Openness to migrate internationally for a job: Evidence from LinkedIn data

This repo contains the code required to process and model data acquired from the Linkedin Recruiter platform. More details on the rationale and preliminary results below. The following authors were involved in this project: myself, André Grow¹, Daniela Perrotta¹, Tom Theile¹, Helga de Valk^2,3, and Emilio Zagheni¹.

¹Max Planck Institute for Demographic Research, ²Netherlands Interdisciplinary Demographic Institute, ³University of Groningen

International migration is an important phenomenon with measurable demographic consequences. Much migration research, though, is plagued by inadequate data and leaves much to be explained as to why people choose to move – or stay – and where they move to [1]. Recent theoretical advancement emphasizes a "two-step" approach to the study of migration, called the "aspiration-(cap)ability" framework, where the decision-making process prior to migration is separated from the migration event (or lack thereof) [2, 3]. In addition to this theoretical advancement, the number of surveys that focus on people who have a migration "aspiration" has increased over the last decade, also enabling empirical advancement [4].

While traditional surveys of migration aspirations are of great use for theory development, they typically have several shortcomings, including an often narrow scope, spatio-temporal restrictions, and lack of inter-survey comparability. To address some of the limitations of existing survey data and to complement the growing literature in the field, we offer new analyses that rely on a novel – and as of yet untapped – data source: aggregate-level information on LinkedIn users open to work-related relocation, as obtained from the LinkedIn Recruiter platform.1 Compared to traditional survey data, the LinkedIn data that we use to study openness to migration are less expensive to collect, are continuously available, have consistently defined variables across 24 different languages, and provide a global snapshot of openness to migration as recent as the latest update to a person’s LinkedIn profile. Furthermore, rather than directly asking a person about their migration preferences, we capture routine job-seeking tendencies, thus removing reactive responses to the survey instrument.

Figure 1: Chord diagram showing the number of LinkedIn users open to relocation between two countries, aggregated to regions. Direction is indicated by the arrowhead and size by the width at the base of the arrow. Proportions have been scaled within each region, with the total rounded to the nearest thousand. As an illustrative example, approximately 30% (800,000) of those currently located in Europe are open to relocating to another country in Europe [5].

From the LinkedIn Recruiter platform, we collected web-scraped aggregate-level data on the aggregate number of people open to relocating for a job in a country that differs from their current location, for a total of 191 prospective countries and territories, every two weeks from July 2020 onward. The data collection process involves a separate search query for each destination country and returns the top 75 current user locations (ranked by number of users). Figure 1 shows a global overview of the size and direction of people open to relocating between regions, showcasing geographic availability and heterogeneity. Interpretation of these raw counts make inter-country comparisons challenging, as differences could be driven by, for instance, the difference in LinkedIn users in a country. Therefore, to assess the relative attractiveness of certain countries, we used a gravity-type model to standardize across a set of geographic and linguistic factors [6]. We focused on Europe in part due to the large number of those open to relocating to and within Europe, but Europe is also an interesting case study due to the establishment of free movement of European Union (EU) nationals across borders. From this, we identified specific countries where openness to migration is different from what we would expect from the model alone, thus providing a potential indicator of future flows of migrants in Europe. Future work can delve into other country-level factors driving attractiveness amongst job-seekers, which can provide information and different scenarios on what the future of these countries could look like if people decided to move.

Our study offers a novel view on openness to international migration by capturing the routine jobseeking behavior of those on LinkedIn. Despite the niche these data can fill, there are a number of challenges which fall into one of two categories: 1) accounting for biases which may distort results from a population of LinkedIn users that is not representative of the general population and 2) adapting a platform designed for professional networking for use in demographic research. With careful consideration of these limitations these data can be used to understand the relationship between economic development and migration amongst job-seekers, stratified by industry, level of education, or field of study, to name a few. In combination with further analysis, these global, continuously collected data can shed light on the extent to which openness to relocation translates into migration decisions and potential future demographic scenarios if these professionals who are open to moving choose to do so.

References

1. Willekens, F., Massey, D., Raymer, J. & Beauchemin, C. International Migration under the Microscope. Science 352, 897–899. ISSN: 0036-8075, 1095-9203 (May 2016).
2. Carling, J. & Schewel, K. Revisiting Aspiration and Ability in International Migration. Journal of Ethnic and Migration Studies 44, 945–963. ISSN: 1369-183X, 1469-9451 (Apr. 2018).
3. de Haas, H. A Theory of Migration: The Aspirations-Capabilities Framework. Comparative Migration Studies 9, 8. ISSN: 2214-594X (Dec. 2021).
4. Aslany, M., Carling, J., Mjelva, M. B. & Sommerfelt, T. Systematic Review of Determinants of Migration Aspirations External Series 2,2 (University of Southampton, Southampton, 2021).
5. Gu, Z., Gu, L., Eils, R., Schlesner, M. & Brors, B. Circlize Implements and Enhances Circular Visualization in R. Bioinformatics 30, 2811–2812 (2014).
6. Cohen, J. E., Roig, M., Reuman, D. C. & GoGwilt, C. International Migration beyond Gravity: A Statistical Model for Use in Population Projections. Proceedings of the National Academy of Sciences 105, 15269– 15274. ISSN: 0027-8424, 1091-6490 (Oct. 2008).