The Algorithmic University: On-Line Education,Learning Management Systems,and the Struggle over Academic Labor

From Unbundling to the Rationalization of Academic Labor

The outside pressures of austerity, entrepreneurialization, and outsourcing on higher education are well documented (Ovetz, 1996, 2015a, 2015b, 2015c, 2017). In the midst of rising costs and declining revenues, neoliberal “disruptors” ⁶ have advocated fragmenting higher education at the level of systems, institutions, nonacademic services, instructional, and professional into separate “primary” (teaching and research) and “support” activities (administrative and support services). The strategy of such “unbundling” (Gehrke and Kezar, 2015: 93, and 119; McCowan, 2017: 737; Sandeen, 2014: 2) is break up, disperse, automate, privatize, outsource, and off-shore each component along the global higher education “value chain” (BCG, 2020a; Ernst & Young, 2012).

To date, all but the professional and instructional components have been mostly unbundled, leaving teaching and other academic services such as counseling, advising, financial aid, tutoring, library support, LMS tech support, American Sign Language, and admissions as current targets for rationalization. Today, there is relentless pressure to expand OLE and integrate telecommunications and AI such as “Packback” discussion and grading chatbots (Delaney, 2019; McKenzie, 2019) in an effort to physically unbundle higher education from place-based to online (Mazoué, 2012: 75). While edtech ideologues are quick to praise the lack of a need to invest in infrastructure and faculty salaries (BCG, 2020a), there is insufficient research demonstrating such cost savings once the fixed technology and staffing costs are included (Gehrke and Kezar, 2015: 129; Sandeen, 2014: 6–7). Preempting opposition to the high capital costs LMS companies are now covering the startup costs in exchange for half or more of students’ fees to recoup their investment (Williamson, 2020).

While there have been three previous phases of unbundling of higher education driven by external pressures (Gehrke and Kezar, 2015: 97–108), the current phase is targeted at unbundling the academic labor of teaching. The rationalization of teaching essentially seeks to fragment, deconstruct, and redistribute its three key elements of design, delivery, and assessment of teaching into as many as nine components no longer under the control of faculty (Gehrke and Kezar, 2015: 104; Sandeen, 2014: 3; Smith, 2008). Gehrke and Kezar describe this unbundling of teaching as “the differentiation of instructional duties that were once typically performed by a single faculty member into distinct activities performed by various professionals, such as course design, curriculum development, delivery of instruction, and assessment of student learning” (Gehrke and Kezar, 2015: 93–94). This has only been made easier by the nearly complete dismantling of the three pillars of faculty academic labor: research, service, and teaching by transforming nearly the entire faculty in contingent “just in time” adjuncts.

With the exception of Noble (2003), the recent research into so-called bundling and unbundling have almost no explanatory power. Lacking a class analysis, such theories are entirely unable to explain what is driving the deskilling of academic labor. The catchy concept of “unbundling” could instead be understood as a euphemism for “deskilling which involved a fragmentation of formerly comprehensive skill sets and the displacement of skilled labour (‘all-round’ academics. . .) by semi-skilled or unskilled workers (semi-skilled para-academics)” both inside and outside academia ⁷ (Macfarlane, 2011: 59; see also Czerniewicz, 2018). Those who have reframed the rationalization of academic labor into “unbundling” have mistakenly represented as an unstoppable monolithic force with no origin whose penetration is leading to a predictable outcome.

In reality, rationalization has a cause that can be explained. Considering the immense effort to impose it by force and the growing struggle of academic workers opposed to it, the outcome is far from predetermined. Rather than “unbundling,” we are better served to understand what is happening as the rationalization of teaching as a strategy to discipline and better control faculty academic labor (The Analogue University, 2019: 1187–1188) in order to produce more unwaged students who are self-disciplined and productive waged labor. The labor of faculty and students are linked. Teaching faculty’s labor is intended to produce disciplined student labor power for capital. To the degree that faculty refuse to discipline and students refuse to be disciplined, teaching becomes unproductive to capital and ruptures the circuit of reproduction.

For nearly a half a century we have been subjected to the neoliberal attacks on higher education for churning out too many students who are unprepared for work and unprofitable to employ. While this complaint is better laid at the feet of students who are engaged in everyday refusals of work, the imposition of work is the driving motivation for rationalization.

From Professor to Appendage of the OLE Machine

To better understand the rationalization of academic labor, we can draw on the work of Marx (1867: 481) who described the deskilling of workers characteristic of a new technical composition of capital. ⁸ Braverman (1974) further applied Marx’s analysis of rationalization to the Taylorization of craft labor at the turn of the 20th century. Bringing both Marx and Braverman into the classroom, Foucault (1977) applied rationalization to education as a strategy for the control and disciplining of academic labor.

