What’s that smell? Bullshit jobs in higher education

Abstract

This study examines the growth of administrative and non-academic staff positions in the United States higher education sector through the twentieth and twenty-first centuries. We first document the growth of these employments relative to increases in faculty positions and student enrollments, as well as common explanations therefor. Finding those explanations wanting, we proceed to synthesize the work of David Graeber, Benjamin Ginsberg, Roberto Michels, and Thorstein Veblen, developing an alternative explanation that focuses on the bureaucratic tendencies of large organizations and the business principles that have underwritten college and university management for more than a century. Next, using Graeber’s typology of bullshit jobs as applied to higher education we generate testable hypotheses related to our explanation. We then conduct our empirical analysis of the incidence of bullshit jobs in higher education. Finally, we summarize and discuss our findings and conclude with suggested topics for future research.

KEYWORDS:

• Bullshit jobs
• managerialism
• information technology
• higher education
• administrative bloat

JEL Codes:

• B5
• J5
• L3
• M5

Introduction

In 1918, Thorstein Veblen described American universities as ‘corporations of learning’ that embrace the model of a well-conducted business organization, tempered only by humanity’s proclivity to idle curiosity and habits of thought favoring matter-of-fact knowledge (p. 152). The principles of competitive expenditure and competitive gain insist, Veblen argued, ‘on business capacity in the executive heads of the universities, and hence also the extensive range of businesslike duties and powers that devolve on them’ (1918, p. 62). And beneath these ‘captains of erudition’ sat the executive staff, managers selected for their ‘administrative facility, plausibility, proficiency as public speakers and parliamentarians, ready versatility of convictions, and staunch loyalty’ (1918, p. 95).

In taking stock of these administrative cohorts, Veblen was characteristically unreserved:

They are needless, except to take care of needs and emergencies to which their own presence gratuitously gives rise. In so far as these needs and difficulties that require executive surveillance are not simply and flagrantly factitious, – as, e.g. the onerous duties of publicity – they are altogether such needs as arise out of an excessive size and a gratuitously complex administrative organization; both of which characteristics of the American university are created by the governing boards and their executive officers, for no better purpose than a vainglorious self-complacency … . (1918, p. 277)

Nearly a century after the publication of this appraisal, Ginsberg (2011) described universities’ bureaucracies as occupied with an extraordinary number of vice presidents, provosts, vice provosts, associate provosts, deans, vice deans, associate deans, and assistant deans, along with armies of ‘other professionals’ (Waugh, 2003) serving as the administration’s ‘arms, legs, eyes, ears, and mouthpieces’ (Ginsberg, 2011, p. 25).

What do college administrators and professional staffers do? Most days they attend several meetings, usually with other senior academic managers. They also spend time planning professional conferences and staff retreats, engaging in strategic planning processes, and fund-raising. While time consuming, the contribution of these activities to the mission of the institution is debatable. According to Ginsberg (2011, p. 41), ‘little would be lost if all pending administrative retreats and conferences, as well as four of every five staff meetings, were canceled.’

The same sentiment can be found in David Graeber’s Bullshit Jobs: A Theory (2018). That volume collects a number of testimonials as well as Graeber’s own experiences as a professor to describe the extent, nature, and causes of jobs ‘so completely pointless, unnecessary, or pernicious that even the employee cannot justify its existence even though, as part of the conditions of employment, the employee feels obliged to pretend that this is not the case’ (2018, p. 19). Though Graeber’s research is not limited to higher education, the work of non-academic college and university employees features prominently.

It would seem, then, that the social contribution of these administrative and professional hierarchies in higher education cannot be taken for granted. Indeed, if Veblen, Ginsberg, and Graeber are correct, many of these employments may represent a significant, long-standing and ever-growing inefficiency in an important part of the community’s social, civic, economic, and moral infrastructure. To explore this further, the present paper inquires into the nature and causes of the growth of administrative and professional hierarchies in United States institutions of higher education. We argue that (1) the rapid rise of these employments over the past several decades is chiefly a consequence of internally generated dynamics, yet reflective of broader social norms and trends; (2) the assessment noted above, buttressed by the arguments of other social scientists, remain relevant; and (3) these conclusions are empirically supported by the self-reporting of individuals working in 2- and 4-year colleges and universities.

The remainder of this paper is organized as follows. First, we document the growth of higher education administration and professional staff since the 1970s and review common explanations and their inadequacy. Second, we synthesize the work of Graeber (2018), Ginsberg (2011), Michels (1915 [Paul and Paul 2001]) and Veblen (1918) to construct an alternative explanation that focuses on the bureaucratic tendencies of large organizations and the business principles that have underwritten college and university management since Veblen’s time. Third, we use the interdisciplinary lens constructed in the previous section and Graeber’s (2018) typology of bullshit jobs to develop empirically testable hypotheses concerning the nature of these non-academic employments in higher education. Fourth, we describe our data, how we measure bullshit jobs in higher education, and report the results of our analyses. Fifth, we provide a discussion of the results, and finally, we conclude.

The growth of higher education administration and bullshit jobs

Interest in administrative staffing, raised by Veblen (1918) a century ago, has carried through to today, in part due to the rapid growth in non-academic employment in higher education. As seen in Table 1, between 1976 and 2018, the number of full-time faculty employed at colleges and universities in the US increased by 92%, during which time total student enrollment increased by 78%. During this same period, however, full-time administrators and other professionals employed by those institutions increased by 164% and 452%, respectively. Meanwhile, due in part to the proliferation of part-time and adjunct faculty, the percentage of full-time faculty decreased from 67% to 54%, whereas the percentage of administrators who were full-time increased from 96% to 97% (National Center for Education Statistics, Digest of Education Statistics 2011 and 2019; see also Champlin & Knoedler, 2017, p. 237).

