In 2022 and 2023, US$27.7 billion—4.8 percent of all official development assistance —was spent on refugee situations in low- and middle-income countries. This assistance provides programs and services, and it also shapes how host communities view refugees. Sometimes it helps to build positive relations, and sometimes it generates resentment and tension.
How can policymakers and practitioners use aid to strengthen social cohesion between refugees and hosts? Our research points to two complementary practices. The first is to include host communities in assistance programs. The second is to clearly communicate that those programs are funded by refugee-related aid.
Many aid programs already include host communities. But hosts don’t always know this—or realize that a project exists because their area is hosting refugees. Adding a brief message to make that link explicit, often at minimal cost, can make a significant difference.
We’ve tested this idea in three experiments. The first two are described in our recent Journal of Political Economy article (the open access working paper and initial blog are available here) and the third in a recent working paper. This blog focuses on what we learned about messaging and how changes in communication can meaningfully shift attitudes.
When aid doesn’t look like “refugee aid”
Aid to refugee situations can benefit host communities through improved public services or direct inclusion in (effective) programs. These practices are often meant to promote social cohesion, but their effects are rarely measured—and even visible improvements may not change attitudes.
Sometimes the link is obvious to the host community, like if they visit a clinic inside a refugee camp. Oftentimes, however, attribution is more difficult. When assistance is channeled through governments—as with the World Bank’s Window for Host Communities and Refugees (WHR), for instance—citizens may see the results as ordinary government services. The local school’s operating expenses could be partially paid by WHR funds, for example, without parents knowing the connection to the refugee presence. Costs of hosting (like refugees using public services) are easy to notice; benefits like foreign aid are not.
Even when hosts and refugees participate together in the same programs, hosts may not realize that the initiative exists because refugees live nearby—or worse, may assume that refugees are taking resources meant for citizens. As a result, hosts may benefit materially, yet still hold negative views.
This communication gap is what we set out to test.
The experiments: Testing aid and messaging
We ran three randomized evaluations in Uganda and Kenya to see whether sharing aid and emphasizing its link to refugees could change host attitudes toward refugees.
1. Kampala, Uganda with YARID
Uganda requires that 30 percent of foreign aid for refugees also benefits hosts, but awareness of this “aid-sharing” policy is low—19 percent of our sample at baseline reported that international aid for refugees is shared with Ugandans like them. We worked with the refugee-led organization Young African Refugees for Integral Development (YARID) to test four programs:
- Information about aid-sharing that partially funded local services
- The same information combined with a one-time US$135 cash grant
- The grant with no explicit message about refugees
- Mentorship from an experienced refugee
The first three improved social cohesion, with the largest effects in the second program. Hosts became more supportive of inclusive policies for refugees (e.g., right to work, freedom of movement, hosting more); more positive about their economic effects; and more comfortable socially (e.g., willingness to have a refugee neighbor). The effects were large—an 11-18 percent increase in supporting refugees’ right to work, for instance—and persisted for more than two years. Mentorship effects, in contrast, faded over time.
2. Western Kenya with RELON Kenya
We then replicated the main findings with the Refugee-Led Organization Network of Kenya (RELON), in a context where baseline support for refugees was lower. A short video message added to small grants—US$10—again increased support for inclusive policies toward refugees.
3. Kampala, Uganda with the IRC
Our third trial, implemented by the International Rescue Committee (IRC), tested larger grants (US$540) and contact between refugees and hosts through mentorship groups. Again, the economic intervention with clear messaging significantly affected social cohesion (here we did not isolate the messaging component specifically), while the mixed groups showed no additional benefits.
Across the experiments—covering two countries, urban and rural settings, and dissemination through conversations and videos—the findings were consistent: aid-sharing helps, and communicating the connection with refugees helps even more.
What to say: A simple script
What information is important to convey? The full scripts from each experiment are available here, and this excerpt from the third experiment illustrates some ideas we think are key:
Our programs work in places that host refugees. We think it’s important to support and host refugees because we want everyone in our community, including refugees, our families, our friends, and our neighbors to be treated with compassion and not feel excluded or suffer discrimination.
The IRC started this program because refugees live here in Kampala, and we want both refugees and Ugandans who live in Kampala to benefit. Refugees and Ugandans are participating in this program, both as mentees and as mentors.
This project is part of the international donations that are shared between refugees and hosts in Uganda. In Uganda, more than 30% of foreign donations for refugees go to supporting Ugandans. In addition to small businesses, these donations are used to support schools and hospitals in areas where there are many refugees, including Kampala. International donors support these buildings and services because Uganda is a generous host to many refugees. Refugees are allowed to live and work in Kampala, and more donations can go to Ugandans because refugees can earn an income.
In short, the message acknowledges the community’s generosity in hosting refugees; explains that the program is in Kampala because refugees live there too; highlights that refugee-related aid is shared with hosts; and connects the program to Uganda’s inclusive policies. Messaging can be delivered through different channels, adding little cost but significant potential for impact.
Shaping positive-sum perceptions
The broader message we believe aid-sharing conveys is one of fairness. Aid in refugee situations often triggers resentment because hosts perceive refugees as receiving more support. Our evidence suggests that sharing aid with hosts helps counter this. Communicating that link clearly goes one step further, reframing refugee-related aid as fair and mutually beneficial.
For programs already including hosts, intentional messaging is likely an easy win: a short script at a project launch can make a measurable difference. For investments in public services, a similar information session at a village meeting or an opening ceremony, perhaps combined with an advertising and/or social media campaign, are some ideas to potentially reproduce that effect.
Hosting refugees doesn’t have to be perceived as zero-sum. When hosts see that aid strengthens their own communities and understand its connection to refugees, we see that support for refugees increases. Humanitarian and development actors can achieve important gains in social cohesion by conveying this message.
