Are scientists sensitive to political pressures? The question is not new, but it has resurfaced with the presence of Donald Trump in the White House.
Since February 2025, the American government has released a list of words that are forbidden to be used in articles published by public research and funding agencies, under the threat of having their funding cut off.
Words considered to reflect the “woke” trend and the belief in the “hoax” of climate change by researchers in the United States: diversity, inclusion, climate change, etc., are now terms to avoid in order to receive funding from the National Science Foundation (NSF) or the Centers for Disease Control and Prevention (CDC), a major public health organization.
Has this financial pressure had an impact on research? To find out, three economists looked into their own discipline. This is an interesting case, as economists often circulate their working papers in “gray” literature before official publication in professional journals, allowing for close monitoring of developments since last year.
A High Proportion of Self-Censorship
Dominic Rohner from the University of Lausanne, Oliver Vanden Eynde, and Philine Widmer from the Paris School of Economics (PSE) delved into working papers published by the American National Bureau of Economic Research (NBER) and the European Centre for Economic Policy Research (CEPR) between January 2020 and December 2025.
By checking which authors used public funding, they were able to gauge the propensity of economists to self-censor. They did this by classifying research based on their dependence on public funding, to see if those more dependent were more likely to constrain themselves. The results they have just published provide a frankly positive answer.
Out of the 14,412 scientific articles they reviewed, 1,991, or 13.8% of them, focused on gender, race, or environmental issues and contained at least one of the banned words.
A gap is emerging between papers published by researchers from the most public funding-dependent institutions and others
Their statistical tests show that before the blacklist was announced, the nature of funding did not influence the likelihood of using the banned words. After the announcement, however, a gap is emerging between papers published by researchers from the most public funding-dependent institutions and others.
In the former situation, the probability of using banned words significantly drops, as does the proportion of papers dedicated to Gender-Race-Environment themes (measured by the presence of these themes in the abstracts). For researchers from less public funding-dependent institutions, no significant change was observed.
Terms are Dropped, Not Necessarily the Subjects
This result does not prove that economists have completely abandoned these subjects, nor that they are publishing less, nor that scientific production has collapsed. The authors found no effect on the length of papers or the number of publications. What changes is the presence of certain terms, in other words, the way research is formulated or presented.
The study does not measure an absolute level of self-censorship, but rather a variation after a political shock, showing a rapid adjustment of lexical content
The study authors also statistically confirm that this adjustment results from individual researchers’ behavioral changes, not from a modification of the author pool’s composition (where, for example, fewer researchers specializing in censored themes would have published). The study does not measure an absolute level of self-censorship, but rather a variation after a political shock, showing a rapid adjustment of lexical content, consistent with a phenomenon of self-censorship.
Finally, the three researchers clarify that the three major areas designated by the banned words (gender, “race”, environment) are equally affected. Likewise, they do not observe any behavioral differences based on gender or researchers’ seniority: men and women, young and experienced, behave similarly.
In theory, science should be independent of political power. In practice, the study does not suggest that economists have ceased studying gender, race, and environmental themes, but rather highlights that, faced with a threat to their funding, they quickly adjusted how they present their work.
In this Orwellian world, there is no need to explicitly forbid research. The mere prospect of losing funding seems sufficient to change behaviors.