Table 1. Changes in full-time staff size and student enrollment, 1976–2011 and 1976–2018.^a

	1976	2011	% change 1976–2011	2018	% change 1976–2018
Faculty	434,000	762,114	+76%	832,119	+92%
Executive/Administrative	97,003	231,602	+139%	255,881	+164%
Other Professionals	150,319	699,867	+366%	830,189	+452%
Nonprofessional Staff	630,511	742,445	+18%	631,675	+0.19%
Student Enrollment	11,012,137	21,010,590	+91%	19,645,918	+78%

Source: National Center for Education Statistics.

^a Calculations relied on various issues (i.e. 2011, 2013, and 2019) of the Digest of Education Statistics (‘Digest’) issued by the National Center for Education Statistics. Specifically, we relied on the information presented in Table 258 (2011 Digest), Table 314.30 (2013 Digest), and Table 303.25 and Table 315.30 (2019 Digest). Additionally, Other Professionals and Nonprofessional Staff occupations were reclassified (i.e. disaggregated) as of fall 2013. Therefore, only the faculty, administrator, and student enrollment categories are comparable with data from 1976 and 2011. To approximate the Other Professionals and Nonprofessional Staff classifications from earlier years, multiple categories from 2018 were combined to approximate the 1976 and 2011 classifications as follows: Other Professionals (librarians, curators, and archivists; student and academic affairs and other education services;) and Nonprofessional Staff (service occupations; sales and related occupations; office and administrative support).

Explanations for this growth in managerial and professional staff often focus on demands put upon colleges and universities from some group other than the administrators themselves. Greater numbers of students, with expanding extracurricular demands; larger, more complex institutions, with new technological capabilities; and the like, for instance, require proportionately greater numbers of non-academic staff. Ginsberg (2011, pp. 27–29), however, finds this argument unconvincing, noting that new fields of teaching and research have also proliferated, without concomitant growth in faculty numbers.

Alternatively, it is commonly argued that external mandates – Title IX, Affirmative Action reporting, and so on – lie at the heart of this ‘administrative bloat.’ However, this argument is belied by the fact that private colleges and universities have seen substantially greater increases in non-academic staffing than their public counterparts (Ginsberg, 2011, pp. 29–32).

Relatedly, some have attributed these trends in higher education employment to the inefficiencies of public sector or nonprofit institutions in general. Zywicki and Koopman (2017), for instance, argue that, despite efforts to ensure that the heads of colleges and universities act like CEOs, they have nevertheless been forced to behave like government bureaucrats for want of the ‘profit and loss feedback signals that guide management conduct in a for-profit business setting’ (p. 30).

While it is a common misconception that private for-profit enterprises are less prone to bureaucratic distension, it is a misconception nonetheless (see Dean et al., in press; Gordon, 1996). Just the same, there is some limited evidence to suggest for-profit models are somewhat ‘leaner’ in higher education. Among postsecondary institutions participating in Title IV aid programs, private for-profit colleges and universities have the highest ratio of full-time equivalent faculty to all staff: 45.0%, compared to 35.1% and 35.7% for public and private nonprofit institutions, respectively. However, this higher faculty ratio is effected through higher numbers of students per faculty – 22.0% for private for-profits versus 15.3% and 10.2% for public and private nonprofits – and drastically lower ratios of full-time faculty to all employees – 18.0% for private for-profit 4-year institutions versus 65.9% for their public counterparts (National Center for Education Statistics, Digest of Education Statistics 2019).

Advocates for an even closer resemblance of colleges and universities to the private enterprise must be careful what they wish for. Modestly lower administrative ratios in for-profit colleges do not appear to reflect more resources to teaching and research, but rather higher payouts to executives and, of course, stockholders, as well as greater spending on marketing. That marketing, in turn, often targets vulnerable populations, not as a means to offer a hand up, but because they represent a largely untapped source of federal financial aid money (Schade, 2014; Watkins & Seidelman, 2017). Indeed, these institutions are the subjects of nearly all fraud claims from former students (98% in 2018; Flannery, 2018).

Hence, while the explanations above for the growth in non-academic employment in higher education may not be entirely without merit, they are weak enough to suggest a look to other sources. In the following section, we argue that bureaucratic growth in US colleges and universities does indeed reflect the normal dynamics of bureaucracies; but, importantly, also that this growth has occurred not for lack of business-like thinking and processes, but because of them.

On the nature and causes of the growth of non-academic employments

Graeber’s (2018, pp. 117–119) analysis describes the external forces pressing upon contemporary institutions, in general, as well as on colleges and universities, in particular. The proliferation of bullshit jobs is viewed as a consequence of structural changes in the economy, including the rise of finance capitalism, aggressive application of market principles, and growth in information industries. At the same time, cultural dynamics emphasizing the social value of work, together with political processes promoting the ideal of full employment have produced pointless, even pernicious, positions – i.e. bullshit jobs – in both corporate and noncorporate bureaucracies (Graeber, 2018).

These systemic processes are significant and necessary to understanding the historical context for the proliferation of bullshit jobs, but such structural dynamics are not the same as the bureaucratic features that determine an organization’s hiring, firing, and job creation. Administrative growth in corporations and in higher education also result from the internal bureaucracies that characterize these large organizations. Therefore, we suggest that unnecessary administrative and staff positions in higher education result from economic, cultural, and political forces overlaid on a highly bureaucratic and hierarchical organizational form. The following analysis of this intersection between structural dynamics and managerial hierarchy is informed by the work of Michels (1915/2001), Veblen (1918), and Ginsberg (2011).

