CONSUMER AND MARKET STUDY IN SOUTHWEST AND WEST NILE REFUGEE- HOSTING AREAS IN UGANDA DECEMBER 2021 SUPPORTED BY: DISCLAIMER © International Finance Corporation 2021. All rights reserved. 2121 Pennsylvania Avenue, N.W. Washington, D.C. 20433 Internet: www.ifc.org The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. IFC encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly, and when the reproduction is for educational and non-commercial purposes, without a fee, subject to such attributions and notices as we may reasonably require. IFC does not guarantee the accuracy, reliability, or completeness of the content included in this work, or of the conclusions or judgments described herein. IFC accepts no responsibility or liability for any omissions or errors (including, without limitation, typographical errors and technical errors) in the content whatsoever or for reliance thereon. 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 concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The contents of this work are intended for general informational purposes only and are not intended to constitute legal, securities, or investment advice, an opinion regarding the appropriateness of any investment, or a solicitation of any type. IFC or its affiliates may have an investment in, provide other advice or services to, or otherwise have a financial interest in, certain of the companies and parties named herein. All other queries on rights and licenses, including subsidiary rights, should be addressed to IFC’s Corporate Relations Department, 2121 Pennsylvania Avenue, N.W., Washington, D.C. 20433. International Finance Corporation is an international organization established by Articles of Agreement among its member countries, and a member of the World Bank Group. All names, logos, and trademarks of IFC are the property of IFC and you may not use any of such materials for any purpose without the express written consent of IFC. Additionally, “International Finance Corporation” and “IFC” are registered trademarks of IFC and are protected under international law. i ACKNOWLEDGEMENTS This consumer and market study was a collaborative effort between IFC and ASIGMA Advisory, a Kampala-based consulting firm. ASIGMA designed the methodology under IFC’s guidance, fine-tuned and contextualized the household survey questionnaire developed by IFC, managed field data collection, analyzed data, and developed some of the recommendations included in this report. The efforts on ASIGMA’s side were led by David Nanambi Wakyiku (Managing Director, ASIGMA) with key team members Prisca Adong (Advisory Associate, ASIGMA) and Joel Kasibante (Project and Data Associate, ASIGMA). IFC consultants Romain Galgani (Statistician) and Polina Lenkova (Evaluation Expert) conducted further statistical analysis and produced the final report. Luba Shara (Senior Operations Officer, IFC), Muhammad Taif Ul Islam (Operations Officer, IFC), Marion Kimani (Operations Officer, IFC), Sandra Leticia Sanchez Flores (Consultant, IFC), Eyoual Demeke (Consultant, IFC), Awena Lebeschu (Consultant, IFC), and Gerald Owachi (Consultant, IFC) reviewed the report and provided feedback at the draft stage. We are grateful to Benjamin Reese (Senior Operations Officer, World Bank), Paul Turner (Senior Private Sector Specialist, World Bank), and Michael Wong (Lead Economist, World Bank), who peer-reviewed the report and provided valuable feedback. We also acknowledge the valuable inputs from Mark Lewis (Global Agricultural Specialist) that helped finalize the report. The team is also very grateful to the Uganda Bureau of Statistics (UBOS) and UNHCR for providing ASIGMA with access to sampling frames for surveying the host community and refugee settlements, and to the Office of the Prime Minister with which ASIGMA engaged during the data collection. We also acknowledge the inputs provided by the key informants as well as the sampled refugees and host community members who gave their time to participate in the study. Photos included in this report were taken by ASIGMA Advisory. Clarity Global Strategic Communications copyedited and designed the report. 1 CONTENTS ACKNOWLEDGEMENTS 1 LIST OF FIGURES 4 LIST OF TABLES 5 ACRONYMS 5 GLOSSARY 6 ABOUT THE REPORT 8 EXECUTIVE SUMMARY 11 1. INTRODUCTION 17 1.1 BACKGROUND 17 2. METHODOLOGY 18 2.1 GEOGRAPHICAL SCOPE 18 2.2 HOUSEHOLD SURVEY 20 2.3 BUSINESS SURVEY 20 2.4 QUALITATIVE DATA COLLECTION 21 2.5 DATA WEIGHTING, STATISTICAL SIGNIFICANCE, AND DATA VISUALIZATION 21 2.6 METHODOLOGY TO ESTIMATE HOUSEHOLD CONSUMPTION EXPENDITURE 21 2.7 STUDY LIMITATIONS 22 3. SOCIOECONOMIC CHARACTERISTICS OF THE SOUTHWEST AND WEST NILE 24 3.1 OVERVIEW OF THE REFUGEE SITUATION IN UGANDA 24 3.2 DEMOGRAPHIC CHARACTERISTICS OF REFUGEES AND THE HOST COMMUNITY 27 3.3 EDUCATION 29 3.4 EMPLOYMENT 29 3.5 INCOME 31 4. ACCESS TO TELECOMMUNICATION AND FINANCIAL SERVICES 34 4.1 TELECOMMUNICATION SERVICES 34 4.2 FINANCIAL SERVICES 36 2 5. HOUSEHOLD CONSUMPTION EXPENDITURE IN THE SOUTHWEST AND WEST NILE 42 5.1 HOUSEHOLD CONSUMPTION EXPENDITURE OVERVIEW 42 5.2 BREAKDOWN OF HOUSEHOLD CONSUMPTION EXPENDITURE 46 5.3 FOOD AND BEVERAGES 48 5.4 NON-FOOD ITEMS 54 5.5 ENERGY EXPENDITURE 55 5.6 HOUSING AND SANITATION 56 6. PROFILE OF BUSINESSES IN THE SOUTHWEST AND WEST NILE 60 6.1 BUSINESS OWNERS’ DEMOGRAPHIC CHARACTERISTICS 60 6.2 TYPES OF BUSINESSES AND THEIR CHARACTERISTICS 61 6.3 REGISTERING A BUSINESS 62 6.4 AVERAGE MONTHLY SALES OF THE SURVEYED BUSINESSES 63 6.5 BUSINESSES’ ACCESS TO FINANCE 64 7. AGRICULTURAL VALUE CHAINS IN THE SOUTHWEST AND WEST NILE 65 8. INVESTMENT OPPORTUNITIES 68 8.1 AGRICULTURAL AND FISHING VALUE CHAINS 68 8.2 ENERGY FOR LIGHTING 71 8.3 FINANCIAL SECTOR 71 8.4 HOUSING AND SANITATION 72 8.5 ARTS AND CRAFTS PRODUCTION AND MARKETING 73 8.6 RETAIL TRADE 73 8.7 PLASTIC RECYCLING 74 8.8 EDUCATION—VOCATIONAL TRAINING 74 ANNEX 75 3 LIST OF FIGURES Figure O.1: Southwest and West Nile districts surveyed 9 Figure O.2: Market sizes of the Southwest and West Nile 13 Figure O.3: Average monthly household expenditure 13 Figure 2.1: Map showing refugee settlements in Uganda 19 Figure 3.1: Uganda’s refugee population, 2000 to April 2021 24 Figure 3.2: Age distribution of respondents 27 Figure 3.3: Country of origin of respondents from refugee settlements 28 Figure 3.4: Respondents’ employment status 31 Figure 3.5: Impact of COVID-19 on monthly household earnings 32 Figure 3.6: Impact of COVID-19 on individual earnings 33 Figure 4.1: Ownership of mobile phones and SIM cards 34 Figure 4.2: Phone types owned by respondents 35 Figure 4.3: Mobile network operators’ penetration 35 Figure 4.4: Usage of phone/SIM card for mobile money transfers 36 Figure 4.5: Bank account ownership 37 Figure 4.6: Sources of loans in the past 12 months 38 Figure 4.7: Saving mechanisms used 39 Figure 4.8: Reasons for saving 39 Figure 4.9: Sources of information on financial matters 40 Figure 5.1: Market sizes of the surveyed districts 43 Figure 5.2: Average monthly household expenditure by gender 44 Figure 5.3: Average monthly household expenditure by nationality 44 Figure 5.4: Impact of COVID-19 on expenditure 44 Figure 5.5a: Annual household expenditure in the Southwest 45 Figure 5.5b: Annual household expenditure in West Nile 45 Figure 5.6: Expenditure on consumer goods and other categories 46 Figure 5.7: Percentage of population that spends on a particular category 47 Figure 5.8: Most popular food and beverage categories 48 Figure 5.9: Most popular cooking oil brands 49 Figure 5.10: Most popular tea brands 49 Figure 5.11: Most popular rice brands 50 Figure 5.12: Access to foods and beverages 50 Figure 5.13: Popular venues for buying food items 51 Figure 5.14: Nationality of vendors of food items 52 Figure 5.15: Typical vendors of home and personal care items 54 Figure 5.16: Sources of energy for lighting 55 Figure 5.17: Willingness to pay for better access to energy 56 Figure 5.18: Willingness to pay more for specific lighting products 56 Figure 5.19: Quality of housing facilities in refugee-hosting areas 57 Figure 5.20: Willingness to pay for improved housing types 57 Figure 5.21: Toilet arrangements used by households 58 Figure 5.22: Better toilet arrangements that respondents are willing to pay for 58 Figure 6.1: Gender of business owners 60 Figure 6.2: Surveyed businesses by sector and geographic area 61 Figure 6.3: Number of employees hired by surveyed businesses 62 Figure 6.4: Businesses’ bank account ownership 64 Figure 6.5: Businesses’ willingness to open a bank account 64 Figure 6.6: Percentage of respondents who obtained a loan in the previous 12 months 64 Figure 7.1: Basic illustration of agricultural value chains in refugee-hosting areas in the Southwest 67 4 LIST OF TABLES Table 2.1a: Refugee-hosting districts surveyed 18 Table 2.1b: Refugee settlements surveyed 18 Table 2.2: Household sample size targets and achievements 20 Table 2.3: Business sample size targets and achievements 21 Table 2.4: Number of focus group discussions 21 Table 3.1: Population of refugee-hosting districts in Uganda 25 Table 3.2: Respondents’ levels of education 29 Table 6.1: Time it takes to register a business 63 Table 7.1: Agricultural value-chain mapping in refugee-hosting districts 65 Table A.1: List of key informants interviewed 75 ACRONYMS BTVET business, technical, vocational education and training CRRF Comprehensive Refugee Response Framework DRC Democratic Republic of the Congo FGD focus group discussion IFC International Finance Corporation NGO nongovernmental organization NRC Norwegian Refugee Council OPM Office of the Prime Minister SACCO savings and credit cooperative organization UBOS Uganda Bureau of Statistics UGAFODE Uganda Agency for Development Limited UN United Nations UNDP United Nations Development Programme UNHCR United Nations High Commissioner for Refugees US United States VSLA village savings and loan association WFP World Food Programme YIDA Youth Initiative for Development in Africa 5 GLOSSARY Camps and Base camp A refugee settlement’s administrative center, hosting the Office of settlements the Prime Minister, United Nations (UN) agencies, and nongovern- mental organizations (NGOs). Refugee A temporary facility built to provide immediate protection for ref- camp ugees. Refugee A designated area in Uganda where refugees live. Unlike refugee settlement camps, which are built to provide immediate protection, Ugandan refugee settlements resemble towns or villages with a high degree of integration with the surrounding host communities. Economic Aggregators Individuals who travel between towns to buy agricultural produce activities, from rural assemblers. Aggregators are mainly based in towns and places, and travel to trading centers for produce, while rural assemblers are products based in trading centers and travel to villages to source produce. Home and Items used in a household such as brooms or utensils, and personal care personal hygiene items such as toothpaste or body lotion. products Housewife A woman whose main occupation is caring for her family, managing household affairs, and doing unpaid housework. She may depend on income from a husband or partner. Kitenge African printed fabric often worn by women. Local Business lines or jobs that involve making items—for example, production manufacturing, artisanry, carpentry, metal fabrication, and mechanical and auto works. Outside of Not employed and not looking for work. the labor force Trading A center where economic activities such as shops, markets, and center restaurants are concentrated. Education Incomplete Student attended primary school but did not complete the seven levels (in primary academic years. ascending education order) Incomplete Student attended secondary school but did not complete the four O-level academic years of O-level education. O-level Ordinary level: Lower secondary education covers the first four academic years of high school. 6 A-level Advanced level: Upper secondary education covers the final two years of high school. Business, Training that provides practical knowledge and skills necessary technical, for employment in a skilled craft or trade. It is usually offered by vocational a high school, technical institute, or college, leading to the award education of a certificate. and training (BTVET) Finances Merry- A rotating savings and credit association collects money from and savings go-round members at each meeting and gives the total to one member on a rotating basis. Savings A savings group owned, governed, and managed by its members, and credit who may share an employer, church, labor union, or social frater- cooperative nity, or live/work in the same community. organization (SACCO) Village A group of people who meet regularly to save together and take savings and small loans from those savings. loan associa- tion (VSLA) Mobile Basic phone A basic phone, or button phone, has no Internet access; it has phone limited functions such as texts and calls. types Feature A feature phone can access the Internet, take photos, and phone store and play music, but lacks the advanced functionality of a smartphone; it may have a small display and an alphanumeric or QWERTY keypad. Smartphone A phone with advanced functions, such as an updatable oper- ating system, a touch screen, and the ability to install and run downloaded apps to send emails, browse the web, edit photos, and so on. Solar Solar home A small system designed with one or more solar panels to absorb products system (solar the sun’s rays as a source of energy for generating electricity or panels) heating, usually for one home. Solar lantern A portable lamp that provides lighting and is recharged using solar energy. 7 About the Report T he International Finance Corporation (IFC) the way in which governments and other stake- commissioned a Consumer and Market holders, including the private sector, respond to Study to explore economic activities, em- forced displacement crises. In particular, it aims ployment trends, consumption levels, and con- to (i) enhance the enabling environment for the sumer preferences of refugees and host commu- socioeconomic inclusion of forcibly displaced nities in Uganda’s largest refugee-hosting areas persons; (ii) enhance access to education and in the Southwest and West Nile regions. protection services for forcibly displaced persons and host communities; and (iii) strengthen the The study covers a gap in existing research on resilience of host communities through inclusive the economic situations of forced displacement, socioeconomic development that also benefits which is often conducted from a humanitarian forcibly displaced persons. perspective and rarely offers the private sector view. The study’s primary target audience is The report is divided into eight chapters. Chapter 1 private companies looking to enter this substan- introduces the study. Chapter 2 outlines the study tial, yet mostly untapped, market. Commercial methodology. Chapter 3 provides socioeconomic and financial data is necessary for private sector baseline data, such as educational attainment, engagement, but there is seldom information employment, and income, comparable by region available on refugees outside of academic, and population group (refugees versus host development, and humanitarian studies. The communities). Chapter 4 explores access to tele- study presents the refugees’ economic activities communication and financial services. Chapter 5 in their distinct roles as consumers, producers, analyzes household consumption expenditure, suppliers, and salaried workers from the view the volume of economic activity, consumer of a private sector firm entering the market. It preferences, and access to finance and tele- builds on earlier research conducted by the communication services. Chapter 6 discusses Uganda Investment Authority, in partnership with findings from the business survey. Chapter 7 the United Nations Development Programme briefly looks at agricultural value chains in the (UNDP), which produced investment profiles for Southwest and West Nile. Chapter 8 presents refugee-hosting districts. investment opportunities in the refugee-hosting districts for the private sector. The study is financed by the Dutch Partner- ship on Inclusive Jobs and Education for Host Communities, Refugees and Other Forcibly Displaced Persons (PROSPECTS). This Partner- ship is a multi-year program that brings together five agencies: IFC, the World Bank Group, the United Nations High Commissioner for Refugees (UNHCR), the International Labour Organization (ILO), and the United Nations Children’s Fund (UNICEF). The Partnership aims to transform 8 Figure O.1: Southwest and West Nile districts surveyed Moyo Yumbe Arua Adjumani Hoima Kyegegwa Kamwenge Kampala, capital city Legend Southwest districts surveyed Isingiro West Nile districts surveyed 9 10 Executive Summary Introduction population. On average there are 6.5 people U per household. West Nile households are larger, ganda hosts the world’s fourth-largest with an average of eight people in refugee set- refugee population and has imple- tlements and seven people in host community mented one of the most progressive areas. The majority of household heads in the refugee management policies, which hinges on sample are female (68 percent among refugees refugees’ self-reliance. Uganda allows refugees and 54 percent among host communities). In to exercise fundamental rights such as being refugee settlements, 18–34-year-olds make employed, having freedom of movement, owning up 54 percent of the sample, while in host property, and establishing a business. Refugees communities they constitute 45 percent of the are included alongside their host communities in sample. The majority of refugees in the South- government policies on health, education, and west are from the Democratic Republic of the livelihood development. Congo (DRC), while most refugees in West Nile are from South Sudan. The Consumer and Market Study explores the economic activities, employment trends, and The household survey shows low levels of consumer preferences of refugees and host education overall—26 percent of household communities in eight refugee-hosting districts in decision-makers have had no schooling at all, the Southwest and West Nile1 regions, offering and only 34 percent completed primary school. potential private-sector entry points in the two Overall, refugees are more likely to have no regions. schooling than members of the host community. There is a significant gender gap in educational Methodology attainment. Overall, 33 percent of women have no The study is based on mixed-methods research, schooling, compared to 11 percent of men, and using quantitative and qualitative primary data this gap is slightly larger in refugee settlements. collection (via household and business surveys, key informant interviews, and focus group dis- Host community members are significantly cussions) and secondary sources. The house- more likely to be employed than refugees, hold survey sampled 2,522 household heads and Southwest residents are more likely to or household members who regularly make be employed than those living in West Nile. purchasing decisions. In host communities, the employment rate (including employees, business owners, and farmers) among respondents is at 80 percent in Household demographics the Southwest and 60 percent in West Nile. In The refugee-hosting areas in both regions have a refugee settlements, the employment rate among high prevalence of female-headed households, the respondents is much lower, at 51 percent in a large average household size, and a young the Southwest and 22 percent in West Nile. 1 For the purposes of this study, the Southwest comprises the districts of Hoima, Isingiro, Kamwenge, and Kyegegwa. The West Nile region comprises the districts of Adjumani, Arua, Moyo, and Yumbe. The study considered old district boundaries because of lack of data on the newly formed districts that was necessary for household survey sampling. 