How Publishing Incentives Shape Scientific Knowledge Production
The modern academic research career is centered around a single output: the peer-reviewed publication. Peer-reviewed publications—especially when accepted to high-impact journals—can make or break one’s career in the modern academy. Specifically in STEM fields, these publications are required of scientists to get hired, receive tenure, and establish credibility. And they are the primary means by which scientific knowledge is disseminated within the academy. Yet although publishing is one effective means to share knowledge, does the centrality of the peer-reviewed publication in the academy impact the type of knowledge that is produced?
I’d argue, yes.
In today’s academic science environment, a variety of incentives shape what research questions scientists ask and the type of knowledge that is ultimately published in the world’s top science journals. There are, of course, myriad additional ways in which the academic publishing system impacts our collective science knowledge, particularly in the space of knowledge dissemination, as I’ve argued elsewhere. But the scientific process starts by conceiving a research question, and the tentacles of journal publishers reach all the way to this initial step and beyond, influencing funding, training, and, ultimately, scientific progress.
Following the Money Trail
If one reads any tenure-track job ad in the sciences, grant funding is clearly a requirement for a successful academic research career. Although many universities offer new assistant professor “start-up funds” to launch their research programs, ultimately professors must win highly competitive federal research grants to fund their labs and career. To obtain these grants, academics must devise the “right” research questions that will yield results of interest, publish findings in the field’s top journals, and attract further funding.
Funding agencies, however, don’t simply fund any well-conceived research program. They fund specific research programs that align with the strategic goals of that agency, specific divisions, and programmatic initiatives. For example, the National Science Foundation (NSF), the premier federal funding agency for basic research, has a labyrinth of specific funding programs one must navigate and align their research proposals with the hopes of procuring funding. This is a problem because professors are now limited to researching topics that are well-aligned with current funding priorities. Without procuring funding, professors can’t train students, buy lab equipment, or travel to conferences, which could jeopardize being granted tenure at their institution and thwart their ascendance of the academic ranks.
Here’s where the publishing system impacts knowledge production: Once a professor has their research funded, they must publish their findings for the grant to be deemed “successful”—and to increase their chances of continued funding. Progress and final reports for NSF grants, for instance, require submission of published articles and conference papers to demonstrate the impact of their funded research. This cycle of funding and publishing further entrenches priority areas of research and continues to constrain open inquiry of research areas that may not fall within the scope of funders and journal editors.
And these impacts are not limited to professors. Graduate students—the next generation of scholars—are equally guided by the funding process. Grant funding allows professors to hire students and postdoctoral associates to work on preestablished lines of inquiry. Thus, the next generation of scholars’ own research paths are already guided by funders’ priorities. The academic publishing system thus shapes not only the questions that research labs investigate but also the training and education of those working in these labs who are ultimately the next generation of scientists. When guided by funding and publishing constraints, where’s the space for true open inquiry?
An Antiquated System
If the incentives of the academic publishing system are negatively impacting scientific knowledge production, what is the alternative? Is the academic publishing system all bad? It is obviously true that the traditional approach to scientific publishing has supported great advancements of knowledge in the past. After all, peer-reviewed publication helps ensure that we do not blindly accept scientific claims without understanding the methods by which such claims are based—a worthy goal indeed!
This argument, however, rests on the assumption that the academic publishing system we know today, which specifically relies on prepublication peer review, is a historically robust system.
But this is not necessarily true.
Coming into its modern form in mid-century America, academic journals were, in fact, the most effective way for scientists to communicate with one another. Peer review became standard across major journals such as Nature only in the 1970s after becoming increasingly common over the course of about a century. The academic publishing system has since morphed into a hugely profitable industry that now instead shapes knowledge production, inhibits dissemination, and can even reduce the impact of research findings because papers take months or years to publish and are then locked behind paywalls and institutional barriers.
The continued existence of this antiquated academic publishing system relies on the structural conservatism of the academy in the sense that it is resistant to change and innovation. This is, in part, why faculty are still hired on the basis of the journals they publishin rather than the quality of their research, lecturing has remained the dominant teaching approach in classrooms for more than a century, and technology continuouslyfails to live up to its promised disruption.