An evolutionary explanation for the rise of bullshit jobs in colleges and universities can be usefully traced to the enrollment growth of the late nineteenth century. As Ginsberg (2011, pp. 148–149) describes, the boom in new students led to an increase in the size, more so than the number, of these institutions, and, coupled with the dominant business principles of the culture, brought the captains of erudition into ascendancy. Hence, by the early twentieth century, colleges and universities were increasingly guided by the principles of competitive advantage and growth – in prestige, fundraising, enrollments, and so on.

As Michels (1915/2001) theorized, all complex organizations, regardless of how democratic when founded, eventually develop into oligarchies. Rule by an elite, or oligarchy, is inevitable as an ‘iron law’ as a consequence of the ‘tactical and technical necessities’ of bureaucracy. Since no sufficiently large organization can function purely as a direct democracy, power is delegated to individuals within that group of oligarchs, elected or otherwise. Since they control channels of communication, administrators can manipulate the flow of information to support their decisions (Hyland, 1995; Michels, 1915/2001). A consequence of the impersonal authority of a hierarchy that operates within the managerial bureaucracy of higher education is that deans, provosts, chancellors, and the like are often selected from and appointed by a small group of familiar administrators (Michels, 1915/2001).

Once these administrators come to occupy a strong position in the managerial hierarchy, they begin to displace organizational goals with their own objectives. For Veblen, this manifests as an irreconcilable conflict between competitive habits of thought and learning, to which ‘competition in any guise or bearing … is detrimental’ (1918, p. 174). Businesslike management of institutions of higher education is bound to focus on publicity, for the institutions as well as their most prominent leadership, and the imposition of a ‘persistent and detailed surveillance and direction of the work and manner of life of the academic staff’ (1918, p. 223). These, in turn, inevitably stifle free inquiry and effective teaching.

Ginsberg (2011) would agree. In explaining the expansion of college and university managerial ranks he points directly to efforts by administrators to aggrandize their roles in the academy and to grab, maintain, and expand their decision-making authority. This view is consistent with the arguments of institutional economist Melman (1983) and could be productively extended to Veblen’s notion of conspicuous waste (see Dean et al., in press). Consequently, administrators ‘have a strong incentive to maximize the power and prestige of whatever office they hold by working to increase their staff and budget. As a means of justification, they often invent new functions to perform or seek to capture functions performed by others’ (Ginsberg, 2011, pp. 32–33).

On the academic staff, the impact is clear. Over a century ago, Veblen saw faculty becoming ‘a body of graded subalterns, who share confidence of the chief in varying degrees, but no decisive voice in the policy or the conduct of affairs of the concern in whose pay they are held’ (1918, p. 92). Ginsberg confirms these observations, which must certainly be truer today. Indeed, having little use for faculty in carrying out administrative necessities – which can be tended to by the armies of ‘other professionals’ – and no shortage of freshly minted PhDs with which to replace retiring or recalcitrant professors (2011, pp. 161–165), it would seem the consolidation of power by college and university administrators is nearing completion.

Investigating the incidence of bullshit jobs – development of hypotheses

When alternative explanations for observed phenomena exist, it falls on empirical investigation to sort the stronger from the weaker of them. This begs the question of how one would ultimately determine the nature of the positions in question. While myriad approaches could be conceived toward these ends, Graeber’s (2018) has a charming, exotic even, simplicity: ask the people in those jobs. Specifically, Graeber created an email account, [email protected], and encouraged people to send in testimonials (p. 33). From the resulting information and previous online discussion, Graeber developed a definition of bullshit jobs as unnecessary to the point that employees themselves cannot defend its existence despite a felt obligation to pretend otherwise (p. 19).

In the next section, we will use Graeber’s framework to test the extent to which the administrative and non-academic professional jobs that continue to inflate the bureaucracies of higher education are in fact bullshit. Before doing so, however, it is worth reviewing the typology of bullshit jobs that Graeber (2018, pp. 27–65) developed, and connecting them more explicitly to higher education. Four of these categories – namely, flunkies, goons, duct tapers, and box tickers – can reasonably be taken to cover the largest source of administrative bloat, the ‘other professionals.’ The last, taskmasters, covers the administrators themselves.

Taskmasters, according to Graeber (2018, p. 49), include both unnecessary superiors, for staff that would be doing the work anyway, and those who ‘create bullshit tasks for others to do,  …  supervise bullshit, or even  …  create entirely new bullshit jobs.’ The former brings to mind Ginsberg’s (2011, p. 164) hypothetical university president who, despite a seven-figure salary, ‘could be kidnapped by space aliens and it would be weeks or even months before his or her absence from campus was noticed.’ The latter would include what Ginsberg (2011, p. 2) refers to as ‘deanlets’ and ‘deanlings’, and might best be represented by Chloe (a pseudonym) whose lack of actual decision-making authority as a ‘Strategic Dean’ led her to ultimately conclude that she, and her many subordinates, had bullshit jobs (Graeber, 2018, p. 50). (Chloe would later realize that, were she given more authority as an administrator she may have in fact been forced to do actual harm, rather than simply useless work (p. 51).)

