Twilight for tenure: even more data exploration

In my last post on casualization at Rutgers, written November 2021, I discussed statistics on the rise of full-time, non-tenure-track faculty, arguing that this was an increasingly significant yet under-discussed aspect of the broader erosion of the tenure track. I promised then that I’d follow up on some of the details about different categories of faculty and of institutions. Would my dire picture of “twilight for tenure” change if I separated non-medical from medical faculty, or if I paid attention to faculty with non-instructional roles? Well, I’m pleased to report the picture is dire no matter how you paint it. I’ve been looking at the more granular information on higher-education staffing found in the Human Resources data from the Department of Education (specifically, the “Employees by Assigned Position” or EAP data files from IPEDS). Here are some tentative explorations, vacillating between being tediously technical and speculatively broad-brush. Skip to the end for my regular “workers of the world” conclusion followed by faculty-casualization league tables for research universities.

The basic question is, what are the terms of employment for people doing academic work in higher education? The EAP data answer that question by classifying workers at each institution as tenured, tenure-track, non-tenure-track, or “without faculty status,” dividing each category into full-time and part-time categories and according to whether they are in medical schools or not. The EAP survey also subdivides academic workers’ duties into instruction, research, and public service—as well as further categories like librarianship, archiving, and “Student and Academic Affairs.” Graduate workers (“graduate assistants”) are treated as another employee category, assigned either to teaching or research (and medical or non-). That leaves us with many possible ways of cross-cutting or subsetting the data about tenured and contingent academic work. (Obviously such categories do not exhaust the interesting variables; for example, the EAP data does not include any demographic information about each category. These are found in other IPEDS components, which however do not subdivide job categories with the same granularity.) What I want to do here is explore—without being exhaustive about it—whether the divisions matter to our understanding of contingency in the academy. I’m hoping that even if my analysis is wanting, the annotated R code on github, as “tidy” as I can make it, might help others get a leg up on working with this data.

Discussions of contingent work in higher education usually focus on the teaching function. For example, the most recent AAUP report on the economic status of the profession restricts its analysis to non-medical instructional faculty. This restriction makes sense for many purposes. It focuses attention on who is doing the work of teaching in higher education—which, after all, is the fundamental purpose of the whole system—and it underlines the impact of the decline of tenure on the college classroom. Leaving out medical faculty is also reasonable in this context. Adjunct medical faculty are often well-compensated clinicians who neither need nor want full-time or tenured faculty status, though the case for protecting the academic freedom of medical faculty is probably pretty much the same as for any other specialist faculty. It is just that the picture is distorted by the ragingly unequal political economy of American health care.¹

Nationally, across all of American higher education, most teachers (71.9%) are not on the tenure track: they are part-time (PT) non-tenured instructors, full-time non-tenure-track (NTT) faculty, or graduate TAs.² The most numerous category of all is part-time instructional staff, who outnumber tenure and tenure-track (T/TT) faculty on their own.³ But these overall numbers conceal a great deal of variation. For example, tenure essentially does not exist in for-profit higher-education; at for-profits, faculty are overwhelmingly part-time (75.8%). The proportion of PT faculty in in non-profits is half that figure, but they are still the most numerous group. On the other hand, among medical faculty, full-time non-tenure-track instructors are much more common; in fact they are the majority. The proportions of tenure-track faculty are comparable between medical and non-medical faculties, but the composition of the contingent majority is rather different. Here are the breakdowns as of 2020–2021, where I have given percentages of all instructors in order to give a sense of the relative size of the groups in question.

Proportion of instructional staff across all title-IV postsecondary institutions. Numbers in parentheses are the proportions within the table row. Source: IPEDS 2020–2021.

instructional staff	grad	NTT	PT	T/TT
in for-profits	0.0% (0.3%)	1.0% (23.5%)	3.2% (75.8%)	0.0% (0.4%)
in non-profits: medical	0.1% (1.6%)	4.3% (50.8%)	1.5% (18.2%)	2.5% (29.4%)
in non-profits: non-medical	11.1% (12.8%)	14.4% (16.6%)	36.0% (41.3%)	25.6% (29.4%)

More about full-time NTTs

Now let us return to the rise of the full-time non-tenure-track faculty. The pictures in my previous post, it turns out, owe some of their drama to the inclusion of medical faculty. Among non-medical instructional faculty, NTTs have a smaller share of the total, but they remain the only staffing category which has been consistently growing in recent years, in relative and absolute terms.⁴ (Click on any of the figures for the full-size version.)