11 Agricultural work is the most common Internet-enabled phones (by more than two- occupation in the Southwest, while owning a thirds of respondents) prevents a significant business is a prevalent occupation in West Nile. portion of refugees and hosts from accessing the The overall distribution of employment type is Internet. Among mobile network operators, MTN similar between refugees and host communities, has the largest market share in both regions. but there are significant regional differences. In West Nile, business owners make up more than a Mobile money drives the adoption of formal third of those employed, compared to 8 percent financial services, especially in the South- in the Southwest. The Southwest has a much west host community and among West Nile larger share of agricultural workers (45 percent) refugees. Sixty-four percent of Southwest host than West Nile (11 percent) because the quality of community respondents use mobile money ser- land is more suitable for cultivation there. Fewer vices (compared to 30 percent of refugees), as opportunities in agriculture in West Nile also do 55 percent of West Nile refugees (compared explain the lower employment rate in the region. to 46 percent of the host community). About a quarter of respondents in the two regions have Host community households earn twice as much bank accounts, with a higher percentage among as refugee households do. Besides employ- Southwest refugees (65 percent) because of ment, remittances contribute significantly to cash transfers. For refugees, increased adoption the household incomes of both refugees and of financial services is driven by mobile money hosts, while humanitarian aid is an important services, and by humanitarian cash transfers income source for refugees. The survey found from NGOs and UN agencies, facilitated by the highest household earnings among West Nile Equity Bank and PostBank (in the Southwest). host community households ($64) and the lowest among West Nile refugees ($31). Compared to Refugees and host community members tend income levels before COVID-19, the average to borrow and save informally. Both communi- income decreased by 15 percent in both regions ties primarily borrow from village savings and in the period immediately following the lockdown. loan associations (VSLAs are the most common In the Southwest, refugees receive humanitarian community savings group in rural Uganda, cash transfers from NGOs and UN agencies. A allowing members to receive interest on savings much smaller share of West Nile refugees receive and access microloans); they also borrow from humanitarian cash transfers, so they tend to friends and family, and buy goods on credit from resell the food (and other) rations that they shopkeepers. Borrowing from banks or from a receive to gain some extra income. Remittances mobile money service is uncommon across all from within and outside of Uganda represent surveyed population groups. VSLAs are also an income source for both refugees and host the most common means of saving (along with communities. Eighteen percent of Southwest keeping cash at home), while saving with finan- refugees and 10 percent of West Nile refugees cial institutions is very uncommon. VSLAs tend receive remittances. Remittances are also an to store money in cash rather than in a bank or income source for 22 percent of host community mobile money account. households in West Nile. Household consumption expenditure Access to telecommunication The market size—measured as total household and financial services consumption expenditure—of the refugee- Mobile phone ownership is high in the two hosting districts of the two regions (1.1 million regions; however, most owners have basic households) is estimated at $485 million annually. (button) phones. Sixty percent of South- The West Nile region accounts for $246 million west refugees and 70 percent of West Nile (51 percent), which is largely driven by the refugees declared owning a phone, compared to host community’s spending, at $201 million. 70 percent of the Southwest host community and The Southwest region contributes $239 million 52 percent in West Nile. Refugees can register a (49 percent) to the total. The market size is SIM card on par with Ugandan citizens, as long slightly larger in West Nile than in the Southwest as they can produce a refugee identification card because of the larger population size in the former or attestation letter from the Office of the Prime (588,000 versus 514,000 households). However, Minister. The widespread ownership of non- per-household monthly expenditure is higher 12 Figure O.2: Market sizes of the Figure O.3: Average monthly Southwest and West Nile household expenditure Southwest Southwest $165 million $74 million $37 $43 Host community Refugees Host community Refugees West Nile $201 million West Nile $45 million $39 $24 Host community Refugees Host community Refugees in the Southwest at $39, compared to $35 in In the Southwest, refugees are more likely to West Nile, a contrast which is even more evident shop from independent sellers/stalls, while for refugee households ($43 in the Southwest grocery stores are more popular with the host versus $24 in West Nile—possibly because many community. Host communities mostly buy from West Nile refugees receive food rations). Overall, fellow Ugandans, although 11 percent also refugee households’ expenditure accounts buy from refugees. Three-quarters of refugees for 25 percent of the total annual expenditure. buy from refugee vendors, while there are The Kyegegwa and Isingiro districts are the regional differences in buying from Ugandan largest markets in the Southwest, whereas Arua vendors—82 percent of refugees in West Nile and Yumbe are the largest markets in West Nile. shop at Ugandan businesses, but only a third do so in the Southwest. The majority of refugees Consumer goods dominate household spending and host communities are able to access most of in both regions, representing 59 percent of the foods and beverages they need. the total expenditure ($286 million). Within consumer goods, food and beverages account Solar home systems are the main source for 45 percent of the total expenditure, home of lighting in the Southwest, as are battery- and personal care for 6 percent, clothing for operated torches in West Nile. National grid con- 5 percent, alcoholic drinks for 2 percent, and nection in the surveyed areas is mostly available lighting for 1 percent. Cooking oil, fruits and in major towns and settlement base camps. In vegetables, rice, and meat are in the top five West Nile, access to the grid is particularly low in purchased items in both regions. Average monthly the host community (6 percent) and non-existent household expenditure on services decreased among refugee households. Willingness to pay by 30 percent (from $21 before the lockdown to for better access to energy for lighting is high $15 after the lockdown). This was largely driven among both refugees and host communities. by a reduction in education expenditure due to Host communities in both regions (59 percent) schools being closed during the data collection are more likely than refugees (45 percent) to be period. willing to pay for improved energy. Furthermore, more than half of respondents are willing to pay Refugees and host community members more for solar home systems. West Nile refugees purchase consumed items in the same places. are especially interested in acquiring solar lanterns They mostly buy from open markets, followed (with about two-thirds being willing to pay). by independent sellers and grocery stores. 13 Semi-permanent housing structures made of Refugee and host community members tend mud or unburnt bricks are prevalent in both to own the same types of businesses. They regions. A typical Southwest house is made of are mostly duukas—small shops selling basic, mud, while in West Nile it is made of unburnt readily available consumer products, followed by bricks. In the Southwest, 38 percent of refugees barbershops, food stalls, and drug stores. In West and 65 percent of the host community are willing Nile refugee settlements, the business survey to pay for improved housing, mainly using burnt also revealed a presence of host community- bricks. In West Nile, 40 percent of refugees and owned businesses, mainly duukas, drug stores, 46 percent of the host community are willing and mobile money kiosks. The average refugee to do so. As most of the sample use a covered business has monthly sales of $220, while host traditional pit latrine without a concrete floor slab, businesses generate monthly sales of $388 willingness to pay for improved toilet facilities on average. Furthermore, businesses in the is significant; it is higher in the Southwest than Southwest generate on average more revenue West Nile, and higher among host communities ($407) than those in West Nile ($253). Half of the than refugees. In the Southwest, there is more surveyed business owners do not employ other willingness to pay for ventilated improved pit people, and most others have only one to two latrines whereas in West Nile the willingness to employees. Refugee-owned businesses that pay is mainly for covered traditional pit latrines have employees often employ Ugandans, while with concrete floor slabs. it is also common for Ugandan businesses to employ refugees. Only a third of the surveyed businesses are fully registered with the Ugandan Profile of businesses authorities; however, most report paying fees to Businesses operating in both regions tend to be local authorities to maintain their businesses. small retail businesses and service providers. 14 Agricultural value chains Investment opportunities The most prevalent agricultural value chains in Agribusiness both regions are cassava, maize, and beans. Bananas (matoke) are also widely grown, 1. Maize aggregation and processing especially in the Southwest. Other major crops Supporting maize aggregation: Currently, grown in the Southwest are tea and coffee. Sweet only small-scale mills exist in the Southwest. potatoes are a major food crop in West Nile, as Setting up slightly larger mills within and close well as sorghum, finger millet, and sesame. West to the settlements could enable the emergence Nile farmers grow tobacco as a cash crop, and of refugee aggregation businesses. Given the there are also sunflower value chains. high production volumes in the refugee-hosting areas, opportunities could be explored for large- Farmers often face challenges in accessing scale aggregators and post-harvest warehouse good-quality agricultural inputs. NGOs some- businesses to establish or extend their operations times supply seeds for free, but commercially in refugee-hosting districts. traded seeds are expensive, especially for refugees. Farmers often use leftover seeds from 2. Cassava processing previous seasons, which results in poor yields. Opportunities include: Another challenge is that often there are no  Setting up cassava chipping points (in refugee outlets selling inputs near to refugee farmers, settlements and host community areas). forcing them to travel to base camps or towns, Chipping machines convert raw cassava to which increases the cost of obtaining the inputs. cassava chips, which are then dried. There is Farmers in the Southwest also reported limited a market for cassava chips in Kampala and knowledge about agricultural inputs, including the neighboring countries of South Sudan improved seeds, fertilizers, and agrochemicals. and the DRC. Cassava chips could be further converted to cassava flour. Beyond small-scale milling in the Southwest, there is minimal value addition once agricultural  Setting up cassava mills for converting freshly produce has gone through the post-harvest harvested raw cassava to cassava flour, handling stage. Produce is generally traded which could be sold locally and outside the immediately after the harvest. There is a lack of refugee-hosting areas. Mobile cassava mills post-harvest storage facilities in both refugee and could be considered to increase the area for host communities. Farmers, especially refugees, sourcing raw material. are dependent on prices set by middlemen and generally have low negotiating power, often 3. Vegetable production because they lack information about the market Supporting the production of vegetables prices and alternative markets that can be that traditionally form part of refugee and accessed. host communities’ diets will increase the availability of nutritious food. A solar-powered A solar-powered drip irrigation drip irrigation system that uses no power and less water (compared to traditional irrigation system that uses no power methods) should be explored to grow vegetables and less water should be throughout the year. A producer supplying larger cities with vegetables and requiring additional explored to grow vegetables production could be engaged to connect hosts throughout the year. and refugees to additional markets. A solar- powered cool chain from the field to market, if developed, would increase the products’ shelf life and reduce post-harvest losses. 15 4. Poultry production Microfinance institutions could improve access Poultry is an inexpensive source of animal to formal borrowing in a context where collateral protein that can be consumed locally. Setting requirements prevent many from accessing bank up a poultry farm in refugee-hosting areas loans. Financial products targeted at farmers could be explored. This should ideally be done could support the development of local agribusi- by an existing larger poultry producer that can nesses. supervise the farm. Crops could be grown locally to supply a simple feed mill that would produce There are opportunities to extend the presence feed for the poultry farm. of commercial banks through the agent banking model. 5. Fertilizer supply Crop yields could be boosted with a supply Plastic recycling of fertilizers to smallholder farmers in refugee settlements and host communities. Farmers There is a large and vibrant retail market in the need to be supported with access to credit to Southwest and West Nile, with most businesses afford fertilizers and with practical training and selling items such as cooking oil, soft drinks, and mentoring on improved agricultural practices. other consumables. These food and beverage Setting up a fertilizer blending plant could also items are mainly packaged in plastic. Oppor- be explored, potentially in the Hoima district tunities could be explored to collect and grade where an industrial park is located. the plastic that is produced and discarded, before delivering it to recycling facilities in Uganda, thereby reducing adverse environmental impacts 6. Fish-cage farming in the refugee-hosting areas. A small-scale opportunity could support the transition from wild capture to commercial cage or pond farming in the Hoima and Arua districts. Retail trade A larger-scale opportunity could also be Despite the market’s vibrancy, retail business explored at Lake Albert (Hoima district), with core activity has yet to reach its full potential because activities including cage and pond fish farming it is largely cut off from mainstream distribution and processing, potentially supplemented with channels. Thus, opportunities could be explored fish feed and fingerling production. for food and personal care brands that offer inex- pensive products to develop retailer networks in Housing and lighting refugee-hosting areas. Investment opportunities exist to sell solar home systems and solar lanterns in both refugee Crafts production and marketing settlements and host communities. In addition, Refugees, especially those from the DRC, tend a growing market for housing made using burnt to be highly skilled at crafts (such as making bricks means there is an opportunity to involve bags, sandals, wristbands, and kitenge) and sell the private sector in providing building materials. their products locally on a small scale. Therefore, an opportunity exists to connect refugees to Financial sector opportunities markets where they can sell their products. Banks could create links to VSLAs. They could market financial products targeted at VSLAs, focusing on messages such as increased safety of savings when stored in a bank account, opportunities to earn interest and access loans, and availability of branches or agents in the vicinity to resolve problems. Banks would need to reach out to community groups and NGOs working with VSLAs in addition to advertising through traditional media. 16 1. Introduction 1.1 Background refugee policies—the Settlement Transformative U Agenda and the Refugee and Host Population ganda hosts the world’s fourth-largest Empowerment Strategic Framework—both refugee population (1.5 million as of of which support refugee self-reliance and September 2021), and the largest refugee co-existence with host communities. The population in Africa.2 The country is implementing incorporation of the Settlement Transformative one of the world’s most progressive refugee Agenda into the 2015/16–2019/20 National protection policies, which emphasizes the self- Development Plan (NDP II) has allowed for reliance of refugees and recognizes the important transformative investments in health, education, role that refugees play in the economies of their infrastructure, livelihoods, and environmental hosting districts. Uganda adopts an integrated protection. These provisions have become part approach toward the planning and delivery of of integrated response plans, such as the Health services in refugee and host community areas to Sector Integrated Refugee Response Plan 2019– avoid duplication and maximize the impact of the 2024, the Education Response Plan for Refugees available government and donor resources. and Host Communities in Uganda 2018, the Water and Environment Sector Response In alignment with the Comprehensive Refugee Plan 2019–2022, the Jobs and Livelihoods Response Framework (CRRF)/Global Compact for Integrated Response Plan for Refugees and Refugees,3 Uganda pursues a non-encampment Host Communities 2020/21–2024/25, and the approach to refugee management. The Refugee Sustainable Energy Response Plan, which is Act (2006) guarantees refugees’ fundamental under development. Furthermore, the 2021–2025 rights, including the right to participate in gainful CRRF Strategic Directions and the 2021–2022 employment; freedom of movement; the right CRRF Nation Plan of Action emphasize to property; the right to establish businesses; mainstreaming the CRRF into national planning and the right to freedom of association, such to address the long-term impact of hosting as joining trade unions. In addition, refugee refugees, the implementation of sector response families receive plots of land for housing and plans, and district-level coordination, among cultivation in designated refugee settlements. other issues. In the settlements, refugees receive monthly food rations and non-food items (essential A Private Sector Strategy for Uganda’s Refugee household items), and they can access social Response Plan was developed by the Ugandan services provided by humanitarian agencies. government in cooperation with Uganda’s The Office of the Prime Minister (OPM) is the CRRF Secretariat in 2020.4 The strategy government entity responsible for coordinating highlights the need for an enabling legal and and managing the refugee response in Uganda policy environment, designing development and is supported by various line ministries and interventions through a private sector lens, several development and humanitarian partners. greater focus on technology and innovation, Uganda also has a multistakeholder CRRF minimizing the risk of market distortion due to Secretariat, led by the OPM, that serves as a humanitarian and development assistance, and platform for strategic discussions to support the improved collaboration and knowledge between framework’s implementation. the government, development partners, and the private sector. In the lead-up to the formal launch of the CRRF in March 2017, Uganda developed two key 2 Uganda—UNHCR data portal, https://data2.unhcr.org/en/country/uga 3 CRRF was born out of the New York Declaration for Refugees and Migrants, adopted by all United Nations member states in 2016. It calls for easing pressure on refugee-hosting countries; building self-reliance of refugees, thereby lessening their dependence on humanitarian aid; expanding access to third-country resettlement; and fostering conditions for voluntary returns. It urges UNHCR to collaborate with a diverse set of partners in addition to NGOs, including the private sector. Subsequently, the UN General Assembly adopted the Global Compact for Refugees in 2018, which aims to strengthen the international response to large refugee movements and protracted refugee situations, including by promoting the self-reliance of refugees. States and other stakeholders subsequently made pledges to implement the Global Compact at the 2019 Global Refugee Forum. 4 Palladium, Private Sector Engagement for Uganda’s Refugee Response Summary Strategy, January 2020. 17 2. Methodology T he study is based on a literature review and when the study was designed in 2019. The primary data collection through quantitative geographic boundaries of the refugee-hosting and qualitative research methods. districts considered in this study are in line with the old district boundaries (before the districts’ boundaries were re-demarcated in 2018–2020). 