagree 18% 10% 22% Strongly disagree 22% 40% 6% Do not know 34% 2% 4% Nearly all of those disagreeing with the statement said that this is because gangs and organized crime are simply not a problem in Mandera and do not exist. Still, some survey respondents note that substance abuse is a growing problem in Mandera, especially alcohol coming from the Ethiopian border, 109 and though this may not be tied directly to any organized criminal activities, this is seen as a growing problem in a town that is homogenously Muslim. 11.1.6 Safety and Security Respondents were in broad agreement that safety and security in Mandera have improved in the past 12 months, with 64% of survey respondents stating as much. However, this percentage still represents a decline compared to the midline, likely a result of cross border security issues mentioned above, as well as the increase in VE attacks in the region over the months just before this research commenced. Over the past 12 months, the number of terrorist attacks in Kenya has decreased. Midline Endline Strongly agree 13% 16% Somewhat agree 20% 34% No change 4% 4% Somewhat disagree 7% 20% Strongly disagree 53% 22% Do not know 4% 4% Respondents shared divergent views when asked whether the number of attacks across Kenya has decreased. While 42% of survey respondents disagreed that the number has reduced, 50% agreed. Although there had been several sporadic attacks in other parts of the country, the perception of the frequency of terrorist attacks in Kenya was lower as compared to the midline when the data collection process coincided with the 14 Riverside attack in Nairobi. This could mean that the level of attention given to terror attacks, or the centrality of the incident, does indeed influence the perception of communities on the threat of terrorism nationally. 110 11.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced The table below compares scores for the sentinel indicators from the midline and endline evaluations on indicators that track communities’ willingness to mobilize in support of peace and security objectives at the county level. While NIWETU does not work to support these indicators, we assume that communities’ willingness to engage with these structures, or their trust in these structures’ ability and relevance, will affect their perspectives on the wider community’s willingness to mobilize around CVE. Individual scores reflect respondents’ personal views on the corresponding statement, while the ‘meta’ scores, or meta perceptions scores, reflect how respondents believe their peers would answer. Research has shown that respondents are more likely to act in accordance with what they think are their peers’ views, or the social norm. This means that when meta scores are higher than the individual scores, views in general tend to be more positive about a particular statement, and vice versa. These findings come from a sample of 40 individuals in two neighborhoods in Mandera Town and should therefore not be considered statistically significant or representative of a wider sample of Mandera residents. Figures presented below are out of a ten-point scale26. Degree to which peace committees are active and responsive. Midline Endline I feel that the peace committees are active and easy to access in my area. Indiv. 7.35 6.45 Meta 6.06 6.43 I feel that the peace committees have an adequate understanding of the issues in my community. Indiv. 7.77 5.95 Meta 5.59 6.08 I feel that peace committees effectively respond to issues in my community Indiv. 8.15 6.40 Meta 5.60 6.10 Overall score 6.75 6.24 Degree to which communities actively engage in barazas. Midline Endline I frequently attend community barazas. Indiv. 4.92 5.50 Meta 4.86 6.00 When I go to barazas, I feel like my opinion is considered and my concerns are responded to. Indiv. 3.99 5.91 Meta 4.43 6.09 I always feel positive after participating in barazas. Indiv. 4.16 5.92 Meta 4.53 6.04 Overall score 4.48 5.91 Degree to which Nyumba Kumi initiatives are active. Midline Endline Nyumba Kumi initiatives are active in my community. Indiv. 4.65 6.62 Meta 5.07 6.25 I feel that Nyumba Kumi initiatives are positively addressing the issues that concern me. Indiv. 4.67 6.10 Meta 5.60 6.35 I have confidence in Nyumba Kumi initiatives and trust that it will help improve peace in my area. Indiv. 5.70 6.19 Meta 2.61 6.13 Overall score 4.72 6.27 Presence of other organizations/actors that work on peace/security. Midline Endline There are numerous other networks and organizations making a positive impact on peace and security in my community. Indiv. 1.96 6.25 Meta 3.66 5.75 Overall score 2.61 6.00 26 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 111 Positive perceptions of the role of peace committees within Mandera declined slightly over the past year, but remain positive, nonetheless. There was a notable improvement in how respondents view engagement with community barazas; while during the midline evaluation, respondents tended to be somewhat negative in their opinions about their participation in barazas, they felt more positive about their participation at the midline, suggesting that barazas might be becoming a more relevant platform for sharing and disseminating information in the community. Views on the effectiveness of Nyumba Kumi initiatives also improved from the midline. Since one year ago, respondents were mostly positive in their views on how Nyumba Kumi relates to local communities, perspectives on the degree to which Nyumba Kumi is active, that it is positively addressing issues of concern, and that residents are confident in Nyumba Kumi and trust that it will improve peace, uniformly improved. There was a dramatic improvement in residents’ perceptions about the presence of actors and organizations working on peace and security in Mandera. This suggests that either more actors are working on these issues – which is a likely explanation, as activity and momentum around the CAP continues to pick up – or that residents are better aware of and informed about the activities of CSOs in the county, and are therefore also more likely to engage with them. NIWETU’s work in Mandera can be credited with contributing to the increase in this indicator, as NIWETU continues to support and push forward CAP implementation, actively collaborating with CSOs and Champions in Mandera on CVE related matters. 11.3 Section III – Skills and Knowledge The following section looks at the skills and knowledge that exist within the stakeholder community on VE and CVE related activities. The index scores are presented first, followed by a more detailed analysis of respondent perspectives on skills and knowledge. The primary focus is on analyzing the different types of training that exist on these topics, as well as the degree to which respondents have participated in such activities already, and what, if anything, they are doing with the skills and knowledge thereafter. 11.3.1 Summary of Findings Participation in training activities was widespread in Mandera, and respondents reported finding these activities to be useful in improving their skills and knowledge on CVE. However, the delivery and facilitation mechanisms could be strengthened through increasing follow-up and refresher sessions, while eradicating the duplication of the training content by CVE actors. Respondents and key informants also stressed the need to reach out to more individuals, while also being more targeted to ensure the most relevant and affected groups are being reached, including the most at-risk. Proposals included increasing training activities in remote areas and engaging more youth, especially returnees and those at-risk. Still, respondents felt that trainings were largely hosted in Mandera town and targeted individuals in leadership positions and failed to reach more deeply into the community. Although there is a decrease in the sharing of information at the endline that is also consistent with a decrease in training activities, the dissemination of the knowledge acquired seems to be consistent at both the midline and endline. This demonstrates that knowledge is spread beyond the activity participants, even though activities are only reaching select groups of individuals. Beyond improving skills and knowledge, the training sessions have created an environment for actors to network and talk more openly about VE – something many were not comfortable doing even two or three years ago. 112 11.3.2 Skills and Knowledge Index Scores The following table shows the average scores on a four-point scale27 from several key questions related to skills and knowledge. Baseline, midline and endline scores are presented side-by-side to highlight changes over the past three-and-a-half years. The scores reflect minimal change across the three evaluation periods, but are still overwhelmingly positive, underscoring the importance of skills and knowledge activities to Mandera stakeholders. Questions Baseline Score Midline Score Endline Score Have you participated in any training on CVE and understanding the causes of VE in the last 12 months? 3.1 3.26 3.18 How much, if at all, do you think these pieces of training have helped you understand VE and how to prevent it? 3.82 3.86 3.83 How much, if at all, do you use the information you have learned from these pieces of training? 3.68 3.75 3.79 Have you shared the information you have learned with others in the community? 3.94 3.95 4.00 Overall average score 3.64 3.71 3.70 11.3.3 Participation in Training 88% percent of survey respondents have participated in CVE related training programs in Mandera. Only 5% of these respondents had not attended NIWETU-supported activities. Those who had not participated in trainings include respondents from all categories, except the security sector, women and youth representatives. Private sector and media representatives had the lowest levels of participation in CVE-related training during the endline. Category Participated Did not participate Community/Religious leader 4 3 CBO 5 2 Private sector 2 2 County administration 5 1 Educational professional 4 1 Media 2 1 Security sector 5 0 Women’s representative 6 0 Youth representative 6 0 Total 39 (80%) 10 (20%) 82% percent of survey respondents report participating in NIWETU-sponsored CVE training run by the following partners: RACIDA; Greenland Air and Development (GLAD); Focused Approach Development Concern (FADC); Mandera District Peace Committee; and Mandera Women for Peace and Development. There also appeared to be confusion among some respondents over who sponsored the training in which they participated. Six individuals referenced programs sponsored by Pact, UNDP, Islamic Relief, Mandera Peace and Security Committee, Life and Peace Institute, District Livestock Marketing Council (DLMC) and Mandera Vital Signs, none of whom are NIWETU partners in Mandera. 27 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 113 According to the survey results, respondents have participated in trainings in a wide range of topics. Over a third have participated in training on countering violent extremism, why people join extremist groups and understanding violent extremism. It was least common for respondents to participate in training on monitoring and evaluation, and advocacy skills. Other topics mentioned include improving community-security relations, involving youth on CVE, and managing CVE in an educational institution. A respondent noted having attended a CVE symposium for schools’ board of management, teachers and students as a training. Types of Training Baseline Midline Endline Understanding violent extremism 80% 86% 81% How to prevent or counter violent extremism 51% 96% 79% Why people join violent extremist groups 57% 89% 79% How to work with other organizations to prevent or counter violent extremism 46% 80% 58% How to work with the government to prevent or counter violent extremism 49% 89% 49% How to talk to others about violent extremism 43% 86% 49% Training others on preventing or countering violent extremism 46% 82% 40% Conflict resolution 57% 61% 40% How to identify recruiters and recruitment methods 29% 59% 35% How to report extremist activities 29% 75% 23% How to lead interfaith discussions on violent extremism 26% 73% 23% Leadership skills 51% 66% 23% Other 14% 5% 19% Advocacy skills 20% 59% 14% Monitoring and evaluating 17% 52% 9% Identifying signs of recruitment 37% - - There was almost a consensus on which training was most helpful. Over a third of respondents stated that trainings on understanding violent extremism were the most helpful, and half pointed to training on why people join violent extremist groups, and how to prevent or counter violent extremism. Least helpful were training programs on monitoring and evaluation and advocacy skills. Participants pointed to several different benefits from the training offered. Nearly half of the respondents mentioned that the skills helped them in their regular work and that they had learned about a new subject that is interesting and important. 48% percent of respondents noted that attending the training has been valuable in helping them meet colleagues working on similar issues in Mandera. To that end, all but three respondents noted that the trainings have helped their understanding of the issues at least a little. Those disagreeing include a youth representative, a community leader, and a journalist. The journalist noted that the topics were not beneficial because he had advanced knowledge on CVE issues, but discussions offered a bit more depth and the opportunity to talk about the issues more openly. All respondents who attended the training note that they use the information learned at least a little, with almost a third of them stating that they apply what they learn a lot. They stated that they have used what they learned to train or talk to others about the issues, especially youth, CBOs and religious leaders within and outside of Mandera town. This has been possible through existing community networks, forums, and social media platforms such as Facebook and WhatsApp. A security sector representative states that he continues to sensitize community members and has since ‘’identified a peace ambassador in every sub-county’’, while a county administrative official notes that he trains community members at the village level and has used the knowledge to identify radicalized youth. A women’s group representative, part of the Mandera Women for Peace and Development, quotes that: 114 “It has helped me in understanding early signs of radicalization and gave me the ability to talk to others about the problem of VE issues as part of cross-border women networks activities.” Others note that the strengthened capacity, in terms of skills and knowledge, has helped them network, and opened up more opportunities to contribute more to CVE dialogues, fundraise better and engage additional stakeholders in the community. A representative from the education institution category stated that: “I become a CVE facilitator for FADC and some other CVE actors.” A county government official likewise said: “The information I gained through these trainings helped me develop better concept notes, proposals to solicit funds as Assistant Director County De-radicalization.” 11.3.4 Improving Training Programs Respondents and key informants highlighted many ways in which pieces of training be more effective. A few clear themes emerged. 1. There is a need to scale the pieces of training to more remote and border communities, so that they are more inclusive. This includes areas like Fino, Lafey, Ramo, and other hotspot areas like Banisa, Asabito, Othas. Proposals on how to do this included leveraging County Commissioner office structures, including Nyumba Kumi, elders, religious leaders, women and youth networks. This would require more training for trainers and facilitation to support logistics and training resources. 2. Respondents proposed the use of audio-visual materials for delivering training, increasing the number of training days and consistency of the sessions to bi-monthly from quarterly. A community leader also proposed the use of religious leaders in delivery of some of the sessions noting that “often training is led by people with very weak religious knowledge and background.” 3. Stakeholders need to push for policies that protect the free flow of information and safety for trainers, and those sharing CVE messages or reporting VE activities because, as a youth leader notes, “the community is afraid of sharing information as they will be victimized.” 4. As also highlighted during the midline, training should occur more frequently and build off lessons already learned. To be sustainable, training cannot be a one-off engagement, according to a few respondents. It should also seek to build upon the knowledge already gained so that individuals participating in multiple trainings learn something new, and the community gains a deeper understanding of the issues. A range of topics emerged as potential priority areas for future training. Specifically, respondents mentioned the need for more training on CVE, how to prevent or counter violent extremism, why people join violent extremism groups, and understanding violent extremism. There was a slight decrease in interest in training on how to work with the government on CVE. This could be because 115 there had been increased engagement on this through the CAP process since the midline. Least important was said to be training on monitoring and evaluation. What types of pieces of training would be most useful? Baseline Midline Endline How to prevent or counter violent extremism 76% 78% 67% Why people join violent extremist groups 62% 60% 67% Understanding violent extremism 68% 66% 65% How to work with the government to prevent or counter violent extremism 60% 67% 40% How to work with other organizations to prevent or counter violent extremism 62% 57% 40% Training others on preventing or countering violent extremism 78% 71% 37% How to talk to others about violent extremism 60% 60% 37% Conflict resolution 54% 54% 28% How to identify recruiters and recruitment methods 46% 59% 19% Leadership skills 54% 59% 16 % Advocacy skills 34% 66% 14% How to report extremist activities 34% 55% 12% How to lead interfaith discussions on violent extremism 60% 52% 12% Monitoring and evaluating 30% 10% 7% Identifying signs of recruitment 50% 0% 0% Other 4% 0.0% 0% 11.3.5 Sharing Lessons Learned Sharing what participants learned from the training in which they participated was a key outcome of the activities. All respondents who participated in training mentioned that they have shared with a wider network of individuals. A majority, 88% of respondents, mentioned sharing acquired skills and knowledge on CVE with fellow community members, followed by community leaders and colleagues. Others mentioned having shared with students in schools, and TVET trainees. A CBO representative mentioned sharing through FM radio stations, with the CEF, and the county government de￾radicalization and CVE department. With whom have you shared your new skills and knowledge? Baseline Midline Endline Community members 46% 80% 88% Community leaders 36% 71% 63% Colleagues in my organization 28% 64% 61% Family members 50% 71% 49% Colleagues at other organizations 12% 55% 40% Friends/school mates 14% 75% 33% People from my mosque/church 34% 64% 23% Colleagues in the local government 16% 57 % 23% Other 36% 0% 16% Business partners 20% 41% 12% I have not shared the information or skills with anyone 2% 2% 0% 11.4 Section IV – Community Networks The following section assesses the existence and strength of community-based networks working to counter extremism in Mandera. A summary of survey responses and the index score is followed by an analysis of the findings is presented below. 116 11.4.1 Summary of Findings Networks between CVE actors in Mandera County are growing at both the community and leadership level. They include youth, women, CBO and multi-stakeholder networks. Such networks have strengthened collective responsibility and action, where community representatives have, to some level, had an opportunity to learn from each other, address common issues and co-resource for CVE activities. The CEF is viewed by many stakeholders as a pillar for this coordination, being more inclusive. However, community members at the grassroots may not be well versed on how it operates, even though they are aware of the leadership convening. The improved coordination by actors is still to be appreciated at the community level, since CBOs continue to implement activities unilaterally. Conflict and disagreement between CVE actors seem to have reduced compared to 12 months ago, but a few issues remain, especially on the security actors and the national government’s management of security issues, and competition over funding for CVE programming. 11.4.2 Community Networks Index Score The table below highlights the average scores of four questions (on a five-point scale28) related to establishing community networks to counter violent extremism. Scores indicate a noticeable improvement in the rate at which organizations are working together on CVE, from baseline and midline. There has also been a steady increase in the score, capturing the perceptions on the usefulness of improved networks among organizations working on issues of CVE. Perhaps most significantly, respondents point to a decline in the frequency of conflicts and disagreements between different actors now, than 12 months ago. Questions Baseline Score Midline Score Endline Do you work with any other organizations on CVE? 3.84 3.97 4.80 Do you collaborate more, less, or the same amount with these actors than you did 12 months ago? 3.77 4.72 3.94 How often are there conflicts or disagreements between different actors working on CVE?29 3.43 3.94 4.18 Overall, do you think improved networks between organizations working on CVE has been helpful? 4.55 4.74 4.86 Overall average score 3.90 4.34 4.45 11.4.3 Collaboration between Actors All survey respondents noted that they work with others on issues of CVE, a sizable increase from both the baseline and midline. Most respondents reported collaborating with youth organizations, at￾risk youth, women’s groups, county and national government counterparts, other community members, and religious leaders. It was less common to work with self-help groups, teachers and business owners. These findings point to frequent engagement between a range of actors working in CVE, amongst those indicating that they collaborate with others, and this also supports responses from earlier sections where respondents credited training and other engagements for networking actors in the county. Others mentioned working with students, social media users and international NGOs. 28 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 29 A higher score for this question means that there are fewer conflicts between actors; on a scale of one to five, higher scores all reflect positive changes while lower scores reflect negative changes. 117 What organizations do you work with? Baseline Midline Endline Other community members 42% 81% 73% Community leaders 48% 77% 73% County or national government actors - 86% 69% Religious leaders 55% 84% 69% Security sector actors 55% 88% 60% Youth organizations 71% 91% 58% Women’s groups 52% 86% 54% At-risk youth - 79% 31% Teachers 16% 58% 29% Business owners 16% 61% 25% Self-help groups 7% 40% 23% Other 7% 2% 8% Survey respondents work with these actors largely because of the connections, influence or authority such actors have in the community. They are believed to be people and institutions that have a degree of influence in the community and are knowledgeable about those most vulnerable to VE. Religious leaders, for example, are thought to be very influential and respected as CVE actors, while youth, women, elders and ward administrators have a strong constituency at the village level. A journalist notes that they choose to work with government is because the government is the ultimate authority on issues of peace and security. “We have established CVE champions composed of elders, women and youth and ward admins chair these committees in every ward.” – A County Administrator “Youth at risk are difficult to reach, to ensure responsibility we reach out-mothers are close to their children.” – A Community Leader At-risk communities are seen as important to engage, because they are not only vulnerable but also provide valuable information in understanding and countering VE in different parts of the county. Asked why he chooses certain actors, a county administrator who chooses to work with youth, women, security sector and business owners, states: “Youth are most affected, women are used as conduits to recruit young people, businesses pay zakat to insurgents as they attempt to secure their goods from Somalia. In general, these are the most important people that can play a key role in CVE and of course - the security sector, who have a security role as their prerogative and without their involvement, our efforts will be futile.” Other reasons mentioned included the willingness of other actors to work together, familiarity with those they chose to work with, or stipulated policies for engaging stakeholders within specific activity plans. A CBO representative notes that his choices are guided by the organization’s plan, stating that: “As a stakeholder, we attempt to approach every actor in terms of their role and needs. In the case of youth, women and self-help groups, we attempt to provide start-ups and vocational skills such as tailoring. We also target teachers and parents to ensure there is the enrolment of out of school children aged between 7-13 years.” “These are members of my social circle and are those whom we work together.” – Private Sector Representative As was highlighted during the midline, others simply stress the importance of a holistic approach to promoting peace and security, meaning that they must work with a wide network of actors to be effective. 118 With the recognition of the importance of coordination on CVE – and broader peace and security matters – 58% of survey respondents who indicated that they work with others, stated that they collaborate at least a little more now than they did 12 months ago. Only three respondents, a journalist, private sector and CBO representative, stated that they collaborate a little less. The rest stated that there has been no change in the degree to which he works with others on these issues. One private sector and CBO representative noted that the reason for working less with others was because there were fewer CVE activities and insufficient funds in the last 12 months. Reduced activities did not necessarily limit opportunities for collaboration; as one CBO representative noted, the county has seen more organizations network and strive to coordinate than before. A key informant from a CBO also added that, due to increased sporadic AS activities, more organizations and community leaders are working with CVE actors as links to more remote areas. “Heightened security operations by AS which has moved from the periphery across the border to the interior. Before the presence of the militia group was common in border areas such as Lafey, Kotulo, and Mandera South but now it is moving Banisa, Rhamu – areas that previously had no AS presence.” – A County Administrator Official A youth representative noted that government and security sector actors are now more accessible and available for meetings on these issues. People are now more aware and willing to contribute to the discussion on these issues. A community leader stated: “There are a lot of programs on CVE. At least every organization in the account has a component of CVE they work on. Because of the increased awareness, people are talking about it more.” A community leader noted that, despite the will to collaborate with the government, there have been a few challenges. He narrates that: “As a Chairman of the Minority communities in Mandera, I often try to call for meetings that target members of our communities and we try to meet though there is a lack of resources to conduct regular meetings. Changes in administration systems and the transfer of government officers affect the partnership.” The midline pointed to somewhat poor coordination, indicating that although there were existing peace networks, the degree to which organizations and other entities collaborate was limited and with a relatively high degree of duplication. One key informant sees the establishment of the CEF as a critical and more inclusive point of coordination of CVE actors in Mandera, quoting: “CEF has been the center of coordination of CVE activities which are run orderly in Mandera. The forum has become the face of CVE in the county as it has provided a central guidance base. The sector of CVE been in high demand stakeholders have been running helter-skelter with no right information.” 11.4.4 Conflict Between Organizations According to respondents, conflict happens relatively infrequently and is a significant reduction from the midline. Only 16% of the survey respondents stated that it occurs at least somewhat often, while a resounding 54% of respondents said that it never happens. Still, 24% of survey respondents either did not know how often this occurs or chose not to evaluate this question. How often do disagreements occur between organizations working on CVE? Midline Endline Very often 5% 2% Somewhat often 12% 14% 119 Rarely 16% 6% Never 37% 54% Do not know 30% 24% The CEF provides a platform for coordination to avoid duplication and collectively address some of the sources of conflicts highlighted above. A key informant noted that the platform brings together stakeholders to update each other on their planned activities or priorities, explore opportunities, and resolve disagreements amongst themselves. Respondents pointed to a few factors that have contributed to disputes between CVE actors. Most noted competition over resources and duplication of activities as the main sources of conflict. A CBO representative noted that organizations and individuals may not be open to speak about conflict and disagreements, for fear of being requested to contribute more in a collaborative process noting that: “Most happen in a silent way caused by a lack of openness in terms of budgets (most organizations conceal their resources and it’s hard to have a joint effort plan) – there are attempts to engage each partner and request them to share their budget.” 120 11.5 Section V – Community-Driven Strategies and Initiatives This section addresses the community-driven CVE strategies that are underway in Mandera. A summary of the survey result area index score is presented, followed by a deeper analysis of the findings related to community strategies and initiatives. 11.5.1 Summary of Findings Community-driven strategies, though still common, are seen by stakeholders to have slightly reduced from the time of the midline assessment, although are still relatively higher from the baseline. While some respondents believe that there are fewer activities taking place, they found formal meetings with community leaders and members, and training, quite effective. Still, both stakeholders and key informants recommend that leaders and networks at the grassroots receive more support to translate the knowledge and skills at the most remote and local levels in the society. This was also seen as a sustainable way to ensure continued engagement and learning. Key informants noted an improvement in community-security engagements, due to the CVE forums that provided community members and leaders with an opportunity to engage security actors on issues such as extrajudicial killings and disappearances – topics that were previously viewed as controversial, or those who raised such concerns were previously victimized. This has also improved information sharing between community and government on issues of VE. Likely as a result of NIWETU’s more active engagement in Mandera over the past year, respondents were quick to point to NIWETU-sponsored activities, including various training programs, the CAP engagements, meetings via the CEF, and other local strategies, such as the CVE Champions for Change program in Mandera town. 11.5.2 Community-Driven Strategies and Initiatives Index Score The table below presents average scores from four key questions, on a five-point scale30, about the effectiveness of community-driven initiatives to counter violent extremism. These findings point to a slight decrease in the degree to which local actors are supporting and engaging in community-driven strategies and initiatives, compared to the midline. While engagement is still frequent, this slight decrease is reflected in all questions asked in this section. Still, the overall score remains relatively high. In earlier sections, respondents agree that CVE activities have declined over the past 12 months and this score could be a reflection of that. It is also worth noting was that, during the midline, the score on the degree to which local organizations were working with other stakeholders on CVE activities was noted to be artificially high, because 11 respondents did not know. Therefore, the decline observed at the endline might not have been significant if the midline score was not skewed by the high number of respondents who chose not to answer the question. Since there were no respondents who mentioned that they did not know the degree to which CVE actors work together on issues of CVE, it is likely that organizations are now more aware of each other, and are therefore better informed in their response, which can also be seen as a success. Questions Baseline Score Midline Score Endline Score Have local organizations worked with other stakeholders to develop programs on CVE in the last 12 months? 3.44 4.82 4.63 How effective do you think these programs are at teaching others about VE and CVE? 3.86 4.52 3.94 30 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 121 Are CBOs more or less active now than they were 12 months ago in implementing such programs? 2.82 4.28 3.84 To what extent do community members engage with these organizations? 3.70 4.17 4.04 Overall average score 3.68 4.45 4.11 11.5.3 Participation in NIWETU-Sponsored Activities Participation of respondents in NIWETU-supported CVE initiatives was higher than the midline, and a majority of respondents, or 86%, indicated participating in such programs. The majority of respondents mentioned participating in either RACIDA, Mandera Women for Peace and Development, FADC or NAPAD. Others mentioned engagements with the County Deradicalization Department and Champions for Change program. Two participants conflated Danish Demining Group, LPI, Islamic Relief, and RASMI (Regional Approaches for Sustainable Conflict Management, formerly PACT) activities for NIWETU-sponsored programming. 11.5.4 Community-Driven Strategies Approximately 92% of survey respondents believe that local organizations are working together to develop CVE strategies and initiatives, much more so than at the time of the midline. Unlike the midline, where almost 20% of respondents did not know if this was occurring, only two respondents were unaware, indicating CVE programing is becoming more inclusive, or simply that the degree to which coordination efforts turn into tangible action is getting clearer. Have local organizations been able to develop programs and activities to teach others about violent extremism in the last 12 months? Midline Endline Yes 77% 92% No 22% 4% I do not know 3% 4% Survey respondents who indicated that these activities are taking place pointed to a wide range of programs addressing violent extremism. Most respondents state that pieces of training on CVE are the most common activities taking place in the county, followed by formal meetings with community leaders and members. There was a significant decrease in formal meetings between the community and government authorities, religious leaders and community organizations. This could be due to reduced CVE activities, but can also be explained by the fact that although CAP engagements have brought stakeholders together, the same vibrancy is not yet reflected at the community level. CVE Activities Occurring in Mandera Baseline Midline Endline Training sessions 38% 70% 62% Formal meetings with community leaders or members 62% 84% 54% Meetings with the security sector 60% 88% 46% Informal conversations in the community 54% 53% 40% Meetings with religious leaders 64% 79% 34% Meetings with other community organizations 28% 79% 32% Meetings with government authorities (local or national) 38% 81% 30% Sports activities 28% 49% 30% Discussions in schools 20% 46% 28% Advocacy campaigns 18% 51% 14% Mentorship programs in schools 14% 37% 12% Do not know 2% 5% 6% Roleplaying 14% 16% 2% 122 Other - 2% 2% Public forums (Chief’s barazas) 14% - - Respondents suggest that there be more engagement outside of Mandera town, especially in areas that are more vulnerable to VE. A key informant adds that, although there may not have been frequent activities at the community level, there are examples that demonstrate the value of prior engagements with the government and security officials. He stated: “Chiefs and their assistants make public barazas twice a month in public places, such mobilizations have been fruitful over time. For example; a mother who was disposing of garbage in the dry river passing center of Mandera came across a suspicious device which was grenade which was planted under a tree. She immediately called the local chief who also contacted his immediate boss in the County Commissioner’s office. After some time, multi-agency security team arrived the scene and pomp experts detonated the device. This was a typical case of where information sharing has worked from grassroots to higher office and action taken timely.” 11.5.5 Effectiveness of CVE Programs Nearly all survey respondents agreed that these programs were effective; only one respondent did not find them effective, while four respondents did not know how effective these programs were. They include two community/religious leaders, a women’s group member and security sector representative. To improve the effectiveness of such activities, a key informant suggested that there be more support to grassroots organizations and community leaders, to ensure the same processes are decentralized to the lower levels of the society. Some of the organizations that implement these CVE activities are not rooted in Mandera and that makes it difficult to sustain continued engagements at the community level. A C4C participant stated: “There is an organization working on CVE in Mandera County whom the C4Cs has not established a connection with because it operates from their office and have no formal presence within the county. They often attend County Engagement Forum but have no physical address to engage them within Mandera County.” Respondents considered training to be the most effective CVE initiatives, and it also worth noting that training programs were identified as one of the most common CVE activities in the last 12 months. As per during the midline, slightly more than half of the respondents maintained their belief that formal meetings with the community leaders and members, equally, was one of the most effective CVE approaches. No respondent pointed to roleplaying as being effective and the preference of this in Mandera County remains relatively low in the baseline, midline and endline. Most effective CVE initiatives31 Baseline Midline Endline Training sessions 24% 40% 58% Formal meetings with community leaders or members 54% 53% 52% Meetings with the security sector 54% 54% 34% Informal conversations in the community 34% 28% 34% Meetings with religious leaders 62% 58% 30% Meetings with other community organizations 24% 46% 26% 31 Respondents were not limited to just one response on this question; they could indicate several initiatives. 123 Sports activities 28% 28% 26% Discussions in schools 16% 25% 22% Meetings with government authorities (local or national) 24% 46% 16% Mentorship programs in schools 16% 21% 10% Advocacy campaigns 14% 19% 10% Other 7% 2% Roleplaying 10% 9% 0% Do not know 6% 0% 0% Public forums (Chief’s baraza) 8% -0% 0% Specific activities were perceived by respondents to be particularly effective for a range of reasons. Generally, most respondents said that they are effective because they include and target the most relevant stakeholders. Others specifically noted that working with religious leaders, in particular, is the most effective approach, because they are respected and seen as influential by community members. Lastly, several respondents acknowledged that meetings with different community leaders helps to build trust and improve relationships that will hopefully lead to more CVE programming efforts in the future. This helps build a collective view of the problem and potential solutions. According to a security sector representative: “When we meet various stakeholders, we learn from each other and identify weaknesses and find out where we need to do better.” A CBO representative notes that informal and formal meetings which involve community stakeholders are easier to manage, while some level of informality was credited as more desirable especially among youth. The respondent states: “These are the activities and the methods that the local groups can do as they lack capacities and frameworks to measure their successes.” 11.5.6 Level of Engagement All but 11 respondents believe that community-based organizations are more active now than they were 12 months ago. However, only 17 respondents believe that they are much more active. Security sector actors tended to have the most negative views; two believed that activities of community-based organizations focusing on CVE have declined significantly. This perception was also echoed by a youth representative. At the same time, 90% of the survey respondents believe that the community engages at least a little with these organizations, indicating that these actors have important relationships with local communities. To what extent do community members engage with these organizations? Midline Baseline Engage a lot 59% 37% Engage a little 41% 53% Do not engage at all 0% 2% Do not know 0% 8% 124 11.6 Section VI – Objective 2: Improved Government Capacity The table below compares scores for sentinel indicators from the midline and endline evaluations related to Objective 2. Scores for the meta perceptions indicators are again on a ten-point scale32 and reflect the same sampling dynamics as the Objective 1 questions above. Scores for the statements regarding legislation and budgets were identified through a review of the CAPs and County Integrated Development Plans (CIDPs) which spell out the county budgets. Level of government and community participation in community￾government dialogues Midline Endline When I express my views about violent extremism to members of the government, I feel like my opinion is considered and my concerns and ideas are responded to. Indiv. 3.22 6.14 Meta 5.96 6.42 The government has made a strong effort at participating in community dialogue over the past two years. Indiv. 6.39 5.77 Meta 5.08 7.75 I feel like I have been adequately included in the county government’s response to peace and security. Indiv. 5.19 6.04 Meta 4.86 6.52 Overall Score 5.12 6.44 Community perceptions of the national and county government on issues relating to peace and security Midline Endline The national government has made a great effort at improving the peace in my area over the last two years. Indiv. 4.62 6.38 Meta 5.65 6.31 The county government has made a great effort at improving the peace in my area over the last two years. Indiv. 5.96 5.69 Meta 6.24 6.12 The security sector has made a great effort at improving the peace in my area over the last two years. Indiv. 6.09 6.20 Meta 5.70 6.06 My community has become more willing to cooperate with the security sector over the past two years. Indiv. 7.00 5.69 Meta 6.47 7.38 Overall Score 5.97 6.23 Midline Endline Presence of line items for CVE activities in county budgets 2.0 2.0 Presence of legislation to support implementation of the CAPs 0 0 Perceptions on interactions with national and county governments’ response to the issues improved across all indicators. While during the midline, respondents still tended to be somewhat positive, particularly around the county government’s efforts at engaging on the issues and communities’ willingness to cooperate with the security sector, positive sentiments on these and other indicators increased or stayed the same from the midline. The most noticeable change can be seen in the overall score around the level of government and community participation in dialogues, suggesting that efforts to bring together community and government actors around peace and security issues have been effective. Such engagements could occur through CAP-related meetings or other discussions that have happened, particularly in recent months, following an uptick in VE activity that has necessitated a response from the community. Perhaps also owing to the personal interest of the Governor and the County Commissioner, government officials have been more active in meetings and discussions on these issues, and more willing to engage with communities than in some other counties. 32 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 125 Mandera is the only county in NIWETU’s area of focus that has line items for CVE activities in the county budget. These funds have already been used to train government Champions on CVE knowledge and skills, using the NIWETU’s CVE facilitation guide. As is seen in other counties, however, there is no county-specific (or national) legislation to support the implementation of the CAPs. 11.7 Section VII – Government Responsiveness to CVE The following section presents findings on the responsiveness of county and national level government to address VE. Summary findings and the result area index score are presented, followed by a more in-depth analysis of the findings to look at both levels of governments’ willingness to engage with communities and the challenges that exist in the relationship. 11.7.1 Summary of Findings Findings from the survey indicate broadly improved perceptions of both the county and national government. The county government, in particular, is believed to have significantly enhanced its understanding of CVE issues and demonstrated some level of commitment in supporting CVE activities most prominently, the CEF. Its support for the National Police Reservists (NPR) is viewed as instrumental in building the capacity of locals in complementing existing security measures, especially in places that are less served. While there was criticism over the national government policies on security in Mandera, and the feeling that they should also fund CVE programs, respondents in the survey acknowledged that a few representatives, such as chiefs and their assistants, were approachable and engaged on issues of CVE, perhaps more so than those in higher levels. It is worth noting that chiefs and their assistants represent the County Commissioner, and therefore this positive review still serves as a credit to the overall national government efforts. Views toward the security sector were overwhelmingly negative, as expressed by key informants and survey respondents, largely because of feelings of profiling and unjustified targeting. This finding suggests possible future areas of intervention in Mandera County. The CAP process is appreciated and seen as a work in progress in coordinating and managing relationships among governments, stakeholders, donors, security actors and community in Mandera. 11.7.2 Government Responsiveness Index Score The below questions are scored on a four-point scale 33, to understand overall government responsiveness to VE at both county and national levels. There has been a slight drop in the score on all questions gauging the perceptions on both county and national governments’ responsiveness to VE, apart from the perception of respondents on the county government understanding of the problem of VE in the community. The perception that the county government’s understanding has improved is perhaps due to the ongoing CAP engagement process and the increased prominence of the County Deradicalization and CVE Department. Nevertheless, views on questions around county and national governments’ involvement in CVE are still positive and show an aggregate increase since the baseline. Questions Baseline Score Midline Score Endline Score How well do you think the county government understands the problem of VE in this community? 2.82 3.33 3.49 How well do you think the national government understands the problem of VE in this community? 3.20 3.51 3.39 33 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 126 How willing do you think county government is to work on issues of VE? 2.84 3.55 3.10 How willing do you think the national government is to work on issues of VE? 3.10 3.46 3.00 How easy do you think it is to work with national government counterparts? 2.66 3.46 3.24 Overall average score 2.92 3.46 3.24 11.7.3 Government Understanding 94% percent of respondents noted that the county administration understands issues at least somewhat well, while only two respondents, a security sector representative and community leader, did not think the county government had a particularly good understanding of the issues. Perhaps predictably, the county administration themselves believed they have a very strong understanding of the issues. How well does the county government understand VE? Baseline Midline Endline Very well 30% 55% 57% Somewhat well 40% 25% 37% Not very well 18% 13% 4% Not at all 6% 5% 0% Don’t know 6% 2% 2% Key informants tended to be relatively positive toward the county government’s understanding of these issues. They acknowledged that the county administration has demonstrated strong leadership in the formation and implementation of CAP. A key informant noted that, although the community may not have a good perception of the county government development record, they believe that the county is truly committed to CVE. He states that: “The community has a good perception of government in terms of the county’s effort to contain violent extremism. The county has trained women, youth and imams, and provided motorbikes to a group of CVE monitors.” Further credit was given to the county government. Firstly, on its commitment to establish and fund various components of the deradicalization and countering violent extremism department as a formal channel for engaging stakeholders on issues of CVE. Secondly, their support of the NPR, who are well respected and knowledgeable about the community and able to provide additional support on security issues, even in more remote areas. It is also worth noting that some of the residents in the North￾Eastern region have felt pressured to join the NPR to prove that they are not AS or AS sympathizers – and for fear that if they do not join, they might be disappeared by KDF or other security actors. Lastly, the county government’s active engagement in the CEF which, according to a key informant, has been a game-changer in connecting actors within and outside of Mandera on CVE issues and priorities. “Include community representatives such as elders, imams, women and youth. The peace actors are represented as well, and relevant government departments such as the County De-Radicalization Unit and special groups like the disabled people.” – CSO Representative 127 How well does the national government understand VE? Baseline Midline Endline Very well 50% 59% 50% Somewhat well 28% 28% 38% Not very well 16% 7% 8% Not at all 4% 2% 0% Do not know 2% 0% 4% At the national government level, respondents held similar views. Only four respondents did not believe the national government had a particularly good understanding of the issues. These respondents were from the youth, private sector, and community leaders’ categories. It is perhaps predictable that some respondents would harbor negative views due to the perceived disconnect between the community, local CVE actors, and national government officials – mostly over the endorsement security policies that are seen to victimize certain sections of the community, especially youth and religious leaders. As it was observed during the midline, Mandera residents tended not to view the security sector as a proxy for the national government. Views on the security sector were expressed separately from views toward the national government. 