In this paper we analyze the phenomenon of quitting academic science and show how quitting differs between men and women, academic disciplines and over time. The approach presented is comprehensive: global, based on cohorts of scientists, and longitudinal – we observe the publication activity of individual scientists over time. Using metadata from Scopus, a global bibliometric database of publications and citations, we analyze the publication careers of scientists from 38 OECD countries who began publishing in 2000 ( N = 142 776) and 2010 ( N = 232 843). The paper tests the usefulness of large bibliometric datasets for a global analysis of academic careers.

The purpose of this article is to determine what can we learn about the changing demographics of scientific personnel around the world and over time using available global bibliometric data sources. We wanted to see how useful global data could be for analyzing the scientific workforce. We tested how the demographic transformation of the global scientific profession can be measured using new data sources, thus transcending the traditional approach in which national statistics from national statistical offices are aggregated to a higher level, as in the case of the scientific workforce databases produced by the OECD, UNESCO and the European Union (Eurostat).