Marx’s detailed analysis of the deskilling of craft workers in the rational organization of industrial production in the factory is entirely relevant to understanding the rationalization of skilled into deskilled academic labor today. As Marx demonstrated in his study of the new technical composition of industrial work, “Not only is the specialized work distributed among the different individuals, but the individual himself is divided up, and transformed into the automatic motor of a detail operation,” thereby transforming the worker into an “appendage” of the machine and the factory (1867: 481–482).

Just how the worker is transformed into the machine tender is illustrated in Braverman’s analysis of the rationalization of industrial labor by the work of engineer Frederick Taylor. As faculty labor is assessed and rationalized, course design, delivery, and assessment (McCowan, 2017: 738) becomes fragmented and the pieces redistributed to nonfaculty academic staff such as content experts, counselors, course designers, technical support, programmers, and outsourced to textbook and software companies.

Take for example how nonprofit publisher Norton’s February 2017 spam email to professors led with the subject line “No time for grading?” promising “our content, your course.” A May 2020 spam email from Packback further promises the use of AI “to improve student engagement for community college students. . .while also automating some of the administrative faculty burden that unfortunately comes with managing discussion.” These two companies are not merely pitching their product to engorge their bottom lines but for rationalizing of academic labor by what Harry Braverman (1974) famously described as the “separation of conception from execution” (Braverman, 1974: 113–114). He noted how this takes place when “the first step breaks up only the process, while the second dismembers the worker as well, means nothing to the capitalist, and all the less since, in destroying the craft as a process under the control of the worker, he reconstitutes it as a process under his control” (Braverman, 1974: 78).

As will be described below in more detail, OLE relies on the “datafication” and “dataveillance” of teaching (van Dijck, 2014: 198; Williamson et al., 2020: 351). By transforming the complex multivariate aspects of teaching into tasks that measure “competency” of students represented in the form of data, OLE serves to operationalize the rationalization of teaching into disassembled components that can be redistributed to specialized staff responsible for highly differentiated technical aspect of the course (Mcfarlane, 2011). What Marx and Braverman have taught us is that the rationalization of labor is not simply about reducing labor costs, although that is of critical concern. The cost of labor is a factor of the control of labor power. Capital must transform labor power from potential into actual work. Rationalization is a strategy for decomposing the power of academic workers in order to discipline and make them work.

Foucault similarly applied Marx’s analysis of the technical composition of labor to education and the body of the student. He meticulously related how “the human body was entering a machinery of power that explores it, breaks it down and rearranges it” (1977: 138). The “learning machine” Foucault described exists for “supervising, hierarchizing, [and] reward” (1977: 147). This is accomplished by breaking down the action of teaching and learning into its key components so that “to each movement is assigned a direction, an aptitude, a duration; their order of succession is prescribed” (1977: 152). Finally, Foucault noted that the labor of the student and faculty are similarly rationalized as the complex supervisory role of “the master” who assesses by the exam is replaced by the serialization and hierarchization of each task into a series along “disciplinary time” (1977: 159). Although he died about a decade before OLE was introduced, Foucault might as well have been describing its impact on teaching and learning today.

OLE is the central organizing principle of the strategy to impose a new technical composition of capital in higher education that is intended to better serve the emerging technical composition of capital. The US labor market is rapidly moving to contingent part-time, temporary contract work in which increasing numbers of workers, as much as 30–40 percent of the US labour force, work remotely and are monitored and managed by information technology (Conlin et al., 2010; The Economist, 2015). This rapid growth of contingent labor is intended to rapidly make the Northern labor force look more like the workers in the South where about 84 percent of India’s 470 million workers, for example, are “casual” or self-employed, e.g. contingent (Ness, 2015: 85). The adjunctification and rationalization of academic labor in higher education is not an exception to this new global division of labor, it is actually the model for it.

On the extreme end is the short-lived MOOC in which tens of thousands students select an on-line class from a higher education “platform” in which an adjunct professor delivers prepackaged standardized lessons, have no interaction with the professor or one another, and take exams “assessed” by a computer program in order to earn a “badge.” Although it has all but disappeared from discussion since its high-profile defeat at San José State University, the MOOC remains the ultimate objective of achieving the professor- and classroom-less “university” by enclosing all higher public education into an Uber-style platform system for distributing courses in which the content specialist is paid by the enrolled student according to surge pricing (Hall, 2018: 22–29). Abandoning earlier overt efforts to privatize higher education, those seeking to rationalize college and university teaching are now effectively taking the “long march through the institutions” by embedding themselves into campus administrations to use crises like the 2008 recession and the COVID-19 pandemic to accelerate the move to OLE.

