Selected Data from the 2021 CBMS Survey, Part I
By David Bressoud @dbressoud
David Bressoud is DeWitt Wallace Professor Emeritus at Macalester College and former Director of the Conference Board of the Mathematical Sciences. davidbressoud.org
Starting in January 2024, Launchings will begin a new practice of coming out on the third Tuesday of the month.
The CBMS surveys of undergraduate programs in the mathematical sciences in the United States have run every five years, starting in 1965. Because of COVID, the 2020 survey was postponed until 2021. For a variety of reasons, the process of obtaining, cleaning, and organizing these data has been especially fraught. The final data tables are just now becoming available, thanks to the hard work of Ellen Kirkman, Rikki Blair, and the team at AMS led by Sarah Bryant.
The full tables should become available in 2024. In the meantime, I have chosen to cherry-pick five examples where the changes from 2015 have been particularly dramatic. It is not clear whether these shifts are a reflection of temporary disruptions caused by COVID or whether they reflect fundamental changes. We expect to be back on schedule with the 2025 survey that will therefore be extremely important in clarifying how many of these observations reflect long-term trends.
These data only relate to four-year undergraduate programs. A selection of data from two-year colleges will be the topic of my January Launchings.
The first graphic looks at total enrollments in the mathematical sciences from 2005 to 2021 (Figure 1). All numbers reflect thousands of students. In Fall 2021 these enrollments were down 30% from 2015, but it is the years 2010 and 2015 that appear to be an aberration. The year 2021 saw total enrollments fall to the level they had been from 1990 through 2005. The most dramatic change was the drop in the number of students at the precollege level as most of those courses have been off-loaded to two-year colleges. Note that Introductory Level includes College Algebra, Trigonometry, Precalculus, and the various credit-bearing but generally terminal mathematics courses (excluding introductory statistics) below the level of calculus. Calculus Level includes Single and Multivariable Calculus (both mainstream and business) as well as Linear Algebra, Differential Equations, and Discrete Matheamtics. Upper Stats Level are statistics courses that normally require a college-level statistics prerequisite. Advanced Level are mathematics courses that include Intro to Proof and otherwise are generally offered to juniors and seniors.
Figure 1: Fall semester course enrollments (thousands of students) by level of courses in 4-year undergraduate programs in departments of mathematics and statistics.
It is not surprising that distance learning exploded in Fall 2021 from 1% to 10% of students taking calculus level classes (Figure 2). It will be interesting to see the extent to which this shift to distance learning persists.
Figure 2: Students enrolled via distance learning (instructor and students in different locations for at least half of the class sessions) for Fall term. Calculus includes both mainstream and business calculus.
The number of full-time faculty in departments of mathematics did continue to expand from 2015 to 2021, but this was almost entirely among Other Fulltime faculty. For women, there was improvement in the ranks of tenured faculty, moving up from 22.4% to 25.4%. The percentage of women in tenure eligible positions dropped slightly from 35.6% to 34.7% (Figure 3).
Figure 3: The number of full-time faculty and post-docs in departments of mathematics in the Fall term.
If we restrict our attention to departments of mathematics offering a PhD (Figure 4), the patterns are similar. Most of the growth was among Other Fulltime faculty. For women, the patterns were comparable to what happened across all departments of mathematics, but at a much lower level. Women in tenured positions grew from 13.9% to 16.4% while the percentage in tenure eligible positions was essentially constant: 26.1% to 26.5%.
Figure 4: Number of full-time faculty and post-docs in doctoral-level departments of mathematics in the Fall term.
There was strong faculty growth in doctoral-level departments of statistics. Tenured faculty grew by 28.8%, tenure eligible faculty by 24.5%, and Other Fulltime faculty almost doubled. For women, the patterns were comparable to what happened across all departments of mathematics, but at a much lower level. Women in tenured positions grew from 21.1% to 25.1% while the percentage in tenure eligible positions dropped from 32.3% to 29.6% (Figure 5).
Figure 5: Number of full-time faculty and post-docs in doctoral-level departments of statistics in Fall term.
Figure 6: What happened to post-docs from spring to fall of the given year. Stayed = remained in the post-doc at the same institution , TT = took a tenure-track position, Non-TT = took another post-doc or other non-tenure-track academic position, Non-acad = took a non-academic position, unknown = unemployed or fall position unknown.
Some of the most dramatic changes occurred among post-docs. From spring to fall 2015 almost two-thirds of post-docs in doctoral-level departments remained in their position for an additional year while only 21% did so in 2021. Just over half of those who were in a post-doc in spring 2021 moved to another post-doc or a non-tenure-track academic position for the following fall. There also was a sharp increase, from 5% to 16%, in the number of people leaving a post-doc in the spring for whom their employment status in the fall was unknown (Figure 6). There is one caveat for these data. They come from a long survey that AMS sent to doctoral departments of mathematics, and only 23% of departments answered this question. However, there is no reason to suspect any bias in which universities answered this question.