(My apologies to Richard III)
Earlier posts have opined on the rise of China’s higher education infrastructure and its increase in research article production in the mid-2010’s, exponential growth that surpassed the volume of US scientific publications.
A report from STINT, the Stiftelsen för internationalisering av högre utbildning och forskning in Sweden
takes advantage of Scopus, a database tracking about 34,000 peer-reviewed journals in life sciences, social sciences, physical sciences, and health sciences. The report studies the period 1980-2021.
First, the report again documented the exponential growth of scholarly publications from China, as shown below in the figure labeled Figure 1. This phenomenon has been the focus of much discussion probing the causes of this change (e.g., incentives to publish), further analysis on the impact of papers (e.g., citation rates), and implications for national competition.
Second, during these years, an important additional trend was developing – greater collaboration among scholars in different countries.
The figure below (labeled Figure 8) from the STINT report, shows the percentage of collaborative publications between US research and colleagues in other countries. The y-axis is the percentage of all papers published by scholars in two given countries that were collaborative across two pairs of country’s scholars.
The first conclusion of the trends is that the tendency of US scientists to collaborate with colleagues in other counties increases during this time period, especially from 1998 onward. There are many changes that took place in this period, including the ubiquitous use of the internet and greater nation-state support of international collaboration.
Notice, however, the sharp decline in the last three years of the USA-China percentages.
Another report, from the US National Center for Science and Engineering Statistics (NCSES) uses data from the same Scopus source. Figure PBS-5, from that report, plots co-authored papers between US scholars and those from other countries. The green line shows the same dramatic increase in US-China collaborative publications as the STINT data, but no dramatic decline over the most recent years that was evident in Figure 8 above.
Why do these two charts imply different trends?
First, the number of scholarly publications is increasing rapidly among Chinese scholars, quite sharp and exponential growth, especially since 2010 or so. Growth in the number of articles by US researchers is also increasing, but at a much lower rate of increase over the years.
The STINT Figure 8 divides the number of co-authored US-Chinese publications by the sum of total publications by US scholars and the total by Chinese scholars. The NCSES Figure PBS-5 divides the number of co-authored publications only by the total number of US authored publications.
When the number of collaborations are divided by the number of total US publications, the percentage of US-scholar articles co-authored with Chinese scholars is not declining. However, when the number of collaborations are divided by the total of US and Chinese scholar articles, the percentages are rapidly declining. In other words, Chinese co-authors are a larger proportion of US-scholars’ work. But US co-authors are a smaller proportion of Chinese scholars work. (Further STINT work shows this clearly.).
By the way, there is some indication that Chinese-European collaboration is also weakening. The STINT report cites new proposed laws in some European countries restricting such collaborations.
Denominators make a difference in conclusions about patterns. In this little example, the country currently exhibiting the largest global research product, China, is doing so with smaller percentages of work collaborative with US scholars. From the perspective of the US scientific product, the percentage of collaborations with Chinese scholars has not declined. But scholarly collaborations have not increased proportionately among US scholars to accompany the rapid growth of Chinese scholarship. Chinese scholars’ product seems less dependent on the US than was true in earlier years.