11.7.4 Willingness to Engage in Programing Views amongst survey respondents were less positive toward the county administration’s willingness to engage on CVE related matters, compared to the midline. Despite this, over three-quarters of respondents at the baseline, midline and endline still believe that the county government is willing to engage on issues of CVE. Two respondents, a security and private sector representative, did not have an opinion on this. How willing are county government is to work on CVE? Baseline Midline Endline Very willing 34% 70% 46% Somewhat willing 36% 18% 30% Not very willing 16% 11% 16% Not at all willing 8% 2% 4% Do not know 6% 0% 4% Perspectives on the national government’s willingness to engage were just as positive, with 78% of respondents agreeing that the national authorities are at least somewhat willing to work on these issues. A fifth of respondents stated that the national government was not very willing, or not willing at all, to work on issues of CVE compared to views about the county government. How willing are national authorities to work on CVE? Baseline Midline Endline Very willing 44% 62% 32% Somewhat willing 32% 26% 46% Not very willing 16% 9% 12% Not at all willing 6% 4% 10% Do not know 2% 0% 0% Key informants noted the increased commitment to engage the community and apply softer approaches to security, as an indicator of broader support from all levels of government to try to support Mandera County in its fight against CVE. 128 Although perceptions on both the national and county government were viewed positively, the significant difference among the surveyed respondents on the willingness of the county and the national government to engage on CVE can be seen in the key informants’ analysis on Kenya’s stand on the conflict between Somalia National Army and the Jubaland militia. The national government is seen to support the Juba land forces which, in turn, makes community members doubt its commitment to enhancing security in the county. This is perceived as a contradiction to the national government commitment to working with community on issues of security. When this was raised at the CEF – a platform meant to provide a safe space for stakeholders to address key VE and security-related challenges – the County Commissioner’s office was seen to suppress diverging opinions seemed critical of the national government’s policy. Key informants stated: “Coming from a situation where the community was accused of supporting terrorism, to now sympathizing with an ‘enemy’ state is a double-edged sword for people of the North-Eastern region. This perception has greatly affected the gains already made.” – Key CBO Informant “While the county government made it a priority to protect the people, the national government was seen as not looking at the public interest. There is growing resentment because you find people are seeing the government wants them displaced from their own homes and it doesn’t care, after all, why is the government looking at the interest of protecting a mere militant?” – Key CBO Informant Besides, key informants were also critical of national government response to threats of AS in the North-Eastern counties and viewed this is as lack of willingness to address CVE issues collectively. A key informant stated that the reaction to AS attacks targeting non-local residents between January and March 2020, went against efforts made in strengthening community-government engagements on CVE. He stated: “The problem of teachers’ exodus started before the entry of Juba-land forces, the teachers were being withdrawn from the county against their wish, most of the teachers never applied for a transfer and this forceful eviction of teachers created despair. The trust and collaboration we built through NIWETU became broken, the forceful eviction of teachers was a major setback in Mandera.” – Key Community Leader Informant The national government’s commitment to the CAP was also questioned, as a key informant noted that there has been little investment made by the national government to support CEF implementation. “We have never seen an activity organized or paid for by the national government.” – A CBO Representative This was seen as being less sustainable, since the process was largely driven by the county government and not sufficiently a joint leadership with the County Commissioner as was envisioned. In contrast to other counties, the County Commissioner appears to be less committed than the Governor. “On part of the national government, there is no significant support. There was a time they did Nyumba Kumi pieces of training and that was the last activity we know. The county government has conducted pieces of training and supported a new batch of champions for change. Recently, the county Governor led a meeting of North-Eastern region governors in Mandera on CVE.” – Key Informant (C4C) Nonetheless, respondents believe that the national government is willing to work on these issues and note that some officials have been keen to re-evaluate their approach to the community. A key informant noted that there has been some improvement in how the government and security forces treat community members and their willingness to listen to criticism. This has helped bridge the trust gaps and improve information sharing. 129 “The government is now responsive, and it listens to people, we have seen security teams accepting their failures and shortcomings and acknowledging the difficulties of reporting mechanism. This was new to us, and from all segments, the call to work together and prevent the threats posed by VE was becoming apparent.’’ – Key CBO Informant “The national government no longer labels the whole community as terrorists. The security agencies have come closer to community and involve locals in information-sharing forums. Although security actors may not always respond to information that is shared by the community, the security agencies do not victimize the community when an attack happens and no longer resort to blanket punishment.” – Key C4C Informant 11.7.5 Engagement with Government Counterparts Engagement with both county and national government counterparts appeared to occur with some regularity. Of those who work closely with both levels of government on this issue, engagement and interaction are frequent, however significant proportions of respondents still engage only occasionally government counterparts. 54% of respondents stated that there are no challenges in working with county government counterparts, indicative of a strong working relationship between local officials and residents. Still, 21% of respondents believe that the county government does value their contribution and imposes CVE and other development programs on the stakeholders without consulting them in the design. Others state that it is difficult to access and schedule meetings with county officials. Frequency of engagement with national government counterparts Baseline Midline Endline On a daily basis 10% 32% 32% Once a week 4% 7% 6% A few times a week 10% 19% 4% Once a month 8% 0% 14% A few times a month 4% 10% 2% Occasionally 32% 9% 26% Rarely 32% 21% 14% Do not know 0% 0% 2% Responses on the challenges working with the national government were even more positive, especially when referring to their engagement with lower levels of the CC’s office, such as chiefs and sub-chiefs. Half of the respondents stated that they had no problems working with national government officials. Perhaps indicative of the broader fear of harassment and intimidation, respondents opted to moderate their responses, as statements when asked which national government officials are easy to work with, mostly chiefs and their assistants were mentioned. Only two respondents noted that the county commissioner and other officials were easy to work with. One of the two disclaimed that access to the County Commissioner is limited and dependent on the official individual effort to engage: “The former County Commissioner had a lot of interest in matters CVE and he would attend our meetings and activities in person. The current County Commissioner seems not to understand how weighty the issue CVE is, and he often sends his deputies and Assistant County Commissioners for crucial meetings as opposed to his predecessor who participated actively and had CVE at heart.” 130 The disconnect between more senior national government officials at the county-level was highlighted as an identity issue. A respondent from the county government category in describing the challenge in engaging the national government counterparts’ states: “They are defensive, don't act promptly on intel shared by the community, we don't sit in security committee, non-locals sit there and yet they don't understand the context.” 12. 131 13. ANNEX V – NAIROBI COUNTY REPORT: NAIROBI COUNTY ENDLINE OVERVIEW Data collection took place in Nairobi over a five-day period in February 2020, carried out by two researchers. In contrast to the baseline and midline, the research was concentrated only in Kibra sub￾county, because NIWETU has not had programming in Kangemi over the last two years. NIWETU facilitated introductory meetings to key stakeholders to explain the goals of the research and to facilitate further introductions to relevant actors. The research team conducted both qualitative and quantitative research with key stakeholders in Kibra as well as the wider community. The breakdown of respondents in the stakeholder questionnaire for each hotspot is presented below. The overall sample size was 43. Category of respondent from Kibra Number of respondents Community/Religious leader 8 Community-based organization 6 Educational institution 1 Media 3 Private sector 7 Security sector 3 Women’s group 10 Youth 5 Total 43 Seven key informant interviews were conducted with key stakeholders including government officials and civil society representatives. 132 14. KEY FINDINGS 14.1 Section 1 – The VE Context The following section presents findings about the broader VE and security context in Kibra sub-county in Nairobi County. This section then discusses the VE context with synthesized findings from key informants and the quantitative survey respondents. Findings will discuss the wider peace and security context, perspectives on youth disappearances, views on extremism and recruitment, organized crime, and the safety and security context in Nairobi County and Kenya more broadly. 14.1.1 Summary of Findings In Kibra, the security context remains precarious. Survey respondents were broadly split on the degree to which safety and security has improved, while the VE threat reportedly decreased. Violent crime continues, and youth – even young boys – remain at-risk of recruitment into a VEO or a criminal gang. Respondents agree that these groups prey on the same vulnerabilities and grievances in recruitment. Therefore, even though most do not believe VE to be a significant threat to Kibra residents at the moment, they caution that the potential for VEOs to make inroads into these areas is strong, given the high levels of poverty and unemployment. Still, several respondents do acknowledge the improved interaction and cooperation between community members, leaders and authorities, and see this as an opportunity to address some of the grievances related to security management in the community, such as extrajudicial killings. 14.1.2 Causes of Insecurity Two thirds of survey respondents believe that theft and high levels of unemployment are among the most significant risks to security in the area, followed by the threat of criminal gangs, which are factors that can be manipulated by both VEOs and gangs by promising jobs or other income opportunities. No respondent mentioned violent extremist groups as a threat, but two respondents, a security actor and youth representative, noted that radicalization was a security threat in Kibra. The youth representative further clarified that radicalization is not overt, but it is believed to be happening, signifying that communities’ perceptions of VE as a threat have reduced even further compared to the midline. While theft has increased, drug abuse, which was identified as the biggest threat during the midline, has reduced significantly, and so too has the threat of national and local politics. Security threats Percentage of respondents Baseline Percentage of respondents Midline Percentage of respondents Endline Theft 42% 38% 67% High levels of unemployment 50% 45% 63% Criminal gangs 69% 62% 51% Drug abuse - 66% 35% High levels of poverty - 30% 35% Local politics 50% 11% 28% National politics 42% 11% 19% Lack of education -- 4% 9% Land disputes 12% 13% 7% Radicalization 0% 13% 2% Inter-religious tension 8% 4% 2% Inter-ethnic tension 42% 17% 2% Youth disappearances 0% 4% 2% Other 15% 6% 0% Violent extremist groups - 2% 0% 133 14.1.3 Disappearances A majority of endline respondents state that youth have rarely or never gone missing within the past twelve months, compared to 32% of respondents from midline. Those who said they were aware of disappearances learned about them in meetings and workshops related to peace. 21% acknowledged that they did not know if disappearances occurred; unlike in North-Eastern regions, our engagements with respondents in Nairobi show that they are less afraid of repercussions from responding to such a question and feel freer to be more open with their opinions on these issues, so it is likely that these respondents truly do not know. Though 56% of endline respondents do not believe that young people frequently go missing, 43% still believe that disappearances happen, pointing to a division in how these disappearances are understood across respondents. How frequently in the last 12 months have young people (including youth and children) gone missing from this community? Kibra Baseline34 Kibra Midline Kibra Endline Very regularly 8% 6% 2% Somewhat regularly 35% 15% 21% Rarely 27% 15% 35% Never 27% 17% 21% Do not know 4% 4% 21% Survey respondents were uncertain about where youth who disappear have gone. While some stated that those who disappear are believed to have travelled abroad for work to countries such as to Saudi Arabia (mostly women), and Qatar (mostly men), a few believed that there were those who went to Somalia after they were recruited by Al-Shabaab. 44% said they believe that those who disappear could have been engaged with unlawful groups such as Al-Shabaab, ISIS or others. Others note that some are arrested or killed by the state for political sabotage, such as disrupting rallies or having been suspected of committing crimes within and outside of Kibra. A religious/community leader stated that: “One of them was found in Tanzania and was to be taken to Somalia.” Another community leader noted that the fear of being killed by the police has forced those warned or suspected of committing capital crimes to seek refuge in rural areas. As stated above, victims of extrajudicial killing have received warnings from security officers in advance, giving them time to seek refuge in their rural home. He stated: “Mostly these are people linked to criminal gang groups. They move to upcountry or to other countries.” 14.1.4 VE and Recruitment Survey respondents were in broad agreement that extremism and recruitment to join extremist groups is less of a problem now than it was 12 months ago; while only 5% of respondents strongly agreed, 63% at least somewhat agreed which is 20% higher than midline. Extremism and recruitment are less of a problem now than it was 12 months ago. Baseline Midline Endline Strongly agree 1% 13% 5% 34 Baseline and midline data presented in this report represents data from both Kibra and Kangemi combined, unless otherwise noted, while endline data only presents findings from Kibra. As such, drawing conclusions from comparisons across evaluations should be done cautiously. 134 Somewhat agree 24% 40% 63% Somewhat disagree 30% 30% 16% Strongly disagree 13% 13% 5% Do not know 22% 4% 12% According to a youth representative: “Cases of youth going to Eastleigh and Somalia have reduced.” Most of the survey respondents who agree that extremism and recruitment is less common now than before, believe that increased awareness and CVE programming has led to enhanced vigilance in the community. The majority of respondents specifically acknowledged that sensitization and training workshops have increased awareness about VE and that parents are more knowledgeable guardians. They stated: “People are more aware, so such activities which used to happen because communities were not aware have reduced.” – Community-Based Organization “People's thinking has changed, and even religious teachings have helped as well as the CVE activities which were not there before.” – Youth Representative “People have been educated and parents are involved in watching their children’s movements and so it has reduced. The dangers of joining the groups are also known.” – Private Sector Representative Other survey respondents simply stated that they have not heard of any cases or incidents of VE in Kibra over the reporting period. “We don’t have the clear statistics, however it has been a while since I heard of any incident or case.” – Community-Based Organization Respondent One private sector representative tended to agree with this opinion, stating that the community has more pressing security concerns and VE is not a major threat at the moment. He quotes: “Cases of youths getting involved in these issues are not as prevalent as they were before. Now, we are dealing with issues of drug abuse.” A women’s group member who acknowledged that she did not know if any recruitment is taking place clarified that the lack of clear information or data on recruitment does not mean that the youth are less vulnerable, because the rate of frustration and unemployment had increased. She quotes: “These are rumors and no facts. So, we can’t say that people are joining AS. Young people are yearning for money, they are vulnerable.” Despite the belief that VE is a negligible problem in Kibra, 16 of those who were interviewed note to have heard of at least one reported case of recruitment within the last 12 months. Therefore, it is important to sustain CVE processes that are credited with improving the situation, because VE groups have in the past manipulated frustrations of young people, especially those that are unemployed and/or abuse drugs in the society. 14.1.5 Gang and Criminal Activity Survey respondents broadly agreed that organized crime and gang activity are less of a problem now than 12 months ago. Just over 70% of respondents stated as much, while 28% of respondents disagreed. 135 The number of respondents who agree point to increased youth engagement programs and enhanced surveillance by security actors in Kibra. Some further stated that, after the by-election held last year, there is less gang violence; but other forms of violent crimes have increased, such as armed robbery and muggings. A community leader noted that they had informed the Assistant County Commissioner on this last year and the government has enhanced its security operation in Kibra as a result. During the midline, a security officer explained this connection, noting that “the politicians manage and run their own gangs during electioneering period, but when there is no campaigning, the gangs dissipate.” Mugging and robbery is more common after 7pm, but there are some zones where those new to Kibra have been robbed during the day. A representative from the women’s group stated that she was recently a victim of a mugging by a very young boy who stole her phone. A community leader claimed that those engaged in crime include school-going young boys, stating that once schools reopen, crime is visibly lower. “As I am talking to you, I am on another assignment and have a tablet in my bag... I don’t feel safe removing the tablet to collect data... mugging is not a problem to these people day or night.” – Women’s Group Representative “’I haven’t heard of any incidents. There was a time you couldn’t walk without being mugged.” – Community Leader The rate of crime in Kibra remains high and, to an extent, has affected business operations because they are compelled to close early enough to reduce the risks of being robbed by youth in the community. A private sector representative said that: “I have to shut down before 7pm. Closing late means you are making lots of money, so the idle hungry youths may come for you or as you walk home late, they can easily attack you.” A youth representative and community leader noted that some new income-generating opportunities are emerging for youth, such as carwash centers that are common with rehabilitated youths. Beyond economic opportunities, several community-based organizations had integrated mentorship and life skills in their youth engagement programs, which is positively influencing the decision of youth about their choices. Another reason given by the 20% of respondents agreeing that organized crime and gang activity are less of a problem than 12 months ago, was the improved information sharing between government and community, as well as “tough” measures placed by security offers. A shoot to kill order by the police has created fear among youth and gang members. Improved community policing and security responses have helped speed up response times when crimes are reported. The mixed perceptions over the increase or reduction in crime in Kibra could be because the levels of security, by sub-areas, are different and therefore respondents may be reflecting based on their familiarity in specific neighborhoods. A community leader stated that: “Increased surveillance i.e. CCTV Cameras (business premises) have been erected and people are aware and afraid of being caught.” This might not apply to neighborhoods that do not host sizable businesses. A community leader noted that, although the threats of gang activity and other crimes are high in Kibra, terrorism and VE is considered a risky crime. Most youth are more fearful of joining groups such as Al-Shabaab than joining gangs, perhaps because gang membership is more of a norm in some of these communities. 136 14.1.6 Safety and Security Most survey respondents agreed that safety and security in Nairobi has improved in the past 12 months; in total, 72% agreed. Respondents credit this to the actions of security officers, including arrests and sometimes killings, improved reporting on crime, and positive opportunities for engaging young people. Views of survey respondents were more mixed on the degree to which the number of terrorist attacks in Kenya has decreased, perhaps predictably so, as at the time of data collection a series of terror attacks had been reported in North-Eastern and Coastal counties (January through February 2020). This could have influenced views about the threat of terrorism. As such, 69% of respondents believe that the number of attacks in Kenya has increased. 74% were more optimistic when asked about speaking about the threat of terrorism in Nairobi. “We still hear every day of attacks and even in our county Nairobi we are not safe from Al-Shabaab.” – A Representative from the Education Institutions Over the past 12 months, the number of terrorist attacks in Kenya has decreased. Midline Endine Strongly agree 11% 0% Somewhat agree 26% 26% Somewhat disagree 51% 60% Strongly disagree 13% 12% Do not know 0% 2% 14.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced The table below compares scores for the sentinel indicators from the midline and endline evaluations on indicators that track communities’ willingness to mobilize in support of peace and security objectives at the county level. While NIWETU does not work to support these indicators, we assume that communities’ willingness to engage with these structures, or their trust in these structures’ ability and relevance, will affect their perspectives on the wider community’s willingness to mobilize around CVE. Individual scores reflect respondents’ personal views on the corresponding statement, while the ‘meta’ scores, or meta perceptions scores, reflect how respondents believe their peers would answer. Research has shown that respondents are more likely to act in accordance with what they think are their peers’ views, or the social norm. This means that when meta scores are higher than the individual scores, views in general tend to be more positive about a particular statement, and vice versa. These findings come from a sample of 40 individuals in two neighborhoods in Kibra and should therefore not be considered statistically significant or representative of a wider sample of Kibra residents. Figures presented below are out of a ten-point scale35. Midline scores have been revised to reflect only scores from Kibra residents, to allow for a clearer comparison between midline and endline; however, the midline sample comprises only 20 individuals. Degree to which peace committees are active and responsive. Midline Endline I feel that the peace committees are active and easy to access in my area. Indiv. 3.54 4.93 Meta 2.70 4.58 I feel that the peace committees have an adequate understanding of the issues in my community. Indiv. 2.37 5.18 Meta 4.03 5.27 Indiv. 2.71 4.90 35 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 137 I feel that peace committees effectively respond to issues in my community Meta 3.14 4.83 Overall score 3.08 4.95 Degree to which communities actively engage in barazas. Midline Endline I frequently attend community barazas. Indiv. 2.68 3.72 Meta 1.83 3.80 When I go to barazas, I feel like my opinion is considered and my concerns are responded to. Indiv. 2.78 3.87 Meta 2.69 4.23 I always feel positive after participating in barazas. Indiv. 2.84 5.23 Meta 3.39 5.46 Overall score 2.70 4.39 Degree to which Nyumba Kumi initiatives are active. Midline Endline Nyumba Kumi initiatives are active in my community. Indiv. 2.47 3.03 Meta 2.18 3.07 I feel that Nyumba Kumi initiatives are positively addressing the issues that concern me. Indiv. 1.70 3.13 Meta 2.50 3.51 I have confidence in Nyumba Kumi initiatives and trust that it will help improve peace in my area. Indiv. 3.09 4.01 Meta 5.92 4.56 Overall score 2.98 3.55 Presence of other organizations/actors that work on peace/security. Midline Endline There are numerous other networks and organizations making a positive impact on peace and security in my community. Indiv. 5.56 5.68 Meta 3.36 6.49 Overall score 4.46 6.09 Across all indicators, perceptions show an increase from the midline evaluation. Perhaps most notably, while all average scores from the midline were below five, suggesting mostly negative perceptions toward these community security structures, respondents now tend to believe that there are numerous organizations and other actors working on issues of peace and security in their community. While this research is not able to ascribe attribution, it is likely that NIWETU’s investments in Kibra, through grant activities as well as the Champions for Change program, have become more visible to local residents. Perspectives on the effectiveness and relevance of peace committees have also shown a marked improvement. A significant change occurred in perspectives on the degree to which residents believe that peace committees have an adequate understanding of issues in their community; while at the midline, individuals expressed rather negative views on this question, at the endline perspectives for both the individual and their peer group were positive, suggesting that peace committees may be becoming increasingly relevant. Across the board, these findings suggest that willingness within the community to mobilize around peace and security issues, and potentially VE, may still be limited. While scores have improved since the midline, they still represent somewhat negative views across the board. Taken with findings presented in the previous section, this finding is not particularly surprising; extrajudicial killings are met with frustration, and more immediate concerns around earning a daily wage likely take precedence. 138 14.3 Section III – Skills and Knowledge The following section looks at the skills and knowledge that exist within the stakeholder community on VE and CVE related activities. The primary focus is on analyzing the different types of training that exist on these topics, as well as the degree to which respondents have participated in such activities already, and what, if anything, they are doing with the skills and knowledge thereafter. Following the summary of findings and presentation of the index score for this result area, this section discusses participation in trainings, the degree to which they have been helpful and why, and how to make them more effective. 14.3.1 Summary of Findings Training on CVE is more widespread than it was during the midline. A majority participated in NIWETU-sponsored trainings, and they note that few, if any, other organizations are offering training on this topic in these areas. There are a number of important ways to strengthen the trainings and reach a wider network of individuals, especially youth. That said, respondents identified a number of concrete ways to improve these activities, most notably, to think of strategies that reach target groups in more innovative and effective ways outside of a meeting room and strengthen local training capacity to ensure the consistency of such activities. 14.3.2 Skills and Knowledge Index Scores The following table shows the average scores on a four-point scale36 from a number of key questions related to skills and knowledge. Baseline, midline and endline scores are presented side-by-side to highlight changes over the past two-and-a-half years and demonstrate a slightly stronger overall improvement in skills and knowledge from the time of the baseline. Questions Baseline Score Midline Score Endline Score Have you participated in any training on CVE and understanding the causes of VE in the last 12 months? 2.76 3.11 3.44 How much, if at all, do you think these trainings have helped you understand VE and how to prevent it? 3.56 3.82 3.70 How much, if at all, do you use the information you have learned from these trainings? 3.63 3.88 3.69 Have you shared the information you have learned with others in the community? 4.00 4.00 4.00 Overall average score 3.49 3.70 3.71 These findings point to a slight improvement in participation in CVE skills and knowledge activities, with a slight decrease in the scores measuring the usefulness of the training. The score on dissemination being constantly higher than the rest from the time of the baseline, is an indication that the culture of information sharing among trainees remains consistent. 14.3.3 Participation in Training 86% of all survey respondents have participated in CVE skills and knowledge related activities, which includes all community-based organizations and a majority of youth, community/religious leaders, education institution and women’s group representatives. Participation was less common amongst private sector actors. 36 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 139 Participated in CVE Trainings by Category No Yes CBO 0 6 Community/Religious leader 2 6 County administration 0 0 Educational professional 0 1 Media 0 3 Private sector 2 5 Security sector 0 3 Women’s representative 2 8 Youth representative 0 5 Total 6 (14%) 37 (86%) Only two of the 37 respondents mentioned not to have attended a NIWETU-sponsored training and could not recall the organizations that hosted or sponsored the training. It is also worth noting that respondents confused training with meetings and other community forums on CVE. Twenty participated in the Muslim Centre for Peace and Reconciliation, nine mentioned Baraza la Walimu Kibra (BAWAKI), while six noted to have attended interfaith training sessions but could not recall the host organization. Other NIWETU partners mentioned included RACIDA as part of the C4C program. One youth respondent acknowledged that there have been other trainings hosted by organizations outside of NIWETU, such as Kenya Muslim Youth Association (KMYA), YADEN East Africa and BRAVE. Others could only recall the venue where the training were held (such as mosques and specific hotels) and the topics covered, not necessarily the sponsor or host. “We went to Makina mosque. We were all women in that meeting.” – Private Sector Representative (who had not attended a training on CVE) Survey respondents identified a range of trainings in which they participated, with more than half engaging in training on understanding VE, how to prevent or counter VE, and why people choose to join VEOs. Less common was participation in training on monitoring and evaluation, how to work with other organizations, to prevent or counter violent extremism and leading interfaith discussions. Types of Training Baseline Midline Endline Understanding violent extremism 26% 68% 86% How to prevent or counter violent extremism 28% 64% 70% Why people join violent extremist groups 26% 58% 61% How to identify recruiters and recruitment methods 17% 49% 49% How to talk to others about violent extremism 28% 46% 40% Training others on preventing or countering violent extremism 9% 42% 23% How to report extremist activities 20% 42% 33% Leadership skills 20% 36% 21% Advocacy skills 13% 36% 23% Conflict resolution 37% 36% 7% How to work with other organizations to prevent or counter violent extremism 11% 33% 5% How to lead interfaith discussions on violent extremism 7% 33% 5% Identifying signs of recruitment 15% How to work with the government to prevent or counter violent extremism 13% 30% 14% Monitoring and evaluating 4% 27% 2% Other 15% 27% 19% Survey respondents shared mixed views on which trainings were most helpful, but there was no broad agreement. 48% mentioned that training on how to prevent or counter violent extremism was the 140 most helpful, and 35% agreed that understanding VE was most relevant for their needs. Other trainings that were highlighted as significant include: identifying recruiters and recruitment methods and understanding why people join violent extremist groups. Respondents noted that they valued the trainings because the knowledge and skills acquired not only helped them improve in their regular work but was also an opportunity to learn new things and networks with other community stakeholders. A representative from the women’s group categories appreciated the training sessions because she could now identify early warning signs of recruitment and radicalization that she could not before. “With the mentorship program I can see the outcome. There are youths who changed their beliefs on violence and are now focusing on their talents. This has been the most helpful so far.” – Community Leader/Religious “We receive young people at my center who are at risk of being recruited and we engage them actively. There are youth who have confessed to have ideas of joining this groups and we helped them avoid it.” – Community Leader/Religious A respondent from the security sector noted that increased training on CVE has helped mobilize more support from community leaders who are often also part of other community security structures. She stated that: “As a community leader, working with others helps in expanding the number of people with knowledge and skills on CVE and means that we will also get more people working on improving security.” All respondents who have attended training sessions noted that they have used the knowledge acquired at least a little, with 78% noting to have applied the skills a lot in their work and interactions in the community. While a number of survey respondents state that they have used the skills and knowledge gained to raise awareness about VE issues and to disseminate messages, others note specific initiatives that they have carried forward on their own, more so than what was found in other counties. A few notable examples emerged from the survey:  Strengthening and spreading counternarratives – A religious leader noted that he gets two sermons every Friday and uses this platform to demystify misconceptions, while a youth representative noted that the knowledge has helped him confidently respond to questions from peers in both formal and informal spaces. “I am a leader in many women’s groups and I talk in those forums and share this information on how we need to be aware about CVE and what we can do as a community.”  Improving parental skills – A women’s group representative stated that the training helps her pay more attention when interacting with her children and their friends. She states that: “I was able to talk to a young boy who was a notorious thief and now he is a mechanic in one of the garages.”  Intervening to support at-risk youth – A community leader mentioned that he had managed to rescue a young boy from his community who was engaging with violent extremist groups and had been promised work in Somalia. He quotes: “I intervened, talked to the mom, and he was moved from Kibra and taken to the village. This happened in 2019.” 141  Using culture and sports to promote CVE narratives – Two respondents noted that they have applied the knowledge to inform their networks through sports and music. A young respondent noted that he has composed a song on violent extremism issues, while a football coach who manages different schools integrates countering violent extremism as part of life skills training before football practice sessions.  Promoting more coordination – A community leader stated that he works more closely with other leaders and government, more specifically, the County Commissioner’s office to link youth to livelihood programs and vocational training opportunities, in an effort to reduce their vulnerabilities to violent extremism and other crimes. Significantly, survey respondents were able to articulate a much clearer and deeper understanding of VE and CVE. At baseline, researchers struggled to ensure that respondents were answering specifically about violent extremism trends, after noticing that most understood VE to refer to gang activity and associated violence. During this evaluation, respondents more specifically related VE to “terrorism”, “Al-Shabaab”, “radicalization” and “indoctrination” of extreme beliefs, focusing mostly on aspects of prevention and de-radicalization in defining CVE. Some even distinguished between radicalization and violent extremism, and further elaborated the vulnerabilities in the communities that would lead young people to join such groups. Some definitions that capture this include: “The use of violence to achieve one’s demands. Mostly they have grievances. Without attention they become extreme and use violence.” “Groups like Al-Shabaab and terrorist gangs.” “‘Itikadi kali’ it means you are willing and ready to get what you want using violent means.” Respondents broadly understood CVE to refer to preventative efforts to combat violence, behavioral change activities, working with youth to offer various forms of support, and raising awareness. Only two respondents viewed CVE through more of a security lens, as compared to the baseline and midline, where a sizable number believed the term refers to the police and security sector’s efforts to stop violence. A youth representative defined countering violent extremism as: “How you handle the situation before people get into harmful acts like terrorism.” 14.3.4 Improving Training Programs Respondents were critical and suggested improvements in the targeting, consistency and content delivery of the CVE trainings in Kibra. On targeting, respondents noted that hosting training sessions in hotels limited the number of participants that could attend such meetings. A community leader stated that trainings tend to mobilize the same people, mostly community and religious leaders. “I think they need to talk to the youths more… so far they target leaders and adults… but those being talked about are missing… Involve the youths. I was in the trainings but I hardly saw young people… yet the issues being addressed targets them. It is also difficult for us businesspeople to close down the whole day to attend a training. Let them have programs that can be conducted near our business locations.” – Private Sector Representative Respondents recommended that the trainings be more inclusive. A youth representative noted that this can be achieved if some training sessions are conducted within the community and as low as the 142 village level, to target groups such as businesswomen and youths who are rarely represented. In addition, this might require that trainers understand local languages such as good Swahili and Sheng to deliver the content in a way that is easily understood and participatory. It is also important to note that Kibra, although an urban slum, is highly heterogeneous and most villages are segmented based on ethnicity. This may affect how different areas within Kibra interpret, connect, and engage in different conversations. A youth leader states: “Use simple local language understood by the community, for youths use sheng… but more so let those being trained know these words [code words] used by youths.” Several respondents recommended that the training be more regular and structured, for example having a calendar for reference, but in a way that ensures more people participate. A community leader noted that one of the biggest challenges in ensuring more people participate is the lack of funding, for most participants expect an honorarium in every session, making mobilization quite costly. A community leader lamented that the training content was very packed and had a lot of information for a week’s training. “CVE is broad, but we did only three days, I believe this was just a formality. Those selected for mentorship program need more information and support to reach out to others. The activity should also be implemented for several years.” – Community Leader “I found them rushed and too detailed. They can be done over a period of time.” – Private Sector Representative A religious/community leader felt that the training and, generally, most CVE activities disproportionately focus on the Muslim community in Kibra – even though violent extremism is considered a threat from all religions and cultures. A community leader notes that: “I attended three meetings, not everyone gets this opportunity, so we need to create more awareness. These meetings have been focusing on Muslim communities and CBOs led by Muslims.” Beyond targeting participation across religions, there are recommendations that, when delivering trainings, there should be clarity in defining violent extremism to ensure that community member do not perceive it as a Muslim problem. This includes addressing other forms of extremism or injustices in the society. A youth representative stated: “Let them describe the facts clearer when defining what violent extremism means and separate it from mainstream Islam… let them say it’s an extremist who is manipulating religious scripts…. include other forms of extremism like those that are as a result of actions by the state.” A woman’s group representative noted that consistency in the trainings also helps strengthen social networks, because it allows people to network and connect with each other even after the training has ended. Social networks in other counties have been identified as key platforms for disseminating knowledge and organizing to counter violent extremism at the community level. This can be a valuable approach in enhancing sustainability of CVE processes in the communities. Two respondents suggested that CVE programming could include a livelihood component, especially when targeting women and youth, given that this is the biggest challenge both groups face. 143 Some respondents also suggested the need to implement “refresher courses” on trainings, to ensure sustainability of the outcomes. Still, most respondents agreed that the effectiveness of these initiatives will be limited without further resources and funding. “I can’t even remember clearly the last time I was in training... they need to be more frequent, to update us. I am a Luhya, in my community they need to be reminded... they tend to forget very fast and take time to understand.” Survey respondents then identified a number of specific training topics that would be most useful. More than a third suggested training on how to prevent or counter violent extremism and understanding violent extremism. As compared to the midline, none of the options provided garnered above 50% as the most useful topics. Respondents in previous questions did highlight the value of building the capacity of mentorship in the community and keep strengthening mentorship skills across the community. Other topics mentioned include engaging radicalized youth, youth entrepreneurship, and counselling or rehabilitating for drug abusers. What types of trainings would be most useful? Baseline Midline Endline How to identify recruiters and recruitment methods 26% 53% 28% Understanding violent extremism 61% 45% 33% Training others on preventing or countering violent extremism 35% 45% 12% How to work with the government to prevent or counter violent extremism 35% 45% 2% Advocacy skills 24% 45% 9% How to lead interfaith discussions on violent extremism 35% 43% 5% Why people join violent extremist groups 46% 43% 21% Leadership skills 33% 43% 5% Conflict resolution 57% 43% 2% How to work with other organizations to prevent or counter violent extremism 46% 40% 2% How to talk to others about violent extremism 54% 40% 19% How to report extremist activities 41% 36% 12% Monitoring and evaluating 17% 36% 0% How to prevent or counter violent extremism 63% 34% 44% Other 28% 26% 12% 14.3.5 Sharing Lessons Learned Survey respondents highlighted the importance of sharing what they learned with others and, as such, all respondents noted sharing what they have learned with others, particularly with other community members and colleagues in their organizations. This support earlier mentions that most trainings involve community leaders and community-based organizations and would point to steady post￾training knowledge transfer. However, there was a decrease in sharing across all categories compared to the midline. This could be because the reduced number of CVE activities as detailed in other sections may have reduced opportunities for interacting with new contacts or those in categories such as local government officials, religious leaders and individuals from other organizations. With whom have you shared your new skills and knowledge? Baseline Midline Endline Community members 35% 70% 84% Friends/school mates 24% 70% 33% Colleagues in my organization 24% 58% 49% Community leaders 35% 58% 23% Family members 20% 42% 26% Colleagues at other organizations 24% 36% 12% Other 11% 36% 2% 144 People from my mosque/church 26% 33% 19% Colleagues in the local government 15% 21% 5% Business partners 24% 15% 5% I have not shared the information or skills with anyone 0% 0% 0% 14.4 Section IV – Community Networks The following section assesses the existence and strength of community-based networks working to counter extremism in Nairobi. A brief summary of survey responses followed by an analysis of the findings is presented below. 14.4.1 Summary of Findings While networks between CVE actors exist to some extent in Kibra, their reach is limited. Hampered by limited funding and competition between groups for what exists, collaboration has not improved over the past 12 months, according to survey respondents. Despite this, there is a growing appreciation of the strength of local stakeholder networks in reaching out to more groups in the community. As such, the same individuals are targeted and involved in these activities and the level of duplication is high. Supporting efforts to further strengthen engagement between different actors working in the peace and security space in Kibra will ensure more efficient and potentially impactful programming. 14.4.2 Community Networks Index Score The table below highlights the average scores of five questions (on a five-point scale37) related to establishing community networks to counter violent extremism. Scores in this section represent a slight decline in the level of collaboration from the time of the baseline across all categories, although findings still indicate there has been an increase in the value from networks created over time. Notably, conflicts or disagreements between actors working on CVE appears to have slightly increased since the baseline, with a decreasing score in that category across the three evaluation periods. Questions Baseline Score Midline Score Endline Score Do you work with any other organizations on CVE? 4.55 4.15 4.07 Do you collaborate more, less, or the same amount with these actors than you did 12 months ago? 4.04 3.59 3.40 How often are there conflicts or disagreements between different actors working on CVE? 3.84 3.63 3.39 Overall, do you think improved networks between organizations working on CVE has been helpful? 4.35 4.26 4.33 Overall average score 4.20 3.91 3.78 14.4.3 Collaboration Between Actors 81% respondents stated that they are working with other organizations on CVE-related activities in Kibra. Over 65% of these respondents work with youth and this goes to show how much the community and the leadership recognize the vulnerability of young people to violent extremism and other crimes. It is also common amongst respondents to work with women’s groups, community leaders and religious leaders. Although few respondents mentioned having worked with self-help groups, it is important to note that most women and youth groups can also fall in the self-help group category locally referred to as chamas, which are simply investment groups. Although most of these groups are not formally registered, they operate in an almost similar way to those registered (as 37 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 145 associations with the Ministry of Gender and Social Services), but with little or no legal and tax obligations. In most cases, community mobilization has focused on these groups in the community, because they are easy to reach and offer a trusted link to other community members. Least common was collaboration with the private sector and security sector. Which organizations do you work with? Youth organizations 62% Religious leaders 54% Women’s groups 44% Community leaders 39% Other community members 26% County or national government actors 23% Self-help groups 18% Security sector actors 13% Business owners 13% Teachers 8% At-risk youth 5% Other 5% A majority of respondents work closely with more than one category to diversify their reach, because different individuals and groups have distinctive constituencies but also offer different perspectives based on their experiences. A community leader acknowledges this stating: “We have different capacities; we may know something that others don’t, sheikh may have a skill we don’t ... just like that...’’ Respondents note that the size of Kibra, and diversity in both ethnic and religious backgrounds, means that anyone who seeks to reach a wide range of community members has to work closely with other leaders who represent the specific groups. For example, in working with the Nubi community, CVE actors are keen to engage Nubi leaders to increase the legitimacy of their engagements among that constituency. A key informant noted that the choice for engaging community leaders more prominently was to get buy-in and enhance sustainability of the activities, since such leaders are more committed to CVE issues because they had made efforts in supporting the community without external interventions, and therefore the probability of continuity is higher with them. She quotes: “CVE is not for everyone, CVE was a new concept, we targeted people already working in the community; people who are passionate about CVE and people who engage the community; we were looking at a broader picture; NIWETU can end, but working with people from the community promises sustainability.” Women’s groups are also considered to be more reliable in that they not only take the issues that affect their children seriously, but also have a strong social network that enables them to share information learned and mobilize more participants in the community. A community leader also notes that women have privileged access to their children and are more likely to know when a child’s behavior has changed. “Because I work in business and I also belong to women’s groups, I also engage with youth because my son has a lot of friends who come to my business.”– Private Sector Representative On women’s groups, a religious leader noted that: “Women are the most accessible and they are also willing to make contributions to help the community.” 146 A respondent further affirms her confidence in community leaders, who she notes are a source of information on what is happening in the community and a link to understand how best to support youth at risk in the community. Community leaders are seen to be more knowledgeable, networked and often called upon to intervene when a mother is seeking help for their trouble child. Through such engagement, community leaders also benefit because they have access to knowledge about the community – a skill expected from them by other stakeholders. “I am known as a community leader. I get invited and I also try to participate actively in community activities like clean up exercises. It is important for me to get updates on opportunities for young people and women.” Respondents also acknowledge that CVE messaging is largely based on religious knowledge and this placed religious leaders and youth in a better place to interpret and share counter narratives, especially those that manipulate religious teachings. People also trust religious authorities on issues related to CVE, especially in Kibra where the CVE engagements are seen as an Islam issue, since most CVE programs tend to target and happen in Muslim spaces such as mosques and Islamic community centers. “We have been engaged in creating awareness for quite some time and religious leaders have the platform to pass information to a large following.” – Women’s Group Member Others chose to work with specific groups and stakeholders in their respective workstations. A representative from the education institution noted that the choice to work closely with youth and teachers was default for that is a requirement. “I am the government administrator at the community level. I do get invites and I honor them. It keeps me updated and allows me to interact with other actors on the ground.” – Security Representative Over half of the respondents who mentioned that they work with other organizations collaborate more now than they did a year ago. While 45%, a higher percentage compared to the midline, stated that they collaborate with other actors less than they did 12 months ago. Respondents who mentioned having collaborated less in the last 12 months note that this was due to reduced opportunities for interaction among CVE organizations due to a decline in the number of CVE activities in Kibra over the last year. A community leader notes that the frequency of meetings and trainings has reduced drastically, reducing the opportunities for networking among most CVE actors. A security actor further explains that activities in Kibra change depending on the issues at hand, or donor priorities. During elections, for example, there is more focus on ensuring peaceful elections and voter education. Organizations leverage the same community leaders for these activities too. He states that collaboration among actors depends on peace and security issues: “Depending on the security challenges the community is dealing with at a given time, that determines who and what programs are active. Most of these organizations depend on donor funding.” A CBO representative also notes that local groups working on CVE were not necessarily based in Kibra, and therefore the sustainability of such activities largely depended on their funding and not the ability of the community to learn and carry activities forward. Community leaders who were actively involved took up the responsibility individually. A religious leader states that it was easier to take up the responsibility after the CBOs because he has an existing platform, i.e. the mosque, where he continues to share knowledge acquired with his congregation. One community leader lamented that there has been less support given to youth who are willing to reform – a fact that discourages some leaders from collaborating further. In an earlier section, a respondent proposed that training in ways which community leaders can handle and support 147 rehabilitated youths or deradicalization strategies, should be backed by law where those involved are protected. Perceived competition among organizations and community leaders has hindered stronger collaboration on CVE. An ACBO representative notes that other organizations seem skeptical about partnering, stating that: “Finding organizations is not easy. Most don’t want partnerships. I think people don’t want others to get know their financial status as an organization.” “Most of our people are motivated by personal interests. It’s hard to come together.” – Community Leader “Organizations are all looking for funding and sometimes they want to outshine each other mainly because of competition.” – Community-Based Organization Representative Respondents who were more positive about the level of collaboration between actors over the past 12 months noted that connecting with others improved their networks, and that the ability to learn about what other organizations are doing was one of the greatest values of the collaborations. Two community leaders noted that their engagements in CVE activities and connections may have strengthened their authority in the community. Their choice to collaborate is driven by demand, as other community stakeholders often invite them to take part in various activities. A football coach notes, due to his engagements in CVE and peace activities, more schools now see the value of integrating some of the life skills sessions in sport, and often invite him to provide mentorships for students. Overall, however, all respondents believe that improved networks have been helpful. 