Universities retain and promote men and women at similar rates in STEM departments, except for mathematics, according to a new study. The median stay is almost 11 years. And the chance that any STEM faculty member, male or female, stays on is less than 50%, according to the researchers. Of those who enter as assistant professors, 64% are promoted to associate professor at the same institution. But because faculty retain their positions for a very long time, it may take 100 years before women reach parity in STEM departments.
Unlike other STEM departments, women in math leave after less than 4 years, compared to men who leave after more than 7 years, a statistically significant difference. Women in computer science stay longer—just over a decade compared to men who left after about nine years.
Faculty churn is expensive for universities, which invest between $110,000 and $1.5 million in startup costs for earlier-career faculty. Furthermore, it interrupts student mentoring, so retention is important for continuity.
Professor Deborah Kaminski of Rensselaer Polytechnic Institute’s Mechanical Aerospace and Nuclear Engineering Department, and Cheryl Geisler, Professor of Communication, Art, and Technology at Simon Fraser University, conducted the study, which was funded by the U.S. National Science Foundation. They analyzed catalogs and departmental Web sites to track the academic careers of nearly 3,000 STEM faculty at 14 East coast American universities since 1990. (Midwest and Western universities’ catalogs are printed every other year and so could not be included.) Their paper, "Survival Analysis of Faculty Retention in Science and Engineering by Gender," appeared in the Feb. 17 issue of Science.
"Ours is the first study that tracks what individuals did, not just intention," says Geisler. There have been previous studies of STEM and facultys’ ‘intention to leave’ with questionnaires asking if they have considered applying for a job or expect to do so in the coming year, or if they have had a job offer. "Those studies indicate women are less satisfied and thinking of leaving," says Geisler, "so it was thought that if a high percent were considering leaving, then there must be a high percent that actually leave." To their surprise, however, Geisler and Kaminski found that attrition did not disproportionately affect women, except for mathematics. "There is lots of literature about math being a gendered issue," says Geisler, "but there isn’t any literature on gender disparities in the field of mathematics regarding retention or promotion. So we think our results should be surprising to math as a field."
The American Mathematics Society’s Associate Executive Director Jim Maxwell expressed surprise at the findings, and noted that the number of institutions studied is "a pretty small set" given the 1,400 math departments in the country. He says that unlike other STEM fields, mathematicians don’t usually get post docs before becoming faculty and are therefore "younger in terms of establishing their research careers." A mathematics department doesn’t lose a $200,000 lab, if one doesn't continue, he notes. On the other hand, Maxwell says the number of post docs in math has grown since the late 1990s, partly in response to the tough academic job market for mathematics Ph.D.s.
Geisler and Kaminski are calling on the Society to look into the situation. "Our results are robust enough to justify the AMS finding out what’s gong on in terms of retention," says Geisler.
Although there has been a substantial increase in the percentage of women hired into STEM faculty positions since 1990, which Geisler and Kaminski find somewhat heartening, they are concerned about the future. They point out that because the whole system lags, it will take a century before the total faculty becomes representative of the gender proportions hired. "Now we’re only hiring at 27% and from 2000 to 2005 we were hiring at 25%. So the percent is rising very slowly," says Kaminski, "It could take 50 years for them to get up to 50% coming in the door and another 40 years before that’s felt in the whole population. Moreover, that scenario includes retaining those faculty, Kaminski says, "It would be even worse if we were losing them."
"If tomorrow we started hiring 50% men and women, it would still take 40 years until there was gender equity. And that's unlikely because the applicant pool doesn’t support it," says Kaminski. "We wanted to be sure that people wouldn’t think, ‘Oh women are being retained at the same rate as men, so everything is fine.’ Everything is not fine. We still have a long way to go." Adds Geisler, "Our analysis is slightly good news, but it doesn’t justify complacency. [It doesn’t mean that the whole problem is solved.]"
Karen A. Frenkel is a freelance writer and editor in New York City specializing in science and technology.