IMPACT MONITORING REPORT. Armed Violence Reduction: Improving Safety in Karamoja

Transcription

1 IMPACT MONITORING REPORT Armed Violence Reduction: Improving Safety in Karamoja Impact assessment - October

2 About Danish Refugee Council - Danish Demining Group The Danish Refugee Council (DRC) is a humanitarian, non-governmental, non-profit organization founded in 1956 that works in more than 30 countries throughout the world. DRC fulfils its mandate by providing direct assistance to conflict-affected populations refugees, internally displaced people (IDPs) and host communities in the conflict areas of the world; and by advocating on behalf of conflict-affected populations internationally, and in Denmark, on the basis of humanitarian principles and the Human Rights Declaration. Danish Demining Group (DDG) is a Humanitarian Mine Action Unit in the Danish Refugee Council (DRC). The DDG mission is to recreate a safe environment where people can live without the threat of Landmines, Unexploded Ordnances and Small Arms and Light Weapons. DDG operates with three strategic objectives that define the fundamental and specific aims of our operations: Enhance Human Security by clearing landmines and unexploded ordnance, reduce the threat from small arms & light weapons and by providing risk education and raising awareness Provide impact oriented, cost effective and innovative solutions in close cooperation with relevant stakeholders Support local structures and institutions with finding sustainable solutions to residual problems and in a manner that enables economic and social development. We work to enable post-conflict communities to (re)gain access to their assets and support the efforts of governments, relief- and development organisation to enable recovery and community transition into social and economic development. Our work is based on the Value Compass of the Danish Refugee Council: Humanitarian approach - people s right to a life with dignity takes precedence over politics and principles Respect - for the equal rights of human beings Independence and neutrality - in regard to our surroundings Inclusion - of the people we work to help Honesty and transparency - for all beneficiaries, donors, partners and others Danish Demining Group C/o Danish Refugee Council Borgergade 10, 3rd floor DK-1300 Copenhagen K. Denmark Tel: ddg@drc.dk;website: and This report was written by the DDG Impact Monitoring team: Ocan Emmanuel and Semine Lykke Brorson 2

3 Table of contents Executive summary 1. Introduction a. DRC/DDG programme in the country b. Description of the AVR project b. Brief description of working context 2. Methodology a. Data collection methods b. Sampling and characteristics of respondents c. Data handling and analysis d. Challenges with data collection 3. Findings a. Immediate objective I i. Level of violence in target communities ii. Types of conflict and perceptions of community conflict management mechanisms iii. Community perception of security b. Immediate objective II i. Relationship between target communities and security providers ii. Community attitude to disarmament 4. Conclusions and recommendations 5. Annexes a. Succes story of Nakele Theresa from Lopelipei b. List of literature c. Survey questionnaire d. Focus group discussion guide 3

4 4

5 Acronyms AVR Armed Violence Reduction CME Conflict Management Education CSP Community Safety Planning CMC Community Management Committee IDP Internally Displaced Person LC Local Council (levels I V) LDU Local Defense Unit OCHA (United Nations) Office for the Coordination of Humanitarian Affairs RDC Resident District Commissioner SALW Small arms and light weapons UPDF Uganda People s Defence Forces UPF Uganda Police Force Glossary 1 Conflict A conflict is a clash between different interests, values, needs and/or understandings. Insecurity The threat or actual experience of violence or damage to person and/or property. It can also include the threat or experience of deprivation in a broader sense, including meeting basic needs like food and shelter. For this report, the focus has been more on the physical side of security. Karachuna Male youth in Karimojong society who implement the collective will of their community s traditional decision-making systems, and often take on the roles of protectors and warriors. Karamoja This refers to the five (old) districts of Abim, Kaabong, Kotido, Moroto, Nakapiripirit, which cover approximately 27,200 km of Northeastern Uganda and the two newly created districts Amudat which was created in December 2009 out of Nakapiripirit District and Napak which was created from Moroto District in Karimojong Refers to the inhabitants of the above districts and includes the ethnic groups (or sub-tribes) of the Dodoth (north); Jie (central); Pokot (Kenyan border) and Bokora, Matheniko and Pian (south). Smaller ethnic groupings that also come under the generic term Karimojong include the Tepeth, Nyakwe, Ik, Ngipore and Ethur. Other common spellings include Karimojong. Kraal Mobile, often fortified, cattle camps (also known as alomar or adakar). Kraals are usually guarded by the ASTU and the UPDF. 1 This glossary was partly derived from Saferworld (2010), Karamoja Conflict and Security Assessment 5

6 Ugandan People s Defence Force (UPDF) UPDF is the Ugandan national army. Local Defence Unit (LDU) A paramilitary unit under the command of the UPDF. LDUs are recruited within the local youth in Karamoja and once trained, they are posted in their local area. The purpose of the LDU is to deter youth from violence/cattle raiding, to provide protection for kraals and to follow of incidents of raids. Manyatta Traditional semi-permanent family habitat of the Karamojong, consisting of huts and granaries. Small arms and light weapons (SALW) Although there is no universally recognized definition of SALW, the International SALW Tracing Instrument describes SALW as any manportable lethal weapon that expels or launches, is designed to expel or launch, or may be readily converted to expel or launch a shot, bullet or projectile by the action of an explosive.[ ] (a) Small arms are, broadly speaking, weapons designed for individual use. They include, inter alia, revolvers and self-loading pistols, rifles and carbines, submachine guns, assault rifles and light machine guns; (b) Light weapons are, broadly speaking, weapons designed for use by two or three persons serving as a crew, although some may be carried and used by a single person. When the assessment uses the term, it does not include other weapons such as clubs, knives and machetes. 6

