This paper argues that differences in citation flows between countries capture a growing global inequality in scientific knowledge production. Using data from the Microsoft Academic Graph, the authors introduce a framework called 'citational lensing' to identify citational distortions – instances where citations are more or less abundant than expected given textual similarity between research outputs. Findings reveal that leading countries increasingly receive more citations than expected, while peripheral countries are undercited, posing challenges to the growth of novel ideas.

Charles J. Gomez, Andrew C. Herman, Paolo Parigi