What motivates scientists to move to new academic institutions, and what factors make them stay put? A National Bureau of Economic Research (NBER) working paper posted earlier this month suggests that family structure plays an important role, at least for “elite” life scientists holding an independent position in the United States. Specifically, these scientists were less likely to relocate during those years when they had children of high school age. The results also suggest that the quality of the peer environment and recent funding are important factors in elite life scientists’ decisions to move or to stay at their current institution.
Moving “is a key career decision that can potentially play an important role in the generation of scientific knowledge by allowing scientists [to] find environments where they are more effective in doing their research,” writes Ina Ganguli, an assistant professor of economics at the University of Massachusetts, Amherst, and a co-author on the paper, in an email to Science Careers. Yet, she adds, “[w]e … understand very little about why scientists move,” especially regarding the factors that tend to deter them from relocating. This study takes a step toward addressing that question, Aldo Geuna, a professor of economic policy at the University of Torino in Italy who was not involved in the work, writes in an email to Science Careers. “Studies like this one are extremely important to better understand scientific career[s],” notes Geuna, who recently edited a book on the global mobility of research scientists, although the study’s results “might not be generalized to the population as the sample includes only ‘stars’” and excludes postdocs.
[E]ven ‘star’ scientists take into account these more personal, family factors in their mobility decisions.
The study team—which also included Pierre Azoulay of the Sloan School of Management at the Massachusetts Institute of Technology in Cambridge and Joshua S. Graff Zivin of the University of California, San Diego—used public records to reconstruct the careers of 10,004 elite life scientists in the United States from the moment they earned their first independent position until 2006. Elite scientists in the field were defined as the best-funded or most-cited scientists; those who hold the most patents; members of the National Academy of Sciences; and early- and middle-career scientists who showed great potential, as demonstrated by a National Institutes of Health (NIH) MERIT award, an early-career prize, or a Howard Hughes Medical Institute investigator position. Altogether, just over a third of the scientists in the sample had moved to a new institution more than 50 miles away at least once during their independent career. For 3154 of the researchers, the authors were also able to track down information about their children, including number and age, so that they could dig into questions of how personal factors such as family constraints play into scientists’ mobility. (Previous studies have focused on professional factors such as individual productivity.)
The most interesting and original result in the study is the impact of children’s age on scientists' mobility, Geuna says. Having at least one child of high school age reduced the likelihood that an elite scientist would move to a new institution. On the flip side, these scientists were more likely to move either just before their oldest child had started high school or just after their youngest child had finished it. Scientists thus seemed to prefer to either anticipate or postpone such professional moves rather than relocate the family when they had children in high school, the paper says. According to social psychology research, high school is usually the time when potential disruptions to the social networks of children are greatest, and so the results suggest that “even ‘star’ scientists take into account these more personal, family factors in their mobility decisions,” Ganguli says. Regardless of their fields and potential, the study sends young scientists “a message … that it is normal to keep impacts on family and children in mind when making relocation decisions and career decisions more generally.”
The study also looked into the impact of the elite scientists’ peers on their decision to move. Using the presence and productivity of former co-authors and potential collaborators as a proxy to measure the quality of the peer environment, the authors show that, “all else equal, scientists are more likely to move (than not) in a given year if the peer environment at home is not very good,” Ganguli says. And vice-versa, “scientists are more likely to move in a given year if the peer environment at other ‘distant’ places is better.”
The findings also agree with previous studies showing that a good publication record increases scientists’ likelihood of moving, likely due to their attractiveness to other institutions. The results however suggest that “this impulse is constrained by recent NIH funding,” probably due to the difficulties of moving grants between institutions, Ganguli says. When applying, scientists should be aware of the possibility that winning a grant may actually temporarily limit relocation opportunities due to the “potential costs associated with moving grants,” she suggests. The paper’s authors urge funding bodies and institutions to increase the portability of research funding. “[E]asing this process of moving grants may help scientists more easily relocate to environments where they may be more effective in doing their research,” Ganguli says.
But P. Kay Lund, director of the NIH Division of Biomedical Research Workforce in the Office of Extramural Research, and Misty Heggeness, a labor economist in the same division, note in a joint statement to Science Careers that the detected impact of newly received NIH funding on mobility is “minimal” and less than that of children’s ages and peer environment quality. They suggest that incentives to stay may be at play. The fact that scientists have just received NIH funding “suggests that the intellectual and practical framework to successfully conduct [the] research” is already in place where they are, they write. Also, “institutions may actively work to retain newly funded investigators, [which] may also impact scientists’ choices.”
One of the major unanswered questions from this study, however, regards the role gender might play in decisions about moving. The study found that being a woman increased the likelihood of taking a position to a new, nearby institution, but it had no apparent effect on the likelihood of going to a new institution so far away that it would require relocating. Women were too few in the sample to investigate further—their representation among the movers and nonmovers was 13% and 16%, respectively—but Ganguli says that “an important question for further research is whether age of children and other factors appear to affect women’s and men’s relocation decisions differently.”
Elizabeth Pollitzer, the managing director of Portia, a U.K.-based organization devoted to improving gender equality in science, finds the paper interesting but would have liked to see the authors address the gender dimension head-on. “I was surprised how ‘gender blind’ the premises, analysis and conclusions were, starting with the definition of what constitutes an ‘elite life scientist,’” which heavily skewed the sample toward male representation, Pollitzer writes in an email to Science Careers. “The authors have not questioned the possibility of inherent gender imbalance in their sample affecting the conclusions,” especially regarding the impact of children’s ages, she adds. They identified “the ages between 14-17 as critical to mobility decisions but others (e.g. EMBO) pointed to pre-school ages as key to women’s careers since it is women who carry the biggest share of childcare responsibility … and men scientists with children are less obstructed in their mobility career choices and decisions (because they are less likely to be married to women scientists who are often the ‘trailing spouse’).”
Better understanding scientific careers—including mobility, funding, and gender aspects—is important if we are to better support scientists and knowledge production, Geuna says. “We know something from the old quantitative sociology but still not enough to come [up] with proper policy suggestions either at the institutional or national level.” Moving forward, he adds, special attention should be given to the vast majority of scientists who are making an important contribution to research but don’t make it to stardom or may never have a chance to do so, including the growing proportion of scientists holding postdoc positions.
Meanwhile, young scientists may find such studies useful to inform their own career choices. “Relocation decisions in any field or career are often complex and difficult both from a professional and personal standpoint given the uncertainty involved in moving to a new organization, peer environment, city, etc,” Ganguli writes. “Knowing what others have done more generally, based on analysis of large datasets like this one, rather than just anecdotally, can be helpful. In this case, seeing evidence about how senior scientists have behaved and knowing that they had extremely successful careers while still navigating these complex career and personal decisions can help one think more carefully about the tradeoffs involved.”