7 Executive summary The north-eastern Karamoja region in Uganda has for decades been a scene for conflict and armed violence. Insecurity in the region is attributed to several factors including: inter-tribal conflict characterized by cattle raiding; conflict within communities, such as domestic violence and petty theft; and tension between communities and security providers due to forced disarmament programs by the state. Insecurity is heightened by easy access to small arms and light weapons (SALW) from the bordering regions of South Sudan and Kenya as well as domestically. Karamoja is home to the country s highest rates of people living in poverty and injuries deriving from incidents with small arms and is a region where development has been halted by security threats. Based on this, DGG has launched an Armed Violence Reduction project in Karamoja that seeks to address conflicts within communities, improve local management of security threats and to build trust and dialogue between communities and security providers and thus contribute to lowering the level of violence in the region. This report assesses impact in communities within the project s area of operation and explores mechanisms of impact in the AVR project. It is based on quantitative data collected in household surveys survey prior to and at the end of the intervention, together with quantitative data collected in focus group discussions and key informant interviews at the end of the project period. In summary, the report makes the following impact assessment findings: There is a tendency of change from armed violence to less grave forms of physical violence in conjunction with trends of increasingly peaceful usage of weapons like knives and spears. Continued low level violence in communities is dominated by domestic violence and incidents deriving from excessive alcohol consumption. At the inter-tribal level, large scale cattle raids has transformed into incidents of cattle theft. A major change has occurred in communities capacity to manage, constrain and prevent conflict. This change is associated with an increase in community members inter-personal skills and a stronger and more effective role of the Local Council I Chairpersons as mediators. There is a significant improvement in community perception of security and a high increase in the sense of safety. In consequence, there is a tendency of reorientation towards issues of development rather than concerns over security. The relationship between community providers and communities is changed from mainly being characterized by fear, intimidation into a more cooperative relationship based on dialogue which supports a positive circle of improved flow of information, improved performance of security providers and improved trust. Community-driven sensitization on the dangers of small arms holds potential to change attitudes and behaviors regarding use of weapons. Overall, significant improvements have been observed in the targeted communities in relation to the project objectives. In this impact assessment report, the findings will be presented in more detail together with analysis of the factors that has contributed to the observed change. 7

8 1. Introduction DRC/DDG program in the country DRC has worked with Sudanese refugees and the Ugandan communities hosting them in West Nile since In 2007 DRC expanded its operation to also include support to IDPs living in camps, and those returning home in Northern Uganda. DRC's role working with IDPs has largely been provision of protection through presence and monitoring, advocacy on behalf of the population and helping the communities to rebuild their own support systems for helping the vulnerable. DDG launched activities in Uganda in 2008, and has been working in mine action in Northern and Western Uganda with technical, financial and logistic support to the national Mine Action Program for a 4 year period. In 2010, Armed Violence Reduction activities were launched in the north-eastern region of Karamoja which by the end of 2012 is the main region of DDG s intervention. Description of the AVR project This impact monitoring report covers the Armed Violence Reduction project implemented by Danish Demining Group in the Ugandan region of Karamoja. The project in question is a second cycle in which project activities of an initial one year project cycle has been continued and expanded. The intervention is based on a participatory needs assessment in target communities, stakeholder consultations with local leaders, government officials and security providers and an extensive research about perceptions of small arms and security provision in Karamoja conducted in partnership with the Small Arms Survey. Also, the intervention has been guided by findings and experiences in DDG s Armed Violence Reduction programs in Somalia and Southern Sudan. Overall and immediate objectives The overall objective of the project is to contribute towards creating a foundation for sustainable peace in Karamoja. In order to contribute this overall objective, DDG s intervention has the following immediate objectives: 1) Level of security is improved in the target areas and communities perceive security to be better 2) Capacity of the security providers to better cater for the needs of the communities is increased Main activities and outputs In pursuit of immediate objective 1, DDG carries out activities to strengthen community conflict management and local security management as a measure to prevent escalation of conflicts into violence and security threats. Three types of activities support this. Firstly, Conflict Management Education (CME) at village level, where 20 individuals have been trained in 137 villages. Secondly, through participatory development of Community Safety Plans (CSPs) relevant safety and security initiatives have been developed in 7 targeted Sub-Counties. During this process 7 Community Management Committees (CMCs) have been elected and tasked with implementation of the Community Safety Plans. DDG has supported this implementation and facilitated trainings in community mobilization, fundraising, proposal writing and Conflict Management Education for 335 CMC members and 183 LCI Chairpersons. Thirdly, small arms sensitization has been carried out in the form of radio messages and community-based dramas developed and performed mainly by young community members. 49 drama performances have been carried out across the 7 Sub-Counties, involving 105 performers. This figure is lower than originally scheduled, and the sensitization activities are scaled up in the 3 month extension of the project to reach the targeted output. In pursuit of immediate objective 2 on security providers capacity, two types of activities have been implemented. Firstly, dialogue forums have been created between communities and the security providers Uganda People s Defence Force (UPDF) and Uganda Police Force (the Police). These Community Regular Meetings (CRMs) have 8

