Gender differences in academic attrition are narrowing and might never have existed in?some key scientific fields, where men and women face almost equal odds of?quitting early, according to?researchers.

Polish researchers say scientists are leaving the profession in?alarming numbers, with about one-third departing within five years of?their initial publications and more than half gone after another five years. But gender has surprisingly little to?do with?it.

¡°This is a loss of talent. It¡¯s not a loss of women; it¡¯s a loss of both men and women,¡± said lead researcher Marek Kwiek, Unesco chair in institutional research and higher education policy at Adam Mickiewicz University, Pozna¨½.

¡°Attrition in science is ever less gendered. This gender difference is getting smaller, and in many disciplines is just disappearing.¡±

³ÉÈËVRÊÓƵ

Professor Kwiek¡¯s longitudinal study, conducted with doctoral student Lukasz Szymula, analysed individual-level data on almost 380,000 scientists from 38 countries who collectively produced about 5?million publications over 22?years.

The research, , compares the outcomes of people who began publishing in 2000 and 2010. In the earlier cohort, after a few years, women proved about one-tenth more likely than men to leave science in any given year. But in the later cohort, there was no gender difference in attrition.

³ÉÈËVRÊÓƵ

And while gender differences in the 2000 cohort were significant in relatively feminised disciplines such as biochemistry and medicine, attrition patterns were ¡°nearly identical¡± in heavily ¡°mathematised¡± fields like physics, computing and astronomy.

Professor Kwiek said this reflected the ¡°exemplary¡± qualities of the women in these ¡°highly competitive¡± and heavily male-dominated fields. ¡°There are so few of them from the very beginning,¡± he told ³ÉÈËVRÊÓƵ. ¡°You need to be enormously good or enormously hardworking compared with men.¡±

The conclusions contradict decades of research characterising early departure from academia as primarily a gendered issue, with women forced out by sexist and hostile attitudes in workplaces offering little job security, low salaries and poor work-life balance.

But his research investigated only attrition, overlooking the ¡°even more primal issue¡± of equal representation. ¡°[To]?leave science, you have to enter?it,¡± he acknowledged. ¡°If?you don¡¯t enter?it, you don¡¯t have a?chance to leave.¡±

³ÉÈËVRÊÓƵ

The analysis found that publication quantity was the primary predictor of a scientist¡¯s prospects of remaining in academia. Publishing a journal article boosted the odds of publishing the following year by 20?per cent, and in some disciplines by as much as 40?per cent.

Regularly publishing in top-rated journals significantly enhanced people¡¯s prospects of remaining in academia, he said. But citation counts appeared to have little influence.

john.ross@timeshighereducation.com