Since 2016, the year Donald J Trump was elected president of the United States, scientific cooperation between that country and China has markedly decreased, says a new paper from the Cambridge, Massachusetts-based National Bureau of Economic Research.
Building a Wall Around Science: The effect of US-China tensions on international scientific research (“Building a Wall”) examines quantitative data in three areas: STEM trainee mobility (ie, the likelihood of a Chinese student doing his or her PhD in the US and remaining afterwards); the likelihood of Chinese researchers in China using scientific research originating from the US and vice versa; and scientific productivity as measured by production of peer-reviewed papers.
In almost every area, the study shows that “geopolitical tensions at a much lower level than the formal expulsion of academics or violent warfare can lead to a significant shift in scientist mobility”.
According to Professor Britta Glennon, who teaches management at the Wharton School (University of Pennsylvania) and is one of the study’s four co-authors, “Building a Wall” was undertaken to examine the state of internationalisation of science.
“Over the past few decades, science has become more international across many dimensions. Science used to be concentrated in the West, but today there is a much more international scientific community.
“For instance, many academics are immigrants, so right there you have a very international community. There are many more international collaborations than there used to be. And, increasingly, citations in papers are from papers produced in other countries.
“Our study demonstrates in quantifiable terms how this large international scientific community is affected by the geopolitical tensions between the United States and China, which, from a scientific standpoint, are the major players.
“We wanted to understand how a reversal in the trend of growing internationalisation impacts science,” said Glennon.
Anti-Chinese sentiment
For their analysis of mobility using the Open Research and Contributor ID (ORCID), on which academics post their curricula vitae, Glennon and her colleagues constructed a database of 836,495 CVs in STEM fields (out of ORCID’s 14 million CVs).
Using machine learning methods that infer ethnicity from names, they were able to classify scientists as being “ethnically Chinese”, which was the treatment group. The control group for those analyses was the non-ethnically Chinese complement, which was drawn from graduate students and professors in UK universities.
The data shows that in between 2016 and 2019, ethnically Chinese students were 16% less likely to attend a US PhD programme than were the control group and more likely to enrol in Anglophone universities in other countries, such as Britain, Canada or Australia.
While “Building a Wall” does not include a qualitative component, an explanation of the empirical political-social position Chinese nationals and, indeed, Chinese Americans, have faced since 2016 contains several strong suggestions as to why a large number of Chinese scholars now shy away from studying in the United States.
The Pew Research Center reported in 2020 that anti-Chinese sentiment had risen 11 percentage points, from 55% to 66% in the five years since Trump began his first campaign for the presidency, which included significant anti-Chinese rhetoric, and the COVID-19 crisis, which Trump blamed on the Chinese calling it “the China virus”.
Glennon and her colleagues also reference the 2021 study, “Racial Profiling among Scientists of Chinese Descent and Consequences for the US Scientific Community”, that found that Chinese scientists “reported considerable fear of US government surveillance” at almost five times the rate of non-Chinese scientists: 50.7% vs 11.7%.
Trump’s ‘China Initiative’
An additional chill discussed in “Building a Wall” was Trump’s “China Initiative”.
Announced by then attorney general Jeff Sessions on 1 November 2018, the China Initiative was the nation’s premier effort to thwart and punish industrial and academic researchers who funnelled economic and military research to China. It was an essential part of former US president Donald Trump’s crackdown on Chinese scientists.
Eighteen months after the China Initiative began, some 5,000 agents were assigned to the initiative tasked with preventing China from stealing technologies that were vital to America’s economic and military interests.
In July 2020, FBI director Christopher A Wray stated that the bureau was “opening a China-related counterintelligence case every 10 hours”.
As of September 2021, federal prosecutors had charged 28 researchers under the China Initiative. Of these, there had been about a dozen convictions or guilty pleas. Of the dozen or so Chinese professors or professors of Chinese descent, the government had convicted only four – none for espionage or theft of trade secrets or intellectual property.