9 taken place at Sub-County level, monthly in most locations and bi-monthly in locations of particular need. In the 7 Sub-Counties a total of 93 Community Regular Meetings have been conducted in the project period. 2 Local governance stakeholders like Sub-County Chiefs, LC IIIs and LCIs have been strongly involved in this project component. Secondly, in pursuit of increased capacity of security providers to cater for the needs of communities Conflict Management Education has been carried out for security providers. In 33 CME sessions, 557 security providers across the UPDF, the Police and Local Defence Units (LDUs) have been trained. To sustain these achievements, 86 security personnel have been trained as Trainers of Trainees with the purpose of continuing training activities beyond the project. During the course of the project, a need was identified for adding an additional activity. As such, 8 peace meetings between conflicting ethnic groups have been supported. In this activity, DDG supported LC IVs, Sub-county Officials, Resident District Commissioners (RDCs) and various community leaders in their role as the facilitators of peace meetings upon their initiative and contributed with resources to conduct the meetings. This support has been given upon the condition that meetings included deliberation on conflict resolution and written documentation of commitment to the outcomes. The meetings have been fully implemented by local stakeholders. Implementation period: September 2011 to September Donor: Norway - Ministry of Foreign Affairs Brief description of working context Karamoja has been characterized as one of the world s most armed violence affected regions, with a small arms death rate nearing 600 per 100,000 of the population 3. The insecurity in Karamoja is characterized by several factors including protracted inter-tribal conflicts in which cattle raiding are a key factor. Furthermore, conflict within communities, such as domestic violence and petty theft, has severely impaired the level of security among the Karamojongs. Additionally, forced disarmament programs by state security providers have created an environment of distrust and insecurity between civilians and security providers. In summary, three levels of conflict prevail in the Karamoja region. The first level regards conflict and an antagonistic relationship between the Karamojong Society and the Ugandan state. The second is seen at the inter-tribal level, with a history of violent clashes between ethnic groups in competition over resources and distribution of wealth. The third is in the form of conflicts between individuals and families within ethnic groups, which has been found to account for a relevant proportion of violence in Karamoja 4. The AVR project in question mainly addresses the issues of conflict between state security providers and communities, and conflicts within communities. The inter-tribal level is included to the extent of support of peace meetings and positive spill-over of conflict management within communities in border areas with other tribes. Targeted areas, districts and communities The AVR Project has been implemented in 7 Sub-Counties in 4 out of 7 districts in the Karamoja region. Project areas have been selected due to their location near district boundaries which marks areas of tension in intertribal relations. The program activities mainly target the levels of villages and Sub-Counties, and activities have been rolled out in the total of 183 villages in 7 target Sub-Counties in the district of Nakapiripit, Napak, Kotido and Moroto 5. 2 In continuation of the structure of the AVR Project first cycle in , Community Regular Meetings have taken place at parish level rather than Sub-County in the target areas of the first cycle in Moroto. These areas account for 30 of the total number of CRMs, whereas 63 have been conducted at Sub-County level in the new project areas enrolled in the second cycle. 3 Bevan J(2008), Crisis in Karamoja: Armed Violence and the Failure of Disarmament in Uganda s Most Deprived Region, Small Arms Survey, Occasional Paper, p.16 4 Kingma, Kees et al (2012). Security Provision and Small Arms in Karamoja a survey of perceptions. Special report by the Small Arms Survey and Danish Demining Group. 5 Part of the activities carried out in Moroto District follows the program structure of the first project cycle, where the level of intervention has been parishes and villages, rather than Sub-Counties and villages as in the remaining project areas. 9

10 Socio-economic characteristics of the area In 2009, 80% of the population of Karamoja was found to live in poverty 6. The history of violence and insecurity in the region has served as a deadlock for economic development, and the region severely lags behind in terms of development in comparison with other regions in Uganda. Demographically, the population is dominated by young people who historically have been the agents of inter-tribal violence, and who are today faced with difficult access to education opportunities and employment. Karamoja is characterized by a high level of ethnic diversity with 7 major tribes as the dividing social structure. The project areas reach out to the Jie tribe in Kotido, the Tapac and Matheniko in Moroto, Pian in Nakapiripit and the Bokora in Napak. The administrative districts in Karamoja are highly aligned with the ethnic borders, whereby the project locations along district boundaries in general target communities with frequent inter-tribal encounters. Characteristics of the beneficiaries The project has three categories of direct beneficiaries. For once, ordinary community members are selected to undergo Conflict Management Education and to carry out sensitization on the dangers of small arms. On a basis of rotation, people from this category also take part in Community Regular Meetings to ensure the spread of information from this forum. Secondly, there is a category of community leaders, which most significantly include the LC-Is and LC-IIIs of the targeted villages and Sub-Counties. By the instatement of Community Management Committees, opportunities are opened for other community members to participate in community leadership in regards of safety and security whereby the category of community leaders is broadened. Thirdly, security providers in the form of UPDF soldiers, Police men and Local Defence Units have been involved in Conflict Management Education for security providers and Community Regular Meetings with leaders and members of the communities. The outreach to these beneficiaries seeks to create impact for a wider category of indirect beneficiaries which comprises the community members in contact with the direct beneficiaries, e.g. household and family members of individuals trained in CME, community members in need of support from local leaders and individuals interfacing security providers. Changes to the working context since start of operation The intervention in Karamoja has taken place in a context of continuous and gradual improvements in the security situation over the past years. As such, activities feed into societal processes of installation of peace in the region and extension of state services and seeks to augment their effects. Since the start of the operation, three major changes have occurred in the working context. Firstly, a change in the mandate of the paramilitary Local Defense Units came into force at the beginning of 2012, whereby security provision at local level has been enhanced. Since the LDUs are under the command of the UPDF, this can have contributed to the heightening of trust in the military as well as satisfaction with UPDF s performance. Secondly, following the tendency of improvements in the general security situation in Karamoja there is a process of change from relief operations to more long term development interventions in the region, along with an increase in the number of NGO s that are active in the region. Thirdly, rain fall has been at an atypical high level in the period of the current project cycle. In consequence, there have been more conducive conditions to pursue livelihood though agricultural activities which in turn can have impacted positively in the level of violence by reducing the incitement to obtain food forcibly. 6 OCHA, Focus on Karamoja, Special Report No.3: October 2008 to January 2009, p. 1 10