The impact of OLE on learning outcomes, “student success,” and adjunctification are well documented elsewhere. Rather, this article focuses on the rationalization of teaching by analyzing how changes in the organization, methods, processes, and strategies for organizing work are intended to decompose the power of academic workers (Ovetz, 2020b: 12). Because the labor-intensive teaching and learning that come from human interaction, social relationships, and emotional and intellectual exchange is lacking in the LMS, teaching is rapidly becoming deskilled into assessment, measurement, and monitoring while learning is being replaced by competency of task completion. Just how this is occurring through OLE is the focus of this article.

This new technical composition can be seen in the rapid expansion of OLE run on the Canvas LMS and the delivery of courses through Zoom, which has seen a rapid expansion of use during the pandemic (Ovetz, 2021) ⁹ . In order to understand the current technical composition of higher education a workers’ inquiry into academic labor will be explored below by examining how learning in higher education is being transformed into competency and “precarity skills,” and how the datafication of higher education is being pursued as a solution to the transformation problem of transforming labor power into work.

From Learning to Competency and “Precarity Skills”

The attack on faculty academic labor is not limited to employment status. It is fundamentally an attack on the very prerogatives of faculty control over teaching. In the past decade, faculty autonomy over course design, content, delivery, and student assessment have been challenged, and even displaced, by the efforts to replace content-based assessment of learning, represented by the grade and degree, with competency-based standards, rubrics, Departmental and Student Learning Objectives, badges, micro-credentials, pathways, and certifications. While there is much to criticize about grades as IOUs on future wages, disciplinary tools, and a mechanism for sorting graduates into a hierarchical labor market, these alternative assessments are not intended to address these concerns.

Rather, alternative assessment methods have been imposed from top down both by corporate funded foundations, task forces, think tanks, advocacy groups, politicians, and accreditation agencies with the intention of profiting coercing changes in higher education policy and teaching in order to profit from their investments in so-called edtech. Considering the commodity value of longitudinal datasets from “edtech,” venture capital funds are “not gifts, however, but tools of data extraction, the real costs of which will be paid with behavioral and cognitive data harvested from teachers, parents, and students, to serve measurable outcomes of ‘impact’ to guarantee ‘returns on investment’ for social venture capital if greed-upon metrics are met” (Marachi and Quill, 2020: 418; see also Marachi and Carpenter, 2020: 1). Due to disruption of the education of about 1.6 billion students in 200 countries by the pandemic, the edtech industry is expected to reap windfall profits estimated to double to $341 billion in total value, with online degree providers doubling in size to $74 billion by 2025 (Business Insider, 2019; Hogan and Williamson, 2020; Holon, 2020).

The objective is not so much to provide a more effective assessment tool but to remove assessment from the control of the faculty who conduct them based on personal and professional interactions with students and evaluations of their learning. Such alternative assessments are pitched with deceptive doublespeak framing of “equity,” “empower,” “flexibility,” “access,” ¹⁰ and most cynically “personalization.” In reality, these reforms, long put on notice for lacking valid research, actually deskill faculty, and privatize, standardize and de-personalize education. The consequences for learning are catastrophic, with the effect that “will fundamentally disempower youth and exploit the very communities that the solutions purport to help” (Sandeen, 2014: 7; see also Gehrke and Kezar, 2015: 105–106, 121–122, 130–131; Marachi and Quill, 2020: 430; Marachi and Carpenter, 2020: 2, 18).

Countering the claims made by advocates of alternative assessment is like watching the folktale of the blind man who talks to the elephant’s ass and wonders why it never responds to him play out in full view. The objective of this “reform” is the same thing as the reformers’ strategy. The intention is not to come up with a “better” assessment of learning but to remove assessment entirely from faculty control. The strategy of measuring competency rather than learning is analogous to the imperative of prior waves of automation, which “would enable management to discipline, deskill, and even circumvent and displace, the machinist, thereby to gain complete control over production” (Noble, 1993: 66).

Removing faculty control of course design, delivery, and assessment are intended to make all academic labor interchangeable regardless of academic content. To achieve such interchangeability, the objective criteria for assessing the student is no longer learning but is shifting to “competency.” The difference between the two are dramatic. Learning presupposes critical thinking, exploration, analysis, intellectual growth, and self-awareness. Competency is the internalization of normative rules, processes, procedures, relationships, and laws. While neuroscience and pedagogical research confirm that learning is about making connections, competency is measured by the completion of isolated fragmented tasks (Gehrke and Kezar, 2015: 122–123).