2.1 Geographical scope As there was insufficient data on the newly formed The districts that form part of the Southwest and districts to guide the research methodology, in West Nile regions for the purposes of this study this study the Hoima district includes the newly are shown in Table 2.1a. The study attempted formed district Kikuube; the Arua district includes to have a balance between older settlements new districts Madi-Okollo, Maracha, and Terego; where refugees live in protracted situations and and the Moyo district includes the new district relatively newer ones. In addition, the selected Obongi. districts were deemed economically promising Table 2.1a: Refugee-hosting districts surveyed Regions Districts Southwest Hoima Isingiro Kamwenge Kyegegwa West Nile Adjumani Arua Moyo Yumbe The household survey was conducted in the following refugee settlements. Table 2.1b: Refugee settlements surveyed Settlements in the Southwest Settlements in West Nile • Nakivale, Isingiro district • Rhino Camp, Arua district • Kyangwali, Hoima district • Bidibidi, Yumbe district • Rwamwanja, Kamwenge district • Palorinya, Moyo district • Kyaka II, Kyegegwa district • Ayilo I and II, Nyumanzi, and Olua, Adjumani district 18 Refugees and Asylum-Seekers in Uganda Uganda Refugee Response 30 September 2021 Countries of origin Refugees per location Total refugees and SOUTH SUDAN 62% BIDIBIDI ADJUMANI 16% 15% asylum-seekers 9% KYANGWALI DR CONGO 29% NAKIVALE 9% KYAKA II 9% BURUNDI 3% RHINO 8% 1,524,352 PALORINYA 8% SOMALIA 3% KAMPALA 6% IMVEPI 5% RWANDA 1% KIRYANDONGO 5% Figure 2.1: Map showing refugee settlements in Uganda RWAMWANJA 5% PALABEK 4% OTHERS 2% ORUCHINGA 0.6% LOBULE 0.4% SOUTH SUDAN ± Bidibidi Palorinya 240,062 126,307 Lamwo Kaabong *Lobule C # C # 5,779C # C Obongi Yumbe # Palabek Kitgum C 58,542 Maracha Terego # Amuru Imvepi C # Adjumani 70,615 Madi Okollo 233,919 Gulu Pader Kotido C Agago Arua # Rhino Moroto 128,492 Nwoya Omoro Abim Zombo Pakwach Otuke Nebbi Napak Kole Oyam Kapelebyong Lira Alebtong Kiryandongo 72,833 Amuria Nabilatuk C Buliisa Apac Dokolo DEMOCRATIC # Kwania Katakwi Amudat REPUBLIC Soroti Nakapiripirit rt OF THE CONGO Kaberamaido be Hoima Al Masindi Amolatar e Ngora k La Bulambuli Serere Kween Lake Kyoga Kumi Bukedea Kikuube Nakasongola Kapchorwa C Bukwo Kyangwali # Buyende Pallisa 133,288 Kyankwanzi Nakaseke Kaliro Sironko Mbale Kakumiro Kibuku Bududa Ntoroko Kagadi Budaka Kayunga Kamuli Butaleja Kiboga Luwero Namutumba Manafwa Bundibugyo Kibaale Luuka Iganga Tororo Kabarole Kyenjojo Kassanda Bugweri Kyegegwa Mukono Jinja Bugiri Bunyangabu Mubende Busia KENYA C Mityana C # Mayuge # Kampala Buikwe Kyaka II Rwamwanja 126,833 Gomba 95,371 Buvuma Namayingo 78,760 Kasese Kamwenge Butambala Ssembabule Wakiso Bukomansimbi Mpigi Mukono Kiruhura Kalungu Ibanda Lyantonde Rubirizi Buhweju Lake Lwengo Masaka Kalangala LEGEND C Edward Bushenyi Mbarara Mitooma Lake Victoria # Refugee settlement Sheema Capital city C Rakai Rukungiri Ntungamo # Nakivale District boundary 145,206 Isingiro *Oruchinga C Kanungu Kyotera Rukiga 8,345 # Source countries of refugees Rubanda South Sudan Kisoro Kabale UNITED DR Congo REPUBLIC OF Somalia TANZANIA Burundi 50km RWANDA Rwanda Other nationalities The boundaries and names shown and the designations used on this map do not imply official endorsement or acceptance by the United Nations. *Oruchinga and Lobule symbols not to scale Population data: proGres (OPM) Author: UNHCR Representation in Uganda Feedback: ugakaimug@unhcr.org For more info, please visit: www.ugandarefugees.org Source: UNHCR, September 2021 19 2.2 Household survey sample. These nationalities are Eritrea, Ethiopia, and Rwanda for the Southwest, and Sudan for A total of 2,522 households were interviewed for West Nile. the study. Given the differences in characteris- tics of the target populations in the two regions, the study adopted a two-stage stratified cluster To collect the data, 35 enumerators were sampling design, with the two regions forming recruited by ASIGMA, all with knowledge of the the strata. The first stage of sampling involved the local context, experience collecting data, and random selection of enumeration areas using the the ability to communicate in at least three host probability-proportionate-to-size method (also and refugee languages in addition to English. called the PPS method), whereas the second In each region, enumerators were supervised stage involved selecting households within each by two field supervisors. The quality of the enumeration area using a sampling interval. data was monitored daily using a dashboard tracking each enumerator’s performance across a set of indicators checking for interview validity, The household survey relied on two sampling duration, and skip patterns, and detecting frames. The survey of the host communities potential inconsistencies. Randomly selected benefited from the master sample of enumeration audio recordings of a subset of questions were areas covering the whole territory of refugee- also monitored to guarantee the data being hosting districts, obtained from the Uganda recorded on the tablets was consistent with the Bureau of Statistics (UBOS), while enumeration responses provided by households. areas for the refugee survey were created with the aid of geospatial data provided by UNHCR. 2.3 Business survey Only respondents aged 16 and above were The business survey sampled owners with and considered eligible for an interview. Enumerators without dedicated business premises. Given targeted household heads or household members the absence of a sampling frame inclusive of in charge of household purchasing decisions. informal businesses, the business respondent In this context, the head of a household was sampling was done in the field. Thus, the findings defined as the person who manages the income are indicative rather than representative of the earned and expenses incurred by the household, business environment. The data collection and who is the most knowledgeable about teams identified businesses at trading centers5 other members and the general affairs of the in refugee settlements and host communities, household. or adopted a snowballing approach based on recommendations by key informants, while To ensure representation of the countries of attempting to spread the sample across different origin in the sample, four enumeration areas business types and geographic locations. The with minority nationalities (defined as those survey teams interviewed 404 business owners making up less than 1 percent of the total or people with in-depth knowledge about the refugee population) were included in the final business, as shown in Table 2.3. Table 2.2: Household sample size targets and achievements Region Population group Target sample size Sample size achieved Southwest Host community 756 777 Refugees 462 469 West Nile Host community 777 824 Refugees 441 452 Total 2,436 2,522 5 A trading center is an area where economic activity is concentrated. It usually features business outlets clustered next to each other. In Ugandan refugee-hosting areas, most boda-boda (motorbike) taxis and larger businesses operate from within trading centers. 20 Table 2.3: Business sample size targets and achievements Region Geographic area Target sample size Sample size achieved Southwest Host community areas 90 101 Refugee settlements 90 100 West Nile Host community areas 90 120 Refugee settlements 90 83 Total 360 404 Table 2.4: Number of focus group discussions Region Population group Farmer FGDs Village savings and loan Youth FGDs associations FGDs Southwest Host community 2 2 2 Refugees 2 2 2 West Nile Host community 2 2 2 Refugees 2 2 2 2.4 Qualitative data collection All graphs in the report show weighted data, excluding “Don’t know” and “Refuse to answer” The team conducted 37 key informant interviews option choices. Sample sizes are indicated in with development partners, local government each graph. representatives, and other stakeholders. The team also conducted 24 focus group discussions with farmers, youths, and village savings and 2.6 Methodology to estimate loan associations (VSLAs), to explore economic household consumption activities, youth employment, and financial expenditure literacy and saving tendencies. The distribution Chapter 5 presents estimates of the total annual of the conducted focus groups is shown in household consumption expenditure for refugee- Table 2.4. hosting areas within the Southwest and West Nile regions, which were derived using the average 2.5 Data weighting, statistical monthly household expenditures (pre-COVID-19 significance, and data and after the lockdown was lifted), as reported by household survey respondents. A country-wide visualization lockdown was imposed on March 31, 2020, and To correct for refugee minorities’ oversampling lasted until July 31, 2020. (see section 2.2), sampling weights were com- puted by considering the minority nationalities To estimate the total annual consumption (described in section 2.2) as one group. The expenditure for each refugee settlement and sampling weight for the minority nationalities the corresponding host community in the two was computed by dividing the known proportion regions, the following steps were taken: of minorities in the refugee population (from UNHCR data) by the proportion of minorities 1. Average monthly household expenditure achieved in the sample. figures for both pre- and post-COVID-19 lockdown were collected through the survey Population estimates reported for the household and cleaned to remove outliers. Weights were survey are statistically significant at a 95 percent applied to account for the oversampling of level, with a margin of error of +/-4 percent at the minority nationalities. regional level. 21 2. The total number of households for each Energy for cooking was not included in the com- refugee settlement was obtained from putation of the market size, as it was not part of UNHCR’s Settlement Refugee Statistics the survey. Update, February 2021. Unless otherwise stated, all estimated annual 3. The total number of households in each expenditure figures were calculated based on refugee-hosting district was obtained from the household expenditure after the lockdown the UBOS National Population and Housing was lifted (on July 31, 2020). Census 2014. All non-refugee households were considered part of the host community. 2.7 Study limitations 4. As the data was collected in Ugandan shillings, The study was conducted during the COVID-19 the exchange rate to the US dollar was consist- pandemic. The fieldwork took place from ently applied at $1 = 3,750 Ugandan shillings. November 2020 to January 2021. When the 5. The total monthly expenditure for a settlement country-wide lockdown was lifted on July 31, and the host community was then derived as 2020, a curfew was still in place, starting from the product of the average monthly house- 7 pm every day. Although the study solicited hold expenditure and the corresponding information on pre-COVID-19 and post-lockdown total number of refugee and host community employment, income, and expenditure, there is households. no conclusive evidence of the long-term impli- cations of the pandemic on these key indicators In the two target regions, there were refugee based on the data collected by this survey. settlements that were not surveyed: the Imvepi Market sizes have been conservatively estimated settlement in the Arua district and the Oruchinga based on the post-lockdown figures. settlement in Isingiro. Given that the household survey mainly inter- For the non-surveyed settlements, it was viewed household heads (81 percent of the assumed that the expenditure figures collected sampled respondents are household heads), through the household survey are representative findings can be considered representative when by region and by nationality. Hence, the following the household is the unit of analysis. In cases additional steps were taken: where the data refers to respondents (such as education level, employment, and mobile phone 6. The average monthly household expenditure ownership), findings are more representative of by nationality in each region was calculated. household heads than of the general population. 7. To obtain the total expenditure in non- Given the absence of a sampling frame to allow surveyed settlements, for each nationality, for random sampling of business respondents, these calculated averages were multiplied the findings from the business survey are not by the total number of households of that intended to be representative of the whole nationality in each non-surveyed refugee business community in the two regions. settlement (with data obtained from UNHCR’s Nonetheless, the team mitigated this challenge Settlement Refugee Statistics Update, by spreading the sample geographically and February 2021), and then added together. including different types of businesses, to enable the survey to capture a more accurate picture of 8. The total monthly expenditure in non- the local business community. surveyed settlements was added to the total expenditure for surveyed refugee areas and host communities derived earlier. Multilingualism posed a challenge to data collection. The team mitigated this issue by 9. The figure was multiplied by 12 to obtain hiring local enumerators who were fluent in the the total annual household consumption most prevalent refugee and host community expenditure. languages. In addition, enumerators were trained to clarify the key questionnaire terms in the languages the respondents could understand. 22 23 3. Socioeconomic Characteristics of the Southwest and West Nile 3.1 Overview of the refugee Figure 3.1 shows the refugee population trajec- situation in Uganda tory from 2000 to April 2021. A slight reduction in R numbers occurred in 2018 because UNHCR and efugees in Uganda live in protracted and the Ugandan government carried out biometric in emergency settings. UNHCR defines verification of all refugees in the country.7 a protracted refugee situation as one in which 25,000 or more refugees from the same According to UNHCR data, as of September nationality have been in exile for five or more 2021, the majority of refugees are from South years in their asylum country. In contrast, an Sudan (61.2 percent) and the Democratic emergency situation is one where the immediate Republic of the Congo (DRC) (29 percent). Other focus is on lifesaving protection and assistance.6 Figure 3.1: Uganda’s refugee population, 2000 to April 2021 2,000,000 1,500,000 1,000,000 500,000 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2013 2015 2016 2017 2000 2012 2014 2018 2019 2020 APR-21 Source: UNHCR Uganda Refugee Portal 2020–2021; Macrotrends data, Uganda Refugee Statistics 2000–2019 6 https://www.unrefugees.org/news/protracted-refugee-situations explained/#What%20is%20a%20protracted%20refugee%20situation? 7 UNHCR 2018, https://www.unhcr.org/afr/news/press/2018/10/5bd72aad4/opm-and-unhcr-complete-countrywide-biometric-refugee-verification-exercise. html#:~:text=KAMPALA%3A%20The%20Office%20of%20the,seekers%20and%20refugees%20in%20Uganda.&text=This%20represents%2075%25%20of%20 the,start%20date%20in%20March%202018 24 refugees hosted are from Burundi (3.4 percent), to accommodate refugees fleeing South Sudan Somalia (3.2 percent), Rwanda (1.4 percent), in large numbers. Refugees make up a large Eritrea (1.2 percent), Sudan (0.2 percent), and majority (71 percent) in the newly created district Ethiopia (0.2 percent). West Nile mostly hosts of Obongi (formerly Moyo), where the Palorinya South Sudanese refugees; Uganda experienced refugee settlement is located. The districts that a sharp increase in the refugee population from host refugees are shown in Table 3.1. 2013 as a result of a civil war in South Sudan that broke out after the country gained indepen- Typically, refugees will settle or be settled with dence.8 The Southwest mostly hosts refugees refugees of the same country of origin, leading from the DRC who flee from ethnic violence and to the informal renaming of trading centers general insecurity in their country of origin.9 according to the largest refugee population (such as New Congo and Little Mogadishu in Nakivale). Refugees mainly live in designated settlements, The host community and refugees generally live although there is a sizeable population of urban in harmony, with community-based initiatives refugees in Kampala. Of the districts covered such as savings groups attracting membership in this study, Adjumani and Yumbe in West Nile from both hosts and refugees.10 Previously host the highest numbers of refugees. Generally, conducted studies on the refugee situation in host community members outnumber refugees Uganda have also found that the relationship at the district level. However, in Adjumani, half of between refugees and the host community is the population are refugees. The district has the generally amicable, which is further encouraged highest number of refugee settlements in Uganda by marriages between Ugandans and refugees. (18 settlements), both older ones (dating back However, COVID-19-related shocks have placed to the 1980s) and ones created more recently pressure on both refugees and host com- to host South Sudanese refugees. The Yumbe munities, which, coupled with reducing refugees’ district hosts the world’s second-largest refugee food rations, may present future challenges. settlement, Bidibidi, which was opened in 2016 Table 3.1: Population of refugee-hosting districts in Uganda District Host population Refugee population Refugee population as a (bold districts were surveyed) percentage of the total district population Adjumani 237,400 233,919 50% Isingiro 616,700 153,551 20% Kampala 1,709,900 95,371 5% Kamwenge 475,600 78,760 14% Kikuube (formerly Hoima) 376,600 133,288 26% Kiryandongo 322,300 72,833 18% Kyegegwa 475,600 126,833 21% Koboko 267,700 5,779 2% Lamwo 145,400 58,542 29% Madi-Okollo and Terego 770,600 199,107 21% (formerly Arua) Obongi (formerly Moyo) 50,300 126,307 72% Yumbe 699,300 240,062 26% 8 Council on Foreign Relations, https://www.cfr.org/global-conflict-tracker/conflict/civil-war-south-sudan 9 World Bank, Informing the Refugee Policy Response in Uganda: Results from the Uganda Refugees and Host Communities 2018 Household Survey, 2019, p.13. 10 UNHCR and World Bank, An Assessment of Uganda’s Progressive Approach to Refugee Management, 2016, p.28. Available at: https://openknowledge.worldbank. org/bitstream/handle/10986/24736/An0assessment00o0refugee0management.pdf?sequence=1&isAllowed=y 25 Traveling to and within a refugee settlement Refugee settlements normally have a base camp, which is the settlement administrative center, hosting the OPM office, UN agencies, and NGOs. Trading centers with economic activities such as shops, markets, and restaurants also tend to be concentrated at base camps. The OPM and UNHCR jointly administer the settlements. The settlement commandant is the highest government official in refugee settlements, reporting to the OPM. Outside visitors to a settlement, such as individuals from private firms, need to obtain a letter from the settlement commandant authorizing entry to the settlement. The road network connecting major towns to refugee settlements is generally poorly maintained, with roads being dusty in the dry season and under water in the rainy season. Air transportation is limited; only the Rhino Camp in Arua (West Nile region) is close to an airfield. The ongoing construction of the Hoima International Airport in Kabaale and roads to oil wells in the Albertine Graben region will connect Kyangwali settlement in the Southwest to major towns. The World Bank-supported Koboko-Yumbe-Moyo road corridor project (West Nile region) is also expected to improve road connectivity in the three districts.11 For travel within refugee settlements, motorcycle taxis (locally known as boda-bodas) are the most commonly used means of transport. People typically use these to visit family and friends, and to go to markets, hospitals, and other locations that are too far to be accessed on foot. Most boda- bodas are clustered in trading centers and other places with economic activity. Settlements in the Southwest have minivan taxis that transport people to host district towns and to Kampala. The most vibrant taxi network is in Nakivale, followed by Rwamwanja, Kyaka II, and Kyangwali. 11 The project was approved in September 2020 and will be completed in December 2025. 26 3.2 Demographic characteristics On average, there are 6.5 people per household. of refugees and the West Nile households are larger, comprising host community eight people on average in refugee settlements and seven people in host community areas. 3.2.1 Household composition The Southwest averages 5.5 individuals in each A large majority of the study sample is household in both refugee settlements and host composed of female respondents. In both the communities. Southwest and West Nile, the share of female respondents reaches 65 percent. The trend is The youth make up a large share of the observed in both refugee settlements and host population. In refugee settlements, 18–34-year- communities. Three potential reasons for the olds make up 54 percent of the sample over-representation of females in the sample (representative of household decision-makers), are: (i) only female members of the household while in host communities they make up were home at the time of the interview; (ii) 45 percent of the sample. The overall average female members were more eligible to answer age is 38, spanning from 35 in West Nile refugee the questionnaire based on their knowledge of settlements to 40 in Southwest host communities. the household’s general expenditure; and (iii) However, respondents were selected based female-headed households are more prevalent on their knowledge of the household’s general in the refugee settlements. Fifty-nine percent expenditure and consumption habits; as such, of household heads in the sample are female the average age is not representative of the (68 percent among refugees and 54 percent in general population. Figure 3.2 shows the sampled the host community). respondents’ age distributions by region and population group.12 Figure 3.2: Age distribution of respondents How old are you? Southwest host community Southwest refugees 60 and above 60 and above Female (n = 484) Female (n = 325) 35-59 35-59 Male (n = 284) Male (n = 141) 18-34 18-34 25% 5% 15% 35% 25% 5% 15% 35% West Nile host community West Nile refugees 60 and above 60 and above Female (n = 495) Female (n = 339) 35-59 35-59 Male (n = 328) Male (n = 110) 18-34 18-34 25% 5% 15% 35% 25% 5% 15% 35% 12 Only respondents aged 16 and above were interviewed. The percentage of 16–18-year-olds is negligible and is not shown in the chart. 