11 This impact assessment report seeks to capture changes in the stated indicators at community level. Other factors can be expected to have made an impact on the project indicators in the period of intervention. DDG welcomes any positive change in the targeted communities, and is aware that it is not possible to control for the effect of external and intervening factors to clearly demarcate impact attribution. In this report, however, an attempt has been made to explore mechanisms within the AVR Project that can explain the ways in which the program is likely to contribute to any observed changes in the indicators, and these findings are included in Chapter Methodology Data collection methods This impact monitoring report is based on a combination of quantitative and qualitative data, whereby it aims at monitoring the descriptive impact in the target communities and to identify mechanisms that has contributed to change in these communities. Quantitative data in the form of household questionnaires has been collected prior to the initiation of this second program cycle in September 2011 and again at the end of the cycle in September The qualitative data is generated at the end of the program cycle in 8 focus group discussions in 6 program locations and in 7 Key Informant Interviews, conducted in the period between15 and 25 September Baseline and endline data has been collected by a team of local field staff, at end-line supported by 2 external enumerators. Data collection has throughout been monitored by the DDG M&E Officer. This report includes a comparative analysis of the baseline and end-line data, and a qualitative analysis of explanatory factors described in the focus group discussions and Key Informant Interviews. Sampling and characteristics of respondents Random sampling in quantitative data collection Sampling for the baseline and end-line household questionnaire surveys has been conducted in a way that seeks to accommodate representativeness for the population in the 7 targeted sub-counties. In the baseline a total of1050 household questionnaires were collected in 30 parishes, covering the total number of 183 targeted villages in 7 subcounties. For practical reasons, the end-line included 554 household questionnaires collected in 55 villages, covering 29 parishes. As such, the end-line data is more geographically clustered towards the parish level than the baseline, and has a confidence interval of +/- 4 while that of the baseline data is +/-3 7. Data collection has been sensitive to the demographic factors of gender and age. Non-random sampling in qualitative data collection The qualitative data has been collected non-randomly with the purpose of assessing explanatory factors of the quantitative findings and to reveal more in-depth information about the impact. With that objective, three types of respondents have been included in 8 focus group discussions and 6 Key Informant Interviews. Firstly, community members with no involvement in the program activities have been included in 2 focus group discussions. This group has been included in order to assess the depth of impact in the community as a whole. Secondly, people with knowledge about the intervention in the community has been included in 5 focus group discussions in order to explain the particularities of the intervention at different locations and its interaction with social structures in the community. Thirdly, Key Informant Interviews have been conducted with community leaders, government officials and representatives of the security providers who hold particular knowledge about the impact of the program on their fields of work in security and community leadership. The Key Informants include the LCI Chairpersons of Nawait in 7 Both baseline and end-line confidence intervals are at a 95% confidence level. The calculation of confidence interval is based on projections of population size from a 2002 national census, and is therefore sensitive to errors in the projections. Furthermore, it is vulnerable to uncertainties about the population size at Sub-County level. 11

