Insights from the Kenya Longitudinal Socioeconomic Study of Refugees and Host Communities 2 Building Evidence Building Evidence to Enhance the to Enhance the Welfare of Refugees Welfare of Refugees and Host Communities and Host Communities Building Evidence to Enhance the Welfare of Refugees and Host Communities BUILDING EVIDENCE TO ENHANCE THE WELFARE OF REFUGEES AND HOST COMMUNITIES Insights from the Kenya Longitudinal Socioeconomic Study of Refugees and Host Communities Building Evidence to Enhance the Welfare of Refugees and Host Communities © 2024. World Bank Group This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank Group concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank Group encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org Photo credits: UNHCR (with consent) Building Evidence to Enhance the Welfare of Refugees and Host Communities Table of Contents Acknowledgements i Abbreviations and Acronyms ii Executive Summary xii 1. Background and Context 1 1.1 Refugee hosting context 2 1.2 The need for better data on both refugees and hosts 6 2. Demographic Profile 9 3. Basic Needs 17 3.1 Non-monetary poverty indicators 14 3.2 Food insecurity 19 3.3 Education 21 4. Socioeconomic Resources to Meet Basic Needs 43 4.1 Humanitarian assistance and remittances 43 4.2 Employment 45 4.2.1 Few camp-based refugees are working 45 4.2.2 Factors associated with increased likelihood of being employed 48 4.2.3 Education and skills trainings are perceived to be key to employment 53 4.2.4 Demand-side and other constraints 54 5. (Dis-)enablers to Sustaining Self-Reliance and Resilience 61 5.1 Social cohesion 61 5.2 Socioeconomic and environmental shocks 67 5.3 Psychosocial wellbeing 68 6. Suggestions for Policy Direction 73 6.1 Short run: Unlocking access to socioeconomic opportunities through movement and complementary measures 73 6.2 Medium-term: Easing the transition with place-based development 74 6.3 Sustaining self-reliance and resilience in the long run 75 6.4 Continuing the implementation of a progressive policy and hosting framework for refugees 75 References 76 Annexes 80 Annex 1: Survey Sampling 80 Annex 2. Multidimensional Poverty 83 Annex 3: Additional Figures and Tables 84 Building Evidence to Enhance the Welfare of Refugees and Host Communities List of Boxes Box 1: Spatial disparities in welfare 4 Box 2: Indices to assess food security 19 Box 3: Administrative data in action: education in refugee camps 39 Box 4: Measuring psychosocial wellbeing 68 List of Figures Figure 1: Absolute poverty rate, 2005/06 – 2021 4 Figure 2: Absolute poverty rate by county, 2021 4 Figure 3: Poverty rate and education 5 Figure 4: Poverty rate, access to water and electricity 5 Figure 5: Conceptual framework 7 Figure 6: Composition by camp 9 Figure 7: Household composition 10 Figure 8: Reasons for leaving country of origin 11 Figure 9: Reasons for leaving camps 12 Figure 10: Reasons urban refugees left the camps 12 Figure 11: Proportion of those who want to move from current place of residence 13 Figure 12: Share of individuals that have ever moved since 2016 14 Figure 13: Where refugees want to move to 14 Figure 14: Where hosts want to move to 15 Figure 15: Share of individuals that have moved for more than 6 months 15 Figure 16: Where those who moved and stayed longer than 6 months went 15 Figure 17: Proportion of households in moderate and severe multidimensional poverty 17 Figure 18: Food insecurity and negative coping strategies (% of the population), by location/ refugee status 20 Figure 19: Highest education level among 25–64-year-olds by gender 22 Figure 20: Highest education level among 25–64-year-olds by location 22 Figure 21: Primary school enrollment 23 Figure 22: Share of primary-age children who have never attended school among those not enrolled 24 Figure 23: Primary enrollment by gender 24 Figure 24: Top 5 reported reasons for children not attending school 25 Figure 25: Early Grade Math Assessment 26 Figure 26: English assessment scores 27 Figure 27: Swahili assessment scores 28 Figure 28: Secondary school enrollment rates 29 Figure 29: School enrollment among secondary-school-aged children 30 Figure 30: Child respondent’s educational aspirations and expectations 31 Figure 31: Caregiver’s educational aspiration and expectation for the child respondent 32 Figure 32: Discrepancy between child respondent’s (CR) educational aspiration and expectation 33 Figure 33: Discrepancy between caregiver’s (CG) educational aspiration and expectation for the child respondent 34 Figure 34: Discrepancy between child respondent (CR) and caregiver (CG) educational aspirations 34 Figure 35: Proportion of caregivers who would want their child to take a job before completing secondary school 36 Figure 36: Important for sons to have more education than daughters 36 Figure 37: Daughters should be sent to school only if they are not needed to help at home 37 Figure 38: Proportion of caregivers who would want child to marry before completing secondary school 37 Figure 39: Net Secondary School Enrollment by Gender 38 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 40: Gross enrolment and net enrolment rates by gender and level of education 40 Figure 41: Pupil-teacher ratio by camp and level of education 40 Figure 42: Classroom utilization by camp and level of education 43 Figure 43: Sources of household income 45 Figure 44: Labor force participation 46 Figure 45: Which sectors do refugees work in? 48 Figure 46: Age at the time of displacement 52 Figure 47: Current labor force participation by age at displacement 52 Figure 48: Sankey chart on education or labor force status pre- and post-displacement (for those currently 18-64 years old) 53 Figure 49: The main support needed by those looking for employment is education, vocational training and contacts with employers 54 Figure 50: The main support needed by those looking to become self-employed is vocational training, business skills and credit 54 Figure 51: Salaries of paid employees and profits from own-account work in non-agricultural businesses 54 Figure 52: Proportion of employed refugees who would like to work more hours if they were paid 55 Figure 53: Proportion of refugees who would be able to work more hours in the next two weeks if offered a job 55 Figure 54: Reasons for not looking for jobs 57 Figure 55: Hours spent on childcare and household chores 59 Figure 56: Perception of trust towards out-group 61 Figure 57: Median number of shocks experienced across locations 67 Figure 58: Perceptions of weather variability within the preceding five years 68 Figure 59: Self-reported experiences that may indicate symptoms of depression and anxiety across locations 70 Figure 60: A dual frame design in which frame B (SES) is a subset of frame A (proGres) 80 Figure 61: Primary school attendance rates by location and gender 86 Figure 62: Primary school attendance rates by strata 87 Figure 63: Secondary school attendance rates by location and gender 87 Figure 64: Secondary school attendance rates by strata 88 Figure 65: Top 5 reasons for not attending school 88 Figure 66: Number of Child respondents who speak neither, only English/Swahili, or both. 89 Figure 67: Proportion of Child respondents who speak neither, only English/Swahili, or both 90 Figure 68: Perception of safety and crime 94 Figure 69: Views on socialization of children 96 List of Tables Table 1: Deprivation prevalence (% of households), by location/status 18 Table 2: Percent of households receiving assistance in past 12 months, by location / refugee status/ poverty 43 Table 3: Food insecurity and negative coping strategies (% of population), by refugee status/ year of interview 44 Table 4: Who works and holds better jobs? 49 Table 5: Determinants of LFP for refugees, including extended set of controls 50 Table 6: Regression - Determinants of employment, hours worked, and earnings for women 58 Table 7: Factors correlated with perceptions of trust 63 Table 8: Regression on correlates of feeling like Kenya is home (refugee children) 65 Table 9: Regression on correlates of feeling like Kenya is home (refugee adults) 66 Table 10: Factors correlated with symptoms of depression or anxiety. 71 Table 11: Host sample breakdown per stratum 81 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 12: Overview of survey respondents, questionnaire modules and sampling approaches 82 Table 13: Multidimensional index composition 84 Table 14: Deprivation prevalence (% of population), by status/ sex of the head of household 84 Table 15: Food insecurity and negative coping strategies (% of population), by status/ sex of the head of household 85 Table 16: Percent of households receiving assistance in past 12 months, by programme and location / refugee status 85 Table 17: Percent of MPI poor households receiving assistance in past 12 months, by programme and location / refugee status 86 Table 18: Regression results on academic achievement Math 91 Table 19: Regression results on academic achievement English 92 Table 20: Regression results on academic achievement Kiswahili 93 Table 21: The self-efficacy scale: 94 Table 22: Factors associated with perceptions of safety. 95 Table 23: Correlates of comfort with children socializing with outgroups. 97 Table 24: Regression on correlates of school preferences among refugee children 98 Table 25: Proportion of respondents that have experienced shocks 98 Table 26: Proportion of respondents using coping strategy among those experiencing the shock 99 Building Evidence to Enhance the Welfare of Refugees and Host Communities i Acknowledgments This report is based on the Kenya Longitudinal The report was prepared under the guidance of Pierella Socioeconomic Study of Refugees and Host Communities, Paci and Rinku Murgai (Practice Managers, World Bank a component of the Kenya Analytical Program on Forced Poverty and Equity Global Practice), Tom Bundervoet Displacement (KAP-FD). KAP-FD is a collaboration between (Lead Economist), Keith Hansen (County Director for the World Bank, United Nations High Commissioner for Kenya, Rwanda, Somalia, and Uganda), Anne Bakilana Refugees (UNHCR) and Center for Effective Global Action (Operations Manager for Kenya, Rwanda, Somalia, and (CEGA) at the University of California, Berkeley (UC Uganda), and Marek Hanusch (Program Leader for Kenya, Berkeley). It is funded by PROSPECTS (Partnership for Rwanda, Somalia, and Uganda). Improving Prospects for Host Communities and Forcibly The work benefited from many helpful comments and Displaced Persons) of the Government of the Netherlands. discussions across the World Bank Group, UNHCR and The principal investigators are Edward Miguel, Nistha UC Berkeley including with Utz Pape, Edward Miguel, Sinha, Precious Zikhali, Theresa Beltramo and Utz Pape. Pieter Ventevogel, Caleb Gitau, Benjamin Reese, Miguel Antonia Delius was the project manager during the design de Corral, Farah Manji, Tom Bundervoet, Esther Owelle, and data collection period of this study. Federica Ricaldi, Sarah Stillman, Mansi Kalra, Madeline The preparation of the report was led by Olive Nsababera Langley, Sonia Gomez, Dania Khan, Andrew Agumba, Filip Pongrac, Edith Ingutia, Philip Kinara, Sateesh Nanduri, with guidance from Nistha Sinha and Precious Zikhali. Suhail Awan, Jedediah Fix, John Wagacha Burton and The core team included Mirko Vintar, Abla Safir, Somara Edward Benson. The team gratefully acknowledges Sabharwal, Sebastian Silva-Leander, Beatrice Daniel, comments from José Cuesta, Johannes Hoogeveen and Gabriel N. Camargo-Toledo. Important inputs were and Stephen Wrinkler who served as peer reviewers received from Antonia Delius, Theresa Beltramo, Ibrahima for the report. The team would also like to thank the Sarr, Masud Rahman, Jackline Makandi, Laura Abril Ríos- Kenya Department of Refugee Services (DRS) for their Rivera, Patrick Mutinda, Anne Kittony, Nicolas Polasek valuable feedback. and Sheah Deilami. Kevin McGee provided key technical assistance on sampling and survey weights. We thank Ipsos Kenya Ltd., for their work on the data collection and Caleb Gitau, Antonia Delius, Layna Lowe, Administrative and logistical support during the preparation and Sarah Stillman for coordinating the effort. Lastly, we of this report was provided by Anne Khatimba and Martin extend our sincere gratitude to the survey respondents, Buchara and is gratefully acknowledged. The team thanks whose experiences have provided valuable insights that Charity Nzomo, Keziah Muthembwa, Sambrian Mbaabu, will inform policy and interventions to improve the welfare and Vera Rosauer, for their excellent communications of both refugees and host communities. support. The report was edited by Neville Otuki and designed by Tracy Bett. ii Building Evidence to Enhance the Welfare of Refugees and Host Communities Abbreviations and Acronyms AMREF African Medical and Research Foundation CBC Competency Based Curriculum CCCM Camp Coordination and Camp Management CEGA Center for Effective Global Action CRA Commissioner for Refugee Affairs CRRF Comprehensive Refugee Response Framework CSI Coping Strategy Index CT-OVC Cash Transfer to Orphans and Vulnerable Children CWPM Correct Words per Minutes DRS Department of Refugee Services EA Enumeration Area EGMA Early Grade Mathematics Assessment EGRA Early Grade Reading Assessments FE Fixed effects GAD-7 Generalized Anxiety Disorder Score GCR Global Compact on Refugees GISEDP Garissa Integrated Socio-Economic Development Plan HSNP Hunger Safety Net Program ICT Information and Communication Technology ILO International Labor Organization IGAD Intergovernmental Authority on Development ISCO International Standard Classification of Occupations KAP-FD Kenya Analytical Program on Forced Displacement KDRDIP Kenya Development Response to Displacement Impacts Project KISEDP Kalobeyei Integrated Socio-Economic Development Plan KCHS Kenya Continuous Household Survey KCSE Kenya Certificate of Secondary Education K-LSRH Kenya Longitudinal Socioeconomic Study of Refugees and Host Communities Ksh. Kenya Shillings Building Evidence to Enhance the Welfare of Refugees and Host Communities iii LCS Livelihoods Coping Strategies LFP Labor Force Participation MPI Multidimensional Poverty Index MSPSS Multidimensional Scale of Perceived Social Support NGO Nongovernmental Organization NHIF National Health Insurance Fund NSSF National Social Security Fund OAU Organization of African Unity OPTC Older Persons Cash Transfer PCA Principal Component Analysis PEELP Primary Education Equity in Learning Program PHQ-8 Patient Health Questionnaire Score PPS Probability Proportional to Size proGres Profile Global Registration System, the UNHCR refugee database PROSPECTS Partnership for Improving Prospects for Host Communities and Forcibly Displaced Persons PTSD Post-Traumatic Stress Disorder PWSD-CT Persons with Severe Disability Cash Transfer RRPS Rapid Response Phone Survey RSD Refugee Status Determination SES Socioeconomic Surveys TVET Technical and Vocational Education and Training UNHCR United Nations High Commissioner for Refugees USAID United States Agency for International Development WASH Water, Sanitation, and Hygiene WFP World Food Program xii Building Evidence to Enhance the Welfare of Refugees and Host Communities Executive Summary The Kenya Longitudinal Socioeconomic refugees in Kenya continue to face poverty and Study of Refugees and Host Communities limited prospects for economic opportunities. (K-LSRH) is the first nationally comparable survey of registered refugees and hosts in Findings from K-LSRH underscore the need Kenya, offering a unique opportunity to inform for targeted policy interventions to enhance the socioeconomic integration of refugees, the integration of refugees and improve the including the Government of Kenya's proposed wellbeing and resilience of both refugees and shift from camps to integrated settlements. host communities. Three key findings emerge: Unlike previous surveys that lacked comparable First, basic needs are unmet for both refugees host-refugee data, K-LSRH includes both refugee and hosts, with Turkana hosts facing significant and host communities, providing comprehensive challenges in various aspects of wellbeing. Despite insights into their living conditions and challenges. high educational aspirations and support for gender The first wave of K-LSRH was conducted between equality in education, secondary school enrollment June 2022 and 2023, covering refugees in Kakuma is hindered primarily by limited financial resources. Refugee Camp, Dadaab Refugee Complex, Second, both hosts and refugees encounter barriers Kalobeyei Integrated Settlement, as well as urban in accessing employment, with women being refugees in Nairobi, Mombasa, and Nakuru. The disproportionately affected. Third, camp residents, survey instrument addresses various household- both refugees and hosts, experience more frequent level, individual-level, and children’s outcomes. shocks compared to their urban counterparts, with It delves into less understood themes such as lower socioeconomic resilience levels. Overall, psychosocial wellbeing and trust, aiming to advance these results underscore the importance of knowledge and programming in these areas. enhancing refugee service provision and resilience as part of a broader strategy to address and reduce The survey is particularly insightful considering spatial inequalities in Kenya. the evolution of Kenya’s policies towards a development approach, aiming for refugee To support the socioeconomic integration self-reliance. The Kalobeyei Integrated Socio- of refugees in Kenya, a phased approach Economic Development Plan (KISEDP), initiated can be considered. In the short run, easing in 2016 in Turkana County, exemplifies the shift restrictions on the right to work and movement aiming to transition from short-term aid to longer- is essential. Streamlining procedures for work term development initiatives. Kenya also has made permits, movement passes, and banking access significant strides in aligning its legal framework with can help refugees transition toward self-reliance, international standards, evident in the adoption of reducing hosting costs over time. Upskilling, job the Refugee Act 2021 and its Regulations, replacing search support, and self-employment opportunities, previous legislation from 2006. The Act permits especially for women and youth, are also critical refugees to open bank accounts, own SIM cards in empowering refugees to integrate into the labor and apply for work permits. market. In the medium term, addressing spatial inequalities through place-based development in Despite these advancements, the results refugee-hosting areas can enhance welfare for both show that refugees still encounter challenges refugees and host communities and strengthen accessing services and improving their welfare. household resilience to shocks. Incorporating Restrictions on movement and employment psychosocial support is also essential to building opportunities outside of refugee camps, coupled resilience in the long run. Lastly, sustained self- with the difficulty of obtaining work permits and reliance and resilience in the long run require business licenses, continue to contribute to their continued implementation of a progressive policy reliance on humanitarian aid. As a result, many framework and robust data. Building Evidence to Enhance the Welfare of Refugees and Host Communities xiii ©UNHCR/Samuel Otieno iv Building Evidence Building Evidence to Enhance the to Enhance the Welfare of Refugees Welfare of Refugees and Host Communities and Host Communities ©UNHCR 01 Background and Context © UNHCR/Mohamed Maalim Building Evidence to Enhance the Welfare Building Evidence to Enhance the Welfare of of Refugees Refugees and and Host Host Communities Communities 1 Background and Context The rise in forced displacement and the recognition of 774, 370 registered refugees and asylum seekers who live in its protracted nature, necessitates a shift from viewing designated camp areas and urban areas (UNHCR 2024b). forced displacement solely as a humanitarian concern The designated camp areas include Kakuma Refugee to better bridging the humanitarian-development Camp and Kalobeyei Integrated Settlement in Turkana nexus. The number of displaced people globally has County and Dadaab Refugee Complex in Garissa County nearly doubled from 59.2 million in 2014 to 117.3 million while the main urban areas include Nairobi, Mombasa, and in 2023 (UNHCR 2024a). Addressing the immediate, mid- Nakuru, among others (UNHCR n.d.). United Nations High term and long-term needs of forcibly displaced populations Commissioner for Refugees (UNHCR) data suggest that remains critical. Recent attention has been directed to the the Kakuma camp and Kalobeyei Settlement host 288,206 need to expand support from the provision of assistance to refugees and asylum seekers while the Dadaab complex ensuring forcibly displaced people and host communities hosts 382,658, constituting 37 percent and 49 percent can develop sustained resilience (World Bank 2023a). At of all registered refugees and asylum seekers in Kenya the same time, there is an acknowledgement that countries respectively (UNHCR 2024b). All the urban areas host provide a valuable global public good by hosting refugees, 103,506, representing 13 percent of all registered refugees and therefore it is crucial that support be targeted at both and asylum seekers. Dadaab and Kakuma refugee camp displaced populations and host communities (World Bank have been operational for decades since 1991 and 1992 2023a). respectively and they are among the five biggest refugee camps globally (UNHCR 2023a). An estimated 49 percent Kenya is a major refugee-hosting nation and is home of registered refugees are female while women and children to some of the world’s largest refugee camps. Kenya together make up 80 percent of registered refugees hosts refugees from protracted displacement situations (UNHCR 2024b). from neighboring countries, some of which have lasted more than three decades.1 As of May 2024, Kenya hosts 774,370 As of May 2024, Kenya hosts 774,370 registered refugees and asylum seekers who live in designated camp areas and urban areas . ©UNHCR/Mohamed Aden Maalim 1 Estimates from global refugee figures in 2022 suggest that Kenya hosts 8 percent of the total Sub-Saharan refugee population (Sarzin and Nsababera, forthcoming). 2 Building Evidence to Enhance the Welfare of Refugees and Host Communities to enhance self-reliance by adopting a ‘whole of society’ 1.1 Refugee hosting context approach. Kenya was selected as one of the pilot countries The refugee policies, laws and the regulatory for the CRRF implementation, which it signed in December environment in Kenya have evolved over time. During 2017, and committed to review its policies and practices the 1950s to 1970s, refugee inflows into Kenya resulted on refugees (O’Callaghan et al. 2019). By this time, Kenya from the struggle for independence across the continent had already begun to implement local area multi-sectoral (Nyanduga 2004), due to tumult in Uganda and civil strife plans which sought socioeconomic integration of refugees in Ethiopia, Somalia, and Sudan (Abuya 2007). During in Turkana County (Dick and Rudolf 2019). this period, the Government of Kenya applied prima facie refugee status determination, an approach used to declare Kenya has also engaged in regional initiatives and refugeehood in situations of mass inflows, especially ratified conventions to enhance protection and when the need for protection is apparent, and individual livelihoods of refugees. Kenya’s ratification of the determination would take time. This approach was easy to ‘Organization of African Unity (OAU) Convention Relating apply in the context of African hospitality and the rise of Pan- to Specific Aspects of Refugee Problems in Africa’ in Africanism. However, in the 1980s, refugee flows increased 1974 showed its commitment to granting refugees those further due to widespread ethnic conflicts and political rights spelt out in the 1951 Refugee Convention. More instability across the region, which strained the existing recently, Kenya has committed to improve the livelihoods lean asylum system at the Thika Reception Center. Many of refugees by agreeing to several declarations led by the other factors, such as reduced donor support for long-term Intergovernmental Authority on Development (IGAD), some refugee assistance, exacerbated the situation. In 1992, of which were partly supported by the CRRF process. The Kenya provided camp spaces and directly partnered with first was the 2017 Nairobi Declaration on Durable Solutions the UNHCR to support the functions of refugee reception, for Somali Refugees and Reintegration of Returnees in registration, and assistance. The UNHCR was expected Somalia which sought to enhance refugee education, to provide humanitarian support to refugees and support training and skills development, the free movement of Camp Coordination and Camp Management (CCCM) in the refugees and granting residence permits or citizenship for country. refugees (Intergovernmental Authority on Development 2017). The second was the 2017 Djibouti Declaration on Kenya has made considerable progress in the Regional Refugee Education in which Kenya committed to international legal arena in terms of domesticating expand its integration of refugees in the national systems refugee-related policies and protocols. In 1966 and of education to include refugees and host communities in 1981, respectively, Kenya ratified the Convention relating education sector plans, and to expand distance learning for to the Status of Refugees and its Protocol. Articles 17 and refugees. Kenya has fulfilled part of these pledges through 18 of the Refugee Convention address the right to work for the adoption of the Recognition of Prior Learning Policy refugees in host countries, while Article 26 requires states which recognizes skills and competencies acquired in non- to accord refugees the freedom of movement and choice formal or informal learning but where certification is lacking of residence. Other international laws and protocols that (Kenya National Qualifications Authority n.d.). Additionally, Kenya has ratified, which directly or indirectly offer refugees in the 2019 Kampala Declaration on Jobs, Livelihoods and right to work and free movement include the Universal Self-reliance, Kenya pledged to enhance the economic Declaration on Human Rights, the International Covenant inclusion and promote the livelihood opportunities to on Civil and Political Rights, International Covenant on achieve the self-reliance of refugees and host communities Economic, Social and Cultural Rights, and the African (Intergovernmental Authority on Development 2019). Charter on Human and People’s Rights (Malombe et al. 2016). More recently, Kenya has made further progress in the transformation of its national legal and policy In recent years, Kenya has demonstrated its environment by enacting the Refugees Act 2021 and commitment to implement international soft law publishing its accompanying Regulations, as well as instruments on refugee protection. The 2016 New York developing the ‘Shirika Plan’ – a broad agenda which Declaration for Refugees and Migrants and the 2018 Global aims to support the transition from camps to integrated Compact on Refugees (GCR), helped pave the way for the settlements and foster socioeconomic integration. The Comprehensive Refugee Response Framework (CRRF) on 2021 Refugees Act and the 2024 Refugee Regulations how to provide more predictable and comprehensive support replace the 2006 Act and its associated regulations. Under during mass inflows of refugees and migrants, including how the new Refugees Act (the Act), the Commissioner for Building Evidence to Enhance the Welfare of Refugees and Host Communities 3 © UNHCR/Pauline Omagwa Refugee Affairs (CRA) heads the Department of Refugee accounts and own SIM cards. The Act also grants refugees Services (DRS) which receives and processes applications rights to self-employment. However, there are apparent on refugee status determination (RSD; The Refugee Act bottlenecks for refugees to apply for business licenses due 2021). Successful asylum seekers are issued with refugee to the existing challenge of documentation. The Refugees identity cards, one of the six types of refugee identification Act requires that refugees are subject to Kenyan laws, but documents gazetted by the government earlier in 2024 for the 2012 Micro and Small Enterprises Act and the 2010 purposes of accessing public services. Refugee identity Constitution of Kenya do not define refugee rights as they cards expire after five years which might roll back the gains relate to access to cooperatives, employment services, a refugee has made in that time (Vuni and Iragi 2023). as well as registration and ownership of businesses (ILO Although the Act specifies that refugees have the right 2022). Nevertheless, refugees have rights and access to to work in Kenya, refugees are required to have Class M the National Health Insurance Fund (NHIF) as well as basic permits - as per the Kenya Citizenship and Immigration education, and those who have SIM cards access mobile Act 2011, but these are challenging to access in practice. banking services, as well as some services at the National Through the Shirika Plan, the government intends to Social Security Fund (NSSF). transform refugee camps into integrated settlements that Kenya is increasingly adopting the sustainable benefit both refugees and host communities. The plan approach to refugee-related interventions by seeks to enhance the welfare for both refugees and hosts progressively transitioning from a humanitarian- through improved infrastructure, services, and economic focused to a development approach. In 2016, the opportunities. Additionally, the establishment of the two Turkana County government, in conjunction with the municipalities of Dadaab and Kakuma, which include the stakeholders, developed the Kalobeyei Integrated Socio- refugee camps, aims to integrate services and governance, Economic Development Plan (KISEDP), which reflected the enhancing the management and support of both refugees initial transition from humanitarian assistance to longer-term and host communities. development interventions (UNHCR 2018). Although the first As part of integration and financial inclusion efforts, phase of KISEDP (2018 – 2022) faced various challenges, the Refugees Act provides that refugees can open bank the lessons learnt can be used to make the second phase, 4 Building Evidence to Enhance the Welfare of Refugees and Host Communities which concludes in 2030, more successful. Garissa County extensive verification processes before issuance. also plans to replicate Turkana County’s model to roll out Refugees and asylum seekers found outside the the Garissa Integrated Socio-Economic Development Plan designated areas face arrest and prosecution. In addition, (GISEDP; O’Callaghan et al. 2019). due to the challenge of documentation, refugees are quite limited in terms of labor force participation and the Despite Kenya’s progress to improve the legal businesses they can operate. They are often employed and policy environments, refugees still face many as ‘incentive workers’ for humanitarian organizations and challenges that curtail their livelihoods. Freedom of receive lower wages than Kenyans for equivalent work movement for refugees is limited as under the Refugees (Vuni and Iragi 2023; Omata 2021). Refugees also face Act 2021 refugees are required to stay in ‘designated significant financial inclusion barriers including lack of areas’ unless they apply for an exemption. To leave the documentation, geographical barriers, limited knowledge designated area, they are required to apply for movement on financial literacy, and access to existing financial passes, which are valid for a specified period and have services (Vuni and Iragi 2023). Box 1: Spatial disparities in welfare Refugee camps and settlements, where most refugee live, are in underdeveloped arid and semi-arid regions with limited infrastructure and services.2 The recent World Bank’s Kenya Poverty and Equity Assessment (World Bank 2023b), highlights that geography is a significant factor in poverty, with arid regions in the North and North-East,3 where most refugee camps are situated, experiencing much higher poverty rates than the rest of the country. In 2021, Turkana, Kenya’s northernmost county, had the highest poverty rate at 78%, followed by Mandera, Tana River, and Garissa, where over two-thirds of residents are poor. The poverty gap between these arid regions and the rest of the country has remained consistent from 2005/06 to 2021. Despite comprising only 10 percent of the population, these areas account for about 18% of the nation’s poor. Figure 1: Absolute poverty rate, 2005/06 – 2021 Figure 2: Absolute poverty rate by county, 2021 60% 49.7% 50% 43.5% 40.7% 38.8% 37.0% 42.9% % of population 40% 46.7% 41.7% 38.6% 36.1% 30% 34.5% 33.6% 34.1% 29.4% 20% 26.0% 10% 0% 2019 2020 2021 2005/06 2015/16 National Rural Urban Source: 2021 Kenya Poverty Report, based on the 2021 Kenya Continuous Household Budget Survey. Notes: According to the official national poverty lines produced by the Kenya National Bureau of Statistics (KNBS), an individual is considered poor if they live in a household with a monthly per adult equivalent consumption expenditure of Ksh 3,947 and Ksh 7,193 for rural and urban areas, respectively, in 2021 prices. 2 While camp areas have historically been poor arid and semi-arid areas even prior to the establishment of camps, the report does not suggest a causal relationship between refugee hosting and poverty rates. Refugee presence may increase pressure on scarce resources but there is also evidence it attracts inflow of resources and creates new economic opportunities which improves conditions for hosting communities. Thus, the direction and magnitude of refugee presence on poverty is not straightforward. See for example Alix-Garcia et al. (2018) and Zhou et al. (2023). Building Evidence to Enhance the Welfare of Refugees and Host Communities 5 Kenya has made notable progress in improving non-monetary welfare dimensions, but spatial disparities persist. Education, water, and electricity access remain significantly lower in arid counties compared to non-arid regions. The Kenya Poverty and Equity Assessment 2023 reveals that arid counties continue to have the lowest rates of live births delivered by skilled providers: Turkana (53%), Mandera (55%), Wajir (57%), Samburu (57%), and Tana River (59%). These counties are also about 30 percentage points behind the national average in primary school enrollment and face lower transition rates from primary to secondary school. While the national transition rate is high at 87%, driven by the Government of Kenya's policy to achieve 100% transition, most arid counties fall below this average. Figure 3: Poverty rate and education Figure 4: Poverty rate, access to water and electricity 100% 100% 80% 80% 60% 60% 4 0% 40% 20% 20% 0% Poverty rate Primary School Secondary 0% Net Enrolment School Net Poverty rate Access to Access to Enrolment improved el ectricity water sources National average Nairobi county Non-ari d/semi-arid Arid county Turkana county Garissa county Source: Ministry of Education Basic Education Statistical Source: Based on the Kenya Continuous Household Booklet (2020). Survey (KCHS, 2021). Limited progress in poverty reduction in arid areas is partly due to agroclimatic constraints, which lead to high vulnerability to climate shocks and generally low agricultural productivity. This is exacerbated by relatively low access to basic services, which hinders human capital development. Thus, improving welfare of refugees and hosts should be part of a broader strategy to close Kenya’s spatial disparities. Host communities also experience high levels of poverty and require both humanitarian and development assistance. 6 Building Evidence to Enhance the Welfare of Refugees and Host Communities 1.2 The need for better data on both refugees and hosts Better data is needed to understand both the situation level outcomes (labor supply and earnings, access to of refugees and hosts and inform appropriate and financial services, social cohesion, movement to and within evidence-based interventions and policies. Previous Kenya, health and psychosocial wellbeing, marriage and surveys provided some understanding of the socioeconomic fertility, gender norms); and children’s outcomes (learning conditions of refugees.4 However, such surveys did not assessments; aspirations; social cohesion). The K-LSRH include host communities and only allowed for a limited is the first representative survey of registered refugees in comparison and understanding of living conditions and Kenya with comparable host-refugee data.11 Its coverage challenges of refugees and host communities. 