How the Fear of Coming in Second Hurts Scientific Progress

The drive to be first in science can turn discovery into fierce competition. Just ask Isaac Newton and Gottfried Leibniz, whose competing 17th-century claims of discovering calculus kicked off a feud that delayed collaboration between British and continental mathematicians for decades. Or Charles Darwin and Alfred Russel Wallace, who both laid out theories of evolution by natural selection.  

For today’s scientists, the race for recognition still shapes research agendas and careers. But while coming in second has its downsides, the penalties are not as severe as many think. What’s more, the rush to be first may be lowering scientific quality overall.

That’s according to a pair of related studies published in the Journal of Political Economy and The Quarterly Journal of Economics by Carolyn Stein of Berkeley Haas and Ryan Hill of Northwestern Kellogg.

“As a graduate student, I learned that other people were working on a much bigger, better, more ambitious version of the exact same project I was working on and I felt like, ‘Wow, this has derailed my whole career,” recalls Stein, an assistant professor of economics. “So, we got interested in this idea of getting scooped.”

Before Stein and Hill could explore that concept, they first had to solve their own research problem. How could they study competitiveness when most scientists either tend to abandon or modify research if they don’t publish first?

The answer came from a surprising source: structural biology’s Protein Data Base (PDB), a worldwide repository of biological macromolecules, which provides near-universal coverage of structure discoveries. Unlike other research efforts, scientists submit structures to the PDB months before publishing, leaving a record of their work—even if they ultimately abandon the project.

“It was a very specific solution to this data problem we had,” Stein says. “But then to actually use that data, we had to learn a lot more about the field, which opened up new avenues and other interesting things to explore.”

The modest benefits of being first

The first avenue Stein and Hill explored were the empirical benefits of being first to publish, as revealed by the PDB. While the study showed some rewards for scooping the competition, second place finishers still garner attention: In fact, scooped papers receive only about one-quarter fewer citations, according to an analysis of more than 1,600 ‘races’ from the PDB.

“It’s meaningful, but it’s not devastating,” Stein says.

Hill and Stein also surveyed 877 structural biologists and found that they significantly overestimated both the likelihood and costs of getting scooped. Respondents guessed that they had a 27% chance of getting scooped by a competitor, when the actual probability was only 3%. They also estimated that a scooped research project would get 59% fewer citations, when in reality the penalty was nearly three times smaller.

But there’s an interesting wrinkle. The degree to which being scooped matters corresponds with how long it took the secondary finding to be published. “If you’re very close behind, it tends to go better, possibly because your paper is already being reviewed when the other one comes out,” Stein posits.

The hidden dangers of competitiveness

Those findings brought up another question the PDB allowed the researchers to explore: What impact does competitiveness have on the quality of the research itself?

As Hill and Stein expected, the more competitive a particular structure-solving race was, the more the researchers rushed their work—resulting in lower-quality findings, as represented by the accuracy of the 3D models and how closely the models matched the researchers’ experimental data.

Furthermore, an analysis of real-life projects in the PDB revealed that the problem was particularly acute among the most important protein structures— for example, those linked to human proteins or diseases. These proteins tended to incite the most interest, and thus competition—therefore leading to more rushing and lower quality.

To further verify those findings, Stein and Hill surveyed a random group of scientists, posing a hypothetical scenario in which they were working on a competitive project. They then asked this group how long they would wait to send their results to a journal for publication and which measures they would take to ensure their results were robust before doing so. Next, they did the same thing with another control group, but this time with the understanding that it was not a competitive scenario.

“The people who got the ‘likely to be competitive’ version of the survey reported that they would finish it much faster and do fewer of these quality-control items,” Stein says.

Wider economic implications

This “race to the bottom” has real consequences. Stein and Hill estimate that the costs associated with improving the low-quality structures that resulted from competitive pressure are between $1.5 billion and $8.8 billion since the PDB’s founding in 1971.

But rather than demonizing competition, Stein hopes the findings encourage other economists to think about the incentives in science, how they differ from other economic models, and the implications of those differences. “Basic science is sort of a weird area, where you invest a lot and you learn something great about the world, but the person who discovered it is not going to necessarily make a lot of profit from doing that,” she explains.

“The individual lab model is very old. It’s probably worth asking what projects it makes sense for, and should there be other ways of structuring science?”

—Assistant Professor Carolyn Stein

As such, she hopes the research makes other scientists consider different ways to organize and incentivize research. For example, the Human Genome project encouraged collaboration with great results. “The individual lab model is very old,” she says. “It’s probably worth asking what projects it makes sense for, and should there be other ways of structuring science?”

Read the full studies:

Scooped! Estimating Rewards for Priority in Science

By Carolyn Stein and Ryan Hill

Journal of Political Economy, March 2025

Race to the Bottom: Competition and Quality in Science

By Carolyn Stein and Ryan Hill

The Quarterly Journal of Economics, February 2025