Nowcasting Daily Population Displacement in Ukraine through Social Media Advertising Data

Douglas R. Leasure, Ridhi Kashyap, Francesco Rampazzo, Claire A. Dooley, Benjamin Elbers, Maksym Bondarenko, Mark Verhagen, Arun Frey, Jiani Yan, Evelina T. Akimova, Masoomali Fatehkia, Robert Trigwell, Andrew J. Tatem, Ingmar Weber, and Melinda C. Mills

Population and Development Review, Volume 49, Issue 2 (2023), Pages 231-254 


The Russian invasion of Ukraine in February 2022 triggered the rapid displacement of millions of refugees into neighboring countries and the displacement of millions of Ukrainians within the country. In conflict and crisis situations such as this one, representative survey data quickly become outdated, primary data collection is challenging if not impossible, and the dynamic nature of population changes requires high-frequency measurement not suited to traditional data gathering techniques.  

This study uses aggregate data from Facebook’s marketing tools, accessible via its marketing application programming interface (API), to estimate daily population sizes and internally displaced populations within Ukraine. Facebook’s marketing API provides estimates of current audience sizes for targeted advertising on the social media platform, including counts of daily and monthly active users within specific age-sex demographic groups and subnational geographic areas.  

The authors combine Facebook data with pre-conflict population data and daily counts of border crossings out of and into Ukraine to estimate: (1) daily population sizes for age-sex demographic groups within subnational administrative units (Oblasts) of Ukraine; (2) daily net changes in these populations relative to pre-conflict baseline population estimates; and (3) the total number of people internally displaced away from their original Oblast each day. 


  • National total internal displacement increased sharply after the Russian invasion on February 24 reaching 5.3 million people by March 14. Inter-Oblast displacement fluctuated between 5 million and 6 million thereafter reaching a peak of 6.2 million people on June 21. This national metric of internal displacement was sensitive enough to detect key events, such as the evacuation of Khersonska Oblast and mass returns of people to their home Oblasts during Orthodox Easter. 
  • Sub-national displacement and demographic patterns quantify large-scale evacuations of major cities in the first few weeks of the conflict, such as Kyiv and Kherson, and east-to-west movements of displaced persons during this period. They also capture the distinct movements of women and children and of men in the initial months following the invasion. 
  • Demographic patterns of internal displacement reveal (1) areas subject to large-scale evacuations where there were reductions in all or most age-sex demographic groups; (2) refugee staging areas (Oblasts with preferred international border crossings) where there were population increases across all demographic groups, but particularly women and children; (3) internal safe havens for nonrefugees (Oblasts with relatively few conflict events but without preferred international border crossings) where men and retirees tended to increase while women and children decreased or remained constant; and (4) irregular dynamics in some Oblasts. 

This work highlights the value of nontraditional data for nowcasting (i.e., estimating in near real-time) population dynamics at high frequency to complement existing data sources and support targeted humanitarian assistance in response to a crisis.