We appreciate the opportunity to discuss the value of science, the topic of this special section. At present, both citizens and scientists are challenging historic barriers that limit equal access to participation in STEM fields. This has led to the development of new methods for assessing equality, and it offers opportunities to consider how to identify and address barriers, increase participation, and more effectively advance scientific research.
We were therefore intrigued by the “data mosaic framework,” developed in part to realize the potential of linked data to “ensure” equity (Chang et al., 2022). In this discussion, we examine the extent to which data linked under the data mosaic framework can be used to assess equality as directed by U.S. Executive Order No. 13985, Section 4 (2021): “whether agency policies and actions create or exacerbate barriers to full and equal participation . . . . with respect to race, ethnicity, religion, income, geography, gender identity, sexual orientation, and disability.” Specifically, we consider how federal funding of doctoral students might contribute to the gender gap in STEM fields, as funding is necessary for prospective candidates to enter programs and earn doctorates. Our examination links data from the Survey of Earned Doctorates (National Center for Science and Engineering Statistics, 2011), OECD.Stat (Organisation for Economic Co-operation and Development [OECD], 2006), and UMETRICS (Chang et al., 2017; Chang et al. 2022).
We find that data produced under the framework do offer insight. Chang et al. (2017) report that funded men outnumber funded women across all fields: STEM fields admit and fund more men than women overall (e.g., in engineering, nearly three men are funded for every one woman), and STEM fields receive more funding overall (cf. Tables 1 and 2). Chang et al. (2017, table 4) indicate that of women and men who earned doctorates, funding occurred at a similar rate within each field (n.b., the rate is used here to measure funding equity after students exit programs rather than when they apply or enter). But funding decisions are typically made before a candidate is admitted, enrolls, and completes a program.
Field | Male | Female |
Agriculture | 2% | 2% |
---|---|---|
Biology | 15% | 19% |
Computer Science | 5% | 2% |
Engineering | 24% | 8% |
Health | 3% | 6% |
Math & Statistics | 4% | 2% |
Physical Sciences | 14% | 7% |
Psychology | 4% | 11% |
Social Sciences | 9% | 10% |
Other | 20% | 32% |
Total | 100% | 100% |
Note. This table shows the distribution of U.S. doctorates earned in 2011 by field for each gender, from the Survey of Earned Doctorates (National Center for Science and Engineering Statistics, 2011). Note that percentages may not sum to 100% due to rounding.
Field | Federally Funded | Not Federally Funded |
Agriculture | 3% | 4% |
---|---|---|
Biology | 24% | 9% |
Computer Sciences | 4% | 2% |
Engineering | 28% | 16% |
Health | 4% | 5% |
Math & Statistics | 5% | 3% |
Physical Sciences | 21% | 6% |
Psychology | 3% | 6% |
Social Sciences | 4% | 14% |
Other | 5% | 36% |
Total | 100% | 100% |
Note. This table shows the distribution of doctorates earned in 2011 by field for those receiving and not receiving federal funds (as represented by seven major U.S. research universities). Adapted from Chang et al. (2017, table 2). Note that percentages may not sum to 100% due to rounding.
We wondered how funding disparity might create or exacerbate barriers facing women in STEM prior to the point at which women and men earn doctorates. Funding disparities could perpetuate inequality by influencing which fields female students pursue and whether female- or male-dominated fields gain prominence. Federal funding decisions do not occur in isolation; they result from a series of potentially gatekeeping policies and actions (e.g., selection committee membership, award criteria, and/or access to networks of support); and such gatekeeping practices occur at all entrance and exit points on the pathway from education to STEM occupations.