As a result, measuring competency is a matter of assessing whether the student (1) follows directions, (2) completes required tasks in sequence, and (3) completes the tasks efficiently and effectively. Macfarlane describes this as a “shift in emphasis in higher education from teaching students to supporting their learning more broadly associated with the associated use of information technology.” The intent is that the measurement of competencies can now be obtained from a wide variety of sources of which faculty are only one (Macfarlane, 2011: 63; see also McCowan, 2017: 739–740). The faculty in effect are transformed from teacher to machine tender. Mazoué is forthright about the strategy for achieving this when he writes that “we need to . . . individualize student learning and standardize faculty practice” in which “teachers monitor academic progress and apply appropriate interventions” (Mazoué, 2012: 79). These interventions now take place with chatbots and technology specialists, which are intended to rationalize, standardize, and reproduce what “good teachers do,” making all actual teachers replaceable (Mazoué, 2012: 87). The CEO of Instructure, which owns Canvas, has touted the use of predictive analytics and machine learning to entirely replace faculty with automated directions to students to carry out school work (Hill, 2019).

The shift to measuring competency effectively lops off the upper half of Bloom’s taxonomy, which can only be assessed through a labor-intensive, subjective, and imprecise pedagogy. Although most students enroll to receive the latter, the reformers appear to know better. Learning is packaged as no longer requiring more than explanation, repetition, and application. Students are no longer expected to engage in creative exploration but simply perform tasks.

The mission is to further subordinate learning and teaching to the prerogatives of employers. In the quest to produce more disciplined labor power, employers seek an assessment tool that can more accurately gauge the productivity of labor and its willingness to submit to work remotely by algorithmic management (BCG, 2020a). OLE is the methodology for teaching and assessing competency by modeling student “learning” to serve the same technical composition of precarious labor managed by big data elsewhere in the workforce.

The shift from learning to competency is made possible by the nearly complete adjunctification of the faculty, perhaps one of the worst defeats of the labor movement in the past three decades. Stripping faculty of control in order to standardize course design, production of content, ¹¹ delivery, and assessment has removed much of the impediments to automation. As more and more faculty integrate the LMS into even their in-person courses, the computer is increasingly being used to assesses and track how students perform tasks rather than how they think and create. Self-disciplined completion of tasks can now be evaluated by objective measurements of disconnected tasks such as length of time spent and quantity of words and other bits of product submitted to complete each task. Teaching is being transformed from “imparting knowledge to one that is focused on creating the conditions that best enable students to learn” based on their own self-discipline (Mazoué, 2012: 75). As learning shifts to competency, faculty are being refashioned as a “monitor” and creator of necessary “conditions” for students to work.

Behind these “objective” measurements lie the emphasis on the new difficult to measure “precarity skills” desired by OLE proponents. Among these “skills” now include adaptability, flexibility, habits, personality traits, and self-direction, which are considered “broader definitions of success, venturing beyond traditional academic measures” (Kaplan nd). EdSurge advocates the “MyWays” framework, which much like other OLE advocates reframes standardization as “personalized.” This framework divides 20 core competencies into four “domains” in which “content knowledge” is only one. Competencies that prepare a student for precarious labor run throughout the other three “domains” including “behaviors,” “perseverance,” “positive mindsets,” “learning strategies,” “responsibility,” “life skills and landscapes,” “surveying work,” “identifying opportunities and setting goals,” and “developing personal roadmaps,” among others (Figure 1). According to EdSurge, these domains “provide insights into new measurement approaches” that answer the “complexities of evaluating growth in non-academic skills” (Kaplan nd).

OPEN IN VIEWER

Figure 1.

My ways competencies (Kaplan nd).

These desired attributes and measurements of competency and “growth” match the growing demand for precarious gig workers who have little direct supervision other than the ubiquitous all-seeing eye of the algorithm. The expected outcome is that workers will face their utterly low wages, insecurity, and oppression with “grit,” “thriving,” and “positive mindsets,” rather than agitate, organize, and strike. The expectation by the corporate sector is that the shift from “subjective” faculty assessment of learning to “objective” data used to measure competency will solve the stubborn “transformation problem” of turning potential labor power into actual productive work (Cleaver, 2019: 112–113). According the cofounder of the Arizona State University GSV Summit for venture capital edtech, real-time big data measurements can replace “less objective measurements” to inform hiring and firing (Marachi R and Carpenter R, 2000: 13).

