On Scaling Back the Enshittification of University Bureaucracy

More than a month ago, I agreed to an offer to be a visiting scholar at a private US university next year. This was no simple matter because of obligations to my own family and (somewhat more unexpectedly) my department. I have not made any public announcement on it yet not because I am especially personally worried by the Trump administration’s policies toward higher education, but because I am still completing (electronic) paperwork and background-checks from the host institution. (It would be bad luck to announce before the process is fully completed. So watch this space.)

Now, by academic standards, I have moved jobs (not always willingly) quite frequently and I have also accumulated quite a bit of visiting positions. I have worked in three different countries and have held all kinds of jobs. So, I am familiar with the great variability in the process by which the (electronic) paperwork for an appointment is completed. When it comes to paperwork before the appointment-process is completed nothing will ever beat Flanders.* But Stateside, I had a rule of thumb that wealthy private institutions are relatively unencumbered by paperwork relative to the state institutions in order to ‘enter’ the system.

Since I had no prior experience with this particular private university and N=1, I shouldn’t make any claims on the basis of it. But since university administrators in the same ecology tend to mimic each other, I would not be surprised if what I am experiencing is part of a wider trend of bureaucratic enshittification [a phrase I am stealing from my friend Tom Stoneham] at US private universities. (I won’t bore you with a graph of the rise of the number of administrators in US universities, but I am not the first to remark on the phenomenon.)

As regular readers know, I have been mulling the significance of Joseph Tainter’s (1988) The Collapse of Complex Societies. A key insight of this book is repeated in the ‘summary’ chapter at the end:

Sociopolitical organizations constantly encounter problems that require increased investment merely to preserve the status quo. This investment comes in such forms as increasing size of bureaucracies, increasing specialization of bureaucracies, cumulative organizational solutions, increasing costs of legitimizing activities, and increasing costs of internal control and external defense. All of these must be borne by levying greater costs on the support population, often to no increased advantage. As the number and costliness of organizational investments increases, the proportion of a society’s budget available for investment in future economic growth must decline. (p. 195)

Tainter is a trained anthropologist not an economist, but he has fully internalized the significance of opportunity costs (and, as the next paragraph makes explicit, marginal returns).

When I first read the quoted paragraph, I read ‘sociopolitical organizations’ as a synonym for ‘state,’ ‘empire’ or ‘society’ (as is clearly intended from wider context). But, upon reflection, one could read the paragraph as scale invariant, and then if one replaces ‘future economic growth’ with ‘future growth in knowledge,’ one has a nice description of the process of bureaucratic enshittification of universities.

Of course, universities are not just about advancing knowledge; they are multi-purpose organizations. But in this they are no different than states/societies/empires. So, the comparison between state and university is not altogether silly. Both require (recall) the art of government in their leadership.

Interestingly enough, it’s pretty clear that many national polities have decided to scale back their investment in universities. And this need not be irrational. For, as Tainter puts it, “investment in sociopolitical complexity as a problem-solving response often reaches a point of declining marginal returns.” (p. 194) For, “as the cost of organizational solutions grows, the point is reached at which continued investment in complexity does not give a proportionate yield, and the marginal return begins to decline.” (p. 195)

Now if we zoom out toward the political scene, we can immediately see that MAGA’s way of scaling back on organizational complexity — breaking and unilaterally changing contracts, punishing perceived political enemies, offering specious justifications, etc. — is exceptionally brusque and in violation of all principles of good governance, but it would be a mistake to treat it as an isolated example. Universities’ budget constraints are getting harsher in much of the OECD, and we should expect, then not just ongoing closing of departments/schools, but also the collapse of whole universities. (The UK’s Higher Education system — which caps tuition — is exceptionally striking example of this process.) So far there is no sign that university bureaucracies/structures have been or will be genuinely decomplexified.

While (recall) I am no friend of the leadership of private higher education and the quality of the educational aristocracy it has produced, I am not defending MAGA higher ed policies because its way of scaling back is so destructive on both the legitimacy and efficiency fronts that it may well undermine societies’ general ability to make the kind of scientific, technological, and social breakthroughs on which the maintenance of social complexity relies. And because MAGA has risen to power (Stateside and elsewhere), in part, through a strategy of sowing distrust not just toward technocratic elites and it is actively destroying non-trivial features of the machinery of government that witnesses truth, makes society legible, and provides society with the public goods that makes all kinds of private transactions possible, they have also undermined their own and their successors’ ability to steer the ship of shape in sensible and cost-effective ways.

Lurking in the background is, alas, AI. Now, the good news is that indeed AI carries the possibility of making the task of information processing faster and more cost-effective in bureaucracies. And let’s stipulate that it allows for greater centralization and fewer layers of bureaucracy where rule-following work is relatively standardized.

But because of the feverish focus on AGI and existential risk and fears of job-loss among white collar and creative labor, it has generally been overlooked that as we move to an AI intensive-society genuine expertise is not made redundant by AI at all as AI displaces the rule-following work even experts often do. For, unless the training and output of AI is monitored by agents that can catch dangerous error, one constantly risks creating junk machines where ‘garbage in’ produces ‘garbage out.’ And if society undermines the sociopolitical organizations that train, certify, and employ experts, then the social benefits of applying AI will not be reaped at all, but society will nudge itself into collapse.+

The previous paragraph suggests, then, that the skillful introduction of AI may well aid society in postponing a dire downward spiral of increasing marginal costs of bureaucracies and other complex systems. But how to get from here to there is by no means obvious even if we had non-corrupt, skillful leadership and could smooth the process with the fruits of economic growth.

As it happens my research proposal centers on the ways in which the division of epistemic labor generates a social demand for a kind of expertise [recall, synthetic philosophy] that helps in navigating the mutual incomprehensibility of the sciences and the black-box systems of AI. So as this post illuminates (I hope), while the electronic paperwork is slowing me down, this slow-down is — thanks to the timely interventions of an administrator/colleague at the private university which is helping me navigate the application process —not preventing me altogether from developing my research plans.

To be continued.

---

*I had to pass an extensive health check (etc.). In Flanders, if one has an unrecognized PhD, one’s dossier also has to be approved by all the other department in one’s field in addition to ordinary paperwork. (One can imagine that does not speed up the process.)

+As Tainter notes that may well be a good thing for many, so here I am not taking a stance on evaluating such collapse