Researchers analyze citation data to document trends in scientific migration and collaboration
Researchers at Elsevier, the academic journal publisher, have used bibliographic data to identify trends in scientific mobility and collaboration across 17 countries. Tracking unique author IDs, they documented authors’ movements from one country to another and identified rates of co-authorship between scientists from different countries. They found that migration and co-authorship are distinct trends, driven by different factors. For example, shared language and geographic proximity drive rates of migration more strongly than they influence rates of co-authorship between countries. The authors also note that political tensions have less of an effect on migration than they do on co-authorship. Relative to rates of co-authorship, the researchers found high rates of migration from Taiwan to China, Iran to the United States, and between India and Pakistan.
In conducting their analysis, the Elsevier researchers mined the Scopus database, which includes bibliographic data for about 20,000 journals (about a fifth of which are published by Elsevier). They considered the research output from 2000-12 for scholars from 17 countries, including 10 emerging and 7 established research powers: Brazil, China, Egypt, India, Iran, Malaysia, Pakistan, Portugal, Romania, and Thailand fall into the former category, and Australia, Germany, Italy, Japan, the Netherlands, the United Kingdom and the U.S. into the latter. The dataset included 100,830 authors from the 17 nations. The Elsevier researchers estimate an error rate below 10 percent (due to errors in Scopus author profiles).
They found that some countries, including Iran, Malaysia, Pakistan, and Thailand, show high levels of temporary, as opposed to permanent, migration. In other words, large numbers of scientists leave the home country and then return. This is good news given longstanding concerns about the problem of brain drain. As the Elsevier researchers, Gali Halevi and Henk F. Moed, write in a recent article in the publication Research Trends, “This type of migration supports the development of the country’s professional skills levels and infrastructure.” By comparison, migration from Germany, India, Japan and the U.S. is more likely to be permanent in nature. Countries like Australia, Brazil and China are in the middle, with a more even balance between permanent and temporary migration.
China and the U.S. both have comparatively low rates of outward migration. Of the 17 countries studied, the U.S. has the largest percentage of researchers who remain in the home country, followed by China. Both nations are important destinations for researchers from abroad. Countries with the highest levels of migration to the U.S. include Australia, Brazil, Canada, China, India, and the U.K., as well as other countries in the inner circle in the first figure below, while countries with the highest levels of migration to China are Australia, Canada, Japan, Singapore, South Korea, Taiwan, the U.K., and the U.S. In the following images, courtesy of Elsevier, the strength of the migration relationship is indicated by distance from the center, with those countries with the lowest levels of migration to the country in question farthest away.
The Elsevier researchers note that while common language and geographic proximity generally drive migration, this is not the case with scientists from the U.K. and the U.S. going to China and vice versa. They also note that China serves as a destination and hub for scholars from throughout Asia.
Halevi, a senior research analyst and program director at Elsevier, said she and Moed would next like to break down the data by discipline to determine if some countries are serving as hubs for researchers from particular fields.
“We found the [broad] trends and that’s great, but I think we need some explanation as to why,” Halevi said. Take researchers from China coming to the U.S.: “What makes them come here? Their country is rich with resources. Do certain research areas attract them to the U.S. and vice versa?”
Such data would be informative for policymakers, Halevi said. "It will make them think, so, we are a center for bioengineering [for example], but our researchers are pulling away when it comes to X, Y, and Z [fields]: what can we do to attract them back?"