14.4.4 Conflict Between Organizations There has been a slight decrease in the frequency of conflicts and disagreement between organizations working on CVE issues in Kibra compared to both the baseline and midline; 28% stated that conflict or disagreement between CVE actors never occurs. Another 28% stated that it rarely occurs, but 23% of respondents stated that they did not know how often this happens. Only 19% believe that it happens somewhat often. This indicates a slight decrease in disagreements that occur between organizations over CVE from the midline. How often do disagreements occur between organizations working on CVE? Midline Endline Very often 0% 2% Somewhat often 24% 19% Rarely 28% 28% Never 22% 28% Do not know 26% 23% Respondents who refute that conflicts and disagreement happen frequently note that, in an informal settlement like Kibra that is saturated with a lot of community-based organizations, very few individuals and organizations focus on CVE and therefore there is less competition over ideas or resources. A community-based organization representative notes that, unlike poverty reduction programs, CVE does not attract a lot of attention because there are few financial resources involved. When disagreements do occur, though rarely, respondents suggested they happen for a variety of reasons: 148 1. Managing community expectations – There are not harmonized rates for transport reimburses or allowances. While some organizations choose to pay more, others may not, and this has affected the ability of less resourced organizations to mobilize in the community. During community consultations, activities that are seen to attract more personal financial benefits are prioritized over those that may bring less, even when they happen to be more effective. 2. Competition over resources – Numerous respondents indicated that funding is limited, and available sources of funding are limited, and organizations compete with each other to get the largest share of the pot. 3. Inter-group tension – A respondent noted that inter-religious tension and tribal conflict often cause disagreements between organizations. He also stated that some community members felt excluded and, often, there is no criteria provided on how participants are selected. A CBO representative noted quoted: “Sometimes because we are a very mixed community with different ethnic and religious backgrounds, we disagree on those matters during meetings. Sometimes we can misunderstand each other by views like Al-Shabaab are Muslims. It’s always difficult to discuss these issues.” This bias was highlighted in the midline, where respondents acknowledged that many organizations choose to work with their friends or family members, excluding other community members more in need, which can lead to some hostility. When disagreements occur, respondents mentioned that religious leaders mediate to find a common ground, which is often an effective strategy at resolving these disagreements. 149 14.5 Section V – Community-Driven Strategies and Initiatives This section addresses the community-driven CVE strategies that are underway in Kibra. A summary of the result area index score is presented, followed by a discussion on participation in NIWETU￾sponsored activities, different types of CVE initiatives taking place, the effectiveness of those activities, and the level of engagement of the community. 14.5.1 Community-Driven Strategies and Initiatives Index Score The table below presents average scores from four key questions on a five-point scale38 about the effectiveness of community-driven initiatives to counter violent extremism. Scores indicate slightly decreased collaboration between community-based organizations and other stakeholders in the community, in developing CVE programs over the past year compared to the baseline and midline. Coinciding with views expressed by respondents in earlier sections, CBOs are less active now than they were 12 months ago, and the perception of the extent to which community members engage with them has reduced. Overall, respondents still agree that these initiatives are effective in teaching others about VE and CVE. 86% of respondents affirmed that the community was much more aware about VE and CVE, as one respondent from the private sector noted that: “We have tried our best to educate people, now we have gone slow because people understand most of the things.” Questions Baseline Score Midline Score Endline Score Have local organizations worked with other stakeholders to develop programs on CVE in the last 12 months? 4.39 4.32 4.07 How effective do you think these programs are at teaching others about VE and CVE? 4.10 4.70 3.86 Are CBOs more or less active now than they were 12 months ago in implementing such programs? 3.83 3.02 2.07 To what extent do community members engage with these organizations? 4.47 4.45 3.65 Overall average score 4.20 4.12 3.41 14.5.2 Participation in NIWETU-Sponsored Activities 60% of survey respondents stated that they have participated in NIWETU-sponsored activities. Most stated that they participated in the MCPR, BAWAKI and Interfaith activities. One respondent mentioned that he was part of the Champions for Change activity. Others did not provide the name of the activity. 14.5.3 Community-Driven Strategies 81% of the respondents stated that local organizations are working together to develop CVE strategies and initiatives. As was in the midline, respondents who interviewed were knowledgeable about CVE activities in the community – referring to those few activities by organizations they have interacted with, known to be working on these issues. Formal meetings with community leaders or members were mentioned as the most common initiatives in Kibra, followed by informal conversations in the community. Mentorship in schools, meetings with religious leaders, and training sessions were also highlighted as CVE activities that are occurring. 38 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 150 Roleplaying, discussions in schools, and sports activities were less common in Kibra. Other activities mentioned include integrating CVE messaging and conversations in community clean up exercises, and talent search events. CVE Activities Occurring in Nairobi Baseline Midline Endline Informal conversations in the community 33% 75% 43% Formal meetings with community leaders or members 61% 62% 49% Meetings with religious leaders 41% 55% 23% Meetings with other community organizations 26% 49% 17% Meetings with the security sector 52% 47% 11% Advocacy campaigns 59% 38% 15% Training sessions 28% 34% 23% Sports activities 44% 34% 9% Discussions in schools 15% 30% 4% Mentorship programs in schools 35% 30% 28% Meetings with government authorities (local or national) 48% 26% 9% Other 13% 17% 11% Roleplaying 17% 13% 4% A respondent from the women’s groups category noted that CBOs’ biggest challenge in mobilizing and engaging communities was navigating the mistrust from community members who assume that organizations have benefitted unfairly from their engagement with them. In most cases community members use this rationale to demand compensation for time spent engaging CBOs. She also quotes: “It’s not easy to mobilize people. They assume the one who brings people together for community initiatives have pocketed money… mistrust is still evident. This affects locally initiated activities.” 14.5.4 Effectiveness of CVE Programs Asked about their perceptions on the effectiveness of CVE programs, respondents mentioned informal conversations in the community, formal meetings with community leaders and members, as well as mentorship programs in schools, to be more effective approaches. Perceptions of the effectiveness of activities listed has declined over time. But then again, Kibra has had few cases of violent extremism and is noted to have other pressing security challenges, it is difficult to gauge effectiveness of the initiatives when there are almost no factors to evaluate against for example youth disappearances or recruitment reports. Most effective CVE initiatives Baseline Midline Endline Informal conversations in the community 4% 39% 21% Formal meetings with community leaders or members 35% 37% 21% Meetings with religious leaders 20% 35% 16% Advocacy campaigns 33% 33% 12% Sports activities 29% 30% 9% Meetings with the security sector 20% 24% 2% Meetings with other community organizations 9% 22% 2% Training sessions 4% 22% 7% Mentorship programs in schools 15% 22% 21% Discussions in schools 2% 17% 0% Other 22% 17% 12% Meetings with government authorities (local or national) 17% 15% 5% Roleplaying 2% 11% 5% 151 Do not know 11% 4% 2% Respondents agreeing that formal meetings were more effective credited this to the fact that leaders involved are respected, and/or have a large following. As such, they are believed to be authorities who can make or influence decisions. “They target leaders. We had a seminar for all chiefs at the sub-county level on CVE. This was an initiative by the national government. And it helped to improve our understanding on VE. We have people who depend on us for security matters... when we are informed then we respond to issues and act better.” – Security Sector Representative Respondents further add that formal spaces enable diverse groups of leaders, who would not have otherwise met to network and connect in a way that feels more intentional. This has improved mutual understanding, especially on issues related to religion, security and politics. A community leader quotes that: “We meet other people from other religions, and we learn from each other… this activity has helped in clarifying misconceptions and foster good working relationships between Muslims and Christians.” Formal meetings are also said to include more people than workshops and trainings. This could be because of the costs attached to mobilizing for workshops/trainings, such as allowances or venue costs that may limit the number of those invited. A community-based organization representative appreciated formal meetings as spaces for sharing information with peers and engaging government representatives, who are much more beauracratic.eu She stated: “We share information. Get to know how to get different resources. We have been having some meetings with the county commissioner, chief, OCS and a youth officer, to update each other and organized for community activities.” A respondent noted that the formalities, venue choice and language used during the formal meetings, does not always encourage youth engagements, and leaders have not made efforts to ensure such meetings sufficiently accommodate views of the youth. Activities such as clean ups and sports were hailed as spaces more effective in targeting youth. A respondent noted that sports activities are more inclusive and less intimidating. He states: “There are no boundaries in informal meetings and sports activities. People will be heard and everyone will be part of it.” Respondents further acknowledge the value of informal conversations in the community, noting that they reach more people in the spaces in which they spend their time in the community, and that people who may not be well networked have an opportunity to learn about CVE. “It’s through a friend and a neighbor that I got new information. I trust her. Furthermore, not everyone gets a chance to attend trainings. It’s only through informal conversations that we get informed.” – Women’s Group Representative “They are the ones I have seen being used here in the community. Just small conversation, working one’s social network.” – Private Sector Representative Clean ups and advocacy campaigns, in particular, were thought to mobilize more community members and bring attention to issues across various neighborhoods. They are also viewed as avenues for 152 engaging county government representatives who are visibly missing in formal meetings and other CVE activities. This is mostly because such activities that mobilize large numbers as seen as an opportunity for strengthening political capital, but also touch on issues that are mandated to the county government. A respondent stated that: “The turn out for clean ups and advocacy campaigns is good. Community members are well represented. Interaction between leaders is face to face. And community leaders are in most cases present, including MCAs and other political leaders.” A community leader stated: “I find sports as the best avenue because it keeps the youth involved and disciplined and it also makes them keep the correct company and not get into groups like Al-Shabaab.” Such activities are also designed to share information more creatively and in simple formats which make them easy to communicate key messaging across a diverse society such as Kibra. Ensuring sustainability of CVE processes was seen as the ability for those participating to cascade the knowledge down to other community members who have not been involved. So far, he stated, mentorship programs are one of the most sustainable approaches to CVE in Kibra. This is because the skills and knowledge transferred to mentors lives beyond the CVE programs. Mentors have taken up their roles in respective spaces, even without external support. 14.5.5 Level of Engagement Survey respondents had divergent views on the degree to which these organizations are more or less active now than they were 12 months ago in leading CVE programming. Less than a third, or 27%, believe that they are at least only a bit more active, while 55% believe they are the same or less active. Perhaps notably, most CBOs themselves, the private sector, and youth representatives, believe that these organizations are less active on this issue than they were 12 months ago. Nevertheless, 79% of survey respondents believe that community members engage with these organizations at least a little bit, while only one respondent thinks they do not engage. The perception of the degree of these engagements has decreased from the midline. To what extent do community members engage with these organizations? Baseline Midline Endline Engage a lot 67% 72% 30% Engage a little 24% 28% 49% Do not engage at all 0% 0% 2% Do not know 9% 0% 19% 14.5.6 Summary of Findings Local organizations are engaging in CVE strategy development in Kibra, but this has reduced over the last 12 months. This is due to a decrease in activities that facilitate stakeholder collaboration, and mistrust from the community who question their credibility and motive. There are mixed views on the best approaches for engaging the community on CVE, but formal and informal meetings are thought to be more effective, simply because they are more inclusive and often mobilize those with authority and influence in the community. On enhancing outreach, respondents proposed that CVE 153 engagement devises more creative messaging and a mobilization strategy that reflects the social and economic diversity of communities in Kibra. 154 14.6 Section VI – Objective 2: Improved Government Capacity The table below compares scores for sentinel indicators from the midline and endline evaluations related to Objective 2. Scores for the meta perceptions indicators are again on a ten-point scale39 and reflect the same sampling dynamics as the Objective 1 questions above. Scores for the statements regarding legislation and budgets were identified through a review of the CAPs and County Integrated Development Plans (CIDPs) which spell out the county budgets. Midline scores were revised to include only Kibra respondents, cutting the sample size by half. Level of government and community participation in community-government dialogues Midline Endline When I express my views about violent extremism to members of the government, I feel like my opinion is considered and my concerns and ideas are responded to. Indiv. 4.65 6.19 Meta 7.03 4.43 The government has made a strong effort at participating in community dialogue over the past two years. Indiv. 7.68 4.77 Meta 4.42 6.18 I feel like I have been adequately included in the county government’s response to peace and security. Indiv. 3.47 3.99 Meta 3.38 4.96 Overall Score 5.11 5.09 Community perceptions of the national and county government on issues relating to peace and security Midline Endline The national government has made a great effort at improving the peace in my area over the last two years. Indiv. 3.43 4.18 Meta 5.65 3.91 The county government has made a great effort at improving the peace in my area over the last two years. Indiv. 5.67 3.80 Meta 5.18 4.96 The security sector has made a great effort at improving the peace in my area over the last two years. Indiv. 5.83 5.48 Meta 5.60 4.82 My community has become more willing to cooperate with the security sector over the past two years. Indiv. 6.35 6.30 Meta 5.45 6.25 Overall Score 5.40 4.96 Midline Endline Presence of line items for CVE activities in county budgets 0 0 Presence of legislation to support implementation of the CAPs 0 0 Respondents’ views of government and community participation in dialogues remain largely unchanged, but there was a noticeable decline in individuals’ views on the effort that government has made in participating in community dialogues. Still, respondents think their peers would express more positive views to this question, suggesting perhaps that the government is still making an effort, but dissatisfaction with these efforts may have increased. Positive perceptions of the national and county government on issues relating to peace and security also declined slightly, though views can still be considered to be somewhat neutral on these questions. Perspectives are most positive on the community’s willingness to cooperate with the security sector, as compared to two years ago, perhaps in part due to NIWETU’s efforts to bring together residents and security agents in positive engagements. Respondents’ views are more negative on the county government’s efforts at improving peace in the community over the last two years. While at the midline evaluation, scores were somewhat positive 39 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 155 on this question, individuals’ views are more negative, while perspectives on peers’ views are simply neutral. Preoccupation with political wrangling over the past year could explain this decrease. There has been no change in scores related to the presence of line items for CVE activities in the county budget or legislation to support CVE at either the local or national levels. As these processes take time, and CIDPs outline budget priorities over a five-year period, it is no surprise that there has been no change in this regard. 14.7 Section VII - Government Responsiveness to CVE The following section presents findings on the responsiveness of county and national level government to address VE. Summary findings are presented, followed by a more in-depth analysis of the findings. The section will cover perceptions of both levels of governments’ understanding of and willingness to engage on VE and CVE issues. It will then look at engagement with various government counterparts, as well as the challenges faced in dealing with government actors. 14.7.1 Government Responsiveness Index Score The below questions are scored on a four-point scale40 to understand overall government responsiveness to VE at both county and national levels. The scores presented below demonstrate a slight overall regression in perceptions of government’s responsiveness to the issues. Perceptions of the county administration are least positive compared to midline. NIWETU supported Nairobi’s CAP development last year. Unlike the other four counties, the Nairobi CAP formation was a rapid process where consultations were done in a few days. The Nairobi CAP, at the time of the endline, was yet to be launched. A key informant noted that a public launch had been scheduled for the 23rd March 2020, but due to the Covid-19 response measures, it was postponed to a later date. Reflecting meta-perceptions scores highlighted above, views on the county government’s willingness to address VE declined as compared to the midline, perhaps because of preoccupation with other political issues at the time. Questions Baseline Score Midline Score Endline Score How well do you think the county government understands the problem of VE in this community? 2.56 2.09 2.21 How well do you think the national government understands the problem of VE in this community? 3.13 3.23 3.26 How willing do you think county government is to work on issues of VE? 1.68 2.12 1.98 How willing do you think national government is to work on issues of VE? 3.21 3.00 3.21 How easy do you think it is to work with national government counterparts? 3.09 3.57 3.02 Overall average score 2.73 2.80 2.73 14.7.2 Government Understanding of VE Survey respondents’ views on the degree to which the county government understands the problem of violent extremism were rather negative. Only 15 respondents totally believe that the county 40 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 156 government is somewhat knowledgeable about the problem of VE, no respondents believe that the county administration understands the problem very well. How well does the county government understand VE? Baseline Midline Endline Very well 35% 6% 0% Somewhat well 45% 23% 35% Not very well 3% 36% 58% Not at all 8% 27% 7% Do not know 10% 6% 0% Key informants agreed that the county government’s understanding of the issues is limited, and the level of their engagement has been minimal during and after the CAP formation. A key informant stated that: “The county government of Nairobi may not have had this as a priority. Most of the engagements were with the national government, especially the County Commissioner’s office. I think that the county government understood CVE as a security matter but not a social issue and thus felt it was a national government role as they are in charge of security. The county did not find the need to invest in the formation and the sensitization process.” Another explanation for the minimal engagement by the county government in Nairobi would be increased political disputes which led to a high turnover of county officials, the governor facing challenges, an impeachment motion, and final takeover of key functions by the national government. A key information noted that there just: “Seemed to be less interest on CVE issues from the county government, e.g. the governor has never engaged in the process, but has sent representatives who were not consistent and often made a technical appearance.” Views of the national government were slightly more positive. Only two survey respondents, or 5%, believe that the national government does not understand the issues very well. The remaining 41 respondents believe that the national government understands the issues well. A key informant noted that, security being a mandate of the national government meant that national government officials were responsible and committed to understanding the problem and engaging stakeholders. 14.7.3 Willingness to Engage in Programming Survey respondents’ perceptions on the county government’s willingness to work on CVE were also negative. Only nine respondents believed that the county government was somewhat willing to work on issues of CVE. How willing are county government is to work on CVE? Baseline Midline Endline Very willing 9% 2% 2% Somewhat willing 26% 30% 30% Not very willing 44% 36% 61% Not at all willing 17% 23% 12% Do not know 4% 9% 5% Once again, survey respondents’ views toward national government officials were more positive. More than three quarters of survey respondents believe that national authorities are at least somewhat willing to work on CVE issues, while the remaining 16% believe they are not very willing, or do not 157 know. No respondent stated that national government officials are not at all willing to address the issues. How willing are national authorities to work on CVE? Baseline Midline Endline Very willing 22% 23% 40% Somewhat willing 67% 53% 44% Not very willing 7% 23% 14% Not at all willing 0% 0% 0% Do not know 4% 0% 2% In considering the national government’s willingness to address the issues, respondents mentioned having engaged national government representatives in various activities at the community level and interacting with them as stakeholders in various CVE trainings and awareness creation forums. At least 70% mentioned specifically engaging with the local chief. Others mentioned the County Commissioner, the deputies, and Officer Commanding Police Station (OCPS). One youth leader acknowledged having worked with the local youth officer. A key informant further noted that other national government departments were supportive and involved in CVE forums, specifically CAP formation process, but were yet to formally sensitize the public on CAP and what is expected of them. A key informant from the national government noted that: “Organizations that work on issues of CVE are aware about CAP and may have engaged the community, but as government we are yet to. I am aware that boda boda leadership which is part of the county engagement forum held meetings with their members.” 14.7.4 Engagement with Government Counterparts As with the midline, engagement with government counterparts, at both county and national levels, was minimal, according to survey respondents. The largest proportion of respondents noted that they occasionally engage with officials, and at the county level, 44% stated that they rarely engage with government counterparts, pointing to a disconnect between communities in Kibra and both local and national authorities, and demonstrating an opportunity to promote more regular engagement. During the midline, it was expected that the CAP formation process would likely shift responses to these questions toward more regular engagement and interaction, but this has not been the case. A key informant stated that: “The CAP process was a presidential directive and that meant that it was done very quickly and even the engagement initially did not include everyone. I found that there were gaps in the participation processes and most of those who led the process were from the security personnel.” Frequency of engagement with county government counterparts Baseline Midline Endline On a daily basis 0% 4% 2% Once a week 0% 2% - A few times a week 2% 4% 2% Once a month 0% 0% 2% A few times a month 2% 13% 7% Occasionally 33% 38% 41% Rarely 44% 38% 44% Do not know 20% 0% 0% 158 To that end, challenges exist in the relationship between government agencies and residents in Kibra. Predictably, 40% of respondents agreed that the county government does not seek to involve them, while 28% noted that they do not attend scheduled meetings when invited. A security actor, CBO and youth representative, note that the county government is conflicted over what roles they should play on issues of security because: “They have no structures to engage on CVE.’’ While security is not a devolved function, there are opportunities for collaborating with the national government on issues of CVE that go beyond strategic security. In addition, the CAP provides platforms for county government engagement and, with political will, it is an opportunity to strengthen awareness of CVE within the county government structures. One respondent also noted that the county government claims not to have resources for engaging on issues of CVE, and instead, they only reach out into communities during elections. Frequency of engagement with national government counterparts Baseline Midline Endline On a daily basis 11% 13% 9% Once a week 9% 6% - A few times a week - 13% 9% Once a month 5% 2% 2% A few times a month 14% 19% 7% Occasionally 36% 45% 42% Rarely 7% 2% 44% In contrast, 79% of respondents stated that there are no problems working with national government counterparts. Although 35% state that national level actors do not show up to meetings when invited, 67% of all respondents noted that this engagement has been through formal meetings. Others noted that their engagement included meeting with security personnel and informal meetings in the community. 14.7.5 Summary of Findings Perspectives on the county government’s willingness to engage in and understand VE and CVE-related matters is comparatively low in Nairobi. The county government has demonstrated minimal interest in engaging in the CAP formation process, due to distraction from ongoing internal political struggles and lack of clear formal guidelines for engaging on security issues, since this is a mandate of the national government. Still, as was expressed in baseline and midline, there remains a high degree of frustration over how government officials act and treat residents in hotspots for extrajudicial killings, and disappearances are prevalent. In this environment, it becomes difficult to work in a collaborative and consultative way with government counterparts; but there has been an improvement in community￾government engagements. 159 15. ANNEX VI – WAJIR COUNTY REPORT: WAJIR COUNTY ENDLINE OVERVIEW Data collection took place in Wajir Town over 11 days in February and April 2020, carried out by two researchers. NIWETU facilitated introductory meetings to key stakeholders to explain the goals of the research and to facilitate further introductions to relevant actors. The research team conducted both qualitative and quantitative research with key stakeholders in Wajir, as well as the wider community. The breakdown of quantitative respondents in the stakeholder questionnaire is as follows: Category of respondent Number of respondents County administration 4 Security sector 10 Educational professionals 6 Youth representatives 4 Private sector 2 Community-based organization 11 Media 7 Women’s groups 1 Religious and community leaders 4 Total 49 Key informant interviews were carried out with eight individuals who include NIWETU staff, county and national government officials, local civil society organizations, and members of NIWETU’s Champions for Change program. 160 16. KEY FINDINGS 16.1 Section 1 – The VE Context The following section presents findings of the broader VE and security context in Wajir County. The section then discusses the VE context, with synthesized findings from key informants and the quantitative survey respondents. Findings will address the wider peace and security context, perspectives on youth disappearances, views on extremism and recruitment, organized crime, and the safety and security context in Wajir County and Kenya more broadly. 16.1.1 Summary of Findings Perceptions of the VE context in Wajir have deteriorated. This change has been influenced largely by an increase in VE attacks, especially in Wajir East sub-county. Despite the fact that the data collection for the endline was conducted amid such attacks, a notable number of respondents still believe the VE context has improved. Key informants viewed this as a turning point that not only validates CVE coordination mechanisms like the CEF, but also challenges the community to engage more. They not only credit this change to intensified CVE activities that have brought different stakeholders together and created more awareness, but also to counter-terrorism measures that act as a deterrent to engaging in VE-related activities. There has been no notable change in the frequency of youth disappearances, where youth are believed to have joined VE groups or been victims of police killings. Unlike the midline, there is a growing acknowledgement that gang violence and organized crime is a problem, though not all that significant. 16.1.2 Causes of Insecurity Perceptions on the main security threats in Wajir point to a wide range of issues, with the most significant being violent extremist groups followed by drug abuse. The relevance of these two threats remains consistent from the midline through the endline. Other significant sources of insecurity mentioned were high levels of unemployment, inter-ethnic tension, radicalization, local politics and high levels of poverty. Less relevant to the Wajir context were criminal gangs, national politics and inter-religious tension, likely because Wajir County is comprised of a rather predominantly Muslim community. Other issues highlighted include human trafficking, lack of access to justice, and resource￾based competition. Security threats Midline Endline Violent extremist 81% 80% Drug abuse 75% 72% High levels of unemployment 93% 66% Inter-ethnic tension 90% 66% Radicalization 81% 64% Local politics 74% 60% High levels of poverty 88% 58% Lack of education 79% 54% Youth disappearances 72% 54% Land disputes 74% 44% Theft 44% 24% National politics 44% 16% Criminal gangs 21% 10% Inter-religious tension 14% 8% Other 5% 0% *Respondents misinterpreted “inter-ethnic” tensions as both inter and intra-ethnic/interclan. However, the data still captured interclan variables. 161 Interviews were conducted in February through March 2020 when Wajir, just like other North￾Eastern counties, was experiencing increased and sporadic AS attacks which mostly targeted non-local professionals, security actors and key infrastructure, such as telecommunications masts. This had a significant impact on the community, as teachers hired from outside of the North-Eastern region began fleeing for fear of further attacks from the terror group. This paralyzed the education sector and sparked protests in Wajir town when the Teachers Service Commission honored a request from hundreds of teachers (from both affected and less affected areas) for transfers. This was a key focus of CVE discussions among stakeholders at the county and national level in the first quarter of 2020. Other significant security threats were noted, such as unemployment, poverty, drug abuse, youth disappearance and inter-ethnic tensions, interpreted by respondents as inter-clan tensions, can be closely linked to terrorism and radicalization as either drivers or catalysts in the context of Wajir. Although VE recruitment in Wajir is not based on clannism, VE groups have previously manipulated clan ties in Kenya and Somalia to operate in both countries. AS has also pushed narratives of victimization and alienation in an attempt to pit Somalis and Muslims against the government and other ethnic communities. This weakens the community’s resilience to VE, because the communities and leaders will be divided along these lines when they should be focusing on collective action. Consequently, the government and other stakeholders have intensified community-security engagement within Wajir and among border communities. 16.1.3 Disappearances Survey respondents acknowledged that youth disappearances still occur, but with a slightly higher level of frequency than they did during the baseline and midline. A majority of 41%, however, note that disappearances are rare. Those who thought that youth disappearances occur very regularly include a security actor, private sector, media, women’s group representatives, and two county administrators. Four of those who believed it never occured were from the security sector and religious/community leaders. Eight respondents either chose not to answer or did not know whether youth disappearances occur in Wajir. How often do youth in Wajir County disappear? Baseline Midline Endline Very regularly 7% 12% 12% Somewhat regularly 25% 18% 22% Rarely 45% 40% 41% Never 20% 18% 8% Do not know 2% 12% 16% According to the respondents, when youth disappear, they were believed to have left for Somalia, Syria, or South Africa, or are trafficked (“Tahrib”) to European countries via Libya in search of employment and other economic opportunities. These responses indicate that communities are knowledgeable about Wajir’s reputation as a transit route for human trafficking. (The Kenya National Commission for Human Rights identifies Wajir, as border county to Somalia, as a major human trafficking transit point.) Others noted that those who go to Somalia and Syria join terror groups such as AS. A women’s group representative who believes youth disappearances occur very regularly explains that those who disappear are victims of extrajudicial killings and further refers to the mass graves discovered in Wajir (in 2015, 2016 and 2017). A private sector representative noted that: “It is possible that the military takes them, they’re killed.” 16.1.4 VE and Recruitment Despite the significant threat posed by VE activity in Wajir, more than half of survey respondents were still in broad agreement that the VE context has improved over the past 12 months. Only 25% of respondents disagreed, while 12% pointed to no change on this subject. 162 Extremism and recruitment to join extremist groups is less of a problem now that it was 12 months ago. Baseline Midline Endline Strongly agree 57% 52% 27% Somewhat agree 18% 21% 29% Somewhat disagree 5% 16% 22% Strongly disagree 12% 5% 14% Do not know 7% 5% 8% Despite the significant threat posed by VE activity in Wajir, survey respondents were still in broad agreement that the VE context has improved. However, a slightly higher percentage than the midline, or 26% of respondents, disagreed. Still, 12% pointed to no change on this subject while others did not have an opinion on this. Key informants also generally agreed but noted that the increase in VE activity seen in the January through March period had been an awakening call to strengthen existing structures, including implementing recommendations under the CAP to tackle the threat of VE. Those who believed that the VE context had improved credited this change to the increased awareness on VE and CVE strategies and strengthened relations across different actors in the county. A community leader noted that the community increasingly acknowledges that recruitment is also a direct security threat to them, because “once this happens, they will come back to the community to commit crimes.” Another community leader who strongly agreed that the VE context has improved credits this particularly to the efforts in engaging youth and women on CVE issues, noting that it has effectively “reduced requirement of joining VE”. Others who also agreed with the statement note that the counter-terrorism measures have been a huge deterrent for engaging in VE activities, especially among youth and VE group members. Respondents noted that fear has pushed community members to share information with security officials, and measures to curb the illegal movement of VE members into Kenya and within the region have been effective. “Extreme security measures have pushed the terrorist to the wall.”– Security Sector Representative “Because of fear, security has improved, some communities are giving out information to government Because if someone is suspected, he is picked by security agencies and never seen again.” – Security Actor Representative Some felt that evaluating the context of VE based on the number of VE incidents can be misleading. A respondent from the private sector cautions that the reduced incidents could be due to fear of reporting, because of the potential for retribution by VE actors or victimization by security agents. In addition, a community-based organization representative suggests that the increased CVE activities have forced VE groups “to change tactics, for example during rainy seasons, security forces decrease and VE recruitments increase” and may not necessarily have reduced. A media representative stated that, at times it is hard to confirm who is behind increased attacks on the police stations: “Attacks on police have increased, but we do not know who the attackers are. They might be AS assailants or locals.” A community-based organization representative who thought that VE context had not deteriorated noted that his rationale was due to the spike in attacks at the time of the interview, while another respondent noted that continued police brutality and extrajudicial killings keep justifying VE narratives. 163 “There has been no change, because if someone is suspected of engaging with VE actors, he is picked by security agencies and never seen again.” – Security Sector Representative Others feel that the factors that increase youth vulnerability to VE, such as unemployment and poverty, are still high in Wajir. A CBO representative who strongly disagreed with the statement that the VE context has improved said, “The local economy has been going down forcing youth to look for opportunities outside the town.” 16.1.5 Gang and Criminal Activity As was observed during the midline, survey respondents at the endline demonstrated broad agreement that gangs and criminal activity are less of a problem now than 12 months ago; however upon further probing, survey respondents were more split over the existence of gangs and organized crime in Wajir than they were during the midline. “There have been no attacks by organized groups in the county in the last year.” – Youth Representative While some noted that gangs and organized crime do not exist and therefore is not a problem, others did mention that they are concerned over illegal trade and contraband that fuels extremist groups and also highlighted the prevalence of banditry activities. A security actor, however, notes that banditry has reduced as a result of interventions by religious and Nyumba Kumi leaders, and increased security surveillance in the county. Organized crime and gang activity are less of a problem now than it was 12 months ago. Baseline Midline Endline Strongly agree 47% 42% 22% Somewhat agree 2% 11% 20% Somewhat disagree 2% 18% 27% Strongly disagree 2% 18% 22% Do not know 45% 12% 8% While a security sector actor stated that “organized crime is not a problem and that Wajir has other big problems”, a youth leader, on the other hand, claimed that crime had increased over the last year, but the research could not establish the extent to which this was gang-related or not. Three respondents acknowledged that there is an increase in gang activities, especially involving the youth and non-locals. A community-based organization stated that: “A lot of non-locals both Somalis and non-Somalis who come from other regions who move to Wajir for economic reasons and fail to find jobs often are involved in crime.” Although illegal human trafficking is not listed as an example of organized crime in this section, a few respondents mention it in explaining where youth who disappear or where they are thought to have gone. 16.1.6 Safety and Security More than half of the respondents disagreed that safety and security in Wajir has improved over the past 12 months. This could have been influenced by a series of sporadic VE attacks in Wajir at the time of the endline between February and March 2020. Despite these attacks, 43% of respondents still believe that the situation has improved over the year. One media representative, community leader and CBO representative each noted that there has been no change. 164 The pessimism was also reflected in the respondents’ perception of the context of safety and security related to VE at the national level. 53% tend to disagree with the statement. The same explanation would apply as the number of terrorist attacks between November 2019 to March 2020 had increased not only in Wajir, but in the North-Eastern and Coastal regions. All incidents that drew national attention affected non-locals coming from diverse counties and escalated conversation on the risks that VE possesses to the country. 165 16.2 Section II – Objective 1: Community Mobilization to Address VE Enhanced The table below compares scores for the sentinel indicators from the midline and endline evaluations on indicators that track communities’ willingness to mobilize in support of peace and security objectives at the county level. While NIWETU does not work to support these indicators, we assume that communities’ willingness to engage with these structures, or their trust in these structures’ ability and relevance, will affect their perspectives on the wider community’s willingness to mobilize around CVE. Individual scores reflect respondents’ personal views on the corresponding statement, while the ‘meta’ scores, or meta perceptions scores, reflect how respondents believe their peers would answer. Research has shown that respondents are more likely to act in accordance with what they think are their peers’ views, or the social norm. This means that when meta scores are higher than the individual scores, views in general tend to be more positive about a particular statement, and vice versa. These findings come from a sample of 40 individuals in two neighborhoods in Wajir Town and should therefore not be considered statistically significant or representative of a wider sample of Wajir residents. Figures presented below are out of a ten-point scale41. Degree to which peace committees are active and responsive. Midline Endline I feel that the peace committees are active and easy to access in my area. Indiv. 6.27 7.19 Meta 7.92 6.52 I feel that the peace committees have an adequate understanding of the issues in my community. Indiv. 6.98 7.61 Meta 7.67 6.72 I feel that peace committees effectively respond to issues in my community Indiv. 6.84 6.94 Meta 7.47 - 42 Overall score 7.19 7.00 Degree to which communities actively engage in barazas. Midline Endline I frequently attend community barazas. Indiv. 5.55 5.24 Meta 7.11 6.34 When I go to barazas, I feel like my opinion is considered and my concerns are responded to. Indiv. 4.20 6.10 Meta 6.20 6.74 I always feel positive after participating in barazas. Indiv. 5.46 7.28 Meta 5.16 7.6 Overall score 5.61 6.55 Degree to which Nyumba Kumi initiatives are active. Midline Endline Nyumba Kumi initiatives are active in my community. Indiv. 4.44 6.80 Meta 6.56 6.64 I feel that Nyumba Kumi initiatives are positively addressing the issues that concern me. Indiv. 4.98 6.87 Meta 5.58 6.52 I have confidence in Nyumba Kumi initiatives and trust that it will help improve peace in my area. Indiv. 5.73 9.41 Meta 5.56 7.58 Overall score 5.48 7.30 Presence of other organizations/actors that work on peace/security. Midline Endline There are numerous other networks and organizations making a positive impact on peace and security in my community. Indiv. 5.72 5.22 Meta 5.92 4.96 41 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 42 Data for this score was lost during data transfer. As such the overall average score for this indicator will exclude this figure. 166 Overall score 5.82 5.09 Community members’ views on the relevance and utility of various peace and security structures in Wajir township were largely positive during both the midline and endline evaluations, but still showed an improvement at the time of the endline. Respondents overwhelmingly agreed that peace committees have been relevant in the Wajir context, likely because of the role peace committees have played in resolving inter-clan disputes in recent years. Perspectives were also largely positive on the role of community barazas. While during the midline evaluation respondents were fairly neutral in their views on barazas, by the endline they tended to feel more positive. Respondents mostly agreed that they feel positive after attending community barazas. Likewise, perspectives on the degree to which Nyumba Kumi initiatives are active and effectively working to improve peace were overwhelmingly positive, particularly with regard to respondents’ confidence in Nyumba Kumi. There was a significant increase in both individuals’ and peers’ perspectives on this question as compared to the midline suggesting that confidence in Nyumba Kumi has grown over the past year; this research did not aim to determine why. These findings generally suggest that existing peace and security structures could be leveraged to support CVE initiatives in Wajir township; if community members trust these structures and initiatives and have confidence in their effectiveness and relevance to addressing security related issues, they are more likely to respect their involvement in CVE activities. Working with individuals who are members of local Nyumba Kumi activities, chiefs who often lead barazas, and members of peace committees are effective entry points to gaining their support for CVE programming. 167 16.3 Section III – Skills and Knowledge The following section looks at the skills and knowledge that exist within the stakeholder community on VE and CVE related activities. The primary focus is on analyzing the different types of training that exist on these topics, as well as the degree to which respondents have participated in such activities already, and what, if anything, they are doing with the skills and knowledge thereafter. Following the summary of findings and presentation of the index score for this result area, this section discusses participation in pieces of training, the degree to which they have been helpful and why, and how to make them more effective. 16.3.1 Summary of Findings Survey respondents view training activities in Wajir as important and effective activities and sharing what they learn with the wider community has been consistent since the midline. Respondents also express a desire to strengthen their understanding of violent extremism, perhaps as a way to keep up to date with changing dynamics in recruitment and VE strategies. A secondary outcome from the training sessions is the strengthened relationships among participants across different categories. At the same time, however, there are clear ways in which these activities could be strengthened, through stronger facilitation, ensuring consistency of the sessions, follow-ups with tangible outcomes and skills, and including a wider network of individuals, especially among security agencies and community members in the more vulnerable areas in the county. 16.3.2 Skills and Knowledge Index Scores The following table shows the average scores on a four-point scale43 from several key questions related to skills and knowledge. Baseline, midline and endline scores are presented side-by-side to highlight changes over the past year and indicate a consistent improvement in the culture of information sharing on CVE issues over time. At the endline, the number of those who participated in training sessions had decreased slightly, though most respondents still participated. This figure translates into the number of those who used the information they learned or thought the training sessions had helped them understand VE and how to prevent it. Questions Baseline Score Midline Score Endline Score Have you participated in any training on CVE and understanding the causes of VE in the last 12 months? 2.75 2.95 2.64 How much, if at all, do you think these trainings have helped you understand VE and how to prevent it? 3.71 3.76 3.60 How much, if at all, do you use the information you have learned from these trainings? 3.71 3.68 3.44 Have you shared the information you have learned with others in the community? 3.90 4.00 4.00 Overall average score 3.51 3.60 3.42 16.3.3 Participation in Training The number of those who participated in the CVE trainings dropped by ten percentage points at the endline compared to the midline survey. There were no representatives from the private sector who 43 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 168 had participated in training. Only one CBO and women’s group representative had participated in CVE training sessions. At the midline, women were under-represented too. Community and religious leaders had the highest representation, followed by security sector representatives. The composition of participants may have been influenced by the themes of the training or the kind community networks leveraged by training partners in the mobilization. Category Midline Endline Participated Did not participate Participated Did not participate CBO 3 2 1 3 Community/Religious leader 4 3 8 3 County administration 6 2 3 1 Educational professional 1 6 4 2 Media 3 2 5 2 Private sector 3 3 0 2 Security sector 7 1 6 2 Women’s representative 3 1 1 0 Youth representative 0 7 3 1 Total 43 (75%) 14 (25%) 31 (66%) 16 (34%) The majority of the respondents who attended trainings indicated that they had attended NIWETU￾sponsored CVE trainings. Asked which organization had led the training, several were mentioned including: Wajir Peace Development Agency (WPDA); African Social Development Focus (ASDEF); Wajir Youth Bunge; RDI, RACIDA; and C4C. Others mentioned activities not sponsored by NIWETU such as, North Link Development Organization (NODO), BRAVE, UN Women, DDG, and AMKA. As seen during the midline, a majority of respondents reported participating in a wide range of training topics, with over one third having participated in training on understanding VE, why people join VEOs and how to prevent or counter-extremism. As was with other counties, fewer respondents participated in monitoring and evaluation training. Types of Training Baseline Midline Endline Understanding violent extremism 50% 97% 80% Why people join violent extremist groups 50% 97% 71% How to prevent or counter violent extremism 43% 97% 78% How to work with the government to prevent or counter violent extremism 38% 79% 58% Training others on preventing or countering violent extremism 23% 74% 56% How to work with other organizations to prevent or counter violent extremism 42% 66% 49% How to report extremist activities 35% 76% 47% Conflict resolution 20% 76% 44% Leadership skills 20% 74% 44% How to talk to others about violent extremism 40% 76% 42% Advocacy skills 15% 61% 38% How to lead interfaith discussions on violent extremism 28% 53% 36% How to identify recruiters and recruitment methods 15% 68% 33% Monitoring and evaluating 8% 32% 20% Other 13% 5% 20% Identifying signs of recruitment 15% 0% 0% Asked why they found the training helpful, most noted that the skills gained helped them in their regular work, while several others mentioned having learnt something new or strengthening existing skills and knowledge on CVE issues. Some also noted having met new colleagues during the training. 