At 26, Albert Einstein rewrote physics. By his sixties, he spent the rest of his life arguing against quantum mechanics — the very theory that would shape physics for the next century. The man who shattered the old scientific worldview eventually became one of its staunchest defenders. It turns out this is more than the story of a single genius. It is a statistical pattern, now measured across 12.5 million scientific careers.
The study at the center of this discussion is Aging and the Narrowing of Scientific Innovation by Lingfei Wu, Yiling Lin, James Evans, and Haochuan Cui, published in Science on May 7, 2026. Alongside Nature, Science is widely regarded as one of the world’s two most prestigious scientific journals, making publication there a strong signal that the findings deserve attention. The paper’s DOI is 10.1126/science.ady8732, allowing it to be located through any major academic database. This article also draws on an earlier study by Park, Leahey, and Funk, Papers and Patents Are Becoming Less Disruptive Over Time, published in Naturein 2023.
The conclusion is both simple and uncomfortable for many advanced economies: the older a country’s scientific workforce becomes, the less likely it is to produce truly disruptive discoveries. China and India, with younger generations of researchers, publish a larger share of work that fundamentally changes the direction of entire fields. The United States and Japan, where scientific communities are considerably older, increasingly produce careful, incremental research that refines existing knowledge rather than overturning it. The authors call this phenomenon the “nostalgia effect.”
Wu’s team analyzed six decades of scientific records, covering the period from 1960 to 2020, and reconstructed the careers of 12.5 million researchers by examining who they cited and who cited them.
For every paper, they calculated two measures.
The first was the disruption index. If future studies cite a paper directly while bypassing the earlier literature on the topic, that paper has genuinely shifted the field. The scale ranges from -1 to +1. Scores approaching +1 indicate a paradigm-shifting breakthrough. Scores around zero or below reflect work that extends existing ideas, while -1 represents research that consolidates previous knowledge without introducing major novelty.
A classic example is Watson and Crick’s 1953 paper describing the structure of DNA. It redirected biology so decisively that subsequent research cited their work directly rather than the earlier literature it had replaced.
The second metric was the novelty index, which captures a different kind of creativity. Papers scoring highly on novelty combine existing ideas in original ways without necessarily overturning established theories. The numbers that emerge are difficult to ignore.
The researchers defined academic age simply as the number of years since a scientist’s first publication. As academic age increases, the probability that a new paper ranks among the top 10% most disruptive declines steadily. Researchers in the first decade of their careers are far more likely to generate ideas that change the rules of the game. Senior scientists are more likely to refine, combine, and improve what already exists.
Even their intellectual references grow older over time. In medicine, early-career researchers cite papers with an average age of 7.9 years. By the end of their careers, that average rises to 10.1 years. In the social sciences, the shift is even more pronounced, with the literature researchers rely on aging by 13 to 17 years over the course of a career.
The study also identifies the mechanism through which this pattern perpetuates itself. Senior scientists run laboratories, supervise junior researchers, and review manuscripts submitted for publication. Wu and his colleagues compared preprints — manuscripts before peer review — with their final published versions.
The pattern was remarkably consistent. After peer review by senior scholars, younger researchers tended to cite significantly older literature. In other words, nostalgia is not simply an individual habit. It spreads through the scientific hierarchy, shaped by the people who ultimately decide what is worthy of publication.
Russell Funk, whose earlier research documented the long-term decline in scientific disruption, puts it bluntly: as the scientific workforce ages, the system increasingly rewards consolidation over breakthrough innovation. James Evans reaches a similar conclusion. Senior scientists remain highly creative, but they are more likely to recombine familiar ideas than embrace fundamentally new ones. Not everyone in the scientific community accepts the methodology without reservations.
Researcher Mikko Packalen argues that citation-based indicators cannot fully capture conceptual originality and should ideally be complemented by direct analysis of the papers themselves rather than citation networks alone. It is a fair criticism. But it does not overturn the central finding. The relationship between academic age and the type of innovation researchers produce appears consistently across an exceptionally large dataset, making it difficult to dismiss as a statistical artifact.
This is where an academic observation becomes a question of national competitiveness. Countries with younger research communities consistently produce a larger share of disruptive science. That is precisely where China and India now hold an advantage. The United States and Japan, where mandatory retirement has largely disappeared and academic careers stretch over increasingly long training periods before researchers gain independence, are moving in the opposite direction. There is another revealing finding.
Immigrant scientists in the United States, who are younger on average than the domestic research workforce, significantly raise the country’s overall level of disruptive research.
The effect has little to do with nationality and everything to do with researchers’ age and scientific trajectory — an inconvenient reality for debates over who truly belongs in a national scientific system. “We need to welcome young talent from around the world,” Wu argues. “Society should remain open to exchange students, international students, scientific talent, and immigration.” His point is straightforward. Young international researchers are often less attached to established schools of thought within a particular country. That distance is an advantage rather than a weakness.
The authors are not calling for older scientists to step aside. Instead, they advocate building flat, intergenerational research teams, where younger scientists work alongside senior colleagues as equals rather than under rigid hierarchical control. They also argue that early-career researchers should be given opportunities to lead their own laboratories much earlier than is common today.
Recent scientific history offers striking examples. CRISPR, the revolutionary gene-editing technology, emerged from relatively young laboratories willing to risk their reputations on an unproven idea. AlphaFold, the AI system capable of predicting protein structures in hours rather than years, was largely advanced by postdoctoral researchers.
The same pattern appears in the development of mRNA vaccines. For years, Katalin Kariko pursued the idea from the margins of academic science rather than its institutional center, long before it became the foundation of COVID-19 vaccines.
Different stories. The same pattern. Disruptive innovation consistently clusters among researchers who are younger and less constrained by established ways of thinking. Wu’s paper is not the first warning signal. Back in 2023, David Park, Erin Leahey, and Russell Funk analyzed 45 million scientific papers and 3.9 million patentsspanning six decades using the same disruption index. Their findings were remarkable.
Average disruption declined by more than 90% for scientific papers published between 1945 and 2010 and by over 78%for patents between 1980 and 2010. Importantly, the absolute number of breakthrough discoveries did not collapse. Instead, truly disruptive work became increasingly diluted by the explosive growth in overall scientific output. Earlier work led by James Evans, published in Nature in 2019, reached another complementary conclusion: small research teams are more likely to challenge established ideas, while larger teams tend to develop and expand existing ones.
Researcher age, team size, and the relentless growth of scientific literature are different expressions of the same underlying reality. Science has accumulated so much knowledge that overturning it has become increasingly difficult. Taken together, these studies point to one unmistakable conclusion. Science is becoming less disruptive on average, and the aging of the global research workforce is closely linked to that trend.
What makes the 2026 study particularly significant is that it extends the argument beyond individual careers to entire national research systems. For the first time, it directly compares countries such as China and India with the United States and Japan, showing how demographic differences within scientific communities translate into differences in breakthrough innovation.
Einstein at 26 and Einstein at 60 are no longer simply two chapters in one extraordinary life. They may also represent the life cycle of modern science itself.
Elvira Baturova, expert at EconomyKZ.org
