Trends in American scientists’ political donations and implications for trust in science

The study addresses the persistent claims of liberal bias in American academia and its potential impact on public trust in science. Conservative critiques often cite the perceived left-leaning nature of scientists as a reason for distrust in scientific findings. However, previous research on faculty politics has suffered from methodological flaws, such as limited sample sizes and biased selection of institutions and disciplines. While some studies suggest a slightly centrist professoriate, it's undeniable that a significant number of scientists favor the Democratic Party. This paper utilizes FEC data, a vast and publicly accessible database, to gain a more comprehensive understanding of American scientists' political contributions. The authors leverage this data, covering donations from 1979 to the present, to track long-term trends and assess the political orientations of scientists across academia and industry. The data offers a unique opportunity to examine political giving as a quantifiable measure of political leaning, overcoming the limitations of relying on party registration or self-reported political views. The growing volume of political donations in recent years further strengthens the value of this approach, making it possible to discern significant trends that were previously harder to detect. This approach helps to avoid the potential biases of smaller, less representative datasets.

The paper reviews existing literature on the political ideologies of academics, highlighting the methodological weaknesses of previous studies that often focused on elite universities or specific disciplines. It acknowledges that robust studies using broader samples have found the professoriate to be more centrist than often portrayed. The authors cite a 2009 Pew survey indicating a significant preference for the Democratic Party among scientists, which is the basis for exploring this political leaning further. The discussion notes the limitations of using monetary donations as a proxy for beliefs, acknowledging that donors constitute a subset of the overall scientific population and that donation patterns have changed significantly due to online platforms and social media campaigns.

The study uses data from the Federal Election Commission (FEC) covering donations from 1979 to 2021. The researchers weigh donations by amount and analyze donations to Republicans, Democrats, and third parties. Donations to candidates with unknown party affiliations were excluded. From 2002 onwards, each record includes the donor's employer and occupation. To identify donations from academic scientists, the researchers filtered data based on employers including “college” or “university” and occupations including “professor,” “faculty,” “scientist,” or “lecturer.” A control group of college and university administrators was also included. To further refine the sample of scientists, they matched donor names to the Scopus database of scholarly abstracts and citations. The analysis also included scientists in the energy sector, specifically focusing on ten major energy companies. To distinguish between executives and scientists within these companies, the researchers filtered by occupation, identifying those with titles suggesting executive positions versus those with scientific or engineering roles. The methodology further included analysis of data from the General Social Survey (GSS) to examine trends in Republicans’ and Democrats’ trust in the scientific community over time using responses to questions about political affiliation and confidence in various institutions, specifically the scientific community. To study the various disciplines the authors matched the FEC records to those from the Scopus database and separated the results into physical, life, social and health sciences. The authors note limitations in their approach such as potential inaccuracies in self-reported names and employment information, and the impact of online donation platforms on donation patterns over time.

The analysis of FEC data confirms a substantial preference for Democratic candidates among scientists who donate to political campaigns. However, this is a relatively recent phenomenon. From 1984 to 2000, donations to Republicans by college and university employees were around 40%, with only a slight Democratic preference. From 2000 to 2021, this dramatically shifted, with less than 10% of donations going to Republicans. By 2016, professors gave even less to Republicans than other university employees, and Ivy League professors gave only about 2%. The total value of donations increased substantially in 2019, coinciding with a historic low in academic donations to Republicans. In the energy sector, support for Republican candidates also significantly decreased, from 60% of donations in 1990-2000 to approximately 30% in 2020. A striking divergence was found between executives and scientists/engineers in this sector, with executives donating to Republicans far more frequently. Donations to third-party candidates remained consistently low throughout the study period. The analysis of GSS data shows a partisan reversal in attitudes toward science since the 1970s, with Republicans starting with higher confidence in science and ending with lower levels than Democrats. This reversal aligns roughly with scientists shifting away from the Republican Party. Across all sectors analyzed, the authors observed a pronounced trend of decreased support for Republican candidates among scientists, irrespective of the employing institution's type or the scientific discipline.

The study's findings confirm existing impressions and survey data showing that scientists, particularly in academia, lean more liberally than the general population. However, the authors argue that this does not necessarily indicate problematic bias. They point out that similar patterns exist in industry, where claims of liberal bias are less common. They further suggest that higher levels of education correlate with more liberal political views. This could explain why scientists, who often have advanced degrees, tend to be more liberal than their supervisors (who may have lower levels of education). The dramatic decline in Republican support among scientists in recent years cannot be fully explained by education alone, however. The authors propose several contributing factors: the Republican Party's increasing hostility toward science in areas such as climate change and public health; the Republican Party's populist turn, which portrays scientists as out-of-touch elites; and increasing trust in science among Democrats. The conflict between the norms of science (communism and universalism) and conservative viewpoints may also play a role. The observed shift in political donations strongly correlates with the partisan reversal in attitudes toward science noted in the GSS data, supporting the argument that the Republican Party's stance toward science has significantly influenced scientists' political affiliations.

The study highlights the increasing polarization between scientists and the Republican Party in recent years. While the authors acknowledge that this may affect public perception and trust in science among conservatives, they contend that the Republican Party's anti-scientific stances are a primary driver of this shift. They suggest that greater public engagement from scientists with centrist or conservative viewpoints could foster trust and improve communication, particularly within the Republican base. Increasing the visibility and voices of scientists who do not align strictly with the Democratic Party could be crucial in bridging the trust gap and countering negative portrayals of scientists as uniformly liberal.

The study acknowledges several limitations. Self-reported data on names and employment information may be inaccurate. Donation patterns have been significantly altered by online donation platforms, making comparisons across time periods challenging. Finally, the study focuses on scientists who donate to political campaigns, which is a subset of all scientists and may not fully represent the political spectrum of the broader scientific community.