The gender gap in science: how long until women are equally represented?

How long will it take for women to be equally represented in science? Luke Holman, Devi Stuart-Fox, and Cindy Hauser at the University of Melbourne decided to answer that question with data rather than speculation. They assembled one of the largest bibliometric datasets ever created — and the answer, for many fields, was: not in our lifetimes.

The dataset

The study drew on 35.5 million author records from 9.15 million PubMed articles (2002 onward) and 1.1 million authors from half a million arXiv preprints (1991 onward). Together, the data covered more than 6,000 journals across 115 STEMM disciplines in over 100 countries. The researchers used Genderize.io to assign gender to authors' given names, keeping only names where the tool returned at least 95% confidence. That threshold preserved 92.4% of the dataset, and validation against manually curated records showed a misclassification rate of just 0.3%.

The projections

The team fitted linear models to each field's trajectory and extrapolated to find when — if ever — it would reach 45% women authors.

The results varied wildly. Nursing, midwifery, and palliative care already exceeded 50% women. Health-related disciplines were approaching parity. But physics was on track to take 258 years. Computer science, surgery, and mathematics were similarly bleak. Across the board, 87 of the 115 disciplines examined had significantly fewer than 45% women authors.

The seniority drop

Women were overrepresented as first authors — the position typically held by the researcher who did the hands-on work — but substantially underrepresented as last authors, the senior position that usually belongs to the lab head or principal investigator. The gap between those two numbers told a story of attrition: women were entering the pipeline but not advancing through it at the same rate as men.

Single-author papers showed the same pattern. Men were far more likely to publish alone, suggesting either greater autonomy or fewer collaborative obligations — or both.

Geography and prestige

Gender ratios varied by more than 30 percentage points across countries. Japan, Germany, and Switzerland showed particularly large gaps despite their wealth and research investment. High-impact journals published fewer women than lower-impact ones. Men received roughly 1.7 to 2.1 times more invitations to submit papers. Open-access journals did slightly better than traditional subscription journals.

The interactive tool

The researchers built a public web application that lets anyone explore the data by discipline, journal, country, and authorship position. It remains live and is one of the most widely used resources in bibliometric gender research.

Impact

The paper has been cited thousands of times and remains the standard reference for anyone studying the STEMM gender gap at scale. Its core finding — that many fields will not reach parity this century without deliberate intervention — shifted the conversation from "is there a gap?" to "what are we going to do about it?"