The late-breaking uptick in the proportion of T/TT faculty in 2020–2021 is an effect of a shrinking denominator, as part-time numbers were greatly reduced in the first pandemic year; the absolute number of T/TT faculty also decreased, but only slightly, from 2019–2020 to 2020–2021. One could not want a clearer demonstration of the difference true job security makes than the ease with which tens of thousands of adjuncts were shown the door in the midst of a global crisis.

Focusing on non-medical instructional faculty, not only are the majority contingent (NTT, part-time, or TAs), but contingent faculty are the majority within most higher-education institutions. However, it turns out there are plenty of institutions (about 15% overall) where there is a T/TT majority of instructional staff, especially among four-year colleges. But even in this category, institutions with at least 75% T/TT instructors are quite rare.

Degree-granting non-profit institutions, by class and % T/TT instructional staff. Source: IPEDS 2020–2021.

Type	< 25% T/TT	25-50%	50-75%	> 75% T/TT	Total
Associate’s/Two-Year	581	338	71	8	998
Baccalaureate Colleges	158	167	188	25	538
Doctoral Universities	59	293	41	3	396
Master’s Universities	203	303	97	8	611
Special Focus Four-Year	367	55	20	22	464
Tribal Colleges	33	1	0	0	34
Total	1401	1157	417	66	3041

Bear in mind, however, that there are almost as many instructional staff (including graduate TAs) at the “Doctoral Universities” as there are at all other institution types combined. Also not evident from the table is the astonishing number of institutions with no tenure track at all, who are here grouped in with those institutions with low but non-zero rates of T/TT instruction. In 2013–14 948 out of 2991 (31.7%) non-profit institutions had no T/TT faculty; in 2020–2021 the figure was 1039 out of 3065 (33.9%).⁵

What about researchers?

But teaching is not the whole story in higher education: what about research? In the EAP dataset, one finds large groups of faculty—and graduate assistants—classified under “research” alone. I noticed this when looking more carefully at Rutgers’s own staffing pattern; at Rutgers, about a third (810 out of 2672) of the full-time NTT faculty are reported as assigned to research and not to instruction; only a very few T/TT faculty have the same classification. On the flagship New Brunswick campus, if we count all faculty assigned to any combination of instruction or research or public service, there are nearly as many non-medical NTTs as there are non-medical T/TT faculty. This by no means the case nationally, even at research universities, but trends in different faculty categories are interesting to compare to the picture of instructional staff above.

With the research function in view, graduate assistants are a much larger group, but the full-time NTT category is still striking in its growth. It is not at all self-evident that the research GA is to T/TT faculty research what the TA is to T/TT faculty teaching. I imagine many of the research GAs are in large science labs at research universities, a very different situation from the humanities doctoral student doing an individual dissertation project while earning money as a TA. The same distinctions apply of course to the full-time NTT category, which (if I am not mistaken) includes the sciences’ large army of postdocs. But this dataset has no information at the discipline level.

The impact of contingent forms of employment on research in higher education seems to me much less discussed than the impact on teaching, even though academic freedom is just as significant to research as to teaching. Instruction is casualized because the majority of instructors are not T/TT, across the whole system and at most institutions. Research is less casualized only by comparison; T/TT faculty are in fact the plurality in the research category, but both the growth of NTTs and the enormous size of the graduate worker category mean that academic freedom in research is still not very well-protected by tenure. No less consequential, however, is the large proportion of instructional staff who are not also researchers. The ideal of higher education is that it combines instruction and research, the transmission and the creation of knowledge, in the same institutions and the same personnel. But the reality is that most instructional personnel are not also assigned research duties, and the more casualized, the less likely to be employed for research:

Proportion of all instructional staff (excluding graduate students) assigned to instruction primarily or instruction, research, and public service, across all non-profit Title IV institutions. Source: IPEDS 2020–2021.

role	T/TT	NTT	PT
Instruction/research/public service	13.2%	3.2%	4.6%
Primarily instruction	20.5%	15.8%	42.7%

Even if you are inclined to be sanguine about the moderately less casualized staffing pattern of the research function, you should be troubled by the way casualization breaks the connection between teaching and research.

