Intersectional inequalities of representation and research topics in science

Abstract

The scientific workforce of the United States is mostly composed of white men. Barriers to entry and participation have been well-studied. However, few have adopted an intersectional perspective to examine the consequences of these inequalities in scientific knowledge. In this work, a large-scale bibliometric analysis is provided on the relationship between intersectional identities, research topics, and scientific impact. Using the Web of Science bibliometric database for US publications between 2008 and 2019, an analysis of over and underrepresentation of different racial and gender identities was conducted, as well as their distribution across different disciplines. The findings reveal a marked underrepresentation of women and marginalized racial groups (Latinxs and Black individuals), with women from these identities being the most affected. Additionally, an asymmetric distribution of racial and gender identities across disciplines is observed, with a notable underrepresentation of women in areas such as Physics, Mathematics, and Engineering. Subsequently, the analysis is deepened by modeling research topics in the Social Sciences and Health. The results show that women tend to publish more in topics such as education, nursing, and gender-based violence, while Black authors are overrepresented in studies on racial discrimination and Latinxs in migration and topics related to Latinx bodies. This distribution in different research topics is in turn related to the academic impact each topic entails. Topics where privileged groups are overrepresented are also the most cited on average, while marginalized groups tend to receive fewer citations in all topics.