12 Lorengedwat and of Lopei Trading Centre, The LCIII Chairperson of Ngoleriet, The Sub-County Chief of Panyangara, the UPDF Chief Administrator in Kotido and representatives of Uganda Police Force in Kotido. In Tapac, a total of 12 LC-I Chairpersons were consulted in a focus group setting. Furthermore, interviews have been conducted with 3 UPDF participants in Conflict Management Education for Security Providers, with the Chairperson of the Community Safety Committee in Panyangara Sub-County in Kotido District and with selected participants in Community Conflict Management Education. Data handling and analysis The quantitative data was immediately after data collection entered into the AVR program s statistical database, utilizing SPSS as the tool for data analysis. The qualitative data was recorded in hand written notes during the focus group discussions and Key Informant Interviews. The data was afterwards coded according to the program objectives and indicators, and is stored digitally in this form. Data reliability has been assessed mainly by the use of source triangulation, whereby analysis of the program indicators cuts across the qualitative and quantitative data sources as well as monitoring data that has continuously been recorded in the form of community Incident Reports of violence and monitoring forms for usage of CME skills at village level. Challenges with data collection During the process of data collection and data analysis two forms challenges has been encountered. Construction of scales The first challenge regards construction of selected answer scales in the household questionnaire. During data collection it became very clear that numerical answer categories in the form on a scale from 0 to 4, how do you rank are not very sensitive to the Karamojong respondents way of thinking. In consequence, answers were in many cases given as a narrative statement rather than a numerical answer, after which the answer was rated numerically by the enumerator who conducted the interview. This leaves an unfortunate room for subjective interpretation of the enumerators, which is exacerbated by the relatively high turn-over of enumerators that followed the use of local field staff in the 7 sub-counties. In this survey this type of challenge mainly affects the measurements of community perception and rating of security providers performance. On the same note of scales, two questions in the questionnaire have answer scales that are constructed with the purpose of prioritizing 3 answers among a list of choices. During the data collection it was difficult for respondents to manage the many answer choices. In the data analysis, the same questions have been affected by the fact that the current construction of the AVR program s database does not enable statistical analysis that reflect ranked prioritization of several answer categories. For these reasons it has not been possible to fully utilize the measurements regarding security concerns in the community and main perpetrators of violence. Aggregation of monitoring data The AVR program is embedded in an ambitious monitoring set-up where household records of violence and usage of the conflict management skills taught in the CME education is continuously collected by the field staff and stored. In order to gain more from this work, it would be useful to have the data entered in a database that enabled systematized analysis at an aggregated level. In this respect it would also be advisable to record data in a more numerical form that would enable statistical analysis of progress and correlations. In this report, it has not been possible to analyze the monitoring data in an aggregated and systematic way. This - in addition to the above mentioned challenge regarding scaling - means that some of the sources of triangulation have been incomplete whereby the mentioned variables is more sensitive to human error. 12

13 3. Findings In this chapter, the impact assessment findings are presented. The findings are structured by the indicators under the two immediate objectives and will present the baseline situation and changes in the situation with reference to the quantitative and qualitative data. Immediate objective 1: Level of security improved in the target areas and communities perceive the security to be better Indicator 1: Change in level of violence in target communities Situation at baseline At the time of the baseline survey a tendency of reduction in the level of armed violence could be traced in the data. While the majority of respondents (55,9%) ranked the level of armed violence as somewhat decreased, 19,4% stated that there had been no change in the situation compared to one year earlier. 9,7 % of the respondents stated that the level of armed violence had increased. 46,1 % indicated that a member of the household had been the victim of a violent encounter in the previous year 51,3% stated that a member of the household had been threatened by someone with a firearm Changes in the situation (impact assessment) At the time of the end-line survey, the tendency of reduction in the level of armed violence in communities is found to be strengthened, together with a qualitative change that makes physical and domestic violence relatively more predominant compared to violence with the use of small arms. A gradual process of reduction in the level of raids has currently reached a point where the pattern of violence is one of small-scale criminal incidents. Evidence The portion of respondents who ranked the level of armed violence as somewhat decreased remains stable at 56,6%, whereas the proportion who state that there has been no change is reduced to 7,1% and the proportion stating and increase in level of armed violence is reduced to 3,1%. The major change is found in the answer category that armed violence has decreased a lot with 33,2% at end-line compared to 15% at baseline. 39,2 % indicated that a member of the household had been the victim of a violent encounter in the previous year, which appears in conjunction with a qualitative change in the graveness of violence due to a tendency away from armed violence towards more physical forms of violence. 40,9 % stated that a member of the household had been threatened by someone with a firearm. In the focus group discussions there has been stated a tendency of fewer injuries occurring from small arms, and a reduction in the number of women who are victims of rape. There is a unanimous tendency in the qualitative data regarding an increase in free movement in the communities both in day time and evening hours, which strongly indicates a lower risk of assaults and threats. There is presented a general picture of increased peaceful usage of weapons like knives and spears in hunting, farming and rituals. On the same note, though, it is indicated that knives, sticks and bows and arrows are the predominant instruments at hand when violence arises in the form of e.g. fights between individuals in a community. 13

14 At community level, it is stated, the majority of violent incidents involve drunken individuals, and it is indicated that problems with alcohol consumption is a driving force of continued low-level violence and risk of escalation of conflicts between individuals. New mechanisms of prevention of violence is highlighted in the form of social constraints of individuals prone to conduct raids, and particularly the enrolment of previously idle and aggressive young men as LDU s under the command of the UPDF. The occurrence of incidents with small arms in the communities is found to correspond with the level of disarmament conducted in the area. As such, the tendency of reduced risk of armed incidents is at a slightly lower in the area of Tapac, where a high presence of small arms is still prevailing. However, a tendency of reduction in conflicts and conflict escalation is found to be at the same level as in the other target areas, as will be described in further detail below. Indicator 2: Change in conflict types and community conflict management mechanisms Situation at baseline A tendency of decrease in the number of conflicts in communities is reflected by 57,5 % of baseline respondents stating that they experience the number of conflicts as somewhat decreased and 10,3% experienced the number as decreased a lot. 22,6% stated that there has been no change compared to one year earlier, and 10% stated an increase in the number. The prominent type of conflict is between families, which accounts for 62,9% of the answers at baseline. This is followed by conflicts between individuals at 22.2%, and inter clan/tribal conflicts at 9.8 % A positive tendency is indicated in relation to improvement of community conflict management mechanisms as 73 % of respondents rate them as better than one year earlier. 23,2% express that mechanisms of conflict management at the time of baseline were the same as 12 months earlier. Changes in the situation (impact assessment) The positive tendency of a decrease in number of conflicts is strengthened and the capacity to manage conflicts within communities is further increased by improvement in the quality and effect of mechanisms applied to resolve conflicts. It appears that conflicts between individuals are less prone to escalate into conflicts between families This impacts positively on that fewer respondents rate family conflict as the most prominent by the time of end-line, although this type still accounts for the largest proportion. This decrease has made clan/tribal conflicts relatively more visible, and an increase is seen in the proportion who ranks inter clan/tribal conflict as most prominent. As the level of conflict is reducing, however, this reflects a proportional rather than numerical increase in the prominence of ethnic conflicts. At community level, there has been a change in the agents and forums of conflicts resolution, which has created a change in the level of conflict management, whereby conflicts are solved by involvement of stakeholders nearer to the conflict and increasingly by non-violent means. At the level of inter clan/tribal conflicts, revenge patterns and threats has been reduced between selected groups following strengthening of the element of conflict resolution in peace. Evidence 14

