Discrimination in Academic Publishing: Choices, Pipelines, and Networks in PS

• In Event: Methods to Study Diversity

Abstract

Discrimination in the production and dissemination of knowledge is a subject of growing and widespread concern in academia and society. Despite the growing concern, the degree of race or gender imbalance, and the extent to which the imbalance is function of choice (such as area of study or methodologies used) is not well-documented or understood. Current research on academic publishing in political science fails to address both the choices made by individual researchers and the underlying pool of scholars who could potentially be publishing.

We move the discussion forward in three ways. First, we introduce a new data set that includes granular information about who is publishing (undergraduate and graduate education, rank, department, institution, race, ethnicity, sex, etc.), the areas in which they are publishing (American politics, international relations, comparative politics, and normative or positive theory, as well as subfields within these categories), and importantly, which methodological tools they are using (mathematical models, statistical analysis, qualitative methods, etc.). Second, we identify the underlying pool of scholars who could be publishing, that is, all individuals who have completed Ph.D.s in political science (and some associated fields). We again collect granular information as well as dissertation abstracts. We link this information to those who eventually realize success in journal publication. Third, we use a topic model on dissertation and published abstracts to understand the kinds of questions and methods that eventually end up as published papers. The granular data plus the topic model allow us to identify “pipelines” and networks, as well as gain greater insight into the choices made both by underrepresented groups and editors.

Authors

• Kevin A. Clarke, University of Rochester

• Lawrence S. Rothenberg, University of Rochester