Many researchers of color are at a disadvantage when applying for postdoctoral positions. That’s one of the main findings of a new study of 22,098 applications for 769 scientific postdoc positions at nine U.S. universities. The results could inform long-standing efforts to diversify science faculty, as postdoc appointments are a key stage in the academic career pipeline—but notoriously difficult to track or study because they are generally managed at the level of individual principal investigators (PIs) rather than departments or institutions.
The study, published last month in Social Science Research, is “very important” and “very needed,” says Damani White-Lewis, an assistant professor of higher education at the University of Pennsylvania who has researched inequity in faculty careers but wasn’t involved in the new work. “The data are just unparalleled,” agrees Mary Blair-Loy, a professor at the University of California (UC), San Diego, and co-director of the Center for Research on Gender in Science, Technology, Engineering, Mathematics, and Medicine (STEMM) who also wasn’t involved in the study.
Lead author Kimberlee Shauman, a professor of sociology at UC Davis, and her colleague analyzed the hiring process for postdoc positions advertised in STEMM fields from 2013 to 2021, dividing it into three stages: being considered for the position, being selected for an interview, and being offered the job. The project planned to draw on data from UC campuses, according to a description of the grant that funded the work. But the study doesn’t specify what universities the data came from, beyond their U.S. location, “to maintain confidentiality,” according to Shauman.
The first step of hiring—whether an applicant was seriously considered, or made the “long list”—was the most selective, with only 9% of applicants making the cut. White applicants were among the most likely to reach that stage, along with women who are Black, Latina, or Native American (BLNA), whereas Asian applicants and BLNA men were the least likely. Of the applicants who were seriously considered, BLNA women were the most likely to be interviewed. But in the final stage of hiring, BLNA women were the least likely of the interviewed candidates to be offered the job. (The job seekers self-classified when submitting their application to the university hiring system, although the people making hiring decisions didn’t see that information. Some of the applicants were nonbinary, but the study didn’t consider hiring disparities for that group because of low sample size.)
Shauman points out the study couldn’t directly test for bias based on gender, race, or ethnicity. That can only come from experimental studies, such as those that alter the names on otherwise identical applications, she notes. But she and her co-author—Jill Huynh, a Ph.D. student at UC Davis—were able to extract potentially useful information from applications, such as the applicants’ educational history, how they heard about the postdoc position, who was listed as their references, and the grants, fellowships, and publications listed on their CV, and test for any effects of those variables. After accounting for that information, the authors found that gender, racial, and ethnic disparities in who was initially considered for a position decreased by up to 64%.
That indicates that structural inequities play a role in determining who gets to later stages of the hiring process, White-Lewis says. “Inequities … compound over time. That, of course, should make us concerned.”
The apparently contradictory results for BLNA women at different hiring stages also raise concerns. “It is interesting that [BLNA women] are advantaged in being seriously considered and interviewed but are less likely to receive an offer,” observes Jue Wu, a postdoc at UC Berkeley who has studied faculty hiring but wasn’t involved in the new study. “It would be great if more nuances regarding the ‘why’ of this result can be revealed, either by digging more into the data or perhaps interviewing the hiring committee.”
Shauman suspects BLNA women’s higher success rates early in the hiring process may at least partly reflect those candidates being highly selective in their applications. BLNA women were underrepresented in the applicant pool compared with their numbers among U.S. Ph.D. graduates, which may mean that many of them are only aiming for positions that are closely aligned with their skills and background. But it’s not clear why their success drops off at the job offer stage. “That could be an indication of bias” among those making the hiring decisions, Shauman says.
As for why Asian candidates didn’t fare as well as white candidates, it may have to do with factors related to their educational background and the network of scientists they interact with. The researchers found that applicants who received their Ph.D. from a U.S. institution and learned of the position through their personal networks—rather than through a website—were more likely to be seriously considered for a position. Asian applicants, however, were more likely than other groups to receive their Ph.D. abroad and less likely to learn of the position through word of mouth.
Blair-Loy says it’s important to look at factors that contribute to postdoc hiring and potential inequities because the postdoc path is “an increasingly required part of getting good jobs, not only in academia, but also in industry, and in nonprofit and policy and government work.”
There are also fewer safeguards in place to prevent bias in postdoc selection compared with graduate student and faculty hiring. Unlike the processes for those two career stages—which often involve standardized applications and diverse hiring committees—postdoc hiring depends more heavily on the idiosyncratic preferences and actions of the PIs who are selecting among applicants. Some positions aren’t even advertised, relying entirely on word of mouth and personal connections—a situation that could exacerbate disparities beyond what was observed in Shauman’s study, which didn’t capture postdoc positions that were filled through informal channels.
“There should be much more organization and transparency about postdoc hiring,” Shauman argues. Universities, she says, should “have more specific policies about hiring and what is required for posting jobs—making things transparent, being specific about qualifications, having job postings open for a period of time.” She also argues that funding agencies should enact their own postdoc selection requirements. Right now, “There’s very little policy written down that says, ‘This is how postdoc hiring needs to happen if you’re going to do it with our federal funds.’ It’s really left up to the funded researcher.”
White-Lewis agrees changes are needed, to an extent. “Institutions can and should do something, but they need to make sure that they’re still preserving some of the features that make postdoctoral hiring what it is,” he says. “Oftentimes, the postdoc market looks different than the faculty hiring market—you need someone more quickly.” He worries PIs might get “grumpy” if they encounter roadblocks that cause significant delays in the hiring process. “I think it’ll require careful consideration about what features of postdoc hiring are outdated and can be revised so that it’s both efficient for PIs and humanistic for applicants.”
Another possible policy remedy, White-Lewis adds, could be establishing programs that promote network connections across universities. “How can institutions get in the business of creating strong bonds for minoritized folks?” he asks. “I think there are some things that can be done there.”
