The rich have been getting richer in the biomedical research enterprise, and the system favors those who are already doing pretty well, according to a new analysis of National Institutes of Health (NIH) grant recipients.
The top 10% of NIH grant winners (by total award) received about 37% of NIH funding in 2015—up from the 32% the top group got in 1985, but down a bit from a peak of 40% of total funding in 2010, according to an analysis published last week by researchers at the Berkman Klein Center for Internet & Society at Harvard University.
In contrast, the bottom 40% of principal investigators (PIs) won approximately 12% of the pot in 2015, down from 16% in 1985 and not a dramatic change from 11% in 2010.
The analysis also found that the amount of grant money PIs in the top and bottom echelons had varied dramatically. In 2010, the median funding of PIs in the top 10% was $1.45 million per PI, whereas the median funding of PIs in the bottom 40% was $140,000 per investigator.
The results were similar when viewed by university. In 2015, the top 10% of grant-winning universities received 75% of the NIH total, whereas the bottom 40% won only 0.6%. That’s only a slight shift from 1985, when the top 10% of universities won 70% of NIH funds, whereas the bottom 40% won 1%.
To produce the numbers, researchers tracked the amount of money NIH awarded to grant recipients between 1985 and 2015, as well as publications and patents related to the awards. Lead author Yarden Katz, a fellow at the center and in the department of systems biology at Harvard Medical School in Boston, says they were interested in examining the relationship between NIH funding and metrics widely used to indicate scientific quality and determine career advancement, such as publication, citation, and patent counts.
The skew in NIH funding that the researchers found is not surprising—NIH has noted the inequality in its own analyses. But the new study serves as a reminder that the inequality has persisted for decades, says Chris Pickett, executive director of Rescuing Biomedical Research in Washington, D.C. The data show “why you can have people at universities where the university ranks 30th or 40th in total NIH funding coming in, and they feel like they’re barely scraping by,” Pickett says. “It’s because there’s a lot of money concentrated in the top institutions.”
Strong funding corresponds to other metrics as well. For instance, PIs in the top 20% of NIH funding had more publications and more citations than those lower on the funding tally. But there were noticeable differences even at the top of the pile: PIs in the top 90th percentile had nearly 1.4 times more publications than PIs one step down in the 80th percentile.
PIs publishing in high-profile journals tended to have more funding, too, the study found. For example, the median funding level of PIs publishing in the biomedical publishing world’s big three—Nature, Science, and Cell—was in the 80th percentile, with many clustering around the top 90th funding percentile. The median funding level for PIs publishing in journals with lower impact factors—such as PLOS ONE and the Journal of Biological Chemistry—was not much lower, about the 70th percentile, but the distribution of funding brackets was more diverse.
For new PIs, those who started strong in winning grants tended to stay on top. Within the first 8 years of their career, PIs whose first grant landed them in the top 10% of NIH funding stayed in that top rung for an average of 5 years. In contrast, PIs starting in the bottom 40% stayed in that rung for 4 to 6 years, on average, before transitioning out, which could include moving up or dropping out. That stagnancy suggests that where early career researchers start on the funding ladder “may have an important effect on their career,” Pickett says.
Patents—a favored metric for scientific innovation and another source of income for institutions—were also concentrated in the hands of a few: Ten percent of institutions hold 80% of the NIH-funded patents. That might be because many universities don’t have the financial and legal resources necessary to pursue a lot of patents, Katz notes. But “if we’re saying patents are required to show that you are an innovative scientist or to get funding, then again this will … privilege this small minority of institutes,” he says.
The study comes in the wake of an NIH decision to drop a controversial plan to cap how much grant funding an individual investigator can win from the agency, in part to free up funding for more early- and midcareer researchers. Katz, for one, was disappointed by the decision, which he sees as the result of strong resistance from biomedicine’s better-off elite. But he says he’s committed to staying out of the metrics race. For example, since 2014, he says he has published only in open-access journals. The fixation on prestige publications and citation numbers for career advancement takes away from the creativity and diversity in science, Katz believes, and he says “I’m not going to ruin my time in science trying to optimize these things.”