This new paper, “Policy Implications of Aging in the NIH-Funded Workforce,” has people talking again about the aging biomedical workforce. The authors’ aim was to investigate how this “graying” population is affecting funding of science in the U.S.
A key finding was that older scientists do not have a greater individual advantage than younger investigators, but that there are just more of them and so older scientists, as a group, receive more funding than younger investigators.
However in some of the reporting of the story, such as, “Why Don’t Young Scientists Get More Grants? Often They Don’t Apply,” there seemed to be an implication that young researchers choose not to apply, whereas in fact it is likely the case that there are fewer young investigators than in the past. To find out more, I spoke to the lead author of the paper, Misty Heggeness, previously a labor economist at the Division of the Biomedical Research Workforce at NIH, and now with the U.S. Census Bureau.
What question(s) were you looking to answer with this research?
I am an economist, but I trained with demographers. When I came to NIH, there was a general anecdotal opinion in the community that the average age of NIH R01-Equivalent Principal Investigators was increasing because those with less experience (usually younger) were getting funded less because funding was harder to attain. The question we were looking to answer is whether or not older principal investigators had higher funding rates than their younger peers.
There seems to be some confusion about the term ‘workforce’ in the paper, and in the Supplementary Materials you refer to the Retirement Age Replacement Ratio as the ‘PI replacement index’. Can you clarify who exactly is being talked about in this ratio, and what the ratio indicates? We aren’t talking about grads and postdocs at all, right?
In this paper, we focus on the established independent researcher workforce, which we define as NIH R01-Equivalent Principal Investigators. We recognize that principal investigators are funded through broader Research Program Grants (RPG), K (mentored-career) awards, F (fellowship) awards, and T (training) awards. While the entire biomedical workforce funded on NIH grants is larger than just PIs of R01-Equivalent Grants, the goal of many scientists is to achieve independence. This is why we were particularly interested in examining this group. The NIH Retirement Age Replacement Ratio focuses on how many independent investigators of the younger age group will be needed to replace the older age group if they were all to retire tomorrow.
So fewer younger investigators, could this be due to the increase in training times – and the oft-cited average age of investigators receiving their first R01 being 42? Do you know if there is any shift in the number of trainees applying for trainee fellowships over time – or are people leaving the system?
I have not looked into whether there is a shift in the number of trainees applying for fellowships overtime, but through other research I am working on we have found an increase in applicants (although not in awardees). This is an excellent question and would be a great follow-up line of research. I have looked in the K (mentored career) awardees and applicants. I can tell you that they do not make up the numbers of younger PIs who are in recent years not applying for R01-Equivalent grants. In other words, people are either lingering longer in postdocs in hopes of finding an academic tenure-track position or they are leaving academia for other employment options.
There was an interesting statement, “Women under age 40, however, have increased their applications since 1980 and the number awarded has remained relatively stable.” This sounds optimistic – what does this trend suggest about gender differences in the PI workforce?
Well, it’s no surprise that the number of women in academic medicine research has been increasing over the decades. This could imply that the dynamics of research for women is changing or that younger women are now finding their way within the competitive research grant application process or something else entirely. This would be another line of very interesting research to conduct in the near future.
In the section, “Factors outside the Direct Control of Funding Agencies,” it seems to come to a major point – that it’s not easier to get grants when you’re older, but that there are more older PIs to apply for them, and that in fact it’s to the institutions we should look to explain the aging PI workforce. Is that the case and can you elaborate?
I don’t think the answer lies entirely in looking to institutions. I do not have a long history with NIH nor am I a biomedical scientist. As a labor economist who is fascinated with this workforce, I come at this from an outsider’s perspective. It seems to me that often the community looks to NIH for leadership and guidance, but also looks to NIH as the root problem. I don’t think either is true. And, in this case, it seems to me that investigators and institutions also play a significant role in the challenging dynamics we now face. NIH does not dictate how investigators should fund their science (e.g. how many staff scientists, postdocs, or graduate students to have on a grant). I think institutions, investigators, and NIH need to work in harmony to self-examine how each plays a role and what each can or should do differently to have a greater impact on creating a sustainable environment and workforce.
Certain names in the authorship have raised concern that there is some underlying policy intention for this paper, are there any intentions for this work?
No. As I mentioned earlier, I was interested in this question because as an economic demographer and labor economist, I was intrigued by what I was hearing in the community about the aging workforce. I felt someone needed to bring the data to bear. Sally Rockey and others at NIH had already conducted work in this area and were open to and interested in developing further analysis. However, as an economist in the federal government, I am not allowed to dictate, create, or develop policy. Additionally, any data-driven research is the opinions of the authors and not necessarily the opinions or beliefs of the agency. Our only intention with this paper is to generate data and metrics that allow leadership and policy makers to make informed, data-driven decisions. We are lucky that NIH leadership have that focus as well.
Nevertheless, is there a possibility this metric could be misconstrued to try to justify a need for MORE trainees? How can we avoid that from happening?
First, I think it is important to understand what your definition of “trainee” is. If you mean NRSA funded trainees, you will have to ask NIH if that is a plan they have for the future. I tread with caution here because by increasing R01 funding or other types of grant funding at NIH, we also indirectly increase “trainee” funding, since most PIs will increase labor slots on their grants if given additional money (or if more grants are given). This is where the vicious cycle lies in terms of training.
R01s are a major source of grant support for hiring trainees. What are the likely implications for the biomedical workforce of a growing group of aging investigators continually hiring trainees? And does this say anything about how reliant the system should be on R01 funding mechanisms?
I can tell you from my time at NIH that this is an issue NIH takes to heart. They understand the complexities of how grants are structured and want to make our biomedical research community the best it can be and develop a system that is supportive, generates great science, and makes the workforce viable and scientific research jobs desirable. I know some institutions, like NIGMS and NCI, are working hard on developing pilot grants that alter the structure of our current system. I don’t have a straightforward answer to this question. All I can say is that it is so essentially important that it requires great thought and foresight.
Any final take-home messages?
We wrote this paper because we wanted leadership and policy makers to have concrete data and metrics with which to base policy decisions. Specifically, we were interested in understanding whether the aging PI workforce was a product of more competitive funding or something else. We don’t find evidence that it is because of greater competition. In fact, it’s very possible (although our paper does not test this) that the Early Stage Investigator (ESI) policies and other NIH policies developed to help young investigators are actually working and working well. Our results are important because it demands a different policy response than if competition is driving a lack of awards to young investigators. Our NIH Retirement Replacement Ratio and the Age Pyramids are two metrics that can be replicated as we move forward and give us a deeper understanding of the underlying dynamics of the NIH-funded PI workforce. I hope the data is useful and that NIH will continue to support the development of these metrics when examining issues related to the biomedical workforce.