The Distance between Us

“An informal discussion between colleagues about the benefits of having a collaborator close by led to agreement that face-to-face interactions – no time-difference, no Zoom meetings – seemed to spur more in-depth exchange,” said Dr Frank van der Wouden, Assistant Professor of Economic Geography and Innovation Studies in the Department of Geography, explaining the origins of the study.

“The pandemic pushed us to limit our face-to-face interactions. In a way, most of the world was, at some stage, collaborating ‘at distance’. Previous research has suggested that proximity leads to more trustful relationships which in turn are more likely to result in knowledge sharing. However, there is little empirical evidence on whether knowledge sharing (or learning) happens more frequently for people who, indeed, are close by.”

Together with Dr Hyejin Youn from the Kellogg School of Management, Dr van der Wouden examined this theory for a population of people who are considered knowledge producers – academic scholars. Scholars collaborate locally and non-locally, and it is possible to track their careers. Using databases from 1975 onwards and involving 17.6 million publications authored by 1.7 million scholars, the team was able to evaluate scholars’ knowledge portfolio, as reflected in their academic output. They then used a clever sampling technique to identify ‘knowledge spillovers’ and analyse whether locally collaborating scholars tend to share and learn more during collaboration, as compared to similar scholars collaborating across distance.

The results show that those collaborating at distance tend to learn significantly less (57 per cent), leading the researchers to conclude that social distancing during the pandemic has very likely limited the flow of knowledge among collaborators.

Insights into the role of information and telecommunication technologies were also revealed. “It has long been argued that developments in these technologies – like email, telephone and the internet – have removed ‘the friction of distance’ on social and economic processes,” said Dr van der Wouden. “Interestingly, our research shows that the positive effects of collaborating locally are now greater than they used to be at the start of our sample in 1975. It has been steadily increasing. Being geographically close matters more, not less, even considering all the recent advancements in these technologies!”

Complex knowledge

Asked to analyse this, he said: “A possible explanation is that novel knowledge is becoming more complex, all the easy stuff is already discovered. To produce complex novel knowledge, we need to specialise and collaborate (like the medical profession does so well – there is a special doctor for everything). To collaborate effectively on novel, complex issues, we need trustful relationships. You will not share your cutting-edge knowledge with people you don’t trust. These relationships are probably not built effectively across digital means. So, the internet still doesn’t substitute for physical interactions and colocation.”

Furthermore, not only does the probability of learning drop with geographical distance but it also corresponds to the number of institutional boundaries crossed during collaboration. “A lot of institutions are geographically bounded,” he said. “For instance, Hong Kong has specific norms, values, but also laws, that govern social interactions. These will differ from New York, or Amsterdam, but will be more similar to those of Shenzhen or Shanghai.

So, as geographical distance increases, more and more of these boundaries are crossed: different cultures, languages and norms affect how to collaborate; different values such as hierarchy in social relationships; and different laws, for example, ‘can we patent this technology?’”

Now that some companies are at loggerheads with their employees about continued working from home, the results have significance. “When people work from home, they don’t build (as easily) the trustful relationships they would when they interact face-to-face repeatedly in the office,” said Dr van der Wouden. “This would mean that working from home results in less knowledge sharing and learning. For businesses, this potentially means barriers not only for creativity and problem-solving but also for talent retention and employer loyalty. Other research has shown, for instance, that people who work from home are less likely to be promoted.”

For academia, he sees two main implications. “First, local and non-local collaborations might serve different purposes. When knowledge exchange is encouraged, local collaboration is best – for instance, between supervisor and PhD student. If the objective is to produce output that relies on complementary knowledge/inputs in which knowledge sharing is not of interest, teams from across the world can be established. Second, to promote knowledge exchange and production, local collaboration is to be favoured over non-local collaboration – all else being equal. This would benefit society-at-large the most.”