Monitoring Cross-Border Refugee Flows from Sudan Using Digital Trace Data

Jackson Mason-Mackay , Max Planck Institute for Demographic Research and University of Manchester

Forced migration is one of the fastest-growing forms of human mobility, yet it remains understudied in demographic research. This paper explores whether digital trace data can serve as an indicator of refugee flows in data-scarce environments, focusing on cross-border displacement from Sudan into neighbouring Egypt and Ethiopia since internal conflict began in April 2023. Most movement out of Sudan involves overland travel, and refugees commonly use online platforms during their journey. I compare cross-border migration records (IOM and UNHCR, April 2023 to May 2025) with patterns of online interest for locations along migration corridors, as captured by page views on Wikipedia (filtered by language) and search volume on Google Trends (filtered by country). Weekly crossings of Sudanese nationals into Ethiopia are positively correlated with Wikipedia views for transit towns along the Sudan-Ethiopia corridor (r = 0.47-0.55, p < 0.001), specifically in Arabic, with weaker or insignificant associations for English, Amharic, and Russian. Granger causality tests found that Wikipedia views for locations on the Sudanese side of the border are predictive of border crossings (1 week lead, p < 0.05), while border crossings predict page views for locations on the Ethiopian side (1–2 week lag, p < 0.05). Preliminary analysis of the Sudan-Egypt corridor found a sharp rise in information-seeking for transit towns at the onset of the conflict - both for Wikipedia views (in Arabic) and Google searches (made in Sudan) - with online activity rising and falling with the first wave of Sudanese refugees into Egypt (April-June 2023).

See extended abstract

 Presented in Session P4. Migration, Migrants, and Mobility