PHOTO BY © 2016 DORTE VERNER Data-Driven Development Response to Displacement Crisis in Uganda: The Displacement Crisis Response Mechanism Chris Mahony and Barry Maher Context Uganda’s Displacement Shock: Uganda experienced a displacement shock in 2016 and 2018. It has one of the fastest growing refugee crises in the world, including the fourth largest number of refugees (the highest in Africa). Uganda is currently hosting 1.43 million refugees and asylum seekers (as of September 2020). South Sudanese (62%) make up the largest refugee population followed by refugees from the DRC (29%), Burundi (3.3%), Somalia (2.7%) and others (3%) from Ethiopia, Eritrea, Rwanda and Sudan. The Office of the Prime Minister (OPM) in collaboration with the United Nations High Commissioner for Refugees (UNHCR) undertakes biometric registration and documentation of the refugee popula- tion.1 The greatest concentration of refugees is in the West Nile region in Northern Uganda and to a lesser extent in the South West, bordering the Democratic Republic of Congo and South Sudan.2 Poorly managed responses to forced displacement threatens development outcomes and has the potential to increase tensions between refugee and host communities. Mitigating grievances between social groups, including between and within refugee and host communities, assists prevention of violent conflict.3 The additional pressures displaced persons may place on the host country, specifically on local hosting communities’ already scarce resources, is associated with risk of violent conflict.4 In Uganda, refugees are hosted in often underserved, underdeveloped areas in environments prone to climate change and natural disasters. A displacement shock can further strain these existing vulnerabilities.5 Uganda is currently hosting South Sudanese 1.43 million refugees and asylum seekers. 62 percent make up the largest refugee population OPM in collaboration with UNHCR undertakes BIOMETRIC REGISTRATION 2 | Data Driven Development Response to Displacement Crisis in Uganda: The Displacement Crisis Response Mechanism World Bank Response Following the inflow of refugees in 2016 and 2018, the World The approach adopted was Bank significantly scaled up support to the Government cross-sectoral, focusing on of Uganda to respond to displacement shocks, as well as more protracted displacement. The approach adopted was cross-sectoral, focusing on education, health, livelihoods, energy, and urban development. Central to this response was the US$50 education agriculture million Development Response to Displacement Impacts Project (DRDIP). DRDIP provided much needed improved social services access and expanded economic opportunities in refugee hosting health energy districts. With the increase in refugee numbers in 2018, the project was restructured to frontload support and expand the geographical coverage from the four initial districts to all 11 ref- ugee-hosting districts in Uganda. The World Bank then mobilized and urban development. US$150 million of additional financing to DRDIP for scaled-up support in these districts, drawing on innovative funding sources Central to this response was the such as the IDA 18 Regional Sub-window for Refugees and Host Communities.6 Two key innovations were added: 1) a focus on US$200 million Development Response to enhancing the management of social risks including Gender Displacement Impacts Based Violence and Violence Against Children given the socially frayed displacement context; and 2) the Displacement Crisis Response Mechanism. The immediate pressing nature of the refugee-related crisis demands that Bank-supported GoU activities are frontloaded and focused on host areas where there is increased need.7 Scarce services, particularly in education health and water, demanded a large share of available resources under DRDIP for the project’s initial year. The Displacement Crisis Response Mechanism (DCRM) Recognizing the need to be able to rapidly respond to future displacement shocks, the Displace- ment Crisis Response Mechanism (DCRM) was established under DRDIP.The DCRM aims to develop and finance a mechanism to support rapid government scale-up of assistance to host communities that are poor and vulnerable in response to a refugee-related displacement shock. The response is directed through development activities related to community infrastructure to build host and refugee community resilience by augmenting basic service delivery and sustaining human capital. To enable rapid response, the DCRM must have an accessible, accurate, transparent, rules-based and accountable (objective) decision making process to disburse resources. As the DCRM will enable a development response to a displacement shock, it will be owned and led by GoU. The rules- based approach to the response aims to de-risk perverse incentives that accompany the financing of humanitarian response. Furthermore, the data informing the response process will be objective, independently and transparently collected, and readily available to rapidly