169 All respondents who participated in training found them helpful, with a majority of 78% indicating that they benefited a lot. Survey respondents also demonstrated a relatively good understanding of violent extremism related concepts, yet at times a biased understanding of the issues, which suggests they may not be framed the right way in training activities to date. While many understood VE to be the violent expression of an ideology, almost all the respondents equated ideology to religious narratives and extremist interpretations of Islam. Only two, a youth and county administration representative, considered socio-political factors as drivers toward extremism. Also, several respondents did mention creating divisions along religious and social identity as a key strategy of violent extremism. A security sector representative considers violence and inhuman treatment from security agencies as a form of violent extremism. He quotes: “Violent extremism can come from two sources i.e. either from terrorist groups or from security agencies. CVE is the process of changing the views and understanding of radicalized persons to influence them to their loyalty from the terror groups to the government.” Further definitions of CVE were also less clear; survey respondents simply mentioned that CVE means “strategies applied to fight ideologies that lead to radicalization”, or “creating awareness on VE.” Several respondents defined CVE by providing examples of activities they had come across, for example, “mentorship programs, workshops and community meetings” on CVE. It is worth noting that the words and concepts around CVE or VE are often presented in English are difficult to define or translate in different languages and cultures. Often when translation fails, it is easier to offer examples rather than explanations. This research illuminated overwhelmingly positive outcomes from the trainings that have taken place in Wajir. Perhaps in contrast to the other counties of research, specific examples emerged of how the trainings have benefited local communities, as well as which trainings have helped attain those outcomes. “I talk to students on the importance of moderation in religion.” – An Educational Institution Representative Improving information sharing was the most significant outcome. Participants mentioned that they have used the knowledge and skills to train and mentor others on CVE within their networks. Media representatives have used the knowledge to create documentaries and improve their content in reporting on CVE in the county. A CBO representative notes that she shares information “everywhere, with all groups (women, mothers, daughters, youth) and speaks about the issues of radicalization in schools.” Security sector actors noted that the training has been instrumental in helping them understand recruitment methods and respond better in managing risks related to VE. One security actor listed the following: “Identification of terror cells, identification for individuals in terror groups, strengthening counter terrorism skills and integrating CVE skills in when responding terrorist-related incidents.” 16.3.4 Improving Training Programs Respondents provided several suggestions on how to improve the training. They focused on consistency, content, inclusivity, ownership and sustainability. Firstly, most respondents noted that CVE training does not frequently cover more remote locations and those close to the border of Somalia, such as Konton and Khoraf-Harar, where communities are more vulnerable to VE. However, it is worth noting that this research took place only in Wajir town and did not reach respondents in 170 these more remote localities, where NIWETU has carried out activities. As such, perspectives of the respondents captured in this report are not reflective of those in remote border towns, who may feel that the activities that have reached them are sufficient. Others generally proposed that such trainings extend to all sub-counties and wards in Wajir. During the midline, fear of retaliation by VE actors was noted to be a hindrance in hosting intensive CVE activities in remote areas of the county. In ensuring that such activities do not harm, the midline proposed that more consideration and consultation needed to be made before “implementing large scale programs in these areas.” Secondly, as was indicated during the midline, respondents emphasized the need for consistency in the training activities. Respondents alluded to the fact that trainings require stronger facilitation and adequate length of time to make sure that the messaging is consistent, and the quality of information delivered is strong, especially if those trained are to effectively interpret and disseminate the knowledge in the local context and language. This is viewed as a more sustainable approach and one that enhances local ownership. A community-based organization representative noted that it was important that such a session is conducted by qualified personnel. “Training should be done in local dialect to make sure everybody understands the concepts.” – CBO Representative Lastly, others mentioned the importance of including security actors to help them appreciate and understand strategies for strengthening community-security relations. A respondent from the security actors’ categories also listed key agencies that need to be integrated, including the Ministry of Immigration, national intelligence services, police, and office of the Director of Public Prosecution. “There should be more trainings with security and community to decrease the mistrust between the two, and more trainings for women and youth.” – Women’s Group Representative Respondents also pointed to specific future training needs. Most respondents would wish to participate further in training around understanding violent extremism, why people join violent extremist groups, and how to prevent violent extremism. Other themes proposed include how to work with other organizations to prevent or counter violent extremism, and how to report extremism activities. Other trainings mentioned include understanding the root causes of violent extremism, dealing with drug abuse in the county, and how to work with the county government on issues of CVE. As was noticed in other counties, respondents were less inclined to learn more about monitoring and evaluation. What types of trainings would be most useful? Baseline Midline Endline Understanding violent extremism 71% 73% 71% How to prevent or counter violent extremism 64% 59% 64% Why people join violent extremist groups 62% 68% 62% How to work with the government to prevent or counter violent extremism 42% 57% 42% How to work with other organizations to prevent or counter violent extremism 40% 52% 40% Training others on preventing or countering violent extremism 40% 70% 40% How to report extremist activities 40% 57% 40% How to talk to others about violent extremism 36% 57% 36% How to lead interfaith discussions on violent extremism 31% 55% 31% Leadership skills 27% 55% 27% Advocacy skills 27% 54% 27% How to identify recruiters and recruitment methods 24% 57% 24% 171 Conflict resolution 24% 55% 24% Monitoring and evaluating 11% 43% 11% Other 7% 14% 7% 16.3.5 Sharing Lessons Learned There is an evident desire for participants across all the categories to share the acquired knowledge on VE with others in both formal and informal spaces. This is a significant increase from the baseline, especially when it comes to sharing information and appreciating the importance of mainstreaming of CVE knowledge at the household level (through family and friends). Only two respondents, a media and educational institution representative, acknowledged that they did not share what they learned. The survey found that the participants mostly shared new skills with their close networks with whom they would have more frequent and organic interaction, such as family and colleagues at work. Others often shared with fellow community members and friends or school mates. The decline in sharing lessons from the midline to the endline is consistent with the reduction in activities as is shown later in section V. This perhaps indicates that the intensity of information sharing is highest immediately after participating in CVE activities, but dissemination also continues long after that. With whom have you shared your new skills and knowledge? Baseline Midline Endline Family members 13% 86% 71% Colleagues in my organization 40% 81% 71% Community members 42% 89% 69% Friends/school mates 10% 81% 62% Community leaders 32% 72% 56% Colleagues at other organizations 20% 50% 44% Colleagues in the local government 22% 56% 31% People from my mosque/church 20% 67% 27% Business partners 12% 42% 18% I have not shared the information or skills with anyone 3% 0% 7% Other 20% 6% 4% 172 16.4 Section IV – Community Networks The following section assesses the existence and strength of community-based networks working to counter extremism in Wajir. A summary of survey responses followed by a discussion of collaboration between actors, and conflicts between organizations is presented below. 16.4.1 Summary of Findings Networks are continuing to strengthen in Wajir while disagreements continue to happen. When they happen, conflicts and disagreement are largely about resource competition, duplication and arguments over ideologies in designing CVE narratives. The CEF is viewed as an essential platform for managing these conflicts and coordinating actors to ensure the limited resources are used efficiently. The forum is quite inclusive as no CVE actors are allowed to work outside the CEF framework. However, the efficiency of the forum can be improved to allow for more synergies in how actors implement, monitor and measure their work as a collective. This coordination could also be reflected among donors to ensure better grant allocation polities that are more transparent to manage conflicts over donors funding among CEF members. 16.4.2 Community Networks Index Score The table below highlights the average scores of five questions (on a five-point scale44) related to establishing community networks to counter violent extremism. The scores indicate that there has been a significant increase in those working with other stakeholders on issues of CVE in the last 12 months. However, there has been an increase in the frequency of conflict between actors since the baseline, likely because of increasing competition for limited resources. Although the overall score on this section is slightly low compared to the midline, the frequency of collaboration with other actors remains comparatively high. Questions Baseline Score Midline Score Endline Score Do you work with any other organizations on CVE? 4.86 3.46 4.29 Do you collaborate more, less, or the same amount with these actors than you did 12 months ago? 3.77 4.57 4.51 How often are there conflicts or disagreements between different actors working on CVE?45 4.14 3.70 2.79 Overall, do you think improved networks between organizations working on CVE has been helpful? 4.29 4.57 4.59 Overall average score 4.27 4.07 4.05 16.4.3 Collaboration Between Actors 86% of respondents reported working with other organizations on issues of CVE, a significant increase from the midline. Also similar to the midline, respondents reported collaborating more closely with youth organizations, security sector actors, women’s groups and religious leaders. Respondents collaborated less with business owners and self-help groups. Other actors mentioned include students, a board of management for both primary and secondary schools, parents and journalists. 44 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 45 A higher score for this question means that the frequency of conflict is decreasing (improving), while a lower score means it is increasing (worsening). 173 What organizations do you work with? Baseline Midline Endline Youth organizations 48% 83% 68% Religious leaders 74% 69% 54% Security sector actors 62% 71% 51% County or national government actors - 66% 49% Community leaders 50% 63% 49% At-risk youth - 60% 49% Women’s groups 45% 66% 44% Teachers 36% 43% 37% Other community members 21% 51% 29% Self-help groups 6% 29% 20% Business owners 29% 43% 12% Other 29% 11% 10% Survey respondents collaborate with these actors for a variety of reasons. Most of the reasons given are based on the respondent’s ability to access the organizations, the impact VE has on certain groups they represent, or the level of authority or knowledge that an organization commands. Several respondents noted that it is common to work with youth and women’s groups for they are considered the most adversely affected by VE, and therefore working with them is vital to reducing the threat and impact of extremist activity. Similar to an observation made during the midline, respondents stressed that: “Working with these organizations strengthens communities and allows for more open dialogue, supporting the notion that more cohesive societies can better mitigate internal and external threats.” They are perceived to command respect, have a higher degree of influence, and a good understanding of the context of the communities they target. Most work with youth on awareness creation and employment programs. “Community leaders and religious leaders are more valued to bring security and social cohesion.” – Security Actor “I work with the national government on issues of local security and vetting of national ID applicants.” – Community Leader “County government actors help create a partnership with the community.” – Youth Representative A community-based organization representative noted that his choice on who to collaborate with is limited by resources and he therefore chooses to work with organizations that are easy to reach with minimal resources. A security actor who worked with the widest array of organizations notes that he prefers working with select organizations because they demonstrate a level of commitment to addressing issues of VE. He quotes that such organizations “are effective and most reliable on matters about violent extremism when youth is involved and need protection.” This indicates the diversity of organizations that are useful in supporting at-risk youth and why coordination across stakeholders remains important. Lastly, security actors and the national government are seen as default to work with, because they have the mandate on issues of security. Besides, the red tape in working with government stipulates 174 mechanisms of collaborating with other actors and this is a reference point for those engaging government within and without. As such, all but four individuals, a representative from CBOs, youth and the private sector, stated that they collaborate more with other actors now than they did 12 months ago, while one youth representative noted that the frequency has not changed. 16.4.4 Conflict Between Organizations Conflict and disagreements between CVE actors still remain a problem. A notable number, or 27%, acknowledge that they do occur with at least some regularity. Worth noting is that a higher number of respondents did not have an opinion or chose not to answer this question. At the endline, more respondents indicated that this happens very often (12%) than they did at the midline (5%). How often do disagreements occur between organizations working on CVE? Midline Endline Very often 5% 12% Somewhat often 21% 14% Rarely 19% 31% Never 33% 12% Do not know 21% 31% These disagreements occur for a variety of reasons. Perhaps predictably, most respondents agreed that limited financial resources and funding cause organizations to compete for what is available, especially when conflicts of interest are not addressed. As a community-based organization representative noted, disagreement has occurred over “nepotism and the complaints were raised severally but there has been no change.” There could also be a limited understanding of the requirements for managing a grant for some grassroots organizations which may not have the capacity. A second community-based organization representative stated: “It occurs when experienced CBOs are left out and grants awarded to less experienced CBOs. The challenge can be resolved by allocating funds to partners by engaging the constituency and key stakeholder in the grantmaking process.” Others believe that there are differences in the methodologies and approaches to dealing with the VE threat that cause tension between actors. A respondent from the educational institution category, who mentioned that conflicts happen very often, notes that actors disagree over issues of ideology and issues of the language used in CVE engagements; while another respondent adds that political interference does create conflict among actors. While some respondents note that few disagreements have gone unresolved, others believe that dialogue among actors has helped create better understanding, especially when resolving conflicts over duplication and resource competition. “Conflict over which organization implements programs and who gets the funds are common and resolved by having the County Commissioner facilitate the mediation process. Each CBO is given the chance to address their disagreement.” – A Media Representative The establishment of the CEF has helped address some of these issues. A CBO representative noted that the CEF is a central platform for coordination and is recognized by all CVE actors stating that “no organization with interest in CVE is allowed to work outside the CEF in Wajir.” 175 “Coordination of the various activities happens in the form of switching locations or geographical area when duplications of activity are reported in the forum. WPDA has sometimes been requested to cancel or modify an activity to switch to the needs at hand, such experience has been useful.” – A Key Informant from a Community-based Organization Beyond coordination, there are times when CEF members have demonstrated solidarity with those affected by VE acts in the county. A key informant stated: “As of today (13.04.2020), reports have been received on an attack where Al-Shabaab ambushed and killed six Macawisley fighters (locals combating Al-Shabaab) in Wajir County. The forum has been coordinating on how to support the victims on burial arrangements. Vehicle and fuel support are being sought.” However, a key informant notes that the forum is still at the “storming and norming stage” of team development, where conflicts that arise are being addressed and members are beginning to understand their roles within the forum. Issues of coordination of CVE activities are discussed during the forum meetings as organizations consult on the design, but when it comes to implementation “each actor focuses on their own campaign.” Overall, all but one respondent, a security actor who only chooses to work with colleagues and at￾risk youth, believes working together on issues of CVE is helpful. 16.5 Section V – Community-Driven Strategies and Initiatives This section addresses the community-driven CVE strategies that are underway in Wajir. A summary of the result area index score is presented, followed by a discussion on participation in NIWETU￾sponsored activities, different types of CVE initiatives taking place, the effectiveness of those activities, and the level of engagement of the community. 16.5.1 Summary of Findings Community-driven strategies in Wajir are increasingly common. The most effective strategies are thought to be a formal meeting that bring together key stakeholders, especially security actors and religious and community leaders. The community and religious leaders are respected, while security actors are viewed as authority figures and vital to tackling issues related to VE. Overwhelmingly, all actors engage with the community, although informal meetings with community members are viewed as less effective, though quite common. Perhaps this could be because such engagement is seen to be too organic to be viewed as an intentional CVE strategy. 16.5.2 Community-Driven Strategies and Initiatives Index Score The table below presents average scores from four key questions from the stakeholder survey on a five-point scale46 about the effectiveness of community-driven initiatives to counter violent extremism. Scores indicate a decrease in the rate at which organizations work with other stakeholders to develop CVE programs in the last 12 months, though collaboration is still strong. However, the extent to which such organizations engage communities on CVE activities has increased, even though the perception of the effectiveness of the programs seems lower than it was during the baseline and midline. Survey findings also indicate that CBOs are less active than they were during the midline which was much higher than the baseline period. 46 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 176 Questions Baseline Score Midline Score Endline Score Have local organizations worked with other stakeholders to develop programs on CVE in the last 12 months? 4.72 4.52 3.86 How effective do you think these programs are at teaching others about VE and CVE? 4.25 4.54 4.02 Are CBOs more or less active now than they were 12 months ago in implementing such programs? 3.72 4.29 3.73 To what extent do community members engage with these organizations? 4.75 4.32 4.55 Overall average score 4.36 4.42 4.04 16.5.3 Participation in NIWETU-Sponsored activities Participation in NIWETU sponsored activities has improved as compared to the midline. 55% of respondents surveyed confirmed having attended NIWETU-sponsored activities. They are noted to have participated in activities hosted by the following partners: RDI, ASDEF, WPDA – Wajir Peace and Development Agency. One respondent mentioned C4C, while a youth representative stated that he had attended a CVE consultative forum. Two respondents could not recall the hosting organization. 16.5.4 Community-Driven Strategies 74% of respondents surveyed acknowledged that local organizations have been working together to develop CVE strategies in the last year. Four respondents, representing the security actors, a media representative and educational institution category, disagreed. Nine respondents did not know if this was happening or chose not to give an opinion about it. Of the many activities taking place, the most common was noted to be formal meetings with community leaders or members. Other activities respondents highlighted, as commonly included, meetings with religious leaders, both local and national government authorities, and security actors. Roleplaying was the least used CVE strategy in the last 12 months. Other activities mentioned included online engagement particularly developing documentaries on issues of CVE targeting their youth. CVE Activities Occurring in Wajir Baseline Midline Endline Formal meetings with community leaders or members 65% 86% 88% Meetings with the security sector 55% 88% 71% Meetings with religious leaders 53% 75% 76% Meetings with government authorities (local or national) 43% 74% 74% Training sessions 60% 77% 63% Meetings with other community organizations 35% 79% 61% Informal conversations in the community 23% 61% 57% Discussions in schools 28% 56% 53% Sports activities 18% 68% 43% Mentorship programs in schools 5% 49% 39% Advocacy campaigns 47% 47% 29% Roleplaying 0% 28% 14% Do not know 10% 4% 0% Other 7% 0% 0% 16.5.5 Effectiveness of CVE Programs All but four respondents praised these initiatives as being at least somewhat effective. The four included three community leaders and a security officer. Formal meetings with the security sector, 177 community leaders and members, as well as religious leaders, were considered the most effective CVE initiatives in Wajir at the time of the endline. These activities were thought to be more effective, because they linked the community to local and national authorities and leaders who are respected by the community to coordinate better in addressing community challenges. This is seen to have also reduced the trust gaps between the community and security actors. Those involved such as religious leaders and security actors have the mandate to implement suggestions made in such meetings. CVE Activities Occurring in Wajir Baseline Midline Endline Meetings with the security sector 22% 59% 59% Formal meetings with community leaders or members 25% 59% 57% Meetings with religious leaders 32% 59% 57% Training sessions 20% 43% 51% Meetings with government authorities (local or national) 7% 45% 47% Meetings with other community organizations 10% 45% 43% Informal conversations in the community 7% 28% 37% Discussions in schools 7% 28% 31% Sports activities 15% 51% 29% Advocacy campaigns 8% 13% 20% Mentorship programs in schools 0% 32% 19% Roleplaying 0% 9% 6% Do not know 15% 4% 4% Other 3% 0% 0% A media representative, on the other hand, believed informal meetings were the most appropriate strategies for targeting at-risk or those affected by VE, perhaps due to fear of exposure to either VE sympathizers, or being victimized by security actors for having access to sensitive information. He states: “Face to face communication with target groups- this will help the target groups to get foist hand information compared to other forms of communication.” A security sector actor mentioned schools as important spaces for engaging students on issues of VE. It worth noting that CVE engagement in schools is guided by government policy and a set of governance structures such as Parents Teachers Associations and Boards of Management that oversee content and strategies to protect students. Otherwise, it will be more difficult for CVE actors to access school-going youth under the age of 18 years at the household level, even though they at risk of VE. “Schools are conducive grounds for recruitment where extremists take advantage of unsuspicious innocent students.” – Security Actor Others mentioned to be more effective are NGOs, human rights defenders, media and elders. 16.5.6 Level of Engagement As it was established during the midline, a large majority, 94%, of respondents noted that community members engage with CBOs working on peace and security issues at least a little bit, pointing to the fact that they are well-embedded and well-established entities in Wajir town. Two respondents said they did not know, which could be true but could also suggest that perhaps they see the facilitators 178 and the informal CVE conversations in the community as part of an ordinary community interaction process. To what extent do community members engage with these organizations? Baseline Midline Endline Engage a lot 84% 66% 67% Engage a little 5% 34% 29% Do not engage at all 3% 0% 0% Do not know 7% 0% 4% 179 16.6 Section VI – Objective 2: Improved Government Capacity The table below compares scores for sentinel indicators from the midline and endline evaluations related to Objective 2. Scores for the meta perceptions indicators are again on a ten-point scale and reflect the same sampling dynamics as the Objective 1 questions above. Scores for the statements regarding legislation and budgets were identified through a review of the CAPs and County Integrated Development Plans (CIDPs) which spell out the county budgets. Level of government and community participation in community-government dialogues Midline Endline 47When I express my views about violent extremism to members of the government, I feel like my opinion is considered and my concerns and ideas are responded to. Indiv. 4.48 5.93 Meta 6.14 5.98 The government has made a strong effort at participating in community dialogue over the past two years. Indiv. 7.36 6.67 Meta 8.21 7.06 I feel like I have been adequately included in the county government’s response to peace and security. Indiv. 6.67 6.20 Meta 7.01 6.99 Overall Score 6.65 6.47 Community perceptions of the national and county government on issues relating to peace and security Midline Endline The national government has made a great effort at improving the peace in my area over the last two years. Indiv. 7.00 5.33 Meta 7.46 6.79 The county government has made a great effort at improving the peace in my area over the last two years. Indiv. 6.04 6.24 Meta 7.00 6.56 The security sector has made a great effort at improving the peace in my area over the last two years. Indiv. 7.92 5.94 Meta 7.64 6.32 My community has become more willing to cooperate with the security sector over the past two years. Indiv. 7.67 6.21 Meta 8.26 6.99 Overall Score 7.37 6.30 Midline Endline Presence of line items for CVE activities in county budgets 0 0 Presence of legislation to support implementation of the CAPs 0 0 Community member views on government-community relations with regard to peace and security issues declined slightly from the midline, though are still largely positive. Perspectives on the government’s efforts at participating in community dialogues over the past two years and the degree to which respondents feel included in the county government’s response remained rather positive (even though they showed a marginal decline). Likewise, respondent views on the national and county governments’ commitment to addressing peace and security still remained positive, though declined slightly from the midline. Perspectives remained most positive on the community’s willingness to cooperate with the security sector; the meta perceptions score was higher than the individuals’ score, meaning that respondents were even more supportive of their peers’ willingness to engage at this level, suggesting that they might follow suit. Again, while still positive on average, views declined around the national government’s effort to improve peace over the past two years. This decline could be in relation to the increase in VE attacks in North-Eastern regions over the reporting period, or to the Kenyan government’s involvement in border attacks in Somalia that has increased insecurity in Mandera, just to the north of Wajir. It is 47 With a score of one representing “strongly disagree”, and ten representing “strongly agree.” 180 worth noting, however, that this decline is not as significant as it was in other North-Eastern counties that faced a similar spate of attacks around the same period. No progress has been made on allocating county budget funding for CVE activities, suggesting that local resources to support such initiatives will remain limited. While funding cycles tend to be largely fixed for five-year cycles, there is still room to negotiate adjustments in annual budgets, however this can often be more difficult. 16.7 Section VII – Government Responsiveness to CVE The following section presents findings on the responsiveness of county and national level government to address VE. Summary findings are presented, followed by a more in-depth analysis of the findings. The section will cover perceptions of both levels of governments’ understanding of and willingness to engage on VE and CVE issues. It will then look at engagement with various government counterparts, as well as the challenges faced in dealing with government actors. 16.7.1 Summary of Findings While views of the government’s engagement on CVE issues were slightly less positive than at the time of the midline assessment, respondents still demonstrated positive perceptions about both levels of governments’ willingness to understand the issues. Although both survey respondents and the key informants express frustration about working with the county government, their engagement with the county government has improved in the last 12 months. This coupled with the CEF provides an opportunity to strengthen coordination and support of the county government. CVE stakeholders are also keen on issues of transparency, especially from both levels of government. More openness and valuable consultation among all actors will strengthen the culture of trust and sense of commitment in addressing issues of CVE collectively though the forum. 16.7.2 Government Responsiveness Index Score The perceptions of the government’s understanding of VE issues in Wajir has improved throughout the reporting period. However, at the time of the endline, scores48 relating to the willingness of both county and national government to work on issues of CVE have slightly reduced. Despite the frustration over the county government’s engagement, scores on the ease in working with the national government remain steady throughout the period. Key informants mostly equated willingness of the government to work on CVE issues with its ability to fund CVE activities. Questions Baseline Score Midline Score Endline Score How well do you think the county government understands the problem of VE in this community? 3.13 2.96 3.14 How well do you think the national government understands the problem of VE in this community? 3.71 3.55 3.69 How willing do you think county government is to work on issues of VE? 3.28 3.13 2.90 How willing do you think the national government is to work on issues of VE? 3.69 3.63 3.37 How easy do you think it is to work with national government counterparts? 3.59 3.46 3.17 Overall average score 3.48 3.35 3.25 48 Please refer to Annex 1, “Methodology, Result Area Scoring Criteria”, for information on scoring approaches for questions on the four and five-point scales. One represents negative responses, and four or five represents positive responses. 181 16.7.3 Government Understanding of VE 78% of surveyed respondents were of the view that the county government understands the problem of VE in Wajir. Only 16% stated that the county government does not understand VE very well, while two respondents, a media and educational institution representative, believed that the county did not understand VE issues in Wajir at all. This is a slight improvement from the baseline and midline. How well does the county government understand VE? Baseline Midline Endline Very well 45% 42% 55% Somewhat well 25% 20% 22% Not very well 22% 24% 16% Not at all 5% 11% 4% Do not know 2% 4% 0% The improved perception of the county government’s understanding could be attributed to its engagement with the CAP process and, consequently, the CEF. This perhaps allows for frequent information sharing and updates on the context of VE between the county and other CVE actors. On the contrary, a key informant from a local CBO expresses his frustration that, although knowledgeable of violent extremism, the county government’s capacity on issues of counter violent extremism is not as good. He quotes: “The county government’s capacity in CVE is very poor. They are required to review their commitment to engaging in CVE with the necessary resources. Their human resource requires trainings and capacity buildings. The county government needs to understand CVE is as important as development. The two-level of governments have been reactive to CVE issues. For a sustainable solution to arrive, the governments have to be pro-active throughout.” Despite this reservation, respondents credit the CEF for helping strengthen relationships between the national and county government, while to a certain extent clarifying the roles both should play in on issues of violent extremism. A key informant from a CBO also acknowledged the complementarity of both government on issues of CVE in Wajir, quoting: “Though security is fully under the central government, community policing and seeking support from the locals to combat sympathizers that can assist the terrorist has been the core function the county government and the CSOs played well. “ Views toward the national government were very positive. All but three individuals believe that national officials at least somewhat understand the issues. Religious and community leaders, media and a community-based representative comprise the group of respondents who think national government understanding of VE issues is not very good. 16.7.4 Willingness to Engage in Programming Although perceptions of the county government willingness to work on issues of VE do not deviate very far from the midline, both respondents and key informants expressed frustrations over the counties level of commitment to funding CVE processes. How willing are county governments to work on CVE? Baseline Midline Endline Very willing 48% 45% 29% Somewhat willing 30% 30% 42% Not very willing 12% 18% 21% Not at all willing 5% 7% 6% Do not know 3% 0% 2% 182 One informant from the national government, however, notes that the county has indeed funded several activities including providing in-kind support. This perhaps could have been channeled through the County Commissioner, the CEF co-chair, and many stakeholders may not be privy to such details. Still, he noted that the county government is caught up in dealing with the intricacies of clan politics and is constantly faced with a dilemma on whether or not to prioritize CVE over politics. He noted that, faced with the choice of condemning VE acts that where perpetrators share the same clan or publicly support issues that are not popular with the communities, the county often chooses to take the side that does not risk them losing any political capital. The key informant from the county administration category says: “Due to this double-faced element, they are unable to act and deal with these groups the same way the national government does. In some instances, they would go lukewarm compromising on the war against terror especially where they feel it would have negative political outcomes for them, especially from the Wananchi for example on extra-judicial killings, police harassment and roadblocks. The county government is not very fond of these subjects since they are political entities. Where they feel that a certain size of the population will feel disenfranchised, they may end up siding with the public even though it's not sensible for security. Because of the stereotype that terror is a localized issue, there is some element of denial in the county government as they don't appreciate the magnitude of the problem even when it exists.” Both governments are criticized for not doing enough to demonstrate leadership by either convening meetings proactively or securing additional funding in a way that ensures that the recommendations made during the CEF meetings are implemented. A community-based organization representative who believes the county government doesn’t demonstrate commitment stated: “The county leadership delegates junior officers to represent them in CEF meetings.” A key informant commends efforts made by Mandera County on CVE and compares it to Wajir noting that: “Wajir County government isn’t as committed and coordinated when it comes to the CAP. It should take ownership and coordinate when it comes to CVE. There is also a lack of funds allocation by the county government.” – Youth Community Leader While discussing the county government, a CBO representative quotes that “They don't see investing in CVE as a priority compared to constructing roads or digging dams and boreholes. They are only reactive especially when a terror attack happens.” The national government, on the other hand, is viewed as not being transparent and accountable enough to share information on how it spends the CVE resources its’ entrusted with to fight VE in Wajir. A key informant representing a CBO acknowledges attempts made by the county government to jointly fundraise for CVE activities in Wajir. He noted that the county government once collaborated with a local CSO and raised about one million Kenya shillings to support emergency relief when VE attacks occur. How willing are national authorities to work on CVE? Baseline Midline Endline Very willing 78% 74% 53% Somewhat willing 13% 19% 35% Not very willing 3% 4% 8% Not at all willing 3% 4% 4% Do not know 0% 0% 0% Though with less confidence than at the midline, a majority of 88% of the respondents think that the national government is willing to work on issues of CVE. 76% further agree that they find it at least 183 somewhat easy to work with national government representatives, particularly the chief and the assistant chief at the village level. Surprisingly, three respondents noted that they have a relatively easy experience of working with security agents, including the ATPU. Wajir is the only county in North￾Eastern Kenya where respondents have pointed this out. However, mistrust between security agencies and communities exists, as was also highlighted during the baseline and midline, largely due to the and unwarranted surveillance of local residents and alleged extra-judicial killings, partly due to discoveries of mass graves of victims believed have been community members suspected of VE acts. 16.7.5 Engagement with Government Counterparts The trend in the engagement with government counterparts is similar to the midline, but with more frequency. Survey respondents stated that they engage with government counterparts with some degree of frequency, though engagement with national government officials seemed to occur with more regularity. This is perhaps because there are frequent interactions between national government representatives, especially chiefs and assistant chiefs, who live with the community and are easier to access than other government representatives. Frequency of engagement with county government counterparts Baseline Midline Endline On a daily basis 17% 21% 20% Once a week 0% 7% 6% A few times a week 5% 9% 20% Once a month 17% 4% 4% A few times a month 13% 4% 10% Occasionally 20% 23% 20% Rarely 30% 32% 16% Do not know 10% 2% 2% A majority of respondents mentioned that one of the biggest challenges to working with the county government is that it does not make an effort or create enough room for consultation with stakeholders and the wider community when developing or implementing CVE related activities. 67% of respondents complained that the county government either does not talk to them, involve them or value their contributions. Only four respondents did not express any concerns, but all those individuals mentioned that they do not have regular engagement with the county. Other reasons given include corruption and lies, politicizing CVE, and lack of resources to engage the county. Frequency of engagement with national government counterparts Baseline Midline Endline On a daily basis 28% 40% 27% Once a week 5% 5% 12% A few times a week 5% 9% 8% Once a month 17% 4% 6% A few times a month 3% 4% 10% Occasionally 20% 14% 25% Rarely 20% 23% 8% Do not know 2% 2% 4% In looking at the challenges experienced when working with the national government, nearly 43% of the respondents stated that there are no issues. Others, however, noted that the national government does not seek to involve or consult other stakeholders (as with the county government) and that they prioritize arrests and law enforcement. Some also noted that they are slow with information sharing and victimize and harass those seeking to report suspicious activity. An educational institution representative stated: 184 “They often ask irrelevant questions when civilians say they’re seen AS members pass by, etc. Because of this, people avoid engaging with the national government.” 185 17. ANNEX VII – CASE STUDY: CHAMPIONS FOR CHANGE NIWETU works through four different entry points in Kenya to achieve its goal of improved CVE capabilities to identify and respond to VE threats: government actors, influencers, communities themselves, and CSO partners. NIWETU’s Champions for Change (C4C) initiative is the primary NIWETU activity to support influencers at the local level. As one of NIWETU’s flagship programs, C4C is designed to improve the leadership skills, connections, and strategic actions of chosen CVE champions and to create a greater awareness of Champions’ (influencers) ability to enhance resilience to radicalization and work with at-risk individuals. This case study looks at NIWETU’s work with influencers, discusses key achievements within NIWETU’s result areas and how these gains can be sustained. 17.1 Overview of Champions for Change Through the C4C program, NIWETU identified existing community CVE influencers with access and credibility in their own local communities in each of NIWETU’s priority counties: Garissa (11), Isiolo (10), Mandera (12), Nairobi (10), and Wajir (10). Since November 2018, NIWETU has supported these 53 CVE local influencers through this innovative activity, with the aim of sustainably expanding their CVE activities within their communities. Involvement in the activity has been purely voluntary; Champions are not compensated for their participation. NIWETU supported the Champions in developing CVE action plans for each of the ten hotspot areas. The action plans seek to sensitize at-risk youth in learning institutions; design and disseminate CVE media products; improve community and security relations; and to include women in CVE dialogues and activities. As part of the year-long C4C program, NIWETU has organized cross-county networking opportunities for the Champions. The first networking activity allowed the Champions to share with one another their action plans and strategies for engaging different stakeholders within their local communities. The second networking activity seeks to encourage cross-county learning, prompt the sharing of best practices, and foster discussions on opportunities for continued CVE activities after the end of the formal C4C program. Influencers Communities CSOs Government 186 17.2 Who are the Champions? Champions were selected based on their access and influence within their hotspot communities. Selection of the Champions was a key step in ensuring the broader success of the activity. If the wrong individuals were selected – those without spheres of influence or access to key state and non-state stakeholders – the influence that the champions could leverage to carry out their action plans would be rather limited. A robust selection process was therefore used to select Champions who met certain criteria:  Informed about C/VE in their communities, either through existing work in this space or through related work in peace, security, or conflict resolution matters.  Respected individuals within the community who are known to and can engage with county government officials. The Champions’ pre-existing influence with community groups was key to strengthening CVE efforts meant that they would be respected and well-received when trying to raise awareness about VE and develop CVE strategies. In Garbatulla, one Champion was selected because of his experience mediating conflict between his community, the government, and neighboring communities as a member of the local peace committee. In this capacity, he also engages with government counterparts (at both the county and national levels) and is therefore an effective interface between his community and government. In Dadaab, one Champion was selected because of his experience in peacebuilding and conflict management; he sits on the Dadaab sub-county peace committee. A Kibra-based Champion began engaging on CVE matters through his work with BAWAKI, a local CSO, that brought together Christian and Muslim religious leaders. After engaging in this activity, he began to more proactively work with others on CVE programs and information sharing within his community to raise awareness about the risks of VE and the impact on youth in particular. These examples show that the Champions have the right levels of access and influence to be effective. Participation in the C4C program is purely voluntary; while some funds are dispersed to help in the initial stages of implementing their action plans, much of the work they are being asked to do is unfunded and will require them to either volunteer their time or source funds from elsewhere, encouraging the Champions to become self-reliant and think from the start about how the activity and their contributions can become sustainable. The selection of both male and female Champions has ensured that key community groups are not left out. Recognizing that men and women have different experiences and perspectives to bring into CVE conversations, selecting both male and female Champions would allow the program to reach across a wider network of community groups and ensure those voices are heard. 17.3 What are the main achievements of the C4C initiative? The C4C activity has strengthened outcomes in three of NIWETU’s result areas: (1) Improved CVE knowledge and skills; (2) CVE networks established; and (3) Community-led CVE strategies and initiatives. As a result, there has been a clear contribution to change at NIWETU’s goal level (improved CVE capabilities to identify and respond to VE threats). Champions’ knowledge around C/VE matters has increased substantially as a result of this activity. They Improved CVE knowledge & skills 187 pointed to a greater understanding of the drivers of extremism, what makes people in their communities at-risk, and the narratives that are used to recruit and radicalize. A Champion in Mandera noted that he has also learned about the important role residents play in ensuring the security of their communities, which encourages more active engagement in CVE activities. One of the Champions from Garbatulla sub-county in Isiolo noted that training on CVE knowledge and skills helped him to better understand how extremism has affected all Kenyans and reached into areas that he previously thought were not affected by VE; previously, he believed that VE only affected communities living in Northeastern Kenya, but he learned that some youth from Garbatulla had been recruited, and that VEOs often rely on local communities to facilitate movement of people and good. Through these lessons, he began to understand the importance of engaging with all community groups in the fight against VE and reaching into areas that might not be considered hotspots. Regular county-level and inter-county networking meetings have helped the Champions to strengthen their networks with other CVE actors locally and across Kenya. One of the key elements of the C4C activity design was to create a network of CVE leaders across NIWETU’s counties of focus. Through monthly Champions’ meetings at the county level, as well as two networking events for all Champions in Nairobi, networking and relationship building within and across counties has been a key element of the activity since its inception and has resulted in a strong network of CVE influencers across Kenya. The newly established networks are creating the space for Champions to learn from each other, share ideas, and expand their networks in support of CVE programming. These networks have been supporting the development and implementation of community-driven CVE strategies and create a space for sharing of different contextual learning and emerging good practice across counties. The networking events also enable champions to come together with politicians or other government officials which has helped to solidify their relationships with government and political counterparts. In one such case, the Chief Administrative Secretary in the Ministry of the Environment, who is from the Northeastern region, met with Champions from Wajir County, where they agreed to work together to engage with women in Wajir County on CVE related issues. Pooling resources, they organized a meeting where the Champions also met the Governor’s wife to discuss women’s role in CVE. The Champions have institutionalized their network through the formation of local CBOs which enables them to compete for funding and develop a track record of implementation. In doing so, they also gain legitimacy as CVE actors within their communities and have a clear platform for engaging with the CEFs and government counterparts. A Champion from Kibra noted that his group of Champions have already begun building relationships with existing organizations that work in the broader peace and conflict resolution space, including with groups like BAWAKI, which bring together religious leaders from different faiths in Kibra. He highlighted the importance of leveraging the networks, relationships and actors already working in this space because of the importance of building trust: “It is easy to work with existing networks, given how crucial it is to understand communities and get the community to support us.” Leveraging the champions’ existing networks to raise awareness about C/VE enabled a wider and deeper reach into their communities. With diverse networks and backgrounds, the Champions are seen as “credible voices” and respected individuals across different community groups. Through this activity, they have been able to solidify their influence within these groups, as well as forming new networks as a collective of influencers. Additionally, the Champions in some counties CVE networks established 188 have already begun engaging in the CEFs. A Champion in Wajir noted that the way the C4C activity is set up encourages them both to work together as a group of Champions, but also engage with other CVE actors within the community in order to keep them informed of their activities and plans. Participating in such initiatives helps to solidify the important role that the Champions play in CVE efforts at the county level. Despite early success in networking across different community groups, challenges in engaging county government remain. In Dadaab sub-county, one Champion noted that county officials often fail to show up to meetings or activities to which they have been invited and have not responded to requests for logistical support to continue carrying out CVE activities. In Garbatulla, a Champion also pointed to difficulties in coordinating with county officials, believing that they have chosen to let CVE issues rest with national government counterparts. These frustrations are not unique to the work of the champions at the county level; as is seen through other case studies and county reports produced for this evaluation, county government generally remains less committed to CVE than their national government counterparts. Some county governments, such as the Mandera county government, are committed to supporting CVE activities, but resource constraints remain a problem. The action plans are the primary output of the C4C activity, with one action plan developed for each sub-county, for a total of ten action plans that aim to disseminate important CVE messages and skills to identified target groups within each sub-county. The action plans were revamped based on research into the Champions’ networks, to understand their levels of access and influence within their network, and in response to regulations imposed by GOK in response to the spread of Covid-19. The action plans are designed to address an identified issue or challenge in CVE programming in the communities in which they live and work, through strengthening engagement between residents and security agencies, sharing information within local schools, or reaching out to groups often excluded from CVE programs, such as women’s groups. The plans are therefore highly contextualized and rely on the local knowledge and experience of the Champions to be successful. So far, positive strides have been made in implementing the action plans. A few examples are presented below:  Students in Garissa township valued their interactions with the Champions and requested them to try to reach out-of-school youth, who are known to be more at-risk. Recognizing that skills and knowledge of youth on C/VE was rather limited, the Champions chose to target students and the wider school infrastructure, including teachers, Parent Teacher Association members, Boards of Management members, and Ministry of Education officials, in their action plans. Students in particular really valued what they learned through this engagement and understood the need for these messages and lessons to be shared more widely, across different at-risk groups.  