This article is by Dashun Wang, associate professor of management and organizations at the Kellogg School of Management, Northwestern University.
Could we have predicted when Picasso would paint his most groundbreaking pieces? Or when Marie Curie would perform her Nobel Prize-winning research on radioactivity? Or when Steve Jobs would have his greatest technology breakthroughs?
The simple answer, suggested by our research, is no. We probably can’t predict the timing of impact in a given field. But we may be able to forecast who is more likely to have the greatest impact.
My fellow researchers Roberta Sinatra, Pierre Deville, Chaoming Song, and Albert-László Barabási, and I studied the careers of thousands of scientists to understand when they did their most impactful, most cited work, and whether any factors, such as productivity, reliably predicted a scientist’s overall impact.
We found first that peak impact can happen at any time in a scientist’s career. This runs counter to the conventional wisdom that most scientists, innovators, and artists are in learning mode early in their careers, peak somewhere in the middle, and decline in impact as the sun sets on their professional lives. Second, there are ways of predicting overall career impact based on early performance, but it has less to do with productivity than many believe. Our findings about that may have implications for fields far beyond science, including business and the arts.
Our research started with this question: When do great scientists do their best work? We knew, as researchers who have been considered for promotion and who have sat on review committees, that the ability to predict the timing of impact could be of value to multiple decisions in academic science, including ones about promotions and the awarding of grants. To study the timing of impact, we analyzed the the publication (in other words, productivity) and citation (impact) records of more than 10,000 scientists in a wide range of disciplines, including physics, chemistry, ecology, and economics. We looked at how many papers the researchers had published overall and when they had published their most cited works, and we looked for patterns related to what came before and after this highest-impact work.
What we found was surprising: The timing of one’s greatest impact in a science career is remarkably unpredictable. What I’d call the Random Impact Rule suggests that peak impact can happen at any time in one’s career, beginning, middle, or end. Moreover, there is no pattern leading up to or following that impact; nothing predicts what happens before or after the highest impact work. This unexpected finding, which flies in the face of conventional wisdom, may provide some solace to late-career scientists who haven’t yet had the effect they’d like.