Level the slopes

It was a crisp sunny morning at the physics workshop in the mountains, and the talks were running behind schedule. A few attendees sat in full ski gear, some glancing at the frosty branches outside, others impatiently drumming their fingers on the pull-out tables, waiting for the chance to hit the slopes. Later that day, a few physicists could be found on top of the mountain enjoying their lunch, networking with such casual ease that they seemed more like old friends than strangers on a business trip. Credit: Illustration by Falco FalcrowThe younger researchers in attendance secretly thanked the good fortune that allowed them to learn how to ski while growing up, and the disposable income to afford the ski pass — eye-wateringly expensive even after the conference discount. After a week in which their privilege had eased their way into the academic elite, each early-career researcher considered how they should continue forming their career path, how to forge strong relationships and whom to forge them with.This fictional vignette attempts to capture how social connections shape science careers. Success depends not only on individual ability but also on resources obtained through relationships — the social capital that determines who gains access to opportunities and recognition. In science, collaboration networks1 act as a form of social capital, mapping who works with whom and how connection patterns shape the circulation of ideas and credit. But access to these networks is far from equitable.Social capital accumulates more easily for those with the background or privilege to be present in the right settings and to be able to effortlessly fit into those settings. Its uneven distribution risks exacerbating existing inequities in science. Indeed, it has been suggested that the gendered gap in productivity — a measure based on publication counts — could be explained by differences in collaboration networks1. However, even if everyone had equal access to beneficial networking opportunities, structural inequalities — for example, gendered differences in recognition2 — would likely persist.In a Comment in this issue of Nature Physics, Weihua Li and colleagues analyse the prominence of men and women in physics, defined as the number of papers that are highly cited within two years of publication. Accounting for the effects of collaboration networks using the methods introduced in their previous work1, they find no underlying differences in individual productivity and prominence between men and women. Yet among elite researchers, women have lower levels of prominence until they receive external recognition, such as election to the United States National Academy of Science.In the years following election, their prominence rises to the same level as men — not because they publish more or with different collaborators, but likely because the community begins to value their work differently. This suggests that women’s contributions to science are systematically undervalued until an external stamp of approval counteracts the negative effect of gender stereotypes on how their work is perceived.Physics is no exception to the broader rule that social capital is important for career success. Similarly, it is plagued by the same inequities and prejudices that exist in wider society. One might argue that encouraging those from marginalized groups to make connections and build collaboration networks may help individual career advancement to an extent. However, this approach places the burden of change on the individual and leaves untouched the wider stereotypes and barriers in place.Additionally, expecting researchers to adapt their behaviour to fit in with the dominant group does little to achieve true diversity within the physics workforce. In an equitable research landscape, every researcher should be able to perform to their own unique strengths and abilities without the need to emulate those who are in the majority.A potentially more effective approach would be to design research environments so that change can occur at both the individual and system level3. Research groups and departments should incentivize collaboration with scientists from different institutions and diverse backgrounds. Universities could widen their recruitment pool through opt-out rather than opt-in promotion processes that shift the burden of initiative away from the individual, gender-neutral objective language in job adverts, or standardized interview questions that reduce the risk of bias3.Journals should expand their reviewing and commissioning pool to include researchers from underrepresented groups to increase their visibility, recognition and chances of successful funding applications. Exploring different reviewing options, such as double-anonymized and triple-anonymized (when even the editor doesn’t see author identities until as late in the process as is practically feasible), could help address known disparities in peer review4. At the system level, the introduction of a standardized measure of undervalued contributions to science, such as mentorship, could rebalance how success is defined, moving attention away from sheer publication volume5.True equity, in other words, requires structural change — a re-evaluation of what and whom science rewards — rather than asking individuals to climb the same ladders under unequal conditions.