Champions have become outspoken advocates around issues affecting conflict and security in their community in Mandera East sub-county. Leveraging their existing influence, Champions in Mandera East have taken to the radio and convened community groups in barazas, other gatherings, and in mosques to discuss these issues. They have focused on the perils of substance abuse, importance of supporting youth education, and disseminating information about Covid-19. One Champion noted, “we are sort of information conveyors to the community” on these issues, which they believe all have implications on VE dynamics in the county. They have been in contact with heads of security agencies in Mandera, building a good Community-led CVE strategies 189 working relationship to collaborate more closely with the security sector on how to combat extremism with softer, more community-centric approaches.  With a new understanding of different gender dynamics and experiences of different community groups, Kamukunji-based Champions are identifying deliberate strategies to effectively reach both men and women. In their activities to target young mothers and other young women, the Champions learned that the most effective approach for reaching women is to work with them in separate gender groups, so they feel more comfortable speaking openly. They also learned that through using other innovative approaches, they can more cost effectively disseminate messages to their specific target groups, such as using different forms of art to reach young people. Because of the ongoing Covid-19 pandemic, however, the Champions noted that many of their activities and engagements have been placed on hold. While they fear that momentum will be lost if restrictions on movement and gatherings last too long, they are still committed to working on their action plans. Champions have utilized the skills gained through the C4C activity in other areas outside of their NIWETU-related engagements. Champions are identifying other ways to independently share important messages and information with other groups. One Champion from Kibra, for example, said that he has begun to use his platform as a religious leader to disseminate important messages through weekly sermons and discuss the issues in more intimate settings with the members of his mosque. He underscored the importance of countering false religious narratives with correct interpretations of religion to push back against VEOs: “You know that one of the major problems we face among Muslims on matters of CVE is that there is a group of people who use religion to justify terrorism. This has been a challenge for us who study and preach the religion, because of the misinterpretations of the Quran. My role has been to repeatedly give the proper translations and debunk the existing ideas about VE. I do this through research and use the facts from religious scholars, as well as what I have learned in my CVE work.” Using their positions of influence, some Champions are also supporting at-risk individuals or returnees in their communities. A Kibra Champion shared that he was mentoring a youth who was being lured into Al-Shabaab. Working with the young man’s mother, they helped him apply for a passport and find a job in Qatar, where he is now working and sending money back to his family. A Kamukunji-based Champion also often regularly mentors and engages with at-risk and returnees, offering emotional support, helping them find work in the community so they can provide for themselves and their families. Outside of the CVE space, the Champions are also working to combat disinformation around Covid-19 and government regulations. Working across NIWETU’s result areas has strengthened the Champions’ knowledge and understanding of C/VE, created platforms for engagement and information sharing, and supported Champions in designing action plans that would target key areas in need of CVE support in their communities. Through the activity, there has been a clear contribution to change at the goal level, whereby the Champions have begun to demonstrate clearly that their capabilities to respond to VE have improved. Through the examples cited in the previous section, the Champions have begun designing and implementing CVE activities and programs and are leveraging their existing relationships and networks to disseminate important CVE messages and information to the wider community, including to groups often not engaged in CVE activities. Improved capabilities to identify and respond to VE threats 190 At the same time, and perhaps an unintended outcome of the C4C activity, the Champions are also leveraging their networks and leadership skills, in particular, to work on other important issues affecting the community. Examples include combating disinformation around Covid-19 and taken a more outspoken and proactive stance on peace and security issues. 17.4 Lessons Learned A number of lessons on how to design holistic and integrated activities to support Champions have emerged through this process. 1. The selection of individuals who already have influence and access ensures the activities are more effective. With pre-existing relationships that they can leverage, Champions who already have networks with groups or individuals an activity is trying to reach are likely to be more effective, because of the influence and respect those groups already have for the Champions. 2. Developing clear strategies for engagement through the formation of action plans solidifies the champions’ work and contributes to sustainability. The action plans are intended to extend past NIWETU’s close-down, through a clear strategy of engagement with local communities and government officials and requiring limited financial resources so that a lack of funding does not become a deterrent for carrying forward their action plans. The action plan approach also helps to strengthen the Champions’ networks and influence on CVE matters, making them key stakeholders in CVE discussions at the county and sub-county levels. 3. Establishing CBOs for the Champions further solidifies their role as CVE stakeholders and enables them to take a more active and legitimate role in CVE discussions. As a registered CBO, they are able to compete for funding and participate in the CEFs, positioning them well within the growing CVE space in each county. It also enables collaboration, and the Champions are able to share what they have learned through their intensive yearlong NIWETU engagement with other CVE stakeholders or actors interested in participating in CVE work. 4. Designing the C4C activity to sit clearly within NIWETU’s conceptual framework ensured that early successes would continue to build off each other, strengthening the capability of Champions – and ultimately the wider community – to understand and respond to the VE threat. As presented above, the C4C activity worked across all of NIWETU’s result areas in order to contribute to NIWETU’s goal of improving capabilities to identify and respond to the VE threat. The activity was designed to work across result areas in tandem, so that early successes could help strengthen work across other areas. For example, networking events helped to expand and deepen champions’ networks, while also providing a platform to learn from each other about different contextual considerations and adaptations. 191 192 18. ANNEX VIII – CASE STUDY: COUNTY ACTION PLANS One of NIWETU’s key achievements over the last four years has been supporting the development of County Action Plans (CAPs) for Countering Violent Extremism (CVE) across four of NIWETU’s target counties, and leading on the development of Rapid CAPs (RCAPs) in Nairobi – NIWETU’s other county of focus – and across the other remaining 36 counties that did not have an existing CAP. Working across the different entry points into the system, NIWETU’s engagement in the CAP formation and subsequent implementation can be seen as a microcosm of NIWETU’s systems-based approach. The CAP formation process and efforts at implementation that have occurred since the launch of the CAPs, have helped strengthen vertical and horizontal linkages between and within county and national level government actors, and between government and civil society actors at the county level. Working at all levels within the county has facilitated the expansion of CVE networks, while creating new opportunities for collaboration and coordination amongst a set of actors who had previously often had contentious relationships. While in many areas the CAPs and RCAPs continue to be under￾resourced and political will remains limited, the process which the counties underwent to develop the CAPs, and the engagements that have taken place after, have solidified networks and strengthened relationships on CVE issues within the county. This case study will identify lessons learned through NIWETU’s support to the CAPs through its systems-based approach, and the effects of resulting linkages and relationships that have been forged. 18.1 Engaging Government 1. Evidence-based advocacy, presented through the right messengers, can lead to meaningful institutional adaptations. Government involvement in the CAP formation process and activities following the launch of the CAPs, happened across a number of different levels in both county and national government. At the helm of the formation process was NCTC, which pushed forward momentum to develop both the CAPs and RCAPs across Kenya. NCTC has led three different iterations of CAP formation processes. The first phase focused on counties along the coast, where CAP development was more haphazard Influencers Communities CSOs Government 193 and there was no clear structure or approach. Acting on lessons learned from the first phase of CAP development, NCTC developed guidelines for developing the CAPs that led to a more streamlined and consultative process, while remaining efficient and timebound. This second phase of CAP development focused on counties in Northeastern Kenya, as well as Marsabit and Tana River. Following the presidential directive in January 2019, which called for every county in Kenya to develop a CAP, NCTC again revised its guidelines and approach to push for an even more efficient process that would allow them to lead on the development of CAPs across the county in a shorter period of time; this was known as the RCAP process. Importantly, the revised approach for the RCAPs also took into consideration findings from previous research, commissioned by NIWETU to identify lessons learned and good practices around CAP formation. While timeline pressures imposed by the President of Kenya limited the degree to which some of those recommendations could be adopted, others were acted upon, including the need to develop shorter and more concise CAP documents, and to work within a shorter timeframe so that progress in implementation could be more easily seen and measured. 2. Relationship building takes time and should focus on officials across the government and political spectrum, including politicians and civil servants at county and national levels of government, to engender sustainable engagement. NIWETU worked closely with NCTC in the formation process in Garissa, Isiolo, Mandera and Wajir counties, while supporting local partners to lead the process within each county. NCTC helped to review drafts and participated in occasional meetings and dialogues organized as part of the formation process. As a result, general awareness about NCTC and its role in the CVE space in Kenya increased, although mostly within the small groups of stakeholders involved, and not more widely into the community. At the same time, NCTC and NIWETU became close collaborators, and through sharing findings and lessons learned from research, NCTC again adapted its guidelines for CAP development to strengthen the process further. Improved relations between NCTC and NIWETU helped to ensure that the formation processes continued efficiently and involved the right sets of stakeholders, or other interested parties at the national government level. 1.1.1 NIWETU-NCTC collaboration in support of the Rapid-CAP development In the months following the January 15, 2019, attack on the 14 Riverside Drive office and hotel complex in Nairobi, the threat of violent extremism across Kenya became more real for citizens and government officials. Recognizing that Kenyans across geographies and socio-economic classes can be vulnerable to recruitment and radicalization, President Uhuru Kenyatta declared on January 22, 2019, that every county in Kenya must develop a CAP by the end of June, in order to support a national effort to combat the threat of VE as it spreads across the country. To accomplish this ambitious goal, NIWETU quickly mobilized resources and logistical support across the country. With close coordination with NCTC, providing technical and governmental oversight, Rapid CAPs were developed in 37 counties in Kenya across a three-month period. Since then, implementation of the RCAPs has largely stalled, in many, but not all, RCAP counties. Without the political will and buy-in, particularly in counties where VE has never been seen as a problem, momentum for implementation simply does not exist. Despite this lack of progress, the RCAP development process was another clear success in relationship building. The coordination required between NCTC and NIWETU helped to solidify relationships between both actors, while also strengthening USAID’s relationship with NCTC because of the support offered to the process. At the same time, NCTC also formed relationships with county governments in areas that have traditionally been seen as low-priority counties for CVE programming, which is an initial step in being able to catalyze momentum for RCAP implementation. NCTC is currently prioritizing revising the current CAPs to match the RCAPs so that implementation is on the same implementation schedule. 194 In addition to close coordination with NCTC, NIWETU worked with county and national governments at the county level throughout the development process. Later, as the CEFs were formed, meetings began more frequently. Engagement with County Commissioners across all counties was instrumental to keeping momentum going in the formation process, as well as in the CEF meetings. In Isiolo, for example, the CC was a clear champion of the CAP. His office and the NIWETU county team had a close working relationship, with regular consultation that carried over into the CAP formation process. After the launch, he was an active participant in the CEF and on CVE issues in general; CSOs in Isiolo recalled strong collaboration with members of his office. More recently in Garissa, the replacement of a CC who had also championed the CAP, with a CC who had no prior experience or real interest in engaging with the issues, meant that NIWETU and other CSOs would need to work hard to garner his buy-in and support. Eventually he began to take a more active role in the CEF, likely as a result of sustained engagement with NIWETU staff, who understood the importance of gaining his buy-in; his buy-in was crucial for the CEF to become more effective with regular meetings and consultations. 3. National and county-level official engagement is instrumental maintaining momentum in the CAP formation process, as well as in the CEF meetings. At the county government level, NIWETU engaged with county Governors who were key to the formation process and who co-chair the CEF with the CC. Engagement with CCs has often been difficult, except during the CAP formation process in Mandera, where the CC had a keen interest in CVE and was eager to be involved. Owing to the personalities of those individuals, there was a clear commitment on both sides to engage productively and effectively with communities to steer CAP development, and also guide its implementation. Since the new CC arrived, however, national government engagement in CVE in Mandera has declined. The Governor’s continued interest in CVE and championing the CEF has meant that the CEF in Mandera is viewed to be particularly effective as a platform for groups to discuss the security threat, share lessons and strategies, and collaborate effectively. County level leadership, and to a more limited extent, national government leadership in Mandera, has therefore shown commitment and capacity in tackling violent extremism in the county through local strategies and mechanisms – key indicators of success within USAID’s Journey to Self￾Reliance. In Wajir, relationships between county government and some CSOs on CVE matters are also somewhat strong, and the CEF borne out of the CAP process has created a platform for further engagement. With some engagement from county officials, CSOs can pick up the slack and push forward. Likewise, in Garissa, county government and national government were reported to both be committed to the process because Garissa has often been a target of VE attacks. However, it is possible that interest in collaboration across actors has slowed. One county government official engaged in CVE issues noted that collaboration has dissipated, a sentiment which is also noticed in the county-level quantitative findings. 4. In places where political will is low, civil society can spearhead the coordination process, but become more reliant on donor/partner funding. Supporting civil society to lobby Members of County Assemblies (MCAs) for development of appropriate funding channels could be one way to mitigate this risk. Mandera presents a unique example, in which both county and national government leaders have taken a keen interest in supporting CVE efforts in the county. Allocating budget funding to support community-driven activities has also ensured that Mandera’s CVE response is tailored to the specific needs and experiences of the communities across the county. 195 18.2 Engaging CSOs During CAP development, NIWETU worked with a local partner in each county, which led stakeholder engagement meetings and served as the focal point for collaboration between various stakeholders and the Malaika Foundation, which led on the drafting of the CAPs. 1. Working through local partners was key to legitimizing the CAP process. While some CSOs welcomed Malaika Foundation’s contributions and technical leadership, others thought that the process should have been more locally led and driven. Lead CSOs from each county therefore helped localize the process and ensure that it was locally owned. These organizations have the capacity to convene the right set of stakeholders as, being based in the counties, they are better placed to understand local dynamics and competition between actors, as well as know which organizations and individuals have been positively engaged on CVE or similar topics, and could help drive momentum for CAP implementation. For the RCAP process, the government selected the stakeholder groups, while Malaika’s role helped ensure consistency in process and outcomes. 2. State-civil society collaboration during CAP development strengthened relationships necessary for implementation. Since the launch of the CAPs, NIWETU has worked alongside CSOs and government to constitute the County Engagement Forums (CEFs). As NIWETU county teams also have local contextual understanding, they were well placed to support government and CSO counterparts in formulating the CEFs and driving forward their initial meetings. NIWETU continues to play an active role on the CEFs in all counties, sitting alongside other CSO and government counterparts. However, as NIWETU is not a leader within the CEF, the sustainability of the platform is not tied to NIWETU’s continuation. Working closely alongside government and other CSO counterparts, NIWETU has been able to strengthen and support the capacity and commitment of local actors, such that existing momentum and interest in CEF engagement can be expected to continue after NIWETU exits, except for in areas where VE is not considered to be a real problem (such as in the RCAP counties). As a result, the CEFs in some counties are believed to be highly effective. According to one CSO representative in Mandera, the CEF plays a critical role in helping actors jointly decide on strategies, so that they “can move as a team to counter violent extremism.” Meetings occur on a monthly basis 1.1.2 CEFs Coordinate Response to Increase in AS Attacks The CEFs in Northeastern Kenya have become important convening bodies around peace and security issues, specifically related to VE. As VE incidents in the region became more frequent in the early part of 2020, the CEFs began to mobilize stakeholders to respond to the growing crisis. In Wajir, the CEF came together to mobilize resources and form a 12-member committee for the districts of Wajir East and Tarbaj, which were the most adversely affected by the increase in violence. The committees were chaired by the Deputy CCs (DCCs), and again brought together a diverse set of state and non-state stakeholders including area chiefs, religious leaders, and other influential community members and leaders from the affected areas. The CEF in Garissa also began having emergency meetings to discuss increasing insecurity, but also focus on the impact on educational services; many schools had closed as a result of the violence. And in Mandera, intelligence and other sensitive information coming from community members has been shared with security agencies via the CEF. The group has also mobilized resources and emergency aid to affected communities. Leveraging the growing relationships between stakeholders borne out of the CAP processes, these platforms have been effective in alleviating tension between residents and security agents and identifying a collective way forward between these groups to respond to the threat. Attacks have since become less frequent in Northeastern Kenya, but the relationships strengthened as a result of close consultations remain. 196 and are effective at ensuring coordination across actors and sectors. It also helps organizations develop more technically sound activities, by providing advice and support to those working in the sector and in the broader peace-building space. 3. Sustainability of the CEFs requires diverse ownership and buy-in, as they serve as catalyzing forces in a systems-based approach to CVE. In Isiolo, IPL’s keen support for the CAP process and engagement in the formation of the CEF, coupled with lack of leadership from the government, led to confusion over who ran the CEF. It was initially seen as a CSO-led initiative, in part because of the Governor’s lack of engagement, but with active leadership from the CC, there is clearer leadership within the CEF, while still maintaining very active participation from CSO members, who also feel as if they own the CEF and are therefore keen to see it succeed. “Initially there was a lot of confusion. The CEF was synonymous with IPL, because it made most of the communications to members. Now, with the expanded CEF, the communication will be from the office of the CC.” Government and CSO stakeholders play an active role within the forum, which helps to improve coordination and build effective working relationships between both sides. Respondents in Mandera also note that the CEF has been highly effective, “the center of coordination mechanisms for CVE activities,” with members utilizing WhatsApp to ensure fast and ongoing communication. However, the CEF discussions and members are all concentrated within Mandera town, and according to one respondent, there is no representation outside in more rural areas. NIWETU participates in the CEFs alongside other CSOs and implementers, leading activities aligned with certain pillars, such as an activity in Garissa that sought to enhance CVE skills and knowledge in learning institutions, which aligns with the education pillar of Garissa’s CAP. By taking the lead on the initial stages of CAP implementation, NIWETU has helped catalyze momentum, showing some early successes in all counties. As NIWETU does not maintain an active leadership role within the CEFs, their sustainability is not tied to NIWETU’s ongoing participation and engagement. 18.3 Engaging Influencers Over the past year, NIWETU has worked closely with identified Champions in each county to strengthen their leadership skills and understanding of VE and CVE strategies through the Champions for Change (C4C) activity. Champions are individuals who are known in their communities to be involved in peace and security issues, human rights advocacy, or in other positions of influence within their communities. 1. Civil society brings the mobilization power, while influential individuals bring in￾depth knowledge about VE at the county level necessary to inform significant change. Many of these individuals were consulted and engaged during the CAP formation process stakeholder engagement meetings, given their prominent roles within the community, they presented perspectives and opinions on behalf of certain community groups. Influencers at this level have a strong grasp of the concerns and issues affecting local communities, and different constituencies within communities, that are necessary to understand in relation to CVE and VE. With experience working in peace committees, as human rights defenders, influential women and youth leaders, or as advocates for other peace and security issues at the local, county or national level, these individuals were able to provide more nuanced understanding about how VE affects communities. Using their positions of influence, 197 they were able to positively contribute in the stakeholder engagements meetings in a way that ensured that a deeper understanding of the issues was factored into the CAP formation process, and subsequent CAP document. 2. Influencers can proactively engage with formal state and non-state structures in ways that complement, rather than duplicate, existing platforms. In addition to the Champions’ involvement in the CAP formation process, platforms have been created for them to continue engaging with the CEF, and on CVE activities aligned with the CAP more broadly. Firstly, the Champions have created action plans that will span one year and are designed to identify and address a particular issue or challenge in their communities, related to C/VE, through consultations, engagement and programming over the year. Engagement with the CEFs is therefore required, and the action plans themselves all clearly align with a particular pillar. For example, Champions from Mandera South and Garissa Township chose to work with students to increase their resiliency to VE narratives in the school setting, in clear alignment with the CAP’s education pillar. Because CEFs are meant to serve as platforms to promote coordination between actors in the space as well as to provide technical support and capacity building, NIWETU’s activities with the Champions encourage them to engage with the CEF. In each county, the Champions form a registered association to enable them to compete for funding for CVE or other peace and security activities, and to ensure the sustainability of the activity. The association also provides a clear platform to engage with the CEF; in Isiolo County, the association the Champions have registered, has become a member of and an active participant on the CEF. 18.4 Engaging Communities 1. Community consultation in CAP formation was key to ensuring diversity of perspectives and geographies. Community consultation was a key facet of the CAP development process. Through various stakeholder engagement platforms that took place over a nine-month period, meetings were organized with different community groups, to extract their perspectives and experiences to inform the CAPs. These consultations were held in both the main towns in each county, as well as in the identified hotspot areas. Consultation across all parts of the counties – some vast with quite remote and difficult￾to-access villages – was not possible in order to keep the process streamlined and efficient. It is likewise important to recognize that inclusivity does not require consensus, but instead should be seen as reaching into diverse areas and gathering diverse perspectives. 2. Feedback loops back to the community are key to ensuring a ‘whole-of-society’ approach to CVE. Perhaps as a result, knowledge about the CAPs and their purpose is not widespread outside of select community groups in hotspot areas. Dissemination of the CAP and its priorities has been limited in most counties. Steps have been taken in some locations, however, to boost community inclusion and participation, particularly at the time of the launch of the CAP. In Wajir County, the CAP launch was hailed as a success; with nearly 350 participants packed into a cramped social hall, with an informal, baraza-style feel, community members and dignitaries organically intermingled. Attendance included youth from local schools, senior government and military officials, women, human rights advocates, and CSO representatives, amongst other stakeholder groups. WAPDA, NIWETU’s local partner in Wajir which led the formation process, conducted a thorough CSO and stakeholder mapping across 198 the county in order to identify and invite attendants for the session. For their part, community members expressed keen interest in participating in the launch. 18.5 Creating Linkages Between Actors The whole of society approach that NIWETU works through is clearly evidenced through its engagement across the different stakeholder communities, in relation to CAP formation and implementation. Activities and engagements have been specifically designed for each stakeholder grouping (government, CSOs, influencers, and communities), which has led to relatively widespread engagement in the formation process and activities in support of implementation. While the pace of CAP implementation has been rather slow in most counties, largely due to resourcing constraints, the process has had a clear impact on the quality of relationships between different stakeholders: 1. Government-to-government: Relationships between county and national government stakeholders have at times been fraught. Resource allocation and political tensions have often been the contentious issues in these relationships. At the same time, security remains the mandate of the national government and is not devolved to the counties; as a result, there remains some confusion over which level of government is mandated to lead on CVE issues. However, the CAP formation process, and now the CEFs, have been co-led by representatives from both county and national government, ensuring that both levels have an equal role to play. Relationships between county and national officials have consequently improved, through collaboration on the development of the CAPs and their equal roles in the CEFs. While county officials’ ongoing participation is still lacking in some counties, communication and coordination with national government counterparts has nonetheless improved. 2. CSO-to-government: There have long been frustrations in the relationships between CSOs and government, around issues of resourcing activities and engagement in the design and implementation of activities. The CAP formation process and the CEFs necessitate collaboration between stakeholder groups and, throughout these processes, CSOs and both county and national government counterparts have become effective partners on CVE issues in the counties. While resources remain limited, and the ability of either actor to implement programming is still constrained, the coordination and relationships built have resulted in forging a unified, county-focused voice in addressing issues of VE. 3. Influencer-to-government: The influencers NIWETU has worked with through the Champions for Change initiative are key stakeholders within their communities. Previous research has shown that they have influence over specific community groups and tend to be well known and have their perspectives valued within their networks. One of NIWETU’s key achievements through the CAP processes, has been strengthening the relationships between these individuals and county and national government officials. Improved relationships at this level helps government officials better understand the needs of local communities, while allowing the Champions to serve as conduits for information sharing between communities and government, and vice versa. 18.6 Lessons Learned The processes surrounding the development and implementation of the CAPs have resulted in a number of lessons learned around how to promote involvement and buy-in for a locally owned collective response against extremism. 199 1. The quality of relationships that have emerged through the process is the clear success from this process; CAP implementation itself is not a marker of success. Relationships within and across stakeholder groups have been strengthened as a result of the consultative CAP formation process and through engagement with the CEFs. Many of these relationships and interactions had been previously fraught with tension, because of competition over limited resources, or longstanding historic tensions between groups. While implementation of the CAPs may have stalled in some places, the resulting relationships that have formed have still remained strong and have been instrumental in allowing for better information sharing and awareness raising. 2. Working through local partners based in the counties builds local legitimacy. Local partners have the trust of the communities in which they work and are more effective in gathering experiences and perspectives; they can solicit more active participation from community members and other stakeholders. Criticism is often lodged against organizations that come from Nairobi and are said not to understand or appreciate the local experience. NIWETU worked through local partners effectively, in order to gain the buy-in of key community groups, which helped to increase support for and engagement in the formation process and other activities since the CAPs’ launch. 3. Leveraging the positions of influence of different stakeholder groups is an effective strategy to encourage active involvement in an activity. NIWETU worked closely with each stakeholder group throughout the formation process of the CAPs and since their launch. Understanding each different group’s role within the system, and the influence they carry with them, helped find effective entry points to bring them together in support of the CAPs. For example, working with the Champions on developing their action plans that align with CAP pillars, while helping them form associations so they can engage as a group with the CEFs, has helped encourage influencers to engage with the CAP related structures. Working closely with CCs’ offices, and helping them formulate the CEFs which they co-chair, has helped to increase positive interaction between the head security actor in the counties and other community-level stakeholders. This positive momentum therefore encourages others to engage in these platforms. 4. Co-ownership of the CAP between different stakeholders can promote buy-in and strengthen cooperation. Confusion over which level of government is mandated to address CVE within communities, has allowed different government officials to shirk responsibility for engaging in CVE activities. The CAP process, however, required both county and national government representatives to engage and lead together, which helps to initially force buy-in, but then generate it organically. While county governments in some areas minimally participate with the CEF, in others county and national government representatives have bought into the idea of the CAP and pushed forward momentum on implementation, as well as even funding some CVE activities from county budget lines. By bringing both levels of government onboard equally, more cooperation naturally happens across government actors. 200 19. ANNEX IX – CASE STUDY: KENYA SCHOOL OF GOVERNMENT NIWETU works through four different entry points in Kenya in order to achieve its goal of improved CVE capabilities to identify and respond to VE threats: government actors, influencers, communities themselves, and CSO partners. NIWETU’s support to the Kenya School of Government (KSG), a training institution for civil servants, in developing a CVE research and training institute, has contributed to NIWETU’s support to government, at the Objective 2 level. This activity fits squarely within Task 2.4 (Develop indigenous CVE research capacity) but sought to leverage NIWETU’s ongoing work with government and work with government actors, to strengthen their ability to research CVE issues to inform policy and decision-making. This case study looks at NIWETU’s work with KSG and discusses how the establishment of an embedded CVE research and training institute will contribute to improved government capacity in responding to the VE threat in Kenya. 19.1 Establishment of the Institute KSG was founded in 2012 as a parastatal entity, meaning that the organization serves the government while maintaining independence. Its mandate is to offer advisory, consultancy, research and training services to the public sector. It provides learning and professional development programs that help to inculcate a culture of public service and ethics within the civil service, to make the sector more responsive to the needs of Kenyan citizens. Given KSG’s positioning within the government and its primary role to serve as an institution of research and learning, KSG was well positioned to establish an institute of research and training on CVE-related issues, in order to meet NIWETU’s Task 2.4 objectives. The establishment of the institute came after two-plus years of relationship building and political negotiation, during which time NIWETU learned a series of important lessons about navigating challenging political circumstances, in support of building sustainable government capacity to understand VE dynamics within the country and coordinate VE policy formation: 1. Given the high prioritization of CVE within the GOK, housing such a research and training institute within government will help ensure its continued relevance and sustainability after NIWETU ends. While the original intention was to establish the Influencers Communities CSOs Government 201 institute at a Kenyan university, a series of decisions by USAID and NIWETU fed into the decision to consolidate the work with the government and establish this institute in a government facility that was already set up to provide research advisory services. As KSG is the official training institution for civil servants who are required to complete certain course work in order to progress to more senior level positions, KSG’s relevance is inherent. Equally, housing the institute within KSG will facilitate the mainstreaming of CVE knowledge, equipping graduates with the understanding of CVE’s importance across all strands of government and policy within Kenya. A KSG staff member also noted that government officials had reservations about establishing such an institute at a university where data security was uncertain, and where information could flow more freely. From their perspective, working with a government (or parastatal) agency was preferred. 2. Leveraging pre-existing and growing relationships with government agencies serves as important entry points into the co-design and development of such an institute. Ongoing consultation and communication with KSG counterparts helped maintain momentum, even when bureaucratic processes caused some delays; discussions around establishing a Memorandum of Understanding (MOU) between NIWETU and KSG, and between USAID and KSG, caused some initial delays, but both NIWETU and KSG continued to meet and plan to ensure that momentum was not lost. These discussions also helped to ensure a strong co-design process, with the establishment of a Technical Working Committee in October 2018, that would guide the design of the work plan and identify the gaps in KSG’s existing capacity on training and researching C/VE matters, which the institute would then fill. The committee also created a precedent of partnership between KSG and NIWETU – and later NCTC – that will be instrumental in guiding the institute’s growth and will help generate buy-in from important stakeholders across government, to ensure that it remains relevant and responsive to evolving needs. 3. Senior-level institutional support can spur greater ownership over the development of such an institute. Initial conversations about the institute began in September 2017, between KSG’s Director General (DG) and other staff at the school, NIWETU and USAID. The DG’s personal interest and experience in international relations and security issues spurred the initial momentum from KSG’s side around this activity. The school also used to be part of the Ministry of Interior (MOI), since splintering off into its own entity; however strong relationships with MOI and the wider security apparatus remain strong. KSG would therefore have automatic buy-in from the relevant agencies who would be amongst the primary targets of the research advisory and training services the institute would offer. The DG also saw the establishment of a research and training institute focused exclusively on security issues as an opportunity to create the first security studies institute in Kenya. It was noted that most academics working in this space are housed within political science, international relations, or other humanities departments, while the one security studies institute in Kenya, domiciled at Egerton University in Nakuru, focuses primarily on military science. KSG’s interest in the institute also stemmed from a desire to create capacity within the school to carry out policy-oriented research on VE in Kenya, and to teach civil servants and other professionals about VE and CVE strategies. The leadership showed the commitment to ensuring that KSG remains responsive to the most relevant issues affecting Kenya, and to 202 providing civil servants with the needed skillsets to be able to adeptly respond. With similar experience establishing research institutes with the World Bank and other donors, KSG was also well positioned to be able to effectively establish, launch and maintain a new institute. 4. Greater credence should be paid to understanding power dynamics and fostering inter-government cooperation on the establishment of such an institute. The institute’s establishment required support from NCTC in order to proceed, because of NCTC’s mandate, including capacity building on CVE within the public sector and coordinating with other government agencies on CVE issues, amongst other tasks. Deconflicting NCTC’s and KSG’s work would therefore be vital, while also ensuring that the relevant government stakeholders buy-in to the institute’s vision to help ensure its success. This process took longer than expected, as NCTC delayed by one year in responding to a letter from KSG requesting a meeting to discuss the institute. NCTC raised concerns over the duplication of mandates that could arise from the creation of such an institute and wanted to be consulted at earlier stages. This delay underlines the importance of understanding and paying credence to political power dynamics, where mandates between security and development arms of government, and national and sub-national arms of government, already muddy the waters. Greater understanding of these political dynamics could enable the development of a nuanced strategic plan for government engagement and relationship building that takes into account personalities, interests and influence for which USAID can proactively engage in support of the program. NCTC ultimately bought into the concept for the institute and reaffirmed its importance in mainstreaming CVE throughout government functions, after responding to the letter from KSG and engaging in some discussions with KSG, NIWETU and USAID counterparts to discuss the vision for the institute, how it can complement NCTC’s existing work, and how KSG and NCTC can work closely to identify and carry forward the institute’s priorities. A Memorandum of Understanding (MOU) was signed between NCTC and KSG, which has resulted in more direct interaction between the two entities and ensured that NCTC is brought into all key decisions that the institute has made. The MOU also serves to minimize potential areas of duplication between KSG and NCTC and formalizes cooperation between the entities. 19.2 Key Achievements KSG co-designed and implemented a work plan to guide the establishment of the institute. Given the good relationships and keen interest mentioned above, KSG took ownership of the process to work with NIWETU on a two-year work plan that would properly establish the institute through financial support to KSG’s early activities. Working with an external consultant, Swordfish, with expertize in establishing similar entities in universities and other academic institutions, NIWETU and KSG engaged in an in-depth, two-day workshop to create a clear vision for the institute, to be known as the Security Management Institute (SMI), and develop a work plan that would achieve those aims. The work plan outlined different work streams, including the development of training curriculum and establishing a certificate course – later to be followed by the development of more intensive curricula for higher level courses – research activities, including case study analysis, discrete research, and validation workshops; and lastly, leading public forums, to present findings of research activities or convene regional stakeholders to discuss CVE issues from a policy perspective. The delays mentioned in signing the grant agreement have required KSG to compress two years’ worth of activities into nine months, placing further limitations on what KSG could reasonably accomplish within the available time frame . Nevertheless, implementation of SMI’s agreed tasks began in earnest almost immediately after the signing of the agreement, and significant achievements can be seen since. 203 SMI at KSG is fully functional, even though it has not yet been formally launched. NIWETU has supported complete furnishing and equipping of SMI, while KSG has also recruited dedicated relevant staff, including a director, four research staff, one IT officer, and one office administrator. The launch of SMI has been delayed because of the Covid￾19 outbreak. The official launch was due to take place in June 2020 upon completion of the institute’s strategic plan, but the ceremony has been postponed to a later date. SMI’s ability to produce clear outputs at an early stage, such as completing research activities and piloting a training course, make it responsive and relevant, and help to garner buy-in. Despite the fact that SMI has not been formally launched, it has still managed to carry out a number of activities under its work plan. In doing so, SMI’s leadership has shown a commitment to ensuring that it becomes a fully functioning institute within KSG, that produces timely and relevant outputs. Achievements such as those presented below signal early wins for SMI, which can motivate other stakeholders within KSG and other relevant government departments to buy-in to SMI and see its clear value. These outputs fall squarely within NIWETU’s Objective 2 (Increased Government Capacity), through training of government officials and other civil servants, in an attempt to mainstream CVE throughout government functions, and research, designed to identify and inform policy gaps and strengthen the government’s response to countering violent extremism. Key outputs and accomplishments to date include: Output Description Two policy briefs One focused on policy gaps in responding to VE in Kenya, and the other focused on gaps in implementation of existing policies. The next step is escalating the findings to higher levels in government, to inform policy making. A research study on the continuum of policy and practice on CVE in Kenya This study established the drivers of VE, determined the level of adoption of the National Strategy for Countering Violent Extremism, assessed the uptake of the CAPs on CVE, and identified the capacity training needs to address VE in the public sector in Kenya. Three validation workshops These workshops took place in January 2020, where KSG presented findings from research already completed. Stakeholders from government and non-state actors participated. Two case studies These case studies focused on returnee policy and capacity building gaps in government. They have become reference materials in the Certificate Course Training. Development of training curriculum The curriculum is based on findings from the research study and policy briefs. Course materials are for the certificate level course, currently targeting government civil servants; a pilot round of training has been completed. The target audience includes mid- to senior-level managers in the public sector. Participants in the pilot training that took place in February 2020 pointed to a significant improvement in their understanding of C/VE. Development of a Training of Trainers Manual and Facilitators’ Guide These outputs are to support the certificate course. An evaluation from that course shows that trainees expressed an increase in their knowledge and understanding of VE and were likely to use what they have learned in their work. A ten-day residential and didactic training This training involved 24 KSG staff from various campuses. It sought to increase their skills on data collection, research tools and “We are keen to ensure that the training and our contribution to policy processes has an impact.” – KSG staff member 204 techniques, research ethics, quality assurance, confidentiality and consent, relevant to research on C/VE. Hosted a two-day workshop for NCTC and the Tony Blair Institute This took place in March 2020 and focused on CVE and the public sector. Collaboration between KSG and other government departments has been vital to the success of these activities. Despite the challenges in establishing KSG’s relationship with NCTC, KSG has still collaborated with other departments, such as the Ministry of Interior (MOI) and other security agencies. While NCTC now plays a key role in all activities implemented by SMI, other government actors have also contributed through validation workshops, information sharing for relevant policy briefs, and participation in training. MOI, and other security agencies, such as the Anti￾Terror Police Unit, National Intelligence Service, and National Police Service, amongst others, have all participated in workshops and validation exercises of the research findings. Perhaps the early success and achievements of SMI can be attributed to the strong and committed leadership within KSG, and from counterparts at NCTC. There is keen interest in seeing SMI succeed from both agencies, which has resulted in a lot of early momentum in implementing the above-mentioned activities and designing the Institute’s strategic plan, which will guide its objectives and activities over the next few years. 19.3 Long-Term Vision To guide the SMI’s forward planning, the institute, with the support of NIWTEUE staff, is developing a strategic plan which will outline the institute’s key objectives and activities; and will develop a strategy for fundraising, which is crucial for ensuring sustainability. An MOU signed between KSG, NCTC and MOI solidifies the relationships between these three key stakeholders and demonstrates the commitment to continue supporting the Institute. The MOU also outlines requirements for civil servants to pass training courses at the Institute in order to progress to higher levels within government. This is significant, because it codifies the importance of understanding C/VE related matters within government and shows that these three entities are committed to strengthening the understanding of the issues across the civil service. KSG has demonstrated a clear interest in expanding the role and relevance of the Institute through its five-year strategic plan. While the strategic plan (2020 – 2025) is still in draft form, it demonstrates KSG’s interest in expanding SMI’s reach to cover a wider range of security￾related issues, with the focus remaining on C/VE. According to one KSG staff member, SMI aims to guide Kenya’s engagement on domestic and regional security issues. This decision also acknowledges that security issues do not exist within siloes; rather, they are often inter-connected and therefore must be understood and acted upon in a more holistic manner. SMI will also seek to incorporate aspects of CVE training into other security related trainings for the police and other security providers. Future collaboration with other stakeholders is a key facet of the strategic plan, which will ensure the Institute remains relevant and responsive to the evolving context. The strategic plan also lays the groundwork for future collaborations with other stakeholder groups, which will help the institute gain support and remain relevant to CVE and broader security discussions in the region. Collaboration with think tanks and other research institutions operating in the region will help SMI to increase its capacity and understanding of the issues and will facilitate information sharing and learning across different research entities. Establishing these relationships will also provide opportunities for co-convening workshops and seminars to disseminate a growing body of research and evidence around VE and effective CVE practices. Because KSG is uniquely positioned within 205 government as a parastatal organization, it can also leverage its existing relationships with government entities, such as parliamentary committees and the Ministry of Foreign Affairs. The strategic plan suggests working with the parliamentary committee for defense and foreign affairs and increasing their understanding of security issues more widely, but C/VE, specifically, is vital. More intensive training courses are being developed for relevant government actors. Executive and diploma courses targeting County Commissioners and Deputy County Commissioners, who need more rigorous training to be able to carry out their jobs effectively, are currently under development. The executive course would target those who do not have the time to sit for longer or more intensive courses, such as magistrates. SMI intends to mainstream CVE in other existing KSG courses. As part of the long-term vision and promotion of sustainability, SMI is working to incorporate relevant C/VE-related information into other existing KSG courses, including within the Strategic Leadership Development Program, Senior Management Course, and Supervisory Skills Development Course. These courses target more senior management levels within the civil service, which participants in the pilot certificate course recognized must be targeted to ensure that the lessons learned and skills gained become institutionalized within various government agencies. 