15 At the time of the end-line survey, the proportion of conflicts occurring between families was reduced to 48,9%. Conflicts between individuals was steady at 21,1 %, whereas the inter clan/tribal conflicts was rated higher at 23,6% compared to 9,8 % at baseline. In the end-line survey, the proportion who had experienced no change in the level of conflict dropped to 6,2% compared to 22,6% at baseline. At the end-line, 2,4 % of respondents express to have experienced an increase in the number of conflicts compared to 10% at baseline. More than 90% express to have experienced a decrease in number of conflicts, divided between 62,4% stating that the number is somewhat decreased and 29.2% stating that the number of conflicts has decreased a lot. In relation to changes in community mechanisms to manage conflicts, three factors have been stressed in the focus group discussions and key informant interviews. Firstly, there is an increase in inter-personal skills within families that enables management and containment of conflicts at a lower level. On the same note, community members increasingly step in as managers and preventers of conflict. Women respondents emphasize that an increase in the level of understanding between husbands and wives together with a strengthening of communication has contributed to reduction in household conflicts. It is also stated that women are becoming less prone to encourage raids due to a change in the risk assessment of male family members resorting to such actions. Secondly, the LC-Is are increasingly empowered and involved as mediators, which contributes to the emergence of a norm of non-violent conflict management as LCIs apply methods of dialogue and rather than physical punishment as often practiced by community elders who previously had a stronger role in enforcing community conflict resolution. Thirdly, new forums of management of conflicts and security issues are arising; village meetings have been strengthened or introduced, and a tendency of stronger cooperation is seen in the form of information sharing and collaboration between LCIs in a sub-county, which enables more effective management of conflicts that cuts across villages. This was particularly highlighted in a focus group discussion with 12 LC-I Chairpersons in Tapac Sub- County, a key informant interview with the LC-I of Nawait in Lorengedwat Sub-County and a focus group discussion with community members in Ngoleriet Sub-County. Respondents attribute the spread of conflict management skills to a trickle-down effect caused by continued informal education conducted by individuals trained in Conflict Management Education, who furthermore function as role models and mediators in a community. The impact of these changes is summarized in the quantitative finding that 88% of end-line respondents state that conflicts management mechanisms in the community work better than one year earlier. Indicator 3: Community perception of security Situation at baseline Security concerns was the most prominent type of concern, listed by 36,8% of respondents as their main concern. The issue of health care was stated as most concerning by 6,3 %, and water, hygiene and sanitation by 11,6%. Education was rated as the main concern by 4% of respondents. 8 91,3% stated that their Manyatta is faced with safety and security concerns 63,4 % expressed concern over the risk of becoming a victim of a violent encounter Changes in the situation (impact assessment) There is a change in the perception of level of security threats towards Manyattas, as there is a decrease in the proportion of people who experience external security threats. In absolute terms, however, the proportion is still high. 8 A quite high proportion of 30,5% stated other issues than the ones listed in the answer choices as their main concern, which indicates that the answer categories have not fully captured the prevailing concerns for the communities. In the end-line, other issues accounted for 14,6% of the answers. 15

16 Percent Concerns over security issues are relatively less prominent compared to development concerns like health care, water & sanitation and to some extent education, which are all becoming more prominent. A significant increase is seen in the sense of safety expressed by community members. Evidence In the quantitative data, it is found that the prevalence of security concerns remains high, as 83,2% stated that their Manyatta is faced with safety and security concerns. This figure is mediated by a tendency in the qualitative data towards a decrease in the intensity of this threat, as the fear of large-scale raids and attacks on a community is diminished. Respondents also generally indicate a strong improvement in sense of safety when moving around in the community. Examples are trading centers that are now approachable without fear, movement in the evening hours and a reduction in feelings of needing to carry a weapon for protection. Furthermore, the quantitative data indicates a relevant change in security concerns relative to other issues, as indicated in the following table: Table 1: What is your main concern for your community? September 2011 (n=1050) September 2012 (n=554) Lastly, the change in the proportion of respondents who express concern over the risk of becoming a victim from 63% at baseline to 24% at end-line strongly suggests an improvement in community member s perception of security. The findings regarding change in perception of safety is positively disproportionate to the indicated reduction in number of violent incidents. This suggests that the major impact in the period of intervention has been in the perception of security, and supports the previous finding that a significant positive change is seen in the form and graveness of violence, while the number of violent incidents has undergone change more steadily (46% indicated that a household member had been a victim of violence at baseline, followed by 39,2 % at end-line). 16