27 3.2.2 Countries of origin Sudan). In the Southwest, the Banyabwisha In West Nile, most surveyed refugees (98 percent) (28 percent), the Kongo (13 percent), the Somali come from South Sudan, whereas the majority of (10 percent), the Hutu (11 percent), and the Bage- the surveyed refugees (79 percent) in the South- gere (8 percent) are the most prevalent ethnicities west are from the DRC. The refugee settlements among sampled refugees.13 in the West Nile region also host Sudanese (from the Republic of Sudan) populations (making up Host community members speak a variety 1.3 percent of the surveyed refugees) and smaller of languages, mirroring the distribution of minorities from Burundi. In the Southwest, ethnicities. Lugbara is spoken by a majority minorities include Somalis (13 percent), Burun- (58 percent) in West Nile, while Runyankore/ dians (3.8 percent), Rwandans (1.2 percent), as Rukiga has the largest percentage of speakers well as smaller proportions of Ethiopians, South (66 percent) in the Southwest. Just under a third Sudanese, and Eritreans. A small percentage of of host community members are able to speak Ugandans also live inside refugee settlements’ English, whereas 11 percent speak Swahili— boundaries, as shown in Figure 3.3. Uganda’s official languages (with more speakers of either of the two in West Nile than in the Southwest). 3.2.3 Ethnicity and languages spoken Both refugee settlements and host community Juba Arabic and Swahili are the most spoken areas are ethnically and linguistically diverse. languages in West Nile and Southwest refugee The host community in West Nile is mainly com- settlements respectively. In West Nile, refugees posed of the Lugbara (47 percent), the Aringa (who are mostly from South Sudan) speak (24 percent), and the Madi (22 percent), while Juba Arabic (62 percent), followed by Kakwa the host community in the Southwest mostly (47 percent), English (27 percent), and Koku comprises the Banyankore/Bakiga (55 percent) (19 percent). In the Southwest, 74 percent and the Banyoro/Batooro (28 percent). In West speak Swahili—the language mostly spoken by Nile refugee settlements, the Dinka (28 percent), refugees from the DRC and Burundi—followed the Luo (22 percent), and the Kakwa (12 percent) by Kinyarwanda (14 percent). Only 3 percent of are the most prevalent ethnicities (all from South refugees speak English in the Southwest. Figure 3.3: Country of origin of respondents from refugee settlements What is your country of origin? Southwest West Nile DRC 79.8% South Sudan 98.3% Somalia 13.4% Sudan 1.3% Burundi 3.8% Uganda (host) 0.2% Rwanda 1.2% Uganda (host) 0.7% Ethiopia 0.7% South Sudan 0.2% Eritrea 0.1% Refugees (n = 468) Refugees (n = 452) 13 The breakdown by ethnicity is based on self-identification by refugees. As such, it may not correspond to existing classifications. 28 3.3 Education five times higher than that of uneducated males, while in the Southwest, it is 2.2 times higher. The household survey shows low levels of These gender differences are slightly larger in education overall—26 percent of household refugee settlements than in host communities. decision-makers have had no schooling at all, and only 34 percent completed primary school. Overall, refugees are more likely to have no The data also reveals notable differences in schooling than members of the host community. education levels between nationalities and eth- In the Southwest, 44 percent of refugees have no nicities of refugees. Forty-four percent of DRC education, compared to 20 percent of the host refugees and 34 percent of South Sudanese community. In West Nile, 34 percent of refugees refugees have no schooling. Among minorities, have no schooling, compared to 18 percent of the 52 percent of Somalis and 27 percent of Burun- host community, as shown in Table 3.2. A large dians have no education. There are differences majority of the population has incomplete primary between South Sudanese ethnicities: the Dinka education. In the Southwest, only 8 percent of have the lowest educational attainment, with refugees and 14 percent of the host community 67 percent without schooling, compared to completed primary education. In West Nile, 14 percent of the Luo. 10 percent of refugees and 12 percent of the host community completed primary education. 3.4 Employment Host community members are significantly There is a significant gender gap in educational more likely to be employed than refugees, attainment. Overall, 33 percent of women have and Southwest residents are more likely to be no schooling, compared to 11 percent of men. employed than those living in West Nile. In host Only a quarter of women completed primary communities, the employment rate (including school, in strong contrast to 51 percent of men. employees, business owners, and farmers) is at The gender gap is considerably wider in West 80 percent in the Southwest and 60 percent in Nile, where the share of uneducated women is West Nile. In refugee settlements, the employ- Table 3.2: Respondents’ levels of education Education level Southwest Southwest Southwest West Nile West West Nile total host refugees total Nile host refugees community (n = 408) community (n = 440) (n = 820) (n = 834) No schooling 28% 20% 44% 23% 18% 34% Some primary 37% 41% 29% 44% 47% 38% school Primary school 12% 14% 8% 12% 12% 10% completed Some secondary 8% 9% 6% 10% 10% 10% education (incomplete O-level) Completed O-level 8% 10% 4% 6% 7% 6% Secondary school 3% 3% 3% 1% 1% 1% completed (with A-levels) University degree 1% 1% 2% 1% 2% 1% completed Vocational/ 3% 4% 1% 2% 3% 1% technical training 29 ment rate is much lower, at 51 percent in the and tending the garden. However, the rates of Southwest and 22 percent in West Nile. Overall, business ownership (of non-agricultural enter- in the Southwest, 70 percent of respondents are prises) are about the same for males and females employed, compared to only 47 percent in West (13 percent and 14 percent respectively in the Nile.14 A large portion of the population is outside host community and 6 percent and 8 percent of the labor force (not employed and not looking respectively among refugees). for work), especially in West Nile, although most of this number are women doing unpaid house- The unemployment rate of the youth work. (18–34-year-olds) among the respondents is much higher in refugee settlements (65 percent) Among the economically active population, than in host communities (34 percent). There agricultural work is the most common are also regional differences: in the Southwest, occupation in the Southwest, while owning a less than a third of the youth is unemployed, business is a prevalent occupation in West Nile. compared to 59 percent in West Nile. According The distribution of employment type between to focus group discussions with young people refugees and host communities is similar overall, in the Kyangwali (Southwest) and Bidibidi (West although the portion of agricultural workers (mostly Nile) refugee settlements, the reasons for unem- farmers) is higher in host communities (33 percent) ployment are low levels of education and skills, than in refugee settlements (24 percent). There limited networks, and low demand for certain are, however, significant regional differences: in services they have the skills for, such as welding West Nile, business owners constitute more than and construction. a third of those who are employed, compared to 8 percent in the Southwest. The Southwest Refugees from the DRC and Burundi are more has a much larger share of agricultural workers likely to be employed than Somalis and South (45 percent) than West Nile (11 percent).15 Better Sudanese, differences that can be partly farming opportunities in the Southwest seem attributed to opportunities available in the to explain the higher employment rate in that respective regions. Fifty-six percent of refugees region compared to West Nile. Non-agricultural from the DRC are employed, as are 56 percent occupations mainly include teaching (mostly of Burundians, compared to just 27 percent of among host communities), and to a much lesser Somalis and 23 percent of South Sudanese. extent, working as company/NGO employees, However, the same differences can be observed boda-boda (motorbike taxi) drivers, construction between the two regions hosting the respective workers, waiters, translators, duuka employees, populations, which means that the differences in and others. employment are likely to depend on the oppor- tunities available (especially in agriculture) rather The gender gap observed in educational attain- than on nationality alone. The rate of business ment is reflected in the labor market. The unem- ownership among South Sudanese and DRC ployment rate for females reaches 50 percent, refugees is similar, at 9 percent and 7 percent against 25 percent for males. The gender respectively, but is slightly lower than in the host difference is higher in refugee settlements than communities, where business owners make in host communities and is higher in West Nile up 13 percent of respondents in both regions. than in the Southwest. More women in refugee The percentage of women engaged in unpaid settlements identify themselves as housewives housework is particularly high among the South (48 percent) than those in the host communities Sudanese, at 49 percent, compared to 17 percent (31 percent). They engage in everyday household among DRC refugees. Finally, the prevalence of tasks such as cooking, taking children to school, agricultural workers is notable among DRC and and fetching water. For women with access to Burundian refugees, but less so among South land, activities may also include cultivating crops Sudanese or Somalis.16 14 The seasonality of informal agricultural work may affect employment figures, especially for West Nile. 15 Likewise, the percentage of people engaged in agriculture is likely to oscillate because agricultural work is seasonal. 16 Low engagement in agricultural activities among these two nationalities is likely because the Dinka (who make up the majority of South Sudanese refugees in the sample) and Somalis traditionally engage in pastoralism instead of agriculture in their countries of origin. Regional differences can also help explain this phenomenon—agriculture is more prominent in the Southwest than in West Nile. 30 Figure 3.4: Respondents’ employment status What is your employment status? Southwest West Nile Agricultural worker 39% Agricultural worker 8% 17% 1% Independent worker 23% Independent worker 22% (non-agricultural) 13% (non-agricultural) 8% Outside of the labor force 16% Outside of the labor force 30% 25% 59% Working for someone as an 13% Working for someone as an 9% employee (non-agricultural) 15% employee (non-agricultural) 4% Unemployed, 2% Unemployed, 5% looking for work 18% looking for work 11% 6% 22% Business owner 6% Business owner 9% Other 2% Other 4% 6% 9% Host community (n = 787) Refugees (n = 408) Host community (n = 838) Refugees (n = 442) Three-quarters of respondents who work for to additional land by renting it from the host someone as an employee would rather work community.19 A larger percentage of people are for themselves if they could. The most common engaged in agriculture in the Southwest as land activities that respondents would be willing to tends to be more fertile there relative to West undertake were opening a duuka (45 percent) Nile. Moreover, Southwest land around refugee and working as a farmer (30 percent). Over settlements is government-owned in contrast to 90 percent of respondents to this question more complex communal land titles in West Nile. mentioned lack of capital as an issue preventing Less than half of refugees rear livestock for their them from starting their own business, while a own consumption—an activity that tends to be significantly smaller percentage also mentioned more land intensive. 20 lack of technical skills and lack of an available space to rent. 3.5 Income Refugees are more likely to engage in 3.5.1 Income levels and sources subsistence than commercial farming because In both regions, host community members of the small plot sizes allocated to them. Refugees earn significantly more than refugees. Average living in settlements generally have access to household earnings in the host communities land (which is usually the land surrounding their have been twice as high as those of refugee shelters), but it is often insufficient to feed a households ($61 versus $32) since the lockdown household.17 The plot size varies depending on was lifted at the end of July 2020. Household the settlement. In West Nile, where access to earnings have been significantly affected by land is more challenging than in the Southwest, the ongoing pandemic, although they have the average refugee plot size is less than half an also been recovering since the lockdown acre per household, while hosts have on average was lifted. Compared to the period before two acres per household.18 Access to land has the COVID-19 outbreak, average earnings become more difficult since 2016 because of an reduced by 15 percent in both regions in the influx of new arrivals. Some refugees gain access period immediately following the lockdown. 17 Norwegian Refugee Council (NRC), Owned Spaces and REACH, Owned Spaces and Shared Places: Refugee Access to Livelihoods and Housing, Land and Property in Uganda, September 2019, p.25. 18 Food and Agriculture Organization of the United Nations, Food Security, Resilience and Well-Being Analysis of Refugees and Host Communities in Northern Uganda, 2018, p.13-14. 19 A study by NRC and REACH in Adjumani found that few refugees who rent land possess formal rental agreements. NRC and REACH, Refugee Access to Livelihoods, and Housing, Land and Property in Uganda, May 2019. 20 NRC and REACH, Refugee Access to Livelihoods, and Housing, Land and Property in Uganda, May 2019, p.75. 31 Before the pandemic, monthly household household’s monthly income. The extra income earnings were higher on average in West Nile is usually received monthly (for 44 percent of ($70) compared to the Southwest ($58). The households), every two months (25 percent), or survey found the highest post-lockdown earnings weekly (10 percent). 21 among host community households in West Nile ($64) and the lowest among West Nile refugees 3.5.2 Influence of demographics ($31). on individual earnings Level of education: Overall, educational attain- Besides employment, humanitarian aid and ment appears to have a statistically significant remittances contribute to household incomes. influence on individual earnings. On average, Humanitarian aid to refugees in the Southwest has individuals with no schooling earn $22 per largely transitioned to cash-based assistance. month, against $27 for those with primary edu- As such, 96 percent of refugee respondents in cation. Both categories were similarly affected the region declared receiving income from cash by the pandemic (with reductions of 33 percent transfers. Conversely, a much smaller share of in individual earnings). The sample of those with West Nile refugees mentioned cash transfers as higher levels of education is too small to draw their income source (29 percent) due to the use conclusions. of food-based rations. On the other hand, more than half of refugees in West Nile (54 percent) Nationality: Disparities in income are more resell their rations to gain additional income. prevalent in the Southwest, where refugees from Remittances from within and outside of Uganda Burundi and Somalia earn a higher income on represent an income source for both refugees average than other nationalities.22 In the case and host communities. Eighteen percent of of Somalis, this may be attributed to com- refugees in the Southwest and 10 percent in munity-based revolving funds and a high level of West Nile receive remittances. A sizeable portion remittances, which can be used as a source of of the host community in West Nile also receives financial capital. 23 remittances, with 22 percent mentioning them as an income source, compared to only 5 percent of host communities in the Southwest. On average, cash transfers and remittances add $33 to a Figure 3.5: Impact of COVID-19 on monthly household earnings24 How much did you earn per month on average before the COVID-19 outbreak, during lockdown, and after lockdown? Southwest West Nile $73 $65 $64 $56 $50 $44 $46 $40 $33 $31 $23 $26 Host community Refugees Host community Refugees (n = 776) (n = 468) (n = 824) (n = 452) Pre-COVID-19 During lockdown Post-lockdown 21 Amounts received and frequencies are not significantly different between regions and type of additional income, and between refugees and host communities. 22 The sample of Rwandan, Ethiopian, and Sudanese refugees is quite limited, making it difficult to draw conclusions. 23 Although the sample of Somali refugees is limited, they made up the largest share of remittance recipients in the sample. 24 Household earnings were derived from asking questions about the individual earnings of each household member. The questions were formulated as follows: “How much did you earn per month on average before the COVID-19 outbreak, during lockdown, and after lockdown?” and “How much on average did he/she [household member] earn per month before the COVID-19 outbreak, during lockdown, and after lockdown?” 32 Figure 3.6: Impact of COVID-19 on individual earnings What were your average monthly earnings before COVID-19, during lockdown, and after lockdown? Southwest $37 $37 $31 $29 $29 $24 $24 $19 $22 $17 $18 $14 $16 $12 $13 $5 $7 $4 Uganda Rwanda Burundi DRC Ethiopia Somalia (host community) (n = 19) (n = 59) (n = 321) (n = 10) (n = 54) (n = 773) Pre-COVID-19 During lockdown Post-lockdown West Nile $58 $45 $37 $25 $26 $25 $23 $15 $16 Uganda South Sudan Sudan (host community) (n = 426) (n = 22) (n = 818) Pre-COVID-19 During lockdown Post-lockdown 33 4. Access to Telecommunication and Financial Services 4.1 Telecommunication services West Nile have mobiles.26 These figures may O be higher than the mobile phone penetration wnership of mobile phones and SIM rates in the general population, given that the cards is high in the two regions. In the survey interviewed household decision-makers. Southwest, mobile phone owners make Refugees can register a SIM card on par with up 66 percent of the overall population (which is Ugandan citizens, as long as they can produce higher than the national average of 59 percent),25 a refugee identification card or attestation letter whereas in West Nile they make up 52 percent. from the OPM. 27 A slightly higher percentage About 60 percent of refugees in the Southwest of SIM-card owners relative to phone owners and about 70 percent of refugees in West Nile own indicates that some mobile phone users may be a mobile phone, whereas in the host community, borrowing a phone from others. 70 percent in the Southwest and 52 percent in Figure 4.1: Ownership of mobile phones and SIM cards Do you own a mobile phone? Southwest West Nile Host Host community 70% community 52% (n = 818) 30% (n = 836) 48% Refugees 59% Refugees 69% (n = 422) 41% (n = 440) 31% Yes No Yes No Do you own a SIM card? Southwest West Nile Host Host community 73% community 58% (n = 817) 27% (n = 835) 42% Refugees 61% Refugees 72% (n = 425) 39% (n = 440) 28% Yes No Yes No 25 Data Portal, 2020, Digital 2020: Uganda. Available at: https://datareportal.com/reports/digital-2020-uganda#:~:text=There%20were%2026.83%20million%20 mobile,60%25%20of%20the%20total%20population 26 The analysis did not find any statistically significant differences in phone ownership by gender. 27 Uganda Communications Commission, Telecoms Operations Guidelines on SIM Card Registration 2020. Available at: https://www.ucc.co.ug/wp-content/ uploads/2020/12/All-Telecoms.Operational-Guidelines-on-simcard-registration.pdf 34 Owning basic (button) phones may prevent MTN, Airtel, and Africell29 are the three major Internet access for a significant proportion of mobile network operators in the refugee- refugees and hosts. 28 Among mobile phone hosting areas. In both the Southwest and West owners, over two-thirds of the population use Nile, most mobile phone users use the MTN basic phones, instead of phones that are Internet- network, which also has the highest country- enabled (smartphones and most feature phones). wide penetration. Among West Nile SIM-card Consequently, mobile Internet use is low, and so owners, 97 percent of the host community and is overall Internet use. (Of respondents who use 98 percent of refugees have MTN SIM cards. In the Internet, 96 percent reported accessing it the Southwest, 84 percent of host SIM owners on their phones.) In West Nile, only 4 percent of and 67 percent of refugee SIM owners have refugees and 9 percent of the host community MTN SIM cards. Airtel also has a significant use the Internet (8 percent overall), with the use number of customers, as shown in Figure 4.3.30 rate comparatively higher in the Southwest— MTN has made it easier for refugees to register 13 percent among refugees and 9 percent among a SIM card, having established tents in some members of the host community (11 percent settlements and sent agents to others. Africell overall). The cost of mobile data bundles is also (which is withdrawing from Uganda) has been a likely barrier. popular in the Congolese community, who find the network most convenient for communicating with relatives in the DRC, where Africell also operates. Figure 4.2: Phone types owned by respondents What kind of phone do you have? Southwest West Nile 70% 73% Basic phone Basic phone 63% 73% Feature phone 19% Feature phone 17% 16% 17% Smartphone 11% Smartphone 11% 11% 15% Host community (n = 482) Refugees (n = 183) Host community (n = 667) Refugees (n = 289) Figure 4.3: Mobile network operators’ penetration Southwest West Nile 84% 97% MTN MTN 98% 67% Airtel 42% Airtel 30% 9% 6% 1% 5% Africell 28% Africell 4% Host community (n = 568) Refugees (n = 293) Host community (n = 478) Refugees (n = 326) 28 A smartphone has advanced functions similar to a personal computer, such as an updatable operating system, a touch screen, and the ability to install and run downloaded apps to send emails, browse the web, edit photos, and so on. A feature phone can access the Internet, take photos, and store and play music, but lacks the advanced functionality of a smartphone; it may have a small display and an alphanumeric or QWERTY keypad. A basic (or button) phone has no Internet access; it has limited functions such as texts and calls. 