206 20. ANNEX X – CASE STUDY: NIWETU ENGAGEMENT IN MANDERA COUNTY 20.1 Mandera in 2017 In response to an increase in attacks in the county, that claimed the lives of 12 people in one month alone in 2016, the Government of Kenya (GOK) imposed a dusk to dawn curfew to address growing insecurity, and it has remained in place ever since. To better understand the evolving VE dynamics in Mandera, and the impact of rising insecurity as a result of increased VE activity in the area, NIWETU conducted an initial actors and opportunities mapping exercise, which was finalized in late 2016. The research showed a worrying trend of increasing attacks in the county, alongside a growing presence of radical elements and recruiters easily crossing through the porous border between Kenya and Somalia. At the same time, harassment, police brutality, extrajudicial killing, and feelings of ambivalence in the judicial system increased tension and hostility between local communities and the state, creating space for extremist narratives to spread. The few individuals who, at the time, may have been willing to positively engage with government and security counterparts to develop a unified CVE strategy, reportedly shied away because of fears of retribution by Al-Shabaab (AS) for working against them. With AS reportedly heavily invested in local businesses, levels of fear were high, as residents feel they are constantly being watched.49 People were even afraid to call AS by its name in public, according to some key informants. Recognizing that the security context in Mandera was deteriorating, NIWETU sought an opportunity to intervene and support the development of localized efforts to counter the growing threat of extremism. Working first through the rapid response mechanism to assess the feasibility of maintaining a longer-term permanent presence in Mandera, NIWETU supported Mandera District Peace Committee to implement a two-day stakeholder dialogue to discuss the VE threat and better understand the challenges faced in generating collective action. Since then, NIWETU has maintained an active, full-time presence in Mandera County, the only CVE program in Kenya to do so. It has allowed NIWETU to remain at the forefront of the CVE debate in the county, working alongside state and non-state counterparts to develop a shared vision and mission for countering the evolving threat in the country. This case study will look at NIWETU’s work in Mandera across sectors over NIWETU’s lifecycle and will identify the key achievements made over this timeframe. It will conclude with a look at Mandera county now, and the impact of cross-border political tensions that have emerged on the CVE landscape in the county, as actors look to continue leveraging lessons learned through engagement with NIWETU to counter the extremist threat in the area. 20.2 Working Through a Systems-Based Approach NIWETU has worked through its systems-based approach in Mandera, identifying strategic entry points into the community, in order to build capabilities to respond to the VE threat in Kenya. NIWETU worked closely with county and national government counterparts to improve coordination between themselves and with communities: with influencers to increase their leadership skills and understanding of CVE so they could be effective CVE advocates within their networks; with CSOs to deliver activities to key community groups to strengthen understanding of the VE threat and the ability of the community to respond; and with communities, through consultations and dialogues to inform the design of the County Action Plans (CAPs) and to engage with government counterparts. Below, we look at NIWETU’s support to each set of actors. 49 Task Order 2. Actors & Opportunities Mapping: Mandera County. December 2016. 207 20.2.1 Civil Society Organizations NIWETU’s first activity in Mandera, with Mandera District Peace Committee, brought together a range of stakeholders, including youth and women’s groups, other CSOs, community leaders, and county and national government, in one of the first such gatherings to discuss the VE threat in the county and identify the challenges faced in trying to counter it. Few, if any, such engagements had happened in Mandera prior, likely as a result of the fear associated with discussing AS and VE openly. Since this initial engagement, NIWETU has continued to work with other CSOs in Mandera on disseminating CVE skills and knowledge to a range of key interest groups, to boost understanding of the VE threat, encourage closer collaboration with security agencies, and support the development of localized responses to the threats of radicalization, recruitment and attacks. Partners have included Nomadic Assistance for Peace and Development (NAPAD), Focused Approach Development Concern (FADC), GLAD, and Women Care and Concern (WCC). Working through local community-based organizations has legitimized NIWETU’s CVE work in Mandera. The fear associated with openly speaking about and trying to counter extremism in Mandera was significant at the start of the activity. As such, trust-building was key to the success of the activity, and to even encouraging participation. At the same time, the problem had to be localized, as Mandera faces a unique set of circumstances; it is home to a rather ethnically and religiously homogeneous population, is affected by cross-border issues with both Ethiopia and Somalia, and is a largely rural county with limited road networks, where commercial flights only began in 2010. At the same time, clan dynamics are fragile, and inter-clan conflict can quickly erupt. Educational outcomes remain low, and partners have pointed out that low literacy rates have at times been a hindrance in their training sessions. As such, working with both staff and partners who understand these dynamics, are embedded within the communities, and are able to work at the community level in contextually relevant ways are key needs for success. NIWETU’s approach to working through these local organizations, that are entirely based in Mandera, helped ensure that the complexities of the context were adequately addressed and that community members felt comfortable engaging. Collaboration across civil society actors on CVE related issues has improved, largely as a result of the CEF. CSO representatives acknowledge that collaboration between actors on issues related to CVE and the broader peace and security agenda used to be limited. Competition over Influencers Communities CSOs Government 208 resources and donor funds meant that every group was keen on earning their share of the funding. Since the formation of the CEF, a key initial outcome of the CAP, which has been championed both by governmental and non-governmental actors, collaboration has naturally increased. While funding remains limited, CSOs acknowledge the importance in working together to develop a shared agenda. The CEF has been credited with facilitating this coordination: “The CEF provides a unique meeting point where CSO stakeholders have discussed issues of mutual benefit professionally, while sharing relevant information and lessons learned. The relationships formed have been used to enhance CVE efforts, rather than competition.” Local CSOs are championing CVE activities across the county. Through their participation in NIWETU-supported activities, either as grant recipients or through engagement platforms such as the CEF, local organizations are beginning to own the CVE agenda in Mandera county. With a greater understanding of the significance of the threat, they are better placed to address it, according to one CSO representative. Another shared that his organization, FADC, has placed a CVE focal point in each sub-county in Mandera, in order to better spread key messages, skills and information to communities that are often left out of CVE initiatives happening within Mandera town. 20.2.2 Government Strong anti-government sentiments persisted in Mandera at the start of NIWETU’s engagement in the county. Extrajudicial killings and other “hard” approaches, in which community members reported being arbitrarily targeted and profiled, were the norm. Reporting suspicious activity or incidents was not common, out of fear of harassment and suspicion by the security sector. Working closely with county and national government counterparts was crucial in Mandera to support community-driven CVE activities. NIWETU effectively leveraged the County Governor’s personal interest in CVE in support of the CAP formation process, CEF, and broader CVE agenda. Previous research has shown political will and buy-in to be key determinants of effective CAP implementation, through the formation and regular engagement of the CEFs, and pushing for use of local resources to fund activities.50 In Mandera, and perhaps in contrast to other NIWETU counties of operation, the Governor has demonstrated keen interest in supporting the county’s CVE agenda. Through NIWETU’s approach of working alongside county governments with constant informal consultation, the Governor’s office became a key partner for NIWETU in championing the CVE agenda within the county and pushing forward momentum on supporting locally driven CVE activities. At the same time, the county government designated a specific department, the Department for De-Radicalization and Countering Violent Extremism, in 2018, under the current Governor’s watch, to spearhead the county government’s response to the threat. The department therefore coordinates the CEF and, according to an official in the department, supports CVE activities at the “grassroots” through their field staff dispersed across the county. The impact of establishing a full-time presence in Mandera is clear when looking at the increased capacity and commitment of the Mandera County government to supporting CVE activities locally. NIWETU’s commitment to opening an office in Mandera helped the program and government counterparts learn to respond and adapt to changing scenarios more quickly. The team was able to understand and respond to the impact of political issues, clan dynamics, community perceptions, and other factors on C/VE quickly. Recognizing that these issues do not exist in siloes, a holistic understanding of the context was crucial to be able to respond and adapt quickly. Working in close collaboration with government counterparts, they too were able to learn about the interconnected nature of various drivers of conflict and insecurity, and therefore develop strategies and approaches specific to Mandera’s context to respond. The establishment of the Department for 50 Research includes Task Orders 17 and 23 on the CAP formation and implementation processes, respectively. 209 De-Radicalization is one example, while the county government’s support to the National Police Reservists (NPRs) is another. The Mandera County government has allocated funds for CVE programming in the county budget and used them to train Champions and fund the NPRs. County government officials took a keen interest in NIWETU’s C4C activity; they observed part of a training session for the Champions and then decided to fund their own Champions program, training 200 sub-county administrators and other front-line community security actors spread across the county, in basic CVE skills and knowledge. Those trained Champions are then meant to train others, to keep cascading skills and knowledge more deeply into communities across the county. The county government also chose to supplement funding to the NPRs, who are key players in maintaining safety and security in the county. The county government has provided stipends and other logistical support to NPRs based in Mandera town and Kutulo, another VE hotspot in the county. While buy-in from national government counterparts has lagged, momentum still continues. The County Commissioner’s office has been less engaged in the CEF and CAP process than the Governor’s office. The previous CC in Mandera, who had championed the CAP formation process and had taken a keen interest in strengthening Mandera’s CVE response, was transferred shortly after the CAP launch; his successor has shown little interest, in comparison. Likewise, an Assistant CC (ACC) who was attached to the Department for De-Radicalization was recently transferred back to Nairobi and has not yet been replaced, leaving the county government to largely work on the government side of the CVE agenda on its own. While one respondent pointed to a “lack of good leadership” on the part of the national government, there is broad support for the county government’s efforts in this area. 20.2.3 Influencers Through the Champions for Change program, NIWETU worked with 12 individuals from across Mandera County, in a year-long training and networking activity, that focused on leadership skills and strengthening knowledge and understanding of VE and CVE in the Mandera context. By supporting influencers from across the county, NIWETU has been able to disseminate CVE skills and knowledge to a wider network of individuals. By strengthening the Champions’ understanding of extremism, including teaching them about the individual or societal factors that make someone vulnerable and at risk of being radicalized and recruited, the Champions were able to identify important target groups to sensitize and teach about extremism. They reached out to vulnerable youth, community elders and leaders, boda boda riders, women’s groups, religious leaders, taxi drivers, female small business owners, students and teachers. NIWETU encouraged the Champions to identify approaches that meet the specific needs of the particular groups targeted. For example, in meetings with teachers – which included members of the school Boards of Management, Parent Teacher Associations, and Ministry of Education officials – the discussions focused on how to influence young learners and teach them through the school system about extremism and the consequences of joining an extremist group. The Champions are leveraging their new leadership skills to spread important public service announcements, in addition to messaging about CVE. NIWETU’s work with the Champions helped them learn how to use their influence for positive change in their communities. In addition to working specifically on CVE issues, the Champions have now begun to leverage their influence and leadership skills to work on other issues affecting the community. Two Champions noted that they are working together to share public health related messages about the Covid-19 pandemic and spread accurate information about GOK’s response, to prevent the spread of false information. He said that the Champions have found this work particularly important at the moment, because it is allowing them to continue reaching out to communities and working together, while trying to address a serious issue that is severely impacting Mandera County. 210 NIWETU’s focus on the importance of volunteerism has encouraged the Champions to find creative ways to spread CVE messages. The C4C activity is built around the concept of volunteerism; Champions understand from the get-go that participation in the activity happens on a volunteer basis only. While NIWETU offers some funding to help the Champions kickstart their activities outlined in their Action Plans, it comes with the understanding that the activities will either continue on a volunteer basis only, or the Champions will be responsible for raising funding on their own. Instilling a culture of volunteerism in the Champions is crucial to ensuring the continuation and sustainability of the outcomes from the C4C activity. NIWETU’s systems-based approach has encouraged close collaboration between the Champions and county government counterparts. Working across different levels within the system simultaneously, means that linkages will be forged between the different actors NIWETU supports. The Champions have thus been able to collaborate closely with the Department for De￾Radicalization; while funds are limited, the parties often share information and discuss ideas for future programming, in line with the CAP priority pillars. At the same time, the Champions often send one representative to the monthly meetings of the CEF, where they report on progress made in implementing their Action Plans, as well as challenges and lessons learned. 20.2.4 Communities Most of NIWETU’s work at the community level happens via other touchpoints within the system in Mandera County. However, community groups have been engaged in consultations for the CAP formation process. Consulting with communities during the CAP formation process ensured that a wide array of perspective and experiences were gathered, to factor into the final CAP document. Mandera County faces a multitude of conflict and security-related challenges, including cross-border issues between Ethiopia and Somalia, inter-clan tensions, and extremist activity. Any response to violent extremism must therefore take into consideration these different dynamics and their impact on local communities. Working with different community groups, such as youth, women, teachers, and others, helped the CAP formation process to better understand these diverse experiences and consider how they should be incorporated into the CAP. Consultations at this level also enabled these different community groups to share their thoughts on areas for prioritization within the CAPs, including faith-based ideology, political factors, and security factors, as the top three priority pillars. NIWETU’s leadership in reaching out to local communities is being matched through local CBOs with support from the county government. A county government official spoke of an activity with a local organization, called GLAD, to reach communities in Banisa, Mandera West, and El Wak sub-counties, focusing on peacebuilding and conflict resolution, with overtones of CVE messages. These efforts point to the county government’s commitment to support locally driven initiatives and to begin to mainstream CVE messaging into activities that speak more generally about peacebuilding or conflict resolution. There is also a growing sense amongst these actors that, working more closely with local communities across the county is an important step in responding to the VE threat, particularly as it continues to evolve and reach into new, previously unaffected, communities. Local Organizations Responding to the Threat Communities in Banisa sub-county previously had no knowledge of VE or CVE. Having never experienced an attack or been affected by extremist activity, no one saw a need to deliver CVE activities in this community. Nevertheless, with support from NIWETU, a local CBO called GLAD began CVE activities in the area, just as a series of attacks hit along the Banisa-Rhamu road. As a result of the activity, and following the attacks, community members mobilized and reported the presence of AS militants that had been spotted in the area. This actionable intelligence information has proved useful for the security sector. This was an inherently risky endeavor, however, as the participants could have been targeted by the militants. Additional training on safety and security considerations for the participants would also have helped to manage that risk. 211 NIWETU’s systems-based approach has encouraged different actors to identify ways to reach communities in more hard-to-access locations. While NIWETU’s activities in Mandera were limited, and many were concentrated in a few accessible hotspot locations, NIWETU’s support to other categories of actors within Mandera has strengthened the system’s ability to disseminate key CVE messages to communities in more hard-to-reach areas. The C4C initiative is a key example of this; the Champions can leverage their influence and existing networks around the county to share what they have learned through the NIWETU-supported trainings and other activities. At the same time, CSOs also have a stronger grasp of the issues, including a better understanding of how to address them, and can work with communities in remote locations to develop relevant and localized CVE strategies. NIWETU’s work in Mandera has contributed to an opening up of the CVE space, facilitating open and honest conversation and dialogue between communities and other stakeholders. One respondent noted that residents now feel comfortable openly saying “Al-Shabaab” in public settings, including community barazas and in the media. An example was also shared of a woman who had been washing clothes in the river in a remote area and came across a grenade. She immediately reported it to security officials. This was hailed as a noticeable success, as prior to NIWETU’s sustained engagement in the county, residents feared reporting anything suspicious, out of fear that they would then be targeted and harassed. 20.3 Mandera in 2020 In February and March 2020, an internal conflict in Somalia between the Jubaland state militia and Somali National Army spilled over into Mandera town. Jubaland forces sought refuge in a police station in Mandera and tension between the Government of Kenya and Federal Government of Somalia escalated, because of a perception that the Kenyan government was aiding the Jubaland rebels. As a result, insecurity along the border has increased and there is a pervasive fear that extremists will take advantage of the security sector’s distraction with the ongoing political crisis. Two respondents also noted that there is now confusion in identifying who is AS and who is part of the Jubaland militia, which creates another entry point for AS forces. They take advantage of poorly patrolled routes on the outskirts of town that allow easy movement in and out of Kenya. Respondents again feared that the crisis would detract from the county government’s and security sector’s focus on CVE, as their resources and attention would be shifted toward addressing the border crisis, creating clear entry points for extremists. While they noted that attention had moved to focus on quelling the row between Kenya and Somalia, drawing resources away from CVE, since the conflict has begun to die down there has been a renewed focus on CVE efforts within Mandera. One key informant suggested that the government’s shifting focus to addressing this issue could negatively impact community members’ perceptions of government actors, based on how the conflict is handled. He further noted that it detracts from the ongoing CVE efforts in the county that had begun to gain momentum: “During the recent fighting by SNA and the Jubaland militia that was housed in Kenya, the grenades struck some houses in Mandera town, cars were burnt, a stray bullet killed one person, and several people were injured. One young man with a bullet lodged on his body was referred to Nairobi for removal of the bullet, but the medics there refused to attend to him, and he was taken to Awassa, Ethiopia where the bullet was removed. This not a healthy environment for CVE programming and the community’s perception of the government could be affected as a result.” This is particularly significant, as substantial gains have been made in Mandera in relation to community￾government and community-security relations over NIWETU’s lifecycle and serves as an important reminder that other issues outside of the control of a particular program or activity can negatively (or positively) affect outcomes. 212 Nevertheless, the support that NIWETU has provided to Mandera County stakeholders has resulted in some sustained gains, and perhaps also contributed to a maintenance of some degree of stability in the region. The county government’s strengthened focus on CVE issues through the Department of De-Radicalization and Countering Violent Extremism has helped to strengthen local security structures; through the department, the county government is supporting 334 NPRs who conduct security patrols at night and work to secure government installations. While funds at the present moment remain limited, the government is still working in collaboration with local partners to design CVE activities should more funds become available, while also working with CSOs with existing active grants on CVE, to strengthen their approaches and support cross-activity learning. Local communities are also demonstrating a keen interest in engaging in CVE programming locally. Perhaps as a result of awareness raising campaigns that have spread across the county, likely with the support of NIWETU’s Champions who hail from different parts of the county and have influence over a wide range of community groups, locals are beginning to identify ways to support peace and security structures within their areas. In Mandera North, for example, where the county government has not yet been able to provide funding to the NPRs, the community instead has identify approximately 100 volunteers with experience in handling guns, who are now working to play a key role in addressing the VE threat locally and ensuring the safety of their respective communities. The county government is also recruiting religious leaders in the areas of Lafey, Mandera East, Mandera South and Mandera North, to conduct important community sensitization activities on VE issues so that they can disseminate messages to their networks. This again presents another case of the county government demonstrating their commitment to the issues and identifying ways to take the lead on locally driven and relevant CVE initiatives. At the same time, key informants underscored the importance of continuing to push forward with CVE programming, stressing the need to consider a wide range of conflict and instability drivers in recognition of the complex operating environment. The Governor’s keen personal interest in pushing forward with CVE initiatives and working with community groups to identify relevant approaches, has encouraged other actors, particularly CSOs, to commit to the same. With regular meetings of the CEF taking place to discuss CVE programming, lessons learned, and broader changes in the operating environment, CSOs are playing an active role in ongoing discussions with government counterparts on the issues. 20.4 Lessons Learned A number of lessons have emerged on how to work through a systems-based approach to strengthen CVE capabilities at the county level. 1. Tailored support and engagement with each stakeholder helps strengthen cooperation and buy-in. By working simultaneously to support the capacity building of different groups of actors within the system, NIWETU has organically created a new space for these groups to begin to work together. NIWETU’s support to these different players is also tailored specifically to strengthen the existing skillsets and networks of each group, meaning that NIWETU is building off existing strengths and relationships to help them identify responses to VE in the county. This approach has proven effective; it takes what is already working well and adapts it to address an emerging and evolving issue. It also seeks to identify existing gaps and fill the void through increased collaboration and support. This approach therefore encourages more buy-in from different actors, as the work they are already doing is simply being adapted to respond to VE, instead of reinventing the wheel and forcing stakeholders to completely shift course. 213 2. Maintaining full-time engagement ensures that the team can more deeply understand and respond to local dynamics in real time. NIWETU’s locally led presence in Mandera has helped the activity gather and analyze information on the evolving context more quickly. Particularly in an area where a multitude of conflict drivers factor into vulnerability and affect the key areas upon which NIWETU has worked, the ability to understand the inter-connections between these issues is vital to developing a relevant and effective response to the VE threat, as evidenced by the current political crisis that has embroiled the county. At the same time, this presence has enabled the NIWETU team to better relate to and understand communities’ needs, in a way that standard intelligence gathering or research does not; this information has proven valuable to USAID and other USG colleagues seeking to understand dynamics in the area, and again has contributed to the development of a more relevant response. 3. Effective CVE strategies must take into account other drivers of conflict and insecurity affecting local communities, as these issues often feed off the tension they each cause. The political crisis between Kenya and Somalia that played out along the border in Mandera is a clear example of why a one-size-fits-all approach to CVE is not effective. It must take into account other factors and causes of insecurity and instability within a community, as these are often the spaces that extremist groups seek to exploit. Working at different levels across the systems enables this outcome; it encourages different stakeholder groups to work together and collaborate, share diverse perspectives and experiences, and design strategies that speak to different groups’ particular situations and needs. This approach therefore inherently leads to the development of CVE strategies that are grounded in the lived reality of communities in the area, and therefore can enable responses to the VE threat that take into account other issues that cause conflict or instability. 214 21. ANNEX XI – VE INDEX SCORES This section presents the VE index scores from baseline and endline for each county to identify changes in the VE context over the past four years. Garissa Statements Level of Agreement Baseline Endline Extremism and recruitment to join extremist groups is less of a problem now than it was 12 months ago. 3.95 2.89 Organized crime and gang activity is less of a problem now than it was 12 months ago. 3.54 3.39 Community members feel safer now than they did 12 months ago. 4.04 2.26 Over the past 12 months, the number of extremist attacks in this county has decreased. 3.82 2.02 Over the past 12 months, the number of extremist attacks throughout Kenya has decreased. 3.87 2.50 Overall average score 3.84 2.61 The VE index in Garissa indicates that respondents feel overall security has declined since the baseline. Qualitative responses reference an increase in the number of VEO attacks in the county, particularly on local schools. Some pointed toward a greater prevalence of Al-Shabaab communication within their communities, often threatening or warning of imminent attacks in order to scare the population. Recruitment of youth by VEOs remains an issue and some noted that individuals who had been ‘disappeared’ were reportedly observed by pastoralists conducting armed raids in rural areas. When discussing gang activity, a respondent said that boda boda drivers are fearful of driving past Bula Mzuri due to a youth gang called ‘Squad Chafu’. Isiolo Statements Level of Agreement Baseline Endline Extremism and recruitment to join extremist groups is less of a problem now than it was 12 months ago. 4.48 3.67 Organized crime and gang activity is less of a problem now than it was 12 months ago. 4.06 3.44 Community members feel safer now than they did 12 months ago. 3.97 4.38 Over the past 12 months, the number of extremist attacks in this county has decreased. 3.85 3.94 Over the past 12 months, the number of extremist attacks throughout Kenya has decreased. 4.68 2.98 Overall average score 4.20 3.68 The overall average VE index moderately declined in Isiolo, while there was a slight improvement in perceptions of community safety and frequency of extremist attacks. Those reporting a decline in security noted a perceived uptick in recruitment activity, noting that some of the organizers of the 14 Riverside attacks were allegedly from Isiolo. Respondents also reported an increase in problems arising from cattle rustling among pastoralists and drug abuse. Conversely, those who said security improved highlighted better police-community relations and more moderate messaging from local religious 215 leaders. Further, some attributed improved levels of safety to more frequent security patrols and more active Nyumba Kumi activity. Mandera Statements Level of Agreement Baseline Endline Extremism and recruitment to join extremist groups is less of a problem now than it was 12 months ago. 2.88 3.84 Organized crime and gang activity is less of a problem now than it was 12 months ago. 1.86 3.77 Community members feel safer now than they did 12 months ago. 3.5 3.47 Over the past 12 months, the number of extremist attacks in this county has decreased. 3.24 3.38 Over the past 12 months, the number of extremist attacks throughout Kenya has decreased. 3.48 3.02 Overall average score 2.99 3.50 There was a notable improvement in the average VE index in Mandera between baseline and endline, spurred by a perceived drop in VE recruitment and organized crime. Several respondents pointed to increased sensitization around VE among religious leaders and community elders as a result of their participation in CVE forums. Many also noted that youth had become more aware of the dangers of radicalization and were therefore less vulnerable to recruitment activities. A few respondents referenced an attack on a passenger bus that occurred around the time of the endline but acknowledged that generally attacks had decreased. There was also a reported decline in organized crime, while petty theft and drug sales persisted. Nairobi Statements Level of Agreement Baseline Endline Extremism and recruitment to join extremist groups is less of a problem now than it was 12 months ago. 2.94 3.55 Organized crime and gang activity is less of a problem now than it was 12 months ago. 2.60 3.37 Community members feel safer now than they did 12 months ago. 2.78 3.60 Over the past 12 months, the number of extremist attacks in this county has decreased. 4.42 3.54 Over the past 12 months, the number of extremist attacks throughout Kenya has decreased. 4.04 2.43 Overall average score 3.36 3.30 The average VE index in Nairobi remained largely unchanged between baseline and endline. However, there was a notable decrease in levels of recruitment, gang activity, and improved perceptions of safety. Conversely, respondents considered the overall frequency of attacks to have increased. Qualitative responses attributed the decrease in VE recruitment to more CVE sensitization among youth and more constructive relationships with security actors. Similarly, respondents linked the reduction in organized crime to an increase in the number of mentorship programs and livelihood opportunities for youth in local communities. The perceived increase in attacks was reportedly due to highly visible incidents, such as the 14 Riverside attacks and the recent extrajudicial killings carried out by police. 216 Wajir Statements Level of Agreement Baseline Endline Extremism and recruitment to join extremist groups is less of a problem now than it was 12 months ago. 4.16 3.47 Organized crime and gang activity is less of a problem now than it was 12 months ago. 4.77 3.11 Community members feel safer now than they did 12 months ago. 4.66 3.28 Over the past 12 months, the number of extremist attacks in this county has decreased. 4.83 2.65 Over the past 12 months, the number of extremist attacks throughout Kenya has decreased. 4.59 2.52 Overall average score 4.60 3.01 Qualitative responses pointed toward a moderate decrease in security due to a perceived increase in VEO activity in Wajir. In particular, respondents referred to an uptick in al-Shabaab attacks in Wajir East, contributing to security concerns in surrounding communities. Feedback indicated that, while instances of VEO activity had increased in some areas, communities were more inclined to report such incidents to security forces. A large proportion of respondents highlighted the lack of employment opportunities and livelihoods for youth as a key issue contributing to instability. 217 22. ANNEX XII - Q3-2019 - QUARTERLY RESEARCH REPORT: COMPARATIVE ANALYSIS OF COMMUNITY AND GOVERNMENT RESPONSES IN THE AFTERMATH OF THE WESTGATE AND 14 RIVERSIDE ATTACKS 22.1 Key Findings The following sections present further details on the findings from the research conducted. We begin by offering an overview of terrorism in Kenya, followed by discussions on the national government’s response and community mobilization in the greater Eastleigh area. The report concludes with a discussion of community perspectives that help understand any changes experienced in the day-to-day lives, vis-à-vis security and government relations, of local residents since the Westgate attack. 22.1.1 A Brief History of Terrorism in Kenya Kenya first experienced terrorism almost two decades after attaining independence in 1963. On New Year’s Eve of 1980, a bomb went off at the Jewish-owned Norfolk Hotel, leaving 20 people dead and another 80 injured. Allegedly, either the Popular Front for the Liberation of Palestine (PFLP) or Palestinian Liberation Organization (PLO) was responsible for the attack. The attack was allegedly organized in retaliation for Kenya’s role in the Israeli-led ‘raid in Entebbe’ operation, which successfully rescued more than 250 hostages in neighboring Uganda following the seizure of an Air France plane by the PFLP.51 Nearly two decades later, Al-Qaeda bombed the United States (US) Embassy in Nairobi in 1998 and the Paradise Hotel in Kikambala, Mombasa in 2002. Up until this point, terror attacks in the country could be linked to Kenya’s ties to Western powers and their involvement in the geopolitical dynamics in the Middle East. Kenya would experience a new wave of AS-linked terror attacks across the country following its military incursion into Somalia in October 2011. Kenya launched Operation Linda Nchi, ‘protect the country,’ in response to a series of tourist abductions by AS along the coastline of Kenya. In perhaps its most brazen and boldest show of terror yet, AS militants attacked the Westgate shopping mall in upmarket Nairobi in 2013, killing at least 67 and injuring scores more. Several other attacks followed in various parts of the country, notably in Mpeketoni, Lamu County, and in Mandera County. AS would then record their deadliest attack yet in 2015 after storming Garissa University in Northeastern Kenya, killing 148 students and staff. On January 15, 2019, AS perpetrated its second largest attack in Nairobi at the 14 Riverside Drive complex. On that fateful afternoon, four heavily armed militants stormed into the complex opening fire and hurling explosives. One of the assailants, later identified as 25-year-old Mombasa native Mahir Khalid Riziki, stood in an open lawn overlooked by a restaurant patio and blew himself up. The explosion killed six patrons of the Secret Garden Restaurant, in the center of the complex.52 Kenya’s disaster response and security apparatus quickly mobilized at the scene. Following a coordinated multiagency tactical response, the Kenyan security forces rescued more than 700 people and neutralized the threat in less than 20 hours. The execution of this operation sits in stark contrast to the operation at Westgate five years prior. The response at 14 Riverside was remarkably better 51 Kiruga, M. “20 Killed in Bomb Attack on Norfolk.” Daily Nation. 15 Sept 2013. See: https://mobile.nation.co.ke/lifestyle/1950774-1993444-format-xhtml-tt6skxz/index.html. 52 Kimuyu, H. “Police Name Dusit Suicide Bomber: Mahir Khalid Rizik.” Daily Nation. 19 Jan 2019. See: https://www.nation.co.ke/news/Dusit-suicide-bomber-named/1056-4941976-94fyg3z/index.html. 218 coordinated, resulting in a swifter resolution to the siege and significantly fewer casualties. AS issued a statement claiming responsibility for the 14 Riverside attack that left 21 people dead.53 Against this background, this report seeks to unpack government and community responses to VE in Eastleigh, using the Westgate and 14 Riverside attacks as reference points for a qualitative comparative analysis. Eastleigh was chosen as the primary focus of this study because of the government’s traditional heavy-handed approach in the community and its reputation for being a hotspot for extremism. 22.1.2 National Government & Security Sector Response Days after Kenya initiated Operation Linda Nchi (meaning ‘protect the country’) in the Juba Valley in southern Somalia, the Assistant Minister for Internal Security, Orwa Ojode, told Parliament in October 2011, “After the Somalia thing is over, I am going to do a mother of all operations here in Nairobi to remove all Al-Shabaab and Al Qaeda.”54 Mr. Ojode continued to describe AS as “a big animal, with the tail in Somalia, and the head of the animal is hidden here in Eastleigh,” articulating the government perception that Eastleigh housed, and perhaps even protected, AS operatives in Kenya, and laying the groundwork for the government’s heavy handed response in Eastleigh after the Westgate attack two years later. From this starting point, there has been a clear evolution in the government’s understanding about the nature of VE in Kenya, as evidenced by their more nuanced and deliberate approach in the aftermath of the attack at the 14 Riverside complex. The GoK response during and in the immediate aftermath of the 14 Riverside attack showed a marked improvement in comparison to the response to the attack at Westgate Mall. The appearance of better coordination between the various security agencies at the scene, the swift response time, and relatively few casualties in comparison to the Westgate attack combine to suggest that the crisis response capabilities have significantly improved. Following the attack, security forces responding within the Eastleigh community took a more targeted and deliberate approach to investigation and arrest, rather than rounding up suspects without clear cause. This section will provide further evidence to support the conclusion of an improved response during and immediately following the 14 Riverside attack, with a focus on security forces’ handling of the situation in the Eastleigh neighborhood. 22.1.3 Westgate attack response and the aftermath The security sector’s response during the Westgate attack was slow and tactically uncoordinated, with no clear chain of command between the different security agencies deployed at the scene. Reports later emerged that highlighted infighting amongst the different security agencies deployed at Westgate, which included the Flying Squad, the Recce Squad, regular police, Kenyan Defence Forces (KDF), Israeli commandos and Kenyan-Indian vigilantes.5 The GoK released conflicting and ambiguous information coming out of the four-day operation, which led to growing public mistrust of government. Westgate storeowners accused KDF personnel of looting shops in the mall, further corroborated by CCTV footage. 55 As a result, the government promised an official inquiry that resulted in the suspension and arrest of some army, police and fire brigade officers56 – the charges against these officers were later 53 Sperber, A. “Al Shabab wants you to know its alive and well.” Foreign Policy. 19 Jan 2019. See: https://foreignpolicy.com/2019/01/19/al-shabab-wants-you-to-know-its-alive-and-well-kenya-somalia-terrorism/. 54 Ngirachu, J. “Govt to weed out Shabaab sympathizers in Nairobi swoop.” Daily Nation. 19 Oct 2019. See: https://www.nation.co.ke/News/politics/Govt+to+weed+out+Shabaab+sympathisers+in+Nairobi+swoop/-/1064/1257988/- /nvftjsz/-/index.html. 5 “Blame game over Westgate attack.” Daily Nation. 26 Sept 2013. See: https://mobile.nation.co.ke/news/Blame-game-over￾Westgate-attack/1950946-2009266-format-xhtml-8ay6qx/index.html. 55 Howden, D. “Terror in Nairobi: the full story behind Al Shabaab’s mall attack.” The Guardian. 4 Oct 2013. See: https://www.theguardian.com/world/2013/oct/04/westgate-mall-attacks-kenya. 56 Kenya National Assembly (2013). Report of the joint committee on administration and national security; and defense and foreign relations on the inquiry into the Westgate terrorist attack, and other terror attacks in Mandera in North-Eastern and Kilifi in the coastal region. 219 dropped under unclear circumstances.57 In total, the siege lasted for more than 48 hours, as security forces struggled to regain control of the mall. Figure 1. An editorial cartoon satirizing the KDF led operation at Westgate/Source: gadocartoons.com The Westgate attack spurred a knee-jerk counter-terror security reaction. During and after the attack, the GoK deployed various security agencies to and around the Eastleigh neighborhood for a massive security operation. The operation sought to round up AS operatives and their accomplices and supporters in the area; the process appeared to be more of a heavy-handed show of state power reminiscent of past security operations. Many local residents in the Eastleigh area described this operation as standard government procedure following acts of terror in Kenya. A community organizer quipped, “This is the way it is with other such attacks” when describing anti-terror operations in the area. An FGD participant in her late teens also recounted the brutality security forces subjected her community to. She described how her own family members were affected by saying: “There was a day so many men in uniform were dropped at our area for a major search and they were beating people up. My father and brother were among those assaulted.” Her fellow FGD participant, a young Muslim man, corroborated her remarks and added that “Eastleigh residents have always been put on the spotlight whenever there is a terror attack.” Respondents speculated that this is largely because Eastleigh is known to host a large community of immigrants, or ‘foreigners’, most of whom come from Somalia. Indeed, locally-born and new Somali migrants mix in Eastleigh, but the association between terrorism and Somali migrants enables others to easily stereotype against all Somalis. A Somali youth said that both the government and the community at large stigmatizes and faults Somalis for acts of terror: “The Mzee (Yassin Mohamud Jama), owner of Kilimanjaro said the truth.58 A Muslim and Somali, he said, ‘I too have lost my children but yet you suspect me to be a terrorist. You see. These are the perceptions of Kenyans. You are accused even when you are a victim.’” 57 Leftie, P. “Two KDF soldiers sacked over Westgate looting.” Daily Nation. 29 Oct 2013. See: https://www.nation.co.ke/news/Two-KDF-soldiers-sacked-over-Westgate-looting/1056-2051746-scuskpz/index.html; “Theft cases against KDF soldier withdrawn.” Daily Nation. 29 Oct 2013. See: https://www.nation.co.ke/news/Theft-case-against￾KDF-soldier-withdrawn/-/1056/2051992/-/ets9xm/-/index.html. 1. 58 THE RESPONDENT WAS REFERRING TO MZEE JAMA’S MEDIA INTERVIEW FOLLOWING THE 14 RIVERSIDE ATTACK. SEE: “FATHER RECOUNTS HARROWING MOMENT SON CALLED DURING DUSIT TERRORIST ATTACK.” CITIZEN TV. 16 JAN 2019. SEE: HTTPS://YOUTU.BE/GIPS4IN_FCC 220 In March of 2014, about six months after Westgate, two more AS-linked terror attacks occurred in Mombasa and Eastleigh. These attacks, which were only a week apart, killed at least ten people and injured scores more, triggering the government to initiate Operation Usalama Watch. The operation targeted Eastleigh and other Somali dominated neighborhoods in Nairobi to allegedly ferret out illegal immigrants and terror suspects, emanating from government’s long-held belief that Eastleigh served as a safe haven for AS operatives and sympathizers. Mr. Ole Lenku, the then Cabinet Secretary for Interior, went on record as saying, “For the last few months we've had heightened insecurity. Time has come for a mop up to restore order.”59 The government deployed about 5,000 security officers to Eastleigh in what some saw as a crackdown on Somalis. During an FGD, a Somali youth said, “[security forces] were doing a door to door searches… It was an around-the-clock operation.” Usalama Watch indiscriminately rounded up about 4,000 people (majority Somalis), alleged to be unregistered aliens or refugees,60 and illegally detained them at the Kasarani Stadium. A police officer justified this knee-jerk security policy response by saying: “You find that there are unregistered foreigners living amongst us. It says a lot when a foreigner chooses not to register himself or herself. Maybe that person is not good for the society. Why then would they not go through the right procedure of registration? And these are the people who definitely need to be arrested during the police operations and be questioned as to why they have not registered as per the laws of the land.” Human Rights Watch estimated that the operation forced more than 1,000 people to relocate to refugee camps and about 100 repatriated to Somalia in contravention to international law.61 Residents also accused security forces of gross violations of human rights. Accusations of police brutality, rape, illegal detention, torture, and extortion arose during the operation.62 Many Somali men went into hiding during this time in order to avoid the indiscriminate crackdown. The operation and its fallout only added to a mounting body of evidence of heavy-handed security approaches targeting an ethnic community that has historically articulated grievances of systematic marginalization against the Kenyan state. 22.1.4 14 Riverside attack response and the aftermath While respondents in this study presented divergent attitudes toward the GoK’s counter-terrorism strategies, they all agreed that there was a clear improvement in the government response to the 14 Riverside attack in contrast to the Westgate attack. Respondents believed that the government demonstrated a heightened degree of control and presented a more unified response, with a more obvious inter-agency coordination and clearer, and more frequent, communication and dissemination of information to the public. Respondents also highlighted the quick response time, relatively short duration of the operation, and minimal casualties, as other key indicators of an improved response and capability. Reports suggest that with the help of international partners, the GoK restructured its security forces and trained them in effective VE response strategies.63 One respondent succinctly summarized this observation as follows: 59 “Kenya sends back ‘illegal’ Somalis after Nairobi Raids.” BBC. 9 Apr 2014. See: https://www.bbc.com/news/world-africa￾26955803#TWEET1095872. 60 “Kenya: Counterterrorism operations undermine rights.” Human Rights Watch. 29 Jan 2015. See: https://www.hrw.org/news/2015/01/29/kenya-counterterrorism-operations-undermine-rights. 61 “Kenya: Halt crackdown on Somalis.” Human Rights Watch. 11 Apr 2014. See: https://www.hrw.org/news/2014/04/11/kenya-halt-crackdown-somalis 62 “Kenya: Police abuse Nairobi’s refugees.” Human Rights Watch. 29 May 2013. See: https://www.hrw.org/news/2013/05/29/kenya-police-abuse-nairobis-refugees 63 AFP. “From Westgate fiasco to Dusit, Kenyan response praised.” Daily Nation. 19 Jan 2019. See: https://www.nation.co.ke/news/Dusit-attack--Kenyan-response-praised/1056-4941836-118nggi/index.html 221 “They were professional. They knew what to do and how to contain the whole situation. This is unlike Westgate, where the government was not organized in their response.” A government representative also suggested that the change in approach was a result of better use of intelligence and investigation, pointing to a more thoughtful and deliberate response. He stated: “When the Dusit64 attack happened, we did not see something like [Westgate]. There were no random arrests, and I think the government was trying to use the intelligence to try and understand where the problem was and the people who were involved in those problems, so we did not feel the Dusit attack like we felt the Westgate attack.” Likewise, in the following days, as government and security forces sought to identify and arrest suspects and others connected to the attack, respondents agreed that GoK officials involved demonstrated a higher level of professionalism and treated the Eastleigh community with a greater degree of respect. In particular, they highlighted the fact that security forces were not deployed to “brutalize” them as they had in the past, and instead believed that in the aftermath of 14 Riverside, the security agencies conducted strategic and targeted arrests based on evidence flowing from investigation.65 A youth leader thought about the measured response of security forces in the days after the 14 Riverside attack and concluded: “I suppose in the past they thought of terrorism as a foreigner problem. They have learned that Kenyans can be caught up in it too. They also realized that ethnic profiling stigmatizes communities and could even contribute to the radicalization of youths who have ethnically been stigmatized… Remember the mass security screening of Somalis at Kasarani sports stadium?” Some respondents also speculated that the profiles of the perpetrators and victims challenged traditionally-held stereotypes that Somalis and Muslims are the perpetrators and non-Somali Christians are the victims. When the profiles of the attackers and the casualties emerged after the 14 Riverside attack (specifically that non-Somali recent converts led the siege that killed two Somali Muslims, in addition to numerous others), some respondents speculated that Kenyans and the GoK started to better understand the ability of violent extremist organizations (VEOs) to reach and recruit from parts of society most thought were immune to such narratives and influences. An Eastleigh resident reflected on why he thought the government was more lenient in its response than in the past: “I think this time round the government was a bit lenient with us, because we were victims of the attack. One of the influential people from here lost his two sons in the Dusit attack.” These sentiments are reflective of the perceptions of research respondents and demonstrate the community believes that the profile of the perpetrators and victims contributed to the GoK’s change in approach. While it is equally important to consider the role of police training and internal learning 64 The 14 Riverside complex houses the Dusit D2 Hotel, a large hotel frequented by international guests. Respondents used Riverside Drive and the Dusit Hotel interchangeably to refer to the attacks 65 Agutu, N. “Two linked to Dusit attack arrested in Ruaka, Eastleigh.” The Star. 19 Jan 2019. See: https://www.the￾star.co.ke/news/2019-01-16-two-linked-to-dusit-attack-arrested-in-ruaka-eastleigh/; Muriuki, B. “7 suspected arrested in connection with 14 Riverside terror attack.” Citizen Digital. 