Exploring the changing geographical pattern of international scientific collaborations through the prism of cities

Abstract

Science is becoming increasingly international in terms of breaking down walls in its pursuit of high impact. Despite geographical location and distance still being major barriers for scientific collaboration, little is known about whether high-impact collaborations are similarly constrained by geography compared to collaborations of average impact. To address this question, we analyze Web of Science (WoS) data on international collaboration between global leader cities in science production. We report an increasing intensity of international city-city collaboration and find that average distance of collaboration of the strongest connections has slightly increased, but distance decay has remained stable over the last three decades. However, high-impact collaborations span large distances by following similar distance decay. This finding suggests that a larger geographical reach of research collaboration should be aimed for to support high-impact science. The creation of the European Research Area (ERA) represents an effective action that has deepened intracontinental research collaborations and the position of the European Union (EU) in global science. Yet, our results provide new evidence that global scientific leaders are not sufficiently collaborative in carrying out their big science projects.

Scientific collaboration, research funding, and novelty in scientific knowledge

“…Due to growing specialization in science labor, decrease in communication and travel costs, and the need to access interdisciplinary ideas and various database [ 10 , 11 ], the number of internationally co-authored papers has grown significantly over the last three decades [ 8 , 15 ]. Internationally co-authored papers take up to about 25% of WoS articles [ 7 ], and the increase in published outputs from Western Europe and United States is primarily driven by the growth of international collaborations [ 6 ].…”

“…Consequently, some studies tried to take account of intercity co-authorships both within and between countries [ 35 – 38 ]. However, most of the studies that deal with city level collaborations focused only on large or top publishing cities, and still, it needs more investigation since collaborations at city and institutional level have been analyzed not much due to data collecting and processing availability [ 15 ].…”

“…Similar results are found at city and institutional level studies. For instance, Csomós , Vida , and Lengyel [ 15 ] showed a positive correlation between a Jaccard index to measure relative strength of intercity collaborations and a binary variable of highly cited papers. Moreover, Abbasi and Jaafari [ 41 ] explored the correlation between institutional collaboration types and their impact on research.…”

“…In fact, previous literatures have shown that co-authored research papers tend to have higher impact when measured by the number of citations a paper gets [ 8 , 12 – 14 ]. Yet, it is only recently that researchers have started to pay attention to city-level or institution-level scientific collaborations due to the availability of collecting and processing big publication data [ 15 ].…”

Geographies of the global co-editor network in oncology

“…In the past few decades, an increasing number of countries (including post-socialist, newly industrialized, and many developing countries) have entered the arena of global science. [ 3 ]. Besides experiencing a robust increase in international scientific collaborations [ 4 , 5 ], global science has become even more multipolar [ 6 ].…”