17 Immediate objective 2: Capacity of the security providers to better cater for the needs of the community is increased Indicator 1: Relationship between target communities and security providers Situation at baseline Assessment of community perception of security providers is measured on variables regarding trust and performance. Respondents are to rank security providers on a scale from 0 to 4, where 0 is the least positive and 4 the most positive. Due to challenges in numerical ranking encountered during data collection the scales were to some extent converted into the narrative answer categories ranging from very bad to very good. Respondents have in the baseline survey given the following rankings of the security providers: Table 2: Baseline relationship between security providers and community members Ind ic a t o r s V e r y Bad (0) Bad(1) F a ir(2) G o o d(3) V e r y G ood(4) 1.1 Trust the Police Efficiency (get the job done/create Results) Accessibility Trust the UPDF Efficiency at dealing with raids/providing security to kraal Consultation with the communities while carrying out disarmament Changes in the situation (impact assessment) A significant positive change is found in the ranking of security providers performance and the relationship between community members, particularly regarding UPDF. There is more trust and cooperation between communities and security providers. A higher flow of information from communities to security providers enables more effective follow up on violent and criminal incidents. A behavior change among security providers particularly UPDF is indicated by communities, whereby law enforcement is conducted with less force and more targeted suspected individuals rather than a whole community. Communities express increased demand for and appreciation of presence of state security personnel. 17

18 Evidence In relation to the Police, improvements are indicated on all categories of trust, effectiveness and accessibility. The smallest change is found in relation to accessibility which corresponds with the fact that police presence remains low in Karamoja. However, it also relates to a qualitative change in communities expectation of police presence in their community. The major changes in the relationship between communities and security providers are found in the perception of UPDF, where the percentages presented in the table below indicates a transformation of the relationship. 77,8% of endline respondents indicate that trust in UPDF is either good (3) or very good (4). 9 The quantitative and qualitatively data indicates that the relationship today is characterized by trust and cooperation, and that communities experience a big change in the approach and level of force applied by the UPDF towards the communities. This change is attributed to a sense of being one with the army, which has emerged from dialogue and interaction in the controlled environment of Community Regular Meetings and the enrolment of local young people as LDU s and recruits in the UPDF who have broadened the interface between UPDF and Karamojong communities. Table 3: Endline relationship between security providers and community members Ind ic a t o r s V e r y Bad(0) Bad(1) F a ir(2) G o o d(3) V e r y G o o d(4) 1.1 Trust the Police Efficiency (get the job done/create Results) Accessibility (easy to find) Trust the UPDF Efficiency at dealing with raids/providing security to kraal Consultation with the communities while carrying out disarmament There is strong evidence in both the quantitative and qualitative data for a significant behavior change of the UPDF soldiers present in communities. The combination of the more qualified level of information that guides security provider s actions and the collaboration with community leaders has led to a changed approach whereby security providers increasingly - to the extent of generally - target suspected individuals rather than the approach of encircling a whole community with force. On the side of security providers, respondents also relate this to an increased feeling of safety while interaction with communities and increased capacity to deal with community members by use of new conflict management skills. The long term effect of the latter is, however, vulnerable to the high level of transfer of security providers. 9 In the end-line survey responses from Panyangara and Nakapelimoru Sub-Counties in Kotido District, there is unlike in the other Districts a tendency of larger change in the level of trust in the Police, than the change in the level of trust in the UPDF. This is to be interpreted within the context that there in Kotido is a history of a particular hostile relationship between communities and the UPDF. 18

19 Percent Data indicates that a key change has taken place in the mode of contact between community members and security providers. There is a unanimous tendency that LC-Is role have become a form of gatekeepers between individuals in a community and security providers. In consequence, LC-Is is the connection point of information regarding incidents and knowledge of individuals in possession of weapons flowing from the community to security providers. The LC-Is function also serves as a form of protection for community members, as security providers have become knowledgeable about the presence of LC-Is and utilize them as a valid entry point to the community. In consequence, local governance s role in security provision has been enhanced and law enforcement is to a higher extent conducted hand in hand with community leaders who hold detailed knowledge about incidents and individuals in a community. In consequence, focus group respondents state that community members are no longer intimidated by the presence of security providers, and that there from a highly antagonistic starting point has been a process of humanization of the perception of each other. Lastly, there are indications of an increase in the accountability of security providers. There are slightly mixed findings regarding the level of corruption, but the tendency appears to be a decrease in the level of bribes in most of the targeted areas. In addition, there are indications in some locations of emergence of complaint mechanisms where communities refer issues to Sub-County officials in case of incidents of violation. Indicator 2: Community attitude to disarmament Situation at baseline In the baseline data there are indications of a positive tendency in people s opinion towards the impact of disarmament on the community. 49,5 % of respondents state that they find the impact positive and 27,1% that it is somewhat positive. At the other end of the scale 3,1 % indicate that the impact is negative or somewhat negative. The remaining 20,4% state that impact is neither positive nor negative. Changes in the situation (impact assessment) In the end-line data is it found that the existing tendency of positive attitudes towards the impact of disarmament in the community is sustained and strengthened. Data furthermore suggest that a change in attitude on disarmament impacts on behavior of potential users of small arms and the likeliness of individuals in a community to report knowledge of possession of guns. Evidence The quantitative data indicates the following sustainment and strengthening of a positive attitude towards the impact of disarmament: Table 4: In your opionion, what impact is disarmament making on your community? September 2011 (n=1050) September 2012 (n=554) 19

