Science’s Safest Path Is Smothering Big Ideas, Major New Analysis Warns

Scientific creativity is not, according to a sweeping new perspective published in Nature Human Behaviour, primarily a matter of individual genius. It is, the authors argue, an emergent property of the academic ecosystem in which researchers operate—and that ecosystem is failing. In a large international analysis led by Phillip J. Haubrock of Bournemouth University and Teun Everts of the Research Institute for Nature and Forest in Belgium, more than fifty researchers from institutions spanning six continents contend that contemporary academic structures progressively constrain high-risk, conceptually innovative research while rewarding safe, predictable, and easily evaluated outputs. The result, they warn, is a science that is measurably less transformative precisely at the historical moment when complex global problems demand the opposite.

The core of the argument rests on a systems-level diagnosis. Modern science, the authors write, is structured around short funding cycles, productivity-based evaluation criteria, and risk-averse frameworks that favour non-transformative research. Peer review and grant allocation, escalating publication costs, and institutional inequalities interact with precarious employment and hierarchical dependencies to create an environment in which strategic conformity becomes the rational career choice. For early-career researchers in particular, whose futures depend on publication records, citation metrics, and the goodwill of senior collaborators, deviation from established methods and fashionable topics carries disproportionate risk. The probability that a young scientist will attempt a genuinely disruptive project has therefore declined, while talent increasingly either abandons academia or clusters within a small number of well-resourced institutions and national systems.

The technical literature the authors assemble paints a quantitatively troubling picture. Drawing on large-scale bibliometric studies, they note that papers and patents have become systematically less disruptive over time, a trend documented in analyses of citation patterns across decades. Research on scientific teams shows that large consortia tend to consolidate and extend existing knowledge while small teams are disproportionately responsible for disruptive advances—yet funding structures increasingly favour the former. Experimental work on grant evaluation demonstrates that reviewers penalise novelty: novel projects are systematically scored lower and funded less often, and low agreement among reviewers evaluating the same application suggests that much of this filtering is closer to noise than to calibrated judgement. Studies of funding distribution further reveal a pronounced Matthew effect, in which already-successful researchers and institutions capture a growing share of resources, compounding advantage and starving peripheral researchers of the capital needed to take intellectual risks.

Employment precarity amplifies every one of these pressures. The authors synthesise evidence that the academic labour market produces far more doctoral graduates than stable positions, with the ratio of PhDs to permanent posts described in earlier modelling as resembling an uncontrolled reproductive number. Fixed-term contracts, repeated relocation, and dependence on individual principal investigators mean that early-career researchers operate under conditions in which a single failed grant cycle or slow publication period can end a career. Longitudinal studies from Sweden show measurable deterioration in mental health among PhD candidates, while global surveys report widespread academic fatigue, burnout, impostor syndrome, and repeated rejection. In such an environment, the authors argue, creativity is a luxury: the psychological safety required for exploratory thought is simply absent when employment security is measured in months.

The evaluation systems that govern careers compound the problem. Publication counts, journal impact factors, and h-indices remain central to hiring, promotion, and tenure decisions in many systems despite long-standing critiques, including the San Francisco Declaration on Research Assessment and the Leiden Manifesto, warning against exactly this practice. The authors cite evidence that prestigious journals struggle to reach even average reliability, that metrics are routinely gamed in ways consistent with Goodhart’s law, and that hypercompetition for publication slots has been linked to declining integrity, including rising retraction rates and exploitative authorship practices such as so-called abusive coauthorship and publication parasitism. When careers are scored on quantity and venue rather than substance, rational researchers optimise accordingly, producing incremental papers aimed at citable niches rather than ambitious work aimed at shifting paradigms.

Economics sharpens the picture further. Open-access publishing models that shift costs onto authors have created a system in which the ability to publish increasingly depends on the ability to pay, with article processing charges of thousands of euros per paper. Analyses cited in the perspective show that payment-based open access systematically biases participation against researchers from the Global South, that African scholars face acute dilemmas over whether they can afford to publish at all, and that transformative agreements between libraries and publishers can reproduce global inequalities under a progressive banner. Rising costs interact with geographic concentration of research funding: R&D expenditure remains heavily skewed toward a handful of wealthy nations, and the authors warn that the privatisation and nationalisation of research agendas risks narrowing the global diversity of scientific questions even as the number of published papers explodes.