As we have seen, however, the decision-makers do not make up the greatest part of the growth of non-academic staff in colleges and universities. That honor goes to the ‘other professionals.’ Insofar as these might in fact be bullshit jobs, the most obvious type would be flunkies, serving chiefly to make their superiors appear or feel important. These would include administrative assistants and other office staff, hired and retained even if they are doing little work, lest the executive’s prestige suffer (Graeber, 2018). The existence of such jobs is neither obscure nor particularly new: Veblen (1918, p. 93) observed even in his time that the captains of erudition ‘will gather a corps of trusted advisors, whose qualifications for their peculiar work is intelligent sympathy with their chief’s ideals and methods and an unreserved subservience.’

Perhaps less obvious, but just as important, are the goons and the duct tapers. The former exist at any given organization simply because they exist at competing organizations (Graeber, 2018, p. 39). In higher education, these are the employees whose work relates to ‘one of the unwritten  …  commonplaces lying at the root of modern academic policy that the various universities are competitors for the traffic in merchantable instruction, in much the same fashion as rival establishments compete in the retail trade for custom’ (Veblen, 1918, p. 89). Graeber developed the latter category, duct tapers, principally from the software industry. These are employees ‘whose jobs exist only because of a glitch or fault in the organization; who are there to solve a problem that ought not exist’ (2018, p. 40). We take the category to cover more broadly those employments which exist because competitive exigencies and managerial dispositions supersede the essentially cooperative goals of teaching and inquiry.

And finally, some portion of the ‘other professionals’ at colleges and universities are surely box tickers, existing ‘to allow an organization to be able to claim it is doing something that, in fact, it is not doing’ (Graeber, 2018, p. 45). Box-ticking tasks range from filling out forms which, although ultimately meaningless, the bureaucracy considers more important than the actual functioning of the organization itself, to creating slick PowerPoint presentations: ‘If the ongoing importance of a manager is measured by how many people he has working under him, the immediate material manifestation of that manager’s power and prestige is the visual quality of his presentations and reports’ (Graeber, 2018, p. 47). Hence, box tickers might characterize some workers in occupations such as personnel, training, and labor relations specialists; accountants, auditors, and actuaries; counselors; as well as administrative assistants.

With these categories in mind, our aim is to test for the existence of bullshit jobs in higher education that the theory and history presented in the previous section leads us to expect. To the best of our understanding, there has been little to no attempt at empirically investigating the incidence of bullshit jobs in higher education from an interdisciplinary perspective that leverages insights from anthropology (Graeber, 2018), institutional economics (Veblen, 1918), political science (Ginsberg, 2011), and sociology (Michels, 1915/2001).

The limited scholarship published on the existence and nature of bullshit jobs (and similar conceptualizations) is reviewed in Dean et al. (in press). Of the literature reviewed therein, only one article stands out as speaking specifically to higher education: Wrzesniewski et al. (1997) report the results of a survey of full-time employees at a public university’s student health services and non-faculty employees of a small private liberal arts college. Notably, 38% of those respondents answered ‘false’ to the statement ‘my work makes the world a better place,’ and only 25% indicated that they would continue their work unpaid if they were otherwise financially secure. Given that the occupations of the respondents ranged from physicians and administrators to computer programmers and administrative assistants, these results indicate we are on the right track in looking for the existence of bullshit jobs in these employments.

Hypotheses

Our analysis is concerned with the underlying characteristics that differentiate bullshit jobs from one another and from academic staff (e.g. faculty, researchers, and graduate students). To that end, and based on the previous discussion, we develop two hypotheses that concern the occupational categories of non-academic staff and bullshit jobs – specifically other professionals (i.e. duct tapers, flunkies, box tickers, and goons) and taskmasters (managerial hierarchy). Both hypotheses assume controls for individual characteristics of the respondents (e.g. age, ethnicity, gender, salary, etc.) and for the institutional/organizational features of the colleges or universities that employed the respondents (e.g. size of employer, public/private organization, etc.).

Hypothesis 1: other professionals (flunkies, goons, duct tapers, and box tickers)

Other Professionals will have the lowest levels of perceived social contribution compared to all other higher education professional staff and administrators. That is, respondents in this category are more likely to believe their jobs are bullshit.

These positions, representing far and away the greatest source of bloat in higher educational bureaucracies (see Table 1), are central to understanding the nature and causes of that bloat. Many or most of these workers may in fact serve chiefly, as Ginsberg (2011, pp. 24–25) argues, as ‘the bulwark of administrative power in the contemporary university,’ ensuring autonomy from faculty (cf. Cottom & Tuchman, 2005), in which case their roles serve some other purpose than the usual academic goals of their institutions. Our hypothesis suggests this is the case, and that this can be observed empirically through the reporting of the workers themselves.

Hypothesis 2: taskmasters and administrative hubris

Management in institutions of higher education, from the captains of the erudition (e.g. presidents, chancellors, and provosts) to those in the many levels of administration below them (vice provosts, deans, and so on), will report higher levels of perceived social contribution than other non-academic staff.

Understanding this hypothesis requires reflection on an important observation Graeber made in his broader study of occupations. Graeber (2018, p. 57) reports that, among the multitude of employments for which he received responses, ‘business owners, and anyone else in charge of hiring and firing’ were among the least likely to report their jobs were ‘bullshit’; in fact, they often expressed outright hostility to the very idea of bullshit jobs. We interpret these findings as suggesting that the decision-making authority of a job, and especially authority over others, adds an additional dimension to the likelihood of an individual believing their job makes a contribution to society. Specifically, just as the ‘captain of industry’ is apt to see himself as ‘a benefactor of the community at large and an exemplar of the social virtues’ (Veblen, 1914, pp. 216–217), so too do the captains of erudition, ensconced in the institutional principles of pecuniary rectitude and legitimate authority, see themselves as champions of the general welfare. If such were indeed the case, the fundamental driver of the growth of administrative positions in higher education, along with their retinues of ‘other professionals’ covered in the first hypothesis above, would be patent.

