More graduated but less promoted: A dynamic perspective of the leaky pipeline of women in academic Science, Technology, Engineering, and Mathematics

• A systemic approach explores the persistence of the gender representation gap • System dynamics simulation applied to the academic career in the United States • The interrelation between gender bias and self-selection drives career disparities • Women underrepresentation affects attraction and retention of female candidates • Interventions at different career stages are necessary to reduce the gender gap Although women attain doctoral degrees in Science, Technology, Engineering, and Mathematics (STEM) at rates comparable to men, and despite ongoing initiatives to promote gender parity, women still experience lower promotion rates to senior academic positions. This underrepresentation of women not only limits their professional growth but also has broader implications for innovation and teams’ operational performance. While previous research suggests that gender differences observed at a specific career stage may result from gender biases and self-selection, there remains limited understanding of how these processes unfold together and may affect women’s career opportunities. We argue that understanding such a gender gap requires shifting from a focus on isolated, specific career stages to a more systemic approach. Using a system dynamics simulation model based on data from the American Higher Education Sector, particularly STEM academia, we build on a dynamic perspective and provide a novel view of the gender representation gap. System dynamics allows us to understand how gender bias and self-selection are associated with changes in the representation of women over time. Our results corroborate that gender bias women encounter during their academic careers affects their retention rates, which serves as a signal for junior female academics of the career opportunities the field offers. Thus, the low retention rate among women is associated with a deterrent effect on prospective candidates. Results further indicate that while one-time interventions at specific career stages contribute to closing the gender representation gap by 2050, adopting multi-intervention approaches help to achieve gender parity more rapidly.