Although this type of conservatism slows innovation in education as an industry, the publishing business occurs outside the walls of the academy and should therefore be more amenable to new opportunities afforded by technologies, such as preprint servers, open source and cloud-based materials and data storage, open post-publication peer review, and much faster and cheaper publishing practices. In fact, we’ve already seen firsthand how these new approaches to scientific publishing afforded by the internet resulted in rapid scientific progress and societal impact during the COVID-19 pandemic.
In early 2020, as the world was gripped with fear and uncertainty about the spread of the SARS-CoV-2 virus, scientists were working to understand the virus and develop treatments and vaccines. Unlike previous global pandemics, however, scientists now had the internet—and they used it to share new research results with scientists, media, and the public at unprecedented speeds.
The star of early SARS-CoV-2 science was the preprint—a publicly accessible, unreviewed research paper that is posted to an online preprint server such as bioRxiv or medRxiv. What makes the preprint remarkable is the speed at which information can be disseminated at scale, discussed, fact-checked, and applied at much greater speeds than traditional prepublication peer review.
Had scientists used only the academic publishing system to disseminate the latest knowledge about SARS-CoV-2, the knowledge landscape would have likely looked different. Even with accelerated peer-review processes in place for COVID-related papers, research articles related to COVID-19 and SARS-CoV-2 still took on average 68 days to be published—a long time in a global pandemic—after being posted as a preprint, compared with 116 days for non-COVID papers in the same period.
Using only the academic publishing system would have been untenable to drive knowledge production and scientific progress needed during this time. Scientists used preprints and social media to amplify new insights and communicate with other scientists and the public. There are obviously trade-offs that come with this rapid and iterative open-access approach, including bad information and sloppy work being housed online. Yet moderation and post-publication and open peer-review processes can work to remedy these downsides.
Freed from the constraints of the academic publishing system, valuable knowledge about SARS-CoV-2 progressed at a far more rapid pace using preprints and social media, which provided actionable insights from scientists directly to public health officials, such as regular updates about the spread of COVID-19 based on genomic analyses. The early-pandemic science period also demonstrated that most scientists understand that the academic publishing system is a constraint on knowledge production rather than a catalyst for it. Scientists continue to push for policy changes at journals that favor preprints, such as allowing formal journal submissions and preprints on servers to coexist. Clearly, rather than a concerted effort to defend traditional publishing, open science and progressive publishing policies are becoming the norm.
The early-pandemic period of research demonstrated the utility and feasibility of a new approach to producing and disseminating insights. This period is representative of a broader movement that has been sweeping across the sciences in the last decade or so. For example, nonprofit organizations such as the Center for Open Science have launched to produce infrastructure to support open science, preprints and post-publication peer review are catching on, and new visions are being shared by scientists for a modern system of publishing.
Despite this progress, however, the centrality of peer-reviewed journal articles in a scientist’s academic career remains and is an exemplar of Goodhart’slaw: The focus of research has largely become to publish in the right journals rather than to pursue open inquiry to better understand and change our world. To loosen the grip of for-profit publishers on knowledge production and research dissemination, continued progress is needed.
Funding models must also progress in lockstep with these continuing changes in publishing norms. Proposals for changes such as lottery systems are promising to more fairly distribute grant funding across a variety of research programs. Other changes could include increased investment in public research universities earmarked for research: Rather than only start-up funds, a more consistent flow of funding for research professors could provide a foundation for diverse research inquiry across labs.
Most important, however, we must remember that systems are made up of people who respond to incentives. Academia has historically rewarded those who publish in the right journals, get the right funding, and work at the right universities. It makes sense, then, that work within academia has become centered around achieving these metrics of success. Many academics still scan curriculum vitaes of job applicants for prestigious journal names and tenure those who have published in the best journals. Too many still train their students to create research programs with the explicit goal of publishing in top journals to get those tenure-track jobs, and many still sit on the editorial boards of the for-profit journals we complain about.
Although policy can contribute to changes of incentives from the top down, people—academics—must lead change locally within their institutions from the bottom up. Individuals must change their own behavior, their own labs, and their own departments to break free from the constraints of a metric-focused academic career. We must not remain active participants. Our individual actions must culminate in new systems that reward open inquiry, scientific discovery, and learning. Although the conservative nature of educational institutions will most likely remain a challenge to progress, I believe we are on the right path toward positive change in the academy.
Your generosity supports our non-partisan efforts to advance the principles of open inquiry, viewpoint diversity, and constructive disagreement to improve higher education and academic research.