Improving Refugee Integration through Data-driven Algorithmic Assignment

Kirk Bansak, Jeremy Ferwerda, Jens Hainmueller, Andrea Dillon, Dominik Hangartner, Duncan Lawrence, and Jeremy Weinstein

Science Volume 359, Issue 6373, 19 January 2018


When refugees are resettled in third countries, resettlement countries do not fully leverage the factors that promote refugee integration such as: (1) geographical context (e.g. economic and social opportunities available in resettlement locations); (b) personal characteristics of refugees (e.g. country of origin, language skills, gender, age, education); and (c) synergies between geography and personal characteristics (e.g. expected employment returns associated with personal characteristics can vary across different resettlement locations). The authors developed a data-driven algorithm that assigns refugees across resettlement locations to improve integration outcomes by leveraging synergies between refugee characteristics and resettlement sites. The algorithm was tested on historical registry data from two countries (United States and Switzerland) with different assignment practices and refugee populations. The proposed approach led to gains of 40-70 percent in refugees’ employment outcomes relative to current assignment practices. The authors suggest that this approach can provide governments with a practical and cost-efficient policy tool that can be immediately implemented within existing institutional structures.