As science faces external attacks, it needs to look within to defend and reform
Rather than just defending the status quo, scientists should use data to understand and fix their own institutional problems
By C. Brandon Ogbunu, Undark
On May 24, Vice President J.D. Vance authored a post on X that highlighted a “reproducibility crisis” in the sciences. Vance offered this amid a series of other critiques of higher education to justify the withholding of federal science funding to universities over the past several months. His post was timed to accompany a White House executive order that invoked the language of open science to introduce sweeping changes to our federal scientific infrastructure. It came just weeks after the release of plans to cut science funding in the 2026 fiscal year budget.
The playbook is standard: Fuse an aggressive political agenda to a more palatable set of criticisms. In this case, many agree that processes within professional science have, for decades, had significant flaws. In my view, politicians in power are using this as a justification to burn it down. And outside of a few higher-education legal efforts to fight back, the scientific community remains shell-shocked, unable to gather the momentum to resist effectively.
But in addition to resisting the changes, there might be other ways that we can navigate an uncertain future. In recent years, a field called “metascience” (often referred to as “the science of science”) has emerged, charged with understanding the processes of science, how it operates, and identifying themes in what is produced. I argue that this area is going to be essential moving forward in stormy times, as it can dispel the myth that science is an ideological leviathan incapable of self-reflection and can help us rebuild science into a craft that interrogates its fragilities.
As described in a 2018 review, the science of science “is based on a transdisciplinary approach that uses large data sets to study the mechanisms underlying the doing of science—from the choice of a research problem to career trajectories and progress within a field.” It asks questions about aspects of the scientific enterprise, including employment, publishing trends, economic incentives, merit, and other forces that influence science in ways that may escape our intuition.
Several notable studies have examined the composition of the workforce. In a sweeping analysis of more than 295,000 tenure-track faculty at U.S. Ph.D.-granting universities from 2011 to 2020, an interdisciplinary team revealed that faculty hiring and retention are governed by entrenched hierarchies and stark inequalities. Just 20 percent of universities produce 80 percent of faculty, with a handful of elite institutions dominating the academic labor market. Attrition further entrenches inequality, as those trained at less prestigious or foreign institutions exit academia at higher rates. The 2022 study further suggests that modest gains in women’s representation are largely driven by retirement of older male cohorts rather than equitable hiring, and underscores that U.S. academia is a self-reinforcing system of academic aristocracy — a finding that subverts notions of a meritocracy.
While gender disparities were among many findings in the hiring hierarchy study, they were a central feature of a more recent study that investigated gender disparities in computer science faculty hiring. Instead of looking at large-scale hiring patterns, the researchers examined how perceptions of research areas — particularly the divide between theoretical and applied work — impacted the underrepresentation of women. Using faculty surveys and hiring data, the authors found that applied subfields, where women are more prevalent, are often perceived as less prestigious or “core” to the discipline than theoretical ones. These perceptions shape hiring decisions and reinforce structural biases, effectively penalizing women for working in areas that are socially devalued within the academic hierarchy. This highlights how a scientific perception (that is, which subfields are more central) can have lasting effects on who gets hired, promoted, and lauded.
Both of the above examples focus on the composition of faculty. But this is only one of many areas for which modern metascience has shone a light on hidden themes at work within the profession. Another theme is the role of incentives in shaping how science is done. Kevin Gross and Carl Bergstrom investigated how different forms of peer review encourage different sorts of research enterprises. And in a follow-up study, they more directly examined how incentives drive risk-aversion in research.
Metascience research involving incentives relates to a growing literature that asks the fundamental question: How innovative is science, exactly? The question is hard to answer, as innovation can have different definitions in different paradigms. Nonetheless, an exciting body of work explores the characteristics of teams that foster scientific innovation and how “disruptive” research is. In one landmark study, James Evans and colleagues studied more than 16 million papers, revealing that flat, or egalitarian, teams consistently spark more disruptive breakthroughs than tall, hierarchical ones. Although hierarchical groups rake in citations quickly by developing existing ideas, that short‐term payoff comes at the expense of long‐term influence and the cultivation of junior scholars. These findings and others support notions that the stuff of innovation is as much about the structure of teams as it is about individual talent.
While it is hard to deny the relevance of the findings in these examples, metascience is not without its critics. For one, some suggest that metascience could benefit from some humility, as it is not the first field to ask these questions. Humanists and social scientists (Bruno Latour, for example) have examined science practice for many decades. In addition, others offer that science is not more interesting nor more complex than any other profession, and so its self-study is akin to navel gazing, the self-indulgent practice of being far too interested in our own craft. Lastly, there is sentiment that metascience is becoming just another career elevator field, where power, prestige, and hierarchy are emerging in the same manner that metascience often critiques.
These points are difficult to counter. Name a movement that started with pure intentions, and at some point, it devolved into becoming a ticket to career advancement or profit. Metascience is likely no different. The popularity of metascience will birth a new field — in other words, “The science of science will beget a science of the science of science,” Santa Fe Institute postdoctoral fellow James Holehouse recalled joking at a gathering. And so on.
These problems notwithstanding, metascience offers a lens that is especially important at this critical moment. Support for science in the face of attacks is critical and necessary. But ironically, one of the best ways to defend the craft might be for scientists to identify the fragilities before the enemy does. We can use data and models, not solely our op-ed voices and social media timelines (though all can be useful). The field is already disabusing us of the notion that science as practiced is based on defensible incentives, neutrality of any kind, or merit, however defined.
Instead, it operates on what looks more like a runaway Matthew Effect, whereby the most established scientists benefit disproportionately from the system of reward — and thus the rich get richer. And the problem isn’t that the flaws exist, but that science’s practitioners aren’t interested in a critical lens towards them.
Metascience won’t fix our problems, but it formalizes ways that we can use to reflect, which may implore us to change science for the better.
This article was originally published on Undark. Read the original article.