Pipeline. (James T M Towill/Geograph)

One of the most commonly used metaphors for describing the solution for growing and diversifying America’s scientific talent pool is the “STEM pipeline.” Major policy reports have called on the U.S. to enlarge it so it does not fall behind other nations. Scholars and the popular press have highlighted the need to fix pipeline “leaks” that result in the disproportionate losses of women and minorities. While this metaphor has been helpful in focusing attention on careers in science, I am increasingly convinced that it fails us because it limits our view of the problems and their solutions. Further, these failures are actually hindering efforts to enhance scientific diversity–that is, cultivating talent, and promoting the full inclusion of excellence across the social spectrum.

Limitations of the “Pipeline”
The “pipeline” refers to the educational pathway–from elementary school through college, graduate school or the postdoc–that students complete in the pursuit of a STEM career. There are (at least) two big limitations with the metaphor. First, it reinforces the notion of a strict, linear sequence for becoming a scientist where none exists. There are of course certain benchmarks and competencies that need to be reached for one to be a capable scientist. However, if science wants to benefit from the talents of people from all backgrounds, then diversity efforts must focus on making sure there are more pathways that allow capable, hard-working people to join and participate in the field.

Second, and maybe most importantly, the linear nature of a “pipeline” means that the only way to enhance scientific diversity is to increase the number of people from underrepresented backgrounds entering the system. That is, pipeline framing focuses attention on the number of scientists from underrepresented (UR) backgrounds, and takes focus away from whether the environments and systems in which they are educated and work are supportive and promote inclusion. A major presupposition of pipeline framing is that if more girls and women, minorities, or whatever UR groups were interested in science and progressed through the system, scientific workforce diversity challenges would be solved. While numbers are of course part of the issue, a study I recently published with my colleague Kimberly Griffin suggests that the reason for the lack of diversity is much more structural in nature.

Two roads diverged. (Paro_for_Peace/Flickr)

Disparate Career Trajectories Among PhDs
Professor Griffin and I have spent the past few years studying science PhD recipients. By definition, PhDs are committed to science–no one does that much schooling if they’re not. Moreover, having reached the end of the educational “pipeline,” a PhD recipient has navigated any potential barrier to access, retention, or persistence. Thus they provide an excellent group from which to test the idea that by increasing the number of trained people from UR backgrounds, we can enhance diversity.

In our work, published in PLOS ONE, we surveyed a large sample of PhDs in the biomedical sciences (my home discipline). We asked them about their career preferences over time, as well as factors known to be important in pursuing a scientific career–mentoring, self-confidence and graduate school experiences. We also included objective measures–for example, the number of scientific publications they had produced and the types of institutions where they were educated. If the “pipeline” framing was correct, then one would assume there would be no differences in career trajectories of these Ph.Ds. across lines of race/ethnicity or gender after accounting for any potential differences in these important factors. However, our results showed just the opposite.

When statistically accounting for any difference in these important factors, including objective measures, women and scientists from underrepresented minority (URM) backgrounds were 36-54 percent less likely than White or Asian men to express interest in a career as a faculty member in a research university upon the completion of graduate school. Further, URM women PhDs were twice as likely as scientists from all other groups to express high interest in a career outside of research.

Let that sink in.

Among science PhDs who are otherwise similar on important metrics such as publication record, mentoring support and self-confidence, we still see differences in the career pathways they show interest in pursuing. This, in my view, means that simply focusing on getting more people into and through the educational system will not be sufficient to solve science’s diversity problems. Instead, efforts must focus on creating a system that highly trained and talented scientists from all backgrounds want to be a part of.

Alternatives. (Daniel Oines/Flickr)

Toward Systemic Reform
To be clear, I support programs and initiatives that aim to increase the numbers of students generally, and from UR backgrounds specifically, entering scientific training. I have benefitted from many programs that support young scientists. These include the North Carolina School of Science and Mathematics, UMBC’s Meyerhoff Scholarship Program, the Leadership Alliance, the National Science Foundation’s Graduate Research Fellowship Program, and many more.

Having more scientifically trained people, no matter what career pathway they take, is in my view incredibly important. However, efforts to increase the numbers of women, minorities and other UR groups in the sciences should be coupled with reforms that make sure the institutions training them, and the funding agencies supporting scientific research, promote inclusion.

In addition to focusing on the number of individuals the system produces, policy efforts must also focus on making sure that all scientists have high quality experiences and are well supported throughout their education, training and career. My hypothesis is that if scientists from all backgrounds felt that they would be well supported in the scientific enterprise–particularly the universities where the bulk of federally-funded research is conducted–then we would start to see greater levels of diversity.

Diversity is a byproduct of a highly functioning system that supports scientists from all backgrounds. Hence, we need to go beyond “the pipeline” and begin to tackle the institutional and systemic structures that lead to the loss of talent from diverse backgrounds in the sciences. In subsequent posts, I’ll share more on reasons why I believe these differences exist, and how we might begin to tackle them.

* The views expressed here are the personal opinion of the author alone, and do not necessarily represent the positions of the institutions with which he is affiliated. To see more of our work, follow the links below:

“Biomedical Ph.D. Career Interest Patterns by Race/Ethnicity and Gender”

“What Do I Want to Be With My Ph.D.? The Roles of Personal Values and Structural Dynamics in Shaping the Career Interests of Recent Biomedical Science Ph.D. Graduates”