5 6 Better of less understood themes such as psychosocial wellbeing, data is critical to inform the socioeconomic inclusion of weather perceptions and social cohesion aims to provide refugees such as the Government of Kenya’s intended data to advance knowledge and inform programming in transition from camps to integrated settlements. It is also these areas. critical to inform humanitarian and development operations aimed at improving the wellbeing of both refugees and host Using the lens of self-reliance and resilience, this communities. report gives an overview of survey findings to inform evidence-based interventions and policies. With the The Kenya Longitudinal Socioeconomic Study of relative increase in and protractedness of displacement, Refugees and Host Communities (K-LSRH) aims to attention is increasingly shifting towards enhancing self- produce comparable socioeconomic panel data on reliance, defined as the ‘social and economic ability of an refugees and host communities to inform evidence- individual, a household or a community to meet its essential based policy, planning and programming. The survey, needs in a sustainable manner and with dignity’ (UNHCR conducted between June 2022 and 2023, covers (i) refugees 2005). Figure 5 presents the conceptual framework for self- in Kakuma camp, Dadaab camp and Kalobeyei settlement; reliance, highlighting fundamental human needs at its core. (ii) refugees in urban areas of Nairobi, (iii) refugees in ‘Other While humanitarian efforts typically focus on meeting these Urban areas’ namely, Mombasa and Nakuru, and (iv) host basic needs, achieving self-reliance entails progressing communities drawn from households within 15 kilometers beyond basic needs by assessing a household's resources of the refugee camps and from urban neighborhoods to meet them. Furthermore, it involves identifying factors where a large share of the urban refugees reside.7 The that enable or constrain households in sustaining these sampling frame for refugee camps is based on UNHCR’s needs over time (outer ring). It should be noted that self- population database (proGres).8 Up to four respondents in reliance goes beyond merely not receiving aid (Leeson 5,841 refugee households and 3,464 host households were et al. 2022). In other words, the aim is not to consider interviewed.9 10 The survey instrument includes questions on individuals as self-reliant solely because they receive no household-level outcomes (demographic characteristics, aid, while neglecting their living standards or protection consumption, dwelling characteristics, food insecurity, needs. Indeed, promoting and viewing refugee self-reliance assistance, income sources, shocks); adult individual merely as a cost-effective strategy could be detrimental to 4 See Sanghi et al. (2016); UNHCR and World Bank (2019; 2020; 2021; 2022); World Bank (2022; 2023c). 5 Of the existing surveys, only the Kenya Continuous Household Survey (KCHS) and the Kenya Rapid Response Phone Surveys (RRPS) included hosts. However, the KCHS only covered Nairobi and is therefore not representative of the camps or other urban areas. The RRPS on the other hand is only representative of individuals with registered phone numbers and only included a reduced number of indicators compared due to its mode of collection. Furthermore, all of these studies rely purely on quantitative approaches. 6 Additionally, besides the RRPS which is limited in representativeness and scope of indicators as noted above, there was a limited understanding of the socio-economic conditions in Dadaab camp which has been under researched as security limits access. 7 The survey does not cover the small percentage (6%) of refugees living outside target strata (See Annex 8.1A). 8 The sampling frame included all registered refugees and asylum seekers in Kenya. Thus, this report focuses on individuals who fall within the legal definition of refugees and asylum seekers even if they may have other reasons for leaving the country of origin (see section 2 on reasons for leaving). 9 For interviews with child respondents, appropriate procedures were followed to ensure adherence to ethical protocols. These included asking for parental or guardian consent to allow the child to participate in the interview, asking the child for assent and explaining that they could decline any part or all of the interview if they want to. An ethical review of the interview plans was also done at the University of California Berkeley and with African Medical and Research Foundation (AMREF). 10 A household respondent, an adult random respondent, a child respondent, and a woman respondent. 11 See Annex A on sampling design. 12 For a detailed critical examination see Easton-Calabria and Omata (2021). Building Evidence to Enhance the Welfare of Refugees and Host Communities 7 refugees’ wellbeing and protection.12 It is thus crucial to To this end, the report provides an overview of the status assess both the meeting and methods of meeting basic of key indicators by location and other key disaggregation needs. In this regard, resilience – the ability, capacity, and levels such as gender for selected issues in each of the flexibility to prepare for, cope with, recover from, and adapt three dimensions (basic needs, resources, (dis-)enablers). to economic shocks – is also important (Barron et al. 2023). Secondly, it explores correlates to shed light on factors that Marginalized groups, such as refugees and poor host may shape observed outcomes with a view to providing communities, are often more exposed to shocks and their policy-relevant insights. Besides the conceptual framework, resilience can determine their recovery from subsequent the selection of themes for each dimension was motivated shocks (Barron et al. 2023). by the need to address less-understood issues such as psychosocial wellbeing and trust (both in-group and out- The analysis therefore aims to foster evidence- group) and for which the survey could be leveraged. While based discussions on policies and interventions to the report does not claim to be exhaustive, it selectively enhance the self-reliance of both refugees and host balances breadth and depth to provide insights that can communities, thereby strengthening their resilience. motivate further analysis using the microdata. Figure 5: Conceptual framework (Dis-) Enablers Resources Basic needs Source: Adapted from the Self-Reliance Initiative.13 The report is structured as follows: The next chapter expectations is also provided in this chapter. The fourth provides an overview of the demographic characteristics of chapter examines the sources of household income. refugee and host communities. The third chapter assesses It focuses on understanding labor-market outcomes in the extent to which households are meeting basic needs. It recognition of the role that employment can play in achieving provides a discussion of non-monetary poverty indicators sustained self-reliance. In the fifth chapter, a discussion of and uses a multidimensional poverty measure to assess the shocks faced by households, psychosocial wellbeing welfare. An extensive examination of educational outcomes and trust is provided. The sixth chapter concludes with for adults and children as well as their aspirations and suggestions for the policy direction. 13 See for reference Leeson, Slaughter, Buscher (2022). 8 Building Evidence to Enhance the Welfare of Refugees and Host Communities ©UNHCR 02 Demographic Profile © UNHCR/Mohamed Maalim Building Evidence to Enhance the Welfare of Refugees and Host Communities 9 Demographic Profile The composition of refugee populations varies homogenously resided by Somali refugees primarily due significantly across different locations. Most refugees to the proximity of Dadaab to Somalia. On the other hand, in the sample live in camps (81 percent), while the refugees from South Sudan make up the highest number remaining 19 percent are in urban areas. In terms of refugees residing in Kakuma and Kalobeyei. of nationality, the Dadaab camp in Kenya is nearly Figure 6: Composition by camp a. Country of origin b. Length of stay in Kenya 100% 100% 17.2% 62.8% 38.3% 94.9% 80% 80% 46.8% 60% 55% 6 0% 49% 97% 40% 40% 71% 2 0% 20% 25% 31% 23% % 0% Kalobeyei Nairobi Kakuma Other Urban Dadaab Kalobeyei Kakuma Dadaab Other Urban Somali South Sudanese Nairobi DRC Ethiopian <2yrs 3-5 yrs 6 -10 y rs 10+yrs Other Source: Authors’ calculations based on K-LSRH 2023. There are also significant differences between camp household heads are more likely to be male (59 percent) and urban locations both among hosts and refugees and never married (45 percent) than other groups. More in terms of age, household composition, and child than half of urban refugee households are single-person dependency ratios. Camp refugees are younger than households while on the contrary more than half of camp urban ones. The median age among camp refugees and refugee households have more than five individuals. hosts is 28 and 33 years respectively, compared to 31 Nuclear families (a woman, a man, and children) are and 30 years for urban refugees and hosts, respectively. more common among hosts than refugees overall. Camp Camp settings also have a larger share of women (55 refugees tend to have several adults with children as their and 51 percent among camp refugees and hosts, more common household composition. Finally, camp respectively) compared to 41 percent and 43 percent settings have a significantly higher child dependency for urban refugees and hosts, respectively. Furthermore, ratio (1.5 and 1.6 among camp refugees and hosts, heads of households are more likely to be female among respectively) than urban settings where dependency camp refugees (54.7 percent). By contrast, urban refugee ratios are less than 1. 10 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 7: Household composition Figure 7: Household composition 100% Elderly only, with children 13 15 Elderly only, no children 27 80% 35 2 adults of the same sex or 3+ adults, 14 with children 2 adults of the same sex or 3+ adults, no children 60% 33 11 1 adult woman & 1 adult man, with children 18 1 adult woman & 1 adult man, no 39 children 40% 1 adult woman, with children 11 15 1 adult woman, no children 20% 38 1 adult man, with children 13 16 18 1 adult man, no children 4 0% Camp Refugees Camp Hosts Urban Refugees Urban Hosts Source: Authors’ calculaZons based on K-LSRH 2023. Source: Authors’ calculations based on K-LSRH 2023. Majority of adults experienced displacement during their forma;ve childhood years and displacement has been protracted, with few aIemp;ng to return. About half of the adults who are currently refugees were less than 18 years of age when they were displaced.14 An es?mated 95 percent of refugees have lived Majority adults of for in Dadaab overexperienced displacement a decade, making duringwhere it the loca?on Although refugees protracted attributeis their displacement displacement highest. Across all to their formative childhood years and displacement several factors, insecurity is the most common reason groups interviewed, aJempts to return to their countries of origin are rare, with only around 3 percent of has been protracted, with few attempting to return. for fleeing their countries of origin. Very few refugees respondents repor?ng aJemp?ng to return to their countries of origin. About half of the adults who are currently refugees were (9 percent) attributed their flight to a single reason; rather less than 18 years of age when they were displaced.14 An the causes of displacement are multifaceted ranging from Although refugees aIribute their displacement to several factors, insecurity is the most common reason estimated 95 percent of refugees have lived in Dadaab safety concerns to climate related reasons.15 Most refugees for fleeing their countries of origin. Very few refugees (9 percent) aJributed their flight to a single reason; for over a decade, making it the location where protracted (73 percent) attributed their flight to at least three reasons. rather the causes of displacement are mul?faceted ranging from safety concerns to climate related displacement is highest. Across all groups interviewed, Nevertheless, lack of safety is the common reason cited. reasons. Most refugees (73 percent) aJributed their flight to at least three reasons.15 Nevertheless, lack attempts to return to their countries of origin are rare, with About half of the refugees knew at least one person in their of safety is the common reason cited. About half of the refugees knew at least one person in their current only around 3 percent of respondents reporting attempting current location, while the other half knew no one. loca?on, while the other half knew no one. to return to their countries of origin. Age at displacement and length of stay is based on the year of earliest arrival of a proGres family member. 14 14 Age at displacement and length of stay is based on the year of earliest arrival of a proGres family member. 15 Three was the maximum number of reasons allowed by the question. The true number of reasons is likely to exceed the maximum allowed by the questionnaire. 15 Three was the maximum number of reasons allowed by the question. The true number of reasons is likely to exceed the maximum allowed by the questionnaire. 19 Building Evidence to Enhance the Welfare of Refugees and Host Communities 11 8:8: Figure Figure Reasons Reasons for for leaving leaving countryof country oforigin origin Refugee relocation program 2% Other/Covid/Declined/Don't know 5% Lack of employment opportunities 9% Lack of access to home/area of housing/area of 9% livelihood/livestock Drought/famine/flood 18% Death of husband/family reasons 18% Lack Lackof accessto ofaccess toeducation services and health education services health services 23% services Increased crime,Risks crime, Increased tosafety Risksto and insecurity safety and not insecurity but not lack of safety 47% lack of safety Lack Lackof of safetyin safety inmy village AND/OR myvillage AND/OR neighboring neighboring 78% villages villages % 10% 20% 30% 40% 50% 60% 70% 80% 90% Source: Source: Authors’ Authors’ calculaZons calculations based based on on K-LSRH K-LSRH 2023. 2023. Among campcamp Among refugees, refugees, those those in Dadaab in Dadaab have different have different experiences more likely to leave regarding the camp (74 the frequency percent do sothey leave at least camps.16 regarding experiences In Kakumathe Kalobeyei, they and frequency over 80 percent leave once ofper the inhabitants year). never Among those who leave the the do leave confines camps, of the the camp. camps. This contrasts In Kakuma with the Dadaab and Kalobeyei, camp, where over 80 percent of refugees most commonare reason more likely to leave is to visit familythe camp (cited (74 by 63 percent percent do so at least the inhabitants neveronce per leave year). the Among confines those of the who camp. 16 do ofleave Dadaab the camps,and refugees the most 35 common percent reason in Kalobeyei), is to visit although This contrasts with the Dadaab camp, where refugees are this is less common among Kakuma refugees family (cited by 63 percent of Dadaab refugees and 35 percent in Kalobeyei), although this is less common (7 percent). among Kakuma refugees (7 percent). ©UNHCR/ Mohamed Aden Maalim 16 Respondents were asked how often they leave the camp. The question did not specify whether to go elsewhere in Kenya or across the border. 16 Respondents were asked how often they leave the camp. The question did not specify whether to go elsewhere in Kenya or across the border. 20 12 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 9: Reasons for leaving camps 100% 7.0% 80% 34.5% Other 63.0% Business / Work 60% Looking for work visit a friend 40% Medical 20% Shopping Visit Family 0% Kakuma Refugees Kalobeyei Dadaab Refugees Refugees Other Business / Work Looking for work Visit Friend Medical Shopping Figure 10: Reasons urban refugees left the camps Visit Fami ly 17.0% 29.7% 23.0% 11.3% 19.0% In search for better economic opportunities Insearch of better In search economic for more opportunities independence Family reasons Medical Insearch reasons of more independence Other Family Medical reasons Reasons Other Source: Authors’ calculations based on K-LSRH 2023. Building Evidence to Enhance the Welfare of Refugees and Host Communities 13 Refugees, particularly those in camps, would like Nevertheless, while conditions may be better in urban to move elsewhere to search for better economic areas, even there a large proportion of refugees (27 opportunities. Among urban refugees, about 1 in 10 of percent) as well as hosts (29 percent) want to move from those in Nairobi and 1 in 3 of those living in other urban there. By contrast, fewer camp hosts (4 percent) express areas previously lived in a camp. By moving to urban a desire to move. Lack of employment opportunities is the areas, camp refugees forego assistance, yet a substantial most frequently cited reason among both refugees and number have done so, indicating the difficult conditions hosts who express a desire to relocate (between 40 to in camps. Urban refugees that used to live in camps left 60 percent across groups). Additionally, camp refugees primarily to search for better economic opportunities commonly cite fear of discrimination or prosecution17 (33 (30 percent) and for family reasons (23 percent). The percent) as reasons for wanting to leave. While hosts desire to leave camps is also evident even among those generally do not wish to leave Kenya, refugees often do who are still in camps, with nearly 40 percent of current with over 90 percent of those wanting to move wishing to camp refugees wishing to leave their current locations. move to a new country. Figure 11: Proportion of those who want to move from current Figure 11: Proportion of place of those who want to move residence Figure 12: Share of in from current place of residence moved 100% 100% 12.3% Share of individuals that has ever moved % of people wanting to leave current 80% 80% 60% since 2016 60% place 37.7% 40% 87.7% 40% 27.1% 29.2% 20% 20% 3.9% 0% 0% Refugees Hosts Refugees Hosts Refugees Camp Urban Cam Never mo Figure 13: Where refugees want to move to Figure 14: Where h 100% 100% 17 Questionnaire did not specify the source of discrimination or prosecution. 8.5% 80% 80% 60% 71.2% 60% 66.4% 14 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 12: Share of individuals that have ever moved since 2016 100% 12.3% 5.8% Share of individuals that has ever moved 16.2% 17.5% 80% 6 0% since 2016 87.7% 94.2% 40% 83.8% 82.5% 20% 0% Refugees Hosts Refugees Hosts Camp Urban Never moved Mov ed Figure 13: Where refugees want to move to 100% 80% 60% 71.2% 98.5% 40% 2 0% 27.6% % 1.1% Camp Urban Go to a new country Go to a new country Move out of camp and apply for residency in Kenya Move to another area in Kenya Move to another country Mov e to another area in Kenya Mov e out of camp and apply for residency in Kenya Source: Authors’ calculations based on K-LSRH 2023. Mov e to another country Building Evidence to Enhance the Welfare of Refugees and Host Communities 15 Figure 14: Where hosts want to move to 100% 8.5% 9.3% 80% 6 0% 66.4% 81.4% 40% 20% 25.1% 0% 9.3% Camp Urban Return to Return to hometown hometown Another areaAnother in Kenya area in Kenya Go to a new country Go to a new country Source: Authors’ calculations based on K-LSRH 2023. Most individuals, whether refugees or hosts, have been lived elsewhere (16.2 percent) than those living in camps living in the same place since 2016, consistent with (12.3 percent). Among those that have lived elsewhere restricted movement under the encampment policy.18 since 2016, camp refugees are more likely to have been in However, there are notable differences between camp and another country (24.6 percent), while most others have only urban refugees. Urban refugees are more likely to have moved within Kenya. Figure 15: Share of individuals that have moved Figure 16: Where those who moved and stayed for more than 6 months longer than 6 months went 100% 100% 80% 80% 60% 75.4% 98.4% 60% 99.5% 92.8% 40% 40% 20% 20% 12.3% 12.0% 24.6% 7.8% 4.4% 16.0% 7.2% 0% 0.5% 0% Camp Camp Hosts Urban Urban Hosts Camp Camp Urban Urban Refugees Refugees Camp Urban Camp Urban Abroad Kenya Source: Authors’ calculations based on K-LSRH 2023. 18 Restrictions of movement generally apply only to refugees. They need to contact the Department Refugee Services (DRS) and apply for a movement pass to move outside of the designated areas. 16 Building Evidence to Enhance the Welfare of Refugees and Host Communities PHOTO/Alamy 03 Basic Needs © UNHCR/Samuel Otieno Building Evidence to Enhance the Welfare of Refugees and Host Communities 17 3. Basic Basic Needs Needs 3. 3.3.Basic 3. Basic Basic Needs Needs Needs Basic Needs 3.1 Non-monetary -monetary 3.1 Non3. poverty indicators Both hosts and refugee communities in camp areas have This Basic sub-sec;on Needs poverty indicators provides an overview of the fulfillment basic of oflevels needs poverty. popula;on, across the survey as 3.13.1 Non-monetary 3.1 3.1 3.1 Non-monetary Non-monetary Non-monetary Non-monetary poverty poverty poverty indicators poverty poverty indicators indicators indicators indicators high multidimensional Figure 17 below This sub-section well as the provisionprovides ofan overview social assistanceof the fulfillment across popula;onshows groups. that there 19isBasica the largeneeds will urban-rural be considered poverty gap. This This This sub-sec;on sub-sec;on sub-sec;on This This sub-sec;on sub-sec;on provides provides providesprovides provides an an an overview overview an anoverview overview overview of of the ofofthe of the fulfillment the the fulfillment fulfillment fulfillment fulfillment ofof basic of basic of of basic basic basic needs needs needs needs needs across across acrossacross across thethe the the survey survey survey survey survey popula;on, popula;on, popula;on, popula;on, popula;on, asas asas as of basic through 3.1 needs the lens Non-monetary across the survey of a mul?dimensional poverty population, indicators as poverty well as index (MPI) groups. 20 Multidimensional, the construc?on poverty is of will up which to 80 is described percentage pointsin well well wellwell wellas asasthe as asthe the theprovision the provision This sub-sec;on provision provision provision ofofofsocial of of provides social social social social assistance assistance assistance assistance assistance an overview across across acrossacross across popula;on of the fulfillment popula;on popula;on popula;on popula;onof 20basic groups. groups. groups. groups. needs 19 19 19 Basic 19 19 19 19 Basic across Basic Basic Basic needsneeds needs the survey needs needs will will bebe will will be popula;on,beconsidered beconsidered considered considered considered asRefugees the more provision detail in of the social annex. assistance The across MPI encompasses population 14 higher indicators in camp across areas six compared dimensions toof urban areas. wellbeing, namely, through through through through through the well the the the the as lens lens lens lens thelensofof ofa of of aa amul?dimensional provision mul?dimensional amul?dimensional mul?dimensional mul?dimensional of social assistance poverty poverty poverty poverty poverty index across index index index index (MPI) (MPI) popula;on (MPI) (MPI) (MPI) 20 , ,the 2020 20,the 20 20 ,,the theconstruc?on the groups. construc?on construc?on construc?on construc?on 19 Basic needs ofof ofwhich of of which which will beis which which described isisdescribed is is described described described considered ininin in in groups.19 Basic educa?on, needs will beenergy, employment, considered through housing, the lens water and of sanita?onliving in other and urban nutri?on. areas outsideFollowing of Nairobi the (i.e., Mombasa conven?on in more more moremore more detail detail detail detail in detail in through in the in the in thethe annex. the the annex. lensannex. annex. annex. The The The TheMPI MPI MPI The MPI of a mul?dimensional MPI encompasses encompasses encompasses encompasses encompasses 14 14 poverty index 14indicators indicators indicators 14 indicators 14 indicators across (MPI) , the 20 across across across across six six dimensions six dimensions dimensions six construc?on dimensions six dimensions of which ofof wellbeing, ofofwellbeing, wellbeing, of wellbeing, iswellbeing, described in namely, namely, namely, namely, namely, a multidimensional poverty index (MPI) 20 , the construction and Nakuru) have the highest urban poverty rate (25.9 the literature, educa?on, educa?on, educa?on, educa?on, educa?on, more households employment, employment, employment, employment, employment, detail in the annex. are energy, energy, energy, considered energy, energy, The housing, housing, housing, MPIhousing, housing, mul?dimensionally water encompasseswater water water water and and and and and 14 sanita?on sanita?on sanita?on sanita?on sanita?on indicators poor and and acrossand and and ifsixthey nutri?on. nutri?on. nutri?on. nutri?on. nutri?on. are dimensions deprived Following Following Following Following Following of wellbeing, the theintheat the the least conven?on conven?on conven?on conven?on namely, onein conven?on third ininin in of which is described in andmore detail in poor annex. the mul?dimensionally The percent). This is still significantly lower than the lowest of weighted the thethe literature, the theliterature, literature, educa?on, literature, literature, indicators households households households employment, households households severely areare are considered energy, are areconsidered considered consideredhousing, considered if deprived and in mul?dimensionally mul?dimensionally water mul?dimensionally mul?dimensionally at least sanita?on poor poor poor andhalf poor poor ififif of they they nutri?on. if iftheythe they they are are weighted are are deprived deprived deprived are Following deprived deprived indicators. in the in at in at in in least atleast least conven?on at at leastone least one one in one one third third thirdthird third MPI encompasses the literature, 14 indicators households across are considered sixifif dimensions mul?dimensionally poverty poorrate in ifoftheycamp are areas, deprived which in at is found least one third Dadaab among of of weighted ofweighted of of weighted weighted weighted indicators indicators indicators indicators indicators and and and and severely severely and severely severely severely poor poor poor poor poor deprived ifdeprived if if deprived deprived deprived inin in atat in in least atleast at at least least least halfhalf half half half of of the of of the the weighted the the weighted weighted weighted weighted indicators. indicators. indicators. indicators. indicators. of wellbeing,of weighted namely, indicators education, and severelyemployment, poor if deprived hosts (73.2 percent). energy, in at least half of the weighted indicators. In urban areas, the difference between Both hosts and refugee communi;es in camp areas have high levels of mul;dimensional poverty. Figure housing, water and sanitation and nutrition. Following the refugees and hosts is statistically significant, although it is 17 Both Both Both hosts below Both Both hosts hosts hosts hosts and and shows and and andrefugee refugee refugee that refugee refugee communi;es communi;es there communi;es is communi;es communi;es a large inin camp in camp urban-rural in in camp camp camp areas areas areas areas areas have have have poverty have have high high high high high levels gap.levels levels levels ofof mul;dimensional of mul;dimensional mul;dimensional Mul?dimensional levels of of mul;dimensional mul;dimensional poverty. poverty poverty. poverty. poverty. poverty. is Figure up Figure Figure to Figure Figure 80 convention Both in hosts and refugee the literature, communi;es households are camp areas have in considered highin limited levels absolute terms by the factpoverty. of mul;dimensional that urban Figure poverty rates 17 17 17below 17 17 below below below below shows shows shows shows shows that that that that that there there there there there is a isis aa is is large largelarge urban-rural urban-rural urban-rural poverty poverty poverty gap. gap. gap. Mul?dimensional Mul?dimensional Mul?dimensional poverty poverty poverty isisis upup uptototo808080 percentage 17 below points showshigher that in camp there isaa alarge large areas urban-rural urban-rural largecompared urban-rural to poverty poverty urban poverty are gap.areas. generally gap. gap. Mul?dimensional Mul?dimensional Refugees Mul?dimensional lower. The living largest poverty poverty poverty in refugee-host other is up to is is urbanup up 80 poverty to to 80 areas80 gap is multidimensionally percentage percentage percentage percentage percentage points points points pointspoor points higher if higher higherthey higher higher ininare in camp camp in deprived camp in camp camp areas areasareas areasin areas at comparedleast compared compared compared compared one to to to urban urban to to urban urban urban areas. areas. areas. areas. areas. Refugees Refugees Refugees Refugees Refugees living living livingliving living inin other in in inother other other other urban urban urban urban urban areas areas areas areas areas outside of Nairobi (i.e., Mombasa percentage points higher in camp areas compared to found and Nakuru) have the highest urban in areas. urbanurban Refugees poverty areas other living rate thanin other (25.9 Nairobi, percent). urbanwhere areas This hosts have is third of weighted outside outside outside outside outside ofof Nairobi of ofNairobi Nairobi of indicators Nairobi Nairobi (i.e., (i.e., (i.e., and Mombasa Mombasa Mombasa (i.e., (i.e., severely Mombasa Mombasa and and poor and and and if Nakuru) Nakuru) Nakuru) deprived Nakuru) Nakuru) havehave have in have have the thethe highest highest highest the the highest highest urbanurban urbanurban urban poverty poverty poverty poverty poverty rate rate rate rate(25.9 rate (25.9 (25.9 (25.9 (25.9 percent). percent). percent). percent). percent). This ThisThis Thisis This isis is is outside of lower s?ll significantly Nairobithan the lowest (i.e., Mombasa and poverty Nakuru) rate have the inacamp highest urban which areas, multidimensional povertypoverty is rate found(25.9 rate ofamong percent). 7.3 Dadaab percent, This is hosts compared to s?ll ats?ll significantly s?ll least (73.2 s?ll s?llsignificantly significantly half s?ll of significantly significantly percent). the weighted significantlyInlowerlower lower lower lower urban thanthan than lower than thanthe the indicators. than areas, the lowest lowest lowest the thethethe lowest lowest lowest poverty poverty poverty differencepoverty poverty poverty rate rate raterate rate in rate between in in incamp camp in in camp camp camp camp areas, areas, areas, refugees areas, areas, areas, which which which which which which and isisisis found found hosts found is is found found foundis among among among among among among Dadaab Dadaab sta?s?cally Dadaab Dadaab Dadaab Dadaab hosts hosts hostshosts hosts hosts significant, 25.9 percent for refugees. (73.2 (73.2 (73.2 (73.2 (73.2 although percent). percent). percent). (73.2 it is percent). percent). percent). limited InIn Inurban In In urban urban inIn urban urban areas, urbanareas, absolute areas, areas, areas, areas, the the termsthe thethe thedifference difference difference by difference difference difference the fact between between between between between between that urban refugees refugees refugees refugees refugees refugees poverty and and and and andhosts and rateshostshosts hostshosts hosts are isis is sta?s?cally isissta?s?cally sta?s?cally is generally sta?s?cally sta?s?cally sta?s?cally significant, significant, significant, significant, lower. significant, significant, The largest although although although although althoughalthough itit it isis it it is limited is is it limited is limited limited limitedlimited in in in in in in absolute absolute absolute absolute absolute absolute terms terms terms terms terms terms bybyby by the bythe by the thefact the the fact fact factthat fact fact that that thatthat that urban urban urbanurban urban urban poverty poverty poverty poverty poverty poverty rates ratesrates rates are rates ratesare are generally aregenerally are are generally generally generally generally lower. lower. lower.lower. The lower. lower. largest The TheThe largest The Thelargest largest largest largest refugee-host refugee-host poverty gap poverty is gap foundis found in urban in urban areas areas other other than than Nairobi, Nairobi, where where hosts hostshave have a mul?dimensional a mul?dimensional refugee-host refugee-host refugee-host refugee-host refugee-host poverty poverty poverty poverty poverty gap gapgap is gap gap isisfound found is is found found found inin urban in urban in in urban urban urban areasareas areas areas areas other other other other other than than than than thanNairobi, Nairobi, Nairobi, Nairobi, Nairobi, where wherewhere where where hosts hosts hosts hosts hosts have have have have have a mul?dimensional mul?dimensional aamul?dimensional aa mul?dimensional mul?dimensional poverty rate ofrate poverty 7.3 percent, of 7.3 percent,comparedcompared to to 25.925.9 percent percent for for refugees. refugees. poverty poverty poverty poverty rate rate poverty rate Figurerate ofof rate of 7.3 of 17: 7.3 7.3 of 7.3 percent, percent, percent, 7.3 percent, Proportion compared compared percent, compared ofcompared compared to households tototo25.9 25.9 25.9 in percent percent percent to 25.9 25.9 percent percent for moderate for for refugees. for refugees. refugees. refugees. for refugees. and severe multidimensional poverty 17: Proportion Figure Figure of households 17: Proportion in of households in moderate and moderate and severe severe multidimensional multidimensional poverty poverty Figure Figure Figure 17: 17: 17: Figure Figure 17: 100% Proportion Proportion Proportion 17: of of Proportion Proportion households ofhouseholds of ofhouseholds households householdsin in moderate in moderate in inmoderate and moderate moderate and and severe severe severe and and multidimensional multidimensional multidimensional severe severe multidimensional multidimensional poverty poverty poverty poverty poverty 100% Moderate 100% 100% 100% 100% 100% Moderate Moderate Moderate Moderate Severe Moderate Moderate 80% Severe 80% 80% 80% 80% Severe Severe Severe Severe Severe 80% 80% 60% 60% 60% 60% 60%60%40% 60% 40% 40% 40% 40%40%20% 40% 20% 20% 20% 20%20% 0% 20% Dadaab Kalobeyei HostsTurkana Urban Hosts Refugees Refugees Refugees Refugees Nairobi Refugees Kakuma Dadaab NairobiHosts Nairobi Hosts Hosts Hosts Urban Other 0% 0% 0%0% 0% 0% Other Dadaab Kalobeyei Turkana Urban Hosts Refugees Refugees Refugees Refugees Dadaab Dadaab Dadaab Kalobeyei Kalobeyei Kalobeyei Turkana Turkana Hosts Hosts Hosts Refugees Refugees Refugees Refugees Refugees Refugees Refugees Refugees Refugees Refugees Nairobi Nairobi Refugees Refugees Refugees Refugees Kakuma Dadaab Dadaab Kalobeyei Kalobeyei Turkana Turkana Hosts UrbanHosts Refugees Refugees Refugees Refugees Refugees Refugees Refugees Refugees Nairobi Nairobi Dadaab Kakuma Kakuma Hosts Refugees Refugees Dadaab Dadaab Nairobi Hosts Hosts Hosts Hosts Nairobi Nairobi Kakuma Kakuma Hosts Hosts Hosts Hosts Dadaab Dadaab Urban Hosts Hosts Nairobi Nairobi Urban Urban Urban Other Hosts Hosts Other Other Other Hosts Hosts Urban Urban Other Other Other Other Other Other Other Other Source: Authors’ calculaZons based on K-LSRH 2023. Urban Urban Urban Urban Notes: Moderate > 1/3rd weighted deprivaZons; severe > 1/2 weighted deprivaZons. Source: Source: Source: Source: Source: Source: Authors’ Authors’ Authors’ calculaZons calculaZons calculaZons Authors’calculaZons Authors’ calculaZonsbased calculaZons based based based on based basedon on on K-LSRH K-LSRH on on K-LSRH K-LSRH K-LSRH 2023. 2023. 2023. 2023. 2023. 2023. There are some significant Moderate Notes: Notes: Notes: Moderate> Moderatedifferences >1/3 1/3 > 1/3 rd rd weighted rd rd weighted rd in deprivaZons; weightedmul;dimensional deprivaZons; deprivaZons; severe severe poverty severe > > 1/2 > 1/2 weighted 1/2 rates weighted by weighted sex of household head deprivaZons. deprivaZons. deprivaZons. Notes: Moderate Notes: Notes: > 1/3 Moderate Moderate >> 1/3 1/3 rd Source: weighted rd rd rd rd weighted weighteddeprivaZons; Authors’ deprivaZons; deprivaZons; calculations severe severe severe based on > >1/2 > 1/2 1/2 K-LSRH weighted weighted weighted 2023. deprivaZons. deprivaZons. deprivaZons. among refugees. In urban areas, female headed refugee households have higher mul?dimensional Notes: Moderate > 1/3rd weighted deprivations; severe > 1/2 weighted deprivations. There There There There There There are are are aresome are aresome some somesome somesignificant significant significant significant significant significant differences differences differences differences differences differences in in in mul;dimensional in mul;dimensional mul;dimensional in in mul;dimensional mul;dimensional mul;dimensional poverty poverty poverty poverty poverty poverty rates rates rates by by rates rates rates by sex by by by sex of of sex sex sex of of of household household household household household household head head headhead head head among among among among amongrefugees. among refugees. 