Towards Durable Solutions: Refugee Inclusion in Social Assistance Across L&MICs

In a sweeping new study published by the World Bank, researchers Alfredo Manfredini Böhm, Pablo Acosta, Jeremy Lebow, and Emanuela Migliaccio present a comprehensive analysis of how low- and middle-income countries (L&MICs) are integrating refugees and their host communities into national social protection systems. Drawing on evidence from institutions such as Oxford Policy Management, the Overseas Development Institute (ODI), the Institute of Development Studies (IDS), and the Partnership for Economic Inclusion, the paper offers a timely review of both humanitarian and government-led efforts. The backdrop to this exploration is the 2017 Global Compact for Refugees (GCR), which encouraged donor nations and international organizations to shift from short-term relief to long-term development solutions. With billions invested by the World Bank and the European Union, the question remains: Are these initiatives working?

Cash Transfers Show Promise but Struggle with Staying Power

Central to the report’s findings is the role of cash transfers in improving refugee livelihoods. Evidence from a wide range of settings, including Ecuador, Lebanon, Türkiye, Uganda, and the Republic of Congo, points to consistent, positive short-term outcomes. These include enhanced food security, increased school attendance, reduced child labor, and greater household consumption. Notably, studies in Ecuador linked cash, vouchers, and food assistance with reduced gender-based violence and improved social cohesion. In Lebanon, similar interventions improved school attendance among Syrian refugee children, while Türkiye’s massive Emergency Social Safety Net (ESSN) lifted thousands out of immediate vulnerability.

However, these gains are often fleeting. Many programs witnessed a decline in impact just months after cash assistance ended. Only Uganda’s “Graduating to Resilience” program, which offered integrated support including savings groups and agricultural training, demonstrated robust and sustained results 30 months after implementation. Refugees and host communities alike experienced boosts in income, consumption, and asset ownership. This underscores the need for comprehensive “cash-plus” models rather than short-term transfers alone.

Refugees and Hosts: Uneven Benefits, Shared Needs

The impacts of social assistance programs are rarely evenly distributed. In several interventions, refugees saw greater benefits than hosts, while in others, the reverse was true. Uganda’s model favored host communities in most outcome areas, except consumption. In the Republic of Congo, ethnicity appeared to have a greater impact than legal status, with Bantu communities and refugees benefiting more than indigenous groups. Such variation highlights the importance of context in shaping program outcomes and reinforces the call to involve host communities by default.

Gender dynamics add another layer of complexity. Refugee women and girls face disproportionate disadvantages in education, employment, and personal safety. Cash transfers have shown promise in reducing intimate partner violence and enhancing women’s autonomy, as documented in Ecuador and Bangladesh. Still, these effects depend heavily on context-sensitive design. Programs that ignore cultural constraints risk excluding women altogether or creating unintended tensions. The authors argue that women should be prioritized as direct beneficiaries and that accompanying measures like childcare, access to information, and GBV response services are crucial to inclusive impact.

Implementation in the Real World: Cameroon, Ethiopia, and Bangladesh

Turning from theory to practice, the report explores three country case studies to illuminate operational realities. Cameroon adopted an inclusive, “refugee-agnostic” model in its safety net program, allowing any poor household in target regions, regardless of refugee status, to participate. While this promoted equity, refugee enrollment lagged due to weak outreach and lack of documentation. The project later shifted to simpler emergency cash transfers, which exceeded targets but left more complex support measures under-delivered.

Ethiopia, one of Africa’s top refugee-hosting countries, integrated refugee communities into its urban safety net through the Refugee and Host Integration through the Safety Net (RHISN) program. Despite a supportive legal framework and multi-phase design, rollout was delayed by conflict, administrative hurdles, and staffing shortages. However, with adaptive learning and stronger coordination with the Refugee and Returnees Service (RRS), implementation accelerated, and a landmark randomized controlled trial is now underway to assess the program’s impact.

Bangladesh took a different route, maintaining parallel systems for its over one million displaced Rohingya population. Humanitarian agencies like the WFP led the refugee-facing programs, while government ministries served host communities. This dual structure enabled rapid scale-up but introduced high overhead costs and sustainability concerns. Nonetheless, the intervention delivered income support to 175,000 Rohingya households, exceeding initial goals. The report stresses that while parallel delivery may be expedient during crises, long-term solutions must build toward integration with local systems.

Building Smarter, Shared Systems for the Future

The final section of the paper reflects on the strategic balance between humanitarian and government-led models. Humanitarian agencies are nimble, technically capable, and laser-focused on refugee needs, but their high costs and limited integration capacity restrict long-term reach. Governments offer scalability, legitimacy, and lower costs, but often lack urgency, especially in remote areas. The authors propose hybrid strategies like “piggybacking” and “shadow alignment,” where humanitarian agencies support governments in building capacity and systems compatible with national delivery chains.

To make these hybrid models work, governments must invest in capacity building in underserved regions, identify institutional champions to sustain momentum, and enlist humanitarian partners to support targeting, communication, and grievance mechanisms. At the same time, humanitarian actors should plan for eventual handoff to national systems. These lessons align closely with the GCR’s vision and will be critical as the next wave of investments from IDA-21 and other donors aim to embed refugee protection within broader development agendas.

In sum, the paper makes clear that while challenges persist, the foundation is being laid for a more sustainable, inclusive, and evidence-based approach to refugee assistance in the Global South. With the right blend of political will, technical support, and adaptive design, safety nets can become lifelines not only for displaced populations but for the fragile communities that host them.