top of page

The Impact of US–China Tensions on US Science: Evidence from the NIH Investigations

  • Writer: Ye Wang
    Ye Wang
  • Dec 2
  • 1 min read

PNAS 2023, with Ruixue Jia, Margaret Roberts, and Eddie Yang

Download the manuscript here


This paper documents how U.S.–China political tensions have reshaped scientific production in the United States. We study NIH-initiated investigations that began in 2018 and targeted undisclosed foreign funding, with most cases involving China. Using large-scale publication data from 2010–2021, we compare U.S. life scientists who had collaborated with researchers in China to similar scientists who only collaborated with scholars in other countries. The paper shows that, after 2018, the productivity of U.S. scientists with China collaborations declined in both quantity and impact, with the drop in citations (roughly 10%) far exceeding the modest drop in publication counts. These effects emerge before the COVID-19 pandemic and remain after adjusting for differences in scientists’ backgrounds, fields, and prior productivity.


Beyond the average effect, the paper reveals that the decline is broad and systematic. It appears across most major research institutions, affects both Asian and non-Asian scientists, and is largest in fields with high NIH funding or strong U.S.–China collaboration before 2018. We also link these declines to aggregate patterns: fields most affected by the NIH investigations showed slower scientific growth in both the United States and China relative to other countries. Interviews with scientists help explain why. Researchers describe lost access to funding, reduced collaboration with Chinese partners, new administrative burdens, and a chilling effect that discourages starting cross-national projects. Together, the evidence suggests that the investigations produced sizable unintended consequences for U.S. scientific progress by disrupting long-standing channels of international collaboration.

Comments


  • Google Scholar
  • github
  • Twitter
bottom of page