19 refugees. refugees. refugees. refugees.In Since we lackIn In In urban Inurban In urbanareas, urban urban areas, areas, urban areas, female areas, areas, a pre-assistance female female female female female baseline headed headed headed headed headed and aheaded refugee refugee refugee refugee refugee refugee non-assistance households households households control households households households have have have higher higher higher have have group, it is not possiblehigher highermul?dimensional mul?dimensional mul?dimensional mul?dimensional mul?dimensional have to assessmul?dimensional what impact social assistance is having on the fulfillment of needs, nor whether social assistance is adequately targeted to those most in need. This is because it is not possible to distinguish whether low observed poverty is due to aid being effective in reducing poverty, or whether it is poorly targeted and reaching the wrong (non-poor) beneficiaries. Since 19 Since Since wewewe 1919 19 19 19 lack a lack lacka a pre-assistance pre-assistance a pre-assistance baseline baseline baseline andand and aa anon-assistance non-assistance non-assistance control control control group, group, group, itit is itisis not not not possible possible possible toto to assess assess assess what what what impact impact impact 1919 19 Since Since Since we lack we we lack lack pre-assistance aa pre-assistance pre-assistance baseline baseline baseline and a non-assistance and and control a a group, non-assistance non-assistance it is not possible to control control assess what group, group, impact social it it is is not not assistance possible possible is having on to to the assess assess fulfillment what what of needs, impact impact nor whether 18 Since social we20 lack assistance The a pre-assistance MPI is adequately approach used targeted to those baseline here most adapts in need. This isa and non-assistance exis^ng because work it is on the not possible controlglobal group, to distinguish MPI by whether it low isthenot observed possible Oxford Poverty poverty to is due toassess and aid being what effectiveimpact Human in reducing social social social social social assistance assistance assistance assistance assistance isishaving having is having is is having havingon ononthe on on the the fulfillment the the fulfillment fulfillment fulfillment fulfillment ofof ofneeds, needs, of of needs, needs, needs, nor nor nor whether whether whether nor nor whether whether social social social assistance social social assistance assistance assistance assistance adequately isisisadequately adequately is is adequately adequately targeted targeted targeted targeted targeted tototo those those to to those social poverty, assistance or whether is having it is poorly Development targeted Ini^a^ve onreaching the and (OPHI) fulfillment the to consider of wrong (non-poor) needs, the nor beneficiaries. context, whether including the social assistance methodological is adequately challenges of es^ma^ng targeted and those tothose those most mostmost most most inin need. inneed. in in need. need. need. comparing This ThisThisis This This isis because monetary because is isbecause because because povertyititis itis it itis not not is is in not not not not forced possible possible possible possible possible to displacement totodistinguish to to distinguish distinguish distinguish distinguish whether whether whether whether whether low low low lowobserved low observed observed observed observed poverty poverty contexts. whether low observed poverty is due to aid being poverty poverty poverty isis due is due is is due dueto due toto aid aid to to aidbeing aid aidbeing being being most in need. 20 The MPI approach This used is because here adapts existing it is work on the possible global MPI to distinguish by the Oxford Poverty and Human Development Initiative (OPHI) to consider the context, being including the effective effective effective in effective effective methodological inin reducing reducing in in reducing reducing reducing challenges poverty, poverty, poverty, poverty,or poverty, of estimating or whether or orwhether or and comparingwhether whether whetheritit monetary is it is itpoorly itis povertypoorly ispoorly is poorly poorly in targeted targeted forced targeted targeted targeted and displacementand andreaching and and reaching reaching reaching reaching contexts. the wrong the thewrong the thewrong wrong wrong (non-poor) (non-poor) (non-poor) (non-poor)beneficiaries. (non-poor) beneficiaries. beneficiaries. beneficiaries. beneficiaries. effective in reducing poverty, or whether it is poorly targeted and reaching the wrong (non-poor) beneficiaries. 20 2020 20 The The 20 20 20 20 The The The MPI MPIMPI MPI MPIapproach approach approach approach approach used used used usedhere usedhere here here here adapts adapts adapts adapts adapts exis^ng exis^ng exis^ng exis^ng exis^ng work work work work work on 24ononthe on onthe the the the global global globalMPI MPI global global MPIby MPI MPI by by the the by by theOxford Oxford the theOxfordPoverty OxfordPoverty OxfordPoverty Poverty Poverty and and andHuman and and Human Human Human Human 20 The MPI approach Development Development Development Ini^a^ve Ini^a^ve Ini^a^ve used (OPHI) (OPHI) (OPHI) here toto consider to adapts consider consider theexis^ng thethe context, context, work context, on the including including including the the global the MPI by the methodological methodological methodological Oxford Poverty challenges challenges challenges of of of Human and and es^ma^ng es^ma^ng es^ma^ng and and Development Development Ini^a^ve Ini^a^ve (OPHI) (OPHI) to to consider consider the the context, context, including including the the methodological methodological challenges challenges of of es^ma^ng es^ma^ng and and Development comparing comparing comparing Ini^a^ve monetary monetary monetary (OPHI) poverty poverty poverty in to inin consider forced forced forced the context, displacement displacement displacement including the methodological challenges of es^ma^ng and contexts. contexts. contexts. comparing comparing monetary monetary poverty poverty in in forced forced displacement displacement contexts. contexts. comparing monetary poverty in forced displacement contexts. 18 Building Evidence to Enhance the Welfare of Refugees and Host Communities There are some significant differences in groups, including hosts, to live in crowded housing multidimensional poverty rates by sex of household (21.5 percent). Refugees living in camps appear almost head among refugees. In urban areas, female headed assured to have access to adequate water sources, with refugee households have higher multidimensional poverty deprivation rates around or below 1 percent in all camps. rates (16.1 percent for female-headed compared to 9.7 Turkana hosts are particularly lagging behind, and percent for male-headed households). Female-headed they have the highest multidimensional poverty rate of households are disproportionately likely to have no the groups considered here (86 percent). Turkana hosts employed household members. The gender difference also stand out in having a significantly larger proportion is statistically significant among both urban and camp of the population living in severe multidimensional refugees. Urban female-headed refugee households are poverty (68.9 percent), compared to other groups. The also significantly more likely to use solid cooking fuel high level of severe multidimensional poverty among (27.5 percent) than male headed ones (12 percent). Solid Turkana hosts is because they have higher deprivation cooking fuel is detrimental to health. rates than other groups for almost all indicators included Education outcomes, housing conditions, and access in the multidimensional poverty index (Table 1). One of to water sources for host communities in camp areas the few exceptions is housing, where Turkana hosts have are lagging behind. Fewer Turkana refugees (Kakuma the second highest deprivation rate (37.9 percent) after and Kalobeyei), for instance, are deprived in education Kalobeyei refugees (53.6 percent). The other exception is (27 percent), compared to most other groups, bar Nairobi employment, where Turkana hosts have lower deprivation hosts (21.8 percent). Additionally, fewer Turkana refugees rates (45.1 percent), i.e., are more likely to have employed live in precarious constructions (<2 percent, compared to household members, compared to most refugee groups, 55.5 percent for Turkana hosts). Refugees in the Dadaab except Nairobi refugees (31.5 percent). camp, on the other hand, are less likely than other rural Table 1: Deprivation prevalence (% of households), by location/status Dimension Deprivation Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other (r) (r) (h) (r) (h) (r) (h) urban urban (r) (h) Education School 27.7% 27.0% 68.5% 79.2% 74.9% 32.9% 21.8% 34.7% 21.3% attendance Employment Paid employment 79.7% 83.2% 45.1% 62.5% 53.8% 31.5% 14.4% 68.2% 11.8% Energy Cooking fuel 99.3% 98.2% 99.3% 97.3% 96.1% 11.7% 2.1% 68.2% 37.6% Electricity 74.5% 64.2% 85.3% 42.7% 44.1% 0.4% 1.0% 0.9% 0.6% Housing Crowding 30.8% 53.6% 37.9% 21.5% 27.3% 15.8% 8.9% 5.8% 4.0% Construction 1.9% 1.2% 55.5% 13.9% 15.7% 0.2% 0.0% 0.0% 0.0% WASH Water 0.2% 1.4% 41.1% 0.7% 4.2% 6.8% 7.6% 17.3% 26.5% Toilet 61.2% 40.4% 83.5% 61.9% 56.5% 7.9% 3.4% 5.1% 18.4% Nutrition Not enough 9.6% 17.6% 27.2% 6.0% 0.9% 14.5% 7.8% 5.7% 1.1% Less preferred 10.9% 17.4% 36.4% 5.2% 2.3% 29.8% 21.6% 11.9% 3.4% Borrowed 2.9% 2.2% 11.4% 2.0% 0.7% 5.6% 3.0% 3.6% 0.7% Smaller portions 9.6% 17.7% 25.1% 3.9% 2.1% 12.7% 10.3% 6.5% 1.5% Adult meals 3.2% 3.6% 14.4% 3.3% 0.8% 4.8% 3.4% 3.4% 0.7% Fewer meals 11.5% 22.2% 39.1% 5.6% 1.5% 14.4% 12.5% 6.9% 1.3% Poverty >1/3rd depriv. 81.5% 82.7% 86.0% 74.6% 73.2% 10.7% 3.8% 25.9% 7.3% >1/2 depriv. 37.9% 38.0% 68.9% 42.2% 39.5% 1.6% 0.1% 4.5% 0.3% Source: Authors’ calculations based on K-LSRH 2023. Notes: r = refugees and h=hosts. Construction of indicators is provided in the Annex 2. Building Evidence to Enhance the Welfare of Refugees and Host Communities 19 © UNHCR/Charity Nzomo hosts have the highest level of food insecurity, as measured 3.2 Food insecurity by the CSI (87.6 percent) and the lowest future coping Turkana hosts also have the highest level of food capacity as measured by the LCS (63.5 percent). Turkana insecurity and lower future coping capacity. Figure 18 refugees, who had similarly high levels of multidimensional below shows food insecurity levels across locations and poverty, also have high levels of food insecurity, with 64.3 refugee status, using the Coping Strategy Index (CSI) and percent and 72.3 percent of the population having a CSI the Livelihoods Coping Strategies (LCS) index.21 Turkana score of 4 or more in Kakuma and Kalobeyei, respectively.22 Box 2: Indices to assess food security This report uses two different indices to assess food insecurity: This report uses two different indices to assess food insecurity: 1. The Coping Strategy Index (CSI) gives an indication of current food security at the time of the survey, This This report usesuses report two two different different indicesindices to assess to assess foodfood insecurity: insecurity: 1. by Thelooking at the frequency Coping Strategy of 6 an Index (CSI) gives common indication coping of current strategies food security related at theto food time consumption. survey, by An index score of the 1. 1.Theoflooking 4 to The Coping at 9 Coping the is frequency meant Strategy Strategy Indexto of (CSI) (CSI) 6 indicate Index common gives the gives an coping anpresence indication indication strategies of current of related of moderate currentfoodfood to food securityfood consumption. security atinsecurity, attime the the time An index whereas of theofsurvey, score a the survey, by of by4 of 10 or higher score to looking looking is meant 9 at indicates to the atsevere the indicate frequency frequency of 6the food of 6presence commoncommon insecurity. ofcoping coping The moderate strategies CSI strategies food is composedrelatedinsecurity, relatedto foodto food of whereas the consumption. consumption. 2nd score a to An the of 10 An or index 6th higher index score score nutritionof 4 of 4 indicators used in 9indicates to theto 9 is meant is meant MPI severe to food indicate to indicate (See Annex insecurity. the presence 2).the The However,of CSI presence of is moderatemoderate these composed food indicators of the food insecurity, 2nd to insecurity, are the whereas nutrition whereas assembled6tha a score scoreinofaindicators 10ofor10 higher slightlyor used higher in the different way to form a MPI above. indicates indicates severe However, severe foodfood these insecurity. indicators insecurity. Theis The CSI are CSI isassembled composed composed of thein ofa 2 slightly nd 2nd to the to the different 6the way indicators nutrition to form th 6th nutrition a composite indicators usedused in the index: in the composite index: # MPI MPI above.above. However, However, these these indicators indicators are assembled are assembled in a slightly in a slightly different different way way to form to form a composite a composite index:index: " # !  % #   $ ∙ !$ " $ $$ $ !" !  %  % $"% ∙ ! ∙ !     $ $"% $"% $ Where is the weight for the  &' nutrition indicator and ! is the number of days in which $ $ &'   &' nutrition indicator and $ $ Where WhereWhere household is is the the is the weight experienced weight weight for forforthe the the the   deprivation   nutrition nutrition in the past indicator indicator week. and Nutrition and ! is ! isnumber the number the is indicators the of2, days number of days in4, inof and which which in Error! 6days in which household Reference household i experienced the deprivation in the past week. Nutritionindicators household source experienced not found. experienced the the receive deprivation deprivation a in weight in the of the past 1 in past week. the week. CSINutrition Nutritionindex, whereas indicator indicators indicators 2, 2, and 4,2, 4, 4,6 ingets 3 and and 6a 6 in Error! weight Error! in Table 13 receive a of 2, and Reference Reference source indicator source 5 gets not not found. a weight found. receive receive aof 3. a weight weight of 1 of 1 inCSI theindex, in the CSI index, whereas whereas indicator indicator 3 gets3 gets a weight a weight of 2,of and and indicator 2, indicator 5 gets 5 gets a weight a weight of 3.of 3. 21 The CSI aggregates 6 coping 2. Thestrategies related to Livelihoods food consumption, Coping Strategies including whether (LCS) index aims the household to had to limit portion give a longer-term sizes, eat indication ofless preferred foods, households’ or borrow money for food. future The LCI aggregates 10 negative coping strategies affecting the household’s longer term coping capacity, including whether in the past 30 days the household had to borrow money, sell 2. 2. The coping The capacity, Livelihoods by looking Coping at Strategiesthe prevalence (LCS) index of aims 10 to common give a negative longer-term coping strategies indication of affecting households’ the future productive assets, or spend any Livelihoods cases, Strategies Coping savings. In both (LCS) a higher index index indicates higher to a aims riskgive a longer-term of food insecurity (seeindication the Annex for households’ ofmore details). future household’s coping coping capacity, capacity, longer byterm by looking lookingproductive at at the the capacity. prevalence prevalence of The 10 LCS of common 10 common is composed of negative negative questions 4 coping coping strategies strategies indicating affecting affecting stress; the the 3 22 CSI score of 4 or greater indicates moderate to severe food insecurity. See Annex for details. questions household’s household’s indicating longerlonger term crisis, term and 3 questions productive productive capacity. indicating LCS emergency. capacity. The The LCS is composed is composed Allof questions 4of are binary 4 questions questions yes/no indicating indicating questions. stress; stress; 3 3 The stress questions questions questions indicating indicating receive crisis, crisis, and 3 a and weight 2, crisis ofindicating 3 questions questions weight a emergency. indicating of 3 and emergency. emergency All questions All questions a binary are weight are binaryof 4. yes/no The yes/noindex questions. questions. value Theis equal The stress stress to the value questions questions of receive the receive highest-weighted a weight of 2,of a weight crisis indicator 2, crisis a weight to3which a weight of of and theemergency and 3 emergency household responded a a weight of 4.of weight Thepositively. The index 4. index value value is equal is equal a. to theStress tovalue (2): the value Inof of the thethepast 30 days, did highest-weighted highest-weighted the household indicator indicator (i) to which to which sell the consumption the household household goods; responded responded (ii) spend positively. positively. savings; (iii) a. a. borrow StressStress (2): (2): Inmoney; In past the the(iv) past 30sell animals? 30 days,days, did household did the the household (i) consumption sell consumption (i) sell goods;goods; (ii) spend (ii) spend savings; savings; (iii) (iii) b. Crisis borrowborrow (3): money; (iv) past In the money; 30 days, sell animals? (iv)animals? sell did the household (i) reduce spending on health and education; (ii) sell 20 Building Evidence to Enhance the Welfare of Refugees and Host Communities 2. The Livelihoods Coping Strategies (LCS) index aims to give a longer-term indication of households’ future coping capacity, by looking at the prevalence of 10 common negative coping strategies affecting the household’s longer term productive capacity. The LCS is composed of 4 questions indicating stress; 3 questions indicating crisis, and 3 questions indicating emergency. All questions are binary yes/no questions. The stress questions receive a weight of 2, crisis a weight of 3 and emergency a weight of 4. The index value is equal to the value of the highest-weighted indicator to which the household responded positively. a. Stress (2): In the past 30 days, did the household (i) sell consumption goods; (ii) spend savings; (iii) borrow money; (iv) sell animals? b. Crisis (3): In the past 30 days, did the household (i) reduce spending on health and education; (ii) sell productive assets; (iii) take children out of school? c. Emergency (4): In the past 30 days, did the household (i) sell land or a house, (ii) sell the last female animal; (iii) engage in begging? Although urban areas have lower multidimensional percent and 22.2 percent for refugees and hosts, poverty rates, food insecurity is also rampant there. respectively) and are less likely to use harmful coping Indeed, after Turkana, the second highest prevalence of strategies (30.6 percent and 21.9 percent). This is despite food insecurity is found in Nairobi, where 67 percent of more than 7 out of 10 persons in Dadaab being classified refugees and 55.9 percent of hosts have CSI score of 4 as multidimensionally poor. It should be noted that receipt or higher. These two groups also deploy harmful coping of food aid is not sufficient to explain the lower levels of food strategies indicating stress, as measured by the LCS (46.6 insecurity among Dadaab hosts, since as will be shown percent and 54.6 percent, respectively). below, they have relatively low food aid coverage (14.2 percent). There are no statistically significant differences Dadaab residents, on the other hand, have between male- and female-headed households in terms of comparatively low levels of food insecurity (48.1 food insecurity (see Table 15). Figure 18: Food insecurity and negative coping strategies (% of the population), by location/ refugee status a. Coping strategy Index b. Livelihoods Coping Strategy 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Hosts Refugees Hosts Refugees Hosts Refugees Hosts Kalobeyei(r) Hosts Hosts Hosts Hosts Refugees Refugees Refugees Kalobeyei(r) Kakuma(r) Kakuma(r) -20% Turkana Dadaab Nairobi Other Turkana Dadaab Nairobi Other urban urban Stress Emergency 4+: i nsecure 10+: severely insecure Source: Authors’ calculations based on K-LSRH 2023. Notes: r = refugees and h=hosts. Building Evidence to Enhance the Welfare of Refugees and Host Communities 21 © UNHCR/Charity Nzomo 3.3 Education 23 Educational attainment among adult refugees is very low overall, particularly among women, but in camp areas refugees and hosts 66% of female refugees report similarly low levels of education. Across locations, 66 percent (aged 25-64) of female refugees (aged 25-64) report having no schooling compared to report having no schooling 33 percent of male refugees. While hosts overall have higher educational attainment than refugees, disaggregating by location shows that this is driven by urban areas. Educational attainment in camp areas lags far behind urban areas and is similarly poor or worse among hosts as among refugees. In Garissa County, more than 87 percent of hosts 33% of male refugees and 84 percent of refugees have never gone to school and in Turkana, (aged 25-64) hosts have less schooling than the neighboring refugees in Kakuma and report having no Kalobeyei. The low educational attainment in these areas reflects broader schooling spatial disparities within Kenya, rather than being entirely attributable to refugee-related issues (see Box 1). 23 The data referenced in this section is based on self-reported school participation in the survey. 60% 35 41 % of pop 40% 21 66 20 20% 33 14 14 20 0% Male Female Male Female 22 Building Evidence to Enhance the Welfare of Refugees and Host Communities Refugee Host No schooling Madrassa Primary Secondary Vocational/technical Higher Figure 19: Highest education level among 25–64-year-olds by gender Figure 19: Highest education level among 25–64-year-olds by gender Figure 20: Highest education level among 25–64-year-olds by location24 100% 100% 11 7 5 % of population 2 80% 80% 28 3521 10 18 4 % of population 60% 12 29 28 35 60% 2123 41 40% 25 24 27 20 46 66 84 87 44 40%20% 14 33 68 49 14 20 Figure 19: Highest 20% 0% 39 education level among 25–64-year-olds 29 by gender 19 24 16 0%100% Male Female Male Female 3 4 11 % of population Kakuma Kalobeyei 80% Turkana Refugee 18 Dadaab Dadaab Nairobi Host Nairobi Other Other 35 Refugees 60% Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban 35 No schooling 21 Madrassa 41 Primary Refugees Hosts 40% 66 20 No schooling Secondary 20% Primary Madrassa Vocational/technicalSecondary 14 Higher Vocational/technical Higher 33 20 0% 14 Source: Authors’ Male on K-LSRHFemale calculations based 2023. Figure 20: HighestMale education level Female among 25–64-year-olds by location24 Notes: Pre-primary is classified as no schooling. Primary, secondary, vocational, and higher represent adults who have 100% Refugee completed at least one year of schooling Host within that level. Madrassa is a faith-based education that integrates secular 7 religious academic education and Islamic 5 2 education (Mwaura and Marfo 2011). No schooling 10 4 Madrassa Primary 80% 21 28 % of population Secondary Vocational/technical 29 Higher 12 28 Although 60% net Figure primary20:school enrollment Highest 23 education is above level among8025 percent –64- in year -Kakuma/Kalobeyei olds by location24 and urban areas, very few children of primary age in Dadaab and among Turkana hosts are in school. 25 25 2424 Garissa and Turkana 40% Figure 20: Highest education level84 among 25–64-year-olds 87 27 by location 46 coun?es have consistently had low enrollment rates compared to the na?onal average (see 44 Box 1). In 68 100% urban areas, Kenyans 49and refugees have7similarly5high enrollment rates, likely due to the government’s 20% 39 4 2 integra?on 80% urban refugees of28 21 10 the na?onal into 29 26 Addi?onally, educa?on system. 19 primary enrollment 24 16 in % of population 0% 12 3 28 4 Kakuma and Kalobeyei is comparable to urban areas which is reflec?ve of large investments in camp 29 60% 23 schools by 25 Kalobeyei Kakuma UNHCR Turkana organiza?ons. and development Dadaab Dadaab Nairobi As of March 2020,Nairobi there wereOther Otherschools in 27 primary Refugees Refugees Hosts Refugees 84 87Hosts 27 Refugees 46 Hosts 24 Urban 44 Urban Kakuma and Kalobeyei (Hure 68 40% and Taylor 2023). Although these camp schools follow the na?onal curriculum and receive some 49 financial support from the Government of Kenya through RefugeestheHosts Kenya Primary 20% 39 Educa?on Equity in Learning Program (PEELP), they 29 are financed 19 16 and managed outsideHigher 24 of the state system No schooling 0% Madrassa Primary Secondary 3 Vocational/technical 4 Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other 24 KCHS at the Data fromRefugees national Refugees level shows Source: Hosts Authors’ that half Refugees Hosts calculations of the Refugees based population 2023.25-64 aged on K-LSRHHosts has completed Urban Urban at most primary schooling. Refugees Hosts Notes: Pre-primary is classified as no schooling. Primary, secondary, vocational, and higher represent adults who have 25 completedEnrollment at least rates one are year ofself-reported. schooling Self-reported attendance is rates are also provided in theintegrates are largely annex andsecular No schooling Madrassa within that level. Primary Madrassa Secondary a faith-based education Vocational/technical that Higher similar to enrollment rates. academic education and Islamic religious education (Mwaura and Marfo 2011). 26 World Bank program appraisal document for a Kenya secondary education equity and quality improvement program (2024). Source: Authors’ calculations based on K-LSRH 2023. Although net Notes primaryis : Pre-primary school classified enrollment as no schooling. isPrimary, 80 percent abovesecondary, in Kakuma/Kalobeyei vocational, Source: Authors’ calculations based on K-LSRH 2023. and higher represent adults andwhourbanhave areas, very completed at least one year of schooling within that level. Madrassa is a faith-based education that integrates secular few children of primary age education academic is Notes: Pre-primary in Dadaab classified as and and among schooling. no Islamic religious Turkana secondary, Primary, education hosts 28 vocational, (Mwauraand and are higherin represent Marfo school. 25 Garissa and Turkana 2011). adults who have coun?es have consistently had low enrollment rates compared to the na?onal average completed at least one year of schooling within that level. Madrassa is a faith-based education that integrates secular (see Box 1). In academic education and Islamic religious education (Mwaura and Marfo 2011). urban areas,net Although Kenyans primary andschool refugees enrollment have above 80 high issimilarly percent enrollment rates, likely in Kakuma/Kalobeyei anddue to the urban government’s areas, very integra?on few children of urban of primary refugees age in into Dadaab the na?onal and among educa?on Turkana hosts system. 26 are in Addi?onally, school. Garissa 25 primary enrollment in and Turkana Kakuma coun?esand Kalobeyei have consistently is comparable had low enrollment to urban areas rates which to compared is reflec?ve the na?onal ofaverage large investments (see Box 1). In in camp urban areas, Kenyans and refugees have similarly high schools by UNHCR and development organiza?ons. As of March 2020, there were 27 primary schools in enrollment rates, likely due to the government’s and of integra?on Kakuma Data from KCHS 24 urban Kalobeyei at the national (Hure into refugees level shows that half of the and na?onal Taylor the population 2023).educa?on aged 25-64 Although system. has completed at most primary26 these Addi?onally, camp schools schooling. primaryfollowenrollmentthe inna?onal Kakuma and Kalobeyei is comparable to urban areas which is reflec?ve of large investments in camp curriculum and receive some financial support from the Government of Kenya through the Kenya Primary schools by UNHCR and development organiza?ons. As of March 2020, there were 27 primary schools in Educa?on Equity in Learning Program (PEELP), they are financed and managed outside of the state system Kakuma and Kalobeyei (Hure and Taylor 2023). Although these camp schools follow the na?onal Building Evidence to Enhance the Welfare of Refugees and Host Communities 23 Although net primary school enrollment is above 80 of Kenya through the Kenya Primary Education Equity percent in Kakuma/Kalobeyei and urban areas, very few in Learning Program (PEELP), they are financed and children of primary age in Dadaab and among Turkana managed outside of the state system by non-governmental hosts are in school.25 Garissa and Turkana counties actors (Hure and Taylor 2023). In contrast to Kakuma have consistently had low enrollment rates compared to and Kalobeyei refugees, only half of primary-school-age the national average (see Box 1). In urban areas, Kenyans children in the Turkana host community are enrolled in and refugees have similarly high enrollment rates, likely school. This gap between Kakuma/Kalobeyei refugees and due to the government’s integration of urban refugees Turkana hosts is consistent with previous reports, despite into the national education system.26 Additionally, primary policies that permit hosts’ access to schools in the camps.27 enrollment in Kakuma and Kalobeyei is comparable to Most of the children that are not enrolled in school have urban areas which is reflective of large investments in never attended school suggesting the need for tailored camp schools by UNHCR and development organizations. interventions, particularly in camp areas for both hosts and As of March 2020, there were 27 primary schools in refugee children who have never attended school before. Kakuma and Kalobeyei (Hure and Taylor 2023). Although In Garissa County, where enrollment rates are the lowest, these camp schools follow the national curriculum and nearly all children who are not enrolled have never been in receive some financial support from the Government school before. Figure 21: Primary school enrollment 28 200 174 174 180 % of primary school aged children 160 140 129 114 111 111 120 100 89 88 86 83 84 81 82 80 62 53 51 60 40 27 17 20 0 Kakuma Kalobey ei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban Refugees Hosts Net primary enrollment Gross primary enrollment Source: Authors’ calculations based on K-LSRH 2023. Notes: Under the new Competency Based Curriculum (CBC) phased in 2017, primary school age children are between 6 and 11 years old. However, at the time of survey, some older students (ages 12-13) were still enrolled in primary school under the prior 8-4-4 curriculum. Primary school age children are thus defined in this calculation and in all other calculations pertaining to primary school, as 6-11 years old (for most children who are enrolled in the CBC curriculum or not enrolled at all), and 6-13 years old (for the few children who are part of the last cohorts enrolled in the 8-4-4 curriculum). 25 Enrollment rates are self-reported. Self-reported attendance rates are also provided in the annex and are largely similar to enrollment rates. 26 World Bank program appraisal document for a Kenya secondary education equity and quality improvement program (2024). 27 For example, at the end of 2019, only 499 host children were enrolled in any school in Kakuma (primary or secondary), compared to 11,805 refugees enrolled in secondary school alone SES Kakuma (2019). 28 Data from the KCHS estimates national primary school net enrolment at 78.1% See Box 1. 24 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 22: Share of primary-age children who have never attended school among those not enrolled 100 97 95 79 81 89 % of not enrolled primary-age 80 57 60 32 14 children 40 13 20 0 Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban Refugees Hosts Source: Authors’ calculations based on K-LSRH 2023. While refugee boys are more likely to be enrolled in primary enrollment among refugees is higher in urban primary school than refugee girls, enrollment among areas where only 75 percent of refugee girls are enrolled, host children is similar across gender. In the camp compared to 94 percent of boys. Among hosts in the areas, the net primary school enrollment rate among camps and urban areas, net primary enrollment is similar refugee boys is 76 percent while refugee girls lag behind across gender, but there is higher over-age enrollment by about 6 percentage points.29 The gender gap in net among boys. Figure 23: Primary enrollment by gender 157 160 137 % of primary-school-age children 140 120 113 114 108 120 94 100 86 84 76 77 72 75 80 70 60 39 39 40 20 0 Male Female Male Female Male Female Male Female Camp refugees Camp hosts Urban refugees Urban hosts Net primary enrollment Gross primary enrollment Source: Authors’ calculations based on K-LSRH 2023. 29 This gender gap in refugee enrollment is consistent with Hure and Taylor (2023) who note that in Kakuma/Kalobeyei “girls continue to be disproportionately disadvantaged, with fewer girls than boys attending school at the upper primary and secondary levels.” Building Evidence to Enhance the Welfare of Refugees and Host Communities 25 The main barrier to primary school enrollment reported cost with 66 percent reporting school fees and 7 percent by hosts and urban refugees is school fees, while reporting other costs associated with school. Barriers to reasons among camp refugees are more varied. In enrollment among camp refugees are more diverse. In urban areas among primary-school-age children who Dadaab, where enrollment is lowest, lack of interest (17 are no longer enrolled or attending school, 87 percent percent) and preference for faith-based education such as of refugees and 73 percent of hosts cite school fees or Madrassa (17 percent) are cited as the main reasons for other costs as the main reason for not attending school. no longer attending. However, it is important to note that Although primary school is free in Kenya, the high number this data comes from a small non-representative sample of children reporting cost as a barrier suggests that the of out-of-school children and does not provide insight into implementation of it has not been fully realized.30 Low low enrollment among children who have never attended enrollment among hosts in camp areas is also due to school.31 Figure 24: Top 5 reported reasons for children not attending school a. Refugees b. Host communities Camp refugees Urban refugees Camp hosts Urban hosts % of population % of population 0 20 40 60 80 0 10 20 30 40 50 60 70 7 School f ees 39 School f ees 62 72 Not interested/did not like Not interested/did not like 16 13 school 2 school 7 Other costs associated with Other costs associated with 5 school (uniforms, textbooks, 7 school (uniforms,textbooks, 9 11 etc.) etc.) Own illness/disability 5 Own illness/disability 11 3 4 School closed due to Parents told me to stop 5 COVID/unaware schools 7 4 0 reopening/camp closure Source: Authors’ calculations based on K-LSRH 2023. Notes: This question is asked about primary school-age children who have attended school in the past but are no longer enrolled or attending. The graph shows both primary and secondary age children due to sample size limitations. Nevertheless, patterns are consistent even when examined separately. 31 Respondents were given distinct options of school fees and other costs associated with schools. Still, it is possible that some respondents may have cited school fees while referring to other costs. 32 Survey data on reasons for not attending is limited to children who have attended school previously and are no longer attending or enrolled. Children who have never had contact with the formal schooling system (who make up most out-of-school children) were not asked this question. 26 Building Evidence to Enhance the Welfare of Refugees and Host Communities Elsewhere in Kakuma/Kalobeyei and in urban areas, children. In the early grade mathematics assessment where enrollment is high, the challenge is overage (EGMA), refugees outperform hosts with 47 percent scoring learners. There is substantial over-age enrollment, full marks compared to 23 percent of hosts. However, these particularly in Kakuma and Kalobeyei where the gross assessments are designed for early grade students (grades primary enrollment rate is 174 percent. High over-age 1-3) and competency in subject matter is defined as scoring enrollment and low primary graduation rates also suggest at least 50 percent.33 With this in mind, Figure 25c suggests that there may be insufficient learning. To examine this, that 31 percent of Turkana host children and 19 percent of early grade reading and mathematics assessments were Dadaab host children in grade 6 are not competent in grade administered to camp refugees and hosts enrolled in their 1-3 level mathematics. Although refugees score higher, last year of primary school (grade 6).32 some still lag behind with 12 percent of Kakuma/Kalobeyei Test scores suggest that some learners have a poor refugees and 5 percent of Dadaab refugees failing to pass grasp of the material, particularly Turkana host a grade 1-3 level test. Figure 25: Early Grade Math Assessment a. Early Grade Maths Assessment Score b. Distribution of EGMA Scores 5 .0 50 4.5 4.02 45 4.05 3.91 3.67 4.0 40 % of child respondents 3.13 35 3.5 30 3.0 mean score 25 2.5 20 2.0 15 1.5 10 1.0 5 0.5 0 0 1 2 3 4 5 0.0 Kakuma Kalobey ei Turkana Dadaab Dadaab Camp refugees Camp hosts Refugees Refugees Hosts Refugees Hosts Source: Authors’ calculations based on K-LSRH 2023. 