The impact of geographical distance on learning through collaboration

There is little question that new communication and transportation technologies have effectively "shrunk the world" for a great many activities. At the same time, the "death of distance" has been greatly exaggerated, especially in fields such as academic scholarship and knowledge development where the positive benefits of knowledge spillovers remain highly distance dependent. We analyze 17.6 million publications authored by 1.7 million scholars to examine how knowledge spillovers between scholars collaborating at different geographical distances impacts their future knowledge portfolios. Our results show that in 1975, scholars collaborating locally were 57 % more likely to learn from knowledge spillovers than similar scholars collaborating non-locally. We identify four factors that structure these findings. Individuals deriving the greatest learning premiums from local collaboration tend to be (1) in earlier stages of their career; (2) associated with lower-ranked institutions; (3) working with fewer collaborators; and (4) in STEM fields. The probability of learning drops with geographical distance and correspond to the number of institutional boundaries crossed during collaboration. We conclude that even in the 21st century, geographical distance still negatively impacts knowledge spillovers through collaboration. These findings have implications for debates in innovation and management studies concerning knowledge spillovers, the spatial organization of (knowledge-intensive) economic activity, regional innovation policies, structuring team-work and working-from-home vs. returning to office.