The roles and limitations of data science in understanding international migration flows and human mobility

Understanding where and why international migrants move, and how the size and composition of migratory flows change over time, has long preoccupied scholars and policymakers. Having a clearer assessment of international migration patterns allows for better planning to support responses as well as insights into how policies can act to shape migration manifestations. As the volume of data generated globally emanating from the expansion of new forms of technology skyrockets, the pressure to quantify every aspect of people’s day-to-day lives intensifies. We are also increasingly witnessing the move from evidence-based policy to data-driven policy, however, in moving in this direction we argue that we are becoming more and more distant from the underlying meaning of data variables, thereby increasing the risk of ineffective data-driven policy that misunderstands and misinterprets social phenomena such as migration. We step through key issues of migration data science highly relevant to policy processes, with particular reference to international migration flows and its intersections with human mobility. We find that while the ongoing search for improved understanding of migration flows remains a pressing issue in scholarly and policy spheres, new technologies are increasingly capturing mobility rather than migration data. There is a need to engage in a ‘back to basics’ approach whereby we critically examine the underlying normative and conceptual settings so that data science is better able to navigate the issues and dilemmas that arise from human mobility analytics, especially those concerning vulnerable groups.