32 The Early Grade Mathematics Assessment (EGMA) and Early Grade Reading Assessments (EGRA) for English and Kiswahili were developed by USAID and RTI and have been validated in several developing contexts, including Kenya. Although these tests are designed for children in grade 1, 2, or 3 (early grades), experts recommended that these assessments would still be appropriate in this study given the low learning in the camps. Due to time constraints, only selected sections are administered from EGMA and EGRA. Mathematic competency is measured using Task 6 of EGMA which consists of word problems. From the English EGRA, sections 5a and b are selected (an oral reading passage and a set of comprehension questions); and for the Kiswahili EGRA, parts 6a and 6b (a listening comprehension) are administered. 33 See for reference Kenya National Examinations Council (2020). Building Evidence to Enhance the Welfare of Refugees and Host Communities 27 c. Proportion of students who pass the 50% competency benchmark 100 95 88 87 90 82 80 69 70 % of students 60 50 40 30 20 10 0 Kakuma Refug ees Kalobey ei Refug ees Turkana Hosts Dadaab Refug ees Dadaab Hosts Source: Authors’ calculations based on K-LSRH 2023. Notes: Early grade maths assessment scores are based on a series of word problems. Each correct answer is equal to one point, with scores ranging from 0-5. Error bars in Figure 25a represent the 95% confidence interval. Refugees in Kakuma and Kalobeyei also have the Somali, the shared language may mean less need to learn highest scores in the English EGRA, while refugees English or Swahili to get by, which is not the case in Turkana and hosts in Dadaab perform the worst. However, due County where refugees are more heterogeneous and hosts to small sample sizes, this analysis is underpowered to and refugees are less likely to speak similar languages. detect statistically significant differences.34 Using the same Additionally, the current education policy stipulates that benchmarks defined by the US Agency for International children are taught in the language of the catchment area up Development (USAID), only 28 percent and 23 percent of until Grade 4. Children in Dadaab are therefore first taught Dadaab refugees and hosts are fluent readers in English, in Somali whereas in Nairobi, Kakuma and Kalobeyei they compared to 41 percent and 54 percent of Kakuma are taught in Swahili. Children of Turkana hosts are first and Kalobeyei refugees.35 Since Dadaab refugees are taught in Turkana, and this may explain their lower English predominantly Somali and the host population is also language scores. Figure 26: English assessment scores a. English Oral Reading Score b. English Comprehension Score 70 5 59 55 50 3.94 mean number of words read correctly 60 3.17 47 49 4 3.44 3.52 3.21 50 3 mean score 40 30 2 20 1 10 0 0 Kakuma Kalobey ei Turkana Dadaab Dadaab Kakuma Kalobey ei Turkana Dadaab Dadaab Refugees Refugees Hosts Refugees Hosts Refugees Refugees Hosts Refugees Hosts 34 Compared to Kakuma/Kalobeyei, only a small number of households in Turkana and Dadaab had a child enrolled in grade 6 at the time of survey. Thus, early grade assessment data for these locations is limited by sample sizes: only 34 Turkana hosts, 41 Dadaab refugees, and 42 Dadaab hosts completed the assessments. 35 The Class 2 benchmarks for reading performance in English are 30-64 correct words per minutes (CWPM) for emergent readers and 65 or more CWPM for fluent readers. At the midline study of class 1 pupils in Kenya, 30% were emergent readers and 18% fluent; while 29% of class 2 pupils were emergent readers and 47% were fluent readers in English. USAID (2017). 28 Building Evidence to Enhance the Welfare of Refugees and Host Communities c. English Reading Fluency 100 90 80 41 70 23 54 28 60 % of students 37 50 40 30 58 55 51 20 39 36 10 0 Kakuma Refugees Kalobeyei Refugees Turkana Hosts Dadaab Refugees Dadaab Hosts Emergent Readers (CWPM=30-64) Fluent Readers (CWPM=65+) Notes: Two sections are administered from the Early Grade Reading Assessment in English to those who say they can have a conversation in English (see Annex: A timed oral passage reading (scored from 0-66 with 1 point awarded for each word read correctly) and comprehension questions based on the passage (scored from 0-5 with 1 point awarded for each correct answer). Error bars indicate the 95% confidence interval. In the Kiswahili assessment, hosts seem to score slightly higher than refugees. However, sample size limitations again prevent the analysis from rejecting that scores are the same across locations. Overall, the EGMA and EGRA scores suggest limited grasp of educational material which merits attention given that gaps in foundational learning in primary school also adversely affect later learning. 36 Figure 27: Swahili assessment scores a. EGRA Swahili Comprehension Score b. Distribution of EGRA Kiswahili Scores 5 35 30 4 3.59 3.59 3.68 3.42 3.31 25 3 mean score 20 % of CRs 2 15 10 1 5 0 0 Kakuma Kalobeyei Turkana Dadaab Dadaab 0 1 2 3 4 5 Refugees Refugees Hosts Refugees Hosts Camp refugees Camp hosts Source: Authors’ calculations based on K-LSRH 2023. Notes: The section administered from the Early Grade Reading Assessment in Kiswahili is a set of 5 comprehension questions asked to those who say they can have a conversation in Kiswahili (see annex): Scores range from 0-5 with 1 point awarded for each question answered correctly. Some existing evidence suggests that learning outcomes for refugee learners in secondary school are very low, with only 3 percent of KCSE candidates achieving a mean grade of 36 C+ or higher and this performance is associated with low foundation learning outcomes in primary school. Building Evidence to Enhance the Welfare of Refugees and Host Communities 29 A child’s self-reported self-efficacy is a significant Compared to primary enrollment, secondary school predictor of academic achievement. Regression analysis enrollment is markedly lower, particularly in camp of factors that may be associated with performance on areas. Enrollment rates in urban areas are higher, but test scores reveals self-efficacy as a significant correlate, refugees lag behind hosts by about 18 percentage points. across all assessments.37 The self-efficacy index used is a In the camps, very few children appear to be transitioning measure of the child’s ability to solve problems, accomplish from primary to secondary school. In Kakuma and their goals, and self-regulate.38 This result is consistent Kalobeyei, net secondary enrollment rates are only 23 with a large evidence base that connects socioemotional percent and 18 percent, respectively, compared to primary skills to foundational learning acquisition. Evidence reviews enrollment rates of 81 percent and 89 percent. Children of over 700 studies show that, on average, socioemotional in Dadaab are even less likely to be in secondary school, skills have a positive impact on academic attainment, with only 9 percent of refugees and 11 percent of hosts equivalent to 4 additional months’ progress.39 Although enrolled. Information provided by UNHCR operations higher self-efficacy could lead to improved test scores, highlights that these low transition rates are due to failure higher academic achievement could also increase a child’s to pass national exams and the lack of capacity in camp confidence and self-efficacy, making it difficult to identify the secondary schools which significantly restricts the number direction of the relationship. Factors negatively correlated of students who can transition to secondary school. with a child’s Kiswahili score include the number of days Across gender, secondary net enrollment rates are largely the child was absent from school in the past week and the similar in the camp areas. However, male refugees and number of students in the child’s class, but these are not hosts have higher over-aged enrollment, especially in the statistically significant predictors of scores on the English camp areas. and mathematics assessments. Figure 28: Secondary school enrollment rates a. Secondary school enrollment rates 100 94 84 82 % of secondary school aged children 76 80 64 65 61 59 60 48 44 36 40 30 28 23 18 18 20 9 11 0 Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban Refugees Hosts Net secondary enrollment Gross secondary enrollment 37 Regression results are reported in the Annex. 39 The index is constructed as the normalized sum of ten questions in the self-efficacy scale (For details on the development of an efficacy scale see Schwarzer and Jerusalem 1995). Full scale in the Annex. 100 40 83 See for reference Education Endowment Foundation (2019). 80 81 l aged children 80 68 65 68 64 57 60 52 43 40 32 20 9 11 % of second 0 Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban 30 Refugees Hosts Building Evidence to Enhance the Welfare of Refugees and Host Communities Net secondary enrollment Gross secondary enrollment b. Secondary school enrollment rates by gender 100 83 80 81 % of secondary-school aged children 80 68 65 68 64 57 60 52 43 40 32 25 20 19 14 14 20 0 Male Female Male Female Male Female Male Female Camp refugees Camp hosts Urban refugees Urban hosts Net secondary enrollment Gross secondary enrollment Source: Authors’ calculations based on K-LSRH 2023. Notes: The section administered from the Early Grade Reading Assessment in Kiswahili is a set of 5 comprehension questions asked to those who say they can have a conversation in Kiswahili (see annex): Scores range from 0-5 with 1 point awarded for each question answered correctly. Low (net) secondary school enrollment is largely due Kalobeyei and urban areas. In these locations, accounting to overage learners in primary school. Across locations, for secondary-school-aged children in primary school many secondary-school-age children not enrolled in shows that almost all children in this age group are in secondary school are still in primary school. This highlights school. However, in the Turkana host community and that insufficient learning in primary schools is a significant Dadaab, only about half of secondary-school-age children barrier to secondary enrollment and that children may be are in school (secondary or primary). starting schooling at older ages, particularly in Kakuma/ Figure 29: School enrollment among secondary-school-aged children 100 % of secondary-school-aged children 23 18 80 44 48 65 61 60 18 9 40 11 74 77 53 49 20 38 40 36 32 35 0 Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban Refugees Hosts Enrolled in primary Enrolled in secondary Source: Authors’ calculations based on K-LSRH 2023. Building Evidence to Enhance the Welfare of Refugees and Host Communities 31 Low enrollment rates are not consistent with except in Dadaab where some child respondents would like educational aspirations, which are exceptionally high. to complete a lower level of education than they realistically Most children report that the highest level of education expect to achieve. The discrepancy between the caregiver’s they would like to complete is university. Caregivers have aspiration and expectation for the child respondent follows similarly high aspirations, with most wanting their child a similar pattern, where aspirations on average are slightly to complete university. Moreover, almost all children and higher except in Dadaab. Although children and caregivers’ their caregivers report that they realistically expect to aspirations are mostly aligned, there are a few children achieve these high aspirations, which sharply contrasts the in Kalobeyei and Dadaab who aspire to complete more low educational attainment in the camps.40 On average, education than their caregivers would like for them to children report slightly higher aspirations than expectations, achieve. Figure 30: Child respondent’s educational aspirations and expectations a. Child's aspirational highest level 100 1 2 9 21 80 Upper Primary % of children 60 Juniour Secondary 89 50 90 96 Senior Secondary 100 100 1100 100 1 211002 2 89 1 9 21 9 92 9 9 40 21 21 21 21 21 University 80 80 80 80 80 TVET % of children % of children % of children % of children % of children 60 60 60 60 60 20 96 96 89 96 89 89 96 89 89 5089 50 90 50 90 96 89 17 90 50 50 90 90 89 89 89 40 40 40 40 40 9 1 20 20 1 20 20 7 20 17 17 17 17 17 6 3 2 3 2 0 2 9 9 619 1 9 19 6 1 1 7 07 7 7 07 6 6 6 0 00 1 2 0 02 1 0 03 021 0 0301 0 2 2 0301 2 0 32 3 0 2 0 03 0 32 0 2 32 0 3 2 23 2 2 Kalobeyei Refugees Refugees Refugees Refugees Dadaab Refugees Refugees Refugees Dadaab Hosts Refugees Refugees Hosts Refugees Hosts Dadaab Hosts Dadaab Hosts Turkana Hosts Hosts Hosts Hosts Hosts Kakuma Refugees Kakuma Refugees Kakuma Refugees Kakuma Refugees Kakuma Refugees Dadaab Dadaab Turkana Turkana Turkana Turkana Upper Primary Junior Secondary Senior Secondary University TVET Kalobeyei Kalobeyei Kalobeyei Kalobeyei Dadaab Dadaab Dadaab Dadaab Upper Upper Upper Primary Primary Primary UpperJunior Upper Primary Junior Secondary Primary Secondary Junior Junior Junior Secondary Secondary Secondary Source: Authors’ calculations based on K-LSRH 2023. Notes: TVET comprises Senior Technical Senior and Vocational Secondary Secondary Senior Secondary Senior University Senior Education University Secondary Secondary and Training; University University Aspirations are measured by asking children for University the highest level of education they wish/would like to complete if they had no constraints and could study for as long as they liked. Caregivers are similarly asked about their aspirations for the child respondent. TVET TVET TVET TVET TVET 40 Expectations could be skewed higher for a few reasons. Firstly, among refugees, education can lead to significant opportunities through scholarship programs that allow children to leave the camps, and in some cases, even study in the US or Canada. Although last year (2023) saw the highest number of applicants and scholarships awarded, these opportunities are still limited to a very select few. Despite the low odds, caregivers may choose to hold on to the belief that their child could achieve high levels of education, given the life-altering advantages this could provide. Secondly, responses may be biased if households believe there is an incentive in reporting to UNHCR that their child is academically inclined and expected to go on to university. Finally, the wording of the questions may also have been ambiguous after translation and the differences between aspirations and expectations may have been unclear. It should also be noted that there is a recently established university in Kakuma, making tertiary education more accessible in these locations. 32 Building Evidence to Enhance the Welfare of Refugees and Host Communities b. Child's expected highest level 100 1 3 0 9 21 80 % of children Upper Primary 60 81 87 77 91 Juniour Secondary 100 100 100 100 11 1 1 22 2 2 99 9 9 62 Senior Secondary 40 212121 21 University 8080 80 80 children of children % of children % of children TVET 6060 60 60 20 96 969696 89 898989 909090 505050 90 50 898989 89 4040 40 40 10 % of 16 10 15 9 % 20 202020 2 17 171717 5 0 0 0 0 0 77 7 7 99 9 9 2 61 16161 6 01 00 0 0 1 201 2 01 2 0 3 2 03 03003 202 00220 33 3 3 22 2 2 Refugees Kalobeyei Refugees Kalobeyei Refugees Kalobeyei Refugees Refugees Dadaab Refugees Dadaab Refugees Dadaab Refugees Hosts Dadaab Hosts Dadaab Hosts Dadaab Hosts Hosts Turkana Hosts Turkana Hosts Turkana Hosts Refugees Kakuma Refugees Kakuma Refugees Kakuma Refugees Upper Primary Junior Secondary Senior Secondary University Dadaab Turkana Kalobeyei Kakuma Dadaab TVET Upper Upper UpperUpper Primary Primary Primary Primary Junior Junior Junior Junior Secondary Secondary Secondary Secondary Figure 31: Caregiver’s educational aspiration and expectation for the child respondent Senior Senior Senior Senior Secondary Secondary Secondary SecondaryUniversity University University University TVET a.TVET TVET TVET 's aspiration for the child Caregiver 120 100 1 1 9 20 0 80 % of Caregivers Upper Primary 60 Juniour Secondary 96 Senior Secondary 87 89 46 81 40 University 16 TVET 20 10 2 18 2 2 6 1 1 2 6 5 0 Kalobeyei Refugees Kakuma Refugees Dadaab Refugees Dadaab Hosts Turkana Hosts Building Evidence to Enhance the Welfare of Refugees and Host Communities 33 b. Caregiver's Expectation for the child 1 16 77 5 Kakuma Refugees Upper Primary 1 8 86 3 Kalobeyei Refugees Juniour Secondary Senior Secondary 13 79 9 Turkana Hosts University TVET Dadaab Refugees 2 9 12 55 21 1 Dadaab Hosts 2 10 80 0 20 40 60 80 100 % of Caregivers Source: Authors’ calculations based on K-LSRH 2023. Notes: Aspirations are measured by asking caregivers for the highest level of education they wish/would like the child respondent to complete if they had no constraints and could study for as long as they liked. Expectations are measured by asking caregivers for the highest level of education they realistically think the child respondent will complete. Missing observations among Dadaab host caregivers reflect “Don’t Know” responses. Figure 32: Discrepancy between child respondent’s (CR) educational aspiration and expectation CR Aspiration CR Expectation Kakuma Refugees Kalobeyei Refugees Turkana Hosts Dadaab Refugees Dadaab Hosts Upper Primary Junior Secondary Senior Secondary University 34 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 33: Discrepancy between caregiver’s (CG) educational aspiration and expectation for the child respondent CR Aspiration CR Expectation Kakuma Refugees Kalobeyei Refugees Turkana Hosts Dadaab Refugees Dadaab Hosts Upper Primary Junior Secondary Senior Secondary University Figure 34: Discrepancy between child respondent (CR) and caregiver (CG) educational aspirations CR Aspiration CR Expectation Kakuma Refugees Kalobeyei Refugees Turkana Hosts Dadaab Refugees Dadaab Hosts Upper Primary Junior Secondary Senior Secondary University Source: Authors’ calculations based on K-LSRH 2023. Notes: Aspirations are measured by asking children for the highest level of education they wish/would like to complete if they had no constraints and could study for as long as they liked. Caregiver aspirations are similarly measured by asking the caregiver for the highest level of education they wish/would like the child respondent to complete if they had no constraints and could study for as long as they liked. Expectations are measured by asking children for the highest level of education they realistically think they will complete. In this figure, TVET (technical and vocational educational training) is excluded as the survey response does not distinguish TVET colleges from vocational training that could be completed in lieu of junior or senior secondary, for example. Low enrollment also does not seem to be driven by level they hope to achieve. Moreover, when examining preferences for working. If their child was offered a job employment among children between 5 and 14 years old, this before completing secondary school, most caregivers report does seem slightly higher among Dadaab households, but that they would not want them to take it. An estimated 78 is still low relative to the proportion of children out of school. percent and 57 percent of refugee caregivers in Kakuma About 4.6 percent of refugee children in Dadaab worked and Kalobeyei and 61 percent of Turkana host caregivers outside the household in the previous week, compared to completely disagree with allowing their child to drop out of less than 1 percent of children in Kakuma and 1.3 percent of school to take a job. Slightly more refugees and hosts in children in Kalobeyei. Working appears to be more of a factor Dadaab would want their child to take the job, but still more for hosts – about 5 percent of Turkana host children and 14 than 60 percent disagree (completely or somewhat). This percent of host children in Dadaab worked the previous week. suggests that refugees and hosts in Dadaab may prioritize However, the proportion of children and caregivers in Dadaab children working slightly more than in other locations. This is who express a preference for working is too small to explain also indicated by Figure 31a, which shows that 21 percent of why half of the secondary-school-aged children in Dadaab are refugees in Dadaab report vocational education as the highest not in school. Building Evidence to Enhance the Welfare of Refugees and Host Communities 35 ©UNHCR /Charity Nzomo 36 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 35: Proportion of caregivers who would want their child to take a job before completing secondary school 100 80 38 39 57 61 % of caregivers 60 78 24 23 40 15 14 16 19 20 15 10 15 2 21 18 10 13 9 0 Kakuma Refug ees Kalobey ei Refug ees Turkana Hosts Dadaab Refug ees Dadaab Hosts Completely ag ree Agree somewhat Disagree somewhat Disagree completely Source: Authors’ calculations based on K-LSRH 2023. Notes: Caregivers are read the statement (“If [child] were offered a good job before completing secondary school, you would let [child] take the job?”) and asked whether they completely agree, somewhat agree, somewhat disagree, or completely disagree. Traditional gender beliefs also do not seem to be a also disagree with keeping daughters out of school to help driving factor in low school enrollment.41 Reported with chores and care at home. While some respondents beliefs in gender-equal education are high across locations, in Dadaab report slightly less gender-progressive beliefs, particularly in urban areas. Almost all respondents in urban the majority still support girls’ education with 73 percent of areas and Kakuma/Kalobeyei disagree that that boys’ refugees and 57 percent of hosts disagreeing that sons’ education is more important than girls’. Most respondents education should be prioritized over girls’. Figure 36: Important for sons to have more education than daughters 100 80 57 60 75 73 82 % of RRs 88 95 89 89 96 40 11 20 9 8 8 31 4 17 19 5 5 8 11 1 1 6 6 0 3 3 Kakuma Kalobey ei Turkana Dadaab Dadaab Nairobi Nairobi Other Urban Other Urban Refugees Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Agree Somewhat ag ree Disagree 41 Since responses are self-reported, respondents may be giving responses to what is expected rather than what they believe or what is done. Other studies suggest gender norms play a significant role in enrollment. Giacomo et al. (2024). Building Evidence to Enhance the Welfare of Refugees and Host Communities 37 Figure 37: Daughters should be sent to school only if they are not needed to help at home 100 80 62 56 60 72 % of RRs 93 86 92 90 98 97 40 8 22 20 11 30 22 7 17 2 0 3 3 7 1 1 2 5 3 8 Kakuma Kalobey ei Turkana Dadaab Dadaab Nairobi Nairobi Other Urban Other Urban Refugees Refugees Hosts Refugees Hosts Refugees Hosts Refugees Hosts Agree Somewhat ag ree Disagree Source: Authors’ calculations based on K-LSRH 2023. Although marriage may cause girls to drop out of slightly higher prevalence of traditional gender beliefs in secondary school, data collected from caregivers Dadaab could therefore be a factor in low enrollment in does not suggest that this is an important factor this location. This seems to be reflected in the gender across locations. When asked if they would want their gaps in secondary enrollment: 6 percent of refugee girls in child to marry before completing secondary school, most Dadaab are enrolled compared to 11 percent of boys, and caregivers completely disagree. However, fewer caregivers 10 percent of host girls in Dadaab are enrolled compared in Dadaab report being opposed to this. An estimated 45 to 12 percent of boys. However, the very low rates of percent of hosts and 17 percent of refugees in Dadaab boys enrolled also suggest that there may be other more caring for girls agree (completely or somewhat) that they significant factors driving low enrollment. would want them to marry before completing school. The Figure 38: Proportion of caregivers who would want child to marry before completing secondary school 100 80 42 48 47 62 68 64 60 % of CGs 82 89 91 100 20 8 40 4 14 34 21 11 12 20 10 41 7 12 3 18 1 0 17 6 9 14 9 11 14 7 0 2 2 0 Male Female Male Female Male Female Male Female Male Female Kakuma Refug ees Kalobey ei Refug ees Turkana Hosts Dadaab Refug ees Dadaab Hosts Completely ag ree Agree somewhat Disagree somewhat Disagree completely Source: Authors’ calculations based on K-LSRH 2023. Note: This figure is disaggregated by the gender of the child respondent 38 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 39: Net Secondary School Enrollment by Gender 80 % of secondary-school-aged children 68 64 61 59 60 56 46 41 39 40 22 25 18 17 17 18 20 12 11 10 6 0 Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Male Female Kakuma Kalobey ei Turkana Hosts Dadaab Dadaab Hosts Nairobi Nairobi Hosts Other Urban Other Urban Refugees Refugees Refugees Refugees Refugees Hosts Source: Authors’ calculations based on K-LSRH 2023. © UNHCR/Charity Nzomo Building Evidence to Enhance the Welfare of Refugees and Host Communities 39 Cost is the main obstacle to school enrollment for hosts children can acquire foundational skills and pass national and urban refugees, while camp refugees struggle exams at the end of Primary school, investing more into to transition to secondary school due to insufficient accelerated learning programs could improve graduation learning and lack of capacity among other reasons. In rates and alleviate the strain of over-age enrollment on urban areas among children who are no longer enrolled or primary school resources. However, this must be met with attending school, 87 percent of refugees and 73 percent increased capacity in secondary schools. In Kakuma and of hosts cite school fees or other costs as the main Kalobeyei, there are 27 primary schools but only nine reason for not attending school. Camp hosts face similar secondary schools, severely constraining the number of cost barriers with 39 percent reporting school fees and 7 students who can transition to secondary (UNHCR n.d.a). percent reporting other costs as the reason for no longer Similarly, in Dadaab, there are 22 primary schools in the attending. School fees are also a reported barrier for camp camp, but only six secondary schools (UNHCR n.d.b). refugees but to a smaller extent. While primary schools While the integration of camps as part of the Shirika plan in Kakuma, Kalobeyei, and Dadaab are free to attend, should help refugees access a larger pool of secondary secondary schools charge a fee of 3,000 Kenyan shillings schools, keeping costs low will be important in ensuring or about US$24 per year (UNHCR and World Bank (2019). higher enrollment among refugees and hosts. As most However, this fee can be waived upon application to the out-of-school children have never attended school before, School Board of Management. Among camp refugees, it is also important to create tailored interventions that the more significant obstacles to secondary enrollment can address cost, distance, and any socioeconomic and include low learning, as evidenced by poor test scores and cultural reasons preventing these children from enrolling large over-age enrollment in primary school. To ensure in school. Box 3: Administrative data in action: education in refugee camps In collaboration with government bodies and implementing partners on the ground, an Education Management Information System (EMIS) has been established within the refugee camps of Turkana and Garissa counties. This system facilitates the real-time monitoring of administrative data on enrolment, facilities, and educational performance within the camps. The following section presents key metrics derived from the EMIS data for Term 1 of 2024. The gross and net enrolment rates across the Dadaab, Kakuma, and Kalobeyei camps highlight differing levels of success in educational access across the camps, with Kalobeyei showing the strongest performance, followed by Kakuma, and then Dadaab. Kalobeyei leads with the highest overall gross enrolment rates, with females at 93% and males at 105%, driven by exceptionally high rates in primary and pre-primary levels, where both genders exceed 100%. Net enrolment rates in Kalobeyei are also robust, with both genders achieving 47%. In Kakuma, gross enrolment rates are strong, particularly in primary education, where males reach 110% and females 88%. Net enrolment rates in Kakuma are also significant, with an overall rate of 41% for males and 39% for females, reflecting relatively good access to age-appropriate education. Dadaab, however, shows lower overall gross enrolment rates, with 34% for females and 44% for males, and net enrolment rates of 10% and 13% respectively, indicating greater challenges in enrolling children, especially in age-appropriate grades. These variations highlight differing levels of success in educational access across the camps, with Kalobeyei showing the strongest performance, followed by Kakuma, and then Dadaab. 40 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 40: Gross enrolment and net enrolment rates by gender and level of education 160% 140% 120% 100% 80% 60% 40% 20% 0% Pre-primary Pre-primary Pre-primary Primary Primary Primary Junior School Senior School Total Junior School Senior School Total Junior School Senior School Total Dadaab Kakuma Kalobeyei Gross Enrolment Rate (%) - Female Gross Enrolment Rate (%) - Male Net Enrolment Rate (%) - Female Net Enrolment Rate (%) - Male Source: Authors’ calculations based on K-LSRH 2023. The pupil-teacher ratios across the camps suggest that while some educational levels in these camps are adequately staffed, others, particularly pre-primary education, face challenges in maintaining optimal teacher-student ratios. In Dadaab, the overall pupil-teacher ratio is approximately 97, with particularly high ratios in pre-primary (146) and senior school (132), indicating potential strain in these areas. Kakuma shows the most variability, with a high ratio in primary schools (162) and a much lower ratio in senior schools (49), leading to an overall ratio of about 88. In Kalobeyei, the overall ratio is close to Kakuma at 89, with pre-primary schools showing the highest ratio (155) and senior schools the lowest (58). These figures suggest that while some educational levels in these camps are adequately staffed, others, particularly pre-primary education, face challenges in maintaining optimal teacher-student ratios. Figure 41: Pupil-teacher ratio by camp and level of education 180 160 140 120 100 80 60 40 20 0 Pre-primary Pre-primary Pre-primary Primary Primary Primary Junior School Senior School Total Junior School Senior School Total Junior School Senior School Total Dadaab Kakuma Kalobeyei Source: Education Management Information System, Term 1, 2024. Building Evidence to Enhance the Welfare of Refugees and Host Communities 41 Figure 42: Classroom utilization by camp and level of education Total Senior School Kalobeyei Junior School Primary Pre-primary Total Senior School Kakuma Junior School Primary Pre-primary Total Senior School Dadaab Junior School Primary Pre-primary 0 20 40 60 80 100 120 140 160 180 Source: Education Management Information System, Term 1, 2024. Classroom utilization varies significantly across the Dadaab, Kakuma, and Kalobeyei camps, with Kalobeyei showing the highest overall classroom-pupil ratio at 139, indicating more crowded classrooms compared to the other camps. Within Kalobeyei, pre-primary education has the highest ratio at 168, followed by primary education at 157, reflecting substantial pressure on classroom space. In Kakuma, the overall classroom-pupil ratio is 132, with primary education experiencing the highest ratio at 165, suggesting high classroom utilization. Dadaab has the lowest overall classroom-pupil ratio at 112, with junior schools particularly underutilized with a ratio of 42, in contrast to the senior schools, which have the highest ratio at 156. These figures highlight the varying levels of classroom overcrowding, with Kalobeyei facing the most significant challenges, particularly at the pre-primary level, while Dadaab's junior schools are relatively less crowded. Cost is the main obstacle to school enrollment for hosts and urban refugees, while camp refugees struggle to transition to secondary school due to insufficient learning and lack of capacity among other reasons. Building Evidence to Enhance the Welfare of Refugees and Host Communities 42 Building Evidence to Enhance the Welfare of Refugees and Host Communities PHOTO/Alamy Socioeconomic Resources 04 to Meet Basic Needs © UNHCR/Samuel Otieno Building Evidence to Enhance the Welfare of Refugees and Host Communities 43 Socioeconomic Resources to Meet Basic Needs 4.1 Humanitarian assistance and remittances The high incidence of multidimensional poverty exists Large gaps remain in social assistance coverage despite almost all refugees receiving assistance among hosts: More than half of all poor / food insecure suggesting that, in the immediate term, withdrawal of households in Turkana do not receive any aid. In other assistance may leave households even more deprived. areas, the situation is even worse, with, for instance, more Table 2 below shows the coverage of aid and food aid across than 9 out of 10 food-insecure hosts in other urban areas locations and refugee status. Almost all refugee households not receiving any assistance. At the same time, many in camps receive aid, and food aid coverage is also nearly non-poor households in camps do receive aid. Such a 100 percent among camp refugees. Social assistance to discrepancy is problematic, and could contribute to creating hosts appears to be targeted to the poorest and/or most resentment against refugees, particularly in areas, such as food insecure households, who are significantly more likely Turkana, where hosts are worse off than refugees. It may to receive both food and non-food aid. One exception are be possible to improve targeting fairness and effectiveness hosts in urban areas other than Nairobi, where there is no by moving from what is essentially a categorical eligibility significant difference between households with different criterion, based on refugee status, to one that is based on levels of poverty or food insecurity. need. Table 2: Percent of households receiving assistance in past 12 months, by location / refugee status/ FOOD AID42 ANY AID43 Non-poor MPI poor Food Food Non-poor MPI poor Food Food secure insecure secure insecure Kakuma (r) 100 99.8 99.8 99.9 100 99.8 99.8 99.9 Kalobeyei (r) 100 99.3 99.5 99.4 100 99.3 99.6 99.2 Turkana (h) 12.7 30.2* 16.1 31.5* 32.1 45.8* 32.1 48.0* Dadaab camp (r) 98.7 98 98 98.9 99.7 98 98.3 99.3 Dadaab host 13.8 14.4 12.7 30.2* 31.8 22.5 24.8 32.2 Nairobi refugee 8.4 7.7 6.9 10.4 14.6 20.7 11.3 20.4* Nairobi host 2.2 9.6 0.9 6.2* 12.7 24.2 9.9 20.5 Urban refugee 9.8 12.3 9.3 14.5 19.9 23.9 19.2 26.8 Urban host 1.3 4.4 1.4 3.4 7.4 4.4 7.1 8.7 Source: Authors’ calculations based on K-LSRH 2023. Notes: r = refugees and h=hosts. *=difference is statistically significant at 5% level. Food related aid includes Bamba Chakula food voucher, Equity Transfer for food (Bamba Chapaa), WFP in-kind food aid, UNHCR firewood, kitchen garden support, Hunger Safety 42 Net Program (HSNP), any other in-kind food aid. 43 Note that households that reported receiving aid more than 12 months ago have been excluded from the statistics on “Any aid”. The food aid question, on the other hand, is a binary yes/no question about the past year that does not specify the time of receipt. Consequently, it is possible that some households answered “yes” to the food aid question but were later excluded from the “Any aid” statistics if they reported receiving aid more than 12 months ago. This explains, for instance, why there are more food aid recipients in Kalobeyei than any aid recipients. 44 Building Evidence to Enhance the Welfare of Refugees and Host Communities The provision of humanitarian assistance may have While the table does not show any significant differences helped to mitigate the impact of food insecurity between 2022 and 2023 among refugees, camp hosts among camp refugees in 2023. Table 3 shows the were significantly more likely to resort to negative coping breakdown of food insecurity by year of interview. The strategies in 2023. Camp hosts also had higher levels of year of interview is relevant given food rations increased food insecurity in 2023 (60.3 percent vs 47.7 percent), from 50 percent to 80 percent of the approximate cost of although the difference is not statistically significant. ideal caloric intake on January 1, 2023, during the survey. Table 3: Food insecurity and negative coping strategies (% of population), by refugee status/ year of interview Index Level Camp refugees Camp Hosts Urban refugees Urban Hosts 2022 2023 2022 2023 2022 2023 2022 2023 CSI 4-9: moderately insecure 18.7 19.7 11.8 12.9 17.6 18.6 18.7 17.3 10+: severely insecure 42.6 42.8 35.9 47.4 48.5 42.9 27.7 27.1 4+: All food insecure 61.3 62.5 47.7 60.3 66.1 61.5 46.4 44.4 LSC Stress 21.5 19.1 22.5 23.3 31.5 39.1 43.6 43.9 Emergency 16.0 17 10.4 31.2* 14.1 9.3 5.3 5.8 Stress + emergency 37.5 36.1 32.9 54.5 45.6 48.4 48.9 49.7 Source: Authors’ calculations based on K-LSRH 2023. Notes: *=difference is statistically significant at 5% level. There is a large reliance on aid among camp refugees on aid places refugees in a vulnerable position, especially and a significant reliance on remittances among urban during periods of fluctuating aid delivery.44 In urban areas, refugees. Aid transfers constitute 79 percent of camp refugees exhibit a better income diversification, with 54 refugees' total income. In comparison, wage earnings percent of income issuing from wage work and 15 percent and profits from self-employed activities constitute only from self-employment, mirroring hosts. Aid reliance is 11 and 3 percent, respectively, of total income, indicating notably smaller, as fewer NGOs extend their programs to that any withdrawal of assistance without expansion of urban areas and refugees migrating out of camps generally income-generating opportunities will leave refugees even forfeit international assistance. Remittances comprise a more deprived (Figure 44). The substantial dependence substantial income share instead (29 percent). 