The Effect of Coauthor Networks and Institutional Context on Scholarly Mobility

Alexandra Rottenkolber, Linköping University
Ola Megahed, Complexity Science Hub Vienna
Gergely Mónus, Corvinus University of Budapest
Jiaxuan Li, Oxford Univsersity, Kellogg College
Jisu Kim, Max Planck Institute for Demographic Research (MPIDR)
Daniela Perrotta, Max Planck Institute for Demographic Research (MPIDR)
Aliakbar Akbaritabar, Max Planck Institute for Demographic Research (MPIDR)

Personal and institutional connections shape an individual's opportunity space. This is especially true for migration, where connections provide information during the process and social support after arrival. Connections also influence the direction of migration by helping individuals explore opportunities without physical proximity. However, it is not clear how different forms of connections interact with each other in their effect on migration. Our study aims to disentangle the effects of individual (personal connections) and institutional (relations between workplaces, schools, etc.) networks, by analyzing international scholarly migration decisions. Scholarly migration serves as a perfect petri-dish for analyzing network effects in migration, because we can reconstruct individual migration trajectories, individual co-authorship networks, and institutional relations from digital trace data. Using data from Scopus we reconstruct the mobility histories of 1.2 million scholars. With multinomial logistic regressions and discrete choice modeling, we find that both individual co-authorship networks and established mobility pathways between research institutions influence scholarly moves. However, with larger individual networks, the importance of the location of co-authors increases when deciding the direction of moves. However, the weight of different connections is not homogenous across the globe, the prestige of sending and receiving institutions interacts with network effects.

Keywords: International Migration, Computational social science methods, Social network methods

See extended abstract.