Psychology’s quiet exodus: women exit research far more often than men

Psychology has long been held up as an exception to the gender imbalances that persist across the sciences. Unlike many STEM disciplines, where women remain a minority among doctoral students and early-career researchers, psychology has for years attracted more women than men into its academic pipeline. More than 60 percent of new entrants into psychological research are now women, a proportion that has held steady across cohorts. Yet a sweeping new international analysis suggests that this apparent success story conceals a deeper problem: the field is remarkably effective at recruiting women and remarkably ineffective at keeping them. The study, led by researchers at the Max Planck Institute for Human Development and published in American Psychologist, finds that women leave psychological research significantly more often than men, and that the divergence begins strikingly early in their careers.

The research team, led by postdoctoral researcher Xinyi Zhao of the Center for Adaptive Rationality at the Max Planck Institute for Human Development, together with colleagues, undertook what the authors describe as the first global longitudinal analysis of when gender differences emerge over the course of academic careers in psychology. Rather than examining a single country or institution, the team traced the publication histories of 78,216 psychologists worldwide who began their academic careers between 2000 and 2014. The data were drawn from the Scopus database, one of the largest bibliographic repositories in existence, which allowed the researchers to follow individual career trajectories across borders, institutions, and decades. Scientific publication activity served as the indicator of an active research career: a researcher who had not published any scientific work for at least five years was considered to have left active research.

The scale of the dataset matters, because attrition from academia is a slow, diffuse process that is easy to miss in cross-sectional snapshots. By tracking cohorts over time, the researchers could pinpoint exactly when the gender gap opens. Their answer: the largest differences appear between the fourth and sixth year after a researcher’s first scientific publication. This is a period during which the risk of leaving research rises substantially for everyone, as the initial excitement of a doctoral project gives way to the pressures of establishing an independent profile. But the rise is sharper for women than for men. The gender gap in attrition is most pronounced about five years after the start of a research career, precisely the window in which many researchers are finishing doctorates, moving between institutions on short-term contracts, and competing for their first permanent positions.

The cumulative effect of this early divergence is stark. By the thirteenth year of their careers, around half of the female researchers in the sample had stopped publishing altogether. Among men, that same halfway point of attrition was not reached until roughly five years later. In other words, even in a discipline where women form the majority of those entering, men sustain active research careers for meaningfully longer on average. This pattern helps explain a puzzle that has persisted for years: women are well represented among doctoral candidates and postdoctoral researchers in psychology, yet remain significantly underrepresented in professorships and other senior positions. The bottleneck, the study suggests, is not at the door of academia but inside it.

Lead author Xinyi Zhao framed the motivation for the work in terms of this paradox. In psychology, she noted, women are well represented among doctoral candidates and postdoctoral researchers, but far fewer reach the highest career levels. The team wanted to understand when women are more likely to leave academia and which factors are associated with this pattern. That framing is important, because much of the public debate about gender in science has concentrated on access—on pipelines, admissions, and the so-called leaky pipeline metaphor. The new findings redirect attention to retention, arguing that the bigger hurdle appears not to be entering academia but staying in it.

To probe possible causes, the researchers went beyond simple headcounts and modeled a range of factors that might plausibly influence whether a researcher stays active. These included publication output, citations, collaboration networks, and institutional characteristics such as the scientific reputation of a researcher’s institution. The strongest predictor of long-term retention turned out to be early publication output. In particular, a high number of first-authored papers—publications for which a researcher carries primary responsibility—was associated with a higher likelihood of remaining scientifically active over the long term. In later career stages, the number of last-authored papers, which conventionally signals scientific independence and leadership responsibility, became the more important indicator. Collaboration networks and institutional conditions were also associated with retention, but these relationships were considerably smaller than the effect of authorship patterns.

Crucially, however, none of these measured factors fully accounted for the gender difference. Even after publication output, citations, networks, and institutional prestige were taken into account, women remained more likely than men to leave research, and the difference was especially pronounced during the early years of academic careers. This residual gap is the study’s most consequential finding, because it rules out the simplest explanations. If women were leaving simply because they published less or worked at less prestigious institutions, controlling for those variables would have erased the difference. It did not. Something not captured in bibliographic data is shaping who stays and who goes.

Zhao was careful to caution against misreading the results. The findings, she emphasized, should not be interpreted as reflecting differences in scientific potential. Rather, they suggest that structural conditions and challenges during certain stages of life influence the career paths of female and male researchers differently. The publication-based analysis cannot, on its own, establish why women are more likely to leave. Information on starting a family, caregiving responsibilities, or individual career decisions was simply not available in the Scopus-derived data. The authors are explicit about this limitation, and it shapes how the findings should be understood: the study maps when attrition happens and what career-level factors correlate with it, but the underlying mechanisms require other kinds of evidence.

Some of that evidence comes from a companion study led by Zhao, published in Higher Education, which surveyed more than 8,000 researchers with children from 119 countries. Its results point toward one plausible mechanism: early parenthood is associated with greater career disruption for women than for men. Women in the survey were more likely to postpone starting a family until after completing their PhD. When parenthood did occur during the doctoral phase, their chances of completing a doctorate, achieving high scientific visibility, and remaining in research over the long term were lower than for men in comparable situations. Zhao described the two studies as complementary: the current work shows when women are more likely to leave research, while the survey findings suggest that the challenges of balancing an academic career with starting a family may be an important part of the explanation.

The research team concludes that science policy should place greater emphasis on the early career phase, the very window—roughly years four to six—where the gender gap in attrition is widest. Mentoring programs, transparent evaluation procedures, and better conditions for balancing research and family responsibilities could, the authors argue, help retain more women in academia over the long term. The findings also draw attention to a previously underexamined aspect of gender equality in academia: while discussions often focus on women’s access to academic careers, the decisive barriers may emerge later, once researchers are already inside the system. The team plans follow-up studies that will survey scientists around the world directly, asking about their career decisions and experiences, in order to move from mapping the attrition gap to understanding it from the perspective of those who leave. For a discipline that studies human behavior and prides itself on evidence-based practice, the message of its own career data is difficult to ignore: the pipeline is not leaking at the entrance, but midway through, and fixing it will require addressing the structural conditions of early academic life rather than only the question of who gets in.

Subject of Research: Gender differences in attrition from academic psychology careers

Article Title: Women leave psychological research more often than men

Article References: Women leave psychological research more often than men. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: psychology, gender gap, academic careers, attrition, Max Planck Institute for Human Development, publication output, early career researchers, parenthood, retention, bibliometrics, American Psychologist, science policy