16 Jan 2019. https://citizentv.co.ke/news/7-suspects-arrested-in￾connection-with-riverside-terror-attack-227171/ 222 activities, respondent perspectives suggest that key differences in the attack itself could have had an effect on how the government chose to react. 22.1.5 Summary analysis of GoK’s response In reflecting on the GoK’s new approach in dealing with the local community, some respondents suggested that Westgate was a major learning point for the government, which was most noticeable in the day-of response while the siege was unfolding, and in their subsequent engagement with the Eastleigh community. According to police respondents: “Yes, definitely there was a difference. The security arm is dynamic and that is why between Westgate and Dusit they had taken some strides. The government had put some strategies in place to avert such things from happening again.” He further noted that because of the changed tactical response, the community gained confidence in the security apparatus’ ability to deal with terrorist attacks. The comparatively low number of casualties, in contrast to the Westgate attack, he argued, was the key indicator of the government’s improved handling of the situation. Community members echoed the same sentiments with a Majengo￾based paralegal saying, “You know, we learn through mistakes,” while reflecting on lessons the government learned from its response following the Westgate attack. A businesswoman from Eastleigh’s 5th Street added, “also the community knows the (communication) channels to use in case such things happen.” Perhaps most significantly, the GoK did not launch another iteration of Operation Usalama Watch, and while individuals were still investigated and arrested, no mass round up of local residents occurred, and no accusations of human rights abuses specifically linked to the post-14 Riverside attack have emerged.66 As discussed above, respondents credited additional factors for the improved response:  Better investigative skills and a greater interest in carrying out investigations prior to making arrests.  The non-Somali origins of the attackers demonstrated that a mass round-up of Somali residents was not an appropriate response.  Somali residents lost friends and family members in the attack, which showed the populace that Somalis can be victims too. 22.2 Community Response In the aftermath of both the Westgate and 14 Riverside attacks, Eastleigh residents waited in fear of a heavy-handed security crackdown. While security agents stormed in and indiscriminately rounded up residents following Westgate, no such crackdown occurred in the aftermath of 14 Riverside, and instead, the community proactively mobilized to show solidarity with the victims and take an outspoken stand against extremism. Perhaps in anticipation of another show of force similar to what they had experienced five years prior, community leaders chose to preempt such actions and shift the narrative around the Eastleigh community and its purported support for extremism. While this research cannot conclusively argue that the community’s response in the aftermath of the 14 Riverside attack successfully shifted the narrative, because of myriad other factors at play, including changes in the security forces’ modus operandi, it does suggest a greater willingness – regardless of the motivation – to articulate a more proactive anti-extremism viewpoint. This section compares the community’s response in the aftermath of both attacks. 66 Extrajudicial killings continue to affect residents in Eastleigh. This report will dive into this topic in subsequent sections. 223 22.2.1 Community mobilization following the Westgate attack By the time of the Westgate attack, the government had established a precedent of violent security operations after terror attacks, largely targeting Kenya’s ethnic Somali population, who comprise the majority of residents in Eastleigh. High levels of tension resulted within the Eastleigh community. The heightened sensitivity within the community came from the narrative that associated Eastleigh and the Somali community with AS. Leading up to and after the events at Westgate, Northeastern Kenya experienced an upsurge in extremist activity, which often appeared to target non-Muslims in the region, and in particular, non-locals working in professional positions, such as health care and education67, strengthening anti-Somali rhetoric among the wider Kenyan population.68 With Assistant Minister Ojode’s pronouncement of AS’s ‘head in Eastleigh and tail in Somalia,’ the government effectively legitimized this narrative and reaffirmed the profile of a young Somali Muslim man as the face of AS in the collective public conscious. Respondents across the board expressed a sense of frustration over the way in which the GoK and Kenyan populace had come to view the Eastleigh community, largely reinforced by the heavy-handed response after Westgate. According to an EBC representative: “The police did not consider that Somalis are affected more by the Al-Shabaab both here and in Somalia…When Westgate happened, we were also affected here, because we are also Kenyans.” The EBC took proactive action at this time to get ahead of the issue and show support for the security agents and Kenya at large. In acknowledging that statements made by Assistant Minister Ojode worsened the situation for Eastleigh residents, strengthening the anti-Eastleigh and anti-Somali narrative, the EBC was keen to show solidarity with and support for the troops sent in to secure Westage Mall. An EBC representative said that during the siege, members of the organization showed up at the mall with water and food “to show everyone that we were together.” Despite this show of unity from the EBC, there was no immediate or overt pro-peace and anti￾extremism response across the community; in the days following the attack, respondents stressed the heightened sense of fear and the heavy-handed security response in Eastleigh. It was only after Operation Usalama Watch kicked in almost half a year later, in April of 2014, that the EBC supported a protest led by the Eastleigh Residents Community Association (ERECA) condemning the police ‘crackdown on Somalis.’ 69 A news report suggested that “every person who owns a shop here contributed Sh5,000 to help us hold the demonstration.”70 The EBC and Kamukunji’s MP Yusuf Hassan Abdi, who has long demonstrated proactive CVE engagement, supported ERECA in this process. The timing of events is important to consider. The show of solidarity only emerged after government forces cracked down in Eastleigh, perhaps in an effort to sway notions of widespread support for extremism amongst the Eastleigh community; it was not a proactive stand in support of the victims. Although the EBC attempted to show solidarity, they did not do so through the Eastleigh community, but rather through outward support to the security forces. The procession one year following the 67 “North-Eastern Kenya: A prospective Analysis” Iris-France. December 2015. See: https://www.iris-france.org/wp￾content/uploads/2016/11/ENG-Observatoire-Prospective-Huma-NORTHEasternKenya-12-2015.pdf 68 Abdille, A. “The hidden cost of Al-Shabaab’s campaign in North-eastern Kenya.” International Crisis Group. 4 Apr 2019. See: https://www.crisisgroup.org/africa/horn-africa/kenya/hidden-cost-al-shabaabs-campaign-north-eastern-kenya 69 Muthoni, S. “City residents hold peaceful demo, condemn agents of terror.” The Standard. 22 Sept 2014. See: https://www.standardmedia.co.ke/article/2000135723/city-residents-hold-peaceful-demo-condemn-agents-of-terror. 1.1 70 “Eastleigh businesses to shut down over crackdown.” Nairobi News. 11 April 2014. See: https://nairobinews.nation.co.ke/news/eastleigh-businesses-to-shut-down-over-crackdown 224 attack was an attempt to bring in the community, but it did not result in any unified or cohesive response. 22.2.2 Community mobilization following the 14 Riverside attack Because of the community’s experience after the siege on the Westgate Mall, fear was rife in the days following the 14 Riverside attack, and the community was on edge: “It was peaceful here, but you had to stay alert in case something happened,” a local male youth reported. The Eastleigh community organized a peace march to condemn terrorism and show their solidarity with victims of the attack, and the EBC led the process with the support of MP Yusuf Hassan Abdi. According to a representative from the EBC, all malls and major shops in the area decided to shut down their businesses for the day: “We decided to lose business but be with Kenyans. We did this with the support of the MP, who we always consult and is always with us. With his input, we did the solidarity march together.” Despite the EBC’s leadership in organizing and mobilizing for the processions, some respondents felt that the march did not reflect the diverse character of the community living in Eastleigh. According to them, the organization of the march was an elite-driven process designed to safeguard the business and security interests of the Somali community. A respondent said: “The demonstrations were not necessarily about the grief in the community… It was because they were afraid. They feared that law enforcement would target and harass them…” A significant degree of speculation around the business community’s motivations emerged from different types of respondents, with most agreeing that there was particular concern to maintain peace and security in order to protect business interests in Eastleigh. One respondent suggested that the members of the EBC who reside outside of the greater Eastleigh area were mostly responsible for organizing the processions, a speculation that contributes to respondents’ suspicious about the different interests and motivations of stakeholder groups in the community (which will be explored in more depth in the following section). He believes that these individuals do not share in the same lived experience as others in the community, and therefore do not have a deep enough understanding of the challenges and concerns of the wider community. A representative of a CBO in Eastleigh observed that the procession identified with only certain segments of the community: “Yes, it was called the one hour of silence for the elites. To say the truth, it was organized by the Eastleigh Business Community to show solidarity fearing that their businesses would be affected in case of police raids… We have wealthy people in this community who are politically shrewd. There are tribal dynamics in how Eastleigh operates. We have the elite Somalis and the other normal people. So probably if I was to conduct my own procession, the community will participate but the elites would not. I would not have the influence to make the shops close. But there are those who have that kind of influence with the shop owners.” The demonstrations after the 14 Riverside attack therefore appeared to be a strategic move aimed at insulating and protecting Eastleigh from a possible heavy-handed securitized response; the organizers were allegedly nervous about how a portrayal of Eastleigh as a ‘security operation’ zone or a hotbed of terror would negatively affect business in the area. An EBC representative spoke of their mandate of “looking after their business interests” and “working for the interests of businesspeople in Eastleigh.” Perhaps in an attempt to more accurately reflect the interests of the wider community, local CBOs also organized processions within their immediate communities in solidarity with victims of the attack. However, claims emerged of insufficient GoK support for the CBO-led initiatives as compared to the EBC-led march, likely because of the EBC’s influence with local government representatives. This influence enabled the EBC to leverage the support required to carry out a major event. A respondent who took part in both marches said: 225 “As compared to the march by the Eastleigh Business Community, the one we organized with some CBOs was not that successful. It lacked media coverage and the chief tried stopping us, though we spoke to the Deputy County Commissioner on phone, who told us to proceed.” In contrast to the Westgate response, the Eastleigh community – and the EBC in particular – perhaps learned that a more proactive show of support to the victims alongside a public anti-extremism narrative would effectively dissuade security forces from an indiscriminate crackdown. However, other factors may have been at play. Some respondents suggested that the different profiles of the perpetrators and victims likely contributed to a more measured public reaction after the 14 Riverside attack. A (non-Muslim) Majengo-based youth leader shared his belief that the events of 14 Riverside further shifted public opinion from thinking Muslims are the only ones susceptible to VE influence to understanding that radicalized individuals are capable of such action, regardless of their background. This is a very subtle but important distinction; that is, the profile of the attackers may have directly challenged narratives held by some Kenyans that VE is a Somali-only issue in Kenya. A community leader suggested: “Here in Eastleigh and Kamkunji, there is a perception that once you talk about violent extremism, it is the Somalis or Muslims who are affected, but now people are coming out saying that this thing is not just for the Muslims alone, because based on the Dusit attack we saw that there was a mixture of tribes71, Kisiis and Kikuyus and so on. So, people’s thinking has changed, and now they see that radicalization can happen to anyone.” Another community leader commenting on the collective mood of his community noted: “After the Dusit attack we realized, as Kenyans, that this was an attack on all of us. Even The Honorable Uhuru Kenyatta said that our Muslims brothers lost their lives in that saga. The terrorists were shooting at everybody.” 22.2.3 Comparative analysis of community mobilization The community mobilization after both incidents differed significantly in the Eastleigh community. While the Westgate procession occurred only after, and perhaps in response to, Operation Usalama Watch, the EBC led the charge on a more immediate solidarity demonstration post-14 Riverside. In reflecting on the two solidarity and peace marches, respondents offered mixed views on their effectiveness, and in particular, the motivations of the different individuals involved in organizing and participating in the marches. An aide to a local politician also suggested that ordinary residents (not members of the EBC) do not have access to resources to mobilize the wider community on their own. “The community cannot organize itself. They must find someone who brings them together.” Statements from the EBC representative supported this notion. He acknowledged the EBC mobilized individuals to participate through the organization’s existing networks. Those not networked in therefore were not necessarily reached. Some respondents even suggested that hand-outs motivated residents to participate in such processes. A government representative agreed, suggesting that incentives play a key role in getting people to mobilize, and perhaps those who are most active, are not those who are most passionate: 71 Interestingly, numerous interview respondents conflated religion and ethnicity as identity markers when describing Somalis. A number of respondents (all non-Muslims) would use ‘Somali’ and ‘Muslim’ interchangeably only making a distinction when context required. This speaks to the complexity and plurality of Somali identity categories in a post-colonial state like Kenya. Afyare Abdi Elmi (2010), citing Ali Abdirhman Hersi (1977), also notes that the notion of Somali identity is inextricably tied to Islam as a religion and a system of values. He demonstrates this by giving a historical account of Islam in Somalia. It is suggested that the Horn of Africa has been practicing Islam for a better part of the last 13 centuries, dating back to a few decades after the inception of Islam in Arabia. 226 “You see, in Kamkunji, if you want something to happen you must be able to provide resources to the people who will make that thing happen. It is not like everyone who will attend the rally will be compensated; only 3/4 of them will be compensated. You must compensate the people who will bring the crowd…to come for the rallies. It's just the resources to make things move.” Even if the processions after both attacks were organized to strategically protect certain interests, the wider community did still actively participate in and support the processions, and perhaps more so after 14 Riverside than after Westgate. This suggests that even if the motivations for organizing or participating in these processions are more strategic in nature, the willingness to do so is ultimately what matters most, and the events can be seen as a success in mobilizing the community around a unified, anti-extremism message even if individuals might be inclined to participate for other reasons. However, varying motivations for leading or participating in these marches have implications on the sustainability of a strong and cohesive anti-extremism response. Activities that are designed without considering the interests of different stakeholder groups are unlikely to be sustainable if different sets of stakeholders are motivated for different reasons. There is therefore a need to gain a deeper understanding of those motivations and goals in order to better leverage support for and involvement in CVE programming, while also bearing in mind that other factors could affect willingness to participate in programming. At the same time, diverging interests, experiences and perspectives within the community itself suggest that perhaps the community is not as cohesive and united as the marches might lead outsiders to believe. Uncovering these different experiences and perspectives could help to identify entry points to engaging with the community and creating more representative engagement in future CVE initiatives. These findings are unpacked further in the following sections. 227 23. ANNEX XIII- Q2-2020 - QUARTERLY RESEARCH REPORT: UNDERSTANDING GENDER-RELATED OUTCOMES OF NIWETU GRANTS, AND A REVIEW OF NIWETU’S GENDER PLAN 23.1 Key Findings The need to be more deliberate about proactively involving women in CVE activities and developing targeted CVE strategies for both men and women has been affirmed by a growing body of evidence that suggests that women have more involvement in VEOs than previously thought; they are recruited through a number of different pathways and take on a variety of roles within the organizations they join. However, identifying any trends or patterns in the recruitment of women into VE has been difficult, and evidence gathered to date has been largely anecdotal, with findings isolated to certain geographic areas. What is clear, however, is that both women and men have unique roles to play in VE as well as CVE, and therefore, CVE activities should bear this in mind when designing and implementing activities designed to reach different demographic groups. In response to this shifting understanding of gender dynamics in CVE, NIWETU began to place a deliberate focus on gender-inclusive programming, recognizing the unique roles that men and women play in both VE and therefore CVE. After the TO8 research, NIWETU issued a call for proposals for CVE activities specifically addressing gender issues, while all other activities were meant to utilize newly developed tools, such as the Gender Scorecard, in the design process to ensure consideration of gender dynamics. The findings presented below relate to activities that were designed after the launch of the Gender Plan and associated scorecard documents. First, we present outcomes identified that specifically relate to gender considerations in CVE. A few general outcomes were also identified in this research that are discussed briefly at the end of this section. 23.1.1 Gender-Related Outcomes At a most basic level, in communities where women were often sidelined, they are now increasingly involved in CVE activities. It was noted in all three counties, but highlighted mostly in Isiolo, that by creating a ‘women’ pillar within the County Action Plan (CAP), the importance of involving women has been institutionalized. Women have been seen attending football matches, trainings and participating in conversations about VE, with much greater visibility than in the past. This has resulted in women opening up about issues affecting them and their families, being more vocal in sharing information on suspected radicalization and recruitment, or disappearances of their children, as well as on what they learned through the CVE activities in which they participated with their neighbors and family. This, according to some, has pushed them out of the role of only acting as caregivers of children at home, as per formative cultural expectations. An Isiolo focus group discussant noted: “In the group we talk. In our Borana culture women were never allowed to talk in front of men... but with the sensitization things are changing. Now, even the men are accepting this new role.” A participant in a Wajir focus group shared: ‘’Women are also more alert and do not want to be hosts or victims of VE. They are now concerned with their children’s whereabouts, and what they are taught.’’ Gender considerations were found to be a deliberate component of activity design in the grants assessed through this research. At a most basic level, NIWETU supported activities sought to 228 include more female participants in discussions around C/VE. CSO respondents pointed to a number of different tactics used to identify and target women for activities, particularly those affected by VE.72 Some tactics used included:  Female staff of the CSO partners reached out to other women in the community about the activities being designed and implemented.  Open consultation spaces, or ‘field desks’ were created where women and youth would come to share issues of harassment by the security forces, which would then be taken up by relevant authorities. Creating an open and safe space was said to have encouraged more engagement of women and youth, whose voices are often sidelined.  Designing and implementing women-friendly activities, such as convening separate meetings for women (as has been done in Garissa and Isiolo), and targeting women’s peer networks, where women feel safer opening up.  Involving female facilitators to ensure that female participants feel comfortable speaking openly about issues affecting them.  Encouraging women to lead on the implementation of an activity, through facilitating a training session using the gender-conscious CVE facilitation guide or managing of the grant for the partner organization.  Ensuring that female staff are present during any counseling sessions to create safe spaces for women to speak on VE issues they experienced.  Planning activities specifically to target women’s participation, such as facilitating discussions focused on women’s security issues in hotspot areas. Shifts in VE recruitment dynamics to also target women reaffirmed the need for partners to target women as well as men. While respondents across the board believe that men are the primary targets of violent extremist organizations (VEOs), there is a growing understanding that women are also beginning to be targeted by VEOs in different ways and for different reasons than men; these methods are still not widely understood, but partners expressed the need to also focus activities on women given shifting recruitment dynamics. In a basic sense, women have become increasingly visible in their communities. In communities where women are often sidelined, they are now increasingly involved in certain community activities, such as attending football matches, trainings and participating in conversations about VE. This has also contributed to women’s willingness to be more vocal about their experiences, as well as to share information with neighbors and other family members. Moreover, women are taking an active role on security matters and are also demanding accountability from the security agencies. According to a NIWETU partner in Isiolo: “For the sporting activities, when we started, we were not sure ladies would participate. But they showed interest, and this made us have more women friendly games.” It is worth noting, however, that women still face limitations in their ability to actively participate in some of these activities, because of strict cultural norms. While women may have been able to attend 72 This outcome demonstrates clear uptake of TO8 research recommendations, which identified deliberate targeting technics as a way that NIWETU could more effectively work toward gender inclusivity in its programming. 229 a football match, participating in a football match may not be possible. As a respondent from the TO23 research on the implementation of the CAPs noted: “We also need to consider the cultural differences of the people in Isiolo especially with the games that we brought on board for participants. For example, the football game was not very appropriate for Muslim women, so we now have other games like sack race.’’ Women’s informal networks are increasingly being utilized to disseminate C/VE information. Women are socially networked in a number of informal spaces, such as in their chama (or savings group), within their religious communities, and with other women and mothers in their neighborhoods, among other ways. In Isiolo, a particularly significant finding was the increased utilization of these networks and informal gatherings for women to share what they learned in their training, or to discuss issues of concern around VE. As respondents from different focus groups in Isiolo noted: “I belong to a women’s group, I told them we have a duty to talk to our children. I shared with them about Al-Shabab, youth disappearances, and employment promises. Women should not put pressure on their children to go look for ways of supporting the family. Protect your home, provide for them. It is by going out to look for how to support the family that they get manipulated.” “After training, it was a Friday, so on Saturday we meet for choir where we have men, women and youth. That was the first meeting where I shared. Most were happy, because they had never heard all that information before. I encouraged all of them to share with others. “Even in Islam we are told that when you learn something you share it. …On one of the Fridays, before we got into prayer, we were cleaning. We were about 50 people, so I informed them, and they appreciated it. I believe they are also doing the same.” Placing women in more visible and prominent roles within an activity helped female participants feel more comfortable to actively participate in discussions and more openly sharing their experiences. It also makes women feel more comfortable participating in the first place, encouraging them to become more active listeners and participants to gain the most out of the activity. This was highlighted in the section above and is worth further discussion. According to one CSO representative from Wajir, “With female trainers, women are at ease to share what they have in their mind. It provides a conducive environment where women were open to give their views on violent extremism.” Another respondent in a Wajir FGD reaffirmed the positive role that female leaders can play in activities: “For example, the lady facilitator of the workshop, Mrs. Ruqiya, has demonstrated to the participants what women can achieve. She had great knowledge and oratory skills to deliver to her audience. She was a testimony to all that women have no limit on what they can contribute in countering violent extremism.’’ This is a simple, yet crucial, step to take, as gender norms in many of these more traditional communities offer little space for women either to serve in leadership positions or to feel comfortable discussing sensitive issues in mixed-gender settings. Participants in an Isiolo focus group discussion noted: “Culturally among the Borana, who are the majority in Isiolo, women are rarely included in activities but some of us are encouraging more women to be included as equal members. For me men and women play a key role and they should be both equally involved.” 230 At the same time, female facilitators were also seen to be effective at reaching children; it is perhaps because of their traditional role as caregivers that they are willing to take appropriate time with children to teach them effectively and support them. As a religious leader noted regarding PGI’s activity in Garissa: “Umusalama is a girls’ school but I am aware that this activity also took part in boys’ school. It’s important to note that the facilitators of the activity were mainly women and they took extra time with the students. We even formed a club at school.” Women are better represented in leadership and decision-making roles within their communities which allows them to promote and represent the experiences of female community members. Their voices are increasingly represented in different decision-making forums and in various structures. While respondents were not able to specify if this change was a direct result of some NIWETU supported activities, this finding may have been influenced by the promotion of female champions on CVE issues within government, such as via the training of female security stakeholders in Isiolo who are now better positioned to represent women’s perspectives and interests in CVE discussions. In general, the research found that female champions of CVE in government and the security sector create links with other women in the community; in Garissa and Wajir, women noted feeling more comfortable reporting to other women and expressing their concerns to female local leaders. Deliberate decisions to separate discussion groups for women and men encourages more active participation for both genders. This finding emerged particularly strongly in Garissa, where traditional cultural norms regarding gender often encourage women to take more passive roles in group or community activities. Instead, by separating men and women during CVE discussions, women more noticeably opened up and were willing to discuss VE issues affecting them. Women acknowledged that when they participated in mixed-gender groups, they often agreed with what their male counterparts said, but they were more hesitant to speak up independently. In separated groups, women were more open to sharing their thoughts with facilitators and other participants. A community leader from Garissa reaffirmed the need to have separate discussions for different genders: “Culturally, women do not speak much when there are men. However, the women who have been involved in CVE issues are proactive especially in Garissa Township… women especially the ones in peace committees and women peace organizations were on the forefront in asking questions on the government response to VE activities.” A CSO representative from Isiolo also concurred that cultural dynamics affect women’s participation in many of these activities: “Among the Borana who are a majority in Isiolo, women are rarely included in activities but some of us are encouraging more women to be included as equal members. For me men and women play a key role and they should be both equally involved.” Activities that speak to women’s and men’s distinct roles within society are better able to disseminate relevant CVE knowledge and skills. While some gender-focused development activities would advocate for trying to shift traditional cultural norms around gender, and the expected roles that men and women should and do play in their communities, it is also important to utilize these traditional norms when trying to effectively disseminate information and share new skills. NIWETU did so effectively, by supporting activities that speak to women’s roles and positions within the family and household life and community at large. According to two focus group discussants who participated in the Pastoralist Girl’s Initiative activity in Garissa, 231 “Women are more open to sharing information and likely to have more impact in the community starting from the household level…” “…if mothers learn about early warning signs of VE, they can detect radicalization in their children.” Some respondents also openly challenged the traditional expectations of men and women in more vocal ways than they would have in the past. A focus group discussant who participated in the gender￾focused grant to the Isiolo County Commissioner’s office suggested that even while women are speaking out on these issues more than they have done in the past, men are also coming out to support women in protecting their children: “The men now have to come out and support women to protect our children’s future… Women also have ended up participating and speaking about the issues affecting them. It is not like before where our faces were covered because of our culture. We have come out to be heard…” Both mothers and fathers are beginning to understand their role as parents in providing positive influence and more support to their children, and how they are instrumental in stopping young people from joining VE. Some NIWETU-supported activities, such as the Isiolo Women of Faith grant and a component of an IPL activity, have specifically focused on reaching families affected by VE, through training, counseling and strengthening peer support. Some respondents believe that this particular activity has had a ripple effect on the community at large wherewith both male and female community members are taking an active role in raising the children and ensuring that both parents know their whereabouts, however this was not possible to verify and is likely the result of a number of compounding variables. Other than having curative measures, the community has also been sensitized on preventive measures to VE. A participant in a focus group in Isiolo stated: “The men now have to come out and support women to protect our children’s future… Women also have ended up participating and speaking about the issues affecting them. It is not like before where our faces were covered because of our culture. We have come out to be heard.” “We have started to understand the causes of VE and our role as parents and older members of the community. We were surprised to learn that children of the rich are also joining the terror groups hence it is not a poor man’s problem.” This finding suggests that activities specifically focused on the family unit are effective in reaching both men and women in their roles as parents and spouses, as they are given skills and tools that are accessible and relatable to their daily lives. Activities that targeted women with skills and knowledge have a cascading effect to other women in the community. This finding emerged particularly strongly in Isiolo, borne out of the grant to the Isiolo County Commissioner’s office to equip female security officers and representatives with CVE skills and knowledge. Training female security agents who then work and mingle within the community played an important role of reaching out to women in the community. And as discussed above, women discuss issues of concern freely in their chamas or in other gatherings, meaning that skills and knowledge taught in a training are likely to be spread more widely across women’s networks. Some respondents also suggested that the different ways in which VE issues affect men and women suggest the need for different approaches to supporting both genders that are more relatable to the issues and challenges they face. The recognition of this issue by activity participants is an important finding, because it suggests that participants have gained a more nuanced understanding of VE issues and how to respond to them. Participants in an FGD in Isiolo gave the example of young men, who have a particularly hostile relationship with the security actors in the area, so activities seeking to reach that demographic should focus on trusted and sustainable relationship building with security forces. 232 Creating safe spaces and guidelines for women and mothers who are communicating with their sons provides more intelligence and information to security actors. Whereas individuals who report suspicious activity or information they have are typically then targeted and harassed by security agents, a noticeable change has occurred in which safe spaces for sharing this information are being created. Mothers often communicate with their sons, even those who have disappeared, and they tend to have information on whereabouts, reasons for leaving, amongst other things, which is useful to security agencies trying to understand recruitment pathways and find a solution to returnee issues. The creation of these new mechanisms which encourage community members writ large, but women in particular, to report useful information will both improve community-security relationships and will facilitate information sharing to provide security agencies with relevant information to better address the issues. The way in which this finding appears to be borne out of NIWETU-supported activities is that women who have participated in CVE trainings learned about who to report to when they have information to share, and what information is worth sharing. An FGD from Isiolo touched on this point: “I have seen now better relations between police and the community. Once in a meeting the affected women went to see the security officers and they said they wanted to be allowed to pick calls which started with +252 (calls from Somalia). In that meeting the women were told they were free to speak to their children who had gone missing, but they needed to report. Previously anyone who picked the calls would be arrested.” In Garissa, women now report feeling more comfortable sharing information with their local chief when they see or hear something suspicious, but when they were not comfortable going directly to the chief, they go to other female leaders who will then pass the information to the chief. They noted that the most trusted networks for reporting are the Nyumba Kumi and peace committee members. Perceptions are beginning to change around the role of women within security agencies. This finding was most notable in Isiolo, likely because the Isiolo CC’s activity targeted female security stakeholders. According to participants of that activity, female officers in the past had not been taken very seriously, now, they note, there is a gender desk and more officers are being trained on gender and CVE. Working closely with female members of community policing structures, the perceptions of women as security actors are beginning to change. Similarly, in Garissa, deliberate efforts to involve women in security related conversations were made. This is notable, because Garissa is a context where culturally, security is run by men at the community and governmental levels. During ROAD’s mobilization, they emphasized to the heads of the peace committees that women must be involved in the meetings but noted where it was better to have separate meetings for men and women because of cultural barriers, such as in Hulugho. This decision was taken after the partner realized that women were not speaking as comfortably when men were present. Outcomes around empowering women and strengthening their voices in CVE discussions emerged most strongly in Isiolo county. While this research identified positive outcomes across the counties of research, the most notable changes appear to have occurred in Isiolo county, where in general, women speak to feeling more empowered to report issues of concern, share what they are learning with others, and pay attention to specific behavioral changes in their children. While outcomes were still positive in Garissa and Wajir, women tended to feel more comfortable sharing and discussing amongst themselves, and mostly engaging with a few trusted leaders to share information or report issues. This could perhaps be due to the fact that Garissa and Wajir counties are slightly more traditional and conservative around gender roles in society, while Isiolo – and in particular Isiolo town – is more diverse ethnically and religiously, lending to a possible shift in cultural norms. This is speculative, however. Encouraging active participation of women in NIWETU-supported activities resulted in a number of secondary outcomes positively affecting the wider community. 233  Mothers learned to identify any behavioral changes in their children and raise alarm as needed. In a society where women are the primary caregivers to children, equipping them with skills to identify potentially suspicious behavioral changes is key to stopping radicalization and recruitment. According to female participants in a focus group in Isiolo: “Our pastoralist’s communities are polygamous. The men have other families and children, but women have only one family and that makes our children our main responsibility…”  Women often gather in groups, either for work or for social connection. Therefore, it is easier and faster to pass on messages through women especially after a training. 23.1.2 Additional Findings In addition to the gender-specific outcomes identified in this research, respondents spoke to a number of more general outcomes that are worth highlighting as well: Community leaders in all counties believe that relationship between the community and the security personnel have improved; more VE cases are currently being reported without fear of victimization, and parents are making a more deliberate effort to know the whereabouts of their children. Generally speaking, during focus group discussions with activity participants, respondents affirmed that NIWETU-supported activities have positively impacted the community in areas prone to VE. Specifically, a consensus emerged amongst respondents in all counties of research that the interaction between the police and the community has improved. This occurred largely through participation in joint trainings and dialogues between community groups and the security sector, as well as through less structured activities, such as football matches between youth and the police. In Isiolo, respondents highlighted the important of the County Engagement Forum (CEF) in becoming a platform through which community members feel represented in discussions with government on CVE issues. Dialogues, training sessions, forums and counseling sessions among community members and youth affected by VE, have empowered community members to openly discuss C/VE without fear. Nearly all respondents agreed that they can now discuss sensitive issues around C/VE more openly than before, including in instances in which they feel unjustly treated by security personnel. This has cut across various social divides, with both men and women feeling empowered to speak more openly and honestly and actively address issues and concerns as they emerge. In Garissa the inclusion of teachers, students and Boards of Management in CVE trainings resulted in the formation of peace clubs in schools that now aim to train and sensitize students and others within the educational system on VE and CVE. At the tertiary education level, university students formed WhatsApp groups where they share and educate others on CVE and how to prevent VE in their communities. Participants felt that education plays a key role in the community and its inclusion of education players will play a central role in the fight against VE as education players can easily share the knowledge received. (These findings emerged specifically as a result of the PGI activity.) Communities are beginning to understand the collective responsibility around vigilance and reporting suspicious cases or incidents of extremism. Perhaps as a result of the strengthening and stabilizing of relationships between communities and security agents, there is more openness in sharing information and discussing the issues without fear. There is a sense that in order to protect the country, Kenyans must be forthcoming with what they know or learn about an individual being recruited into VE, even at their own expense. According to a focus group participant in Isiolo: 234 “We have learnt we are Kenyan first and if my child goes missing, however painful it can be, the country comes first and I must report …VE has no gender, no religion and no age. Anyone can be targeted and affected.” 23.2 Recommendations from Component 1 The analysis leads to the following recommendations for strengthening gender-specific programming and ensuring uptake of good practices and lessons learned. 1. Partners should learn to think about gender considerations in activity design. Some activities, such as football matches, exclude women from participating because of traditional cultural norms. While these activities might specifically seek to target male youth, discussions should be had on who is excluded from these activities and why, and if that is an acceptable choice to make given the intended outcomes or goals of the activity. 2. Leverage women’s informal networks in designing programming specifically targeting women. Women traditionally gather in their chamas, religious groups, or informally within the neighborhood. These are effective networks to try to leverage in order to disseminate messages to women, many of whom are likely more reluctant to engage in a formal program on a sensitive topic such as CVE. 3. Ensure activities targeting affected families are gender disaggregated and involve knowledgeable individuals who can employ sensitive approaches in dealing with the two separate groups. 4. Utilize more female facilitators in activities targeting women as well as in mixed￾gender groups. Female facilitators will make female participants feel more comfortable speaking openly and sharing their views in both women-only and mixed gender groupings. 5. Continuously engage and support partners in interrogating the importance of gender dynamics within the communities. By seeking to continuously update their knowledge on local VE dynamics as they relate to both men and women, partners will be better placed to design and implement activities that respond specifically to the context. 6. Work with partners to deepen their understanding of the concept of ‘gender’. The interviews highlighted that while a more conscious effort is being made to target women in effective and relevant ways, partners view the concept of gender to primarily focus on women, ignoring the fact that men should also be targeted in their roles as men within the family and wider community. Helping partners to understand this nuance will further strengthen how CVE programs seek to address gender considerations in their work. 235 24. ANNEX IV- Q3-2018 - QUARTERLY RESEARCH REPORT: THE COUNTY ACTION PLAN FORMATION PROCESS: ISIOLO, GARISSA AND MOMBASA 24.1 Overview of the Research Research Data Table 2 summarizes the nature of the data examined by the Q1-2019 Research, as set out by the CAP Cycle Framework. Table 1: Key 1. Very low focus on this type of data 2. Low focus on this type of data 3. Moderate focus on this type of data 4. Strong focus on this type of data 5. Very strong focus on this type of data Table 2: Q2 Research Data Quarterly Research Tracker NIWETU INTERVENTION RESEARCH METHODS COUNTIES 1.2 – Sub-Grants 1. KII / FGD 1. Nairobi 1.3 – Capacity Building 2. Quantitative Survey 2. Garissa 1.4 – Community Champions 3. Most Significant Change 3. Wajir 1.5 – Strategic Communications 4. Process Monitoring of Impact 4. Isiolo 1.6 – Rapid Response 5. Outcome Harvesting 5. Mandera 2.1 – Research Products 6. Participatory Systemic Inquiry 6. Kilifi 2.2 – TA to County Govt 7. Stakeholder Feedback 7. Kwale 2.3 – TA to National Govt 8. Network / Outcome Mapping 8. Lamu 9. Other 9. Tana River 10. Mombasa VE CONTEXT CVE & RESILIENCE CONTEXT INTERVENTION CONTEXT 1. Incidents of VE 1. Other CVE initiatives 1. Theory of Change 2. Perceptions of VE 2. CVE leaders & influencers 2. Intended outcomes 3. Macro / push factors 3. Inter & intra group relations 3. Unintended outcomes 4. Individual pull factors 4. Economic capacity 4. Causes / attribution 5. At-Risk Groups 5. Strength of civil society 5. Actors 6. Pathways to VE 6. Communication channels 6. Gender 7. VE actors & influencers 7. Community competence 7. Boundaries 8. VE ideas & narratives 8. County government 8. Relationships 9. VE & wider links 9. Security actors 9. Perspectives 10. Community support 10. National government 10. Other 11. Financing 11. Inter-governmental 12. Reintegration 12. Other 13. Institutions 24.1.1Research Context and Objectives In September 2018, NIWETU participated in the launch of the Isiolo CAP, completing the first of four CAP processes currently funded by NIWETU. The process was largely hailed as a success by the 236 county government, national government counterparts, and other local-level stakeholders. At the time of research, processes in Garissa and Wajir were ongoing, but have since concluded, while Mandera CAP process is still underway. In Nairobi, initial conversations with different stakeholders are underway in advance of a larger stakeholders’ forum that will discuss and agree upon a process. The research presents an investigation into these processes to draw out lessons learned, what has or has not worked and why, and will pull together good practices that can be applied to future CAP processes in other counties in Kenya, including Nairobi. To that end, research also took place in Mombasa. Mombasa is a non-NIWETU county, that during the formation of its CAP from a variety of donor programs. Mombasa saw a successful CAP launch and ongoing proactive engagement in CAP related planning and activities over the past year. 24.1.2Areas of Inquiry The below table summarizes the lines of inquiry for this research initiative: 1. Brief overview of the status of CAP development across Kenya  Finalized CAPs and updates on status of implementation  Overview CAP processes currently underway  Forthcoming processes  Alignment with the National Strategy for Countering Violent Extremism (NSCVE) 2. What do the different CAP development processes look like?  Overview of the counties of research and operating contexts  Comparative analysis of the different approaches to developing the CAPs in the research areas  Analysis of stakeholder engagement  Analysis of coordination between county and national government 3. What is/is not working well in the development of the CAPs and why?  Identification of positive outcomes from the processes  Analysis of differences across counties, with consideration of differences in the contexts  Analysis of factors contributing to success73  Analysis of degree of broader community engagement, particularly with marginalized and at-risk groups 4. What are the implications for NIWETU, US Government and Government of Kenya engagement?  Identification of lessons learned and good practices  Opportunities to adjust processes  Ways to strengthen networks between stakeholders  Considerations of approach and management of risks  Recommendations for approaches to implementation of the CAPs based on lessons learned from the development processes 237 24.1.3Research Methods This research employed an approach that relies both on semi-structured and unstructured interviews. Recognizing that the personality of key individuals involved in the CAP formation process significantly influences relationships and therefore perceptions of inclusivity and transparency, a less structured research approach helped to pull out nuanced differences in the personality of these individuals and how that affected the processes in each county. Semi-structured interviews drew out lessons and good practices in the overall process to trends and patterns within and across groups. 24.1.3.1Respondent Sampling Research specifically targeted individuals who have participated in CAP-related stakeholder engagement workshops and other elements within the process. A full list of the respondents from every county detailing their position or role can be found in Annex 1. Generally though, respondents included representatives from the following categories in Isiolo, Garissa and Mombasa:  National government officials  County government officials  Local civil society, including local community-based and faith-based organizations  Youth representatives  Women’s representatives  Religious leaders  Community leaders The research teams obtained interviews with between 20-22 respondents in each county, with close to equal numbers of male and female respondents. In addition, interviews with three stakeholders in Nairobi, in order to gain a cross-cutting perspective of the different county CAPs: with the Deputy Secretary for the Ministry of Interior and Security; the NCTC’s Head of Devolution; and a senior consultant working for Malaika Foundation, a Nairobi-based organization providing technical advisory during the CAP formation in the different counties. In total, 65 individuals were interviewed for this research. 24.1.3.2Research Tool For the interviews, researchers were given a set of Key Assessment Areas to guide their discussions with respondents. These contained sample questions relating to the areas of inquiry, but it was stressed to the researchers that the interviews did not need to follow a particular order or structure and that it was up to their discretion to use the sample questions or introduce their own questions as they saw fit, in order to encourage the flow of conversation and get to more complex issues around, for instance, personality politics. 24.1.3.3Research Locations Research for this activity took place in Isiolo, Garissa, and Mombasa in October-November 2018. In each county, the research centered on the county capital town or city, though in Isiolo, the sub-county of Garbatulla was also visited by the researchers. 24.1.3.4Limitations The below limitations arose during this fieldwork. The research approach and analysis sought to mitigate challenges and limitations that arose to the extent possible. 1. This research draws heavily on perceptions and opinions of process, inclusivity, and approach to the design and launch of the CAPs. As such, the degree to which the findings can be extrapolated to participants and stakeholder groups not interviewed for this research is somewhat limited. The research sought to reach a sufficient number of respondents within 238 various categories to triangulate perceptions, yet views are not representative of at-risk individuals or communities, for example. 