The Pressure to Publish Is Producing Science Nobody Needs — and Some That Never Existed

In 2023, the journal Retraction Watch recorded its one hundred thousandth retracted paper in its database — a milestone that would have been unthinkable two decades earlier. The number itself is not straightforwardly damning; better detection tools and a more vigilant post-publication review culture account for some of the increase. But the underlying trend is harder to explain away. Retraction rates have grown faster than publication rates. Fraudulent or severely flawed papers are appearing with greater frequency not despite the pressure to publish, but in significant part because of it. The system designed to advance knowledge is, at an accelerating pace, producing knowledge that must be recalled.

This is one face of a broader pathology. The other is quieter and in some ways more consequential: the vast accumulation of published research that is methodologically sound, correctly reported, and entirely irrelevant to anything beyond the narrow tenure case of the researcher who produced it. Together, these phenomena — the fraudulent and the merely purposeless — represent the full cost of a culture that has confused output with achievement.

The historical roots run deeper than most researchers realise. The phrase "publish or perish" is commonly attributed to sociologist Logan Wilson, who used it in a 1942 study of academic life — but the pressure it names did not become structurally determinative until the late 1970s and 1980s, when universities began adopting quantitative metrics as proxies for research quality. Bibliometrics, citation indices, and impact factors were developed as analytical tools for librarians making collection decisions. They were gradually repurposed as instruments of career evaluation, a transformation their original architects warned against explicitly and institutions largely ignored.

As these metrics hardened into institutional policy through the 1990s, the incentive gradient shifted in ways that are now well-documented. Publishing frequently in indexed journals became not merely one indicator of productivity but the dominant one. Positive, novel results were what journals wanted and what careers required. Negative results — the failed experiments, the non-replications, the careful null findings that are essential to the actual functioning of science — became career liabilities. Researchers stopped submitting them. Journals stopped publishing them. The literature began to fill with results that were systematically more positive, more novel, and less reliable than the underlying reality of the science warranted.

The mechanism linking pressure to fraud is not primarily one of deliberate bad actors, though those exist. It is one of accumulated small compromises. A researcher who rounds a p-value, omits a failed experimental run, or frames an ambiguous finding as definitive is usually not committing what they would recognise as misconduct. They are responding to a system that has made ambiguity professionally dangerous. Over time, across thousands of researchers making thousands of small adjustments, the literature drifts away from truth in ways no individual feels responsible for.

The irrelevance problem operates through a different mechanism but springs from the same source. When publication is the goal rather than the means, research topics are selected for their publishability — their fit with current journal priorities, their amenability to clean experimental design, their location within well-funded and therefore well-cited fields. Topics that are genuinely important but methodologically messy, underfunded, or simply unfashionable are systematically underexplored.

The result is what some science policy researchers have called a "crowding effect": resources and talent concentrate in areas where publishing is easy and rewarding, while pressing real-world problems — rare diseases, agricultural challenges in low-income countries, materials science for decarbonisation — struggle to attract sustained attention because the career returns are uncertain. The literature grows dense in some areas and thin in others, not because of where the problems are, but because of where the publications are.

For the individual researcher, the consequences arrive as a creeping professional dread. The scientist who suspects her most recent paper will not be cited, will not inform policy, will not change a clinical practice or seed a useful technology, has no good language for that suspicion within the current system. The system does not ask whether the work matters. It asks whether it was published, and where, and how often. Mattering is not a metric.

The cumulative toll — on researchers' sense of purpose, on the trustworthiness of the literature, and on the relationship between science and the public that funds it — is difficult to quantify precisely because it operates through absence. Papers not written because the finding was negative. Questions not asked because the answers were unpublishable. Trust not extended because too many previous findings turned out not to hold.

The corrective that the system most urgently needs is not another layer of post-publication review, nor a new impact metric designed to replace the flawed ones. It is an honest, external signal of real-world value — applied early, before publication pressure shapes the direction of the work.

A rigorous market check — not in the narrow commercial sense, but in the broader sense of asking who actually needs this finding, what problem it solves, and whether the world would be materially different if it existed — gives researchers something the current system does not: a purpose-oriented reference point that competes with the publication incentive. When researchers know that their work has demonstrable value beyond a journal's acceptance criteria, they are better equipped to resist the pressure to oversell, to selectively report, or to pursue topics chosen for their CVs rather than their consequences.

This is the work of an honest partner who understands both the science and the landscape outside the laboratory: someone who can conduct a genuine market and impact assessment, translate it into a compelling case for funders and institutions, and help a researcher move from the exhausting calculus of publish or perish toward something more defensible — and more motivating. Not "was this published?" but "does this matter, and to whom, and how do we make sure the right people know it?" That reorientation does not require dismantling the academic publishing system. It requires, for each researcher who is ready, a way out of the logic it imposes — and toward work whose value can be demonstrated in terms the world, not just the journal, can understand.