Datafication of Higher Education as a Solution to the Transformation Problem

Under pressure to produce more efficient and productive trained waged labor, higher education has undergone immense pressure to demonstrate quantifiable, replicable, transferrable, and “interoperable” measurements of output and outcomes of nonacademic operations. This aspect of the neoliberal assault is already well documented and need not be recounted here. What is often overlooked in these accounts, however, is that “measure as a category of struggle suggests a basis from which to link or circulate struggles both within and outside the university” (De Angelis and Harvie, 2009: 28). The product and productivity of higher education to capitalism—producing disciplined student labor power available for waged labor—is widely overlooked. Higher education’s inability to produce the disciplined labor sufficient for the needs of capital is the central motivating factor of higher education “reform.” Because the academic labor of faculty is responsible for course design, delivery, assessment, advising, and criteria for graduating student labor power, teaching itself must be controlled.

Struggle as one might, there is no single accepted definition of what it means to teach and learn. Without venturing down into the rabbit hole of this debate, we can be definitively sure of one thing about teaching and learning: faculty are expected to control the holistic process of teaching and evaluate learning. The long struggle of educators to control the “art” of determining what is taught and how it is assessed has allowed teaching to long escape rationalization. As a result, employers must still address the stubborn transformation problem that is attributed to what educators call “academic freedom” in the classroom. The transformation problem is what Marx described as the capitalist’s struggle to transform the wage paid for labor power into work and surplus value. In the case of education, the transformation problem is one of converting grades for schoolwork into work for wages (Cleaver, 2019: 259, 305, 385, 429).

There have been many attempts to rationalize and standardize higher education according to what is called “outcomes based performance management,” which are reflected in quantified measurements of “pathways” to graduation and transfer, productivity measurements based on student demographics and units, grades, and degrees, departmental and program rankings, and quantifiable student opinions of teaching (Berg, 2019; Berg et al., 2016: 1–2; The Analogue University, 2019: 1184). Until today, these efforts at datafication ¹² (van Dijck, 2014: 198) have yet to successfully entirely infiltrate the domain of faculty autonomy to choose their pedagogy for teaching and method of assessment, if any.

That has changed with the use of OLE to rationalize teaching. Nine discreet components of teaching have been identified for rationalization, all of which make it actual, not merely possible. These components, a variation of which is portrayed in part in Figure 2 (The Unbundled University, 2017: 3), include instructional design, subject matter, development, delivery, interaction, grading, improvement, and advising (Gehrke and Kezar, 2015: 104). The central strategy for pursuing the rationalization of teaching is OLE in which the course is either taught by a single faculty member or a team of faculty and/or “subject experts.” Content is most commonly written by adjunct faculty or a textbook company, which is delivered through an LMS designed by a course developer. Although there are diverse models, the fundamental premise is that students complete various measurable tasks, assignments, and exams within the time frame of the term of a quarter, semester, or even asynchronous course “modules” that may begin every week.

OPEN IN VIEWER

Figure 2.

Online education program and course unbundled services (The Unbundled University, 2017).

Datafication makes it possible to turn teaching and learning into discreet standardized tasks that are interchangeable and transferable according to same objective measurement criteria. This is what is meant by rationalizing teaching and learning so that “they’re becoming more granular, multifaceted, and multimodal,” allowing for “flexible pathways. . .ease of access—ease of movement, portability, mobility” (Czerniewicz, 2018).

The modern equivalent of Taylor’s “time-motion man” in the early industrial assembly line factory are the “Distance Education” staff such as the course designer, developer, software companies, content producers, and textbook publishers. These technicians have been busy analyzing, assessing, measuring, and rationalizing teaching in order to fragment it into its component parts so that they can be automated, redesigned, and redistributed to low-skilled support staff, nonacademic technical workers, or administrative management. What remains is the curriculum “content” developed by faculty, although not exclusively, to be delivered through an LMS to students.

Datafication of learning and teaching in OLE is a strategy for solving the elusive transformation problem. In other words, it is a new tool to reassure employers of the reliability of grades to assess the efficiency and productivity of the labor power they purchase with a wage and whether that labor power will be sufficiently disciplined enough to be able to be converted into work. Evidence of student resistance to school work and faculty refusal to impose it are reflected in enrollment, grade, and degree inflation since the 1960s (Cleaver, 2019: 181, 259, 385, 460; Ovetz, 1996; The Wages for Students, 1975). The integration of data from within and outside the classroom with Solutionpath’s StREAM (Student Retention, Engagement, Attainment, and Monitoring) that provides a real-time “engagement score” for students based on in class activity, RFID card swipes, attendance, and even library checkouts provides a ubiquitous surveillance of productivity of all aspects of a student’s life (Williamson, 2020).

