Gender in the Journals, Continued: Evidence from Five Political Science Journals
Nadia Brown, Purdue University
David Samuels, University of Minnesota
This special report considers potential gender bias in internal editorial processes at five political science journals: The American Political Science Review (APSR), Comparative Political Studies (CPS), World Politics (WP), Political Behavior (PB) and International Studies Quarterly (ISQ).
These works were inspired by Teele and Thelen’s (2017) exploration of “Gender in the Journals,” the relative presence of women as authors of articles in top political science journals. Teele and Thelen documented a significant “gender gap” in publication rates of peer-reviewed articles between men and women: Women were under-represented relative to their numbers in the discipline, and did not appear as coauthors as frequently as men. The authors also speculated that top journals might be biased against the sorts of work that female scholars are more likely to engage in, whether in terms of substantive questions asked or methods employed.
Teele and Thelen simply counted authors by gender. Their findings raised important questions, but cannot explain why women are under-represented, and why women are under-represented more or less at certain journals. Journals’ tables of contents reflect several factors, especially the pool of submissions and the editorial process. Most obviously, if a journal receives relatively few submissions from women, its table of contents will not reflect women’s relative presence in the discipline. Likewise, Teele and Thelen’s findings also cannot tell us whether actual bias—conscious or not—exists in the editorial process. Do editors discriminate by gender (or in some other way)? Perhaps the fact that most journal editors are male leads to biased outcomes, due to selection bias in the internal or peer review stages of the process. In any case, journal editorial processes are non-standardized and remain something of a “black box” to outsiders. Teele and Thelen’s data cannot pinpoint where bias might occur, if it does occur.