29 Africell has announced that it is withdrawing from Uganda in October 2021 (see http://www.news.cn/english/africa/2021-09/08/c_1310174074.htm). 30 Some respondents subscribe to more than one network provider, which explains why the sum is higher than 100 percent for most categories. 35 4.2 Financial services mobile owners and 80 percent of host com- munity mobile owners use their phones for mobile 4.2.1 Bank and mobile money money transfers. In the Southwest, 88 percent account ownership of host community mobile owners use mobile Humanitarian cash transfers to refugee money; however, only 48 percent of refugees who settlements have helped broaden refugees’ own a mobile phone also use mobile money. This financial inclusion and attract private actors to puts the overall mobile money penetration rate settlements.31 NGOs and UN agencies, which in the Southwest at 30 percent among refugees have implemented cash transfer programs, and 64 percent among host communities, and primarily collaborate with MTN and Airtel to in West Nile, 55 percent among refugees and disburse bulk payments (for example, Mercy 46 percent among host communities. There are Corps and International Rescue Committee in only minor gender differences in mobile money Bidibidi), whereas the World Food Programme use: 71 percent of female mobile phone owners (WFP) collaborates with two banks. Beneficiaries and 81 percent of male mobile phone owners in can cash out their transfers from a mobile money the Southwest use mobile money, and 76 percent agent in the settlement. The WFP partners with and 80 percent respectively do so in West Nile. Equity Bank and PostBank. The former has implemented agent banking in settlements to About a quarter of respondents in the two enable beneficiaries to cash out,32 whereas regions own a bank account, with a higher PostBank uses a security company to deliver percentage among refugees in the Southwest. physical cash to designated distribution points, Two-thirds of refugees in the Southwest own with beneficiaries cashing out from their prepaid a bank account, which is a consequence of cards.33 Some refugee-led organizations, such bank-facilitated humanitarian cash transfers by as the Moral Brotherhood and Neighbourhood the WFP. In contrast, bank account ownership by (MOBAN) SACCO in Nakivale and the Youth refugees in West Nile is low, at only 12 percent. Initiative for Development in Africa (YIDA) in Kyaka For the host community, the data on bank II, are working with banks as agents to facilitate account ownership is comparable between the deposits and withdrawals of cash transfers. two regions, at 18 percent in the Southwest and 20 percent in West Nile. Most host community Among financial services, mobile money has respondents have a bank account with Centenary the highest coverage, especially in the South- Bank (55 percent of account owners), while most west host community (64 percent) and among refugees have their accounts with Equity Bank West Nile refugees (55 percent). In West Nile, (42 percent) or PostBank (50 percent). Six percent the use of mobile money services is high among of refugee account owners, mainly in West Nile, mobile phone owners (77 percent of refugee also have a Centenary Bank account. Figure 4.4: Usage of phone/SIM card for mobile money transfers Do you use your phone/SIM for mobile banking or mobile transfer? Southwest West Nile Host Host community 88% community 80% 12% 20% (n = 595) (n = 487) Refugees 48% Refugees 77% (n = 251) 52% (n = 317) 23% Yes No Yes No 31 https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2017/11/Humanitarian-Payment-Digitisation.pdf 32 The WFP is the largest disburser of cash to refugees. 33 GSMA, 2020. Available at: https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2020/01/Uganda_Mobile_Money_CVA_Case_Study_Web_Spreads.pdf 36 Men are more likely to own a bank account than women, and this discrepancy is greater A significant percentage in West Nile than the Southwest. In West Nile, of those who do not yet men are twice as likely to own a bank account possess a bank account than women. On the other hand, in the South- west, 31 percent of women own a bank account reported willingness to compared to 40 percent of men. Bank-facilitated open one. cash transfers may explain this regional differ- ence. Women are often targeted by humanitarian organizations as recipients of cash transfers, while female-headed households are prevalent in refugee settlements. Thus, more women in the Southwest have bank accounts than in West Nile, likely due to the higher percentage of bank account ownership among female refugees. Figure 4.5: Bank account ownership Do you own a bank account? Southwest West Nile 65% 40% 34% 31% 20% 24% 16% 18% 12% 11% Total Hosts Refugees Women Men Total Hosts Refugees Women Men Sample size: Southwest total n = 1,117; West Nile total n = 930 Other financial institutions have also recently and refugees prefer Centenary Bank (57 percent), entered the refugee market. Notable actors are followed by PostBank (12 percent), and then the Uganda Agency for Development Limited Equity Bank (8 percent). Respondents mostly (UGAFODE) and Opportunity Bank, both of which cited Centenary Bank’s accessibility and popu- have set up a branch in the Nakivale refugee larity in the area, as well as low account charges settlement in the Southwest. Another actor is and low credit interest rates. Kiva, an international NGO, which partners with banks to finance refugee loans (group, individual, 4.2.2 Access to credit and small enterprise loan products) to help de-risk the loan portfolio of the bank. Borrowing is mostly informal among both refugees and hosts. Host communities and refugees mainly rely on VSLAs—community A significant percentage of those who do not yet savings groups that allow members to receive possess a bank account reported willingness interest on savings and access microloans. to open one. In the Southwest, both refugee Just over a third of respondents in both regions (41 percent) and host community respondents claim to have borrowed from this source in the (39 percent) are willing to open bank accounts. previous 12 months. A third of respondents also In West Nile, 36 percent of host community borrowed from family and friends. Borrowing respondents are willing to open a bank account; from shopkeepers, such as buying goods however, fewer refugee respondents (18 percent) on credit, is also common in both regions. are willing to do so. Of those who would like to Borrowing from banks or mobile money services open a bank account, the majority of both hosts 37 is very uncommon, as seen in Figure 4.6. Only cial cards35 from the Bank of Uganda at a cost of 2 percent of respondents had borrowed from a $9.50 (35,000 Ugandan shillings), costs associ- bank in the previous 12 months. Once registered ated with opening a bank account: $2.70 (10,000 for a mobile money service, refugees and host Ugandan shillings), and an initial deposit of $27 communities can access digital loan products, (100,000 Ugandan shillings) before qualifying for such as Airtel Wewole and MTN MoKash. a loan.36 With the latter, users can borrow from 3,000 to 1 million Ugandan shillings ($1 to $267) at a 4.2.3 Saving mechanisms facilitation fee of 9 percent.34 However, the use of this service appears very low, possibly due Host communities and refugees use traditional to lack of awareness or preference for informal methods of saving—most use VSLAs or save borrowing (see Figure 4.6). cash at home. VSLAs are the most common way to save and are especially popular with women. People join VSLAs to save for establishing or Loan collateral requirements influence the expanding business ventures, have coverage choice of financial service mechanisms. Lending for emergencies, and take out microloans. prerequisites are cumbersome at commercial Saving with financial institutions is uncommon. banks and microfinance institutions, according In West Nile, only 4 percent do so (1 percent of to an interview with YIDA, a refugee-led micro- refugees and 5 percent of the host community). finance organization in the Kyaka II refugee In the Southwest, 6 percent do so (2 percent of settlement. Formal financial institutions (such as refugees and 7 percent of the host community), banks, SACCOs, and microfinance institutions) despite higher ownership of bank accounts typically require collateral. As most of the land in there. According to focus group discussions refugee settlements is owned by OPM, the valu- with saving groups, VSLAs themselves do not ation of collateral is that of the buildings (refugee normally deposit their funds with financial insti- shelters) on the OPM-provided land rather than tutions; instead, they often store money in boxes, that of the land and the buildings. Valuation of protected by padlocks. While mobile money ser- semi-permanent housing structures only allows vices are widely used in both regions, transfer refugees to borrow smaller amounts of credit. charges and occasional unavailability of agents Other prerequisites for borrowing include finan- Figure 4.6: Sources of loans in the past 12 months Which of the following sources have you borrowed from in the last 12 months? Southwest West Nile VSLA 37% VSLA 37% 35% 35% Family and friends 34% Family and friends 32% 32% 34% Shopkeeper 15% Shopkeeper 18% 23% 17% Bank 5% Bank 1% 3% 3% SACCO 3% SACCO 4% 3% 4% Microfinance 3% Microfinance 2% institution 1% institution 2% Mobile money 1% Mobile money 1% 1% 1% Money lender 1% Money lender 1% 1% 2% Pawnshop trader 0% Pawnshop trader 0% 1% 0% Host community (n = 555) Refugees (n = 224) Host community (n = 725) Refugees (n = 325) 34 Source: MTN Uganda website. 35 In Uganda, all borrowers need to register and obtain a financial card. 36 Source: Key informant interviews with Opportunity Bank and UGAFODE in the Nakivale refugee settlement. 38 Figure 4.7: Saving mechanisms used Over the past 12 months, which of the following have you used to store or save money for more than one day? Southwest West Nile VSLA 42% VSLA 40% 28% 37% At home 36% At home 37% 16% 37% Financial institution 7% Financial institution 5% 2% 1% Merry-go-round 4% Merry-go-round 0% 7% 0% With a family member 1% With a family member 1% 1% 0% 1% 1% With a friend 5% With a friend 1% Host community (n = 782) Refugees (n = 376) Host community (n = 835) Refugees (n = 440) in refugee settlements have been mentioned goods, emergencies, medical expenses, and in focus group discussions as reasons for not education. Refugees are more likely to save using mobile money for group savings. Finally, for food than host communities are—a third of merry-go-rounds are also used in the Southwest. refugees in the Southwest (22 percent of host These are community savings groups, primarily communities) and 60 percent of refugees in West aimed at entrepreneurs, that collect money Nile (43 percent of host communities) do so. Host from members and give the total amount to one communities are more likely than refugees to be member on a rotational basis. saving for longer-term needs, such as education, setting up a new business, or expanding an Refugees and host communities save for the existing one (see Figure 4.8). same main needs: buying food or household Figure 4.8: Reasons for saving Over the last 12 months, what have you been saving for? Southwest West Nile Food 22% Food 43% 33% 60% To provide for the family To provide for the family in case something 30% in case something 47% 24% 41% happens to me happens to me 25% 29% Medical expenses 21% Medical expenses 38% Education 22% Education 33% 12% 22% Household goods 20% Household goods 15% 11% 25% To develop an 12% To develop an 14% existing business 5% existing business 8% To set up a business 8% To set up a business 14% 6% 5% Travel 4% Travel 7% 0% 4% To pay off a loan faster 3% To pay off a loan faster 2% 1% 2% Retirement/old age 3% Retirement/old age 1% 1% 1% Wedding 0% Wedding 1% 1% 1% Host community (n = 776) Refugees (n = 468) Host community (n = 824) Refugees (n = 452) 39 4.2.4 Sources of information on financial matters Refugees and host community members mainly learn about financial matters from their family Refugees and host or friends and community announcements. community members mainly Refugees and hosts use similar sources for information on financial matters, mainly based learn about financial matters on communication with their peers or community from their family or friends and announcements. A large share of host community members—51 percent in the Southwest and community announcements. 45 percent in West Nile—also learn about Refugees and hosts use financial matters from mass media (radio and television). By contrast, refugees use the mass similar sources for information media much less, especially in the Southwest— on financial matters, mainly 7 percent report doing so in that region, based on communication against 23 percent in West Nile. Mobile-based communication on financial matters appears to with their peers or community be more prevalent in West Nile (see Figure 4.9). announcements. Given that both communities seem to receive information primarily from their peers, financial institutions could reach new customers by engaging with community groups (especially savings groups) to market their services. Figure 4.9: Sources of information on financial matters What are your sources of information on financial matters? Southwest West Nile Community 18% Community 21% announcements 37% announcements 37% 35% 52% Family 32% Family 52% Co-workers 24% Co-workers 18% 5% 4% 53% 61% Friends 51% Friends 63% Mass media (radio, 51% Mass media (radio, 45% TV) 7% TV) 23% 7% 13% Mobile phone SMS 3% Mobile phone SMS 17% Education 6% Education 2% 5% 1% 2% 3% Internet 1% Internet 1% 4% 6% No information 3% No information 14% Host community (n = 783) Refugees (n = 354) Host community (n = 836) Refugees (n = 440) 40 41 5. Household Consumption Expenditure in the Southwest and West Nile 5.1 Household consumption the same region ($201 million versus $45 million), expenditure overview as shown in Figure 5.5. The Southwest also has T a higher monthly household expenditure overall he expected market size—measured than West Nile ($39 versus $35). The contrast as total annual household consumption between refugee households in the two regions is expenditure—of the two regions is esti- evident: in the Southwest, their average monthly mated at $485 million (1.1 million households). household expenditure reaches $43, compared The refugee-hosting districts in West Nile to $24 in West Nile (possibly because many West account for $246 million (51 percent), which is Nile refugees receive food rations as opposed to largely driven by the host community’s spending, cash transfers). at $201 million annually. The refugee-hosting districts in the Southwest contribute $239 million annually (49 percent) to the total.37 The market The Kyegegwa and Isingiro districts are the size is slightly larger in West Nile than in the largest markets in the Southwest, whereas Arua Southwest because of the larger population size and Yumbe are the largest ones in West Nile. in the refugee-hosting districts of the former Arua (253,630 households) and Yumbe (190,206 (588,000 versus 514,000 households). households) have the largest populations in West Nile, and total annual household consumption The market size of the host community is expenditure is estimated at $101 million and larger than that of refugees because of the $67 million respectively. However, Adjumani larger population size and higher average (West Nile) has the largest expenditure per monthly household expenditure ($38 versus household: $54 monthly, versus just $33 in Arua $33). In the Southwest, the market size of and $29 in Yumbe. In the Southwest, total annual the host community is twice that of refugees household expenditure is highest in Kyegegwa ($165 million versus $74 million). The market size ($73 million), as is average monthly household of the West Nile host community, on the other expenditure ($44), followed by Hoima ($58 million hand, is more than four times that of refugees in in total and $41 average monthly figure). 37 Estimates for consumption expenditure in the Southwest and West Nile refer to the surveyed refugee-hosting districts only (see Methodology section for more details). 42 42 Figure 5.1: Market sizes of the surveyed districts Moyo Yumbe $25M $67M Adjumani Arua $53M $101M Hoima $58M Kyegegwa Kamwenge $73M Kampala, capital city $37M Legend Isingiro $71M Southwest districts surveyed West Nile districts surveyed A few socio-demographic characteristics The pandemic had a more pronounced impact appear to affect consumption patterns. While on consumption expenditure in the Southwest there are no significant differences with respect to (with a drop of 16 percent) than in West Nile gender in the host communities, female-headed (with a drop of 10 percent). In the Southwest, households in Southwest refugee settlements household expenditure dropped by 17 percent spend less than male-headed households ($47 since the beginning of the pandemic in refugee versus $56) whereas West Nile male-headed settlements and by 15 percent in host commu- households spend less than female-headed nities. In West Nile, host communities saw their ones ($21 versus $29).38 Differences can also be consumption expenditure decrease by 12 percent observed between nationalities: Somali refugees with the pandemic.39 However, West Nile refugee spend relatively more monthly than refugees households did not experience a decline in from other countries ($68), while South Sudanese spending, as theirs was already significantly low spend the least ($27), with the host community prior to the pandemic. At the district level, the spending an average of $44 monthly (see Figures pandemic had a stronger impact in Kamwenge 5.2 and 5.3). (27 percent drop), Isingiro (22 percent drop), and Arua (19 percent drop). 38 Statistical significance tests were conducted for gender-related differences. 39 The percentage decrease was calculated based on the following questions: “How much did your household spend per month on average before the pandemic?” and “How much did your household spend last month?” 43 Figure 5.2: Average monthly household expenditure by gender 40 Southwest West Nile $56 $45 $45 $47 $46 $40 $29 $21 Male Female Male Female Male Female Male Female Host community Refugees Host community Refugees (n = 605) (n = 386) (n = 652) (n = 399) Gender: There is little difference in spending between female- and male-headed households in the host communities, while refugee households present regional differences. Figure 5.3: Average monthly household expenditure by nationality $68 $47 $44 $42 $39 $27 Burundi DRC Somalia South Sudan Uganda (host) Total (n = 48) (n = 263) (n = 52) (n = 383) (n = 1,253) (n = 2,046) Nationality spending: Somali refugees spend the most, while South Sudanese spend the least.41 Figure 5.4: Impact of COVID-19 on expenditure Southwest West Nile -9% Kikuube 2% Adjumani -22% Isingiro -19% Arua -27% Kamwenge 2% Moyo -6% Kyegegwa -8% Yumbe 40 The gender and nationality spending analysis was conducted on a subsample of household heads making up 81 percent of the sample. 41 The sample of Ethiopians, Rwandans, and Sudanese is too limited to draw conclusions. 