A look into physics programs at liberal arts institutions: Part 2

As I previously shared here, the announced changes to a physics program in a Christian liberal arts institution have inspired me to look more closely into different metrics of physics programs. In my first post, I shared metrics of Coalition of Christian Colleges and Universities (CCCU) schools, focusing on number of bachelor's degrees in physics granted in 2015-2017 compared to full time equivalent (FTE) numbers. The goal was to provide concrete information about how many students graduate from CCCU institutions with bachelor’s degrees in physics, in terms of both absolute number of graduates and as proportion of broader educational activity at the institutions.

In this post, I will expand the analysis to include more groups of institutions, as well as consider several financial metrics. My goal is to understand the numbers of physics majors graduating from programs from a broader scope of institutions and in the context of financial viability of them. I again use the AIP and IPEDS data referenced in the first post.

The first step in expanding the scope was to identify institutions which compare to CCCU schools in some way. I identified five comparison groups:

(1) CCCU peer: institutions which have a similar identity and mission as CCCU schools (and may in fact have previously been members of the coalition),

(2) Chicagoland private institutions,

(3) New England private institutions,

(4) PhD institutions, and

(5) other private liberal arts institutions.

For categories 2 and 3, I was interested in potential regional effects in making comparisons, as Wheaton is in Chicagoland and Gordon is in New England. For category 4, I chose two private PhD institutions, Brigham Young University and Vanderbilt University. I selected these because Brigham Young has an explicit faith-based mission and also a strong physics department, and I am very familiar with Vanderbilt from my time as a graduate student there. In category 5, I chose a collection of other liberal arts institutions from their familiarity to me and my knowledge of them as strong liberal arts institutions, often based upon my connections to faculty there who I know are engaged in physics education research or who are committed to making use of best practices identified by the research done by others. As in my first post, I show here data only for institutions who reported the number of bachelor's degrees in physics they granted in 2015-2017, in all three years, even if that number was zero. Institutions which did not report the number of bachelor’s degrees granted in any of these three years were excluded.

Figure 1: Total number of physics bachelor's degrees granted in 2015-2017.

The blue bars in Figure 1 are the same as Figure 2 in the previous post. I have added the numbers for the other groups, and have also drawn a horizontal line to indicate the number for Wheaton College (34), which is not too much larger than that for Gordon (30). Thirty-four out of these 47 institutions granted fewer bachelor’s degrees in physics than did Wheaton, but that is not much different than for Gordon College: 33 out of 47 institutions granted fewer bachelor’s degrees in physics than did Gordon. My interpretation of this figure is that both Wheaton College and Gordon College granted a strong number of bachelor's degrees in physics compared to these comparison institutions. It is not surprising that the PhD granting institutions in this set granted more bachelor’s degrees in physics than many of the liberal arts institutions, but it is notable that the total number of physics degrees granted by Vanderbilt University (an R1 institution), 46, is not all that much than those granted by Wheaton and by Gordon, and five other liberal arts institutions granted more degrees in physics than did Vanderbilt, during this three-year period. With this figure, I hope to establish for those interested the actual number of physics graduates for a range of liberal arts institutions. The terms “small” or “large” need to be discussed within context of actual numbers.

The next subject of investigation I considered for this broad scope of institutions is the relationship of revenue from graduates of bachelor’s degrees in physics compared to instructional expenses for these graduates. A more nuanced investigation would consider how many full-time faculty there are in each department, but this information does not appear to be available from AIP. AIP does report that most institutions granting bachelor’s degrees in physics as their highest degree offered have fewer than 10 faculty in the department (83% of all institutions in 2016, link here) and that the average number of faculty in all these departments in 2016 was 6.3 (link here). These numbers do not give information about support staff or contingent faculty coverage, and clearly personnel costs are a significant issue when considering cost of maintaining any academic program. However, even with these limitations, we can use the categories of “revenues from tuition and fees per full-time equivalent student” (which I will abbreviate as “revenue”) and “instruction expenses per full-time equivalent student” (which I will abbreviate as “expense”) to look at the balance of revenue and expenses for the physics graduates each year. To create Figure 2, I multiplied the revenues and expenses data (which is reported per full-time equivalent student) by the number of physics graduates. Note that a limitation of this is that students enrolled in the major but not graduating in each academic year are not accounted for. Another limitation is that this makes use of an average expense for each full-time equivalent student, a factor which would surely vary between students of different majors. Finally, this figure assumes that all graduates were full-time equivalent, which may not be the case.