inform a response. Data Driven Development Response to Displacement Crisis in Uganda: The Displacement Crisis Response Mechanism | 3 Importantly, as the DCRM is a global innovation, it has challenges and opportunities. The unprec- edented mechanism enables rapid government response to a displacement shock using its own systems based on (and triggered by) objective, accurate, accessible, independent and transparent data. However, the evolving and context-specific nature of vulnerability in refugee and host com- munities is complex and difficult to monitor. To measure vulnerability, therefore, the DCRM employs multiple sources, consensus, and pragmatism. The principals underpinning the DCRM’s design draw on the emerging evidence (from safety net beneficiaries) of developing shock-responsive mechanisms that target poor and vulnerable households. The enabling principles include: (1) Objectivity (data triggering disbursement should be independent, manipulation-free, regularly collected and representative); (2) Accountability (clear rule- bound, time-bound, rapid, transparent and accountable decision making at each disbursement stage; (3) Pre-Financed (dedicated, rapidly mobilized resources); (4) Development-driven (community-driven activities implemented in line with community development plans and not humanitarian response). Objective, accessible, accountable, pre-financed, and development-driven principles are critical to investment prioritization in resource-poor contexts. It is well documented that such an approach is critical to avoid social group specific discontent, which can increase tensions and may cause insecurity.8 This approach advances capacity to identify, manage and mitigate associated project risk and advance a ‘do no harm’ approach.9 Importantly, given the mechanism’s innovative nature, rigorous development impact analysis will be undertaken throughout implementation with findings added to the global knowledge base on crisis response and peacebuilding. The Displacement Crisis Response Mechanism (DCRM) Activities The DCRM disburses funds to scale up government provision of water, healthcare, and education services. Identification of indicators to inform disbursement must therefore be identified. A pragmatic, iterative and sector-specific approach to the identification of metrices for triggering was adopted given: i) the complexity of refugee and host community vulnerability; ii) indicator range without development community consensus on the ‘best’ indicators, including for both host and refugee communities; iii) the DCRM’s innovative nature, and; iv) the importance of perception monitoring and a community-driven approach, including continuous stakeholder engagement in managing implementation risks.10 One indicator for healthcare, education and water is agreed with the GoU, enabling monitoring and evaluation of metric representativeness, accuracy, objectivity, accountability and accessibility. Furthermore, the indicators will attempt to unpack the similarities and differences between refugee and host communities.11 This approach builds on ongoing data strengthening and oversight undertaken by the GoU and UNHCR.12 The Indicators: 1. Education: Addressing inequalities of access to education between social groups contrib- utes to preventing violent conflict.13 Both UNHCR and the Ministry of Education collect data on primary and secondary enrolment and pupils per teacher specific to refugee and host communities.14 4 | Data Driven Development Response to Displacement Crisis in Uganda: The Displacement Crisis Response Mechanism a) Suggested indicator: School-aged children per classroom in primary schools. This appears to be the most representative indicator for level of need, and should be employed, based on data accessibility and scope for GoU data collection. An Iterative Beneficiary Monitoring (IBM) system using rapid phone interview data collection during displacement alongside geolocated and timestamped data collection via tablet may indicate this metric. 2. Healthcare: The GoU (OPM) and UNHCR monitors, evaluates, and reports on the number of community outreach activities, children treated, children referred, children vaccinated, health workers trained, health workers deployed, consultations, patients admitted, immunization, women’s post-natal and other gender-specific and family-specific health care.15 The Uganda Bureau of Statistics (UBOS) household survey identifies illness and injury rates and distance to medical centers. b) Suggested indicator: Persons per health center. The number of persons per health center is both accessible and representative of medium-to-long term development (rather than humanitarian) need. An IBM alongside geolocated and timestamped data collection via tablet may also monitor the number of persons attending health centers. 