Of course, justice would not be achieved by assigning all existing part-time adjuncts and NTTs to research as well as instruction. As things stand, it would be far better to expand tenure beyond the research function and protect academic freedom with tenure systems for all faculty appropriate to their existing duties: teaching tenure, research tenure, and so on. But it must be said that making it possible for more higher-education teachers to be teacher-researchers is also an important project, which would be at best only indirectly advanced by campaigning for job security for all.

Rutgers: always casual Friday

I started down this rabbit-hole in an attempt to better understand the situation at my own institution. Rutgers has a large medical faculty and a large group of faculty and grads who are assigned to research and not instruction. On campus, these individuals matter a great deal. How much does counting or omitting them, separately or together, affect the overall casualization ratios, at Rutgers and at its broad peer group of “Very High Research Activity” research universities (a.k.a R1 universities)? Here is a lot of stuff on a chart:

The main use of this graphic is to verify that overall trends are broadly comparable with or without medical faculty and with or without the non-instructional faculty and grads. Public universities (faint blue lines) have notably larger grad populations, especially if research graduate assistants are included. The T/TT trend lines are also striking compared to the other staffing categories in their relative lack of variation, both across time and across universities, around the 35% rate.⁶ By contrast to the long-term picture of decline, this looks a little like a short-to-medium-term equilibrium: perhaps 35% T/TT is the point at which the remaining power of the tenured rump to reproduce itself meets university administrators’ desire to maintain the prestige value of having some “traditional” professors around, as long as they don’t cause too much trouble. But news from Florida and Georgia shows that right-wing politicians are ready and able to disrupt any such equilibrium. The moment may soon be ripe for a red-state tenure-elimination wave which would bring many public universities down to 0% T/TT.⁷ In the meantime, advocates of tenure had better commit themselves to the slow boring of hard boards, faculty line by faculty line.

Dismal forebodings aside, highlighting Rutgers on the chart shows my university’s exceptional reliance on part-time faculty. No matter how you slice it, Rutgers’s thousands of adjuncts make Rutgers quite unusual among R1 universities. Restricting to non-medical instructional faculty where the use of PT faculty is most intense, Rutgers ranks 4th among public R1 universities and 8th among all R1s in its proportion of part-time instructional staff. I’ll give the table of all R1s at the end of the post. Rutgers’s high rate of adjunctification corresponds to a smaller proportion of tenured and tenure-track faculty (as well as of TAs). At the same time, Rutgers is also clearly working on a “solution” to adjunctification in its increasing use of full-time NTTs, to the ultimate detriment of both teaching and research. Many other universities are moving along the same trajectory. If an alternate path is to be found, it will require a strategy of organizing across the distinction between research-focused and teaching-focused workers, and across the related but not identical divisions between more secure and less secure job categories as well. Locally at least, it seems to me we have not yet exhausted our capacity to fight back. But to renew tenure at the national level would require a real increase in the share of social resources invested in learning and research, taking the reins of our public institutions back from an indifferent or contemptuous ruling class and the penny-pinching privatizers who serve them as college and university managers.

R source code and instructions for reproducing the plots and numbers in this post may be found at https://github.com/agoldst/ru-casual2022, together with some possibly helpful notes and code about handling the IPEDS data.

R1 universities, ranked by contingency in instruction

(For simple further exploration, I also created a tab-separated values file downloadable from Github with these staffing category breakdowns for each Title IV university. The format is suitable for a spreadsheet program.)

Types of non-medical instructional staff at R1 universities in 2020–2021, arranged by increasing proportion of T/TT faculty. Public universities in italics.