Peer review, the gatekeeping mechanism at the heart of scientific quality control, receives particularly pointed treatment. The authors document a deepening peer-review crisis: reviewer pools are shrinking, twenty-one years of data from one major ecology journal show quantifiable declines in reviewer availability, and the unpaid labour that underpins the system has been estimated to represent a billion-dollar annual donation of researcher time. At the same time, evidence indicates that reviewers show systematic conservatism, that unprofessional reviews disproportionately harm underrepresented groups in science, and that grant peer review suffers from widespread distrust and low predictive validity. Experimental proposals to mitigate these failures—including lotteries and modified lotteries for funding allocation, double-blind review to reduce bias, and even professionalising the reviewer role—are discussed as evidence that the community recognises the dysfunction but has, so far, reformed around its edges rather than its core.

What would meaningful structural reform look like? The authors argue that scientific creativity must be treated as an ecosystem property, which implies interventions at the level of institutions rather than individuals. Their proposals include promoting stable, long-term research pathways that decouple employment from short-term productivity; decentralising decision-making so that resources and agenda-setting power are less concentrated in elite institutions and senior hierarchies; and reforming evaluation frameworks to explicitly recognise collaboration, originality, persistence, and nonlinear career trajectories. They point to existing models as proof of concept: sustainable employment structures piloted in the German higher education system, funding-by-lottery trials at the Swiss National Science Foundation, article-level metrics that would give most researchers more recognition than journal-level metrics, and slow-science approaches that protect the long timelines over which foundational work matures. Fifty-year ecological studies, they note, cannot be built in a three-year grant world.

The stakes, the authors conclude, extend well beyond the welfare of individual researchers. Without systemic change, they warn, science risks stifling the capacity of early-career researchers to move beyond existing methods and deliver transformative advances, limiting their ability to change our understanding of the world or to benefit society. The perspective explicitly asks whether there could be another Einstein under current conditions and answers, in effect, that the question is misframed: exceptional creativity has always depended on exceptional conditions, from the research freedom documented at institutions such as the Max Planck Society to the protected time that historical figures enjoyed. Rebuilding those conditions at scale—through stable careers, distributed power, honest metrics, and tolerance for risk and failure—is, the authors argue, not a matter of nostalgia but of scientific necessity. The alternative is a research enterprise that produces ever more papers and ever fewer revolutions.

The intellectual roots of this argument reach back further than the current funding crisis. The perspective situates itself within an economics-of-science tradition stretching to the 1990s, when scholars first described basic research as a public good whose incentives could not be left to market logic alone. That framing has acquired new urgency as commercialisation pressures and national competitiveness agendas increasingly shape which questions are asked, and by whom, in laboratories around the world.

Bibliometric work offers a concrete sense of what lost creativity looks like in the data. Analyses of atypical combinations in references show that the highest-impact papers tend to draw on unconventional pairings of established literatures, yet such combinations have become rarer as career incentives push researchers toward safer citation neighbourhoods. Studies of scientists’ research strategies similarly distinguish between tradition-oriented and innovation-oriented projects, finding that the most celebrated work often emerges from riskier strategies whose payoffs are delayed and uncertain—precisely the profile that short evaluation cycles systematically disfavour.

The career consequences are not merely hypothetical. Research tracking early-career trajectories suggests that an early setback, such as missing out on a first major grant, can have lasting effects on whether a scientist remains in academia at all, filtering out individuals in ways unrelated to their eventual scientific potential. Combined with the well-documented expansion of doctoral training since the early 2010s, this creates a selection regime in which persistence, not originality, becomes the decisive trait.

The authors also point to quieter structural frictions that rarely make headlines. The growth of administrative roles in universities, the burden of unpaid peer review falling disproportionately on junior researchers, and the rise of precarious professional staff taking on faculty-critical functions all divert time and autonomy away from the exploratory work on which conceptual breakthroughs depend. Their conclusion is that restoring creativity will require attending to these mundane institutional details as much as to headline reforms of funding and metrics.

Subject of Research: How contemporary academic structures, funding systems and career precarity constrain scientific creativity among early-career researchers

Article Title: How contemporary academic structures constrain scientific creativity and hold back early-career researchers

Article References: Haubrock, P. J., Kouba, A., Soto, I., Abreo, N. A. S., Vilizzi, L., Błońska, D., Tarkan, A. S., De Santis, V., Sousa, R., Nogueira, A. B., Heeren, S., Nogueira, J. G., Cano-Barbacil, C., Cremella, B., Strubbe, D., Silva, J. P. D., Poloni, R., Deschepper, P., Thoré, E. S. J., … Everts, T. (2026). How contemporary academic structures constrain scientific creativity and hold back early-career researchers. Nature Human Behaviour. https://doi.org/10.1038/s41562-026-02575-5

Image Credits: AI Generated

DOI: 10.1038/s41562-026-02575-5

Keywords: academic careers, scientific creativity, early-career researchers, research funding, peer review, publish or perish, research evaluation, job precarity, open access costs, Matthew effect, science policy, disruptive research