3. Access to water: Data representing the number, communal nature, and functionality of water points is collected by district water officers and OPM. Access to WHO designated ‘improved’ or ‘unimproved’ water sources data, requires field surveying.16 c) Suggested indicator: Persons per functioning water point. This indicator may obfuscate water access (distance and whether the water point is communal or available only to one or specified households) and quality but best represents average intake. Access to data combining quality and quantity of water intake per person should be explored. For each indicator identified above, a threshold level, pre-agreed by GoU, will be established in each district based on sector standards. Should the indicator breech the threshold level, resources will be rapidly mobilized into the sector in question to finance the relevant development activity in the district. For example, a threshold of 70 school-aged children per classroom might be established. If, in a given district, the number of school-aged children per classroom exceeded 70, resources would be made available from the DCRM to expand the educational facilities in that district. Exploration of further or complementary indicators will accompany the initial phase of DCRM implementation. Human need is complex and the most appropriate metrics for its determination is not settled. Technological innovation, such as remote sensing and image recognition software, as well as geolocated and timestamped data collected by observers, is enhancing data integrity that can complement orthodox data sources and strengthen disbursement processes. A phased imple- mentation approach enables core monitoring, evaluation and learning functions that test different approaches to data collection and disbursement while building GoU data, technical, operational, and institutional capacity. Data on these indicators will be collected and monitored by GoU and closely coordinated with key humanitarian actors. Pre-existing water, education, and health data collection methods will be used, as identified above. GoU will monitor identified indicators on a monthly basis and in response to reported displacement episodes. Where a threshold breech is identified, a response will be mobi- lized. To ensure coordination with the broader humanitarian response, GoU will coordinate regarding the design and disbursement of resources from the DCRM with key humanitarian actors including the UN and NGOs through the Comprehensive Refugee Response Framework (CRRF) Secretariat. Data Driven Development Response to Displacement Crisis in Uganda: The Displacement Crisis Response Mechanism | 5 PHOTO BY © UNHCR/H.COUSSIDIS Conclusion The DCRM proceeds on the principles of Objectivity, Accountability, Pre-Financing and Devel- opment-focus to ensure shock-responsiveness targeted at poor and vulnerable households in funding scarce environments. The DCRM prioritizes accuracy and accessibility of targeted, mer- it-based metrics that enable rapidly disbursing finance to activities of greatest impact when needed most – at the time of a displacement shock. The mechanism advances a ‘do no harm’ approach by preventing insecurity via an inclusive approach that mitigates social groups’ grievances about exclusion from access to services.17 An inclusive approach to a displacement crisis requires the response trigger to be accessible, objective, representative and transparent. The number of school-aged children per classroom represents access for both host and refugee communities and is accessible via already collected data. Similarly, the number of persons per health center objectively and accessibly represents health care across social groups. Finally, the number of persons per functioning water point, while not maximising the representativeness of quality of water, best weighs the accessibility of data, with objectivity and representativeness. The chosen metrics are all dependent on accessibility, including data already collected by the GoU, UNHCR and other relevant actors alongside enhanced capacity to identify demographic change via biometric registration. Survey data may inform the use of IBM phone surveys to test pre-identified representative samples of refugee and host communities. While acknowledging that the selected metrics make the most of pre-existing data collection processes, the project’s establishment and enhancement of GoU capacity to collect data will also have developmental impact beyond the project. 6 | Data Driven Development Response to Displacement Crisis in Uganda: The Displacement Crisis Response Mechanism PHOTO BY © EDGAR BATTE The complexity of need demands continuous monitoring and evaluation of the representativeness, accuracy, transparency and accessibility of selected metrics and the data that inform them. Beyond M&E, and as a consequence of the novel innovation of employing ‘need’ as a trigger, capacity is also made available as part of the project to explore metrics of optimal transparency, representativeness, objectivity, and accessibility. Testing the efficacy and integrity of data and metrics will require collaboration with stakeholders and experts in the field of development indices. These collaborations may include exploration of the representativeness of proxies for ‘need’, for which data, because of technological innovation, is increasingly available. 