RU-VH University	TA	NTT	PT	T/TT
CUNY Graduate School and University Center	65.5%	1.8%	21.7%	11.0%
University of Louisville	31.3%	11.2%	38.0%	19.5%
Temple University	19.1%	20.4%	38.0%	22.6%
University of California-Berkeley	57.3%	6.6%	13.0%	23.1%
Columbia University in the City of New York	32.9%	9.5%	34.4%	23.2%
Boston University	33.4%	25.0%	18.1%	23.5%
New York University	0.2%	21.9%	53.3%	24.6%
Tufts University	25.2%	23.1%	27.0%	24.7%
The University of Texas at Arlington	29.5%	18.5%	26.5%	25.5%
University of Maryland-College Park	53.1%	9.2%	11.9%	25.8%
University of Cincinnati-Main Campus	31.3%	15.2%	27.7%	25.8%
Florida International University	18.4%	21.4%	34.2%	26.0%
George Mason University	23.9%	13.7%	36.4%	26.0%
Arizona State University Campus Immersion	39.2%	19.2%	15.4%	26.2%
Drexel University	10.5%	23.4%	39.9%	26.2%
University of Illinois Chicago	37.7%	20.7%	15.2%	26.3%
The University of Texas at Dallas	41.8%	18.0%	13.8%	26.4%
George Washington University	18.7%	13.5%	41.3%	26.5%
University of California-San Diego	60.6%	6.7%	6.1%	26.6%
University of Pittsburgh-Pittsburgh Campus	29.7%	26.0%	17.4%	26.8%
University of California-Los Angeles	47.8%	10.8%	14.5%	26.9%
Yale University	47.2%	11.3%	14.5%	27.0%
University of South Florida	38.3%	15.3%	19.2%	27.2%
University of Nevada-Las Vegas	20.8%	12.1%	39.3%	27.9%
University of Mississippi	43.6%	18.5%	9.8%	28.1%
University of California-Santa Cruz	55.9%	5.2%	10.6%	28.3%
Rutgers (all campuses)	17.5%	16.9%	36.8%	28.8%
University of North Texas	42.9%	12.9%	15.3%	28.9%
Georgetown University	12.5%	17.4%	40.9%	29.3%
University of Wisconsin-Milwaukee	34.2%	13.2%	23.2%	29.4%
University of Colorado Denver/Anschutz Medical Campus	18.6%	22.2%	29.7%	29.5%
University of Southern California	16.3%	22.4%	31.5%	29.8%
Virginia Commonwealth University	20.3%	22.0%	27.7%	30.0%
Oregon State University	42.6%	16.6%	10.7%	30.1%
The University of Texas at Austin	46.1%	14.8%	8.5%	30.6%
Tulane University of Louisiana	24.9%	18.7%	25.5%	30.9%
University of Michigan-Ann Arbor	31.1%	26.7%	10.9%	31.3%
Florida State University	48.2%	11.5%	8.9%	31.4%
Wayne State University	20.3%	14.9%	33.1%	31.6%
New Jersey Institute of Technology	20.4%	13.2%	34.7%	31.6%
Stony Brook University	34.0%	10.1%	24.3%	31.7%
University of Colorado Boulder	31.1%	12.0%	24.9%	31.9%
University of Connecticut	36.1%	12.1%	19.9%	31.9%
The University of Texas at El Paso	24.5%	19.6%	23.9%	32.0%
Georgia Institute of Technology-Main Campus	46.8%	10.7%	10.4%	32.1%
University at Buffalo	34.3%	12.0%	21.5%	32.2%
University of California-Davis	52.7%	8.5%	6.4%	32.4%
University of Missouri-Columbia	31.3%	18.7%	17.6%	32.4%
The University of Alabama	35.0%	17.0%	15.2%	32.8%
University of California-Santa Barbara	55.9%	6.1%	5.0%	33.0%
University of Houston	40.5%	16.2%	10.2%	33.1%
University of Massachusetts-Amherst	46.1%	11.4%	9.2%	33.2%
University of Illinois Urbana-Champaign	51.3%	10.9%	4.5%	33.3%
Northeastern University	19.5%	26.5%	20.8%	33.3%
SUNY at Albany	34.7%	6.7%	25.3%	33.3%
University of Alabama at Birmingham	29.0%	20.0%	17.6%	33.4%
University of Oregon	44.2%	9.2%	13.2%	33.4%
University of California-Riverside	56.0%	4.9%	5.5%	33.5%
Brandeis University	32.3%	15.7%	18.4%	33.5%
The University of Tennessee-Knoxville	43.5%	16.0%	6.8%	33.7%
University of Wisconsin-Madison	45.5%	13.6%	7.2%	33.7%
Boston College	25.4%	14.1%	26.7%	33.8%
University of California-Irvine	51.2%	6.1%	8.4%	34.2%
Texas Tech University	38.7%	19.0%	8.1%	34.2%
