NIH Grants Show Racial Divide

Black scientists who apply for grants from the National Institutes of Health have less chance of receiving funding than their white counterparts. This is the conclusion of an NIH-commissioned study, which has prompted intense discussions about potential racial bias in NIH’s peer review process and has sent agency officials scrambling to find the causes of and solutions to the problem.

The study, published last month in Science (DOI: 10.1126/science.1196783), revealed that, between fiscal 2000 and 2006, self-identified black and Asian investigators were less likely to be funded by NIH than whites. The difference in success rates between Asians and whites, however, disappeared when only U.S. citizens were considered in the sample population. In contrast, black scientists were still 10 percentage points less likely than whites to receive an NIH grant, even after controlling for factors such as former training, publication record, and institutional affiliation. The results showed no apparent differences between NIH success rates for Hispanics and whites (C&EN, Aug. 22, page 8).

NIH subsequently looked at other data sets, including data from years since 2006, and found similar results, NIH Director Francis S. Collins says. One reasonable explanation for the difference between Asians and whites may be that Asians who are not U.S. citizens are often not native English speakers, Collins notes. But the agency is still scratching its head, trying to explain why there is such a large difference in success rates between black and white investigators.

Calling the data “deeply troubling,” Collins says NIH will take immediate steps to better understand the causes of the problem and seek interventions. He points out that only about 1% of NIH grant applicants are black. In contrast, blacks represent about 10% of the U.S. population, he says. NIH has “failed to recruit the best and brightest minds,” Collins stresses.

NIH considers the new study an opportunity to address a problem that has existed for many years. “We now have data that can guide us in terms of actions,” Collins says. Those plans are outlined by Collins and NIH Principal Deputy Director Lawrence A. Tabak in a commentary in Science that accompanies the study (DOI: 10.1126/­science.​1211704).

One of the actions described in the commentary is evaluation of the accessibility of mentoring programs. “We are concerned about whether mentoring is as available to junior faculty members in all situations and for all individuals,” Collins says. “NIH will assess the value of providing additional assistance to applicants in grant preparation and supporting innovative approaches to encourage more extensive and effective local mentoring of junior faculty by the academic institutions that support them,” he notes.

NIH will also develop an early-career program to encourage promising junior faculty to participate on a peer review panel. Service on review panels has been shown to correlate with success in receiving NIH grants, but investigators typically do not serve on such panels until they are far along in their careers.

In addition, NIH will conduct experiments to address the possibility of bias in its peer review process, Collins notes. Peer reviewers do not see information about an applicant’s race and ethnicity, but they can usually figure out such information with a quick Internet search of the applicant’s name and affiliation.

Reactions to the study results and NIH’s plans to address the problem have been mixed. Some people, including the study authors, say they were surprised by the results. Others say they were not.

“I thought some of the differences would be explained by factors like the types of institutions where minority scholars tend to be, which may not be the most competitive in terms of research intensity,” says study coauthor Raynard S. Kington, president of Grinnell College and a former deputy director of NIH. “I was disappointed that we weren’t able to explain the differences,” he tells C&EN.

“Either there aren’t real opportunities for the best minds to have successful ­careers in science, or our systems of review need to be improved,” Kington says. “Both of those are serious concerns.”

Kington agrees that the next step is to try to understand the cause of the problem and to develop an array of solutions. At the same time, he says, more needs to be done to increase diversity in the biomedical workforce. Many of the programs designed to do that, some of which have existed for 30 years, “just haven’t achieved what we had hoped they would achieve,” he notes.

Other researchers weren’t as surprised by the study’s results. “I was actually impressed that somebody bothered to quantitate the difference, but I wasn’t surprised by the results,” says Malika Jeffries-EL, an assistant professor of chemistry at Iowa State University. “Many people don’t have the mentoring” needed to successfully navigate the NIH system, she says.

Several chemists tell C&EN that they have benefited greatly from a mentoring program for junior faculty in organic chemistry and chemical biology sponsored by NIH’s National Institute of General Medical Sciences, which funds more chemistry-related research than any of the agency’s other 26 institutes or centers. The program was launched in 2005 by now-retired NIGMS program director John M. Schwab and Michael P. Doyle, a chemistry professor at the University of Maryland, College Park.

The mentoring program brings in junior faculty “to show them what actually happens at these NIH review panels,” says Richmond Sarpong, an associate professor of chemistry at the University of California, Berkeley. “They do mock reviews, and senior faculty who have experience with NIH are brought in to give direction,” he explains.

Sarpong attended the workshop as a new faculty member and later served as one of its mentors. He says that other institutes within NIH should consider launching similar programs.

“You have to serve on review panels,” echoes Jeffries-EL. “It gets you oriented to what is acceptable and what is important, including the subtle things you think are benign that might build allegiance or tick off a couple of reviewers,” she says.

Insidious bias in NIH’s peer review process, however, cannot be ruled out as a potential cause for the large difference in success rates between black and white scientists. “Well-meaning Americans can’t believe it, but there is this subtle bias that people have,” Jeffries-EL says.

NIH officials say they are committed to addressing the possibility of bias in the NIH peer review system. “We will conduct innovative experiments to shed light on possible sources of such bias and to develop appropriate intervention,” NIH’s Collins notes. “We will do experiments where we remove all the identifying information from applications and assess whether that has an effect on the priority score,” he says.

Some researchers say the problem goes deeper than just differences in grant success rates. The real issue is “how to get underrepresented minorities into positions at the top institutions,” Sarpong says. Others agree with Collins that it is important to increase the number of proposals submitted by blacks and other minorities.

Achieving a diverse grant pool has been a focus for the National Science Foundation, which collects detailed data on its grant applicants and awardees. In response to the Science study, NSF reports that it sees less of a difference in success rates among races. “We have in our merit review process a well-established system that ensures that we take into consideration the issues regarding diversity,” says NSF spokesman Bobbie Mixon.

To ensure participation by black scientists, NSF goes to colleges and universities and teaches researchers who don’t typically apply for NSF grants how to submit proposals to NSF, Mixon tells C&EN. “We also have a course on diversity that is available to our program managers and our program directors,” he says. In addition, NSF includes diverse members of the research community on review panels that evaluate proposals, he notes.

NSF reports data annually on its grant success rates for underrepresented minorities. Interestingly, blacks are about 4% less likely than whites to receive NSF grants. That difference has been pretty constant from year to year, Mixon says.

“NIH would do well to consult with their colleagues at NSF to learn how they are providing greater opportunities for underrepresented minorities. The answer may lie in NSF’s merit review process,” says Glenn S. Ruskin, director of the Office of Public Affairs at the American Chemical Society. “As NIH searches for solutions,” it should consider looking at successes both inside and outside of government, he says.

Some members of Congress are urging NIH to rapidly evaluate its peer review process and implement steps to ensure equal opportunities for all qualified researchers. “NIH, which receives $30 billion annually in federal taxpayer dollars, has a responsibility to ensure that its grant review process is transparent and equitable, and that its research workforce is diverse,” Sen. Benjamin L. Cardin (D-Md.), a strong advocate for NIH funding and member of the Senate Budget Committee, said in a statement.

NIH’s Collins is optimistic that the agency and the biomedical research community will come together to solve the problem. “Because if we fail to do so,” he says, “we can be confident we have lost out on the opportunity to bring some of the best and brightest minds to help us move forward with medical research at a time of great opportunity.”