Datafication is but one aspect of the latest phase of the neoliberal effort to further subordinate higher education for service to capital rather than to the public good. Earlier forms of datafication of educational outcomes range from standardized admission tests, treating departments and colleges as “cost units” responsible for demonstrating quantifiable outputs such as enrollment (Ovetz, 1996), measurement, and monetization of research (Ovetz, 1996), quantifiable measurements of productivity for the purposes of tenure and promotion, and, of course, as a return on investment of tax revenue, tuition and fees, and opportunity costs of “investing” in a degree. Clearly, datafication is hardly new with big data.

What is different with OLE is that datafication is critical to the effort to rationalize teaching and assessment of learning so that it can be usurped from faculty control. The rationale for doing so is that corporations are better able to both determine the content of teaching and the assessment of learning as a tool for hierarchizing, disciplining, and sorting student labor power. Solving the transformation problem would be impossible without controlling the ability to define the output and outcomes. To do this requires a shift from learning skills, which are notoriously difficult to assess, to competent completion of tasks, which are immensely easy to observe, record, and measure.

The Struggle Over the Algorithmic University

Resistance to these “reforms” has been primarily leveled at the external factors and the impact on loss of “quality,” declining “outcomes,” and cost while almost entirely missing the primary attack on academic labor. The implications of the rationalization of faculty academic labor has been apparent since Troutt first pitched the professor-less classroom more than four decades ago in which “an unbundled system assumes learning can transpire without students having to purchase the teaching function” (Troutt, 1979: 255). Today, it is common to read about the “automation of the profession” in which AI is paired with an entirely precarious faculty “machine tenders” delivering “digitally mediated rebundled teaching” (Czerniewicz, 2018). OLE is transforming teaching to be “focused more on coaching and mentoring and less on content delivery” (Sandeen, 2014: 5). The professor-less virtual classroom is attractive to universities that wish to be “swapping expensive lecturers for cheap, versatile machines that don’t go on strike don’t need sleep, and respond to students within nanoseconds” (Haw, 2019). Higher education faculty and unions have not yet grasped the full implications of expanding OLE, datafication, and dataveillance (Ivancheva, 2020). What is overlooked about edtech advocates is that they are not merely proposing to outsource rationalized teaching merely to make money but to reorganize higher education to better subordinate it to global capital accumulation.

To achieve this objective, it is first necessary to break the power of academic workers over teaching and learning. As Mazoué bluntly puts, “If we assume learning is dependent on teaching, and that teaching is an inherently labor-intensive activity, then we will never be able to increase productivity, improve quality, and lower cost simultaneously” (Mazoué, 2012: 80). As long as the faculty control teaching and assessment of learning, faculty labor is a critical choke point for disrupting the reorganization of higher education.

OLE is only the latest “reform” effort, which is intended to rationalize and measure academic labor. The outcome of a university education is not preordained because the struggle over measurement is a continuation of the struggle over the uses of academic labor. As De Angelis and Harvie remind us, “capital’s constant struggle to impose and reimpose the ‘law of value’ is always a simultaneous struggle to impose (a single, universal) measure” (De Angelis and Harvie, 2009: 27). As the anonymous academics writing as aptly named The Analogue University put it, “we need to do more than merely reveal the darker side of these transformative neoliberal relations; we need to find ways to mobilize and actively resist them” (The Analogue University, 2019: 1186).

Which Use Value Will Prevail?

What is blocking the reformers’ path is that a university education has different contested use values for faculty, students, and capital. For some faculty and students, education is a time for making connections, challenging assumptions, growing self-awareness, emotional and intellectual development, engagement, and learning to change the world. For many, perhaps the vast majority of students and faculty, higher education is about second set of use values. In the absence of a well-organized working class, many students understandably engage with higher education as a box to check off to get a “good job,” for example, avoid starving in a dead-end low-waged “shit job.” For possibly the supermajority of faculty, the use value of teaching is well-paying work that retains some level of autonomy in the workplace lacking in almost all other kinds of work. A minority find a use value in their teaching for service to capital and the state, the correlate to students pursuing a higher wage.

The use values of higher education for capital reside in access to knowledge, skills, and disciplined labor that can serve accumulation and domination. Capital’s use value of a university education is socializing a larger number of workers to exchange their labor power for a wage and discipline them to work with little or no supervision. To the degree it succeeds, higher education is valuable for reproducing the existing relations of production putting even more people to work to exploit their labor power.