University of Minnesota-Twin Cities	36.0%	11.9%	17.8%	34.3%
Colorado State University-Fort Collins	29.7%	18.2%	17.7%	34.4%
University of Central Florida	29.5%	17.1%	19.0%	34.4%
University of North Carolina at Chapel Hill	36.8%	14.4%	14.4%	34.4%
University of Nevada-Reno	31.1%	11.2%	23.0%	34.7%
University of Kansas	42.5%	13.9%	8.4%	35.2%
Washington State University	31.6%	18.7%	14.5%	35.2%
Indiana University-Bloomington	38.8%	17.3%	8.6%	35.3%
Montana State University	26.5%	9.3%	28.8%	35.5%
Georgia State University	19.5%	25.8%	19.0%	35.7%
Case Western Reserve University	24.9%	21.5%	17.9%	35.8%
Binghamton University	41.0%	10.1%	13.0%	35.9%
Northwestern University	37.9%	15.6%	10.5%	36.0%
The Pennsylvania State University	22.5%	24.2%	17.1%	36.1%
University of Arizona	38.7%	13.6%	11.5%	36.2%
Ohio State University-Main Campus	34.7%	5.5%	23.6%	36.3%
University of South Carolina-Columbia	30.4%	13.3%	19.9%	36.4%
University of Iowa	41.6%	15.8%	6.1%	36.5%
Syracuse University	31.7%	9.7%	22.2%	36.5%
Oklahoma State University-Main Campus	41.9%	15.1%	6.5%	36.5%
Rensselaer Polytechnic Institute	40.9%	15.5%	7.0%	36.6%
Harvard University	40.3%	15.6%	7.2%	36.9%
University of New Hampshire-Main Campus	30.3%	15.9%	16.9%	36.9%
Texas A & M University-College Station	34.6%	20.4%	7.7%	37.4%
University of Washington-Seattle Campus	36.4%	12.6%	13.3%	37.7%
West Virginia University	29.7%	17.0%	15.5%	37.8%
Vanderbilt University	23.8%	23.4%	14.7%	38.1%
University of Oklahoma-Norman Campus	35.9%	14.3%	11.2%	38.6%
University of Miami	23.4%	22.7%	15.2%	38.7%
Washington University in St Louis	20.7%	17.0%	23.5%	38.8%
Clemson University	40.8%	16.5%	3.9%	38.8%
Virginia Polytechnic Institute and State University	39.9%	14.6%	5.7%	39.7%
Louisiana State University and Agricultural & Mechanical College	36.3%	15.9%	8.0%	39.8%
University of Arkansas	31.6%	20.1%	7.9%	40.4%
Duke University	26.6%	26.9%	6.0%	40.5%
North Carolina State University at Raleigh	38.7%	11.5%	8.6%	41.2%
University of Utah	34.0%	21.6%	3.2%	41.2%
Michigan State University	30.3%	19.6%	8.5%	41.6%
Cornell University	42.0%	12.9%	2.6%	42.6%
Auburn University	32.7%	14.9%	9.5%	43.0%
University of Notre Dame	33.0%	16.0%	7.8%	43.2%
Massachusetts Institute of Technology	29.7%	14.5%	12.3%	43.4%
University of Nebraska-Lincoln	32.5%	12.4%	11.5%	43.6%
Stanford University	36.6%	14.4%	5.5%	43.6%
Purdue University-Main Campus	41.3%	7.1%	7.9%	43.7%
Princeton University	35.2%	10.6%	9.9%	44.2%
Kansas State University	32.5%	16.0%	7.2%	44.4%
University of Florida	30.5%	16.3%	8.7%	44.5%
University of Hawaii at Manoa	30.2%	8.2%	16.6%	45.1%
Carnegie Mellon University	0%	32.7%	21.5%	45.8%
Mississippi State University	27.6%	14.4%	11.9%	46.1%
Iowa State University	32.1%	11.9%	9.3%	46.7%
University of Delaware	32.9%	18.9%	1.4%	46.8%
University of Virginia-Main Campus	30.0%	19.1%	3.1%	47.8%
University of Kentucky	32.8%	8.8%	10.2%	48.2%
Dartmouth College	31.0%	10.1%	10.5%	48.4%
Brown University	32.1%	11.9%	7.2%	48.7%
University of Southern Mississippi	11.6%	19.4%	18.4%	50.7%
Emory University	0%	20.3%	28.8%	50.9%
California Institute of Technology	33.7%	8.1%	5.2%	53.0%
Johns Hopkins University	17.2%	27.7%	1.9%	53.1%
University of Georgia	24.3%	16.5%	5.6%	53.6%
University of Chicago	8.3%	19.5%	17.5%	54.8%
University of Rochester	12.3%	27.1%	5.6%	55.1%
University of New Mexico-Main Campus	15.0%	21.2%	6.9%	56.8%
Rice University	0%	18.5%	16.9%	64.6%
University of Pennsylvania	0.0%	22.6%	1.2%	76.2%