20 Across the focus groups discussions conducted, there are expressed strong positive attitudes towards disarmament. Given the political context of small arms in Karamoja, this statement is very vulnerable to social desirability bias. It does, however, appear genuine when in several focus group discussion, participants have stated that is has taken time to accept disarmament as a positive phenomenon in the community. In the majority of qualitative responses the indication of a positive attitude is accompanied an argument of that a lower presence of small arms leads to fewer gun-related incidents and increased sense of safety. In the qualitative data, community sensitizations on small arms is strongly highlighted as a contributing factor to a change in attitude to and usage of firearms and other weapons. Since community sensitization was far from fully implemented at the time of end-line data collection it is surprising that the activity is highlighted to the extent that it is. Focus group responses indicate that sensitization activities has transcended impact on attitudes and also affected behavior of individuals prone to resort to use of weapons. The reasons stated is a combination of relevance of the substantial approach, where messages highly resonated with the daily lives of beneficiaries and the involvement of young, potential users of small arms as advocates of the negative consequences of use of firearms. 4. Conclusions and recommendations In this impact assessment report it is found that a positive change in the type and graveness of violence can be observed in the targeted areas. A continuation of low-level violence is found, which within communities is characterized by physical and domestic violence along with incidents involving alcohol affected individuals. At the inter-tribal level, violence is at the level of incidents related to petty thefts involving a limited number of individuals unlike earlier incidents of large clashes between ethnic groups. A major change is traced in the conflict management capacity within communities where conflicts are better managed, constrained and prevented. Strengthened inter-personal skills and community leadership appears to account for this change. A significant improvement is seen in community perception of security and in the sense of safety, which leads to a reduced fear of assaults and an increased focus on development issues rather than security concerns. A crucial change is found in the relationship between security providers and communities, where a positive circle of increased trust, increased cooperation and improved effectiveness in following up incidents has led to increased community appreciation of the presence and work of security providers, which further impacts positively on the prospective demand for state security services and attitude to continued disarmament. Lastly, findings regarding changes in usage of weapons and attitudes towards small arms indicate that communitydriven sensitization activities holds a major potential of significantly impacting risk assessments of potential offenders and motivators of violent actions, and thus prevention of violence. In conclusion, there has been observed a successful augmentation of existing positive trends of peace building in Karamoja. Recommendations: In order to enhance the impact of a potential continuation of the AVR Project in Karamoja, it is suggested that: The area of operation is extended to neighboring Sub-Counties in currently targeted Districts and to other Districts in the Karamoja region. Community-driven sensitization activities are scaled up and outreach of Conflict Management Education is extended to more participants in targeted villages. Involvement of community members in development of Community Safety Plans is broadened. The Parish level can be utilized in bridging efforts at village and Sub-County level. 20

21 Non-formal leaders in the communities are included as a target group in Conflict Management Education Community Safety Committee Members are involved more in cooperation and security issues across Sub- Counties in order to strengthen the preventive dimension of their work and increase synergies between their work. A needs assessment regarding cross-border community activities is carried out and findings are integrated in a potential 3 rd cycle A livelihood component is added to be implemented by a partner with expertise in livelihood support, in acknowledgement of the strong connection between livelihood opportunities and violence reduction The strong entry point in the targeted communities is utilized to direct complementary interventions by other actors to the area, e.g. responses to the challenges of domestic violence, sexual violence and violence arising from excessive alcohol consumption. Advocacy for inclusion of Conflict Management Education in the general training systems in UPDF and UPF is carried out to address the long term challenge of transfer of security providers and make a long term impact on Uganda s security provider s conflict management capacity. Advocacy towards the Kenyan and Ugandan government regarding cross-border cooperation to halt the flow of small arms across countries in the region. 21

22 Annex A: Success story of Nakele Teresa In the mountainous area of Tapac in Moroto Districts lives 23 year old Theresa Nakale. She is a mother and wife, and one of the 20 people in the village Lopelipel that has undergone training in Conflict Management Education. She explains that the training was highly relevant to her daily life, and that she experiences a change in the interaction between men and women in the community. About her own experiences, she says that Before, the husband would cane you because you had not cooked even though it is because there is no food. She adds, In that case the brothers would have to step in to punish the man or it would be presented for the elders who would get young boys to collect cains and straight away punish the man and wife Today, she says, Things are solved in dialogue. The LCI chairperson steps in when there are conflicts. He makes people discus and talk, maybe involves relatives to sort out issues. Neighbors and those who have been given skills intervene when they see someone quarreling. She tells that she experiences that men in her community show women more respect and that husbands and wives communicate more and better. She also says that the increased dialogue has helped to create more cooperation within a family: You now see men taking care of the children when the women is not around. Nakele Theresa was interviewed upon her participation in a Community Regular Meeting with security providers. On the topic of women s role in security, she says that women are now aware of what the men talk about. We feel included and pass on the information between us. She points at a change of behavior that has been instated because of access to new knowledge: Before, women could accelerate the raids from an expectation that the man can provide cows. Now you look more at the risk of your man getting injured or killed, and have become aware of other approaches than forcefully getting what you want 22