Measuring bullshit jobs in higher education – data and methods

Data

Our study uses the 2017 National Survey of College Graduates (‘NSCG’). The 2017 survey is composed of individuals chosen from the four panel cohorts of the American Community Survey between 2009 and 2015. A unique characteristic of the NSCG is the inclusion of workers’ self-reported satisfaction of their job’s contribution to society. Specifically, survey respondents were prompted as follows:

When thinking about your principal job held during the week of February 1, please rate your satisfaction with that job’s contribution to society.

Individuals rated their satisfaction level on a Likert scale: very satisfied, somewhat satisfied, somewhat dissatisfied, and very dissatisfied. We narrowed the sample to individuals who met the following criteria: (1) were working for pay or profit during the survey week; (2) had an annualized salary of greater than zero dollars; (3) were employed at a 2- or 4-year academic institution; and (4) the academic institution was either public or private. Furthermore, we collapsed the respondents’ satisfaction level into two categories: (1) very satisfied, and (2) otherwise, similar to prior work by Nikolova and Cnossen (2020), and Dean et al. (in press). These steps led to a sample containing 5323 observations.

The NSCG data have a major limitation. The survey includes individuals with a variety of employment statuses across all industries. The survey was not constrained to individuals employed in higher education only; thus, we made inferences about individuals, their job status, and their likely presence in the higher education industry. The inference is based on a respondent’s job category and employment in an academic institution. The job categories we included were based on Graeber’s (2018) and our understanding of the higher education industry as members of that very industry.

The job categories our analysis focused on were: (1) faculty; (2) researchers; (3) graduate students; (4) taskmasters; (5) goons; (6) duct tapers; (7) box tickers; (8) flunkies; and (9) unclassified. Table 2 defines our variables and the individual and institutional controls we used.

Table 2. Variables and definitions.

Variables	Definitions
Individual characteristics
Age	Natural log of a respondent’s age in years
Wages	Natural log of a respondent’s wages per hour worked
Female	1 – respondent is female
	0 – otherwise
White	1 – respondent is white
	0 – otherwise
Married	1 – respondent is married with spouse present
	0 – otherwise
Masters^a	1 – respondent’s highest degree attained is a masters degree
	0 – otherwise
PhD	1 – respondent’s highest degree attained is a doctorate degree
	0 – otherwise
Supervisor	1 – respondent has a supervisory role
	0 – otherwise
Institutional characteristics:
Four-year	1 – respondent’s employer is a four-year college or university
	0 – respondent’s employer is a two-year college or university
Private school	1 – respondent’s employer is a privately owned academic institution
	0 – respondent’s employer is a publicly owned academic institution
0–499 Employees^b	1 – the number of employees at the respondent’s employer is 499 or less
	0 – otherwise
500–999 Employees	1 – the number of employees at the respondent’s employer is between 500 and 999
	0 – otherwise
1000–4999 Employees	1 – the number of employees at the respondent’s employer is between 1000 and 4999
	0 – otherwise
5000–24,999 Employees	1 – the number of employees at the respondent’s employer is between 5000 and 24,999
	0 – otherwise
Job categories:
Taskmasters	1 – respondent’s principal job code is consistent with Graeber’s taskmasters variety of bullshit jobs^c
	0 – otherwise
Other Professionals	1 – respondent’s principal job was consistent with Graeber’s flunkies, goons, duct tapers, and box tickers^d
	0 – otherwise
Faculty	1 – respondent’s principal job code was post-secondary teaching, they were not currently enrolled in school, and their primary work activity was teaching
	0 – otherwise
Graduate Students	1 – respondent’s principal job code was post-secondary teaching or research, they were currently enrolled in a graduate program, and their primary work activity was teaching or research
	0 – otherwise
Researchers	1 – respondent’s principal job code was basic or applied research, did not teach, and were not enrolled in school
	0 – otherwise
Unclassified	1 – respondent’s principal job did not satisfy any of the prior job categories or did not fit into Graeber’s typology
	0 – otherwise

^a Reference category is bachelors and professional degree holders for Masters and PhD.

^b Reference category is employers with 25,000 or more employees.

^c These included the following principal job codes including top-level managers (e.g. CEO, COO, CFO etc.), education administrators (e.g. provost, registrar), mid-level managers (e.g. natural science managers) as well as occupation codes that had managerial roles according to their reported primary and secondary work activities.

^d These included principal job codes closely aligned with the varieties of bullshit jobs in Graeber (2018, Chapter 2).

Preliminary data analysis

Figure 1 shows the distribution of the 5323 respondents given the outcome variable of one’s satisfaction with the jobs’ social contribution. Overall, 3379 out of 5323 (i.e. 63.5%) were very satisfied with their jobs’ societal contribution. The job category with the most observations and the most satisfaction with their jobs’ social contribution is Faculty (1920 out of 2701 or 71.1%). The job category with the highest number of respondents who were not very satisfied with their jobs’ social contribution was Other Professionals: 304 out of 585 of these respondents (or 52%). Unclassified jobs were those principal job activities that did not easily align with Graeber’s varieties of bullshit jobs or other job categories in higher education such as Faculty. They included jobs such as other health occupations, other engineering technologist, and diagnosing/treating practitioners.