President Joe Biden’s justice department closed down the initiative in February 2022.
Although Glennon and her co-authors focus on American policy, they note that Chinese international and domestic policies have also contributed to the disruption of international scientific exchange.
While China’s president Xi Jinping has not repudiated the US-China Science and Technology Cooperation Agreement signed in 1979, his policies, including the Made in China 2025 strategic plan, which is designed to push China to achieve independence from foreign suppliers in a number of strategic areas, is cited in the report as an example of Chinese policies that lessens the flow of students to the United States.
The chilled environment has resulted in a 3.6-point drop in probability of Chinese scientists who stay in the United States after graduating from their PhD programmes, which, Glennon explained, works out to a 4% decline relative to the mean. In some cases, these students move to other Anglophone countries like the UK, Australia and Canada.
“From the US perspective, you are losing a lot of talent,” said Glennon.
“There’s substantial literature that shows the huge innovative impacts of immigrants and the spillover effects they have on ‘native scientists’. Immigrants tend to produce more patents; they tend to produce more startups; they tend to be more innovative – and they cause Americans to be more innovative,” she added.
Going to international conferences and reading journals is not the same, she explained, as having these scientists at Stanford University, the Massachusetts Institute of Technology (MIT) or the University of Chicago.
Impact on US citations
The second area “Building a Wall” examines is if the geopolitical reality affected “cross border knowledge”. The database for this part of the article was constructed from Dimensions, a database that contains more than 140 million publications and 1.8 billion citations in the bibliographies of those publications.
“The database has all kinds of information: Who are the authors of an article? What institutions are they affiliated with? What fields do they work in? What journal is the article published in? Who’s citing them? Taken together, this information allowed us to examine both productivity and track cross border knowledge transfers,” Glennon told University World News.
To determine if China-based scientists changed their usage of US produced science, they examined the references of 2.1 million publications from Chinese research teams.
The study documented “a sharp decline in Chinese reliance on US science as measured by citations. Specifically, among Chinese publications, the share of references citing US research declined by about 4-6% after 2016.
“The impact is even more striking for the share of citations to recently published articles, where the share of citations on Chinese publications to US publications declined by 10-12%”.
Interestingly, the database of 2.8 million publications by US and UK scientists showed that US scientists citing Chinese publications has not declined.
Productivity impacts
The final part of “Building a Wall” examined productivity. The researchers drew another subset from Dimensions from publications between 2008 and 2019. Like the other subsets, this was of STEM publications. Using the authors’ addresses at the time they published, the researchers were classified as being US or China-based.
US-reliant researchers in China (the treatment group) did not on average experience any productivity decline relative to UK-reliant researchers in China (these last were the control group), but there was some variation by field, Glennon explained.
Engineers registered the largest productivity loss -5.7% while physicists saw a 17.5% increase in productivity, as measured by the number of publications. Other STEM fields registered no significant difference.
The sample of US-based researchers in this category contained 231,296. The researchers were able to determine that 29,587 of these were ethnically Chinese and 129,032 were “unique non-ethnically Chinese individuals”.
“The average US-based ethnically Chinese researcher experienced a 2% decrease in overall productivity and a 6% decrease in production of impact-weighted publications in US journals relative to their non-ethnically Chinese colleagues,” states the paper.
Downstream effects
Glennon stressed that there are two points that the study did not look at but are important to examine.
The first is the welfare effect, taking into account national security concerns. The history of the impact of the China Initiative is illustrative but does not indicate how national security concerns – and fears about the enforcement of national security – enter into the practice of science on a day-to-day level.
She also told University World News that further research has to be done into the “downstream effects”.
For example, the effect of there being fewer Chinese PhDs on both the future of America’s cadre of scientists and on MA and undergraduate students who won’t have the same opportunity to be taught by the brightest Chinese scientists who came to America that earlier cohorts of students had.
University World News