44 Total annual household consumption expenditure in selected refugee-hosting areas Total annual household expenditure pre-COVID-19 $557 MILLION Total annual household expenditure post-COVID-19 lockdown $485 MILLION Figure 5.5a: Annual household expenditure in the Southwest Southwest TOTAL ANNUAL HOUSEHOLD DISTRICT CONTRIBUTION TO TOTAL EXPENDITURE ANNUAL HOUSEHOLD EXPENDITURE $250 $194 $200 $165 $58M $73M $150 Million $100 $89 $74 $50 $37M $71M $0 Host Refugees community Pre-COVID-19 Kikuube Isingiro Post-COVID-19 lockdown Kamwenge Kyegegwa Figure 5.5b: Annual household expenditure in West Nile West Nile TOTAL ANNUAL HOUSEHOLD DISTRICT CONTRIBUTION TO TOTAL EXPENDITURE ANNUAL HOUSEHOLD EXPENDITURE $300 $250 $230 $201 $200 $101M $53M Million $150 $100 $45 $45 $25M $50 $67M $0 Host Refugees community Pre-COVID-19 Arua Moyo Post-COVID-19 lockdown Yumbe Adjumani 45 5.2 Breakdown of household When measured as a percentage of the popu- consumption expenditure lation, refugees and host communities regularly spend money on home and personal care items, Consumer goods dominate household spending health, airtime, transportation, and education. in both regions, representing 59 percent of Ninety-three percent of all households buy food the total expenditure ($286 million). Within and non-alcoholic beverages every month. consumer goods, food and beverages account Among non-food expense categories, 64 percent for 45 percent of the total expenditure, home of households regularly buy home and personal and personal care for 6 percent, clothing for care items (with a higher percentage in host com- 5 percent, alcoholic drinks for 2 percent, and munities); 64 percent also regularly spend money lighting for 1 percent. Non-consumer goods on health (with a higher percentage among ref- represent a market of $199 million and include, ugees) and airtime (particularly hosts). About a as shares of total expenditure, education quarter of households regularly spend money on (12 percent), transportation (5 percent), airtime/ education and transportation. It is relatively rare Internet (4 percent), housing (1 percent), and to spend money on rent, as most refugees and entertainment (0.2 percent). hosts live in their own houses. Spending money on alcoholic drinks and entertainment is particu- The pandemic had a more pronounced impact larly uncommon among refugees. on the consumption of services than on expen- diture on consumer goods. Average monthly household expenditure on non-consumer goods (housing and services) decreased by 30 percent, from $21 before the lockdown to $15 after the When measured as a lockdown. This was largely driven by a reduction in education expenditure due to schools’ closure percentage of the population, during the data collection period. Monthly refugees and host communities expenditure on consumer goods increased by 17 percent, from $18 to $21, in West Nile because regularly spend money on of an increase in prices of essential consumer home and personal care items, goods during the lockdown period. The average monthly household consumption expenditure on health, airtime, transportation, consumer and non-consumer goods is indicated and education. in Figure 5.6. Figure 5.6: Expenditure on consumer goods and other categories On average, how much did your household spend per month before the COVID-19 outbreak? How much did your household spend last month? Southwest West Nile $24 $23 $22 $21 $21 $18 $16 $14 Non-consumer Consumer Non-consumer Consumer goods goods goods goods Expenditure before COVID-19 Expenditure after lockdown lifted 46 Figure 5.7: Percentage of population that spends on a particular category Which of the following categories does your houshold spend on monthly? 73% 70% 54% 55% 47% 30% 32% 26% 27% 23% 25% 16% 15% 13% 10% 12% 12% 2% 3% 0% Home/personal care Lighting Transportation Education Housing Health Airtime/Internet Clothing Alcoholic drinks Entertainment Hosts Refugees 47 5.3 Food and beverages While overall consumption patterns are similar across both regions, two items stand out 5.3.1 Most purchased foods and beverages among the top purchases: maize flour/posho Cooking oil, fruits and vegetables, rice, and in the Southwest, and tea and coffee in West meat are in the top five purchased items in Nile. Evidence from focus group discussions both regions. In the Southwest, there are no indicates that Southwest refugees have mostly significant differences in the consumption of transitioned to cash aid, which they use to buy foods and beverages between refugees and food (mainly maize), while those living in West the host community for the top three purchased Nile still receive food, and maize grain is one of items (cooking oil, maize flour, and rice), but host the items distributed by the WFP. In addition, rice communities are more likely to consume meat, may be less popular among refugees in West Nile spices, tea or coffee, and milk than refugees. because the South Sudanese majority relies on In West Nile, host communities are more likely other grains, such as maize from the WFP. The than refugees to buy tea and coffee (85 percent relatively large share of people buying tea and versus 65 percent), cooking oil (86 percent versus coffee in West Nile could be related to the South 54 percent), meat (73 percent versus 49 percent), Sudanese culture of drinking tea. and rice (64 percent versus 36 percent). Figure 5.8: Most popular food and beverage categories Which of the following categories does your household use or buy? Southwest West Nile 81% 86% Cooking oil 84% Cooking oil 54% Maize flour/posho 79% Tea and coffee 85% 84% 65% Fruit and vegetables 73% Meat 73% 60% 49% Rice 67% Fruit and vegetables 68% 69% 81% Food flavors (spices) 64% Rice 64% 15% 36% Tea and coffee 63% Drinks (soft drinks) 30% 22% 13% Meat 56% Maize flour/posho 27% 37% 29% Milk 43% Confectionery 25% 27% 10% Drinks (soft drinks) 39% Milk 14% 14% 11% Alcoholic drinks 17% Food flavors (spices) 10% 3% 15% Pasta (spaghetti) 10% Alcoholic drinks 8% 6% 3% Confectionery 8% Milk powder 5% 0% 2% Snacks 7% Wheat flour 4% 1% 5% Wheat flour 7% Pasta (spaghetti) 3% 8% 1% Milk powder 2% Cereals 2% 1% 2% Cereals 1% Snacks 2% 1% 5% Host community (n = 776) Refugees (n = 468) Host community (n = 824) Refugees (n = 452) 48 5.3.2 Retail brands’ presence in products) are popular because Mukwano, like refugee-hosting districts BIDCO, has a distribution network that reaches The most popular cooking oils are Ugandan refugee settlements. products. Fortune Butto, a palm oil product of BIDCO, enjoys the highest popularity among The most bought tea brands are household refugees and host communities in both regions. names in Uganda. Rwenzori Tea is a product This affordable brand is sold in various sachet of Mukwano, Garden Tea is a product of Eagle sizes ranging from 50 ml to 1 liter. In addition, Investments, and IGARA Tea is a product of Igara. BIDCO has distribution networks from Kampala From focus group discussions in the Kyangwali and district capitals that reach refugee settlement in the Southwest and Bidibidi in West settlements. For example, there is a depot in Nile, refugees’ strong knowledge of local brands the Mbarara district from where the product is is attributed to robust distribution networks that distributed to the Nakivale refugee settlement. bring the tea products from Kampala and district Similarly, Mukwano, Roki, and Sunseed capitals to the refugee settlements and host vegetable cooking oils (all Mukwano Industries communities. Figure 5.9: Most popular cooking oil brands Which of the following cooking oil brands have you bought in the past four weeks? Southwest West Nile 57% 52% Fortune Butto 30% Fortune Butto 32% Mukwano 25% Mukwano 28% 14% 20% Roki 4% Sunseed 15% 27% 15% Sunseed 1% Roki 8% 1% 18% Host community (n = 776) Refugees (n = 468) Host community (n = 824) Refugees (n = 452) Figure 5.10: Most popular tea brands Which of the following tea brands have you bought in the past four weeks? Southwest West Nile 13% 26% Garden Tea 1% Garden Tea 59% 21% 8% IGARA Tea 35% IGARA Tea 6% Rwenzori Tea 54% Rwenzori Tea 18% 49% 14% Host community (n = 776) Refugees (n = 468) Host community (n = 824) Refugees (n = 452) 49 Figure 5.11: Most popular rice brands Which of the following rice brands have you bought in the past four weeks? Southwest West Nile Super 53% Super 78% 45% 68% Pakistani 17% Pakistani 23% 40% 48% Kayiso 18% Upland Rice 9% 19% 28% Upland Rice 10% Kayiso 4% 5% 1% Host community (n = 776) Refugees (n = 468) Host community (n = 824) Refugees (n = 452) Figure 5.12: Access to foods and beverages Are there any food and beverage items not available in the settlement/district that you would like to buy? Southwest West Nile Host community 4% Host community 4% 96% 96% Refugees 9% Refugees 6% 91% 94% Yes No Yes No The most popular rice brands in refugee-hosting host community areas include certain fruits and areas are also household names in Uganda. vegetables (apples, avocados, watermelons, and In both regions, Super and Pakistani are the carrots), Irish potatoes, fish, and milk. most popular rice brands, followed by Kayiso in the Southwest and Upland Rice in West Nile. 5.3.4 Channels through which Focus group discussions also highlighted the households purchase food items robustness of rice brands’ distribution networks Refugees and the host community mainly buy from Kampala and district towns to refugee the food items that they consume from open settlements. markets, followed by independent sellers and grocery stores,42 as shown in Figure 5.13. 5.3.3 Availability of foods and beverages Overall, refugees and host community members The majority of refugees (94 percent in the purchase consumed items in the same places, Southwest and 92 percent in West Nile) and although refugees in the Southwest are more host communities (96 percent in the Southwest likely to shop from independent sellers/stalls and 96 percent in West Nile) said that they can (31 percent) than at grocery stores (16 percent), access most of the foods and beverages they while the host community prefers grocery need, as shown in Figure 5.12. Some items stores (44 percent) to independent sellers/stalls reported as not available in some settlements or (29 percent). 42 Open markets are held weekly and attract traders from the host community and beyond. A grocery store is a typical small shop with permanent structures selling general merchandise and non-perishables. An independent seller/stall is a makeshift structure that mostly sells perishable and small-value items. 50 Figure 5.13: Popular venues for buying food items Where do you most often shop for food? Southwest West Nile 96% 88% 90% 67% 44% 31% 31% 34% 29% 21% 12% 16% 14% 14% 7% 11% 9% 5% 1% 5% 0% 0% 1% 1% 0% 0% 0% 0% Host community Refugees Host community Refugees (n = 777) (n = 469) (n = 824) (n = 452) Open market Grocery stores Independent sellers/stalls Relatives/friends Nearby town outside settlement/district Shipped from Kampala Other 51 Host community members mostly purchase from Ugandan vendors—82 percent of refugees consumer goods from fellow local sellers; in West Nile shop at Ugandan businesses, but however, some purchase from refugee vendors only a third do so in the Southwest. In refugee too. While almost all host community respondents settlements, respondents are more likely to buy buy from Ugandan businesses, a relatively size- food items from refugees with whom they share able portion of the host community (9 percent in a nationality, as shown in Figure 5.14. West Nile the Southwest and 12 percent in West Nile) also refugees likely do not purchase goods from purchase from refugee vendors. Three-quarters vendors of a different nationality because of the of refugees buy from refugee vendors, while there homogeneous setup of their settlements, which are regional differences with regards to buying mainly host South Sudanese. Figure 5.14: Nationality of vendors of food items Who are your typical vendors? Host community Refugees 98% 32% Local Ugandans Local Ugandans 82% 99% 9% Refugees 55% Refugee vendors 12% (same ethnicity) 80% 0% Refugees 74% Other 3% (same nationality) 75% Other refugees 18% 5% Other 1% 0% Southwest (n = 769) West Nile (n = 817) Southwest (n = 468) West Nile (n = 452) 52 Case Study: Bukere Market, Kyaka II Refugee Settlement The household survey showed that open markets are the most common venue for refugees and host communities to obtain food, beverages, and home and personal care items. The purpose of this case study is to give insight into market activity by looking at one of the weekly open markets. The case study is based on key informant interviews with the Youth Initiative for Development in Africa (YIDA) in the Kyaka II refugee settlement in the Kyegegwa district in the Southwest, as well as anecdotal insights from households. It focuses on people that participate in the markets and the nature of commodities traded. Traders come from outside the refugee settlement The open market takes place in Bukere, the main trading center in the Kyaka II refugee settlement. It typically attracts traders from outside the settlement, who bring items such as electronics, textiles, solar products, live animals, hardware, mattresses, and food items. The traders come from the Kyegegwa district, the nearby district of Mubende, and Kampala. Live animals brought in include cows, goats, pigs, chickens, and ducks. Goats and cows are usually purchased for slaughter, while chickens, ducks, and pigs are usually purchased for rearing. Frozen fish (transported in cold storage trucks) are brought in by traders from Mubende. Wholesale distributors On a market day, up to four large trucks can be observed offloading items at large wholesaler outlets. Some of the items that these distributors bring include soft drinks, alcoholic drinks (particularly beer), salt, cooking oil, soap, and flour. Participation by women Women traders are also active at the market, especially in fish trading. Fish traders usually travel outside the settlement to buy smoked fish in bulk for sale at the market. Other items that women sell include vegetables such as tomatoes, onions, and white cabbages. These are typically sourced from backyard gardens in the settlement and host community. 53 5.4 Non-food items 5.4.2 Channels through which households purchase home and personal care items 5.4.1 Popular home and personal care brands As with food items, respondents mostly buy non- The most popular brand among home and food items from open markets and independent personal care items in both the Southwest and stalls/stores. Refugees in the Southwest are more West Nile is Radiant (with about half of refugee likely to purchase home and personal care items and host respondents buying it in the four weeks from refugees of their nationality (73 percent) before the survey), by Ugandan company Movit than from local Ugandans (24 percent). However, Products, which manufactures and distributes in West Nile, most refugees buy these items body and hair care products (such as soaps, from local Ugandans (81 percent), followed by lotions, creams, and jellies). Toothpaste brands refugees of their ethnicity (78 percent). Almost all (Colgate in the Southwest and Fresh-up in West refugees and host community members in both Nile) and sanitary pads (Always) come next. regions have access to the home and personal care items that they need (as mentioned by 98 percent of respondents). Figure 5.15: Typical vendors of home and personal care items Who are your typical vendors? Host community Refugees 98% Local Ugandans 24% Local Ugandans 81% 98% 7% Refugees 52% Refugee vendors 9% 78% (same ethnicity) 0% Refugees 73% Other 4% (same nationality) 72% Other refugees 11% 2% Other 1% 1% Southwest (n = 772) West Nile (n = 816) Southwest (n = 468) West Nile (n = 452) 54 5.5 Energy expenditure A third of refugee households and 18 percent of host community households also use charcoal. It Solar home systems and battery-operated is more common for households to buy charcoal torches are the main sources of lighting in than to produce it, and refugees are more likely the Southwest and West Nile respectively. to buy charcoal because of restricted access National grid connection is mostly present to forests for firewood collection. Less than in major towns in host community areas and 1 percent of households in rural refugee-hosting settlement base camps. In West Nile, access areas use liquefied petroleum gas.44 to the grid is particularly low—at 6 percent in the host community and non-existent among Willingness to pay for better access to energy refugee households. The limited access to for lighting is high among refugees and electricity43 is driving the adoption of off-grid host communities. Host communities (overall, solutions, mostly solar home systems, solar 59 percent) are more willing than refugees lanterns, and dry cell battery-operated lamps. (45 percent) to pay for improved energy. Further- Refugees are significantly more likely to own more, more than half of prospective consumers a solar lantern than the host community, while are willing to pay more for costly solar home the reverse is true with respect to solar home systems compared to solar lanterns, as solar systems. The high prevalence of solar lanterns home systems generally provide better value among refugees can be explained by the free for money due to better durability and capacity. distribution of these products by UNHCR and Besides providing lighting, solar home systems partners. Better penetration of solar home can power multiple electronic devices at the systems in the Southwest could be an indicator same time. In West Nile, refugees are especially of better availability of this product and higher interested in acquiring solar lanterns (with about household purchasing power. Open wick lamps two-thirds reporting their willingness to pay), (tadooba) are mainly used in the host community. which may be based on their experience or awareness of this product (a third of refugees In refugee-hosting areas in Uganda, firewood in West Nile possess a solar lantern, as seen and charcoal are the main energy sources for earlier). Similarly, willingness to pay for access cooking among both refugees and host com- to grid-quality electricity seems low among munities. Another IFC-commissioned study refugees in West Nile, where there is no access focusing on energy access in rural Ugandan to this source, compared to the Southwest— refugee-hosting areas (excluding Kampala) 9 percent were willing to do so in West Nile found that firewood is used by 80 percent of against 52 percent in the Southwest. Willingness host households and 67 percent of refugee to pay for different lighting products by region households, who primarily collect it on their own. and population group is shown in Figure 5.17, and by product type in Figure 5.18. Figure 5.16: Sources of energy for lighting What is your main source of energy for lighting? Southwest West Nile 48% 38% 35% 32% 32% 27% 29% 18% 15% 15% 16% 15% 13% 13% 11% 11% 4% 4% 5% 6% 5% 2% 0% 1% 0% 1% 0% 0% 1% 0% 0% 0% Host community Refugees Host community Refugees (n = 777) (n = 469) (n = 824) (n = 452) Battery-operated lamp Solar lantern Tadooba Candle Solar panel Grid electricity Storm lamp Other 43 Electricity is defined as power distributed through the national grid or a private mini-grid. 44 Energy for cooking was not surveyed for this study. The information in this paragraph comes from another report commissioned by IFC: EED Advisory and Centre for Research in Energy and Energy Conservation, Final Report: Opportunities for Private Sector Engagements in the Refugee-Hosting Areas of Uganda under the Partnership for Improving Prospects, 2021. 55 Figure 5.17: Willingness to pay for better access to energy Would you be willing to pay for improved energy services? Southwest West Nile 62% 56% 45% 45% Host community Refugees Host community Refugees (n = 819) (n = 425) (n = 836) (n = 440) Figure 5.18: Willingness to pay more for specific lighting products For which products would you be willing to pay more? Southwest West Nile 69% 68% 62% 58% 53% 52% 52% 28% 13% 16% 8% 9% 0% 1% 0% 1% Host community Refugees Host community Refugees (n = 776) (n = 468) (n = 824) (n = 452) Solar panels Solar lamps/lanterns Electricity Other 5.6 Housing and sanitation housing (made of burnt bricks or cement blocks), compared to only 3 percent of refugees, which The survey revealed the prevalence of semi- can be explained by policies against permanent permanent structures, made of substandard housing construction in refugee settlements. materials, which are not long-lasting. Most According to a study conducted by NRC and people in the two regions live in houses that they REACH in refugee settlements across Uganda, own. For refugees, housing is often an expense refugee shelters tend to be prone to damage and at the time of arrival, when they receive land leakage, and some refugees reported that they from the OPM and construct their shelters. In were not allowed by the OPM/UNHCR to build the Southwest, most refugees (74 percent) and permanent structures using burnt bricks.46 many host community households (52 percent) live in houses made of mud and wattle. In West Nile, most houses are made of unburnt bricks There is a growing market for commercial (66 percent of refugees and 52 percent of host and residential real estate in refugee-hosting communities live in such a house).45 Just under half areas, specifically for houses made from burnt of host community households live in improved bricks.47 While most of the refugees and host 45 Housing styles in Uganda vary from one region to another for reasons such as availability of materials, income levels, culture, lifestyle, or weather conditions. 46 NRC and REACH, Owned Spaces and Shared Places, Refugee Access to Livelihoods, Housing and Property in Uganda, September 2019. 