Figure 1. Perceived social contribution of a job by position category.

Figure 2 shows the Pearson’s correlation coefficients (in percentages), for the variables examined in the logistic regression model. A negative percentage in the correlogram indicates a negative correlation between the two variables, and a positive percent in the correlogram indicates a positive correlation. The deeper shading indicates a stronger correlation. To assess multicollinearity, we also examined the variance inflation factors (VIFs) of the predictor variables. The VIFs range from 1.04 to 3.44 and confirm the independence of the variables used in our estimation.

Figure 2. Correlogram.

Empirical method

Following the preliminary analysis of the data, we employ a logistic model that takes the general form: $p_i\equiv \mathrm{Pr}\left(y_i=1|x_i\right)=\frac{e^{x^{\mathrm{\prime }}\beta }}{\left(1+e^{x^{\mathrm{\prime }}\beta }\right)}$ where $y_i=1$, is the probability that a respondent is very satisfied with their job’s social contribution; $x^{\mathrm{\prime }}$ is a vector of right-hand variables including controls for individual (i.e. age, wages, gender, race, marital status, educational attainment, and job categories) and institutional characteristics (i.e. two- versus four-year colleges, private versus public, and size of the employers), and whether or not a respondent was engaged in supervisory work; and β are the probabilities associated with the right-hand variables.

We generate odds ratios to interpret the likelihood of a respondent in our data to be very satisfied (i.e. $y_i=1$) with their job’s social contribution. The odds ratio represents the increase (or decrease) in the odds of being very satisfied with a job’s social contribution when an independent variable increases by a unit, all else constant. Consequently, an odds ratio of one implies equal odds of being either very satisfied ($y_i=1$) or otherwise ($y_i=0$). An odds ratio greater (or less than) one implies higher (lower) odds of a respondent being very satisfied with their job’s social contribution (Cameron & Trivedi, 2010, pp. 459–460).

Results

With respect to our first hypothesis regarding Other Professionals, controlling for individual and institutional characteristics, we find these employees are less likely to be satisfied with their jobs’ social contribution with an odds ratio, 0.494, considerably below one (see Table 3). Graeber’s assertion and our hypothesis are supported.

Table 3: Results – other professionals.

Variable	β	Odds ratio
Other Professionals	−0.705***	0.494
	(0.093)
	(0.093)
Individual characteristics:
Age	0.513***	1.671
	(0.106)
Wages	0.049	1.051
	(0.048)
Female	0.247***	1.279
	(0.059)
White	0.187**	1.205
	(0.065)
Married	0.087	1.089
	(0.064)
Masters	0.262***	1.300
	(0.074)
PhD	0.155	1.167
	(0.082)
Institutional characteristics:
Four-Year	−0.481***	0.618
	(0.090)
Private School	0.078	1.081
	(0.080)
0–499 Employees	−0.064	0.938
	(0.102)
500–999 Employees	−0.071	0.931
	(0.120)
1000–4999 Employees	−0.159	0.853
	(0.120)
5000–24,999 Employees	−0.178	0.837
	(0.095)
Constant	−1.459***	0.232
	(0.389)
n	5323

Notes: Standard errors in parentheses.

*p < 0.05.

**p < 0.01.

***p < 0.001.

With respect to our second hypothesis, taskmasters and administrative hubris, once we control for individual and institutional characteristics, we find that Taskmasters compared to other bullshit jobs are likely to be very satisfied with their jobs’ social contribution (see Table 4). Moreover, responding in the affirmative to the question ‘Did you supervise the work of others as part of the principal job’ is also associated with an odds ratio greater than one. Graeber’s assertions and our hypothesis are supported: the captains of erudition, as Veblen (1918) would put it, as well as the other administrators, see themselves as benefactors to the community, and this is positively correlated with playing a supervisory role over others.

Table 4. Results – taskmasters and administrative hubris.

Variable	β	Odds ratio
Taskmasters	0.408**	1.504
	(0.137)
Faculty	0.939***	2.557
	(0.106)
Graduate Students	0.696***	2.005
	(0.113)
Researchers	0.385**	1.469
	(0.137)
Unclassified	0.388*	1.474
	(0.184)
Individual characteristics:
Age	0.383**	1.466
	(0.117)
Wages	0.035	1.036
	(0.050)
Female	0.244***	1.277
	(0.060)
White	0.147*	1.159
	(0.066)
Married	0.062	1.064
	(0.065)
Masters	0.247**	1.281
	(0.075)
PhD	0.030	1.031
	(0.091)
Supervisor	0.314***	1.369
	(0.067)
Institutional characteristics:
Four-Year	−0.417***	0.659
	(0.094)
Private School	0.061	1.063
	(0.082)
0–499 Employees	−0.086	0.917
	(0.103)
500–999 Employees	−0.089	0.915
	(0.121)
1000–4999 Employees	−0.167	0.846
	(0.096)
5000–24,999 Employees	−0.155	0.856
	(0.096)
Constant	−1.739***	0.176
	(0.448)
n	5323

Note: Standard errors in parentheses.

*p < 0.05.

**p < 0.01.

***p < 0.001.

Additionally, the other job categories, Faculty, Graduate Students, Researchers, and Unclassified, all have odds ratios greater than one, indicating that, all else constant, each of these job categories is likely to have a respondent who perceives that their job contributes to society. The largest odds ratio pertains to Faculty, who are 2.557 times more likely than Other Professionals to perceive that their jobs contribute to society – an unsurprising finding.