To the degree that faculty control the curriculum and assessment, and have a role in determining the use of resources and administration of the campus, the first set of use values will remain prevalent and take precedence over the second set. This will make it possible to subordinate the wage to living life rather than living life for the wage. It will also make it possible to organize a mass movement that can envision and practice implementing a different system for organizing life.

Which set of use values prevail in the university has been periodically contested, most vividly in the 1960 and 1970s. The post–World War II “multiversity” reached its apex when for a few years young scholars and students embraced the first set of use values for higher education and rejected the second. Over the past four decades the pendulum has swung entirely back to favor higher education as not only a use value for capital but as an industry for the reproduction of disciplined labor power.

However, the strategies that preceded OLE were far from entirely successful in restoring higher education to its productive role to capital. Varying forms of everyday refusal of school work by faculty and students in continued grade and credential inflation, long delays in matriculation, cheating, and college graduates whose work is useless to their employers are still widespread. The “struggle over measurement” simultaneously illustrates the central role higher education plays in not only the reproduction of labor power and capital accumulation process but its continuing vulnerability to choke points of potential disruption by faculty and student academic workers.

The struggle over measurement is nothing less than a struggle over the imposition and performance of academic labor. Graduate student unionization and struggles over tuition hikes, privatization, and austerity since the 1980s transformed graduate students into the adjunct faculty who today are engaged in waves of unionization and strike-related action. As the majority of adjuncts teach at more than one campus and institution, they have carried the struggle over measurement from campus to campus along the very circuits of academic capitalism. The struggle over measurement dramatically illustrated in the 2016 University and College Union strike in the UK in which Newcastle University faculty refused to grade in an effort to directly disrupt a university outcome-based performance management plan (The Analogue University, 2019: 1199–2002). The refusal to grade has become an increasingly prevalent tactic used in in a wave of wildcat strikes of graduate students on nearly half the University of California system campuses refusing to submit their grades over two terms at the start of the pandemic, by CUNY adjuncts, and precarious faculty at the University of the Mirail in Toulouse, France, in Winter 2019 to Spring 2020.

Tactical Defiance and Strategic Rigidity

Resisting the rationalization of academic labor in all its guises from adjunctification to OLE will require devising new tactics, strategies, and objectives to circulate the struggle among more academic workers. To date, because there has been little attempt to assess the current composition of academic labor, the outcome is of yet uncertain. ¹³ With the arrival of the COVID-19 pandemic and global “online-ification,” OLE is now central to the struggle over academic labor. As Noble reminded us, “The ultimate viability of these technologies under the present mode of production depends, in the final analysis, upon the political and economic conditions that prevail and upon the relative strengths of the classes in the struggle over the control of production” (1979: 40). Online-ification is not a foregone conclusion.

Unfortunately the struggles of academic workers continue to follow ineffective tactics and strategies because they lack an analysis of the current technical composition of what Rhoades and Slaughter call “academic capitalism” (Rhoades and Slaughter, 2004; Slaughter and Rhoades, 2004). We commonly mark the connection between worsened academic working conditions to overcrowded classes, the lack of available courses, the rise in tuition, fees, and housing costs, and the push to online-ify more and more of higher education against the wishes of faculty and students. The predominant approach is to attack the neoliberal strategy for channeling the tax burden downward while increasing the costs to students paid by growing lifelong debt and work to repay it.

What is fatally missing is an effort to connect struggles over paid academic labor of faculty with those of students’ unpaid labor of schoolwork. In my teaching I show how academic workers have been “proletarianized” (Harvie, 1999: 105) by explaining to my students how I am also a “cloppener,” a precarious worker who closes the “shop,” in this case an evening class, and opens it up with the morning class, sometimes without a key, shared private faculty office space, or a parking space. This is an intuitive contrast for the many students who have similar contingent service jobs. Understanding and identifying the commonalities of precarious academic labor of the professor and student is the first step toward recomposing the power of all academic labor.

These connections need to be informed by an analysis of the role of higher education in capitalism in which faculty academic workers “co-produce new labor power” of new waged workers who “will in turn be employed to produce value and surplus value” (Harvie, 2006: 12). This class analysis will make explicit that reforms such as datafication, OLE, and performance measurements are each “a concrete expression of capital’s social drive to enhance the quality of human labour power” while driving down the costs to reproduce it (Harvie, 2006: 4, 14, 17). It is critical to make explicit that measurement of student work is the flip side of the assessment of academic labor of faculty. As Williamson et al observed, “measures of student performance, sentiment, engagement, and satisfaction are also treated as proxy measures of the performance of staff, courses, schools, and institutions as a whole” (Williamson et al., 2020: 34). The shift to OLE, datafication, and performance-based measurements are in reality a shift to continuous assessment and control of work both inside and outside of higher education.