Figure 2: Revenues from tuition and fees for students graduating with a bachelor’s degree in physics versus their associated instruction expenses, by academic year ended as indicated. “WC” stands for Wheaton College..

In Figure 2, the diagonal line of unity gives a visual cue about the ratio of revenue to expenses: if the data marker for an institution is above the line, then the revenue was greater than the expenses. If the data marker is below the line, then the expenses were greater than the revenue. At first glance, this figure may seem to give information about financial viability of an institution -- after all, income should be greater than expenses -- but the information must be considered in the context of endowment, which provides additional funds for each student’s education.

Figure 3: Endowment per full-time-equivalent student at end of given academic year versus ratio of revenue from tuition and fees to instructional expenses. The shading for each data point is normalized to the total number of bachelor’s in physics granted that year.*

Figure 3 makes use of IPEDS data on endowment at year end per FTE versus the ratio of revenue from tuition and fees to instructional expenses. Values of endowment is a bit more obvious to interpret -- the more endowment per student, the more passive income that can contribute to the educational enterprise -- although the individual institutions may have different target spending percentages. The ratio of revenue to expenses can give a sense of dependence of an institution upon tuition and fees: the higher the the ratio, the more dependent the institution may be upon them. This may explain the downward trend of the figure for these values. If you can imagine a vertical line at ratio of revenue to expenses of 1, the data appear to show a trend that most institutions for whom this ratio is less than 1 have higher endowments. Institutions that nominally have better financial health are in the upper left-hand quadrant of the figure.

Figure 3 includes an additional axis of data: the values for each data point are shaded with respect to each institution’s number of bachelor’s degrees in physics granted that year.

Up until this figure, it seemed to me that most CCCU institutions tracked reasonably closely to each other and to the other institutions in the comparison groups. Figure 3 displays a very rough correlation of financial health of an institution with number of graduates in physics. However, the correlation is weak and it cannot be definitively said that a higher number of graduates is associated with stronger financial situation for an institution. Likewise, it is not possible to say that fewer number of graduates is associated with a weaker financial position. But it also highlights some institutions:

(1) institutions with a relatively low number of bachelor’s degrees granted in physics and less dependence upon tuition, and thus may be sustaining these lower numbers due to endowment (lightly shaded data points in the upper left-hand quadrant),

(2) institutions with higher dependence on tuition and lesser support from endowment, for which fluctuations in physics enrollment may be problematic (lightly shaded points in the lower right-hand quadrant), and

(3) institutions with a relatively high number of bachelor’s degrees granted in physics which are more dependent upon tuition (darker shaded data points in the lower right-hand quadrant).

The latter group of institutions are producing a relatively higher number of physics graduates with potentially more dependence upon tuition and it may be helpful to better understand the practices of these departments.

These figures as a whole don’t explain why some institutions may evaluate the role of physics programs within them and consider them for closure or revision. As discussed above, there are a number of factors not available in this analysis that are needed for complete nuance. However, the figures may give a sense of relative position for the revenue and expenses for physics graduates with respect to endowment at these institutions, and offers concrete information for the number of graduates from these programs.

Having established what the current numbers are for numbers of students graduating from liberal arts physics programs and the context of real-dollar revenue and expenses of instruction for the graduates, along with institutional endowment, the next topic of exploration on the question of the viability of physics departments at liberal arts institutions is what expenses may be involved in the running of them. That will be the focus of the next post in this series. I will endeavor to look further into the institutions in the lower right hand quadrant of Figure 3 with higher number of bachelor’s degrees granted in physics. I am also interested in hearing from other physics professors: what are best practices for determining and controlling expenses of a liberal arts physics department? How do you determine investment in introductory courses, which can serve both the physics major and majors in other departments, compared to what is needed for upper-level courses, in terms of both physical and personnel resources? I’d love to hear from you. Feel free to comment below or send me a note.

*The normalization was done with respect to the second-highest number, since the number of graduates at Brigham Young was so much greater than all other institutions in the group, and the normalization for BYU was additionally set to one. Thus, in each figure, the shading of the data marker for two institutions is always one: BYU and the second-highest institution.