44 Since the conclusion of the survey, funding shortfalls forced food rations to be cut to 60 percent of the caloric measure in July 2023 and again to 50 percent in February 2024. The impact of these cuts will be investigated with future survey waves but are expected to worsen food insecurity in camps. Building Evidence to Enhance the Welfare of Refugees and Host Communities 45 Figure 43: Sources of household income 100% 2% 1% 6% 11% 20% 29% 80% 19% % of household income 60% 79% 15% % aid 26% % remittances 40% % agricultural profits 68% % nonagricultural profits 54% 20% 43% % wage income 6% 3% 11% 0% Camp Camp hosts Urban Urban hosts refugees refugees Source: Authors’ calculations based on K-LSRH 2023. 4.2 Employment 4.2.1 Few camp-based refugees are working case of Kalobeyei, a small plot of allocated land, which cannot be sold or officially purchased, complicating efforts The encampment model limits refugees' employment to relocate elsewhere (Betts, Omata, Rodgers, Sterck opportunities. Refugees do not have the official right to and Stierna 2018). Mobility is further constrained by the work outside the camps, unless a work permit is obtained in requirement for advance travel permits, which are only advance, for which a recommendation from a prospective issued for a limited set of reasons and involve lengthy employer must be accompanied by a letter from the DRS application procedures. Moreover, refugees forfeit their confirming refugee status. This process is complex and entitlement to food and cash assistance upon relocating in practice, permits are rarely issued (UNHCR and World to urban areas, where NGO support is very limited. Bank 2020). Moreover, fleeing their countries under duress often means refugees lack essential documents such as Consequently, few refugees are engaged in an birth certificates, IDs, or educational certificates, which are economic activity, with the majority outside the labor essential for accessing Kenyan services or demonstrating force (Figure 44). Only 14 percent of camp refugees are their qualifications to potential employers. While employed, which is significantly below the 43 percent level refugee entrepreneurs have the opportunity to establish for camp hosts. Most refugees instead remain outside the businesses within the camps by acquiring a business labor force. Women especially face greater challenges, license from the local county government, the annual with only 7 percent employed in Kakuma compared to 14 renewal requirement, coupled with charges contingent on percent of men. A total of 64 percent of refugee women business size, effectively functions as a tax on economic remain outside the labor force. Unlike men, significantly activity (Betts, Sterck and Omata 2018). Refugees lack fewer are inactive due to being in education (12 percent the freedom to settle freely or own land. Upon registration compared to 30 percent for men). with UNCHR in camps, they receive shelter and in the 46 Building Evidence to Enhance the Welfare of Refugees and Host Communities Although refugees in urban areas are almost four women are particularly higher than in camp settings, with times more likely to participate in economic activities on average 45 percent engaged in an economic activity. compared to those in camps, they still lag behind However, similar to the situation in the camps, fewer women Kenyan nationals in terms of employment rates. are employed than men and more remain outside the labor 54 percent of urban refugees are employed, which is force as compared to men. This is true both for refugees encouragingly higher than in camp areas and a reflection and hosts, which may reflect gender-based and cultural of the better availability of jobs in Nairobi.45 Employment norms that refrain women from engaging in economic of refugees, however, still trails behind that of nationals, 63 activities while prioritizing non-paid care and domestic work percent of which are working. Employment rates among (UNHCR and World Bank 2021). Figure 44: Labor force participation 100% 9% % of 18-64 year old population 21% 5% 27% 35% 4% 80% 38% 44% 5% 50% 3% 64% 9% 5% 14% Inactive 60% 11% 5% 3% 6% Education 30% 6% 40% 79% Unemployed 12% 64% 55% 14% 48% 45% Employed+Education 20% 11% 39% 16% Employed 12% 0% Women Men Women Men Women Men Women Men Camp refugees Camp hosts Urban refugees Urban hosts Source: Authors’ calculations based on K-LSRH 2023. Camp refugees who do manage to work do so as or administrative sectors (16 percent, Figure 46).46 Among “incentive” workers for international organizations in wage workers, 90 percent are involved exclusively or the service sector. Wage work is the predominant form of partially in such ‘incentive’ jobs. Although incentive work employment for refugees in camps, encompassing nearly may appear more appealing due to its association with 63 percent of the employed. Those with higher education large international organizations, it is still precarious in and language proficiency can seek special employment nature. It is often on short-term contracts, with few benefits, opportunities with international organizations, including and workers receive reduced 'incentive pay' (Betts et al. roles such as translators, enumerators, community 2018; Vuni and Iragi 2023). Those not engaged in incentive mobilizers, community health workers, and food work mostly rely on the informal service sector, which is assistance distributors, predominantly in the education especially true for urban refugees who do not have access (30 percent), human health and social work (24 percent), to incentive opportunities. 45 In Nakuru and Mombasa the share of employed is notably smaller, with only 29 percent of refugees in employment compared to 58 percent in Nairobi. 46 43 percent of exclusive incentive workers are ‘higher skilled’ based on the International Standard Classification of Occupations (ISCO). This compares to 21 percent not working as incentives. They are also generally more educated, with 24 percent of exclusive incentive workers having completed university and 38 percent completed secondary school, compared to 5 and 12 percent respectively for non-incentive workers. Building Evidence to Enhance the Welfare of Refugees and Host Communities 47 © UNHCR/Pauline Omagwa 48 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 45: Which sectors do refugees work in? 100% 0% 2% 6% Other industries 4% 1% 2% 80% 24% 7% 5% 9% Activities of hou seholds as employers 15% Human h ealth & social w ork % of employed 60% 30% Education 9% Administrative & support 40% 44% 64% ICT 38% Accommodation & food services 20% 5% 20% Transportation & storage 7% 0% Other services Wage worker Wage worker Wholesale & retail trade Exclusive incentive Ow n-accou nt Ow n-accou nt Construction Manufacturing Agriculture, forestry, fishing Camp refugees Urban refugees Source: Authors’ calculations based on K-LSRH 2023. 4.2.2 Factors associated with increased likelihood of being employed High levels of education are positively associated Refugees who have obtained a work permit or with employment outcomes, particularly with holding movement pass are significantly more likely to be incentive jobs. Regressions with simple demographic engaged in the labor force than those without work controls reveal higher levels of education are generally documentation. Refugees are legally entitled to formal associated with more labor force participation (LFP) and employment as they are technically able to access work better employment outcomes (Table 4). Tertiary education permits, seek and gain employment, and start a business. in particular remains a significant predictor for LFP even Yet in practice, administrative hurdles, travel distances to when location, nationality, pre-displacement labor force registration centers and associated costs often prevent status, networks, household composition and mental refugees from obtaining work permits or movement passes. health are controlled for (Table 5). While these results Only less than 1 percent of refugees in the sample have a are largely illustrative and lack the exogenous variation to work permit. Those who do obtain work permits or movement infer causality, they point towards an association between passes have significantly better employment outcomes, higher levels of schooling and LFP. Especially in camps, suggesting there is scope to improve participation by this is likely to be the case due to the high prevalence simplifying administrative processes (Table 5), although it of incentive jobs which favor refugees with higher is also possible that individuals that go through the hurdles educational attainment.47 of obtaining documentation are also more predisposed to engage in the labor force. 47 A similar regression on the sub-sample of camp refugees suggests that among camp refugees, those with tertiary education have a 37 percent higher chance of having an incentive job compared to those without education, while having a work permit is negatively associated with having an incentive job, likely because permits are not needed for incentive work. Building Evidence to Enhance the Welfare of Refugees and Host Communities 49 Table 4: Who works and holds better jobs? (1) (2) (3) (4) LFP Employed Log monthly Log hourly VARIABLES earnings earnings Demographics Age 0.066*** 0.045*** 0.352*** 0.164*** (0.005) (0.004) (0.078) (0.039) Age squared -0.001*** -0.001*** -0.004*** -0.002*** (0.000) (0.000) (0.001) (0.000) RR is a woman -0.041* -0.014 0.277 0.210 (0.024) (0.019) (0.298) (0.161) RR is married 0.049 0.015 -0.058 -0.054 (0.033) (0.029) (0.278) (0.146) Woman x Married -0.170*** -0.163*** -2.032*** -0.784*** (0.040) (0.034) (0.538) (0.277) Educational attainment (no education=omitted) Primary 0.055** 0.036 1.226*** 0.251 (0.027) (0.022) (0.293) (0.162) Secondary 0.052* 0.004 0.788** 0.130 (0.028) (0.022) (0.391) (0.193) Tertiary 0.239*** 0.240*** 0.889** 0.231 (0.047) (0.046) (0.395) (0.193) Constant -0.851*** -0.604*** 0.127 -0.102 (0.092) (0.071) (1.600) (0.777) R squared 0.113 0.104 0.082 0.049 Observations 5542 5542 1801 1792 Source: Authors’ calculations based on K-LSRH 2023. Notes: Modes (3) and (4) were estimated on working refugees only and results are applicable to that subsample.48 Regression results are weighted using survey weights. *** p<0.01, ** p<0.05, * p<0.1. Informal networks matter for finding job opportunities. points (Table 5). Networks also matter in terms of how jobs Refugees whose friendship group includes Kenyans have are acquired. Most paid wage employees found their jobs a 5 percent higher likelihood to participate in the labor directly by approaching employers (38 percent) or through market and 4 percent higher likelihood to be employed, being connected through friends and family (22 percent). when demographics, location, documentation, and mental Fewer submitted a written application (10 percent) and health are controlled for. Having friends outside one’s own almost nobody had a job arranged via their school or country of origin also positively affects LFP by 4 percentage training institution. 48 Models 3 and 4 were also repeated including those who do not work and setting their wages to zero. Results remain the same. 50 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 5: Determinants of LFP for refugees, including extended set of controls (1) (2) (3) (4) VARIABLES LFP Employed Log monthly earnings Log hourly earnings Demographics Age 0.050*** 0.029*** 0.212*** 0.099*** (0.005) (0.004) (0.071) (0.037) Age squared -0.001*** -0.000*** -0.003*** -0.001** (0.000) (0.000) (0.001) (0.000) Is a woman -0.053** -0.038** 0.292 0.176 (0.023) (0.018) (0.321) (0.169) Is married 0.068** 0.034 -0.083 -0.042 (0.033) (0.028) (0.331) (0.168) Woman x Married -0.133*** -0.117*** -1.493*** -0.559** 0.050*** 0.029*** 0.212*** 0.099*** Educational attainment (no education=omitted) Primary 0.006 0.018 0.121 -0.186 (0.029) (0.022) (0.324) (0.174) Secondary 0.015 0.003 -0.121 -0.238 (0.030) (0.023) (0.382) (0.196) Tertiary 0.088* 0.109** -0.217 -0.175 (0.046) (0.043) (0.432) (0.213) Controls Has worked before displacement 0.106*** 0.124*** 0.303 0.056 (0.040) (0.040) (0.466) (0.231) Length of stay (years) 0.001 0.008** 0.001 0.002 (0.005) (0.004) (0.063) (0.035) Length of stay (years) squared -0.000 -0.000** 0.001 0.000 (0.000) (0.000) (0.002) (0.001) Is a single mother 0.008 0.030 -0.017 -0.037 (0.033) (0.025) (0.367) (0.182) Lives in a nuclear family -0.099*** -0.051** 0.027 0.077 (0.028) (0.023) (0.336) (0.174) Lives in a mixed adult household with children -0.080*** -0.075*** -0.629** -0.307* (0.024) (0.018) (0.314) (0.157) Possesses a refugee ID card 0.062*** 0.038** 0.154 0.071 (0.019) (0.015) (0.276) (0.141) Possesses a Movement pass 0.267*** 0.279*** -0.047 0.035 (0.056) (0.059) (0.722) (0.340) Building Evidence to Enhance the Welfare of Refugees and Host Communities 51 Table 5: Determinants of LFP for refugees, including extended set of controls (Continued) (1) (2) (3) (4) VARIABLES LFP Employed Log monthly earnings Log hourly earnings Possesses a Work permit 0.193** 0.065 0.611 0.556 (0.097) (0.098) (0.953) (0.482) Has Kenyan friends 0.055** 0.037** -0.146 -0.228 (0.022) (0.017) (0.323) (0.166) Has refugee friends from different countries 0.036* 0.025 0.785*** 0.348** (0.021) (0.018) (0.282) (0.144) Kenya feels like home (10 on likert) -0.078*** -0.030* -0.041 -0.045 (0.020) (0.018) (0.304) (0.148) PHQ-8 (depression indicator) 0.003 -0.002* -0.063*** -0.029** (0.002) (0.001) (0.023) (0.012) Constant -0.614*** -0.462*** 4.698*** 2.172*** (0.110) (0.082) (1.522) (0.789) Country of origin FE Yes Yes Yes Yes Stratum FE Yes Yes Yes Yes R-squared 0.231 0.273 0.209 0.152 Observations 5508 5508 1792 1783 © UNHCR/Charity Nzomo 52 Building Evidence to Enhance the Welfare of Refugees and Host Communities Older refugees are more likely to be engaged in the non-educational purposes (33 percent; Figure 49). They labor force and command higher wages. With each year entered Kenya with disrupted education and little practical of age, refugees are on average 5 percent more likely job experience and may delay entry into employment in to engage in the labor force and 3 percent more likely favor of resuming their education. Presently, 33 percent to be employed, although returns decrease over time. of camp refugees who were aged 0-18 years at the time They also earn significantly more (Table 5). This is in part of displacement are enrolled in school, while 45 percent due to more experience, but also reflects displacement are not participating in the labor force (Figure 47). Only 10 experiences. Most of the current working age population percent have secured employment which is half the rate of were displaced during their childhood when they were those displaced as adults. primarily engaged in education (42 percent) or inactive for Figure 46: Age at the time of displacement Figure 47: Current labor force participation by age at displacement 100% 100% 17% 13% 32% % of 18-64 year old refugees 23% 22% % of 18-64 year old refugees 30% 29% 80% 80% 37% 23% 46% 45% 54% 28% 61% 66% 65% 60% 35% 60% 48% 40% 49% 40% 49% 62% 65% 99% 20% 41% 20% 38% 27% 20% 18% 15% 7% 10% 6% 0% 6% 2% 3% 7% 0% Not 0-18 18-29 30+ Not 0-18 18-29 30+ Kakuma Refugees Kalobeyei Refugees Dadaab Refu gees Nairobi Refu gees Other Urban Refugees born born Camp refugees Urban refugees Age at displacement Emp loyed Emp loyed + Education Unemployed Education Not born 0-18 years 18-29 years 30+ years Inactive Source: Authors’ calculations based on K-LSRH 2023. Because most refugees were children at the time without job experience and with disrupted education, the of displacement, there is no clear evidence of same dynamic is less likely to occur among those who were occupational downgrading unlike in other contexts. In adults pre-displacement, only 11 percent held a job before Colombia, recent research documented higher educated coming to Kenya. Furthermore, less than half of those who Venezuelan refugees disproportionately entering positions held a job are employed post-displacement (Figure 48). typically employing less-educated natives, leading to higher The subset of individuals who have worked pre- and post- competition and lower wages for the host population and displacement and where occupational downgrading could overall reduced aggregate productivity (Lebow 2024). Due occur is thus very small and unlikely to have a large effect to the majority of refugees entering Kenya as children on the host community.49 49 In addition, the encampment model bars refugees from engaging in certain protected occupations reserved for hosts. Refugees cannot engage in agriculture (except for self- sufficiency use of small kitchen gardens in Kalobeyei), the collection of firewood, or in pastoralist activities among others, which creates a legal barrier to increased competition with the native community. Building Evidence to Enhance the Welfare of Refugees and Host Communities 53 Figure 48: Sankey chart on education or labor force status pre- and post-displacement (for those currently 18Figure Sankey 49:old -64 years ) chart on education or labor force status pre- and post-displacement (for those currently 18-64 years old) Before displacement Before displacement After displacement After displacement Inactive Inactive Education Employed Unemployed Education Age<5 Unemployed Employed Source: Source: Authors’ Authors’ calculations calculaZons based based onK-LSRH on K-LSRH 2023. 2023. 4.2.3 Educa,on and skills trainings are perceived to be key to employment 4.2.3 Education Educa;on and skillsand skills trainings training is the mostare men;oned perceived to be key type to employment of support needed by refugees – both for employment and self-employment. 32 percent Education and skills training is the most mentioned of seeking technical a refugees looking for or job like to gain would training vocational further and 25 percent type of or technical voca?onal support skillsby needed and 23 percent refugees wouldfor – both like to complete desiring their educa?on business (Figure training (Figure 50). 50). Similarly, These results among those intending employment to establish businesses, and self-employment. 32 percent thereof is a significant indicate a needdemand for skills for accessible development, and skillswith education training 26 percentlooking refugees seeking technical for or voca?onal a job would training like to gain further and 25 percent programs desiring among business refugees training currently seeking(Figure 51). employment. technical or vocational skills and 23 percent These results indicate a need for accessible educa?on would like and skills training Approximately a quarter programs refugees amongentrepreneurs of prospective to complete currently their education seeking employment. (Figure 50). Similarly, a Approximately among quarter also of prospec?ve express a need for capital, also entrepreneurs express a underscoring a need need for those intending to establish businesses, there is a for capital, underscoring a need for ini?a?ves to facilitate access to finance. initiatives to facilitate access to finance. significant demand for skills development, with 26 percent 53 54 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 49: The main support needed by those looking Figure 50: The main support needed by those Techn ical or Techn ical or for employment is education, vocational training and looking to become self-employed is vocational vocational s k contacts with employers vocational s kills training, business skills and credit training training 4% 3% 3% 3% Bu siness trai Continue or 5% complete 8% education 26% 32% Securing contacts 12% Loan or cred 11% with other emp loyers Support with Support with 19% 25% access to markets 25% access to ma 23% Securing other Childcare documentation opp ortu nitie Technical or vocational skills- support with access to Technical or vocational support with access to training markets skills-training markets help with ho Continue or complete education Securing other Childcare opportunies/ Securing work Business Training Con tinu e or dodumentation help with housing Securing contacts with other permit Loan or credit complete employers Securing work permit Continue or complete education education Source: Authors’ calculations based on K-LSRH 2023. 4.2.4 Demand-side and other constraints Wages are low overall but there are also some wage below rates for hosts. Although urban refugees engage disparities between refugees and hosts. The median in similar work to hosts, mostly including wholesale and monthly wage for refugees was Ksh. 8,000 compared to retail trade, their earnings average Ksh. 11,900 a month Ksh. 15,000 for hosts, nearly twice the amount (Figure 51). compared to Ksh. 21,000 for hosts. This divergence signals Earnings for camp refugees are especially low, averaging that refugees are often working in precarious situations around Ksh. 6,800 per month in Kakuma and Kalobeyei and and suggests a need for policies addressing the economic Ksh. 8,400 in Dadaab, which in both cases is significantly integration and wage parity of refugees in urban areas. Figure 51: Salaries of paid employees and profits from own-account work in non-agricultural businesses 35,000 30,000 25,000 20,000 Ksh. 15,000 10,000 5,000 0 Sal ary Profit Sal ary Profit Sal ary Profit Sal ary Profit Sal ary Profit Sal ary Profit Turkana refugees Turkana hosts Dadaab refugees Dadaab hosts Urban refugees Urban hosts Mean Median Source: Authors’ calculations based on K-LSRH 2023. Note: Monthly salaries and profits depicted are of the primary activity only. Building Evidence to Enhance the Welfare of Refugees and Host Communities 55 Most employed refugees would like to work more hours. 65 percent of employed refugees work more than 35 hours a week, yet an overwhelming majority would want to work more hours if paid (Figure 52). This is true for both men and women, although notably fewer women can start additional work within two weeks if they were offered a job (Figure 53). Figure 52: Proportion of employed refugees who would like to work more hours if they were paid 100% 78% 83% 76% 76% 75% 80% 71% 70% 70% % of employed 60% 40% 20% 0% Women Men Women Men Women Men Women Men Camp refugees Camp hosts Urban refugees Urban hosts Source: Authors’ calculations based on K-LSRH 2023. Figure 53: Proportion of refugees who would be able to work more hours in the next two weeks if offered a job 100% 87% 85% 82% 81% % want to work more 80% 70% 70% 66% 58% 60% hours 40% 20% 0% Women Men Women Men Women Men Women Men Camp refugees Camp hosts Urban refugees Urban hosts Source: Authors’ calculations based on K-LSRH 2023. Many refugees outside the labor force are discouraged a lack of jobs is still the primary reason for inactivity. In by the lack of available jobs. 85 percent of refugees addition, access to documentation emerges as a critical without an economic activity were not looking for a job or barrier to employment, with 27 percent of men citing it an opportunity to start a business in the last four weeks. as their primary obstacle. Overall, a lack of opportunity Among men, 61 percent cite a lack of jobs as the main appears to be affecting all groups, including those living in reason for not looking, while among women, pregnancy, urban areas and hosting communities. Given the difficult childcare and household responsibilities are the most labor market from the demand side, policies increasing cited reasons, followed closely by a lack of jobs (Figure labor supply alone are not necessarily going to increase 54). While urban areas offer relatively more opportunities, employment. 56 Building Evidence to Enhance the Welfare of Refugees and Host Communities © UNHCR/Charity Nzomo 53% Sickness/injury 30% % of ina Unable to find work matching s 39% 43% Lack of capital 20% 31% 3% No jobs available 28% 10% Pregnancy, childcare, househol 10% Building Evidence to Enhance the Welfare of Refugees and Host Communities 57 16% Full time student 3% 1% 8% 10% 5% 4% 3% 4% 0% 0% Figure 54: Reasons for not looking for jobs 100% 90% 80% % of inactive, out-of-school population 11% Other reasons 70% Retired/too old Lack of documentation 60% 31% 23% Unable to work (incapacitat 50% 61% 40% Sickness/injury 40% 65% Unable to find work matchi 34% 53% Lack of capital 30% 39% 43% No jobs available 100% 20% 31% 3% Pregnancy, childcare, house 28% 10% 90% 10% Full time student 16% 3% 1% 8% 10% % of inactive, out-of-school population 80% 11% 0% 5% 4% 3% Other reasons 4% 0% 70% Women Men Women Men Retired/too Women old Men Women Men Lack of documentation 60% Camp 23% refugees Camp hosts Urban refugees Urban hosts 31% Unable to work ( incapacitated) 50% 61% 40% Sickness/injury 40% 11% Other reasons 65% Unable to find work matching skillset 34% 53% Lack of capital 30% Retired/too old 39% 43% Lack of documentation No jobs available 20% 31% 23% 3% Pregnancy, childcare, household responsibilities 28% Unable to work ( incapacitated) 10% 10% Full time student 16% Sickness/injury 3% 1% 8% 10% 0% 5% 65% 0% 4%to find work Unable 4% skillset 3% matching 34% Women Men Men Women Lack Women of capital Men Women Men Source: Authors’ calculations based on K-LSRH 2023. 43% Camp refugees hosts CampNo Urban refugees jobs available Urban hos ts 31% Pregnancy, childcare, household responsibilities 10% Full time student 8% Women 10% many of the same challenges as men but face 1% only spend 18 hours on childcare, almost half the time as 3% 4% also higher childcare and household responsibilities. women. In urban areas, the difference is even stronger. Women Men Women Men 39 percent of women in camps and 43 percent in urban After controlling for demographics, location, networks and Urban refugees Urban hos ts settlements reported pregnancy, childcare and household mental health, there is a significantly negative relationship responsibilities as the most important reason for staying between hours spent on childcare and employment albeit outside the labor force (Figure 54). Overall, women in camps small in magnitude – each hour spent on childcare reduces spend 32 hours each on childcare and household chores per the likelihood of being employed by 0.1 percent (Table 6). week (Figure 55). In comparison, male refugees in camps 58 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 6: Regression - Determinants of employment, hours worked, and earnings for women (1) (2) (3) (4) Log monthly Log hourly VARIABLES LFP Employed earnings earnings Demographics Age 0.037*** 0.026*** 0.271** 0.146** (0.008) (0.005) (0.122) (0.058) Age squared -0.000*** -0.000*** -0.003** -0.002*** (0.000) (0.000) (0.001) (0.001) Is married -0.066** -0.088*** -1.576*** -0.642** (0.030) (0.024) (0.543) (0.280) Educational attainment Primary -0.005 0.014 0.049 -0.260 (0.036) (0.025) (0.487) (0.256) Secondary -0.042 0.002 -0.765 -0.603* (0.043) (0.032) (0.696) (0.355) Tertiary 0.070 0.179** 0.644 0.159 (0.087) (0.083) (0.606) (0.331) Childcare Hours spent on childcare -0.002*** -0.001*** -0.002 -0.003 (0.000) (0.000) (0.008) -0.005 Controls Has worked before displacement 0.068 0.130** 0.847 0.372 (0.067) (0.063) (0.854) (0.442) Length of stay (years) 0.004 0.015*** 0.038 0.046 (0.008) (0.005) (0.095) (0.050) Length of stay (years) squared -0.000 -0.000*** 0.000 -0.000 (0.000) (0.000) (0.002) (0.001) Is a single mother 0.094** 0.093*** 0.092 0.024 (0.041) (0.031) (0.400) (0.205) Lives in a nuclear family -0.009 -0.003 0.252 0.126 (0.045) (0.036) (0.691) (0.346) Lives in a mixed adult household with children -0.007 -0.020 -0.222 -0.138 (0.037) (0.031) (0.411) (0.222) Possesses a refugee ID card 0.046* 0.019 0.618 0.408 (0.027) (0.021) (0.509) (0.270) Possesses a Movement pass 0.362*** 0.336*** -0.029 0.001 (0.078) (0.079) (1.078) (0.488) Has Kenyan friends 0.032 -0.005 0.424 0.028 (0.032) (0.024) (0.523) (0.272) Has refugee friends from different countries 0.032 -0.005 0.424 0.028 (0.032) (0.024) (0.523) (0.272) Kenya feels like home (10 on likert) -0.130*** -0.071*** 0.318 0.132 (0.025) (0.021) (0.447) (0.212) PHQ-8 (depression indicator) 0.007*** 0.002 -0.105*** -0.039** (0.002) (0.002) (0.037) (0.019) Constant -0.453*** -0.501*** 1.602 0.128 (0.157) (0.102) (2.789) (1.321) Country of origin FE Yes Yes Yes Yes Stratum FE Yes Yes Yes Yes R-squared 0.166 0.243 0.256 0.215 Observations 2787 2787 738 735 Source: Authors’ calculations based on K-LSRH 2023. Notes: Regression results are weighted using survey weights. *** p<0.01, ** p<0.05, * p<0.1. Building Evidence to Enhance the Welfare of Refugees and Host Communities 59 Figure 55: Hours spent on childcare and household chores 40 35 32 32 Hours spent in last 7 days 30 25 23 22 20 20 19 20 18 15 15 15 11 10 7 8 6 5 3 3 0 Women Men Women Men Women Men Women Men Camp refugees Camp hosts Urban refugees Urban hosts Inactive population Childcare Chores Source: Authors’ calculations based on K-LSRH 2023. 60 Building Evidence to Enhance the Welfare of Refugees and Host Communities 05 © UNHCR/Mohamed Aden Maalim (Dis-)enablers to Sustaining Self-Reliance and Resilience © UNHCR/Pauline Omagwa Building Evidence to Enhance the Welfare of Refugees and Host Communities 61 (Dis-)enablers to Sustaining Self-Reliance and Resilience 5.1 Social cohesion Social cohesion, defined as "the degree to which bonds (25 percent in Kakuma and 34 percent in Kalobeyei) view exist within a community or society, fostering trust among hosts as trustworthy. By contrast, in Garissa County, 80 unfamiliar individuals, a willingness to collaborate, and percent of hosts around Dadaab camp reported feeling faith in institutions” (De Berry and Roberts 2018; Barron that refugees are generally trustworthy. Among refugees, et al. 2023) may enhance integration efforts and provide over half in urban areas and the Dadaab camp expressed a mutually enabling environment for both refugees and strong trust in hosts, as well as among refugees from hosts. This sub-section therefore focuses on trust and different tribes or clans (Figure 56, Figure 57). Even after safety as components of social cohesion given its potential controlling for other factors, location remains associated to enable or limit self-reliance and resilience. with trust – Dadaab and urban hosts demonstrate higher levels of trust in refugees compared to their counterparts Trust perceptions are low in Turkana County including for hosts towards refugees, refugees towards hosts in Turkana. These findings are in line with those of Betts and even among refugees. Only slightly over a third of et al. (2022), who observes that urban hosts in East Turkana hosts view refugees as trustworthy. Similarly, Africa are more likely to perceive refugees as friendly or only a small proportion of refugees in Turkana County trustworthy. Figure 56: Perception of trust towards out-group a. Refugees are trustworthy (hosts' perceptions) b. Hosts are trustworthy (refugees’ perceptions) 100% 100% 100% 19.9% 19.9% 21.2% 90% 22.7% 22.7% 21.5% 26.2% 90% 21.5% 12.3% 80% 38.8% 80%80% 19.9% 19.9% 52.8% 52.8% 70%70% 58.8% 58.8% 22.7% 22.7% 21.5% 21.5% 60% 32.2% 30.9% 60%60% 27.0% 50%50% 40% 78.0% 40%40% 13.6% 13.6% 16.0% 16.0% 68.2% 68.2% 42.8% 46.5% 3 0%3 0% 58.3% 57.7% 20% 58.3% 57.7% 34.2% 20%20% 33.6% 33.6% 0% 25.3% 10% 25.3% 10% Turkana Dadaab Nairobi Other hosts hosts hosts urban hosts 0% 0% Kakuma Kakuma Kalobeyei Kalobeyei Dadaab Dadaab Nairobi Other Nairobi Other Strongly disagree/disagree refugees refugees refugees refugees refugees refugees refugeesurban refugees urban refugees refugees Neither di sagree/agree Strongly agree/agree Strongly Strongly disagree/disagree disagree/disagree Neither Neither di sagree/agree di sagree/agree Strongly Strongly agree/agree agree/agree 62 Building Evidence to Enhance the Welfare of Refugees and Host Communities c. Refugees from different ethnic groups are trustworthy (refugees' perceptions) 100% 14.0% 14% 26% 20% 26.1% 20.3% 80% 41.8% 42% 38.8% 39% 30.7% 31% 25% 21% 25.4% 60% 21.1% 18.2% 22% 22.1% 18% 40% 55% 53% 54% 55.3% 52.8% 54.4% 20% 36% 43% 42.9% 36.1% 0% Kakuma Kalobeyei Dadaab Nairobi Other urban refugees refugees refugees refugees refugees Strongly agree/agree Neither di sagree/agree Strongly disagree/disagree Source: Authors’ calculations based on K-LSRH 2023. Besides location, social networks and perceived finds them positively associated with trust (Fajth et al. 2019; social support are positively associated with higher El-Bialy et al. 2023). perceptions of trust towards individuals from groups Employment, poverty status and education level are different from one’s own (out-groups). Regression also associated with trust perceptions among refugees analysis (Table 7) suggests that all else being equal, having (Table 7). Being employed and having higher education more friends from the out-group is significantly associated are associated with decreased trust in both neighbors with higher trust in members of the out-group. Specifically, and fellow refugees, possibly suggesting that limited among refugees, having more Kenyan friends is associated opportunities for refugees creates a situation where the few with higher perception of trustworthiness of hosts. Similarly, that get opportunities may see others as a threat. Similarly, for hosts, having more refugee friends enhances their trust low household welfare, proxied by multidimensional in refugees. Across all groups, individuals with a higher poverty, is associated with lower trust in both neighbors self-reported social support are more likely to report higher and fellow refugees even among those of the same tribe. perceptions of trust. Active membership in associations On the other hand, being married is associated with higher and community groups is also associated with higher trust in hosts and refugees of different tribes compared to trust of refugees towards hosts and refugees of different being unmarried or divorced. Possible hypotheses for the origin. However, among hosts, active membership in a few association between marital status and trust in out-group associations and community groups is associated with lower could be that individuals that are married are predisposed to trust in refugees. Overall, the results are largely consistent trust to begin with, being married may create opportunities with research in displacement contexts which highlights to foster trust and/or individuals may be married to spouses social networks and group memberships as fundamental from a different group, which increases interaction with out- elements of a cohesive and collaborative community and groups and possibly fosters feeling of trust. Building Evidence to Enhance the Welfare of Refugees and Host Communities 63 Table 7: Factors correlated with perceptions of trust Hosts trust- Hosts Refugees Refugees Refugees Refugees ing neigh- trusting trusting trusting trusting trusting ref- bors refugees neighbors hosts refugees ugees from from different same tribes tribes Employment status 0.088 -0.059 -0.362*** -0.098 -0.195* -0.247** Age 0.009 0.004 0.000 -0.002 0.005 0.003 Gender (female_dummy) 0.053 -0.060 -0.083 -0.064 -0.091 -0.011 Marital status (married_dummy) -0.108 -0.109 0.097 0.305*** 0.198** 0.066 Multidimensional poverty -0.530* -0.020 -0.591*** -0.076 -0.300** -0.591*** Receipt of assistance -0.035 -0.120 -0.190 -0.135 -0.508*** -0.170 Migration experience (Ref: <2 yrs) Migration experience (3-5 yrs) 0.003 -0.024 -0.628 0.095 Migration experience (6-10 yrs) 0.334 0.034 -0.307 0.353 Migration experience (10+ yrs) 0.599 0.220 -0.066 0.518 Secondary education or higher 0.132 -0.070 -0.185 -0.157 -0.177* -0.341*** Location (Ref: Turkana Refugees) Dadaab Refugees 0.538*** 1.511*** 0.809*** 0.358*** Nairobi Refugees 0.337 1.570*** 0.351* 0.836*** Other Urban Refugees 0.554** 1.832*** 0.374* 0.667*** Location (Ref: Turkana Hosts) Dadaab Hosts 0.363* 2.179*** Nairobi Hosts -1.792*** 0.543** Other Urban Hosts -0.478 0.668*** Social Network % of refugee friends (>50%) 0.267 0.673*** % of Kenyan friends (>50%) -0.197* 0.204* 0.056 -0.321*** % of refugee friends from the same 0.090 -0.180* -0.183* 0.215** country (>50%) % of refugee friends from different -0.185 -0.101 0.186 -0.137 country (>50%) % of refugee friends from different tribe 0.030 0.115 0.164 0.074 (>50%) Membership to groups (Ref: None) Active membership (1 to 3 groups) -0.240 -0.353*** 0.153 0.018 0.124 0.202* Active membership (4 to 6 groups) -0.707 0.708 1.241*** 0.921** 1.052*** 0.794** Active membership (7 or more groups) 0.527 0.548 0.846 2.068*** 1.448*** 1.234*** Multidimensional Scale of Perceived Social Support (Ref: Low support) Moderate support 0.193 0.263 0.442*** 0.395*** 0.290** 0.509*** High support 0.723** 0.861*** 1.249*** 0.934*** 0.845*** 1.305*** Observations 3,023 3,010 4,552 4,551 4,550 4,551 Source: Authors’ calculations based on K-LSRH 2023. Notes: The multidimensional scale of perceived social support (MSPSS) score is derived from the summation of all items in the MSPSS (from special person, family, friends); Low support” corresponds to MSPSS scores ranging from 0 to 2.99. “Moderate support” corresponds to MSPSS scores ranging from 3 to 5. “High support” corresponds to MSPSS scores ranging from 5.01 to 7; Migration experience refers to length of stay in Kenya; Regression results are weighted using survey weights; *** p<0.01, ** p<0.05, * p<0.1 64 Building Evidence to Enhance the Welfare of Refugees and Host Communities © UNHCR/Samuel Otieno Perceptions of safety are low in Turkana County as combination of factors including socioeconomic challenges, well as in Nairobi. While across all locations a majority resource scarcity and strained communal relations in these of individuals feel safe during the day, perceived nighttime areas. safety is low particularly in Turkana and Nairobi (Figure Socialization and views on integration 68). More than 7 out of 10 refugees in Turkana County do not feel safe walking at night compared to 3 out of 10 For child refugees, having Kenyan friends, feeling safe refugees in Garissa County. Hosts in Dadaab report highest at school, and being treated fairly by adults at school is significantly associated with their sense of belonging safety perceptions with more than 6 out of 10 feeling safe in Kenya (Table 8). Similarly, for adult refugees, being walking even during the night. All else being equal, women part of community groups and feeling supported by others are more likely to feel unsafe walking alone at night (Table is positively associated with a sense of belonging. The 22). The perception of crime is also higher among refugees results suggest that refugees tend to feel more at home in Turkana and Nairobi hosts with over half saying crimes when they have support networks in place and underscores are common in their respective neighborhoods (Figure the importance of social connections in fostering a sense 70). This heightened perception of crime may be due to a of belonging, regardless of the physical location (Table 9). Building Evidence to Enhance the Welfare of Refugees and Host Communities 65 Table 8: Regression on correlates of feeling like Kenya is home (refugee children) Kenya feels like home (Children) Multidimensional poverty -0.578* Care giver’s employment status 0.042 Gender 0.295 Age -0.031 Location (Ref: Dadaab Refugees) Kakuma Refugees -0.570 Kalobeyei Refugees -0.269 Number of Kenyan friends (Ref: None) Number of Kenyan friends (One) 0.103 Number of Kenyan friends (2-5 ) -0.057 Number of Kenyan friends (>5 ) 0.931** Number of refugees friends from different country (Ref: None) Refugees’ friends from different country (One) -0.123 Refugees’ friends from different country (2-5) -0.039 Refugees’ friends from different country (>5) -0.373 Feel safe at school 0.954*** Teachers support me when I need them -0.062 Adults at my school treat students fairly 0.614*** Observations 443 Source: Authors’ calculations based on K-LSRH 2023. Notes: Regression results are weighted using survey weights; *** p<0.01, ** p<0.05, * p<0; Feel safe at school, Teachers support me when I need them and Adults at my school treat students fairly are dummy variables (1=Strongly agree/agree, 0=Strongly disagree/disagree). 