We are not doing so well

Editor’s note: This was updated on April 26. The original had a few errors. Thank you to Tom Barr at the AMS for help getting the correct numbers.

Ok, so that title doesn’t tell you much.

Two reports came out recently, both based on data collected by the National Center for Science and Engineering Statistics (NCSES). Both interesting, one downright troubling.

This post falls into the category of “I think you should know this,” and I ask you to think about what you can do about the implications of the reports.

One is referred to as an “InfoBrief” and titled “Higher Education R&D Spending: Spending and Funding Sources Differ by State.” The other is a biennial report on “Women, Minorities, and Persons with Disabilities in Science and Engineering.”

The InfoBrief

The InfoBrief shows–for each state–where funding originates for research and development (R&D) at colleges and universities. Data are from the Higher Education R&D Survey, from the National Center for Science and Engineering Statistics within the National Science Foundation (NSF).

Academic institutions depend on a variety of funding sources — including the federal government, state and local governments, businesses, nonprofit organizations, and institutions’ own funding. Table 1 shows the total contributions from these sources, and that there is great variety from state to state.

Table1

Sixty-six percent of funding from nonprofits supported R&D in the life sciences. State and local funds have largely been used to support life sciences (61%) and engineering (17%). Businesses funds have also largely been used to support life sciences (61%) and engineering (25%). And, institutional funding largely supports the broad fields of life sciences (54%) and engineering (13%).

In 2016, the federal government provided the majority of funding in life and physical sciences, engineering, computer sciences, geosciences, mathematics, and psychology.

Where does this (from all sources) investment go, by field?

Table2

Note that further funds went to multi-disciplinary work.

The Report

The report provides statistical information about the participation of the three groups of the title in science and engineering education and employment. NSF reporting on this topic is mandated by the Science and Engineering Equal Opportunities Act (Public Law 96-516). It includes data on the participation of women and minorities that are not favorable for the mathematical sciences.

Recently I was perusing the Annual Survey of the Mathematical Sciences. I was specifically looking to find out what percentage of tenured faculty–at a particular research university–I might expect to be female. Among tenured faculty in PhD-granting mathematics departments, about 14% were women; among tenured faculty in master’s and bachelor’s-granting math departments about 27% were women; among tenured faculty in stats/biostats, the percentage was about 24%. The school I was looking for is a university classified as “Math Private Small”; at this group of 28 departments just under 14% of the full-time tenured faculty are women. This is for 2016. Ugh.

In the back of my mind is the idea that we are doing worse in academia than in other employment sectors, and worse in mathematics than in other fields. Wishful thinking on my part; this report shows otherwise.

Figure 1

Of all science and engineering (S&E) degrees awarded in 2016, women earned about half of bachelor’s degrees, 44% of master’s degrees, and 41% of doctorate degrees, about the same as in 2006. Figure 1 shows that while more women are getting PhDs in mathematics and statistics than were twenty years ago, we are actually doing worse than we were a decade ago. At the bachelor’s degree level, it has been a slow decline.

Even for women who are “making it” in S&E professions, we see salary discrepancies. Among scientists and engineers working full time in 2017, women generally made less than men did. Overall, women’s median annual salary was \$66,000, whereas the median salary for men was \$90,000. In the mathematical sciences, it is \$70,000 and \$81,000, rosier but not exactly a fact of which we should be proud. According to the US Census Bureau, women’s median earnings as a percentage of men’s median earnings was 80.7% in 2017, which is worse than in the mathematical sciences.

What about with minorities? In 2016, underrepresented minority students received 22% of all S&E bachelor’s degrees and 9% of all S&E doctorate degrees. The share of S&E bachelor’s degrees awarded to Hispanics or Latinos has increased over the past 20 years. Over the past two decades, African Americans have seen increased shares of bachelor’s degrees in psychology, social sciences, and biological sciences, but their degree shares have declined slightly in the other fields. Important to us, the share of bachelor’s degrees in mathematics and statistics earned by African Americans declined from 7% to 4%. It is worth noting that the top US baccalaureate institutions of S&E recipients from various groups (2013-2017) are, in this order, Howard University, Spelman College, and Florida A&M University.

According to the Annual Survey, 39 African American men and 14 African American women received PhDs in the mathematical sciences during the period July 1, 2015-June 30, 2016 (the most recent year data are available). This means that African Americans earned roughly 2.7% of math PhDs during that time.

The primary purpose of the report is to serve as a statistical abstract with no endorsement of or recommendations about policies or programs. So what is to be done?

In an ideal world, the demographic of the mathematics community would be closer to that of the US as a whole. What it takes to work toward this will be institution dependent. I urge you to figure out what can be done at your own institution to move the needle, and work with your colleagues to take action. Even small achievements are successes. Awareness is a first step.