06/27/2014 01:58 pm ET Updated Aug 27, 2014

Stemming the Leaks and Backflow in the STEM Diversity Pipeline

John Wildgoose via Getty Images

In late spring, around the same time President Obama hosted a science fair to promote middle and high schoolers' interest in science, technology engineering and math (STEM) careers, the tech firms Google and Yahoo! released corporate reports displaying the lack of ethnic and gender diversity in their ranks. These earnest gestures of STEM advocacy and transparency poignantly illustrated the problem with STEM diversity in the United States today: We stuff as many students as possible into the proverbial STEM pipeline while failing to address the leaks and backflows within, even when many students -- especially minorities -- don't emerge from the other end with STEM PhDs and jobs.

Built into this pipeline metaphor is the assumption that the pipeline lays flat and that the momentum of every individual going through is unchanged. The data tell a different story.

At the same time as the science fair and the industry diversity reports, we presented a performance analysis of the STEM pipeline over the last decade in a paper published in the journal BioScience. We showed that the pipeline inclines upward rather than lays flat, given how many students struggle and how few get through. The problem remains particularly acute among traditionally underrepresented minorities (URMs).

In 2000, when students entered the pipeline in college, whether they were URMs or not, a bit more than a third of them expressed intentions to study STEM. But URM students proved less likely to graduate as STEM majors than non-URM students. The data show that while 24.1 percent of U.S. college freshmen came from URM groups in 2000, only 18.5 percent of bachelor's degree recipients did in 2004.

The STEM-specific attrition of URMs became most evident after college graduation, according to National Science Foundation statistics. After college in 2009, 36 percent of the URM students holding STEM bachelor's degrees left the field instead of taking a STEM job or entering graduate study compared to 30 percent of non-URM graduates. Among those who did enter graduate school, many URM students struggled further up the pipeline. They earned only 18.3 percent of the STEM bachelor's degrees in 2004 and only 12.1 percent of the STEM doctorates awarded in 2010.

Reinforcing the prevailing reality seen at Google and Yahoo! that the STEM workforce lacks diversity, our assessment showed that URMs composed only about 10 percent of the STEM workforce in 2010.

These results tell us that educators and policymakers need to counteract the downward forces that prevent better advancement through the pipeline rather than simply stuff the pipeline with students. In BioScience, we proposed four ideas to help. These ideas are derived from our own experiences as minorities working in academic biology research, from our careful review of the education literature, and from results of our own NIH-funded experiments promoting STEM diversity at Brown University.

Our ideas don't call for infusion of "new money." Instead, we believe changes should be made that address endemic factors of academic life. We need better alignment of culture and climate, partnerships between research and minority-serving universities, a critical mass of underrepresented groups in STEM programs, and more faculty (rather than just administrative) engagement in diversity. These strategies can apply in the U.S and abroad to many underrepresented groups in STEM and non-STEM fields and in the future when the make-up of underrepresented groups may change.

Universities are large institutions with many opportunities for misalignments between the culture of articulated values, such as the importance of diversity, and the climate in which things actually happen (or don't). When these disconnects emerge with respect to diversity, they disillusion minority students. Such disconnects need to be identified and rectified. We therefore recommend annual, confidential surveys -- by third-parties rather than by university administrators -- to enable unbiased discovery of when an institution's diversity values aren't put into practice.

Those working in STEM diversity have encountered many cases of young graduate students who demonstrate great potential at their undergraduate-focused minority-serving institutions (MSIs) but who struggle at a graduate research institution. At Brown, we've addressed this challenge in part by forging partnerships with several MSIs to build relationships that go beyond merely recruiting students to increase diversity. We work with our partners to identify and fill curricular gaps so that students enter our graduate programs prepared. These partnerships, funded with support from the National Institute of General Medical Sciences, have other added benefits. They get MSI students and faculty involved in more advanced research at the undergraduate level and provide opportunities for research-institution faculty to develop "cultural competency" in working with students and faculty at MSIs.

Partnerships also allow institutions to recruit not just a bulk of broadly categorized URM students, but a critical mass of students with shared backgrounds. Critical mass is important because it ensures that students have an environment that fosters a feeling of inclusion. It also provides a support system where their shared experiences further validate their presence and contribution. This is something we all inherently rely on and universities must recognize that in order to maximize URM student persistence and retention, they need to increase the critical mass of peers. Focused recruiting can help ensure such communities.

An often underappreciated reality of academia is that the administrative lifespans of presidents and provosts are too short to be effective in instituting transformative changes in institutional diversity practices without significant faculty involvement. Directly instilling the value of STEM diversity among faculty will make diversity programs more effective. Universities should consider explicit incentives for diversity that are 'on par' with research and teaching in faculty merit and promotion considerations.

Challenges of culture and practice are among the most difficult to resolve. Given the financial investments made to date, it is clear that more money alone will not address the challenge of achieving STEM diversity. More is required. As surely as STEM research and education are valuable, they are even more so when we all have genuine opportunities to participate in them.

Andrew G. Campbell is Associate Professor of Medical Science in the Department of Molecular Microbiology & Immunology of the Warren Alpert Medical School at Brown University in Providence, Rhode Island. Stacy-ann Allen-Ramdial holds a PhD and Masters in Molecular Virology from Brown University.