66 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 9: Regression on correlates of feeling like Kenya is home (refugee adults) Kenya feels like home (Adult) Employment status 0.048 Age -0.007 Gender 0.113 Marital status (married_dummy) -0.123 Poverty status (multidimensional) -0.211* Receipt of assistance -0.030 Migration experience (Ref: <2 yrs) Migration experience (3-5 yrs) 0.311 Migration experience (6-10 yrs) 0.246 Migration experience (10+ yrs) 0.669 Secondary education or higher -0.093 % of friends that are Kenyan (>50%) 0.206* Membership in groups (Ref: None) Active membership (1 to 3 groups) 0.037 Active membership (4 to 6 groups) 1.101*** Active membership (7 or more groups) 1.906*** Multidimensional Scale of Perceived Social Support (Ref: Low support) Moderate support 0.742*** High support 1.405*** Location (Ref: Turkana Refugees) Dadaab Refugees 0.887*** Nairobi Refugees 0.885*** Other Urban Refugees 1.247*** Observations 5,514 Source: Authors’ calculations based on K-LSRH 2023. Notes: Regression results are weighted using survey weights; The multidimensional scale of social support (MSPSS) score is derived from the summation of all items in the MSPSS (from special person, family, friends); Low support" corresponds to MSPSS scores ranging from 0 to 2.99. "Moderate support" corresponds to MSPSS scores ranging from 3 to 5. "High support" corresponds to MSPSS scores ranging from 5.01 to 7; Migration experience refers to length of stay in Kenya; *** p<0.01, ** p<0.05, * p<0.1 Location, membership in groups and social networks to have less positive views on this matter. are also associated with adult’s attitudes towards Preferences for integration are strong among refugee children socializing with out-groups. Generally, children. Most refugee children would prefer to attend a respondents in Dadaab and urban areas express greater mixed school (refugees and hosts), one of many elements comfort with children socializing with the out-group of integration. The preference is associated with both the community (Figure 69). Regression analysis (Table 23) caregiver's education level and their connections within the reveals that for hosts, besides location, some memberships Kenyan community. Children whose caregivers have higher in social groups and having refugee friends increases the education levels are more likely to prefer mixed schools likelihood of feeling comfortable with children socializing compared to those with only primary education or no formal with refugee children. The results point to the potential education. Having at least one Kenyan friend increases role of contact and interactions in building social cohesion the likelihood of preferring mixed schools. These findings (Hewstone 2015). Refugees who have more friends from underscore the importance of both caregiver education and the same country tend to have more positive views about social networks in shaping school preferences for children their children associating with host children. Conversely, (Table 24). those who have more friends from different countries tend Building Evidence to Enhance the Welfare of Refugees and Host Communities 67 5.2 Socioeconomic and environmental shocks Location matters more than host/refugee status in respondents reporting having experienced a flood and terms of having experienced a shock and in terms of drought, respectively, compared to less than 20 percent the type of shock. Refugees and hosts in camp areas of urban respondents having experienced a flood and experience nearly twice as many shocks as those living less than 20 percent a drought. This can be explained by in urban contexts. The difference in the kind of shocks the fact that camps are in arid and semi-arid regions. On experienced among camp and urban respondents the other hand, respondents in urban areas report having show that camp locations are more susceptible to the experienced theft, unemployment, and loss of land/rental environment. Floods, droughts, landslides, pests, weeds, property at a higher rate than camp locations. Although livestock disease, and water-borne disease outbreaks, overall, in camp areas, there is no statistically significant are all more common in camps than in urban locations. difference between refugees and hosts in the aggregate The difference between urban and camp areas is number of shocks, the type differs. Camp hosts are more particularly stark in the experience of floods or droughts affected by pests, weeds, and livestock disease, compared – about 50 percent and around 70 percent of camp to camp refugees. Figure 57: Median number of shocks experienced across locations 6 5 4 3 2 1 0 Camp Refugees Camp Hosts Urban Refugees Urban Hosts Source: Authors’ calculations based on K-LSRH 2023. Notes: Shocks include excessive rains/flooding, Drought, Landslides/erosion, Crop disease, Weeds affecting crops, Livestock disease, Human disease from contaminated water, Theft or destruction of assets, Delay in food assistance/ aid, increase in food prices, loss of land/rental property, Loss of employment, Death of household member. Resilience to shocks is low. Food inflation is the most reported coping strategy across shocks was doing nothing common type of shock experienced with over 90 percent followed by using savings (Table 26 in Annex). There is thus of all respondents reporting facing high food prices. scope to strengthen resilience to shocks particularly in camp The most common strategy used to cope with increased areas to ensure access to diversified coping strategies. food prices was reducing food consumption (more than a Additionally, across all locations, individuals share the third of respondents in all locations), followed by reducing perception that the weather has become hotter and non-essential expenses. Using savings was reported more drier. Approximately 73.4 percent of individuals say there frequently by urban respondents than camp respondents, is less rain, 66.9 percent say there are more droughts, with 30 percent and 18 percent of camp refugees and hosts 80.5 percent report more hot days. Planting trees is a respondents reporting using this strategy, while 39 percent common strategy across locations in response to hotter/ and 47 percent of urban refugees and hosts used their drier weather, although reducing the number of livestock is savings. For other shocks (Table 25 in Annex), the most common in Garissa. 68 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 58: Perceptions of weather variability within the preceding five years 100% 80% % of respondents 60% 40% 20% % Turkana Dadaab Nairobi Other urban Less rain More droughts more hot days Source: Authors’ calculations based on K-LSRH 2023. 5.3 Psychosocial wellbeing Previous research reports that depression and depressive symptoms among refugees compared to the anxiety are particularly common mental challenges host population. Similarly, Klabbers et al. (2022) reported among refugees, often related to the loss of loved higher rates of post-traumatic stress disorder (PTSD), ones, homes, jobs, social connections, and a sense depression, anxiety, and lack of social support among of hope for the future.50 The K-LSRH explores self- refugees and asylum seekers compared to nationals in reported experiences that may indicate symptoms of Nakivale refugee settlement, Uganda. Refugees are also depression and anxiety (Box 4). Overall, refugees are more likely to show symptoms of anxiety. Urban refugees, more likely to experience symptoms of depression than particularly in Nairobi, are more likely to show symptoms hosts across all surveyed locations except in Dadaab. of moderate/severe anxiety than camp-based refugees These results are consistent with the few existing studies (Figure 59) which may reflect the precarious status on psychosocial wellbeing among refugees in East Africa. refugees have outside camps. Pozuelo et al. (2023) finds higher prevalence of elevated Box 4: Measuring psychosocial wellbeing The K-LSRH uses the Patient Health Questionnaire (PHQ-8) for screening symptoms of depression. The PHQ-8 is a self-reported measure used to screen depression symptoms and it is not a diagnostic tool. The scale includes eight questions with overall scores ranging from 0 to 24, scores of 0-4 indicating no symptoms of depression, 5–9 = mild, 10–14 = moderate, 15–19 = moderately severe and ≥20 = severe symptoms of depression (Kroenke et al. 2001). Anxiety is screened by the Generalized Anxiety Disorder (GAD-7) scale which is not a diagnostic tool. It is made up of seven questions each with responses ranging from 0-3. Scores of 5,10 and 15 are taken as the cut-off points for mild, moderate, and severe prevalence of anxiety symptoms, respectively (Spitzer et al. 2006). 50 See for example Gaynor (2020). Building Evidence to Enhance the Welfare of Refugees and Host Communities 69 ©UNHCR /Charity Nzomo 70 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 59: Self-reported experiences that may indicate symptoms of depression and anxiety across a. Self-reported symptoms of depression 100 80 % of population 62 62% 64 64% 61 61% 68 60 78 78% 83 83% 76 76% 68% 82 88 88% 82% 40 24 24% 19 19% 22 22% 20 24 16 16% 16 16% 23% 8 8% 12 12% 14 14% 12 12% 12 12% 8 8% 7 7 0 5 5% 1 5 5% 5% 5 4 4% 3% 3 3% 3 2 7% 7% 1 5 5% 1% 0 0% 0 0% 0 0% 2% 1 1% 1% 1 1% Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban Refugees Hosts Severe Moderately severe Moderate Mild None/ Minimal b. Self-reported symptoms of anxiety 100 90 80 70 % of population 64 64% 66 80% 61 74% 66 81% 60 66% 80 61% 74 66% 91% 89% 81 91 89 50 40 25% 30 25% 23% 25 28% 20 25 23 20% 28 20 16% 10 16 16% 10 10% 16 8 8% 10 10% 7 7% 9 9% 5 5 0 3 3% 1 1%4 4% 2 2% 1 1% 4 4% 5% 5% 1 2 2% 1 1% 1% Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban Refugees Hosts Severe Moderate Mild None/Minimal Source: Authors’ calculations based on K-LSRH 2023. Building Evidence to Enhance the Welfare of Refugees and Host Communities 71 Marital status, food insecurity and household size are such as their concentration in an area where they share some of the factors associated with psychological closer cultural ties with the host population in Garissa wellbeing. Regression analysis (Table 10) suggests County or, alternatively, a particular stigmatization related that all else being equal, individuals who are divorced, to reporting symptoms of psychological distress.61 On the widowed, or married report higher likelihood of depression other hand, being a host rather than a refugee and better and anxiety symptoms compared to those who have never living conditions, as proxied by number of habitable rooms been married. Individuals facing higher food insecurity per household member, appears to be linked with lower also show higher prevalence of symptoms of distress depression and anxiety symptoms. It is important to note when controlling for other factors. Additionally, living in that since psychological wellbeing outcomes are based urban areas and being from South Sudan, DR Congo, and on self-reported symptoms, results may be biased due Ethiopia is associated with higher self-reported depression to potential stigma and unobserved contextual factors. and anxiety symptoms compared to being from Somalia. Understanding the cultural context around perceptions of Lower reported self-rates of distress among Somali psychological wellbeing is an avenue where qualitative refugees may be linked to cultural and contextual factors research can provide further insights. Table 10: Factors correlated with symptoms of depression or anxiety. Variables Symptoms of depression or anxiety Age 0.002 Female 0.044 Is household head 0.068 Ever attended school -0.036 Employed -0.073 Marital status (Ref: Never married) Divorced/Separated 0.460*** Widow 0.288** Married/Couple 0.156*** Urban 0.228*** Host -0.237*** Country of origin (Ref: Somalia)* South Sudan 0.392*** DR Congo 0.498*** Ethiopia 0.492*** Others 0.293*** Female headed household 0.032 Household Size -0.028*** Number of habitable rooms per household member -0.217*** Food insecurity index 1.090*** Observations 8,570 Source: Authors’ calculations based on K-LSRH 2023. Notes: *** p<0.01, ** p<0.05, * p<0.1; The outcome variable is based on the GAD-7 score and PHQ-8 score using Principal Component Analysis. Re- gression results are weighted using survey weights. *Kenya omitted as country of origin since it fully overlaps with the host community indicator. 61 Pozuelo et al. (2023) also find that refugees from Somalia were less likely to report depression symptoms compared to those from DR Congo and South Sudan. 72 Building Evidence to Enhance the Welfare of Refugees and Host Communities 06 PHOTO /Alamy Suggestions for Policy Direction Building Evidence to Enhance the Welfare of Refugees and Host Communities 73 Suggestions for Policy Direction The Kenya Longitudinal Socioeconomic Study of more freely. This approach can help refugees transition Refugees and Host Communities (K-LSRH) is a towards self-reliance and expand economic opportunities, pioneering survey offering comparable data on ultimately leading to reduced hosting costs in the long run refugees and hosts in Kenya. The first wave of the panel (World Bank and UNHCR forthcoming). In Uganda, allowing was conducted between June 2022 and 2023, covering refugees to work and move freely is estimated to lower the refugees in Kakuma Refugee Camp, Dadaab Refugee cost of assistance by an estimated US$225 million annually Complex, Kalobeyei Integrated Settlement, as well as urban (Atamanov, Hoogeveen and Reese 2024). In Chad, the refugees in Nairobi, Mombasa, and Nakuru. It focused on policy of inclusion and dispersion caters for 54 percent of household, individual, and children’s outcomes, as well the cost of basic needs of refugees (Coulibaly, Hoogeveen as understudied themes such as psychosocial wellbeing and Savadogo 2024). and trust among hosts and refugees. This survey can be Upskilling and job search support are essential informative for policy including Kenya's shift from refugee for empowering refugees, equipping them with camps to integrated settlements. the education and skills necessary for successful Despite policy advancements, refugees still face integration into host communities. Many refugees, significant challenges, such as restricted movement especially those who were children when displaced, have and limited employment opportunities. Key findings experienced interrupted education, and lack practical highlight reliance on humanitarian assistance, unmet basic skills upon arrival in Kenya. There is a strong demand needs, barriers to education and employment, and greater for education among refugees, with a significant portion vulnerability among camp refugees and hosts. Below we of the potential labor force currently enrolled in school provide suggested policy directions underscoring priorities and expressing a need for technical and vocational skills in the short run (movement to where opportunities are to secure employment. Supporting refugees in their job available); factors that will help ease the transition (place- search, particularly those with limited connections, is vital. based development) and investments that are essential for Research into informal job acquisition networks could sustained self-reliance and resilience (such as education provide valuable insights for policy aimed at improving and psychosocial wellbeing). labor market outcomes for refugees. By focusing on both upskilling and job search assistance, Kenya can help 6.1 Short run: Unlocking access to refugees build pathways to economic self-sufficiency and socioeconomic opportunities through successful integration. movement and complementary measures Demand-side solutions, including self-employment, are Easing restrictions on the right to work and movement crucial for addressing job scarcity in refugee camps, is crucial for enabling refugees to achieve a dignified despite movement restrictions and limited resources. life. Refugees without work and movement passes face Supporting microbusinesses and establishing market significant barriers to economic activities. Furthermore, linkages beyond the camps can enhance job opportunities. access to formal finance is nearly nonexistent, with 89 Both refugees and host communities consistently cite percent of refugees lacking a bank account and less than the scarcity of jobs as a major obstacle to employment, 1 percent having a formal loan. Streamlining procedures particularly within the camps. Efforts should focus on for accessing work permits, granting freedom of movement, creating jobs that benefit both refugees and hosts, exploring ensuring refugee IDs are accepted for government and strategies to facilitate self-employment while addressing private sector, facilitating access to banking and phone barriers such as limited access to finance. Current self- registration will allow refugees to participate in the economy employment ventures tend to be small-scale and informal, 74 Building Evidence to Enhance the Welfare of Refugees and Host Communities and this pathway is likely to be a realistic pathway for arid regions with limited access to infrastructure and some population segments, both within refugee and services and susceptibility to shocks. Turkana and host communities. Parallel efforts can be pursued for the Garissa counties record some of the highest poverty rates generation of formal employment although such initiatives in the country with little progress in poverty reduction. may be more suitable for urban settings. Policy interventions The host communities tend to face high levels of poverty, should aim not only to transition the unemployed into self- with these communities also in need of assistance. For employment but also to scale ventures and generate more example, education, access to water and electricity lag non- jobs within the labor market. arid counties. Food insecurity is also high. While allowing greater mobility towards areas with greater economic Targeted policies for women, such as expanding opportunities (as outlined in 6.1) can help refugees move to childcare and early childhood development services, where opportunities are and alleviate pressure on hosting can relieve family responsibilities and enable women areas, the lagging levels of welfare in refugee hosting to participate in the labor market while nurturing the counties also suggest scope for place-based policies. Thus, skills of future generations. Many refugee women are improving welfare of refugees and hosts should be part not looking for work due to childcare, household, and other of a broader strategy to close Kenya’s spatial disparities. family responsibilities. Among those actively seeking work, Implementing place-based development approaches 63 percent are pursuing self-employment opportunities, alongside regional development policies can benefit both which offer flexibility to balance work and childcare refugees and host communities. These interventions target responsibilities from home. Despite a demand for daycare specific geographic areas, addressing local challenges services, few are available, leading women to rely on while leveraging local assets and community participation. informal community arrangements, which can be difficult for Investments in refugee-hosting locations should benefit both refugees who are not well-connected. Both government and refugees and hosts, fostering social cohesion by directing private sectors have an opportunity to enhance access to development resources to these areas. Evidence suggests affordable childcare by increasing the availability of daycare that allocating resources to benefit nearby communities can centers, workplace nurseries, or home-based childminders. mitigate potential backlash against refugees and enhance This can ease the burden of combining work with childcare social cohesion (Zhou et al. 2023). The Kenya Development responsibilities and enhance women’s participation in the Response to Displacement Impacts Project (KDRDIP) labor market. exemplifies a place-based initiative that supports both Integrative initiatives offer promising pathways for refugees and host communities. It employs community- increased youth engagement in the labor market by driven approaches by involving local stakeholders in providing refugees with essential skills, resilience, planning and decision-making, ensuring interventions meet capital, and networking opportunities. This includes specific needs (World Bank 2017). various strategies that, after communication campaigns Strengthening household socioeconomic and an intake process, provide youth with labor market resilience to shocks information and counseling to help them choose between different types of support, in particular between Strengthening household resilience to shocks is a entrepreneurship, for which they would receive financing priority for both hosts and refugees in lagging areas. support, and wage employment, for which they would Shock-responsive social protection remains an important receive skills training and job search support (World Bank strategy to provide immediate relief during crises and 2023d). prevent vulnerable households from falling deeper into poverty. The data shows that a large share of vulnerable 6.2 Medium-term: Easing the transition with hosts miss out on assistance, and while nearly all camp- place-based development based refugees receive assistance, welfare levels are still poor. Location-specific social protection strategies, Addressing spatial inequalities such as those seen in the Hunger Safety Net Program Refugee camps and settlements in Kenya, where most (HSNP), can enable the most vulnerable refugees and refugees live, are in under-developed arid and semi- hosts to better cope with shocks and thrive, fostering social cohesion and integration. Other strategies include Building Evidence to Enhance the Welfare of Refugees and Host Communities 75 livelihood diversification, supporting the development 6.4 Continuing the implementation of a of multiple income streams through vocational training, progressive policy and hosting framework for entrepreneurship programs, and access to microfinance, refugees as well as facilitating access to banking services, savings accounts, and microloans to build financial resilience and While the country has made significant strides cope with economic shocks. To the extent that assistance is towards a policy framework oriented towards necessary while households progress towards self-reliance refugee self-reliance, addressing bottlenecks in and resilience, interventions can focus on improving the implementation is necessary. Hosting countries and allocation of aid, especially food aid, including targeting, the the international community have a shared responsibility amount/quantity and quality. in enabling and empowering refugees to achieve self- reliance and resilience. Without the freedom to earn a 6.3 Sustaining self-reliance and resilience in livelihood or relocate to areas with better opportunities, the long run critical investments in basic services and in a conducive policy framework, the goal of fostering self-reliance Education as a cornerstone and resilience will remain unattainable. In this regard, Kenya’s intended transformation of camps into integrated Addressing barriers to education access is essential settlements through the Shirika plan holds potential for for improving and sustaining the prospects of both improved socioeconomic inclusion of refugees while also refugee children and host communities. Over 80 percent improving the welfare of hosting communities. To ensure of children of primary school age in Kakuma/Kalobeyei effectiveness, national and sub-national laws and policies and urban areas are enrolled in school, but few children should not only be harmonized but also consistently applied attend school among the Turkana host community and in across spheres of government. This approach will foster a Dadaab due to cost barriers. Secondary school enrollment cohesive legal environment. In this regard, establishing a is even lower. Immediate integration of refugee children robust monitoring and evaluation framework is crucial. It into educational programs upon arrival in Kenya is vital to will enable the identification and documentation of lessons prevent the loss of valuable education years. Key strategies learned, thereby supporting continuous improvement. For to enhance children's educational attainment include instance, good subnational practices offer opportunities for tackling cost and language barriers, offering accelerated scaling up to national level. One example is the municipality and adult education for older learners, and implementing plan in Turkana County, which aims to provide integrated targeted interventions for unenrolled children with no prior services and infrastructure to both refugees and host schooling experience. Additionally, additional resources communities. Improved cooperation and coordination are needed in national education systems to smoothly among humanitarian agencies, development partners and integrate newcomers and enhance the quantity and quality line ministries, are essential, including ensuring predictable of schools in affected areas. support and resource streams are channeled in municipality Incorporating psychosocial wellbeing in health budgets for integrated service provision. services and programming More, better, and timely availability of comparable Healthcare services tailored to refugees’ unique data on refugees and host communities is key to needs are crucial. As highlighted in Chapter 5, refugees, monitor and assess progress. Continued efforts by the particularly those in camp areas, often experience higher government, humanitarian agencies, and development rates of depression symptoms compared to hosts and partners to improve the coverage, accuracy, reliability, their urban counterparts. Mental health support, ensuring quality, and comparability of data will provide the necessary language, cultural sensitivity, and gender responsiveness analytical foundation for informed policy decisions. among healthcare providers, and establishing efficient coordination and referral mechanisms could help meet refugees’ healthcare needs and support their overall wellbeing and integration into host communities. 76 Building Evidence to Enhance the Welfare of Refugees and Host Communities References Abuya, E. O. (2007). Past reflections, future insights: African asylum law and policy in historical perspective, International Journal of Refugee Law, 19(1), 52. Alix-Garcia, J., Walker, S., Bartlett, A., Onder, H., & Sanghi, A. (2018). Do refugee camps help or hurt hosts? The case of Kakuma, Kenya. Journal of Development Economics, 130, 66-83. Atamanov, A., J. Hoogeveen, & Reese B. 2024. “The Costs Come before the Benefits.” Policy Research Working Paper 10679, World Bank, Washington, DC. Barron, P., Cord, L., Cuesta, J., Espinoza, S., & Woolcock, M. (2023). Social sustainability in development: Meeting the challenges of the 21st century. World Bank Publications. Betts, A., Geervliet, R., MacPherson, C., Omata, N., Rodgers, C., Sterck, O., & Stierna, M. (2018). Self- Reliance in Kalobeyei? Socio-Economic Outcomes for Refugees in North-West Kenya. Oxford, UK: Refugees Studies Centre, ODID, University of Oxford. Betts, A., Omata, N., & Sterck, O. (2018). Refugee Economies in Kenya. Oxford, UK: Refugee Studies Centre, University of Oxford. https://www.rsc.ox.ac.uk/publications/refugee-economies-in-kenya Betts, A., Stierna, M. F., Omata, N., & Sterck, O. (2022). Social Cohesion and Refugee-Host Interactions Evidence from East Africa. World Bank Policy Research Working Paper 9917. https://documents1. worldbank.org/curated/en/598681643291911822/pdf/Social-Cohesion-and-Refugee-Host- Interactions-Evidence-from-East-Africa.pdf Coulibaly, M., Hoogeveen, J., Jourdan, E., & Savadogo, A. (2024). Responsibility Sharing and the Economic Participation of Refugees in Chad (No. 10727). The World Bank. De Berry, J. P., & Roberts, A. (2018). Social Cohesion and Forced Displacement: A desk review to inform programming and project design (No. 128640, pp. 1-41). The World Bank. http://documents. worldbank.org/curated/en/125521531981681035/Social-cohesion-and-forced-displacement- adesk-review-to-inform-programming-and-project-design Dick, E., & Rudolf, M. (2019). From global refugee norms to local realities: Implementing the Global Compact on Refugees in Kenya (Briefing Paper No. 19/2019) Easton-Calabria, E., & Omata, N. (2018). Panacea for the refugee crisis? Rethinking the promotion of ‘self- reliance’ for refugees. Third World Quarterly. Advance online publication. https://doi.org/10.1080/0 1436597.2018.1458301 Education Endowment Foundation. (2019). Improving social and emotional learning in primary schools: Guidance Report 2019. El-Bialy, N., Aranda, E.F., Nicklisch, A., Saleh, L., & Voigt, S. (2023). No man is an island: trust, trustworthiness, and social networks among refugees in Germany. Journal of Population Economics, 36(4), 2429– 2455. https://doi.org/10.1007/s00148-023-00969-7 Fajth, V., Bilgili, Ö., Loschmann, C., & Siegel, M. (2019). How do refugees affect social life in host communities? The case of Congolese refugees in Rwanda. Comparative Migration Studies, 7(1), 1–21. https://doi. org/10.1186/s40878-019-0139-1 Gaynor, T. (2020, October 10). Q&A: Before the pandemic, refugee mental health was severely overlooked. Now it’s a full-blown crisis. UNHCR. https://www.unhcr.org/news/stories/qa-pandemic-refugee- mental-health-was-severely-overlooked-now-its-full-blown-crisis Building Evidence to Enhance the Welfare of Refugees and Host Communities 77 Giacomo C., J. Drzewiecki, & Jesuit Refugee Services (2024). Breaking Down Barriers: Increasing Girls’ Access to and Completion of Secondary Education in Kakuma Refugee Camp. https://jrs.net/wp- content/uploads/2024/03/FINAL-4-April-2024-Kakuma-Gender-Analysis-Report_to-print.pdf Hewstone, M. (2015). Consequences of diversity for social cohesion and prejudice: The missing dimension of intergroup contact. Journal of Social Issues, 71(2), 417-438. https://doi.org/10.1111/josi.12120 Hure, M., & Taylor, J. (2023). The critical role of refugee teachers in the COVID-19 education response: Supporting their continued professional contributions and leadership in displacement and durable solutions. Education and Conflict Review, 4, 37-44. https://discovery.ucl.ac.uk/id/eprint/10175039 Intergovernmental Authority on Development. (2017). Nairobi declaration on durable solutions for Somali refugees and reintegration of returnees in Somalia. Nairobi, Kenya: Intergovernmental Authority on Development. https://data.unhcr.org/en/documents/details/58248 Intergovernmental Authority on Development. (2019). Kampala Declaration on jobs, livelihoods and self- reliance for refugees, returnees and host communities in the IGAD region. Kampala, Uganda: Intergovernmental Authority on Development. https://igad.int/kampala-declaration-on-jobs- livelihoods-and-self-reliance/ International Labour Organization. (2022). “Review of national policy, legislative and regulatory frameworks and practice in Kenya: A baseline study on the right to work and rights at work for refugees”. https://www.ilo.org/publications/review-national-policy-legislative-and-regulatory-frameworks-and- practice-2 Kenya National Examinations Council (2020). Monitoring Learner Achievement at Grade 2 Level of Primary School Education in Kenya: Early Grade Mathematics Assessment (EGMA) Endline Study. https:// www.knec.ac.ke/wp-content/uploads/2020/09/FINAL-EGMA-Endline-Report-17.08.2020-Copy.pdf Kenya National Qualifications Authority. (n.d). Recognizing Prior Learning. https://www.knqa.go.ke/index. php/recognizing-prior-learning/# Klabbers, R. E., Ashaba, S., Stern, J., Faustin, Z., Tsai, A. C., Kasozi, J., Kambugu, A., Ventevogel, P., Bassett, I. V., & O’Laughlin, K. N. (2022). Mental disorders and lack of social support among refugees and Ugandan nationals screening for HIV at health centers in Nakivale refugee settlement in southwestern Uganda. Journal of Global Health Reports, 6, e2022053. https://doi. org/10.29392/001c.39600 Kroenke, K., Spitzer, R. L., & Williams, J. B. W. (2001). The PHQ-9: Validity of a brief depression severity measure. Journal of General Internal Medicine, 16(9), 606–613. https://doi.org/10.1046 /j.1525-1497.2001.016009606 Lebow, J. (2024). Immigration and occupational downgrading in Colombia. Journal of Development Economics, 166, 103164. https://doi.org/10.1016/j.jdeveco.2023.103164 Leeson, K., Slaughter, A., Buscher, D. (2022). Better Lives Now: A Conceptual Framework for Understanding and Measuring Refugee Self-Reliance. Refugee Self-Reliance Initiative. Leeson, K., Bhandari, P. B., Myers, A., Buscher, D. (2020). Measuring the Self-Reliance of Refugees. Journal of Refugee Studies, 33(1), 86-106. https://doi.org/10.1093/jrs/fez076 Malombe, D., Mavenjina, M., Medi, M., & Mbataru, S. (2016). Kenya’s regional and international human rights obligations (pp. 11-14). Kenya Human Rights Commission. Mwaura, P., & Marfo, K. (2011). Bridging culture, research, and practice in early childhood development: The Madrasa Resource Centers in East Africa. Child Development Perspectives, 5(2), 134-139. https:// 78 Building Evidence to Enhance the Welfare of Refugees and Host Communities ecommons.aku.edu/eastafrica_ihd/9 Nyanduga, B. T. M. (2004). Refugee protection under the 1969 OAU Convention Governing the Specific Aspects of Refugee Problems in Africa. German Yearbook of International Law, 47, 86. O’Callaghan, S., Manji, F., Holloway, K., & Lowe, C. (2019). The Comprehensive Refugee Response Framework: Progress in Kenya (HPG Working Paper). London: Overseas Development Institute. https://www.odi.org/sites/odi.org.uk/files/resource-documents/12940.pdf Omata, N. (2021). Refugee livelihoods: a comparative analysis of Nairobi and Kakuma Camp in Kenya. Disasters, 45(4), 865-886. Pozuelo, J. R., Bradenbrink, R., Stierna, M. F., & Sterck, O. (2023). Depression, violence and socioeconomic outcomes among refugees in East Africa: Evidence from a multicountry representative survey. BMJ Mental Health, 26(1), e300773. https://doi.org/10.1136/bmjment-2023-300773 Sanghi, A., Onder, H., & Vemuru, V. (2016). " Yes" in My Backyard? The Economics of Refugees and Their Social Dynamics in Kakuma, Kenya. World Bank. Spitzer, R. L., Kroenke, K., Williams, J. B. W., & Löwe, B. (2006). A Brief Measure for Assessing Generalized Anxiety Disorder: The GAD-7. Archives of Internal Medicine, 166(10), 1092. https://doi.org/10.1001/ archinte.166.10.1092 Sarzin, Z., & Nsababera, O. Forthcoming. “Forced Displacement: A Stocktaking of Evidence.” Background paper prepared for 2024 Africa companion report of the World Development Report 2023: Migrants, Refugees, and Societies. World Bank, Washington, DC. Schwarzer, R., & Jerusalem, M. (1995). Generalized Self-Efficacy scale. In J. Weinman, S. Wright, & M. Johnston, Measures in health psychology: A user’s portfolio. Causal and control beliefs (pp. 35-37). Windsor, UK: NFER-NELSON. Refugees Act. 2021. http://kenyalaw.org:8181/exist/rest/db/kenyalex/Kenya/Legislation/English/ Acts%20and%20Regulations/R/Refugees%20Act%20-%20No.