47 Source: A focus group discussion with savings groups in Kyaka II. 56 communities live in houses that they own, a market for commercial and residential real estate There is a growing market is emerging, especially close to trading centers. for commercial and The preference for improved housing is backed residential real estate in by the data on willingness to pay, especially for burnt brick housing. In the Southwest, 38 percent refugee-hosting areas, of refugees and 65 percent of the host community specifically for houses are willing to pay for improved housing; in West Nile, 40 percent of refugees and 46 percent of made from burnt bricks. the host community indicated willingness to pay. The vast majority of respondents would like to pay for housing made of burnt bricks, as shown in Figure 5.20. Figure 5.19: Quality of housing facilities in refugee-hosting areas What are the walls of the residence made of? Southwest West Nile 74% 66% 52% 52% 42% 36% 18% 15% 7% 9% 3% 0% 3% 3% 3% 3% 4% 1% 1% 0% 0% 0% 0% 1% Host community Refugees Host community Refugees (n = 775) (n = 467) (n = 824) (n = 452) Burnt bricks Clay/mud Concrete/cement blocks Tent Unburnt bricks Wood Figure 5.20: Willingness to pay for improved housing types Which housing improvements are you willing to pay for? Southwest West Nile 93% 95% 88% 62% 25% 25% 19% 9% 8% 4% 4% 5% 4% 0% 0% 1% 1% 2% 0% 0% 2% 0% 2% 3% 1% 1% 2% 3% Host community Refugees Host community Refugees (n = 776) (n = 468) (n = 824) (n = 452) UNHCR tent Other tent Hut Unburnt bricks Mud and wattle Burnt bricks Tenement (muzigo) 57 Covered traditional pit latrines without a Willingness to pay for improved toilet facilities concrete slab are the most common type of is significant, and is higher in the Southwest toilet arrangement. The most common latrine than West Nile, and among host communities has walls made of bricks, wood, or tarpaulin, than refugees. Fifty-six percent of respondents but without a concrete floor slab, as used by in the Southwest are willing to pay for better toilet 60 percent of Southwest and 64 percent of West facilities. In West Nile, the figure is lower, but still Nile households. The use of water-based toilet significant, at 39 percent. The Southwest host systems is limited to just 5 percent of Southwest community has higher willingness (66 percent) and 3 percent of West Nile households. A few than refugees (37 percent), possibly because households with no access to toilet facilities UNHCR and its partners provide latrines to ref- practice open defecation in the bush (8 percent ugees at no cost. In West Nile, willingness to pay of refugee households and 4 percent of host is similar among host and refugee households community households). (40 percent and 37 percent respectively). In the Southwest, there is more willingness to pay for ventilated improved pit latrines, whereas in West Nile the willingness to pay is predominantly for covered traditional pit latrines with a concrete floor slab (see Figure 5.22). Figure 5.21: Toilet arrangements used by households What toilet arrangements does your household use? Southwest West Nile Bush/no toilet 4% Bush/no toilet 4% 8% 8% Flush toilet connected 0% Flush toilet connected 0% to sewage system 1% to sewage system 1% Flush toilet with a septic tank 1% Flush toilet with a septic tank 1% 0% 0% Pour flush toilet 1% Pour flush toilet 1% 11% 0% Covered traditional pit 22% Covered traditional pit 18% latrine with concrete slab 22% latrine with concrete slab 37% Covered traditional pit latrine 63% Covered traditional pit latrine 71% without concrete slab 56% without concrete slab 51% Ventilated improved pit latrine 8% Ventilated improved pit latrine 5% 0% 1% Host community (n = 819) Refugees (n = 425) Host community (n = 824) Refugees (n = 452) Figure 5.22: Better toilet arrangements that respondents are willing to pay for What would be better toilet arrangements for you? Southwest West Nile Flush toilet connected 4% Flush toilet connected 1% to sewage system 11% to sewage system 0% 10% 9% Flush toilet with a septic tank 15% Flush toilet with a septic tank 3% Pour flush toilet 4% Pour flush toilet 4% 11% 5% Covered traditional pit latrine 22% Covered traditional pit latrine 51% with concrete slab 18% with concrete slab 57% Covered traditional pit latrine 8% Covered traditional pit latrine 18% without concrete slab 7% without concrete slab 13% Ventilated improved pit latrine 51% Ventilated improved pit latrine 27% 37% 22% Host community (n = 534) Refugees (n = 153) Host community (n = 334) Refugees (n = 164) 58 59 6. Profile of Businesses in the Southwest and West Nile T his chapter presents findings from the 6.1 Business owners’ business survey and secondary sources demographic characteristics review. It starts with demographic Most businesses in the sample (61 percent) characteristics of business owners, followed are male-owned. Overall, two-thirds of refugee by business registration procedures, average business owners are male across both regions. monthly sales, and access to finance. Among the surveyed host community-owned businesses, the prevalence of female-owned businesses is higher. In the Southwest, host community-owned businesses are more likely to be female-owned (at 53 percent) than in West Nile (at 37 percent). Figure 6.1: Gender of business owners 68% 65% 57% 57% 43% 43% 32% 35% Female Male Female Male Host community (n = 249) Refugees (n = 155) Southwest (n = 201) West Nile (n = 203) 60 60 The business survey identified a significant The business survey shows no direct number of Ugandan businesses operating relationship between the educational level of in West Nile refugee settlements. In line respondents and business ownership. In both with the West Nile nationality make-up, most Southwest and West Nile refugee settlements, entrepreneurs in the refugee settlements most business owners have no formal education are South Sudanese. There is also a notable or only attended primary school. Refugee presence of Ugandan businesses (36 percent of and host community business owners have the sample). In Southwest refugee settlements, similar levels of formal education. Twenty-nine 64 percent of the business owners interviewed percent of refugees have not completed primary are from the DRC and 15 percent are from education, compared to 23 percent of host Burundi, while the remaining 21 percent business owners. Business owners in the West includes Ethiopians, Eritreans, Rwandans, Nile host community have somewhat higher Sudanese, South Sudanese, and Somalis. In levels of formal education, with 81 percent of host communities, a large majority of surveyed business owners having some formal education businesses are Ugandan (96 percent in West Nile versus 70 percent in the Southwest. and 97 percent in the Southwest). 6.2 Types of businesses and In West Nile, most business respondents speak English, while in the Southwest, they mainly their characteristics communicate in three languages—English, Most surveyed businesses are small-scale Swahili, and Luganda. In West Nile, 83 percent retail traders or services providers, which have of business owners in the sample speak English. been in operation for 3.5 years on average. Arabic is the most spoken language among West Refugee and host community members tend to Nile refugees (with 83 percent speaking it). It is run the same types of businesses. Stores mostly also spoken by a significant number of Ugandan include duukas—small shops selling multiple business owners (22 percent). In the Southwest, products (the most prevalent type of business), Swahili is the most prevalent language, with food stalls selling fresh food, and drug stores, mostly refugees speaking it. A quarter of refugee as well as larger stores selling groceries or business respondents speak English, and so do hardware materials. Service businesses include a third of the surveyed Ugandan entrepreneurs. barbers, motorbike taxis, medical clinics, Half of Ugandan business respondents in the restaurants, mobile money kiosks, tailors, and Southwest also speak Luganda, another major others. Ugandans running businesses in West language often used for business communication Nile refugee settlements operate duukas, drug in Uganda. stores, mobile money kiosks, and food stalls or restaurants. Figure 6.2: Surveyed businesses by sector and geographic area What line of business are you in? Southwest West Nile Store/trader 38% Store/trader 42% 38% 35% Services 49% Services 48% 42% 50% Local production 5% Local production 2% 7% 0% Agriculture 4% Agriculture 2% 9% 2% Other 2% Other 8% 4% 13% Host community (n = 101) Refugees (n = 99) Host community (n = 120) Refugees (n = 83) 61 Many of the food items sold in duukas are 6.3 Registering a business basic, readily available consumer products. Most of the surveyed businesses are not regis- They match the food items consumed at the tered with the Ugandan administration. Sixty- household level, as seen in Chapter 5. The two percent of all businesses are unregistered, top six food items sold by these shops in both with the registration rate similar among refugee regions are cooking salt, cooking oil, maize flour, and Ugandan businesses: 42 percent of refugee confectionery snacks, beans, and alcohol. The business owners and 35 percent of host-owned typical basket of non-food items includes non- businesses are registered. Although business drug medical items, soap, and sanitary pads. entities often pay fees to their immediate admin- istration units (such as local councils and refugee Half of the surveyed businesses are sole welfare committees48) to operate, a business is proprietorships with no employees, while a only officially recognized as registered if it was quarter have only one employee. A relative issued a license by a mandated district authority, majority of businesses in both regions do not usually at the sub-county level. While only a third employ other people. Most (83 percent) of the of businesses are registered, a higher percentage respondents who do not employ staff said that (61 percent) of business owners pay fees to main- this is because they operate on a small scale. tain their businesses. A large majority of those Others have one to two employees, often who pay such fees reported paying formal fees working as shop attendants and part-time casual to the district sub-county office, refugee welfare laborers who help with loading and offloading committee, or another relevant authority (such as merchandise. the National Drug Authority for those owning a drug store). Only a small minority of businesses Refugee-owned businesses with employees (less than 8 percent) declared paying informal often employ Ugandans; it is also common fees to the village chairperson. for Ugandan businesses to employ refugees. Seventy-two percent of refugee-owned The vast majority (85 percent) of those registered businesses with employees employ Ugandans, with the Ugandan authorities declared followed by refugees of the same ethnicity registering a business under their business (34 percent) and refugees of different ethnicity entity name, with the findings being consistent (16 percent). On the other hand, a third of for refugees and host communities. For half Ugandan businesses with employees employ of the sample, it took less than three weeks to refugees. Women constitute, on average, register their business. In both regions, refugee- 39 percent of employees in the businesses owned businesses faced shorter registering surveyed. The share of women employed is the processes than the businesses in host same between businesses owned by refugees communities. In the Southwest, the median and host community members. Figure 6.3: Number of employees hired by surveyed businesses How many employees does this business have? Southwest West Nile Four 1% Four 5% 3% 5% Three 8% Three 8% 6% 7% Two 11% Two 10% 12% 23% One 24% One 29% 25% 27% None 54% None 48% 51% 37% Host community (n = 100) Refugees (n = 99) Host community (n = 120) Refugees (n = 83) 48 A refugee welfare committee is an elected refugee administrative structure governing a particular block or zone within a settlement. 62 duration is six weeks in the host community,  Certificate of incorporation registration compared to three weeks in the refugee forms—issued by the sub-county authority. settlements. In West Nile, the median duration is three weeks for host community business owners  Business registration forms—issued by refugee and two weeks for refugee entrepreneurs. A small welfare committees in some settlements number of businesses indicated that they faced (for those applying to operate in a refugee a very long process to register (almost two years settlement) and local village authorities. for some, up to three years for one business). 6.4 Average monthly sales of Table 6.1: Time it takes to register a business the surveyed businesses Median time it takes to Refugee-owned businesses on average register a business (in weeks) generate less revenue than those owned by members of the host community. Refugee Southwest business respondents make average monthly Host community 6 sales of $220, while host community business Refugees 3 respondents generate average monthly sales of $388. Furthermore, the average business West Nile in the Southwest generates more revenue Host community 3 ($407) than one operating in West Nile ($253). Refugees 2 In the Southwest, refugee-owned businesses make $247 per month, more than half as much as host community businesses ($544) in the The process typically involves the applicant vis- same region. In West Nile, refugee-owned iting relevant authorities (such as a sub-county businesses generate, on average, $182 per office, a refugee welfare committee, or local month, while average monthly sales of host village authorities) to submit the following forms: community businesses reach $282. However, host community-owned businesses appear  Trading license registration form—issued by to have been more affected by the COVID-19 the sub-county authority. crisis in terms of decreased sales revenue. The average sales of refugee-owned businesses  Registration form for certificate of registration— declined from $228 to $220 (-4 percent), while issued by the sub-county authority. those owned by host community members saw their average sales decrease from $444 to $388 (-13 percent). 63 Figure 6.4: Businesses’ bank account ownership Figure 6.5: Businesses’ willingness Do you own a bank account? to open a bank account Would you be willing to open a bank account? 78% 61% 56% 55% 47% 37% 21% 21% Southwest West Nile Southwest West Nile Host community (n = 101) Host community (n = 145) Host community (n = 83) Host community (n = 76) Refugees (n = 96) Refugees (n = 57) Refugees (n = 40) Refugees (n = 44) 6.5 Businesses’ access to finance borrow money in West Nile appears much lower, at only 27 percent of refugees and 16 percent Refugee business owners in the Southwest and of host community owners. A majority of those host community entrepreneurs in West Nile who did not borrow in the past 12 months said reported the highest bank account ownership. they did not need a loan (64 percent), while In the Southwest, the rate of bank account own- the rest mainly cited high interest rates or lack ership is high among refugee business owners of collateral, followed by unavailability of credit. (56 percent) and low among host community VSLAs were the most frequently mentioned business owners (21 percent). The inverse is true source from which respondents obtained a loan. in West Nile, as seen in Figure 6.4. Most of the surveyed business owners who do not yet possess a bank account are willing Figure 6.6: Percentage of respondents who obtained a loan in the previous 12 months to open one. Willingness to open an account is Did you obtain a loan in the past 12 months? particularly high among refugee owners in the Southwest (see Figure 6.5). Business respon- dents would like to open an account with banks 57% 52% that are already present near them, such as 27% Equity Bank and PostBank, which facilitate cash 16% transfers to refugees, and Centenary Bank, which has extended its operations to most areas. Southwest West Nile Fifty-five percent of respondents in the South- Host community (n = 83) Host community (n = 76) west borrowed money in the past 12 months to Refugees (n = 40) Refugees (n = 44) support their business operations, compared to only 18 percent in West Nile. Businesses take out loans to expand their business operations, replenish inventory, and cover unforeseen busi- ness expenses. In the Southwest, 57 percent of refugee-owned businesses and 52 percent of host community-owned ones said they take out loans. The percentage of business owners who 64 7. Agricultural Value Chains in the Southwest and West Nile T he prevalent value chains in both regions According to focus group discussions with are cassava, maize, and beans. Bananas farmers in West Nile, another crop grown com- (matoke) are also widely grown, especially mercially is sunflowers. There have been efforts in the Southwest, and primarily by the host to involve the region’s refugees in sunflower pro- communities.49 Other major crops grown in the duction. Mukwano Industries, a leading Ugandan Southwest are tea and coffee. McLeod Russel, cooking oil producer, engaged refugees in the largest tea producer in the country, has two Bidibidi in sunflower cultivation through a joint tea estates in Hoima. Sweet potatoes are a major project with Danish Church Aid.51 food crop in West Nile, as are sorghum, finger millet, and sesame. West Nile farmers also grow tobacco as a cash crop50 (see Table 7.1). Table 7.1: Agricultural value-chain mapping in refugee-hosting districts52 Maize Cassava Beans Banana Ground- Sesame Tea Rice Finger Sweet Sorghum Tobacco Fruit Coffee Cotton nuts millet potato Southwest Hoima        Kyegegwa          Kamwenge            Isingiro   West Nile Yumbe          Adjumani            Moyo          Arua           49 Host community members grow bananas, whereas refugees are not allowed to grow perennial crops, as mentioned in World Bank and Kilimo Trust, Markets and Value Chains Assessment Kyegegwa, 2020. This is also in line with findings from focus group discussions conducted with farmers for this study showing that members of the host community own banana plantations, while refugees sometimes provide labor on those plantations. 50 The 2020 study by the World Bank and Kilimo Trust identified sesame as a crop grown by the host community in Adjumani, with few refugees involved in sesame cultivation. Refugees tend to grow okra, which is an important element in the diet of South Sudanese refugees. World Bank and Kilimo Trust, Markets and Value Chains Assessment Adjumani, 2020. 51 UNDP, Understanding Land Dynamics and Livelihoods in Refugee-Hosting Districts in Northern Uganda, 2018, p.5 52 This information is drawn from UNDP and Uganda Investment Authority, Uganda Refugee Hosting Districts—Investment Profiles, 2017. 65 Focus on the tea value chain in the Southwest McLeod Russel Uganda Limited is the Area under tea cultivation: country’s largest tea producer and exporter. The company has six tea estates in the country, two of which are in Hoima, which hosts the Kyangwali refugee settlement. 805 383 HECTARES HECTARES The company uses a plantation-based tea IN BUGAMBE IN KISARU factory model, meaning that its supply chain TEA ESTATE TEA ESTATE largely relies on tea estates that it owns. A small portion of the tea supply chain is made up of a network of outgrowers in Hoima. Refugees work as pluckers on these tea 2.7 MILLION KG 1.67 MILLION KG plantations. They usually do short-term work OF TEA, OF TEA, with the goal of saving a targeted amount of BUGAMBE KISARU TEA money before they move on. TEA ESTATE ESTATE Source: McLeod Russel Agricultural inputs reach producers through two Beyond small-scale milling in the Southwest, main routes—the development partner route there is minimal value addition once agricultural and the commercial route; refugee farmers often produce has gone through the post-harvest struggle to afford the inputs. The development handling stage, as focus group discussions with partner route comprises actors like UNHCR and farmers revealed. Produce is generally traded NGOs, such as Action Against Hunger and the immediately after the harvest. There is a lack of Lutheran World Federation, supplying improved post-harvest storage facilities in both refugee seeds for free to associations and community- and host communities, combined with sub- based groups. The groups and associations optimal post-harvest handling techniques (such then distribute the seeds to their members. as drying grain on bare ground). Farmers, Figure 7.1 shows agricultural value chains in ref- especially refugees, are dependent on prices ugee-hosting areas in the Southwest.53 The main set by middlemen and generally have low inputs that flow through the commercial route negotiating power, often because they lack are herbicides, pesticides, agrochemicals, farm access to information about the market prices equipment, and low-grade seeds (often coming and alternative markets for their produce. In from Container Village in Kampala—a trading focus group discussions conducted in West hub for agricultural inputs). Farmers who partici- Nile, host community farmers mentioned that pated in focus group discussions in both regions they understand the value of bulking, but often reported that they find the seeds expensive; they cannot afford to hold on to the produce for long, often use leftover seeds from previous seasons, because it is their only source of income. Focus which results in poor yields. Another challenge group discussions also revealed that young for refugee farmers who want to access seeds people in the Southwest and West Nile refugee is that often there are no nearby outlets, so they settlements are actively involved in bulking, must travel to base camps and nearby hosting aggregation, and trade in most agricultural value district towns, which increases the cost of chains. They move from house to house collecting obtaining inputs. Farmers in the Southwest also beans, sorghum, and cassava from smallholder reported limited knowledge about agricultural farmers. Village collectors often lack bicycles inputs, including improved seeds, fertilizers, and and have to transport the produce to the bulking agrochemicals. centers on foot. 53 The illustration was drawn from insights derived from focus group discussions in the Nakivale and Kyangwali refugee settlements and the host communities of the mentioned settlements. 