Few are as vulnerable to disaster risks as the over 121 million people across the globe who have been forcibly displaced due to conflict and violence. For host countries, responding to the needs of the forcibly displaced in a way that accounts for those risks remains a persistent challenge. This is made even more difficult by the fact that host countries, nearly three in four of which are low- and middle-income countries, often have limited capacity to absorb the shock brought about by a sudden influx of forcibly displaced people. 

A case in point is the current situation in Burundi, a country that has long grappled with fragility, conflict, and violence (FCV). Most recently, between January and April 2025, Burundi saw the arrival of over 70,000 people fleeing the armed conflict in the Democratic Republic of Congo (DRC). The Global Facility for Disaster Reduction and Recovery (GFDRR) and the World Bank have supported analytical work that has painted a fuller picture of the displacement situation in Burundi, including the interaction between displacement and disaster risk. The main goal was to support Burundi and development partners in responding to the crisis, including by ensuring that resources are directed where they are needed most. 

For starters, the technical team analyzed social connectedness— the extent to which people are connected within their social networks—to estimate the likely number of displaced people in different areas of Burundi. Extensive research has shown that social connectedness is among the most influential factors driving the movement of people. 

By analyzing social media data from Facebook Data for Good Program, particularly the Social Connectedness Index, the team was able to estimate that, as of May 2025, Bujumbura Province was likely to host around 24,000 displaced people because of its high social connectivity with the DRC as well as its proximity to the border. Drawing on the same analysis, the team was also able to estimate the likely number of displaced people in four other provinces of Burundi. A partnership with Harvard University’s CrisisReady initiative enabled the integration of data from Facebook’s Data for Good Program into the analysis. 

In addition, the technical team also drew on an analysis of geospatial risk data to assess the level of risk faced by displaced people across provinces. Alongside natural hazards such as floods and landslides, FCV risks were also assessed. These assessments have been visualized through maps that co-locate risks from natural hazards with FCV risks. To cite just one of the findings, the team found that while displaced people in Ruyigi and Cankuzo provinces are likely to face very low risks from floods and landslides, by comparison, people from the capital city of Bujumbura and its outskirts are likely to face high risks from both of those natural hazards. The team also found that displaced people in urban areas such as the city of Bujumbura are likely to face higher risks of violence.