Our academic unions, Balkanized on our campuses, sometimes in as many as 16 unions such as at UC-Berkeley, have complacently settled on the dual strategy of collective bargaining and lobbying in partnership with administrators. The reliance on contract unionism that swaps higher wages and benefits for conceding control over academic work has tied the hands of academic workers wishing to counter the encroachment of OLE. We will need to identify new forms of tactical defiance and strategic rigidity that complements the recent progress in organizing adjunct faculty (Ovetz, 2015a, 2017; Rhoades, 2013).

There are currently a range of quiet everyday forms of faculty resistance already in existence in the form of reducing laborious coursework and refusing to take waitlisted students to resist the speed up ¹⁴ (GradeInflation.com, 2016). Student tactics have included course hopping, dropping work heavy courses, cheating, and mutual aid such as sharing notes and study groups. Before the epidemic, many flocked to OLE courses under a widely shared assumption that they are “less work,” flexible, and subject to cheating. After the ubiquitous shift to all online during the pandemic, a survey found 16–63 percent of faculty either reduced the amount of work or indirectly inflated grades in several ways. Because the rates were higher for faculty who had never taught on-line before this suggests that faculty who teach on-line had already done so in their OLE courses ¹⁵ (Bay View Analytics, 2020). The limitations of these atomized actions are obvious. In some ways they might even fall into the trap of those pushing “competency” as faculty replace rigorous inquiry with “project based learning,” skills, and make work students resist.

The focus of student resistance is to escape schoolwork. But another aspect of the resistance is to being forced into waged work, often more than one contingent service job to survive, which is the case for an estimated 80 percent of students. This may explain the growing popularity of plans to make higher education free again and abolish student debt. Unfortunately, these demands fall short by failing to advocate for abolishing student loans and converting grant based aid such as tuition and fee waivers, and Pell and state grants into a wage. If students can get wages for the schoolwork, they already do they not only have a basis to resist waged work but it would also reduce pressure on their families members who support them and are similarly trapped in waged work. A wage for schoolwork would also provide students with leverage to resist the speed up such as pressure to matriculate by progressing through “pathways” that channel them into OLE courses and degree programs strictly based on their expected future wage and not what they would rather not do if they had a real choice. Wages would not simply pay for schoolwork but provide a tactical basis to resist its imposition (Cash and Boyce, 2017: 77–91; Grant, 1976; The Wages for Students, 1975).

Ultimately, current tactics of rigidity will need to develop into various forms of refusal. In the struggle over measurement, De Angelis and Harvie point to the rising frequency of tactics including work to rule, refusing unwaged tasks, fabricating documentation, or more often engaging in mindless “tick-boxing” when feedback is required under the “façade of compliance” (2009: 14–15). These everyday forms of resistance to both faculty imposing and students doing schoolwork establish the necessary social relations that precede and hint at organized disruption, grade strikes, and other forms of action at critical choke points in the new division of academic labor (Alimahomed-Wilson and Ness, 2018; Bonacich, 2003; Empire Logistics, 2016).

Even at the level of governance, faculty have immense power to diffuse, disrupt, or slow online-ification. Efforts are being made to “rebundle” academic labor (Czerniewicz, 2018) by keeping faculty in charge of designing, delivering, and assessing their own unique OLE courses. But more can be done. As long as academic senates still retain powerful roles in campus governance, the following tactics could be used:

These are just a few of the countless possibilities for expanding faculty intransigence and rigidity to slow down the process of online-ification and protect academic workers. OLE acolytes openly admit that “deeply entrenched” (BCG and ASU, 2018: 7; Czerniewicz, 2018; Young, 2018) faculty resistance is the greatest threat to further expansion and openly call for removing faculty control over OLE either by breaking shared governance and faculty unions or coopting faculty through stakeholder engagement and professional development (Young, 2018). Faculty should be escalating their tactics and deploying strategies to make this potential impediment a reality.

It is urgent to offer forms of the refusal as acts of solidarity between faculty that is increasingly contingent, deskilled, and managed by the algorithmic black box and students destined for the global labor market characterized by precarious low-waged work similarly managed by Odin’s algorithmic eye. Resistance to the role of higher education in producing disciplined labor power for exploitation points us to a way out of capital’s endless colonizing all of life as work.