1. https://www.unhcr.org/refugee-statistics/ 23 October 2020UNHCR, Uganda starts biometric verification of refugees, UNHCR spokesperson Babar Baloch, 2 March 2018, Geneva. 2. Most refugees are accommodated in the West Nile region – one of the poorest areas in Uganda. The remaining refugees are settled in the western and southwestern parts of the country and also in the capital Kampala. Continued refugee inflows will increase pressure on land and services, and limited resources. See: UN High Commissioner for Refugees (UNHCR), Uganda: UNHCR Presence and refugee locations | May 2017, 10 May 2017. 3. United Nations. World Bank, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washington DC, 2018, p. 65; Bahgat, Karim, Kendra Dupuy, Scott Gates, H. Mokleiv Nygård, S. Aas Rustad, Håvard Strand, Henrik Urdal, and Gudrun Østby. “Inequality and Armed Conflict: Evidence and Data.” Conflict Trends 2 (2017). 4. Increases in malnourishment, strains on healthcare, land and water access, and diminished access to education, for example, have been shown to increase the likelihood of violence in communities hosting refugees. United Nations. World Bank, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washington DC, 2018, p. 26-27, 64. 5. These risks are particularly acute in Northern Uganda, where grievances residing around exclusion drove the country’s civil war. Situations of protracted displacement may also increase refugees’ vulnerabilities, particularly in circumstances of weak support for their self-sufficiency. For each added percentage point in undernourishment, violent conflict likelihood increases by 0.24 percent per 1,000 population. Similarly, declining access to healthcare can occur for both host and refugee communities. Grievances relating to diminished access to education are particularly associated with risk of conflict. Land and water access, particularly pastoral land, require mutually beneficial negotiated access to avoid conflict. These issues are linked to one another and to environmental degradation, which also increases the frequency and intensity of conflict.. United Nations. World Bank, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washington DC, 2018, p. 6. The IDA18 Refugee and Host Community Sub-Window demands that these resources reflect: Relative refugee numbers; The impact of refugee flows, and; The authorities’ commitment to achieving the objectives of the sub-window. The Ugandan situation meets these criteria. 7. World Bank, Uganda Board Consultation Note on Eligibility for IDA 18 Regional Sub-Window for Refugees and Host Communities (internal document), August 24 2017, p. 8. 8. United Nations. World Bank, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washington DC, 2018 9. Anderson, Mary B. Do no harm: how aid can support peace--or war. Lynne Rienner Publishers, 1999. 10. Implementation risks include effectiveness, safeguards, fiduciary risks, and low (and slow) disbursements, as well as low government ownership important for joint metric monitoring and disbursement processes. World Bank, Uganda Board Consultation Note on Eligibility for IDA 18 Regional Sub-Window for Refugees and Host Communities (internal document), August 24 2017, p. 7. 11. Host communities are more than three times as likely to have access to credit as refugee communities - Food and Agriculture Organization and Office of the Prime Minister, 2018. Food Security, Resilience and Well-being Analysis of Refugees and Host Communities in Northern Uganda 12. UNHCR, Uganda starts biometric verification of refugees, UNHCR spokesperson Babar Baloch, 2 March 2018, Geneva 13. United Nations. World Bank, Pathways for Peace: Inclusive Approaches to Preventing Violent Conflict, World Bank, Washington DC, 2018, Chapter 5. 14. The GoU collects data on recruited and paid teachers, enrolled students, and available classrooms and UNHCR reports on access to educational materials, school management, parent association training, enrolment rates, and classroom and teacher availability. See: Government of Uganda (Ministry of Education and Sports), Education Response Plan for Refugees and host Communities in Uganda, May 2018, Kampala, p.37-38; UNHCR, Operational Update: Uganda: South Sudan Situation, reliefweb, 01-31 December 2017. 15. This includes monitoring maternal mortality and infant-mortality rates. Government of Uganda (Ministry of Health), Uganda National Integrated Health Response Plan for Refugees and Hosting Communities 2018-2023, (Draft) June 2018, p. 21-23 16. Geo-tagged water source mapping may further inform average distance to water source. See: Humanitarian Open Streetmap Team, Community Mapping in North Uganda, 18 January 2018. 17. Anderson, Mary B. Do no harm: how aid can support peace--or war. Lynne Rienner Publishers, 1999. Data Driven Development Response to Displacement Crisis in Uganda: The Displacement Crisis Response Mechanism | 7