Digital and Computational Demography
At a Glance
Studying International Migration of High-Skilled Professionals Using Large-Scale Digital Trace Data
Daniela Perrotta, Diego Alburez-Gutierrez, Tom Theile, Emilio Zagheni; in Collaboration with Carlos Callejo Penalba (Aalto University, Finland), Kiran Garimella (Massachusetts Institute of Technology, Cambridge, USA), Ingmar Weber (Qatar Computing Research Institute (QCRI), Doha, Qatar)
High-skilled migration represents an important and increasingly large component of international migration streams, with substantial socioeconomic implications for both sending and receiving countries. Standard data on migration are often costly, coarse-grained, and inconsistent across countries. As a result, they are typically inadequate to improve our understanding of high-skilled migration. Alternative data and methods are thus needed to routinely capture the ever-growing and ever-changing phenomenon of high-skilled migration and to provide a clearer picture of the spatiotemporal and sociodemographic aspects of the international migration of professionals. In this context, digital trace data from social media implicitly provide a large amount of information that can be exploited to identify mobility patterns at an unprecedented scale.
In this project, we retrieve digital trace data from the social networking service LinkedIn. The aim is to capture the migration patterns of professionals across countries in Europe. Specifically, we use the advertising platform LinkedIn Ads, which provides an estimate of LinkedIn users who meet specific targeting criteria, such as gender, age group, city, industry, and level of seniority. This enables us to retrieve the trajectories of high-skilled LinkedIn users from the place where they studied and from the place where they currently work (according to their LinkedIn profiles) across different groups determined by age, gender, and industrial sector. Remarkably, LinkedIn data provide us with high spatiotemporal resolution data and additional insights that would otherwise be difficult to obtain, such as skills, interests, and years of experience. But there are also challenges as digital data are generally characterized by systematic biases, such as self-selection and measurement biases, which must be taken into account and corrected.
We carefully analyze LinkedIn data to extract relevant information on high-skilled migrants and to complement the new data thus gained with traditional data sources. Furthermore, we aim at developing mobility models to assess the extent to which there are imbalances in migration flows across countries, also with respect to different sociodemographic characteristics (including age and gender) or the field of occupation. We expect to be able to evaluate the impact of socioeconomic, political, and geographical factors in shaping flows of professionals in Europe and to contribute to the theories on the causes and consequences of high-skilled migration.
Economics, Employment, Retirement, Migration