The Branching Pipeline: Understanding How Career Switches Affect Gender Gaps across Fields

• In Event: 3-119 - Gender Inequality in Developmental Psychology

Integrative Statement

Women are underrepresented in many academic fields, and even more so as one moves up the ranks in a field. To meaningfully address this problem, we must first uncover the mechanisms that generate it.

Prior research on women’s underrepresentation has typically focused on static snapshots of this phenomenon—e.g., explaining the low number of women who obtain PhDs in science, technology, engineering, and mathematics (STEM). While informative, this approach misses valuable information about the dynamics of individuals’ career trajectories: It would be informative to know, for example, what percentage of women who obtained a college degree in physics (and were thus in this field’s “pipeline”) then went on to complete a PhD program in other fields, especially if we can also compare this number with the frequency of switches in the opposite direction (from bachelor’s degrees in other disciplines to a PhD in physics). However, little research to date has examined the “branching pipeline” of career trajectories in a comprehensive way.

The present research adopted this novel approach. We examined career trajectories using the single largest dataset of academic resumes: the public data from ORCID.org (Open Researcher and Contributor ID), which contains resumes for approximately 1.1 million researchers from all over the world. We devised algorithms for detecting whether a field change was reported, and if so, the origin and destination fields. Existing algorithms were used to infer genders of individuals based on their first and last names. We then merged this dataset with an existing dataset on various characteristics of 30 academic fields (e.g., the extent to which their members believe that success depends on “brilliance”; Leslie, Cimpian, et al. 2015).

Using this merged dataset, we were able to examine what characteristics of a field predict an imbalance in the gender of people switching into and out of it. Six field-level characteristics were examined: the aforementioned “brilliance” beliefs, the workload of the field, whether it is a STEM field, its emphasis on mathematics (as measured by average GRE scores of applicants), its selectivity (at the level of PhD admissions), and the extent to which it requires “systemizing” vs. “empathizing” (Baron-Cohen, 2002; see Leslie, Cimpian, et al., 2015, for additional detail). We included these variables simultaneously in our statistical models and found that the most powerful predictor of a gender imbalance in field switches was the relative strength of the brilliance beliefs of the origin vs. destination fields: Relative to men, women were more likely to switch from high- to low-brilliance-beliefs fields (e.g., physics to earth sciences) and less likely to switch in the opposite direction. The only other significant predictors were whether a field is in STEM (relative to men, women switched out of STEM fields more often) and a field’s selectivity (relative to men, women switched into more selective fields).

The present findings contribute valuable information on the field-level factors that affect the dynamic trajectories that women and men’s careers take across time. This approach can be used to devise new solutions to the problem of women’s underrepresentation.

Authors

• Andrei Cimpian, New York University
  Presenting Author

• Aniko Hannak, Vienna University of Economics and Business
  Non-Presenting Author

• Kenneth Joseph, University of Buffalo
  Non-Presenting Author

• Daniel Larremore, University of Colorado Boulder
  Non-Presenting Author