%2010%20of%202021/docs/ RefugeesAct10of2021.pdf United Nations High Commissioner for Refugees. (2005). Handbook for Self-reliance. https://www.unhcr.org/ us/media/handbook-self-reliance-complete-publication United Nations High Commissioner for Refugees. (2018). KISEDP: Kalobeyei Integrated Socio-Economic Development Plan in Turkana West. Nairobi, Kenya: https://www.unhcr.org/ke/kisedp-2 United Nations High Commissioner for Refugees. (2023a). Inside the world’s five largest refugee camps. https://www.unrefugees.org/news/inside-the-worlds-five-largest-refugee-camps/ United Nations High Commissioner for Refugees. (2023b). Outcomes of the Global Refugee Forum 2023. Retrieved from https://www.unhcr.org/media/outcomes-global-refugee-forum-2023 United Nations High Commissioner for Refugees Kenya. (n.d.). Urban areas. https://www.unhcr.org/ke/ urban-areas United Nations High Commissioner for Refugees and World Bank. (2019). Understanding the Socioeconomic Conditions of Refugees in Kenya: Volume A-Kalobeyei Settlement: Results from the 2018 Kalobeyei Socioeconomic Survey. Washington, DC: The World Bank Group. United Nations High Commissioner for Refugees and World Bank. (2021). “Understanding the Socioeconomic Conditions of Refugees in Kenya. Volume B: Kakuma Camp. Results from the 2019 Kakuma Socioeconomic Survey.” Building Evidence to Enhance the Welfare of Refugees and Host Communities 79 United Nations High Commissioner for Refugees and World Bank. (2020). “Understanding the Socioeconomic Conditions of the Stateless Shona Community in Kenya. Results from the 2019 Kakuma Socioeconomic Survey.” United Nations High Commissioner for Refugees and World Bank. (2021). “Understanding the Socioeconomic Conditions of Refugees in Kenya. Volume B: Kakuma Camp. Results from the 2019 Kakuma Socioeconomic Survey.” United Nations High Commissioner for Refugees, and World Bank. (2022). “Understanding the Socio- Economic Conditions of Urban Refugees in Kenya. Volume C: Urban Refugees.” United Nations High Commissioner for Refugees. 2024a. Global Trends. https://www.unhcr.org/us/global- trends. United Nations High Commissioner for Refugees. (2024b). Operational Data Portal. https://data.unhcr.org/ en/documents/details/94275 United Nations High Commissioner for Refugees. (n.d.a). Education. Retrieved June 7, 2024, from https:// help.unhcr.org/kenya/dadaab/education/ United Nations High Commissioner for Refugees. (n.d.b). Education – Kakuma. Retrieved June 7, 2024, from https://help.unhcr.org/kenya/kakuma/education/ United States Agency for International Development (USAID) (2017). Tusome External Evaluation – Midline Report. https://shared.rti.org/content/tusome-external-evaluation-midline-report Vuni, F., & Iragi, B. (2023). Refugees’ access to work permits and business licences in Kenya. Refugee-Led Research Series Report no.7. World Bank. 2017. “Development response to displacement impacts project (DRDIP) in the Horn of Africa.” Project Appraisal Document PAD2163, World Bank, Washington, DC. World Bank (2022). COVID-19 Rapid Response Phone Survey with Households 2020-2022, Panel. World Bank. https://doi.org/10.48529/tch6-jx12 World Bank. 2023a. World Development Report 2023: Migrants, Refugees and Societies. Washington, DC: World Bank. World Bank. (2023b). Kenya Poverty and Equity Assessment 2023: From Poverty to Prosperity: Making Growth More Inclusive. Washington, DC: World Bank. World Bank. (2023c). Socioeconomic conditions of refugee and host households in Nairobi. Washington, DC: World Bank. World Bank. (2023d). Interventions to Foster Youth Employment in Refugee and Host Communities in Kenya: A Feasibility Study. Washington, DC: World Bank. World Bank. (2024). Kenya Secondary Education Equity and Quality Improvement Program: Program Appraisal Document. World Bank and UNHCR (World Bank and United Nations High Commissioner for Refugees). Forthcoming. Economic Participation and the Global Cost of International Assistance in Support of Refugee Subsistence Needs. World Bank, Washington, DC. Zhou, Y. Y., Grossman, G., & Ge, S. (2023). Inclusive refugee-hosting can improve local development and prevent public backlash. World Development, 166, 106203. 80 Building Evidence to Enhance the Welfare of Refugees and Host Communities Annexes Annexes Annex 1: Survey Sampling Refugees were sampled from two Annex frames, 1: Survey the UNHCR Sampling registra?on database Socioeconomic Surveys (SES) conducted Refugees in Kakuma, Kalobeyeifrom were sampled and Nairobi. two frames,The U Annexes database proGres holds informa?onSocioeconomic on phone numbers Surveys and conducted inloca the approximate (SES) Kak families registered with UNHCR Kenya and therefore database proGres represents a perfect on holds informa?on sampling phone frame consists of the samples of the Kalobeyei SES (2018), families registered withKakuma UNHCR Refugee Kenya Camp SES and ther (2020-21). SES in Nairobi data and is Ita holds subset rich socioeconomic frame of the consists first, baseline of as depicted the data and insamples Figure of 62. the 52 is a Kalobeyei subset of th S Annex 1: Survey Sampling in Figure 62.52 In thefirst In the stage,a firststage, asubsample" subsample of of proGres Λ(PG)!(2020-21). SES in Nairobi proGres Itfamilies holds rich families is drawn from socioecono Refugees were sampled from two frames, the frame UNHCRusing implicit is drawn stra?fica?on from the by in Figure sub-county entire proGres62.and 52 In na?onality. framethe using first stage,53 Inathe implicit subsample Λ(P second stag subsample registration database proGres and the Socioeconomic Λ(SES) stratification ! is drawn from the by sub-county SES frame frame and using for each implicit In nationality. 53 of the strata stra?fica?on with the secondby sub-county the exis?n Surveys (SES) conducted in Kakuma, Kalobeyei who haveandarrived stage,in Kenya an independent subsample before thesubsample SES, represent Λ(SES) the isisdrawn ! overlap drawn from group from the SES which frame has a po being Nairobi. The UNHCR registration database proGres holds selected into the SESthe sample frame forthrough each who of both the have the strata SES arrived withframe in the and Kenya existing through before SES. theproGres SES, dire repres of households, information on phone numbers and the approximate location weights are adjusted using being an overlapping selected Households who have arrived in Kenya before the SES, into frames the sample correc?on. through both the of all refugee families registered with UNHCR Kenya and of households, weights represent the overlap group which has a positive probability are adjusted using an ove Figure 62: A dual frame design in which frame of being selected into the sample through both the SES B (SES) is a subset of frame A therefore represents a perfect sampling frame. The second Figure frame and through proGres directly. For this subset of 62: A dual frame design in which frame consists of the samples of the Kalobeyei SES (2018), Kakuma Refugee Camp SES (2019) and Urban SES in households, weights are adjusted using an overlapping Nairobi (2020-21). It holds rich socioeconomic baseline frames correction. Figure 60: A dual frame design in which frame B (SES) is a subset of frame A (proGres) Hosts were sampled using a two-step cluster design. The sampling frame of host hou communi?es living near refugees who are poten?ally Hosts were sampled usingby impacted a their presence. two-step The cluster de camp refugees in Turkana and Dadaab are defined communi?es as those living living within near refugees a 15 who are km r poten? borders, while host communi?es of camp refugees in urban refugees are defined Turkana neighborhoods as Dadaab and are defi percent of proGres families living in each city of the stratum. Sampling follows borders, while host communi?es of urban a two-ste refu whereby a set of EAs and replacement EAs is drawn with probability propor?onal percent of proGres families living in each city of to th whereby a set of EAs and replacement EAs is dr 52 Hosts were sampled using a two-step cluster design.Observations The in the of urban were originally SES refugees drawn from are defined as proGres, neighborhoodsand hence households who wer hosting before the sampling frame of host households consists of communities SES had both a chance to be sampled 52 for the Observations SES in or the for SES at least 10 percent of proGres families living in each city K-LSRH were directly. originally of We abstrac drawn from that some households have left Kenya or moved before within the SES Kenya had since both a the timeto chance the ofbe SES, as we sampled forkt living near refugees who are potentially impacted by the stratum. Sampling follows a two-step clustered design, registration data that this group is very smallthat (~1%). some households have left Kenya or moved withi their presence. The host community of camp refugees 53 in Note: proGres whereby familiesa are set not of EAs andthe always replacement samedata EAs is drawn as households with in standard su defined registration that this group is very small (~1%). Turkana and Dadaab are defined as those living within a household probability is defined proportional as a “person or group toof 53 size people Note: (PPS)living proGres from in the the universe same families are not of compoundalways(fenced or un the same to the same head 15 km radius of the camp borders, while host communities and sharing eligible EAs.54a common household source of food/share is defined as cooking arrangement.” a “person Meanwl or group of people refer to the group of people in which refugees are registered to the same to UNHCR. head and sharing A household a common source ofcanfo proGres families, and the selection probability to the to refer needs be adjusted group of people accordingly. in which refugees are proGres families, and the selection probability needs 76 52 Observations in the SES were originally drawn from proGres, and hence households who were registered in Kenya before the SES had both a chance to be sampled for the SES or for K-LSRH directly. We abstract from the possibility that some households have left Kenya or moved within Kenya since the time of the SES, as we know from the UNHCR registration data that this group is very small (~1%). 53 Note: proGres families are not always the same as households defined in standard surveys. For K-LSRH, a household is defined as a “person or group of people living in the same compound (fenced or unfenced); answerable to the same head and sharing a common source of food/share cooking arrangement.” Meanwhile, proGres families refer to the group of people in which refugees are registered to UNHCR. A household can comprise of multiple proGres families, and the selection probability needs to be adjusted accordingly. 54 For camp hosts, eligible EAs are located at least partly within the 15 km radius around the camp borders. For urban settings, eligible EAs cover the neighborhoods hosting at least 10 percent of proGres families. These are Kayole, Eastleigh North, and Kasarani in Nairobi, Old Town and Majengo in Mombasa, and Lanet, Shabaab, and Langalanga in Nakuru. Building Evidence to Enhance the Welfare of Refugees and Host Communities 81 Subsequently, a listing exercise is carried out to eliminate sampling. This results in a sample of 2,000 households ineligible households.55 In the second stage, for each for the host community of camp refugees and of 1,500 selected EA, 10 households and a list of 5 replacements households for the host community of urban refugees for the case of non-response is drawn using simple random (Table 11). 56 Table 11: Host sample breakdown per stratum Strata Number of EAs Number of HHs per EA Total Turkana 75 10 750 Dadaab 125 10 1,250 Greater Nairobi 75 10 750 Mombasa and Nakuru 75 10 750 Total 3,500 Source: Authors’ calculations based on K-LSRH 2023. Modules on individual level characteristics, women respondent also completed the Woman Empowerment empowerment and child education were administered Module; otherwise, a woman aged 15 years and above to targeted respondents. Most questionnaire modules, was selected among the women in the household based in particular those relating to household characteristics on simple random sampling. Finally, a module on child or household member level data, were elicited from education and aspirations was administered to a cohort of a knowledgeable household member, the Household upper primary school children and their main caregivers. Respondent (HR). A separate set of questions on individual One Child Respondent (CR) was randomly selected level characteristics such as labor force participation or among 5th-grade students currently attending school of personal opinions, were administered to a randomly selected any age up to and including 17 years (to also capture over- household member aged 18 or above, the Representative age students) in selected households.57 The child’s main Respondent (RR). In addition, the survey included a module caregiver (CG) was also interviewed to capture parental on women’s empowerment, administered by trained female aspirations and main barriers to education. The Caregiver enumerators to randomly selected female respondents refers to the household member responsible for making the aged 15 and above. If the RR was a woman, the same educational decisions for the child.58 55 This includes households not living within the boundary of selected EAs and households not providing consent to the listing interview. Households where one member was registered with UNHCR as a refugee or asylum seeker were not interviewed, to rule out overlap with the refugee sampling frame. 56 Further details on replacement methodology available upon request. 58 This should consider students moving into Grade 5 in July 2021, who should then be in Grade 6 in April 2022 and Grade 7 in January 2023. 59 Unlike other modules, the Education Module was only implemented in the Kakuma, Kalobeyei, and Dadaab refugee strata and the Turkana and Dadaab host community strata. 82 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 12: Overview of survey respondents, questionnaire modules and sampling approaches Respondent Modules Sampling Household Respondent (HR) Roster data on household members One knowledgeable respondent in the household. Dwelling characteristics Consumption Food insecurity & assistance Household member labor data Transfers & remittances Household shocks Representative Respondent (RR) Extensive labor module One randomly selected adult (age 18+) Savings & credit Migration Displacement & return Social cohesion Health & wellbeing Marriage & fertility Ethnicity & religion Education perceptions for boys and girls Weather variability perceptions Women Empowerment Respondent (WER) Migration and marriage mapping One randomly selected woman (age 15+). If the RR is a woman, she is auto- Women’s empowerment index matically selected as the WER. Child Respondent (CR) Schooling attitude & beliefs One 5th-grade student currently attend- ing school of any age up to and including Learner academic development 17 years Beliefs of own academic ability Self-efficacy Health & wellbeing Child Caregiver (CG) Caregiver module The main caregiver of the Child Respon- dent is selected as the CG. povertycut-off poverty cut-off , ,indica?ng indica?ngthe theminimum minimumnumber numberof ofweighted weightedindicators indicatorsin inwhichwhichthe thehouseholdhouseholdhas hasto to be deprivedin bedeprived orderto inorder tobe consideredmul?dimensionally beconsidered mul?dimensionallypoor. poor.A Apersonpersonmay, may,for instance,be forinstance, temporarily betemporarily deprivedin deprived nutri?onbecause innutri?on becausethey theyare arefas?ng, fas?ng,in inwhichwhichcase casethey theyshould shouldnot notbe beconsideredconsideredpoor. poor.But Butif if theyare they simultaneouslydeprived aresimultaneously deprivedin nutri?onand innutri?on andhealth healthand andeduca?on, educa?on,for instance,it forinstance, itis unlikelyto isunlikely tobe be voluntary. In voluntary. In thisthis paper, paper, we we have have used used two two different different mul?dimensional mul?dimensional cut-offs, cut-offs, = =0.33 0.33 for for moderatemoderate mul?dimensionalpoverty, mul?dimensional poverty,and and Building = =0.5 Evidence 0.5 Annex forforsevere severe 2.mul?dimensional Mul,dimensional to Enhance mul?dimensional the Welfare of Refugees poverty, poverty, Poverty and meaning meaning Host Communities thataahousehold that household 83 willbe will consideredseverely beconsidered severelypoor poorif iftheytheyare aresimultaneously The simultaneously mul?dimensional deprived deprived index used ininat atin least least this ½ ½ of weighted ofweighted document, uses indicators. indicators. the Alkire Foster (2007) "" TheAlkire-Foster The Alkire-Fosterheadcount headcountpoverty povertyra?o, ra?o, indices, , ,forfor where aasample sampleeach of of indicator individuals, individuals,of depriva?on canbe can beis wriJen wriJen given a as: as: weight, reflec?ng its impor and an indicator-specific depriva?on threshold, defining the level ajer wh considered 1 1 $ $ deprived in that indicator. For instance, for nutri?on indicators, the d "" Annex 2. Multidimensional Poverty set = to = 5 +out + 7, ## meaning - - ## ≥ ≥ that 00 a household needs to have suffered from, say, food s of the past #%&7 days to be considered deprived in that indicator. #%& The multidimensional Where Where # # is index isthethenumber used number in this document, ofofweighted weighted uses depriva?ons depriva?ons the experienced experiencedexperienced byby by household householdand household and and(∙) (∙) is is an is anan indicator indicator indicator Alkire Foster (2007) func?on, func?on, method taking taking for the the value constructing value1 1if household ifhousehold indices, where is In is deprived addi?ondeprivedfunction, toin inat at indicator-specific taking least least the weighted weighted value indicators, 1 indicators, depriva?on if household and and thresholds, i0 0 is otherwise. otherwise. the Alkire deprived in Foster index req The The index index contains contains each indicator of deprivation is given a weight, reflecting its 14 14 indicators, indicators, which which are are poverty grouped grouped cut-off into into 6 6 ,different different indica?ng at least weighted indicators, and 0 otherwise. dimensions. dimensions. the minimum ItIt uses usesnumber a a nested nested of weigh?ng weigh?ng weighted indicators in wh system, system, where where importance in the overall index, and an indicator-specific each each dimension dimension weighs weighs 1/6 1/6 of of the the overall overall index, index, be deprived in order to be considered mul?dimensionally poor. A person may, forandand each each indicator indicator hashas equal equal weight weight within within thedimension. the dimension. Dimensions Dimensions with with more more indicators indicators The will will index therefore therefore contains give giveless 14less indicators, weight weight which are grouped into Annexdeprivation 2. Mul,dimensional threshold, Poverty defining the level after which deprived a in nutri?on because they areto toeach fas?ng, eachspecific specific in which indicator. indicator. case they should not b 6 different dimensions. It uses a nested weighting system, household willTable The mul?dimensional Table index be 1313 used considered provides provides in this deprived anan overview document, overviewin that of uses ofthe the indicator. the index index Alkirethey For composi?on composi?on are simultaneously Foster (2007) andand method defini?ons defini?ons deprived ofofdepriva?ons. for construc?ng depriva?ons. in nutri?on and health and educa?on, for ins voluntary. where In each this paper, dimension we have weighs used two of 1/6 the overall different index, mul?dimensional cut-offs where each indices,instance, indicator of depriva?on for nutrition indicators, the deprivation thresholds is given a weight, reflec?ng its importance in the overall index, The The educa?on educa?on indicator indicator isis only only available available forfor households mul?dimensional and households each with with indicator poverty, children children has and agedequal aged = 7-15, 7-15, weight 0.5 for which which within severe means means the dimension. that that mul?dimensional other other poverty, m and an are indicator-specific depriva?on set to 5 out 7, meaning that a household needs to have threshold, defining the level ajer which a household will be nnex 2.Mul,dimensional Mul,dimensional families families are are Poverty measured measured on on 13 13 indicators indicators instead instead of of Dimensions 14. 14. TheThe alterna?ve alterna?ve with more isis indicatorstoto count count will households households therefore give without without less nex 2. considered deprived suffered from, say, food in that Poverty indicator. For instance, shortages in at least 5 out of the for nutri?on will be indicators, considered the severely depriva?on poor thresholds if they are are simultaneously deprived in at least sional set emul?dimensional Poverty mul?dimensional out 7, to 5 past Annex 7 meaning index days 2. index to children children that Mul,dimensional beused used ain considered in asasnot household this notdeprived document, thisdocument, deprived deprived needs in in into educa?on. educa?on. Povertyuses thathave uses suffered the indicator. The theAlkire The The Alkire same same from, weight Alkire-Foster Foster Foster issue, issue, say, food but but to shortages (2007) (2007) lesssevere, less each headcount severe, method specific method in affects affects poverty at for indicator. least employment employment ra?o, forconstruc?ng5 out construc?ng Table " , for 13(18-64). a(18-64). provides sample of an individuals, ca ex used of in this the past document, 7 days to be uses considered the Alkireis deprived Foster in that (2007) indicator. method overview for construc?ng of the index composition and definitions of dices, ces, where where each each The In indicator mul?dimensional addition indicator to The Theof of depriva?on depriva?on mul?dimensional mul?dimensional indicator-specific index is used deprivation given given poverty poverty in a a this weight, weight, approach approach thresholds, document, reflec?ng reflec?ng was waspreferred the uses preferred its its deprivations. the importance importance toto Alkire thethemore more Foster in in the conven?onalthe conven?onal(2007) overall overall method index, monetary monetary index, $for approach, approach, construc?ng as as ator d an an of depriva?on is given indicator-specific indicator-specific a provision weight, depriva?on depriva?on the the provision reflec?ng threshold, of threshold, of in-kind in-kind its assistanceimportance defining defining assistance makes makes thethe in it level it the level difficult difficult overall ajer ajer to compare tocompare which index, which a household a household consump?on consump?on "will levels levels will =between 1 be betweenbe hosts hosts and and In addi?on indices, Alkire where to indicator-specific Foster index each indicator depriva?on requires a of depriva?on thresholds, multidimensional the is Alkire poverty given Foster a weight,index requires reflec?ng a its mul?dimensional importance in + the overall # - # ≥ index,0 depriva?on nsidered sidered deprived deprived threshold,in in that that defining indicator. refugees, refugees, indicator. and and the ForFor across across Annexlevel instance, instance, different differentajer 2. Mul,dimensional forfor camps, which camps, nutri?on nutri?on a depending depending household indicators, indicators, The on Poverty on where education where the will the be depriva?on assistance assistance depriva?on indicator isis is provided provided only thresholds thresholds available in in cash/in-kind. cash/in-kind. areare for households While While itit povertyand cut-off cut-off an indicator-specific , indica?ng the minimum indicating the minimum depriva?on number of weighted number threshold, of weighted indicators defining in which the aged level the householdajer has to a household will be which #%& tat to toindicator. 55out out be 7, For meaning 7,meaning deprived instance, in order that thatin to isis a for apossible be nutri?on possible household household considered toAnnex to impute impute The2. indicators, needs needs Mul,dimensional mul?dimensional mul?dimensionallythe theto to value value have have the ofof depriva?on in-kind in-kind suffered suffered index poor.Where A assistance, Poverty used assistance, from, from, person thresholds with inis #say,say, may, this children itit the document, requires requires food food for number are shortages instance, shortagesdetailed detailed of uses 7-15, weighted be the knowledge in which knowledge in temporarily at Alkire at means depriva?ons least least Foster of 5 of5out that what what out (2007) other services services experienced method familiesare are by for construc?n household a considered indicators deprived which being being the household provided provided inThe that in indices, mul?dimensional in each each indicator. has where loca?on. loca?on. to be For each Given Given instance, deprived index indicator the the used in magnitude in magnitude for of depriva?on this are nutri?on document, measured of of the the is indicators, on phenomenon, phenomenon,givenuses 13 atheweight, indicators Alkire the itit wouldreflec?ng would depriva?on Foster instead also also of its (2007)14. make make The thresholds importance methodthe the alternative poverty poverty forin theare overall construc?ng index at a thepast he household deprived past 77daysdays intoneeds to nutri?on bebe to considered have because considered suffered theydeprived deprived from, are fas?ng, inin say, thatthat food which in indicator. indicator. shortages case func?on, they in should at taking least not be considered 5 out value 1 if household poor. But is if deprived in at least weighted indi set order to to5be out 7, considered meaning es?mates es?mates multidimensionally indices, very very that and a household wherean poor. indicator-specific each needs indicator A person of to have depriva?on depriva?on suffered threshold, is given from, weight, aimputa?on. say, defining food reflec?ng the shortages level its ajer importance in at6deprived which least in the 5 aaout a household overall index,will bI onsidered they deprived are may,simultaneously for ininstance, that indicator. deprived be sensi?ve in sensi?ve temporarily nutri?on toto and deprived modelling modelling health in and assump?ons assump?ons nutrition The educa?on, index is to used contains forused count in instance, in 14 thethe households indicators,imputa?on. it is unlikely without which On On to arethe the children be other other grouped as hand, hand, not into using using different dimensions. voluntary. of the In this past paper, 7 days mul?dimensional mul?dimensional to be and considered considered an approach indicator-specific approach deprived deprived totopoverty poverty in in that depriva?on means means that system, indicator. indicator. that that inwhere threshold, weweare education. For instance, arerestricted each restricted dimensiondefining The for same toto nutri?on the the weighs indicators issue,level indicators 1/6 indicators, but ajer of the for for less which the which which overall severe, depriva?on a household data data index, are are affects thresholds will and each indicato be ar addi?on ddi?on toto indicator-specific because indicator-specific they are we fasting, havein depriva?on depriva?on used which considered set two thresholds, case thresholds, to 5 different out deprivedthey 7, should the meaning mul?dimensional the Alkire not Alkire in that indicator. that be Foster Foster a household cut-offs, index index For instance, requires needs requires = to 0.33 for that have nutri?ona a for moderate mul?dimensional mul?dimensional suffered indicators, from, say, thewill food depriva?on shortages thresholdsin at leastare5 ou mul?dimensional poverty, available available and ininthe the K-LSRH. K-LSRH. for severe mul?dimensional the dimension. employment poverty, Dimensions meaning (18-64). with more a household indicators therefore give less weight ecific verty depriva?on cut-off erty cut-off considered , ,indica?ngthresholds, indica?ng poor. the the But the minimumif minimum = Alkire they set 0.5are to of Foster number the number 5 simultaneously out past 7, index7of of meaning days requires weighted weighted todeprived that bea a considered indicators household mul?dimensional indicators deprived needs in in which which to inhave that the the household indicator. suffered household from, say, hashas food toto shortages in at least 5 out In addi?on to indicator-specific will be considered severely poor if they are simultaneously deprived in depriva?on thresholds,Table 13 provides the at least Alkire an overview ½ of weighted Foster of index the indicators. index requires composi?on a mul?dimensional and defini?ons of depri ng the minimum deprived deprived inin in nutrition order order number toto be be and ofhealth considered considered weighted and of the indicators education, mul?dimensionally past 7 mul?dimensionally " days in for towhich instance, bepoor. considered poor. the A A it household person person The deprived may, multidimensional may, in has that forfor to indicator. instance, instance, poverty be be approach was preferred temporarily temporarily poverty cut-off The Alkire-Foster headcount , poverty indica?ng ra?o, In the addi?on minimum , for a sample towe indicator-specific number of individuals, of to weighted the depriva?on can be more indicators thresholds, wriJen as: in monetary conventional which the household approach, as has the to considered prived rived in innutri?on nutri?on mul?dimensionally is unlikely to be voluntary. because because they they poor. are are In A fas?ng, fas?ng, person this paper, in may, inwhich which forhave case instance, case used The they they educa?onbe should should temporarily indicator notnot be be is consideredonly the considered available Alkire for poor. poor. Foster households But But index requires if if with a mul?dimensiona children aged 7-15 be deprived in order to be considered poverty tocut-off mul?dimensionally , indica?ng the provision minimum poor. of number A in-kind person assistance may, weighted of Alkire for makes instance, indicators it difficult be in14. temporarily compare to The which the household ause they are two fas?ng, different in multidimensional which case In addi?on they cut-offs, should indicator-specific $ not =0.33 be considered for families depriva?on are poor. thresholds, measured But ifonthe 13 indicators Foster index instead requires of a mul?dimensional alterna?ve is has to co t ey y are simultaneously aresimultaneously deprived deprived deprived in nutri?on in in nutri?on nutri?on because poverty be and they deprived cut-off and 1 health health are in , order indica?ng fas?ng, and andto be the educa?on, educa?on, in which considered minimum consumption casefor for number instance, instance, mul?dimensionallythey levels ofin should weighted it it between is isnotunlikely poor. unlikely indicatorsbeA hosts to considered person to and be be may, in which refugees,for poor. instance, the less andBut household be if temporaril deprived in moderate multidimensional nutri?on and health and poverty, educa?on, " =and + =0.5 for for ≥ instance, mul?dimensional severe children 0it as not is unlikely deprived to be educa?on. The same issue, but severe,has to em affects luntary.In untary. In this this paper, paper, we wehave have used used twotwo different different # - mul?dimensional cut-offs, cut-offs, = =0.33 0.33 forfor moderate moderate they multidimensional are simultaneously poverty, be deprived deprived deprived meaning that in in in order a nutri?on nutri?on household # to be considered because and will health they across are mul?dimensionallyand fas?ng, different educa?on, in camps, which poor. for case depending A instance, personthey onshould may, where itfor is not unlikely instance,be considered assistance be to is be temporarilypoor. But we have used two ul?dimensional ?dimensional poverty, poverty, different and and mul?dimensional forfor severe cut-offs, #%& mul?dimensional =deprived 0.33 forpoverty, moderate meaning that aahouseholdhousehold Where voluntary. be considered is the number In this severely of weighted = =0.5 paper, 0.5 deprived poor we if severe they depriva?ons have they are in mul?dimensional simultaneously nutri?on are used simultaneously experienced two because different The they are bymul?dimensional household poverty, in mul?dimensional provided fas?ng, nutri?on in and meaning in and health which cash/in-kind. poverty (∙) is case an that cut-offs, approachthey While indicator and educa?on, should it was is possible = not 0.33 preferred for beto instance, for tomoderate considered impute the more is unlikely it poor. the But if conven?ona to b y, llbeand beconsidered = considered 0.5 # for severe severely mul?dimensional poor if they voluntary. are poverty, In this meaning paper, we that have a used household two different mul?dimensional cut-offs, for moderat simultaneously deprived inat at least ½ of weighted indicators. severely poor if they they are are simultaneously simultaneously deprived in least ½ of weighted indicators. = 0.33 func?on, mul?dimensional deprived taking in at the value least 1 if ½ poverty, of household weighted and is = 0.5in indicators. deprived for The at deprived severe Alkire- least the in provision weighted nutri?on mul?dimensional value ofof in-kind indicators, and in-kind health assistance assistance, and and poverty, 0 otherwise.educa?on, makes it meaning requires for it difficult instance, that detailed to a compare knowledgeit is unlikely household consump?on to be l ly poor eAlkire-Foster Alkire-Fosterif they are headcount headcount simultaneously poverty poverty deprived voluntary. ra?o, ra?o, mul?dimensional " " In ,, in ,forfor at this least asamplepaper, sample ½ poverty, weof weighted have refugees, of and used individuals, = and two indicators. 