2. Desirability bias was observed, in that many respondents saw the researchers as being directly linked to funding providers such as NIWETU and BRICS, and therefore at times their responses appeared skewed in ways that they felt could invite further funding. During the analysis, researchers noted particular instances in which desirability bias was noticed, and these considerations factored into the overall analysis. 3. Due to time and resource constraints, the capacity of the research team to interview members of wider community in each county, as well as actors in Nairobi, was somewhat limited. As such, the lines of inquiry for this research were limited to particular issues and questions targeted to more knowledgeable stakeholders. 25. FUTURE CAPS: TEN ESSENTIAL INGREDIENTS This report contains two key sections: this first section distills a number of ‘key ingredients’ going forward for the CAPs in future, while the second section unpacks the research findings, broken down as they relate to each stakeholder group in the three research counties. The research identified a number of key themes and lessons learned from the CAP formation processes under review. Specifically, findings around activities and actors that positively contributed to the process emerged, while at the same time the research identified a number of key areas for improvement in the development of future CAPs in Kenya. Ten key lessons that coalesced around the following, largely-cross-cutting and process-oriented areas that are applicable to future CAP processes are outlined below: 1. Developing legislative frameworks: The Mombasa CAP (MCAP) is the furthest along of the three counties in this research, offering several important lessons for the end-to-end process. While the MCAP is commonly perceived to be the most ‘successful’ of the existing CAPs, still, in the period since its May 2017 launch, implementation has not yet begun. This is because, although the county government pledged 100 million Kenyan Shilling (KES) for implementation, the legislation at the county level allowing funds from Mombasa’s County Integrated Development Plan (CIDP) to be channeled to the MCAP was not in place. The process for developing this legislation began after the launch, has been slow and complex, and at the time of writing was still not complete. For more recent and future CAPs, the process of developing appropriate legislative frameworks to permit CIDP funding towards implementation should begin proactively at the outset of the CAP formation process, in order for the subsequent activities to begin swiftly after the launch and ensure momentum is not lost. 2. Political will and buy-in: The deterministic nature of political will in the CAP formation and implementation processes cannot be underestimated. The personal and political leanings of key leaders, particularly within county governments, can be a deal-maker or a deal-breaker. Isiolo is an exemplary case: Upon finalization of the Isiolo CAP (ICAP), respondents in Isiolo still expressed doubt and concern over the extent to which the county government would take up its implementation. Respondents felt that, in the absence of a major VE attack in Isiolo so far, the county government leadership saw other issues such as the LAPSSET74 pipeline and 2022 elections as more politically expedient than CVE. This finding is significant when one considers President Uhuru Kenyatta’s announcement that all 47 Kenyan counties will be expected to develop a CAP. The vast majority of those counties have 74 Lamu Port-South Sudan-Ethiopia Transport Corridor, a planned oil pipeline transecting northern Kenya. 239 not yet experienced a significant VE incident, and therefore we can assume that, as in Isiolo, the county governments may not treat CVE as a priority. For future CAPs, especially those not currently regarded as VE ‘hotspots’, finding ways to build political will towards CVE will be essential. Crucially, NCTC and other partners should reframe discussions around VE from focusing primarily on VE attacks to VE recruitment. As has become especially clear since the January 2019 14 Riverside attack, Al-Shabaab increasingly recruits its militants from Kenya’s non-‘hotspot’ counties and communities. Couching VE discussions with county governments in this context should encourage greater political traction and ownership. Supporting this, CVE partners should provide resources to conduct in-depth research in each county on local VE recruitment patterns, conducted in advance of the commencement of any formation processes. 3. Identifying CAP Champions: The significance of personalities and the value of social/political capital in developing and implementing the CAPs applies beyond the government to all stakeholder groups. In particular, for the leading civil society organization (CSO) partner(s) in each county, it is essential that, in addition to being effective convenors and providing technical support, they possess the social, personal, and political capital to maintain positive influence and traction within the national and county governments, as well as amongst other stakeholders. This is a tall order for any organization; nonetheless, development partners and the government should be deliberate about identifying and supporting such actors and organizations in each county with the capacity to do so. 4. Managing expectations: Consistently throughout the three CAP processes, the need for upfront and firm expectation management for all stakeholder groups emerged, regarding the extent and nature of participation for those stakeholders in the process, and the respective responsibilities toward the CAPs going forward. This is particularly true of CSOs in the counties, where an apparent lack of expectation management by the national and county governments and the lead partners meant that the wider net of CSOs involved were disappointed or dissatisfied at certain points in the process. At a basic level, making it clear that there are limits to the total number of CSO members that can be physically present in the consultation meetings, as well as the nature of their role – that they will be expected to actively contribute to the CAP content but may not be given official ownership of (and logos on) the final document – should help to mitigate resentment and competitiveness down the line. Likewise, it should be stated from the outset that they should not expect monetary or other material gains or incentives from their participation. Making all potential CSO participants aware of this in advance of their participation should help ensure that the right organizations choose to participate for the right reasons. Equally important is expectation management regarding who will be responsible for implementing the CAPs, made clear from the earliest stages of the process. Under the NSCVE, county governments are principally responsible for funding and overseeing implementation. However, in all three counties in this research, an ongoing lack of clarity or misunderstanding over which parties should be responsible for providing the resources and assuming responsibility for implementing the CAPs – with respondents’ suggestions ranging from NCTC, the County Commissioners’ offices, the county governments, and the lead CSO partner in the county, to NIWETU, BRICS and USAID - has fed into a level of confusion among stakeholders and resulted in delays to progress. 5. High-quality facilitation: Closely related to the previous point, is the emerging importance of high-quality facilitation of CAP formation meetings and consultations. Of the CSOs in Isiolo, Mombasa and Garissa who participated in the CAP formation, to varying extents, it became clear that many did not take away a clear sense of the purpose or content of the CAPs, and presumably therefore may not have fully contributed in a constructive way during those meetings. This trend raises questions about the level of clarity and structure of consultation meetings and predicts a 240 lack of understanding of the purpose and mandate of the CAPs down the line. Going forward, partner organizations such as Malaika Foundation and county-level lead partners, should work on ensuring high-quality and well-structured facilitation of meetings and events. This will help to streamline and better capture stakeholder inputs into the CAP documents and encourage their technical understanding and buy-in for the implementation phase. 6. Engaging security forces: In Mombasa, the involvement of certain actors from the security services was praised as a positive input into the MCAP. Meanwhile in Isiolo and Garissa, respondents from a variety of different stakeholder groups complained that security sector participation was limited. This points to an important finding: In contexts of poor community￾security force relations, there is an appetite for more security inclusion, not less. At the same time, it should be recognized that the success of doing so is heavily contingent on the particular personalities of security actors brought to the table. It is essential to bring different sectors of the security services into the process to encourage their buy-in, including Anti-Terror Police Unit (ATPU), National Police Service (NPS) and Directorate of Criminal Investigations (DCI), but this must be thoughtful and intentional; filling a room with security actors is still likely to put off other stakeholders and prove counterproductive, whereas carefully identifying particular individuals or sub-departments who are well positioned to engage could yield positive results. Development partners and government could begin by consulting the County Security and Intelligence Committees (CSIC) for guidance, and then facilitate introductory informal meetings between community leaders and security actors at the beginning of the CAP processes. 7. Enhanced community inclusivity: A greater emphasis on community consultation is necessary and should be treated as a priority going forward. More community consultation events need to be conducted outside of the main county towns, in more peripheral or remote locations. Event conveners should be deliberate about including non-CSO affiliated community members, with an emphasis of women, youth and at-risk or marginalized groups. In keeping with the General CAP Guidelines, conveners should seek a balance between ‘social influencers’ (charismatic community members) and ‘gatekeepers’ (CSO members) in the process. For development partners, ensuring that sufficient resources are clearly ring-fenced for broad community consultation is necessary. Community-sensitive design of such events is crucial in order for them to be meaningful. Linked to the prior recommendation, consultation events, particularly those outside the main towns, should be held in local languages and/or have translators on hand, as well as translated versions of associated documents. Subcounty and ward administrators and chiefs should be brought into this process. 8. Concise and practical CAP documents: The Isiolo, Garissa and Mombasa CAPs contain the nine pillars set out in the NSCVE and General CAP Guideline documents, with the addition of three-to-four county-specific pillars each. Beyond a small number of key stakeholders though, knowledge of all of the pillars of each CAP was very limited. The majority of respondents felt that the county-specific pillars should be prioritized over the core nine for implementation, because they were chosen by the county according to their particular needs. This begs the question of whether using the nine core pillars plus the county-specific pillars creates documents that are simply too long and unwieldy, which risks spreading scarce resources for implementation and the attention and focus of key actors far too thinly. Revisiting the NSCVE and General CAP Guidelines, with the view that the nine core pillars could be reduced or even dropped in order to focus on the pillars that are likely to gain more traction in each county, could be considered by NCTC and partners. Relatedly, the existing CAP documents themselves are very long because they contain large quantities of heavily theoretical or overly academic frameworks for analysis of VE and CVE. Malaika Foundation’s inputs have contributed to this theory-heavy approach. While it is important to 241 ground CVE thinking in an underlying theory, a more practical approach would be appropriate going forward for the CAP documents themselves. Future CAPs should be more concise and pragmatic, grounded in local needs and priorities, as well as the realities of county-level capacity, expertise and resource availability. Shorter and more user-friendly CAPs will also make dissemination easier. 9. Broad-based dissemination: Dissemination of all three CAPs in this research has at best been confined to government and CSO partners – as seen in Mombasa – and at worst has not been undertaken at all – as is the case so far in Garissa and Isiolo. In all three counties, dissemination to the wider community has been demonstrably lacking. For future CAPs, a bigger push will be needed though to disseminate the CAPs to the wider communities. Large-scale and innovative dissemination activities help to generate popular buy-in. Development partners or government bodies should allocate funds and time in grant activities for dissemination. It should follow the same process as outlined above for community consultation, in that it should be conducted outside of the main county towns, using translators for local languages and reaching a wider network within the community, including into at-risk or marginalized groups. Making use of public barazas, TV, radio and widely distributing simplified versions of the CAP document in hard copy should be considered. Bringing subcounty and ward administrators and chiefs into this process will facilitate wider dissemination across the county. 10. Transitioning from formation to implementation: This relates closely to the need for political will. At the outset of the CAP formation phase, national governments via the County Commissioners’ offices have taken a leading role. At the same time, the process is supposed to be run in partnership with the county governments, with responsibility for the implementation being handed over to the county government upon finalization. In reality, this collaborative relationship and the process of transition to the county government is vulnerable to disruption, stemming from wider frictions between the national and county governments. In this context, NCTC could play a central role in mediating between the two levels of government where necessary and supporting county government participation and ownership from the outset, so that this is not left too late. 26. STAKEHOLDER PERSPECTIVES AND ANALYSIS This second section of the report unpacks the research findings in more detail, as they relate to each key stakeholder group in the CAP formation processes, with reference to all three research counties. 26.1 NCTC In September 2016, the Kenyan government’s NCTC adopted the NSCVE. The resulting document outlined the government’s understanding of the drivers and means for radicalization of Kenyans towards VE, as well as approaches to prevention. Specifically, the NSCVE outlined nine key pillars of work through which activities should be directed: psychosocial; education; politics; security; faith-based and ideological; training and capacity building; arts and culture; legal and policy; and media and online. Subsequently in March 2017, NCTC released a document that set out the process by which the nine pillars of the NSCVE should be adopted and implemented at the county level by VE-affected counties. The document, known as the General CAP Guidelines, outlines the aims and process of forming a CAP to be used as a reference point by all counties developing a CAP. Both the NSCVE and General CAP Guidelines have been used in the counties in this research as a template and process guide, by national and county government, development partners, Malaika Foundation (whose role is detailed in the following section) and key local partners. 242 Consistent with the NSCVE and the General CAP Guidelines, in Isiolo, Garissa and Mombasa NCTC worked with the County Commissioner in each county to initiate the CAP formation processes. As such, each county’s CAP contains at a minimum the nine pillars contained in the General CAP Guidelines. In the research counties, additional pillars were included in response to locally-expressed demands. Generally speaking, NCTC engaged with the different national government counterparts in each of the county CAP formation processes at particular points, namely the County Commissioner. All three counties in this research have successfully finalized their CAP documents – a considerable achievement in a country-wide process that is relatively new. Mombasa’s CAP formation began in January 2016 with consultations and was launched in May 2017. In Isiolo, a first version of the CAP formation began independently in 2016 but was stopped by NCTC to ensure that the guidelines, particularly around inclusivity, were followed. The process was then restarted in December 2017 and the final CAP launched in September 2018. Garissa’s CAP formation began in March 2018 and was launched shortly after the field research for this report concluded, in November 2018. During this period, NCTC was highly instrumental in the proceedings. Mombasa’s CAP formation began before the General CAP Guidelines had been published, and therefore did not follow the same exact process as subsequent CAPs; the process, spanning 16 months, was somewhat slow and stilted. However, NCTC’s consistent physical presence and inputs during the consultations ensured that the MCAP reached completion successfully. Meanwhile in Isiolo, local CSO Isiolo Peace Link (IPL) in 2016 begun conducting a CAP development process without the guidance of NCTC, which proved to be insufficiently consultative, and therefore in 2017 NCTC rejected this early version of the ICAP and restarted the formation process. Isiolo and Garissa’s CAP formation processes thereafter followed the General CAP Guidelines, and the process ran more smoothly and efficiently, reaching completion in eight and nine months respectively. In Mombasa, the awareness of NCTC among stakeholders and its role in driving the MCAP formation was well known and commonly cited by stakeholders from all sectors as an instrumental force in the process. A seemingly greater physical presence by NCTC officials in key meetings encouraged this, with more non-governmental stakeholders in Mombasa being knowledgeable of what NCTC’s role was. From the perspective of local partners and other stakeholders in Isiolo and Garissa though, NCTC’s presence was less well-known. Commonly, when asked about their awareness of the role NCTC played in the CAP formation process, interviewees from local CSOs in those counties appeared unaware of NCTC’s specific role and spoke in general terms of “people from Nairobi” attending meetings. This begs the question of how NCTC’s role was presented to participants of formation meetings for the ICAP. Of the respondents who were knowledgeable about NCTC, a notable perception emerged that there was a slight disjuncture between NCTC and county-level stakeholders over which of the pillars of the CAPs should be prioritized: Specifically in Isiolo, local partners who had worked directly with NCTC during the CAP formation process noted that while NCTC wanted to prioritize the nine pillars outlined in the national General CAP Guidelines, at the county and community level, stakeholders saw the three Isiolo county-specific pillars which had been included on top of the core nine as being of greater priority. 26.2 National Government in the Counties The national government has demonstrated a proactive and effective engagement in the CAP formation process in all three counties. Consistent with the NSCVE and General CAP Guidelines, the national government is mandated with spearheading the CAP formation proceedings, via the County Commissioners’ offices. 243 In particular, the extent to which the County Commissioner played a strongly proactive role in the formation process, beyond simply a supportive one, proved significant. This is especially true of Isiolo, where the incumbent County Commissioner’s inputs throughout the formation process, which included attending community consultation events himself (noteworthy, given most senior government leadership, including the Isiolo Governor, would oftentimes send representatives in their place), are widely credited with ensuring that the CAP document was successfully finalized. The Assistant County Commissioner was also praised for his active support of the CAP. The previous county administration on the other hand was said to have had a tenuous relationship with the county government, which meant that during the previous incumbency, progress on the CAP had been slow (though as mentioned above, this was also due to the first version of the ICAP being scrapped), whereas under the current dispensation of county and national government ushered in by the 2017 elections, it was widely agreed that the formation process had noticeably gained momentum and was run more effectively between the two levels of government. This observation should be caveated though, as in the timelines set out for the ICAP formation, the bulk of meetings were planned to take place after the 2017 elections anyway. For the most part, the same was said of the Mombasa County Commissioner, who was similarly proactive in the formation process, in spite of competitive frictions in the earlier stages of the process between the national and county governments. A county government representative interviewed commented: "The County Commissioner played a big role in ensuring that there was sanctity and sanity in the whole process. His decision to own and run the process made other CSOs and stakeholders take the process seriously.” In Garissa, on the other hand, while no interviewees levelled negative comments towards the County Commissioner’s office vis a vis the CAP formation process, neither did any offer particularly explicit or enthusiastic praise. Rather, several interviewees from local civil society noted the scarcity of resources at both levels of government, which means that both sides are likely inclined to push the responsibility for the process off onto the other. That is, while the county government saw the CAP as principally a security issue and therefore to be resourced as a national responsibility, the national government saw VE as a localized issue and therefore the county’s responsibility to resource. At the same time, the formation processes should, according to the General CAP Guidelines, be co-chaired by the respective county governments, led by the Governor, leading to parallel inputs from both levels of government in each CAP. Across the three counties, the respective roles of the national and county governments, and moreover their relationship to one another, emerged as a key determining factor in the progress and the perception of the final CAP document. Put simply, at points when the county and national governments do not get on, progress on the CAP development stalls. Sustained engagement with national government in the counties is needed: Development partners as well as the central national government should engage with and provide support to County Commissioners in driving the formation process, as they are best placed to convene and provide momentum, while recognizing that the nature of engagement – which could include technical advisory, convening support or material support - will likely need to look slightly different in each county dependent on the background, capacity and personality traits of the County Commissioner. At the same time, too much investment in the County Commissioner’s office could prove unfruitful, given the high rate of turnover among government officials. 244 26.3 Security Agencies Certain departments in Kenya’s security agencies were cited as having the potential for instrumental and positive inputs into the CAP formation processes, even in a context where relations between communities and security agencies are often poor and need to be handled very sensitively. The appetite for some involvement of the security agencies among county￾level stakeholders is noteworthy. The NPS and ATPU were mentioned most commonly in the counties as the security agencies that should be better represented in the formation process. In Isiolo, there was confusion among respondents over the extent to which the ATPU was involved, with some complaining that the ATPU only attended the ICAP launch, which was an opportunity lost to engage throughout the process. However, an interviewee from the CSO Isiolo Women of Faith mentioned that she saw an inspector of the ATPU attend the two meetings in Isiolo, and also saw on a WhatsApp group that people were saying that since he started attending the ICAP meetings, there was improved reporting of VE cases to the police. Some complaints from the police as well as other CSO stakeholders in Isiolo indicated that in general, the NPS involvement was insufficient at the higher￾ranking levels, and moreover, it would have been beneficial to involve lower-ranking police officers, who deal with VE in their day-to-day work. In Garissa on the other hand, the police and other security agencies appeared to be minimally involved in the CAP, to its detriment. A county government respondent observed that: “The intelligence agencies were not there [with the community] like the ATPU and DCI representatives, as there are big gaps between the community and the security sectors…the security departments were not [utilized] well in not being the same room with community activists, [discussing] how to improve community security issues. The security departments don’t want to accept their fault for the use of a hard CVE approach to the community.” Even without the presence of security agencies, the climate of securitization in Garissa still impacted the CAP community consultations, with community elders being reluctant to participate because they saw the process as an intelligence gathering exercise. Conversely, in Mombasa, which also has a history of negative community-police relations, a respondent from the county government praised the inputs of the NPS: “The NPS was very active in the process and specifically the Changamwe OCPD (Officer Commanding Police Division) Mr. Peter Onamwa, other senior officials in the NPS were present to give their input on the security aspect.” One can surmise that the relative success of police and security actor involvement in Mombasa and Isiolo versus Garissa is heavily personality driven, given that both in Mombasa and Isiolo, praise of the security services for their involvement centered around particular individuals (for example: Inspector Bundi in Isiolo and OCPD Onamwa in Mombasa). 26.4 County Governments As mentioned, county Governors are intended to co-chair the CAP formation processes with the County Commissioners. However, as county government leadership is elected, their position is by nature, political-interest driven – for better or for worse. The impact that this dynamic has on the CAPs development process across the counties is varied according to the political landscape and the key personalities involved, but across the board appears to become more significant when the CAPs reach the point of implementation. 245 In Isiolo, several respondents referred to county government officials as “the political class,” and the overall majority expressed some degree of skepticism towards the county government’s genuine and lasting commitment to the ICAP. Stakeholders from local organizations and community participants in the ICAP formation commented that the county government leadership were inconsistently engaged throughout the process, with their presence only being clearly noticed at the ICAP’s launch. Respondents noted: “Many of the stakeholders do not understand their roles in the ICAP. For a county to implement anything, it has to be in the CIDP, there was a plan for the county government to allocate money for CVE to the ward administrators, but I think it will be misappropriated and never reach its intended use.” Respondents felt that the Isiolo county government saw CVE as a low political priority, in the absence of a major VE attack on the county thus far. While the county government is officially mandated to implement the ICAP, other issues were assumed by respondents to be of greater priority – a local youth leader commented: “Ideally, it should be the county government implementing the ICAP. However, the attention of the county government is on LAPSSET and 2022 politics.” “The county government cannot implement because they are a political institution. Radicalization is not important to the county government, and we doubt whether it has been ever discussed in the county assembly. The governor promised the ICAP would be included in the county's CIDP, but nothing has happened since.” As such, a majority of interviewees expressed positive views towards the County Commissioner’s office for its role in the ICAP’s formation and several went on to say that the national government and CSOs should lead the implementation of the ICAP, rather than the county government. Garissa meanwhile has been directly affected by major VE attacks as well as relatively high levels of VE recruitment, and as such CVE is viewed as a political priority as well as a security issue, and therefore invites a greater degree of support from both the national and county governments. Several respondents commented on the cross-cutting interest in the CAP by the present national and county government: “Both the county and the national government collaborated on the CAP formation; especially with the current county government as opposed to the previous one.” However, there remains a tendency for the Garissa county government to push some of the responsibility of implementing the CAP onto the national government, and vice versa: “The county government works with the national government very well although there was no clarity on whose responsibility it was on security since security is under national government but also affects locals who are under the county government.” Regarding the Garissa CAP’s (GCAP) implementation, though, most respondents seemed uncertain as to whether the county government or another institution would be responsible (though this confusion is to be somewhat expected, given that Garissa’s CAP at the time of research had not been finalized or launched.) Indeed, the provision of resources for implementation by NIWETU/USAID (the two were normally spoken of interchangeably by respondents) was commonly suggested, and in general, NIWETU seems to have had greater visibility in Garissa than Isiolo. Even the Deputy County Commissioner posed that: 246 “Both the county and the national government to be responsible for its funding, however [we have] requested the USAID to support the initial six months of its implementation seeing the process through then let the county work on it later.” In Mombasa, the role of the county government in relation to the national government during the MCAP formation process proved disruptive in the earlier stages, against a backdrop of long running frictions between the two levels of government. The lead local partner in the MCAP’s formation, HAKI Africa, as well as various members of the respective offices, are credited with smoothing over the tensions and ensuring that the process went ahead, holding a series of separate and then joint meetings with the County Commissioner and Governor. One interviewee explained: “The rivalry between the previous County Commissioner Marwa, then County Commissioner Achoki and Governor Joho delayed the MCAP by six months. Minister of Education Mr. Tendai played a key role and the Assistant County Commissioner Mrs. Esther. To resolve this problem, we requested the two offices to second two people who will represent them if they could not seat together. We sent Mr. Tendai to the governor and gave him an ultimatum of two weeks and he came back after a week with clearance and assurance from the governor." Other parts of the county government meanwhile were said to have been constructively supportive during the MCAP formation, with several respondents both outside and within government citing the education department, because it recognized that schools can play a highly significant role in CVE, which should be reflected in the education pillar of the MCAP. In Mombasa, where the MCAP has already been launched and implementation is supposed to have begun, confusion and concern over the resourcing of its implementation remains. The county government has established a secretariat to oversee and coordinate the resource mobilization, implementation and monitoring and evaluation of the MCAP. This mechanism should serve as an important body in the implementation, however it appears to have made minimal progress so far. According to respondents, the 100 million shillings committed by the Mombasa county government has not materialized because they do not have the correct policy in place to facilitate the allocation of the funds in this way. Several respondents from local CSOs argued that the national government should be responsible for the MCAP implementation and resourcing instead of the county government. During its formation, the MCAP received funding from the BRICS program (mainly via HAKI Africa) and a grant of $12,000 from the Strong Cities Network – some respondents hoped that this might continue. 26.5 Lead Partners Beyond the involvement of central government bodies in the CAPs, Nairobi-based CSO Malaika Foundation also provided technical advisory and support during the formation of the CAPs, following the NSCVE and General CAP Guidelines. For the most part, Malaika Foundation was described in positive terms by respondents in all three counties, as being an important partner in the formation of the respective CAPs. Despite being Nairobi-based and not from the counties, many respondents saw them as a key “local partner” alongside the three mandated local partners (Isiolo Peace Link, ROAD and HAKI Africa; their roles are discussed below). On the other side of this coin, a few interviewees belonging to local CSOs in the Isiolo county expressed some resentment of Malaika’s role. For example, a member of a local women’s organization complained that it would have been more appropriate for a local, “non-Nairobi”, organization to facilitate discussions, and that the Malaika consultants were “too academic and didn’t work well with the other stakeholders.” Nonetheless, having a single cross-cutting organization intimately 247 involved in all of the CAPs presumably provided an important and holistic perspective for the other stakeholders, and for this research team. In Isiolo, the lead local partner to the ICAP formation was IPL. Generally, respondents agreed that IPL had done a very good job of convening stakeholders and ensuring the momentum of the process, with several noting that IPL had actually initiated a CAP formation process. Inevitably, IPL’s taking a central role led to other CSOs in Isiolo who were not chosen to lead this activity to feel marginalized, with two respondents from different local CSOs saying that they IPL had “pushed others out” to some extent. Despite these claims, IPL’s role appears to have been instrumental in ensuring the ICAP was successfully finalized. In Garissa, ROAD was the implementing partner. During interviews, while respondents mentioned ROAD alongside other CSOs and actors as having supported the CAP formation, the sense that ROAD was a central driving force in the process did not emerge as strongly. Often, SUPKEM was cited as having been equally involved, as well as the Malaika Foundation, though details of the nature of their involvement were not provided. The main local partner in Mombasa’s CAP formation was HAKI Africa, having received funding support from the United Kingdom’s BRICS program. In the early stages of the MCAP formation, HAKI Africa’s position was perceived negatively by many stakeholders, but was ultimately said to be highly instrumental. At the outset, the organization faced challenges due to prior poor relations with the national government in Mombasa75 which had to be reconciled. HAKI Africa was very deliberate about demonstrating their centrality to the MCAP formation. Even though this role was mandated as the leading MCAP local partner, this still generated some initial resentment among other CSOs. For example, respondents belonging to smaller CSOs complained that HAKI Africa would put their logo alongside those of the national and county governments on all documents produced throughout the MCAP formation, while no other CSOs were able to include their logo. A representative from HAKI Africa himself noted: "At first, most CSOs felt that the document belonged to HAKI Africa and they did not feel like they owned the process or document, but when the County Commissioner took up the process, all the CSOs accepted the process and were comfortable to give their views.” A member of a local youth organization similarly commented: "There was the battle of ownership of the process but HAKI Africa came out strong because they had the resources and the capacity to convince the government security bodies to be part of the process. [But] small organizations had the risk of "Kutolewa" (removed from the process).” Indeed, as noted above, respondents said the director of HAKI Africa was a key actor in mediating between the county and national government and ultimately ensuring that these tensions did not derail the MCAP. Here, as with IPL in Isiolo, one can see the double￾edged sword of having a single strong central organization as the key local partner; in both cases despite resentment generated along the way, it seems to prove a key factor in successfully reaching the finalization stage of a CAP. 75 In 2015 the national government attempted to de-register HAKI Africa as a legal NGO entity, and in 2016 HAKI Africa released a report containing evidence that the security services had committed extra-judicial killings. 248 26.6 Local CSOs In each county, the involvement of a wide net of local CSOs in the CAP formation meetings was treated as a priority by the lead organisers, a valuable step towards building inclusivity. Certainly, the participation of a relatively large number of CSOs ensures a greater degree of inclusivity of the CAP formation, which was appreciated by interviewees in this research who shared a sense that “we were all there”. The dynamics associated with working with a multiplicity of non-governmental organizations as well as multiple government levels and departments produced varied dynamics across the counties. The other side of the coin is that the high numbers of designated local partners and other CSOs in the CAP formation processes can lead to overcrowding. This tendency led a respondent from the Ministry of Interior to comment: “The CAP formation processes were slower than they had anticipated…because of competition from CSOs involved in the development processes. A great challenge faced in the development of the CAPs in the implementation phase [will be] in duplication of CVE activities by actors.” In Isiolo a large number of other CSOs with religious, youth-focused and other backgrounds also participated at various stages of the process, including the interfaith group, Takuma Youth Group, the Council of Imams, NOWSUD76 and SCORES77, among others. Several of these groups were listed by interviewees as having provided valuable inputs into the ICAP’s content. The simple expedient of a WhatsApp group to allow rolling communication and notification of events was said to have been useful in ensuring information sharing and participation across the numerous groups and actors. However, a few complaints were also lodged in Isiolo by smaller CSOs such as NOWSUD and the Amani Club, regarding civil society participation, including that Christian leaders were not involved, and that it was difficult for school-based organizations to participate in events because the ICAP meetings happened during school hours. For those CSOs in Isiolo that did participate in the ICAP formation, a pattern emerged during interviews whereby a number of these groups did not differentiate between the ICAP formation meetings and wider CVE activities. For example, one youth leader, when asked about his involvement in the ICAP formation, responded that he had helped organize football matches between the police and local youth. A number of other local CSO members made similar comments. While the conflation of the ICAP with other CVE activities is not necessarily problematic in itself, this trend does raise questions about the level of clarity with which the ICAP activities were publicized and facilitated, which could then lead to a lack of understanding of the purpose and mandate of the ICAP down the line. Indeed, beyond a small number of key stakeholders, knowledge of the 12 pillars of the ICAP among wider CSO participants tended to be vague. In Garissa, a sizable contingent of other local CSOs were said to have participated by attending formation events, apparently via the Garissa Civil Society umbrella group, under which 23 CSOs sit. As was seen in Isiolo though, many of the local CSO members who had participated in GCAP activities that were interviewed for this research seemed unable to provide much detail or insight into the specifics of the GCAP and its content. Again, this raises questions over the quality and clarity in facilitation of GCAP activities. Put simply, this implies that, while giving multiple CSOs a seat at the table is an important first step towards inclusivity, this does not necessarily guarantee their active and constructive participation. Related to this, it was also noted in Garissa that a common issue plaguing civil society engagement processes across the region arose, relating to expectations of monetary 76 Nomadic Women for Sustainable Development. 77 Sensitization of the Community on Radicalization and Empowerment Solutions 249 incentives for participation (though it should not be assumed that CSOs are the only parties showing these expectations). An interviewee from the Garissa county government observed: “The only challenge was managing expectation of the local partners, as they used to get handouts in every function they attended. Also, some of them were briefcase organizations who are not actually working on the ground though they were few who were involved in the process.” Moreover, it was posed that such organizations do not fully represent local communities in Garissa’s sub-counties. This is presumably part of a wider trend of professionalization of ‘grassroots’ CSOs, such that they become distanced from their ‘roots’: "Most organizations are briefcase, with most having no physical location. If you call for a meeting of CSOs, 100 of them will show up but have no tangible grassroots touch with the communities here." In Mombasa, local CSOs involved in the MCAP were Muslims for Human Rights (MUHURI), Human Rights Agenda (HURIA), Likoni Development Community Programme (LIKODEP), Sauti ya Wanawake and Kenya Community Support Centre (KECOSCE), as well as the Kenya Red Cross. Efforts to involve these CSOs were considerable – the researchers were told that during the MCAP formation process, 57 consultation meetings were held. Amid other comments from NCTC that the MCAP formation was overcrowded, this indicates that there is a ‘bell curve’ trend to CSO participation: both too little, and too much inclusivity, can prove counterproductive. Ultimately, the collective inputs of CSOs and HAKI Africa appear to have been constructive and collaborative, in spite of the crowded space and related early rivalries over who should lead, or be seen to be leading, the formation process. A ward representative in the county government concluded: "All the CSOs who made their contributions on the MCAP supported the process fully and owned the process throughout up until it was launched. Actually, the CSOs took lead in all of the 11 pillars in the MCAP." Indeed, a respondent from the local (non-CSO affiliated) community felt that the CSOs were so active in the process that they absorbed the role of government: “The government has no commitment; they are being "carried" by CSOs. They are not committing any finances, they are joy riding on the good-will of CSOs…The government gave a lot of empty promises after the launch.” 26.7 Wider Communities In addition to creating space from inputs from a wide range of CSOs, the CAP formation processes are intended to be inclusive of the wider community as well. In the three counties, efforts to this effect were made to varying extents, though it became clear that community consultations were not prioritized by the government, lead local partners or development partners to the same extent as local CSO participation. As such, stakeholders from all sectors in the three counties claimed each had room for improvement when it came to community inclusivity. In Isiolo, community consultations and outreach were conducted within Isiolo town, plus two other areas – Merti and Garbatulla. While the sheer size of Isiolo county means that these two additional locations still leave a large expanse uncovered, arguably by at least taking the consultations to more remote locations, an effort towards greater geographical inclusivity was made. 250 Still though, stakeholders interviewed expressed dissatisfaction with the extent of community consultation during the process in Isiolo. Members of national and county government interviewed, as well as some CSO members themselves, complained that the events’ attendees were still largely made up of CSO members as opposed to members of the wider community. Of those community members who did attend such events, their understanding of the ICAP content and purpose was minimal. This would indicate that the events need to be designed and facilitated in a more community-centric way, which are sufficient to genuinely share information and to encapsulate community perspectives on VE and CVE. An interviewee from NIWETU commented that: “The groups targeted for consultation were between 30-40 women, 30-40 elders and 30-40 youth. They were not able to reach out to others due to unavailability of resources and was not certain that those reached out to would disseminate to their communities.” A small number of well-informed respondents in Isiolo observed that the consultation events in the county failed to reach a crucial audience: communities affected by or potentially at risk of VE. For instance, a member of the NIWETU team in Isiolo observed: “[There were] stakeholders not involved [such as] grazing committees, yet pastoralists are also vulnerable to recruitment.” A member of a local CSO observed: “In places outside of Isiolo [town], they were not aware of the CAP and its contents. Participation of people from outside Isiolo was minimal, and even for the people in Isiolo town, the document's knowledge was with a few people, mostly elites.” This audience could have included members of communities in known VE recruitment hotspots or micro-communities in Isiolo, or people whose immediate peers, schoolmates or family members have joined VE groups or been otherwise directly affected by VE. As has been observed through research78 conducted in Kenya, a deliberate engagement of at-risk and vulnerable communities can ensure both that more accurate information on the nature and experience of VE by local communities is captured, but also serves as a CVE approach in itself, by acting as a counter￾balance to historical or contextual patterns of exclusion of such groups or individuals. In Garissa, three consultation events were held outside of Garissa town – in Fafi, Ijara and Dadaab. While these efforts should be commended, it emerged through this research that none of the respondents were aware of these three events. This raises questions around the extent to which they were publicly known, the quality of facilitation used, or whether their outcomes were disseminated. In any case, this resulted in a unanimous criticism of the GCAP formation emerging from respondents – that its formation had not been inclusive of representatives or community members from Garissa’s other sub-counties, particularly those most heavily affected by VE. A county government representative told the researchers: “Local administrators like ward administrators, chiefs, youth, women and elders at local levels were not involved and not consulted…The process was focusing in Garissa town only.” For those consultation events that took place within Garissa town, several barriers to their effectiveness were also reported, including that the events were conducted in English and Swahili by “non-local facilitators,” which meant that many members of the largely Somali-speaking community struggled to participate. Furthermore, the events, at three-to-four hours long, were said 78 See, for example, the December 2018 report released by the BRICS programme: Preventing Violent Extremism: Understanding At-Risk Communities in Kenya. 251 to be too short and rushed for participants to fully understand, but at the same time, day-long events would create other barriers to participation. In the earlier stages, community elders were said to be reluctant to participate because they saw the events as a security intelligence gathering exercise which could put their communities at risk and required considerable persuasion from the event conveners. For Mombasa, mixed views were expressed regarding the extent to which the MCAP’s formation was sufficiently consultative of the wider community. The use of public barazas with the community, which were attended by members of the county and national government as well as CSOs, was noted as a positive way of bringing the community into the process. However, many respondents commented that actually, such events were dominated by ‘grassroots’ CSOs, which ostensibly represent the community but also can fall prey to the same ‘briefcase’ tendencies as other CSOs, as opposed to non-CSO members of the wider community. A member of the county government told the researchers: "It seems like we have the problem with mobilization. The community was involved, but it is the same [CSO] people who have been used over and over again. I wish we were working with the real beneficiaries. It also seems like the CSOs really duplicate initiatives, as one person can be invited to ten meetings when nine other people should have been involved. The target community was not consulted in the process." Another respondent – a member of local CSO Manyatta Youth Entertainment - made a similar comment: “The community was involved during drafting and even the validation process, though not sufficiently…even if it is said that the community was involved, it was the usual suspects that you would find in all the forums but not the real targeted communities affected by VE.” As with Isiolo and Garissa, minimal deliberate efforts to capture the views of particular vulnerable community groups, such as at-risk communities, women, or youth, appear to have been made during the MCAP consultations. Here we see an area in which the process applied in the counties did not fully encapsulate the approach proposed by the General CAP Guidelines. In the Guideline document, the terms ‘gate-keeper’79 and ‘social influencer’80 are mentioned. The former refers to members of local CSOs, whereas the latter denotes charismatic members of the (non-CSO affiliated) community. The document notes: “Both have their place in the CAP but it is extremely important to not confuse the two. The CAP should ensure both are included in the development and implementation process.” This being the case, one can broadly conclude that the CAP formation processes prioritized gate-keepers, at the expense of social influencers. The lack of community inclusivity in the consultation phases has carried over following the finalization and launch of CAPs, into the dissemination phase. In Isiolo, a number of barriers exist to the ICAP content being known and understood by stakeholders and the wider community. Foremost is that since the launch event, there have been no significant dissemination activities. According to local partners, this is because while NIWETU provided local 79 According to the General CAP Guideline document, gate-keeper characteristics include: “Easy to find; often engaged with political parties and politicians; often have a formal position within a social, religious or cultural group; they are entrenched in their ways and a certain way of engaging people; their greatest reach is with an older demographic; they restrict access and censor opinions; are often not influential with the youth; and sometimes have a direct financial interest in their social/religious/cultural role.” 80 Meanwhile, social influencer characteristics were said to include: “Likable; active in a particular activity; able to get their message out; vocal and have strong opinions; charismatic; admired by a specific constituency or age-group; often not united with gate keepers and can be in conflict with them.” 252 CSOs with grants to participate in the ICAP’s formation, the grant did not include resources to conduct dissemination activities. Some local partners say they are trying to do some level of dissemination attached to their other activities, but their capacity for more concerted activities is limited. At the time of this research, the county and national governments also have not undertaken dissemination of the ICAP. Other barriers to a wider awareness of the ICAP include that the events were apparently conducted in English, meaning non-English speaking communities struggled to participate, and that both soft and hard copies of the finalized ICAP document remain hard to access. As such, local ownership and buy-in of the ICAP remains minimal. In Mombasa some respondents, emphasizing the need to make a more comprehensible and digestible version for the wider community, suggested a range of means by which the dissemination of the MCAP could be expanded. According to a representative from Human Rights Agenda, a local organization: “Dissemination needs to be done in a more vibrant way. There is need to do a simple version for the local citizen as at the moment, it is difficult for them to understand it. The community needs to be sensitized on the document and the different stakeholders should all be involved in the dissemination process as they all have the capacity and strength to disseminate it in their own capacities and spaces. Sensitization also has to be continuous. The document needs to be simplified as it would be very complex for the local mwananchi (average Kenyan person) to understand it. Another thing that has to be done is to translate it into local dialect.”