How a nation sees its relationship to refugees can have a profound impact on social services and even infrastructure, as shown in new research on Uganda’s unique refugee-hosting model.
By Yang-Yang Zhou
Unlike the US borders many of us are familiar with—fortified by walls, check points, and border patrol officers—the border between Tanzania and Burundi was practically invisible when I visited the Kigoma region of Tanzania in 2015. Only a few small stone markers hidden by grasses and shrubs demarcated the two countries.
Like many African borders, this one had been drawn up by colonialists in Europe. For the Ha, the indigenous people living throughout this borderland, that colonial boundary-making did not affect their daily lives. This group of people continued to use the same markets, intermarry, and have relatives on either side of this border. However, as I soon learned, when cousins from the Burundian side crossed over as “refugees” and were forced to live in camps, then this once-invisible border began to feel more tangible and consequential.
Strangers in a Familiar Land
In the summer of 2015, I visited Kigoma to learn about how government policies shape the ways refugees and locals interact. A few months before I arrived, a political crisis had broken out in Burundi when President Pierre Nkurunziza changed the constitution to run for an additional term. Protests that started in the capital spread into violence in other parts of the country. Carrying the memories of the 1993 genocide, hundreds of thousands of Burundians quickly fled across the border into Tanzania. Although many of these Burundians had family members in Tanzania, the Tanzanian government required them to live in United Nations-branded tents in an overcrowded refugee camp, Nyarugusu, more than eighty miles from the border.
When I began conducting interviews and community focus groups with Tanzanian officials and citizens, I presumed that because the refugees had close ethnic, cultural, and kinship ties to locals, there would be expressions of empathy and solidarity for them. Indeed, several locals did acknowledge these ties. One community elder told me “there is a close relationship between the Burundians and Tanzanians at the border. We have villages on either side of the border that use the same river. Of course we know each other. You will find a Burundian with uncles in Tanzania.”
However, across dozens of interviews and focus groups, the majority of sentiments expressed to me were negative, often portraying the refugees as strangers. One participant said, “These strangers come by illegal ways and try to stay here in natives' houses… That is the kindness of us Tanzanians, others take advantage, we cannot chase them away.” Others spoke of the refugees as criminals and bringers of disease. And many emphasized their national identity as Tanzanians and how that made them distinct from the Burundian refugees nearby.
This surprised me. Why was this border community, which had regular cross-border exchanges, now speaking about their relatives fleeing conflict as if they were strangers—or worse, criminals? Although antirefugee hostility has become commonplace in many parts of the world, coethnicity in this region should have predicted a more hospitable reaction.
Refugee-Hosting and Coethnicity in the Global South
Each year, the number of people experiencing forced displacement—refugees, asylum-seekers, internally displaced and stateless people—has been steadily rising due to new and ongoing conflicts (e.g., in Ethiopia, Afghanistan, Ukraine) and environmental disasters (e.g., recent earthquakes in Turkey). The UN Refugee Agency considers more than 100 million people around the world as forcibly displaced.
This growing number of migrants has unfortunately been met with public backlash. Research from wealthy Western democracies shows that communities hosting more refugees tend to react negatively, such as increasing hate crimes and voting for anti-immigrant far-right political parties. In these contexts, scholars have mostly focused on Latin American and Asian immigrants entering the US and Canada, and on African and Middle Eastern immigrants entering Europe. Xenophobic reactions in these contexts are rooted in stereotypes and fears over racial, ethnic, and religious differences. This problem of social conflict between migrants and host communities has become so pressing that the World Bank has dedicated its 2023 World Development Report to the topic.
Although most migration research has focused on receiving countries in the Global North, border regions of the Global South host the vast majority of people affected by displacement, since most refugees cross the nearest international border. Lower- and middle-income countries are expected to host and integrate larger refugee populations for longer periods of time. While host governments in these contexts tend to be under-resourced, the local host communities typically share ethnic, cultural, and linguistic ties to arriving refugees. Absent cultural differences, we might expect then that refugees would experience a warmer welcome by their coethnics.
And this is largely the expectation of international policy makers. One UN official I interviewed echoed these sentiments: “Depending on the ethnic affinities across the border, in most instances these populations move to areas with similar cultural, religious, linguistic affinities. With the way that the colonial boundaries were drawn in Africa, you find that the populations are very very similar.”
However, my research investigates this assumption. The interviews and community focus groups I conducted in Tanzania, complemented with large scale surveys, reveal that host communities often felt it necessary to distance themselves from the refugees, especially if they are coethnic. Considering how the government has segregated refugees into camps and—along with the media and national society as whole—portrays them as security risks, it is not surprising that these local host citizens, who may be ethnically marginalized themselves, want to reject any association and assert their national identity.
My book project argues that in contexts where refugees (and other migrants) are labeled by political rhetoric and the media as dangerous (e.g., bringing conflict, crime, disease), coethnic host citizens will fear being mistaken as migrants by the state and other non-coethnic citizens. These fears, in turn, will lead to further out-group distancing of migrants even amongst their own ethnic kin.