Discussion

In discussing potential explanations for the proliferation of bullshit jobs, Graeber (2018) proposes an intersection between the financialization of the economy (with aggressive application of market principles and technology), cultural dynamics (emphasizing the social value of work), and political processes (promoting full employment). Our own explanation, specific to higher education, has emphasized the general bureaucratic tendencies of large organizations as well as the competitive, businesslike way in which these institutions are run. The results for taskmasters and other professionals (i.e. box tickers, duct tapers, flunkies, and goons) in the previous section, while consistent with Graeber’s (2018) and our arguments, warrant further discussion. Why do taskmasters believe they contribute more to society than other bullshit jobs, while other professionals report the lowest levels of social contribution?

As Michels (1915/2001) argued, all large bureaucracies eventually come to be run by a ‘leadership class’, i.e. paid administrators and executives (taskmasters). Far from being responsible to the organization's membership, this ‘leadership elite’ rather than the rank-and-file, inevitably come to dominate the bureaucracy's power structures. With power comes the ability to determine what procedures the organization follows when making decisions.

By controlling access to information, those in leadership positions can centralize their power, often with little accountability, due to the apathy, indifference, and non-participation most rank-and-file members have vis-à-vis their organization's decision-making processes (Michels, 1915/2001). In this bureaucratic structure, the rank-and-file include other professionals (box-tickers, duct tapers, flunkies, and goons), whose jobs are created by administrators. Box-ticker positions appear when paperwork requesting certain actions become more important (to management) than the actions themselves. When administration finds it more difficult to fix a problem than to continue dealing with its consequences, duct-taper jobs are created. Because those in positions of power view subordinates as symbols of status, they will often surround themselves with flunkies. And when rivals add new positions goon jobs are developed to ‘keep up’ with the competition (Graeber, 2018). These ‘other professionals’ in higher education, then, find themselves in the unenviable position of having to repeatedly perform tasks that they believe are unnecessary. Therefore, it should come as no surprise that employees in this category report on average the lowest level of social contribution compared to all other higher education professionals (e.g. faculty, researchers) and administration.

Our results for the taskmasters themselves require a more nuanced explanation, and the limitations of our approach must be recognized. As indicated above, the cultural norms that associate power, prestige, and pecuniary gain with social beneficence are likely to extend to the captains of erudition, just as they do the captains of industry and finance. Indeed, this is probably amplified in the former by the common perceptions of the value of higher education itself. Consistent with Graeber’s (2018) observations, we expected these sentiments would outweigh any felt dearth of social contribution among the upper-tier administrative employees in our sample.

The same habits of mind may extend down into the lower managerial ranks, or the organizational culture itself may attenuate reports of bullshit work. As described by Veblen (1918, p. 92), the university executive will employ a small number of advisers who are in sympathy with leadership ambitions, and who will ‘form an unofficial council, or cabinet, or “junta,” to whom [the administrator] can turn for informal, anonymous and irresponsible, advice and moral support.’ The university executive staff’s ‘first duty is a loyal obedience; … they must utter no expression of criticism or unfavourable comment on policy, actions or personal characteristics’ (p. 91). It may therefore be due to the resulting organizational culture that lower level taskmasters, as well, have an incentive to underreport their work as ‘bullshit’ (Graeber, 2018).

Finally, it must be noted that our taskmasters category has only limited capability in distinguishing between those administrators with real decision-making authority and those without. That employees whose jobs explicitly involved supervising others’ work carried a statistically significant odds ratio greater than one supports our understanding that such authority will, on average, lead to more sanguine beliefs about the social value of one’s work. However, administrators without, e.g. powers of the purse, such as Chloe, one of the few managers identifying her job as bullshit in Graeber’s (2018) research, cannot be readily differentiated from the others. As such, distinctions between the two may be lost in the average.

These arguments help to explain our finding that taskmasters, compared to other bullshit jobs (i.e. other professionals) in higher education, believe they contribute more to the social good. However, the data and our models have limitations, and further research is in order. These will be discussed in more detail in the concluding section.

Conclusion

The reasons individuals create or perform bullshit jobs are not the same as the historical factors that account for the growth of unnecessary employment in higher education. Structural forces that transform economic, political, and cultural institutions are not the same as the organizations and people that respond to these dynamics. By applying a multi-disciplinary explanatory approach, our research brings new insight to the way the phenomenon of bullshit jobs in capitalist societies may emerge from multiple levels of analyses. However, the present paper is not without its limitations, and more research is, of course, needed.

The empirical analysis of the present study relies on self-reporting of social contribution. This is both a strength and a limitation. As to the former, when attempting to assess the nature and causes of work and organization it is essential that social scientists look, in part, to the people actually engaged in those activities. However, an individual’s perceptions are inevitably influenced by cultural norms, requiring interpretation of such self-reporting through an understanding of the relevant norms and their personal history (see Dean et al., in press). One potential line of future research is the relationship between perceptions of social contribution and the experiences and habits of thought that shape those perceptions.

Moreover, it is possible to verify empirically whether the perception of a job’s social contribution is confirmed or contradicted by others within the organization. For instance, Graeber notes that, while managers almost ubiquitously see themselves as contributing a good deal, one survey found ‘that 80% of employees feel their managers are useless and that they could do their job just as well without them’ (2018, p. 49). Although such cross-assessments of occupational usefulness are not available in our dataset, future empirical work along these lines would most certainly be interesting.

It should likewise be noted that our dataset only surveys college graduates. Because nonprofessional staff have not been a significant part of the growth of employment in higher education (see Table 1) and we assume that the remaining occupations under scrutiny here generally require at least a Bachelor’s degree, this has not been a significant hurdle for our purposes. Nevertheless, analyses that cover the full range of work done in colleges and universities remain in order.