66 Figure 7.1: Basic illustration of agricultural value chains in refugee-hosting areas in the Southwest UNHCR and other Community-based development partners Rural groups and associations assembly and aggregation Wholesalers Agri input Retailers in in major towns Trade and dealers in trading centers Production in host transport Kampala in host district community Retailers in main trading centers of Market settlements 67 8. Investment Opportunities T his chapter outlines investment oppor- 8.1 Agricultural and fishing tunities based on the findings presented value chains in this report. These opportunities span a The survey identified a high prevalence range of sectors, including agribusiness, lighting, of agricultural workers, particularly in the and the financial sector. Southwest, with many engaging in subsistence agriculture. The host community has better A distinction has been made to indicate whether access to land than refugees, as do Southwest each opportunity is within the settlement and households compared to those living in West the immediate host community (in-settlement), Nile. Previously, refugees received arable land or including an area outside of the settlement for agriculture in addition to residential plots. but still within a refugee-hosting district However, the amount of arable land available (ex-settlement). “Immediate host community” to allocate has significantly reduced in recent refers to the district sub-county in which a refugee years owing to the influx of refugees. Complex settlement is located. Opportunities beyond the community land titles pose additional challenges settlement and the immediate host community in West Nile. have been explored because the former are likely to be small in scale and thus cannot solve the unemployment problem. In addition, the The agribusiness investments in the refugee- recommended ex-settlement opportunities have hosting districts should focus on production significant potential to create jobs and develop systems designed around smallholder farmers small businesses owned by refugees and host with secure land tenure and prioritize selling communities. agricultural produce in local markets, with a portion of produce set aside for other markets in Uganda and for export. Opportunities in solar- powered drip irrigation should be explored to improve yields, particularly in smaller refugee- owned plots, and ensure year-round crop production. Farmers also need access to finance, as well as training in improved agricultural practices. 68 68 8.1.1 Maize value chain opportunities in the refugee-hosting districts. This could Maize is one of the most common crops grown potentially lead to the emergence of refugee by farmers in both regions. It is one of the nine aggregation businesses. priority crops in Uganda’s National Development  National maize aggregation businesses Plan III, with several ongoing projects contribu- (ex-settlement opportunities): Given the ting toward boosting the maize value chain. high production volumes in refugee-hosting Investment opportunities include: areas, opportunities could be explored for large-scale aggregators and post-harvest  Maize mills (in-settlement and ex-settlement warehouse businesses (such as Aponye, opportunities): Maize prices are volatile, yet Supreme Millers, Tru-Trade, and Equator local livelihoods often depend on maize sales. Seed) to establish or expand their operations Setting up smaller-sized mills that absorb in these areas. maize produced by refugees and the host com- munity would contribute to stabilizing prices.  Feed mills (ex-settlement opportunities): Maize mills would produce maize flour as a key Maize is an important ingredient in the formu- output and maize bran as a by-product. Maize lation of animal feed, especially for poultry. flour has a ready local market—about half of Opportunities can be explored to set up feed respondents in the household survey reported mills for poultry feed production. This would buying maize flour in the four weeks prior to the require a significant market from poultry pro- survey. Opportunities to set up slightly larger ducers in the two regions, which is an area for mills within and close to the settlements could further research. also be explored given the lack of such mills 69 8.1.2 Cassava processing 8.1.4 Poultry production Cassava is widely consumed by refugees as it Poultry is an inexpensive source of animal is a staple food in the DRC, the country of origin protein that can be consumed locally. Setting of most refugees in the Southwest. Focus group up a poultry farm in refugee-hosting areas could discussions with refugee farmer groups in West also be explored. This would ideally be a farm Nile revealed that some refugees grow cassava where day-old chicks are reared to 45 days despite having small plot sizes. Cassava does not old and then sold to households with improved require storage, although it must be processed livability. The farm should be operated by an within 48 hours of harvesting. Cassava is also existing larger poultry producer who can house one of the crops discussed in Uganda’s National parent stock, hatch chicks, and supervise the Development Plan III, which promotes several farm. Crops could be grown locally to supply a ongoing support programs intended to boost simple feed mill that would produce feed for the the cassava value chain. Opportunities in the poultry farm. cassava value chain include: 8.1.5 Fertilizer supply and blending plant  Setting up cassava chipping points (in-set- Crop yields could be boosted with a supply tlement and ex-settlement opportunities): of fertilizers to smallholder farmers in refugee Cassava chipping machines are used to settlements and host communities. Farmers convert raw cassava to cassava chips, which need to be supported with access to credit to are then dried. There is a market for cassava afford fertilizers and with practical training and chips in Kampala as well as the neighboring mentoring on improved agricultural practices. countries of South Sudan and the DRC. An opportunity to set up a fertilizer-blending Cassava chips could be further converted to plant could also be explored, potentially in the cassava flour. Hoima district where the Kabaale industrial  Setting up cassava mills for converting park is located. If deemed viable, a fertilizer- freshly harvested raw cassava into cassava blending plant in the district will not only address flour, which could be sold locally and outside constraints associated with poor crop yields but the refugee-hosting areas. Mobile cassava also provide job opportunities for both refugees mills could be considered to increase the area and the host community. There is also potential for sourcing raw material. to serve other areas of Uganda with fertilizers. 8.1.3 Vegetable production 8.1.6 Fish cage farming Supporting the production of vegetables Refugees participate in very small-scale wild that traditionally form part of refugee and capture fishing along the River Enyau in the host communities’ diets will increase the Rhino Camp and Lake Albert in Hoima. However, availability of nutritious food. A solar-powered the Ugandan government is discouraging wild drip irrigation system that uses no power and capture fishing to preserve the available fish. less water (compared to traditional irrigation Thus, commercial fish farming that uses cages methods) should be explored to grow vegetables and ponds could be explored as an alternative. throughout the year. A producer supplying larger A small-scale opportunity would support the cities with vegetables and requiring additional transition from wild capture to commercial cage production could be engaged to connect hosts or pond farming, including through a set-up of and refugees to additional markets. A solar- small plastic-lined ponds, and would serve the powered cool chain from the field to market, if local market. A large-scale one would involve developed, would increase the products’ shelf setting up a proper aquaculture park with core life and reduce post-harvest losses. activities including cage and pond fish farming and processing, potentially supplemented with fish feed and fingerling production. This could potentially be done at Lake Albert in the Hoima district, as it is deep enough for setting up such an operation. A hatchery is a complex operation 70 and must be set up by an experienced investor. bank accounts, compared to only 16 percent Support activities in the park would be in storage in West Nile. The limited use of banking does and warehousing facilities, office spaces, and not reflect a low demand for financial services transportation services, among others. Refugees but is the result of challenging prerequisites would participate through the provision of labor affecting the most vulnerable people: 46 percent and the production of fingerlings as outgrowers. of the population would turn to formal financial institutions if they could, and half of these report lack of collateral as the main obstacle. In addition, 8.2 Energy for lighting lack of adequate capital is the main reason Significant investment opportunities exist in 95 percent of salaried employees who would like renewable energy, especially for solar home to start their own company are prevented from systems and solar lanterns, in both refugee doing so. With simplified banking procedures settlements and host communities. Overall, and no request for collateral, microfinance 54 percent of the population in refugee-hosting institutions can provide services to respond to areas are willing to pay more for improved energy that demand. Microfinance institutions should services. Of these, 62 percent are willing to pay learn from the relatively successful experience of more for relatively more expensive solar home VSLAs, as well as develop products targeted at systems, while 22 percent would pay more to farmers, who often struggle to access credit. buy solar lanterns. Solar home systems generally provide better value for money due to their In addition, an opportunity exists to extend durability and multiple functionalities, but the banking services through the agent banking market is flooded with cheap, counterfeit solar model. The Financial Institutions (Agent Banking) equipment that has a high failure rate, forcing Regulations (2017) allow banks to extend the consumers to make new purchases regularly.54 services traditionally offered in bank branches Therefore, there is an opportunity to scale up through agents. Agents are allowed to facilitate the provision of genuine, higher-quality solar the following services: products in the refugee-hosting areas. These could be financed using pay-as-you-go models  Collecting and forwarding applications, that allow consumers to pay for solar products in information, and supporting documents for small installments using mobile money. An IFC- opening accounts or for payment instruments commissioned energy study assessed that about a third of refugee households that own solar  Cash deposits and withdrawals products (in all refugee settlements in Uganda) received their solar lanterns for free, while two-  Payment services, including bill payments thirds purchased them.55 Therefore, even taking  Money transfers into account possible distortionary effects of the distribution of free products to refugees,  Loan disbursements and repayments there appears to be a significant market for solar lighting products.  Receiving and forwarding documents relating to loans and leases and any other permitted products 8.3 Financial sector  Paying retirement and social benefits In a context where access to finance is mostly informal, opportunities exist for  Account balance enquiries microfinance institutions to fill the gap. In both refugee settlements and host communities,  Providing account statements. VSLAs, as well as family and friends, are the main source of credit for most people. In the Southwest, 34 percent of the population have 54 GIZ, Sustainable Use of Natural Resources and Energy in the Refugee Context in Uganda, 2018. 55 EED Advisory and Centre for Research in Energy and Energy Conservation, Final Report: Opportunities for Private Sector Engagements in the Refugee-Hosting Areas of Uganda under the Partnership for Improving Prospects, 2021. 71 Equity and Centenary banks already practice media and traditional communication channels, agent banking in refugee-hosting areas. Through or using agents for community outreach, will the Uganda Bankers’ Association, banks can use be key to connecting to VSLAs. Banks could a shared agency banking platform, meaning that also connect with NGOs that already work with the same agent can be used by multiple banks, VSLAs, such as Lutheran World Services, to help which lowers the costs of recruiting, training, and market financial products targeted at VSLAs. onboarding an agent network. 8.4 Housing and sanitation To encourage banks to set up or expand oper- ations in refugee-hosting areas, IFC could Opportunities exist to supply construction de-risk the cost of extending financial products materials for improved housing provision in both to refugees. Other development partners such refugee settlements and host community areas. as UN Capital Development Fund, ABI Finance, As described in Chapter 5, there is generally high and NUTEC Financial Services have been able willingness (over 40 percent) to pay for improved to de-risk financing to high-risk sectors through housing structures in both regions. Among blended financing and credit guarantee schemes. those who do not yet live in improved housing, willingness to pay for structures made of burnt Another opportunity is for banks to create links bricks stands at more than 85 percent in the with community savings groups. While VSLAs host community and more than 60 percent in the are the most common means of saving and credit refugee settlements. Furthermore, focus group for both refugees and hosts, community savings discussions revealed that there is a growing are often stored in cash boxes rather than a bank interest in refugee-hosting areas for houses where they could earn interest and increase. made from burnt bricks. As such, there is an Banks could market opportunities targeted at opportunity for the private sector to be involved VSLAs, focusing on messages such as increased in providing building materials for constructing safety of savings stored in a bank account, housing on the land that hosts and refugees own. opportunities to earn interest and access loans, However, building permanent housing seems to and the availability of branches or agents in the be discouraged in some refugee settlements, vicinity to solve problems. Advertising through which may pose a challenge. 72 Besides housing materials, there is a potentially of connecting these products to markets is significant demand for improved toilet facilities Fashionomics Africa, an online trading platform in refugee-hosting areas. In the Southwest, for handmade African crafts—a flagship initiative 66 percent of the population in host communities of the African Development Bank. and 36 percent of refugees are willing to pay for improved toilet facilities, specifically for venti- lated pit latrines; in West Nile, 40 percent among 8.6 Retail trade the host community and 37 percent of refugees Developing retailer supply channels: Despite are willing to pay for improved toilet facilities, the vibrancy of the retail market, retail business with a preference for traditional pit latrines with a activity is yet to reach its full potential because concrete floor slab. it is largely cut off from mainstream distribution channels. Thus, opportunities could be explored for food and personal care brands that offer 8.5 Arts and crafts production inexpensive products to tap into this retail trade and marketing potential by developing retailer networks. This Refugees from the DRC tend to be highly skilled may, however, require support from development at crafting products such as sandals, bags, wrist- partners that could engage several non- bands, necklaces, and other products from local competing brands to establish robust supply materials such as banana fiber, hides, and beads. channels. Agents could be sent out to different They are also skilled at making the African fabric areas to take orders from small shops, which kitenge. Such products are highly marketable to would allow the brands to test the market. international markets. However, these products are mainly sold within settlements at much lower It is important to note that retail trade might not prices, as links to better markets are yet to be create many new job opportunities, especially if established. total household consumption expenditure does not increase yearly. However, creating robust Therefore, there is an opportunity for trade supply chains will help build more resilient facilitation to connect refugees’ products to businesses as consumer product sourcing will larger and more profitable markets. An example be less affected by external shocks such as possible travel restrictions and lockdowns. 73 8.7 Plastic recycling 8.8 Education—vocational training There is a large and vibrant retail market in the Opportunities in business, technical, vocational Southwest and West Nile. The majority of busi- education and training (BTVET) could be nesses in refugee-hosting areas are retail-based explored to support building skills for formal and trade items such as cooking oil, soft drinks, job opportunities. From the household survey, and other consumables. These food and beverage only 3 percent of all respondents have received items are mainly packaged in plastic material. vocational training (dropping to 0.64 percent for Additionally, in-kind aid to refugees, especially refugees). BTVET can have a positive impact the non-food items, is often packaged in plastic. on employment and boost entrepreneurship: An opportunity exists to collect and grade plastic 58 percent of respondents are unemployed, that is produced and discarded, and partner with while 80 percent of those working for someone existing recycling facilities.56 A similar model has indicated they would rather work for themselves. been implemented in Dadaab Refugee Complex However, the existing BTVET landscape in Kenya, where plastic waste from the com- should also be carefully considered, with free munity and public places is collected, sorted, training opportunities offered to refugees by and graded before being transported to recycling NGOs. Furthermore, affording such services plants. Some of it is recycled and converted into is a challenge for most of the population, with mats, vases, baskets, and handbags by groups households already spending an average of in the community. The same models can be 12 percent of their household expenditure on considered in Ugandan refugee-hosting areas, education. Therefore, the sustainability of BTVET incorporating lessons learned from previously investments will require working with private implemented projects. In addition, bricks and businesses and tailoring the courses offered to tiles required for improved housing construction local demand. Offering the right skills will create could also be produced using plastic. incentives for local medium-to-large businesses looking for skilled labor to subsidize BTVET. This opportunity needs to be researched further. 56 One such facility is Century Bottling Company Ltd (Coca-Cola Uganda’s bottler), which is working in other areas of Uganda, having set up a waste-cleaning and processing plant in Kampala. 74 Annex Key informant interviews Table A.1: List of key informants interviewed Southwest West Nile Development partners Development partners 1. United Nations High Commissioner for 1. United Nations High Commissioner for Refugees (UNHCR), Mbarara sub-office Refugees (UNHCR) 2. Association of Volunteers in International 2. World Food Programme (WFP) Service (AVSI) Foundation 3. Finn Church Aid (FCA) 3. Finn Church Aid (FCA) 4. Finnish Refugee Council (FRC) 4. Finnish Refugee Council (FRC) 5. DanChurchAid 5. World Food Programme (WFP) 6. Adventist Development and Relief Agency Government of Uganda representatives (ADRA) 6. Office of the Prime Minister, settlement 7. USAID—Power Africa commandants (four interviews) 8. Nsamizi 7. Local government officials (four interviews) Government of Uganda representatives Others 9. Office of the Prime Minister, settlement 8. Refugee Welfare Committee in each of the four commandants (four interviews) settlements 10. Group interview with local government officials (host community areas) in each of the four districts studied in the Southwest Private sector actors/local NGOs 11. FUZU (NGO working with youth) 12. Advance Smart Microfinance Limited 13. Opportunity Bank, Nakivale branch (phone interview) 14. UGAFODE, Nakivale branch (phone interview) Refugee-led organizations 15. Youth Initiative for Development in Africa (YIDA) 75 76 77 About IFC IFC—a member of the World Bank Group—is the largest global development institution focused on the private sector in emerging markets. We work in more than 100 countries, using our capital, expertise, and influence to create markets and opportunities in developing countries. In fiscal year 2021, we invested $31.5 billion in total commitments, including $23.3 billion in long- term finance and $8.2 billion in short-term finance, to private companies and financial institutions in emerging and developing economies, leveraging the power of the private sector to end extreme poverty and boost shared prosperity. For more information, visit www.ifc.org.