0.5 across for different can severe different be mul?dimensional mul?dimensional wriJen camps, depending poverty, cut-offs, as:½ of weighted indicators. providedon where meaning = 0.33 for assistance that a moderatehousehol is The index will Foster be contains headcount "considered 14 indicators, poverty severely ratio,are which will grouped poor if a for be considered they a into sample are simultaneously 6 differentof of severely individuals, poor of dimensions. what if 7979 they services deprived are Itcan uses be simultaneously are wriJen a nested inbeing at least as: provided weigh?ng deprived in each in atmeaning location. least ½ of Given weighted indicators unt poverty ra?o, where system,individuals, , eachcan for a dimension sample mul?dimensional weighs of as: 1/6poverty individuals, of the overall poverty, can index, be and wriJen is possible = 0.5 as:to for severe impute the mul?dimensional value of in-kind poverty, assistance, it requires that a household detailed knowle ,and each be written theindicator magnitude has of equal weight can the phenomenon, within it would also make the The Alkire-Foster headcount will The be Alkire-Foster considered ra?o, severely "poor headcount being forif a poverty sample they provided are ra?o, in of simultaneously each " individuals, , for loca?on. a sampledeprived Given of the in be at wriJen individuals, magnitude least ½ can of as: weighted thebe wriJen phenomenon, as: indicators. it wou the dimension. Dimensions with more indicators $ $will therefore give less poverty weight estimates " to each very specific sensitive indicator. to modelling assumptions Table 13 provides an $ overview of the index The Alkire-Foster "" 1 1 headcount composi?on and defini?ons poverty es?mates ra?o, used very of depriva?ons. sensi?ve , in the imputation. for a sample to modelling of individuals, assump?ons On the other hand, using a can be used wriJen in the as:imputa?on. On 1 = = + + # - # - # # ≥ ≥ 0 mul?dimensional0 $ approach $ to poverty means that we are restricted to the indi " = + # -# ≥ 0 " 1 multidimensional " $ 1 approach to poverty means that we are #%& #%& with =childrenavailable + in # - the ≥ K-LSRH. =0 + # -# ≥ 0 The educa?on indicator is only available for households aged # restricted 7-15, to1which the means that indicators for other which data are available in here ere isis the the number number of of #%& weighted weighted depriva?ons depriva?ons experienced experienced by by household household " = + and and ##%& -(∙) (∙)# ≥ is is an an indicator 0 indicator #families # are measured on 13 indicators instead of 14. The alterna?ve #%& the is K-LSRH. to count households without of nc?on, weighted taking depriva?ons the value experienced by Where household# is the number and (∙) of weightedis an indicator depriva?ons experienced by household and (∙) is an indicator c?on, children taking the value # 11 ifif household household numberof is is deprived deprived in in at atleast least weighted weighted indicators, indicators, andand00otherwise. otherwise. #%& as Where not deprived is the in educa?on. number The same weighted issue, but depriva?ons less severe, affects employment (18-64). 1 if household Where is deprived is the in at Where least func?on, of weighted # is the weighted taking number deprivations the value indicators, of weighted 1and if experienced household 0 depriva?ons otherwise. is deprivedby household experienced in at least and by household weighted(∙) is and an(∙) indicators, indicator is an and 0 otherwise. indicator e index index contains contains func?on, 14 14 indicators, indicators,taking the which which value areare func?on, 1 grouped grouped if The index household taking into contains the into 6 value 6 is 141different different deprived indicators, if household dimensions. dimensions. in which at is least are deprived It grouped It usesuses weighted in into at a a nested nested least indicators, 6 different weighted weigh?ng weigh?ng dimensions. and indicators, 0 otherwise. It uses and a0 nested otherwise. weigh?n cators, stem, em, where The wherewhich aredimension mul?dimensional each each grouped dimension into poverty weighs weighs 6 approach different 1/6 1/6 of of was dimensions. the the preferred overall overall to Itthe index, index, uses more and anda nested conven?onal each each weigh?ng indicator indicator monetary has has equal equal approach, weight weight aswithin within the provision The index ofthe in-kind contains assistance 14 indicators, The makes system, index contains it difficult which where 14 are each indicators, to compare grouped dimension which consump?on into weighs are 6 different 1/6 of the grouped into levels between hosts and dimensions. overall 6 differentindex, It and dimensions.useseach a nested indicator It uses weigh?ng has a equal nested weight withi weigh?ng nsion e weighs dimension. dimension. 1/6 Dimensions Dimensions of overall with with moreindex, more and indicators indicators thewhere each dimension. indicator will will therefore thereforeDimensions has give equal give with less lessweight more weight weight within indicators to toeach each will specific specific therefore giveindicator. indicator. less weight to each specific indicato refugees, system,and across where different each dimension camps, system, depending weighs each on where1/6 dimension of theweighs assistance overall is 1/6 index, provided of the and in overall each cash/in-kind. index, indicator andWhile each has it indicator equal weight has equal within weight within ns ble le 13with more indicators provides 13provides an an overview overview willof therefore of thethe index index the give Table composi?on dimension. less composi?on 13 weight provides Dimensions and and an each tooverview defini?ons defini?ons with specific more of the of of indicators indicator. index depriva?ons. depriva?ons. composi?on will therefore and give defini?onsless weight of depriva?ons. to each specific indicator. is possible the to dimension. impute the value Dimensions of in-kind with assistance, more it indicators requires detailed will therefore knowledge give of what less weight services to areeach specific indicator. view of the index composi?on andTable defini?ons 13 magnitude provides of depriva?ons. being provided Table 13 in eachprovides loca?on. an Given overview the of the an index of overview the composi?on phenomenon, of the index andit wouldcomposi?on defini?ons also make andthe ofdefini?ons poverty of depriva?ons. depriva?ons. The educa?on indicator is only available for households with children aged 7-15, which means that othe educa?on eeduca?on es?mates indicator indicator very sensi?ve is isonlyonly to available available modelling for for households households assump?ons used withwith in the children children imputa?on. aged aged 7-15, On 7-15, thewhich which other means means hand, using that that a other other only available for households with The educa?on childrenfamilies aged are measured indicator 7-15, is only which on available 13 means indicators for that instead households other ofwith 14. The children alterna?ve aged 7-15, is to 79 which count households means that other withou milies ilies are mul?dimensional aremeasured measured on approach on13 13indicators to indicators poverty means instead instead that of of 14. we 14. are The The restricted alterna?ve alterna?ve to the is indicators is to to count count for which households households data are without without The educa?on indicator is families only children are available as measurednot for deprived on households 13 in educa?on. indicators with The instead children same of 14. issue, aged The but 7-15, less severe, alterna?ve which is to means affects count that other employment households (18-64). without n 13 indicators ildren dren available in instead asasnot notdeprived deprived the K-LSRH. in of ineduca?on. 14. The The educa?on. alterna?ve The same is tobut issue, count but less households severe, affects without employment (18-64). families are measured on same children 13 as indicators not issue, deprived less instead in educa?on. severe, of 14. affects The The same employment alterna?ve issue, but less is to (18-64). severe, count affects householdsemployment without (18-64). n educa?on. The same issue, but less severe, The affects mul?dimensional employment poverty (18-64). approach was preferred to the more conven?onal monetary approach, a children as not deprived in educa?on. The same issue, but less severe, affects employment (18-64). mul?dimensionalpoverty emul?dimensional povertyapproach approachwas The the waspreferred mul?dimensional provision preferredto of in-kind tothe poverty themore assistance moreconven?onal approach makes conven?onalmonetary was it preferred difficult to monetaryapproach, to the compare more approach,as consump?on conven?onal as levels monetary between approach, hosts as an vertyprovisionof eprovision approach of was in-kind in-kind preferred assistancemakes assistance to the the makes more refugees, provision ititdifficult conven?onal difficult and of in-kind to across tocompare compare monetary different assistance camps, makes consump?on consump?on approach, depending it difficult levels levels as toon where compare betweenhosts between assistance consump?onhosts is and approach, as and provided levels in between cash/in-kind. hosts While and The mul?dimensional poverty possible is consump?on approach to impute was thepreferred value of to theassistance, in-kind more conven?onal monetary ssistance makes it difficult to compare refugees, and across levels different between camps, depending hosts on whereitassistance and requires detailed is provided knowledge in cash/in-kind. of whatWhile services it ar ugees,and ugees, andacross across the provision different different camps, of camps, in-kind depending depending assistance being on provided on where makes where in each it assistance assistance difficult loca?on. to is Given iscompare provided provided the magnitude in cash/in-kind. incash/in-kind. consump?on of the levels phenomenon, While While itit between it would hostsalso and the are make povert rent camps, depending on where is possible assistance to impute is provided the value in cash/in-kind. of in-kind assistance, While it of it requires detailed knowledge of what services possible ossible to impute toimpute refugees, thethevalue value and of across ofin-kindin-kind different being assistance, assistance, es?mates provided camps, very in each it it requires depending requires sensi?ve loca?on. to detailed ondetailed modelling Given where theknowledge knowledge assistance assump?ons magnitude of the of is used what what provided phenomenon, the services in services in cash/in-kind. imputa?on. itareare On would also the While makeother the itpoverty hand, using value ing ng of provided provided in-kind in in each each assistance, loca?on. loca?on. it requires Given Given the the detailed magnitude magnitude knowledge of of the the phenomenon, phenomenon,of what services it it wouldwould are also also make make thethe poverty poverty is possible to impute the es?mates value mul?dimensional of in-kind very sensi?ve assistance, approach to modelling to poverty requires itassump?ons means detailed that used we in knowledge are therestricted imputa?on. of to the what On theservices indicators other for hand, are whichusing data a ar mates Given ca?on. ?mates very the verysensi?ve sensi?ve magnitude to to modelling modelling of the mul?dimensional phenomenon, assump?ons assump?ons available in used theit used would approach K-LSRH. inin also the the to make imputa?on. imputa?on. poverty the poverty means On On that the the we other are otherrestricted hand, hand, to using using the a a indicators for which data are being provided in each loca?on. Given the magnitude of the phenomenon, it would also make the poverty to modelling assump?ons ul?dimensional ?dimensional approach approach toto used poverty poverty in available the means means imputa?on. in the K-LSRH. that that we we On are are therestricted restricted other to hand, tothe the using indicators a for forwhich which data are hand, using a es?mates very sensi?ve to modelling 79assump?ons used in indicators the imputa?on. On data the are other ch to poverty means that we are restricted to the indicators for which data are 84 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 13: Multidimensional index composition Dimension Deprivation Weight Definition Education School attendance 1/6 1+ hhd member aged 7-15 is not currently attending school Employment Paid employment 1/6 No hhd members aged 18-64 are in paid work or studying Energy Cooking fuel 1/12 Household uses at solid fuels: wood, farm residue, charcoal, waste Electricity 1/12 Household does not use electricity (grid, generator, solar) for lighting Housing Crowding 1/12 Household has more than 4 members per habitable room Construction 1/12 House has unimproved roof (grass/dung/plastic/tin cans, other) and unim- proved floor (earth, dung, planks, palm, other) and unimproved walls (none, cane, plywood, mud, other) WASH Water 1/12 Household uses unimproved water source: not piped, borehole, protected well/spring, rain, bottle, water booster Toilet 1/12 Household uses unimproved toilet (not flush, ventilated, slab) Nutrition 1.Not enough 1/36 Household had not enough food for at least 5 out of the last 7 days 2.Less preferred 1/36 Household had to eat less preferred foods for at least 5 out of the last 7 days 3.Borrowed 1/36 Household had to borrow food for at least 5 out of the last 7 days 4.Smaller portions 1/36 Household had reduced portion sizes for at least 5 out of the last 7 days 5.Adult meals 1/36 Household had to reduce adult intake for at least 5 out of the last 7 days 6.Fewer meals 1/36 Household had eaten fewer meals for at least 5 out of the last 7 days Annex 3: Additional Figures and Tables Table 14: Deprivation prevalence (% of population), by status/ sex of the head of household Camp refugees Camp Hosts Urban refugees Urban Hosts Dimension Deprivation Male Female Male Female Male Female Male Female Education School 40.7 37.4 70.7 73.2 33.0 33.4 22.7 20.6 attendance Employment Paid 67.7 81.2* 42.0 58.1 29.0 44.5* 9.4 19.2 employment Energy Cooking fuel 98.7 98.6 97.5 97.7 12.0 27.5* 10.6 12.9 Electricity 61.5 67.9 57.6 68.8 0.5 0.4 0.8 0.9 Housing Crowding 28.5 34.5 32.8 31.7 12.8 15.9 7.2 8.1 Construction 4.9 4.3 26.9 41.5 0.2 0.1 - 0 WASH Water 0.6 0.4 17.0 25.5 8.9 7.6 12.1 13.4 Toilet 62.6 55.5 66.7 71.3 6.1 8.9 6.4 8.5 Nutrition Not enough 9.0 10.6 9.6 16.6 14.1 12.4 7.2 4.5 Less preferred 11.0 10.4 13.0 23.3 28.8 25.6 19.6 13.1 Borrowed 2.8 2.4 3.6 7.8 4.3 6.3 2.4 2.3 Smaller portions 7.9 10.5 11.0 14.5 12.6 11.1 7.4 8.5 Adult meals 3.1 3.4 5.5 8.7 4.8 4.4 2.4 3.0 Fewer meals 11.3 12.2 15.5 22.4 12.1 14.5 10.0 8.8 Poverty >1/3rd depriv. 74.8 83.1 73.7 84.6 9.7 16.1* 3.0 7.0 >1/2 depriv. 36.9 40.1 45.0 61.4 0.8 3.3* 0.3 0.0 Source: Authors’ calculations based on K-LSRH 2023. Notes: *difference is statistically significant at 5% level. Building Evidence to Enhance the Welfare of Refugees and Host Communities 85 Table 15: Food insecurity and negative coping strategies (% of population), by status/ sex of the head of household Index Level Camp refugees Camp Hosts Urban refugees Urban Hosts Male Female Male Female Male Female Male Female CSI 4-9: moderately insecure 20.6 18.3 12.9 11.5 21.6 14.7 18.9 17 10+: severely insecure 42.3 42.9 34.7 45.9 46.6 44.6 27.4 27.6 4+: All food insecure 62.9 61.2 47.6 57.4 68.2 59.3 46.3 44.6 LCS Stress 19.5 20.9 24.1 21.5 33.8 37 44.1 43.4 Emergency 17.1 16.1 13.4 23.3 12.3 11 5.5 5.4 Stress + emergency 36.6 37 37.5 44.8 46.1 48 49.6 48.8 Source: Authors’ calculations based on K-LSRH 2023. Note: *difference is statistically significant at 5% level. Table 16: Percent of households receiving assistance in past 12 months, by programme and location / refugee status Turkana Dadaab Nairobi Other urban Kakuma Kalobeyei Hosts Refugees Hosts Refugees Hosts Refugees Hosts (r) (r) Bamba chakula 99.3% 16.8% 97.9% 1.2% 6.2% Bamba chapa 5.6% 85.4% 3.6% 0.9% 1.2% WFP (food) 96.5% 19.5% 59.4% 1.9% 3.9% UNHCR (shelter) 5.6% 5.9% 18.2% 1.6% 1.1% UNHCR (in-kind) 28.9% 34.8% 37.8% 0.4% 2.6% UNHCR (cash) 59.7% 19.1% 7.5% 1.7% 2.3% Fuel/wood 20.9% 11.1% 34.0% 0.2% 2.2% Kitchen garden 1.1% 2.4% 7.1% 0.0% 0.1% HSNP 20.9% 7.9% 0.4% 0.7% OPTC 9.3% 4.5% 0.3% 0.3% CT-OVC 5.8% 5.4% 0.2% 0.3% PWSD-CT 0.6% 2.0% 0.1% 0.1% Linda Mama 9.9% 1.4% 7.3% 2.9% Food aid 4.1% 6.1% 18.9% 36.7% 8.6% 6.0% 2.1% 3.3% 0.8% Health care 2.8% 9.6% 16.0% 34.7% 2.1% 2.9% 5.1% 0.9% 1.2% Hygiene 4.2% 11.1% 1.5% 36.7% 9.4% 1.8% 2.0% 0.8% 0.8% Education 2.2% 3.1% 2.6% 23.3% 8.2% 2.5% 4.5% 3.8% 0.9% Any other 5.8% 19.0% 9.8% 46.6% 12.2% 11.4% 4.3% 8.6% 2.2% Source: Authors’ calculations based on K-LSRH 2023. Notes: r=refugees. The following host transfers are being considered: Hunger and Safety Net Program (HSNP), Older Persons Cash Transfer (OPTC), Cash Transfer to Orphans and Vulnerable Children (CT-OVC), Persons with Severe Disability Cash Transfer (PWSD-CT) and Linda Mama. 86 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 17: Percent of MPI poor households receiving assistance in past 12 months, by programme and location / refugee status Turkana Dadaab Nairobi Other urban Kakuma Kalobeyei Hosts Refugees Hosts Refugees Hosts Refugees Hosts (r) (r) Bamba chakula 99.3% 16.1% 97.5% 1.2% 6.0% Bamba chapa 4.8% 87.1% 3.1% 1.2% 2.8% WFP (food) 96.5% 18.7% 56.7% 2.1% 3.3% UNHCR (shelter) 5.7% 5.5% 13.8% 1.0% 1.0% UNHCR (in-kind) 29.9% 33.3% 35.1% 0.0% 4.5% UNHCR (cash) 59.7% 18.0% 6.4% 0.0% 2.3% Fuel/wood 22.8% 9.9% 31.0% 0.0% 0.6% Kitchen garden 1.1% 2.2% 7.0% 0.0% 0.0% HSNP 23.2% 8.2% 0.0% 2.9% OPTC 10.6% 4.8% 0.0% 3.5% CT-OVC 6.5% 5.1% 0.0% 0.0% PWSD-CT 0.8% 2.2% 0.0% 0.0% Linda Mama 9.9% 1.8% 5.2% 1.4% Food aid 4.1% 6.6% 20.9% 32.5% 8.3% 5.2% 9.6% 5.2% 1.5% Health care 2.1% 9.6% 15.4% 34.7% 1.7% 4.0% 14.4% 0.6% 0.0% Hygiene 3.6% 11.6% 1.6% 33.3% 7.1% 2.9% 8.1% 3.0% 0.0% Education 2.2% 2.8% 2.8% 22.6% 6.4% 1.0% 0.0% 4.1% 0.0% Any other 99.3% 16.1% 97.5% 1.2% 6.0% Source: Authors’ calculations based on K-LSRH 2023. Hunger and Safety Net Program (HSNP) Figure 61: Primary school attendance rates by location and gender 180 156 % of primary-school-aged children 160 135 140 118 113 120 111 108 100 93 86 83 76 76 71 73 80 70 60 39 39 40 20 0 Male Female Male Female Male Female Male Female Camp refugees Camp hosts Urban refugees Urban hosts Net primary attendance Gross primary attendance Source: Authors’ calculations based on K-LSRH 2023. Building Evidence to Enhance the Welfare of Refugees and Host Communities 87 Figure 62: Primary school attendance rates by strata 200 180 173 174 % of primary-school-aged children 160 140 128 120 112 111 110 100 89 88 86 84 81 81 82 80 61 60 53 50 40 26 17 20 0 Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban Refugees Hosts Net primary attendance Gross primary attendance Source: Authors’ calculations based on K-LSRH 2023. Figure 63: Secondary school attendance rates by location and gender 90 79 90 77 76 % of secondary-school-aged 80 79 77 76 66 % of secondary-school-aged 8070 63 61 6266 70 54 63 61 62 60 54 60 47 50 40 47 50 40 children 40 31 children 40 31 24 3030 19 19 24 20 19 19 14 13 20 14 13 1010 00 Male Male Female Female Male Male Female Female MaleMale Female Female Male Male Female Female Camp Camp Camp refugeesCamp refugees hosts hosts Urban Urban refugees refugees UrbanUrban hosts hosts Net secondary Net attendance secondary attendance Gross secondary Gross attendance secondary attendance Source: Authors’ calculations based on K-LSRH 2023. 88 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 64: Secondary school attendance rates by strata 100 90 % of secondary-school-aged children 90 79 77 80 75 70 62 63 60 60 55 50 46 40 40 35 28 27 30 23 18 17 20 11 8 10 0 Kakuma Kalobeyei Turkana Dadaab Dadaab Nairobi Nairobi Other Other Refugees Refugees Hosts Refugees Hosts Refugees Hosts Urban Urban Refugees Hosts Net secondary attendance Gross secondary attendance Source: Authors’ calculations based on K-LSRH 2023. Figure 65: Top 5 reasons for not attending school a. Primary-age refugees b. Primary-age hosts 0 20 40 60 80 0 10 20 30 40 50 60 70 3 57 School f ees School f ees 71 66 Other costs associated with Other costs associated with 6 11 school (uniforms, textbooks, school (uniforms, textbooks, etc.) 16 etc.) 7 17 7 Madrassa Own illness/disability 2 14 11 3 Own illness/disability Insufficient teachers/classrooms 3 14 Not interested/ did not like 11 6 Teacher absenteeism school Camp refugees Urban refugees Camp hosts Urban hosts Building Evidence to Enhance the Welfare of Refugees and Host Communities 89 c. Secondary-age refugees d. Secondary-age hosts 0 20 40 60 80 0 10 20 30 40 50 60 70 9 30 School f ees School f ees 73 60 Not interested/ did not like 15 Not interested/did not like 21 school 3 school 9 Other costs associated with 11 5 Own illness/disability school (uniforms, textbooks, 4 etc.) 9 Other costs associated with 5 7 school (uniforms, textbooks, Parents told me to stop etc.) 8 6 9 4 Marriage Own illness/disability 0 0 Camp refugees Urban refugees Camp hosts Urban hosts Source: Authors’ calculations based on K-LSRH 2023. Notes: This question is asked about primary-school-age children who have attended school in the past but are no longer enrolled or attending. In figure (a), N=152. In figure (b), N=39. Figure 66: Number of Child respondents who speak neither, only English/Swahili, or both. 200 180 160 140 131 133 number of CRs 120 100 80 60 35 40 67 18 25 16 0 38 20 15 22 3 13 1 0 2 0 3 2 1 Kakuma Refugees Kalobeyei Refugees Turkana Hosts Dadaab Refugees Dadaab Hosts Neither Only English Only Swahi li Both Source: Authors’ calculations based on K-LSRH 2023. 90 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 67: Proportion of Child respondents who speak neither, only English/Swahili, or both 100 90 80 44 47 70 64 64 60 % of CRs 0 90 50 40 30 44 53 17 20 33 10 12 0 3 5 0 1 2 6 6 3 5 Kakuma Refugees Kalobeyei Turkana Hosts Dadaab Refugees Dadaab Hosts Refugees Neither % Only English % Only Swahi li % Both % Source: Authors’ calculations based on K-LSRH 2023. Building Evidence to Enhance the Welfare of Refugees and Host Communities 91 Table 18: Regression results on academic achievement Math Early Grade Maths Score (0-5 (1) Score (2) Score (3) Score Kakuma refugees 0.991*** (4.13) 1.148*** (3.92) -0.0107 (-0.04) Kalobeyei refugees 0.771** (3.24) 1.008*** (3.52) -0.301 (-1.04) Dadaab refugees 0.949** (3.14) 1.315*** (3.46) Dadaab hosts 0.677* (2.04) 0.825 (1.78) Age -0.0316 (-0.84) 0.0102 (0.19) -0.00709 (-0.17) Female -0.275* (-2.12) -0.099 (-0.57) -0.352* (-2.45) No. days absent -0.0499 (-0.81) -0.137 (-1.34) -0.0709 (-1.07) Self-efficacy index 0.234*** (3.45) 0.301*** (3.54) 0.203** (2.93) Health issues 0.0835 (0.53) 0.0669 (0.33) 0.0917 (0.56 No. students in class 0.00123 (0.64) 0.00103 (0.44) 0.00212 (1.02) Female CG -0.101 (-0.66) -0.327 (-1.75) -0.104 (-0.57) CG ever attend school 0.16 (1.19) 0.0578 (0.33) 0.138 (0.95) No. HH shocks 0.0615* (2.58) 0.038 (1.03) 0.0640* (2.49) CG moderate anxiety/depression 0.149 (0.76) Length of stay in Kenya -0.0787 (-0.69) Constant 3.174*** (5.36) 2.680*** (3.52) 4.082*** (4.92) Observations 513 234 439 t statistics in parentheses; * p<0.05, ** p<0.01, *** p<0.001 Source: Authors’ calculations based on K-LSRH 2023. Notes: The outcome variable is the raw early grade maths assessment score between 0-5. Correlate coefficients are measured in units of 1 point (equivalent to answering 1 more question correctly). No. days absent represents the number of school days in the past week that the child was absent. The self-efficacy index is the normalized sum of ten questions measuring the child’s self-efficacy with regards to school. Health issues is an indicator set to 1 if the child reports having any health problems that regularly affect them in school (e.g. difficulties with sight or hearing). We use the number of students in the child’s class as a proxy for school environment. CG ever attend school is an indicator for the caregiver having ever attended school. No. HH shocks represents the number of (unique) shocks experienced by the household in the past 5 years. Length of stay in Kenya uses proGres data for refugees to count the number of years (categorized) from the earliest date of arrival in the household to the interview date. CG moderate anxiety/depression is an indicator set to 1 if the caregiver reports symptoms of moderate anxiety (i.e., a GAD-7 score greater than or equal to 10) or moderate depression (i.e. a PHQ-8 score greater than or equal to 10). Note that as the GAD-7 and PHQ-8 scales are only in the representative respondent module, this data on caregiver mental health only exists if the caregiver is also the randomly selected RR 92 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 19: Regression results on academic achievement English Early Grade Reading English Z-Score (1) BZ-Score (2) Z-Score (3) Z-Score Kakuma refugees 0.737* (2.17) 0.235 (0.57) 0.820** (3.92) Kalobeyei refugees 0.478 (1.37) 0.00404 (0.01) 0.491* (2.15) Dadaab refugees -0.0421 (-0.11) -0.264 (-0.61) Dadaab hosts 0.045 (0.13) -0.0361 (-0.08) Age -0.0340 (-1.25) 0.00147 (0.04) -0.0429** (-1.54) Female -0.278** (-2.66) -0.311* (-2.11) -0.267** (-2.61) No. days absent -0.0129 (-0.23) 0.0173 (0.27) 0.0187 (0.42) Self-efficacy index 0.200*** (4.18) 0.158* (2.32) 0.164*** (3.62) Health issues 0.0818 (0.55) 0.256 (1.38) -0.00521 (-0.04) No. students in class -0.00257 (-1.70) -0.00126 (-0.75) -0.00171 (-1.13) Female CG -0.0448 (-0.30) -0.182 (-0.89) -0.195 (-1.37) CG ever attend school -0.108 (-1.03) -0.0640 (-0.42) -0.136 (-1.25) No. HH shocks 0.0281 (1.65) 0.0202 (0.83) 0.0459* (2.51) CG moderate anxiety/depres- sion 0.0665 (0.33) Length of stay in Kenya -0.023 (-0.27) Constant 0.309 (0.61) 0.277 (0.43) 0.411 (0.72) Observations 372 169 313 t statistics in parentheses; * p<0.05, ** p<0.01, *** p<0.001 Source: Authors’ calculations based on K-LSRH 2023. Notes: The outcome variable is the combined Z-score of the English oral reading passage and comprehension. Correlate coefficients are in standard deviation units. No. of days absent represents the number of school days in the past week that the child was absent. The self-efficacy index is the normalized sum of ten questions measuring the child’s self-efficacy with regards to school. Health issues is an indicator set to 1 if the child reports having any health problems that regularly affect them in school (e.g. difficulties with sight or hearing). We use the number of students in the child’s class as a proxy for school environment. CG ever attend school is an indicator for the caregiver having ever attended school. No. HH shocks represents the number of (unique) shocks experienced by the household in the past 5 years. Length of stay in Kenya uses proGres data for refugees to count the number of years (categorized) from the earliest date of arrival in the household to the interview date. CG moderate anxiety/depression is an indicator set to 1 if the caregiver reports symptoms of moderate anxiety (i.e., a GAD-7 score greater than or equal to 10) or moderate depression (i.e. a PHQ-8 score greater than or equal to 10). Note that as the GAD-7 and PHQ-8 scales are only in the representative respondent module, this data on caregiver mental health only exists if the caregiver is also the randomly selected RR. Building Evidence to Enhance the Welfare of Refugees and Host Communities 93 Table 20: Regression results on academic achievement Kiswahili Early Grade Reading Kiswahili Score (1) Score (2) Score (3) Score Kakuma refugees 0.192 (0.85) 0.400 (1.38) 0.496 (1.65) Kalobeyei refugees 0.0823 (0.33) 0.161 (0.49) 0.463 (1.43) Dadaab refugees -0.203 (-0.71) -0.431 (-1.04) Dadaab hosts -0.0752 (-0.24) 0.0856 (0.22) Age -0.0443 (-0.89) 0.00342 (0.06) -0.0542 (-0.94) Female -0.339* (-2.25) -0.135 (-0.67) -0.448** (-2.69) No. days absent -0.197*** (-3.32) 0.301** (-3.03) -0.235** (-3.03) Self-efficacy index 0.276** (3.30) 0.316** (2.92) 0.247** (2.87) Health issues 0.358 (-1.81) -0.256 (-0.84) -0.406* (-2.00) No. students in class -0.00496* (-2.14) -0.0107** (-3.14) -0.00378 (-1.46) Female CG 0.495* (2.45) -0.0828 (0.28) 0.399 (1.73) CG ever attend school 0.178 (1.04) -0.220 (-0.93) 0.174 (0.95) No. HH shocks 0.0140 (0.50) 0.0404 (1.01) 0.0162 (0.50) CG moderate anxiety/depression 0.0632 (0.28) Length of stay in Kenya 0.248 (1.91) Constant 4.154*** (5.55) 4.253*** (4.93) 3.196** (3.02) Observations 454 211 386 t statistics in paratheses; * p<0.05, ** p<0.01, *** p<0.001 Source: Authors’ calculations based on K-LSRH 2023. Notes: The outcome variable is the raw Swahili comprehension score between 0-5. Correlate coefficients are measured in units of 1 point (equivalent to answering 1 more question correctly). No. days absent represents the number of school days in the past week that the child was absent. The self-efficacy index is the normalized sum of ten questions measuring the child’s self-efficacy with regards to school. Health issues is an indicator set to 1 if the child reports having any health problems that regularly affect them in school (e.g. difficulties with sight or hearing). We use the number of students in the child’s class as a proxy for school environment. CG ever attend school is an indicator for the caregiver having ever attended school. No. HH shocks represents the number of (unique) shocks experienced by the household in the past 5 years. Length of stay in Kenya uses proGres data for refugees to count the number of years (categorized) from the earliest date of arrival in the household to the interview date. CG moderate anxiety/depression is an indicator set to 1 if the caregiver reports symptoms of moderate anxiety (i.e., a GAD-7 score greater than or equal to 10) or moderate depression (i.e. a PHQ-8 score greater than or equal to 10). Note that as the GAD-7 and PHQ-8 scales are only in the representative respondent module, this data on caregiver mental health only exists if the caregiver is also the randomly selected RR. 94 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 21: The self-efficacy scale: The self-efficacy scale consists of ten questions to which respondents are required to respond with a score ranging from 0 (not at all true) to 4 (completely true). The questions are: 1. You take good notes during class instruction |___| 2. You can get yourself to study when there are other interesting things to do |___| 3. You can always concentrate on school subjects during class |___| 4. You know how to study to perform well on tests |___| 5. You can remember information presented in class and textbooks |___| 6. You can always manage to solve difficult problems if you try hard enough |___| 7. It is easy for you to stick to your aims and accomplish your goals. |___| 8. You are confident that you could deal with unexpected events. |___| 9. You can remain calm when facing difficulties because you can rely on your coping skills. |___| 10. When you are confronted with a problem, you can usually find several solutions. |___| Figure 68: Perception of safety and crime a. Feel safe walking alone in area/neighbourhood at night 100% 90% 19.6% 80% 39.9% 36.8% 31.7% 47.9% 44.6% 70% 65.8% 63.1% 60% 75.4% 50% 40% 30% 69.4% 49.7% 52.0% 20% 40.8% 42.3% 42.3% 10% 26.0% 28.5% 17.0% 0% Turkana Hosts Kalobeyei Refugees Dadaab Refugees Dadaab Hosts Nairobi Refugees Other Urban Refugees Nairobi Hosts Other Urban Hosts Kakuma Refugees Strong agree/agree Neither di sagree/agree Strongly disagree/disagree Building Evidence to Enhance the Welfare of Refugees and Host Communities 95 b. Are crimes common in your neighborhood/areas? 100% 80% 32% 36% 33% 33% 35% 40% 46% 46% 40% 60% 40% 55% 52% 54% 47% 20% 41% 42% 43% 45% 28% 0% Turkana Hosts Kalobeyei Refugees Dadaab Refugees Dadaab Hosts Nairobi Refugees Other Urban Refugees Nairobi Hosts Other Urban Hosts Kakuma Refugees Strong agree/agree Neither di sagree/agree Strongly disagree/disagree Source: Authors’ calculations based on K-LSRH 2023. Table 22: Factors associated with perceptions of safety. Feeling safe walking alone during the night Age 0.005 Gender (female_dummy) -0.292*** Marital status (married_dummy) -0.045 Secondary education or higher 0.095 Location (Ref: Turkana Refugees) Turkana Hosts 1.497*** Dadaab Refugees 1.376*** Dadaab Hosts 2.534*** Nairobi Refugees 0.700*** Nairobi Hosts 1.148*** Other Urban Refugees 1.115*** Other Urban Hosts 1.617*** Observations 8,918 Source: Authors’ calculations based on K-LSRH 2023. Note: Regression results are weighted using survey weights; Dependent variables values range from 1 (Strongly disagree/disagree) to 3 (Strongly agree/agree); *** p<0.01, ** p<0.05, * p<0.1 96 Building Evidence to Enhance the Welfare of Refugees and Host Communities Figure 69: Views on socialization of children a. Comfort with children socialize with a. Comfort with children socialize with refugees children (hosts’ perceptions) refugees children (hosts’ perceptions) 100% 100% 13.7% 12.2% 16.4% 90% 27.9% 80% 10.1% 10.8% 80% 25.1% 70% 14.0% 60% 60% 50% 40% 4 0% 78% 76.2% 77.1% 74% 30% 61% 58.1% 58.5% 51% 20% 20% 35% 10% 0% 0% Turkana Dadaab Nairobi Other Kakuma Kalobeyei Dadaab Nairobi Other refugees refugees refugees refugees urban hosts hosts hosts urban hosts refugees Strongly disagree/disagree Strongly agree/agree Neither di sagree/agree Neither di sagree/agree Strongly disagree/disagree Strongly agree/agree c. Comfort with children socialize with refugees’ children from different ethnic groups (refugees’ perceptions) 100% 80% 60% 4 0% 68.2% 62.6% 57.1% 57.4% 46.9% 20% 0% Kakuma refugees Kalobeyei refugees Dadaab refugees Nairobi refugees Other urban refugees Strongly agree/agree Neither di sagree/agree Strongly disagree/disagree Source: Authors’ calculations based on K-LSRH 2023. Building Evidence to Enhance the Welfare of Refugees and Host Communities 97 Table 23: Correlates of comfort with children socializing with outgroups. Hosts Refugees Children socializing Children socializing with Children socializing with ref- with refugee children host children ugee from different tribes Employment (employed_dummy) 0.031 0.175 0.044 Age -0.005 0.001 -0.002 Gender 0.074 -0.129 -0.131 Marital status (married_dummy) 0.063 0.075 0.070 Secondary education or higher 0.138 -0.054 -0.110 Location (Ref: Turkana Host) Daadab Hosts 0.859*** Nairobi Hosts 0.763*** Other Urban Hosts 0.094 Location (Ref: Turkana Refugees) Dadaab Refugees 0.0676*** 0.392*** Nairobi Refugees 1.362*** 0.695*** Other Urban Refugees 1.288*** 0.474** Active membership (Ref: None) Active membership (1 to 3 groups) 0.025 0.558*** 0.457*** Active membership (4 to 6 groups) 1.080*** 1.103*** 1.960*** Active membership (7 or more groups) - 0.276 1.514*** 1.989*** Social networks % of refugee friends (>50%) 0.960*** % of Kenyan friends (>50%) 0.012 -0.157 % of refugee friends from the same coun- 0.318*** 0.182* try (>50%) % of refugee friends from different coun- -0.254** -0.006 try (>50%) % of refugee friends from different tribe 0.020 -0.145 (>50%) Observations 3,017 4,546 4,547 Source: Authors’ calculations based on K-LSRH 2023. Notes. Regression results are weighted using survey weights; Dependent variables values range from 1 (Strongly disagree/disagree) to 3 (Strongly agree/agree); Social network variables (% of friends) is a dummy variable 1(>50%) & 0 (<50%) *** p<0.01, ** p<0.05, * p<0.1 98 Building Evidence to Enhance the Welfare of Refugees and Host Communities Table 24: Regression on correlates of school preferences among refugee children Mixed school Child age 0.018 Child gender 0.036 Caregiver gender 0.009 Caregiver education level 0.137** Location (Ref: Dadaab refugees) Kakuma refugees 0.002 Kalobeyei refugees 0.037 Number of Kenyan friend (Ref: None) Number of Kenyan friends (One friend) 0.174*** Number of Kenyan friends (2-5 ) -0.001 Number of Kenyan friends (>5 ) 0.083 Multidimensional poverty 0.055 Constant 0.418 Observations 434 Source: Authors’ calculations based on K-LSRH 2023. Notes: Regression results are weighted using survey weights; *** p<0.01, ** p<0.05, * p<0.1 Table 25: Proportion of respondents that have experienced shocks Camp refugees Camp hosts Urban refugees Urban hosts floods 57.7% 89.3% 19.4% 27.6% droughts 24.3% 21.3% 3.6% 4.0% landslides/erosion 76.0% 61.9% 96.3% 88.8% pests 3.4% 0.0% 0.0% 0.0% weeds .0% .0% .0% .0% livestock diseases 88.5% 75.6% 83.7% 85.2% disease outbreaks 65.0% 38.9% 46.5% 41.8% theft 24.6% 21.7% 10.9% 5.1% loss of land or rental property 34.6% 31.0% 59.7% 50.0% loss of employment 34.6% 31.0% 59.7% 50.0% death of household member 39.3% 45.1% 34.0% 32.5% Source: Authors’ calculations based on K-LSRH 2023. Building Evidence to Enhance the Welfare of Refugees and Host Communities 99 Table 26: Proportion of respondents using coping strategy among those experiencing the shock loss of death of land- livestock disease land or loss of flood drought pests weeds theft household slide diseases outbreaks rental employment member property Other .5% 1.2% 1.2% 2.7% 2.2% 1.1% 1.5% 1.0% 4.2% .3% .5% New work 1.1% 3.8% 1.6% 2.9% 3.4% 1.3% 2.5% 2.5% 1.8% 13.1% 2.0% Child dropped school/ 1.4% 1.6% 2.4% 2.4% 1.8% 2.5% 2.4% 1.5% 1.4% 2.2% 2.7% looked for work Sold assets 2.0% 3.3% 3.7% 2.4% 2.5% 2.5% 2.1% 1.1% 2.2% 4.5% 2.4% Leased 2.2% 2.4% 5.3% 3.9% 4.3% 2.7% 1.4% .6% 27.9% 1.4% 2.1% land Cheaper 2.4% 5.3% 5.5% 3.4% 3.4% 3.6% 2.8% 4.1% 23.7% 8.9% 2.0% housing Took loan 3.2% 6.7% 6.7% 8.7% 4.7% 6.2% 4.6% 5.0% .3% 6.3% 6.7% Migrate/ send kids/ 4.2% 3.7% 3.8% 3.1% 3.4% 4.3% 2.5% 1.6% 13.1% 1.8% 2.9% relatives Spiritual 4.5% 10.4% .0% 5.3% 6.3% 5.8% 16.7% 8.9% 8.9% 5.8% 28.2% efforts Reduced non-essen- 6.4% 11.3% 3.3% 5.7% 6.3% 5.2% 14.0% 17.8% 4.5% 27.9% 5.6% tial expense Aid from 7.7% 5.5% .0% 11.2% 9.5% 6.6% 7.4% 2.7% 4.6% 4.6% 7.2% govt Sold/killed 10.3% 10.2% .0% 14.2% 11.7% 20.0% 4.7% 1.7% 6.3% 4.2% 5.9% livestock Reduced food con- 19.9% 13.7% .0% 9.1% 8.1% 8.3% 10.2% 16.1% 9.8% 23.7% 6.8% sumption Help from friends/fam- 20.8% 7.0% 100.0% 10.2% 13.1% 7.8% 9.1% 5.3% 5.8% 9.8% 15.0% ily/church Used sav- 23.1% 25.8% 21.0% 30.1% 34.1% 34.4% 28.8% 25.3% 17.0% 38.1% 32.5% ings Nothing 34.4% 33.4% .0% 33.7% 34.0% 29.0% 30.7% 40.5% 38.1% 17.0% 20.6% 100 Building Evidence to Enhance the Welfare of Refugees and Host Communities Building Evidence to Enhance the Welfare of Refugees and Host Communities 101 Building Evidence to Enhance the Welfare of Refugees and Host Communities