In addition to the Tanzania case, I draw from original interviews and surveys from multiple contexts—with Somali Kenyans reacting to Somali refugees, Pashtun Pakistanis reacting to Afghan refugees, Colombians living on the Colombia-Venezuela border, and Asian- and Hispanic-Americans reacting to new immigrants from China and Central America—and discover similar dynamics. When coethnic host citizens fear being “migrantized” by majoritized citizens and the state (e.g., the police), they will dampen their ethnic identity, emphasize their national identity, and they may even repeat anti immigrant rhetoric themselves.
So if even coethnic host communities—the very people whom refugees would expect to welcome them—are instead calling them strangers, then are refugees doomed to be rejected wherever they go, or is there another way?
Inclusive Hosting Policies May Foster Kinship
Moving north from Nyarugusu camp in Tanzania to Nakivale, a refugee settlement in Uganda, we can see an alternative pathway. Instead of a camp like Nyarugusu that segregates refugees from local communities and pushes refugees to return, Nakivale largely resembles nearby Ugandan communities. It is not fenced in, and it is dotted with bustling markets and schools, allowing refugees and local Ugandans to interact and live together.
As of 2022, Uganda is the largest refugee-hosting country in Africa and the fourth largest in the world. Uganda is home to over 1.5 million refugees—a substantial number for a country of forty-five million. Most of the refugees in Uganda come from South Sudan, the Democratic Republic of the Congo, Somalia, and Burundi.
Uganda’s model of hosting refugees is considered uniquely progressive for several reasons. First, its open-door asylum policy means that refugees and asylum seekers are not turned back at the border, forcibly repatriated, or deported. Second, Uganda allows refugees to self-settle across thirty refugee settlements or in the capital Kampala. Refugees are permitted to find employment, start businesses, and access local public services, such as schools and health centers. Refugees are even given a small plot of land to grow food. And importantly, politicians and the media refer to refugees as “brothers and sisters,” not criminals and strangers. Although refugees in Uganda still face many challenges, there are notable differences between their treatment and the treatment of refugees in most other countries.
An official from the Office of Prime Minister further explained, “in Uganda, there is high political will, missing in many other countries. You run into our country, we shall open our borders… There is no backlash in hosting refugees in Uganda. The success of Uganda is what we call inclusion and participation of refugees and host communities in development.”
In other words, due to the inclusive treatment of refugees by the Ugandan government, host communities need not fear stigmatization by association. Furthermore, since refugees have access to local public goods, international humanitarian aid from organizations like the World Bank can improve those services that Ugandans jointly use—everyone “wins.” Thus, while we might expect more inclusive migration policies to generate a larger public backlash, that is not what we observe in this case.
The data support this observation. In an article recently published in World Development, Guy Grossman (University of Pennsylvania), Shuning Ge (MIT), and I combine fine-grained georeferenced panel data on refugee settlements, local development outcomes, census, and surveys from the Afrobarometer and the Demographic and Health Surveys (DHS) Program.
We find that across a multitude of public goods and services—access to public primary and secondary school, county health clinics, infant and maternal health services, and roads—parishes with greater refugee presence experience substantial improvements across the years, especially after more refugees from South Sudan entered in the 2010s. Note that these parishes at baseline tend to be worse off, since refugee settlements are typically located in poorer border regions.
Through nationally representative surveys of Ugandan citizens, we also find no evidence of a public backlash against refugees or migration policies in these more affected areas. Thus, our findings show that social conflict between refugees and host communities is not a foregone conclusion. How policies, institutions, and elite rhetoric structure the ways refugees are hosted matter.
Returning to the original research questions that drive my book project: do these same citizens view nearby refugees in an instrumental light, as a means for better public services? Or do they recognize them, particularly if they are coethnic, as their “brothers and sisters”? To answer these questions, we plan to conduct our own panel surveys in Uganda this fall with refugees and host communities. This research has the potential to inform real-world policies that international organizations and host governments enact.
Contributor Bio
Yang-Yang Zhou is an Academy Scholar with The Harvard Academy for International and Area Studies at the Weatherhead Center for International Affairs. She is an assistant professor of political science at the University of British Columbia. Her research interests include migration; refugees; conflict; political economy of development; nationalism; identity politics; and the Global South.
Captions
- The sign welcoming people into Uganda from the border with Rwanda. Cars drive on the right side of the road in Rwanda and need to be reminded to switch to the left side for Uganda. Credit: Bryan Lupton, AP Fellow in Gulu, Uganda July 2009 Partner: Survivor Corps Uganda (CC BY-NC-SA 2.0)
- Number of People Experiencing Displacement in OECD and non-OECD Countries from 1995 to 2020. OECD countries are associated with the Global North, and non-OECD countries with the Global South. Credit: Yang-Yang Zhou. Data source: UNHCR
- Refugee Settlement Locations and Degree of Exposure to Refugees by Parish in Uganda (2020). This 2020 map of Uganda shows refugee settlements in purple, with surrounding host community parishes shaded a darker orange when they experience greater refugee presence. Credit: Yang-Yang Zhou, Guy Grossman, and Shuning Ge. Data sources: UNHCR Uganda, Uganda census
- Marketplace in Nakivale Refugee Settlement, Uganda. The Nakivale settlement in Uganda hosts more than 100,000 refugees from Burundi, Democratic Republic of the Congo, Eritrea, Ethiopia, Rwanda, Somalia, Sudan, and South Sudan, July 25, 2020. Credit: © UNHCR/Esther Ruth Mbabazi