We conduct a thought experiment using the enrollment of undergraduate students in 26 OECD countries by field. In our experiment, we ‘fund’ these undergraduates according to the rates U.S. doctoral students receive federal funding as reported by Chang et al. (2017), and we measure disparity using the relative rate—the percentage of funded students who identify as female divided by the percentage of not-funded students who identify as female. For example, when U.S. undergraduates in 2006 are funded like U.S. doctoral students, this yields a disparity of .85. The results of our thought experiment are shown in Figure 1. The x- and y-axes denote the hypothetical percent female among the ‘not funded’ and ‘funded’ respectively. The size of each point corresponds to the number of students enrolled. The slope of the best fit line (green) estimates the disparity across countries to be .8 in 2006. (The dotted line reflects no disparity, a slope of 1.) The red ‘x’ denotes the funding disparity among U.S. doctoral students in 2011 calculated from Chang et al. (2017) (.75), which fits along with the other points, suggesting it is of the same size. Repeating the thought experiment for 2000 (red points) and 2012 (blue points) produces nearly identical results to 2006 (green).
In summary, we find the gender gap in federal funding reflects a preexisting disparity that is large and stable across OECD countries and periods (as of 2012). Our interpretation is that funding perpetuates persistent and widespread barriers for women in STEM. We conclude that in order to meaningfully meet the directive of Executive Order No. 13985 (2021) in assessing equity, researchers must create robust data sets that enable longitudinal and multidimensional analysis—a goal of the data mosaic framework (and Lane et al., 2022, and Smith, 2022, more generally). This would allow for data disaggregation and intersectional analysis to account for nuanced and pronounced disparities (see, e.g., Interagency Working Group on Inclusion in STEM, 2021). To meet the Executive Order’s call to improve agency policies and actions for equity, a sustained investment and strategic use of resources is necessary. We believe such an investment has the potential to enhance the value of science and further the advancement of scientific research.
We thank Richard Robbins for helpful comments.
Jonathan Auerbach and Catherine Elizabeth DeLazzero have no financial or non-financial disclosures to share for this article.
Chang, W.-Y., Cheng, W., Lane, J., & Weinberg, B. (2017). Federal funding of doctoral recipients: Results from new linked survey and transaction data. Report No. 23019. National Bureau of Economic Research.
Chang, W.-Y., Garner, M., Basner, J., Weinberg, B., & Owen-Smith, J. (2022). A linked data mosaic for policy-relevant research on science and innovation: Value, transparency, rigor, and community. Harvard Data Science Review, 4(2). https://doi.org/10.1162/99608f92.1e23fb3f
Exec. Order No. 13985, 86 Fed. Reg. 7009 (2021). https://www.federalregister.gov/documents/2021/01/25/2021-01753/advancing-racial-equity-and-support-for-underserved-communities-through-the-federal-government
Interagency Working Group on Inclusion in STEM, Federal Coordination in STEM Education Subcommittee, Committee on STEM Education. (2021). Best practices for diversity and inclusion in STEM education and research: A guide by and for federal agencies. National Science and Technology Council. https://www.whitehouse.gov/wp-content/uploads/2021/09/091621-Best-Practices-for-Diversity-Inclusion-in-STEM.pdf
Lane, J., Gimeno, E., Levitskaya, E., Zhang, Z., & Zigoni, A. (2022). Data inventories for the modern age? Using data science to open government data. Harvard Data Science Review, 4(2). https://doi.org/10.1162/99608f92.8a3f2336
National Center for Science and Engineering Statistics. (2011). Survey of Earned Doctorates. https://ncsesdata.nsf.gov/builder/sed
Organisation for Economic Co-operation and Development. (2006). Education at a glance 2006: OECD indicators. OECD Publishing. https://doi.org/10.1787/eag-2006-en
Smith, T. (2022). Demonstrating the value of government investments in science: Why anecdotes alone are not enough. Harvard Data Science Review, 4(2). https://doi.org/10.1162/99608f92.d219b2ce
©2022 Jonathan Auerbach and Catherine Elizabeth DeLazzero. This article is licensed under a Creative Commons Attribution (CC BY 4.0) International license, except where otherwise indicated with respect to particular material included in the article.