A new paper from Marek Kwiek and Lukasz Szymula was published on the BioRxiv preprint server.

„Leaving Science: Attrition of Biologists in 38 OECD Countries”.

Abstract

This study examines biologists leaving science in 38 OECD countries in the past two decades. We use publication metadata from a global bibliometric database (raw Scopus data at the micro-level of individual scientists). In a cohort-based and longitudinal fashion, we follow individual men and women scientists over time, from their first to their last publication (N=86,178). We examine four academic disciplines: AGRI (agricultural and biological sciences), BIO (biochemistry, genetics, and molecular biology), IMMU (immunology and microbiology), and NEURO (neuroscience). We apply survival analysis, conceptualizing scientific life as a sequence of scholarly publishing events. Our Kaplan–Meier survival analysis shows how women disappear from science: in BIO, about 60% are still in science after 5 years, 40% after 10 years, and only 20% by the end of the period examined (i.e., after 19 years). The percentages are substantially higher for men: approximately 70%, 50%, and 30%, respectively. Kaplan–Meier estimations indicate that women in the largest discipline (BIO) are 23.26% more likely to leave science after 10 years and 39.74% more likely to leave science at the end of the study period. Gender difference in attrition, slightly visible after 5 years, increase consistently in later career stages. The probability of surviving for women after 15 years varies considerably, from 47.8% in AGRI to 27.6% in IMMU; for men, the probability is about a fifth higher. Our data show that with the passage of time, women disappear from science in ever-larger proportions compared to men. Gender differences in attrition in the four disciplines have been and continue to be high, but comparing the 2000 and 2010 cohorts, have slightly decreased over time.