In addition to the above-noted lines of future inquiry, we believe that time series data analysis would also shed light on the evolution of bureaucratic structures, intra-organizational power, and bullshit jobs. There remains as well a great deal to explore in the intersection of race, gender, and a host of other dimensions with occupation and social contribution in our institutions of higher education. We hope that all of these and surely many other lines of research will be pursued in the future.

We conclude by situating our study in the broader cultural context. In his time, Veblen indicated that the guidance of higher education under principles of competitive business remained incomplete. However:

If these business principles were quite free to work out their logical consequences, untroubled by any disturbing factors of an unbusinesslike nature, the outcome should be to put the pursuit of knowledge definitively in abeyance within the university, and to substitute for that objective … a consummate (‘sweat-shop’) scheme … of low-cost perfunctory instruction, high-cost stage properties and press-agents, public song and dance, expensive banquets, speech-making and processions. (1918, pp. 170–171)

Ginsberg (2011) might like to add administrative retreats and strategic plans, but otherwise the list appears to cover well the various non-academic concerns to which so much of college and university resources are now directed.

Still, while it would be difficult to deny that a more complete business-like management of the higher learning has developed since Veblen’s time, it would likewise be imprudent to miss the hold-outs, those who still look to our institutions of higher education for opportunities to learn and to engage in free inquiry. Certainly, this would include most faculty as well as students. But, given the number of workers in the other professional category expressing dissatisfaction with their work’s social contribution, it would be reasonable to conclude that many of these employees long for a less bureaucratic, less pecuniary, less corporate higher education. Indeed, even the captains of erudition will sometimes hold a sincere, if vague, appreciation for the traditional academic ideals (Veblen, 1918, p. 173) – perhaps a more common sentiment as one moves down the administrative hierarchy and hence closer to the actual points of teaching and scholarship.

A number of forces and institutional dynamics push against these desires to align institutions of higher education with their own missions. We have argued for the importance of two in particular: the bureaucratic tendencies of large organizations and the logic of management of colleges and universities under the principles of competitive business. Yet, as a final word, it is worth returning to Graeber, who suggests that at a very fundamental level contemporary culture maintains and proliferates bullshit jobs through a ‘balance of resentments’:

The main political reaction to our awareness that half the time we are engaged in utterly meaningless or even counterproductive activities – usually under the orders of a person we dislike – is to rankle with resentment over the fact there might be others out there who are not in the same trap. (2018, p. 13)

Those with actually useful – yet, more often than not, poorly compensated – work begrudge those with job security, high incomes, and prestige – and especially those who demand or indeed enjoy such perquisites in addition to a putatively beneficent role in society at large. Meanwhile, the masses of flunkies, et al., toiling away in bullshit work, resent those – for instance, primary and secondary school teachers – who have been afforded an opportunity to truly contribute to society, even if it has meant sacrificing material comfort and leisure (Graeber, 2018, pp. 175–180).

Perhaps a similar balance of resentments is present in the halls of the academy. Veblen (1918, p. 173) noted that part of the abiding appreciation for teaching and scholarship above all else among faculty was a ‘touch of envy [for] those among them who are so driven to follow their own scientific bent, to the neglect of expedient gentility and publicity.’ Champlin and Knoedler (2017, p. 240) likewise observe a counterpoise between contingent and traditional (full-time, tenured or tenure-track) faculty. The former resent the latter for their generous pay, benefits, and job security; while the latter resent the former for the additional service work and so on required of themselves as colleges and universities whittle down the number of traditional professors available to do that work. Is it possible that, along the lines described by Graeber (2018), faculty, in general, resent those in the bureaucratic behemoths that have usurped their control over institutional governance, while at the same time the latter resent the former for their direct and salient contributions to their institutions’ core missions?

If there is any truth to this ‘moral envy’ hypothesis of Graeber’s (2018, p. 176), then we hope the research presented above makes a useful, if small, contribution to dismantling the balance of resentments that foster administrative bloat and related issues of concern in higher education. While we have foregone any policy recommendations in this paper, it would seem, at heart, that a first step toward improving the organization and governance of our institutions of higher education would require an honest and careful appraisal of what the employees actually do, no matter how malodorous the inquiry might ultimately prove to be.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that supports the findings of this study are openly available in the Scientists and Engineers Statistical Data System at https://ncsesdata.nsf.gov/datadownload/. Additionally, the code used to generate the results presented in this paper are available from the authors at [email protected].

Additional information

Notes on contributors

Michael Delucchi

Michael Delucchi is an Independent Researcher and retired Professor of Sociology. He received his PhD from the University of California - Santa Barbara. His areas of research and writing include sociology of higher education, organizational sociology, and the scholarship of teaching and learning.

Richard B. Dadzie

Richard B. Dadzie is an Assistant Professor of Economics at Seattle Pacific University. He received his PhD from the University of Missouri - Kansas City. He is a heterodox economist with interest in higher education, development economics, and international tax policy.

Erik Dean

Erik Dean is an Instructor of Economics at Portland Community College and researcher at the Global Institute for Sustainable Prosperity. His core expertise is in heterodox production theory, institutionalist methodology, and pedagogy in economics. His recent research covers a range of topics, including the nature of the modern occupational structure and the place of the corporation in money manager capitalism and the ramifications thereof.

Xuan Pham

Xuan Pham is an independent researcher. She is interested in using microdata and data science to study labor economics, economic sociology, and the social provisioning process.