Iron Laws and Moral Mazes

Placeholder for consideration of Moral Mazes, principle agents problems, collective action problems, Vast Bureaucracies etc in modern large organizations.


Zvi’s Moral mazes summary again:

What has made these mazes so much more powerful than in the past?

Some factors are real and inevitable. We need more large organizations for our civilization to function, than did prior civilizations, and in many ways they get to better leverage big data. machine learning and the internet, giving them an edge. As we grow safer and wealthier, our demand for the illusion of security rises, and mazes are relatively better positioned to provide that illusion.

There are many other reasons that are less real, and less inevitable. We protect organizations from disruption, especially in times of crisis. We see rent seeking of all kinds as increasingly legitimate. Mazes have gotten sufficiently powerful to cause a vicious cycle, as mazes reward and support other mazes and structure things to favor mazes. Our laws and regulations favor mazes over non-mazes, far beyond what is necessary due to civilizational complexity. Our educational system trains people for the maze, so much so that the people have largely forgotten what mazes are and what the alternative to them might be. We have been so atomized, and their ordinary human needs so delegitimized, that we do not see what we are giving up.

What might we do to change things for the better? Regulatory reform, health care reform, tort reform, ending corporate welfare or even forcibly breaking up large corporations would help. So would educating people on what mazes are and the dangers they pose, especially to their employees. We could work to change consumer behavior, to lower the status and aura of legitimacy of mazes and those who work for them. And we could work to lower demand for the illusion of security.

For a given project, the best defense is to focus on the core elements, and thus do less things and be smaller, while minimizing interaction with other mazes. One should also seek separately to minimize levels of hierarchy, provide skin in the game and soul in the game, and be extremely careful with people. Hire, promote, evaluate and fire them with a keen eye. Anyone making you more like a maze needs to go, no matter how painful that is. You must fight for your institutional culture.

It is hard to do quantitative research on this and easy to rant/speculate/complain, so there are lot of essays on this theme with slightly different irreconcilable and dubiously verified models.

Observations on these themes are not new; See the Robert Michels’ 1911 Iron law of oligarchy.

Here are some more.


The Gervais principle, which divides people into archetypes: Sociopaths, Losers and Clueless. The major dataset for this study is the American version of The Office, with all that entails. This model is fun to read but more stilted and stylised.

Same category, maybe, is David Graeber’s theory of ‘bullshit jobs’:

It’s as if someone were out there making up pointless jobs just for the sake of keeping us all working. And here, precisely, lies the mystery. In capitalism, this is precisely what is not supposed to happen. Sure, in the old inefficient socialist states like the Soviet Union, where employment was considered both a right and a sacred duty, the system made up as many jobs as they had to (this is why in Soviet department stores it took three clerks to sell a piece of meat). But, of course, this is the sort of very problem market competition is supposed to fix. According to economic theory, at least, the last thing a profit-seeking firm is going to do is shell out money to workers they don’t really need to employ. Still, somehow, it happens.

welcome to my org chart

There is a whole spin-off book about that now, although I have not read it, just the original article. The invective is fun, as far as being fun to complain about like the Peter Principle or one of those other semi-humorous texts that self-classified renegades like to imagine is the problem with those pencil-pushers at the office. I am not sure it is helpful though. Self classified renegades often find themselves vulnerable to thinking systemic mayhem is soluble using off-the-shelf business leadership.

Clay Shirky, The Collapse of Complex Business Models:

Bureaucracies temporarily suspend the Second Law of Thermodynamics. In a bureaucracy, it’s easier to make a process more complex than to make it simpler, and easier to create a new burden than kill an old one.


SMBC model of organisational evolution

Yudkowsky’s Inadequate Equilibria is an attempt to turn this around. Rather than arguing that Moloch is inevitable, it suggests ways to identify low-hanging fruit that might be plucked despite the tendency to Molochian gridlock. Sometimes even someone acting alone might pluck such a fruit. Zvi’s commentary on chutzpah has been commended to me as substitute replacement for the chapters 6 and 7 of that book.

Perhaps it would be better to consider crack this egg from a different perspective, not by looking for a universal theory of how humans can be suboptimal in their organisations, but more by wondering how we could design better mechanisms to avoid organisational waste?

But also, too much emphasis on unilateral individual renegades is, I reckon, dangerous.


  • Busy Simulator

  • What We Learned About Bureaucracy from 7,000 HBR Readers

  • Yes, You Can Eliminate Bureaucracy

    …the most promising solution may have emerged in an unlikely place: the world’s largest appliance maker, Haier. Under a renegade CEO, it has been divided into 4,000 self-managing microenterprises. About 250 are market facing (“users”), and the rest (“nodes”) supply them with components and services like IT and HR support. Users can hire and fire nodes—or contract with outside providers—as they see fit, and nodes’ revenues are tied to their users’ success.


Graeber, David. 2014. What’s the Point If We Can’t Have Fun? The Baffler.
———. 2018. Bullshit Jobs: A Theory. 1st Edition. New York: Simon & Schuster.
Hubbard, Douglas W. 2014. How to Measure Anything: Finding the Value of Intangibles in Business. 3 edition. Hoboken, New Jersey: Wiley.
Jackall, Robert. 2009. Moral Mazes: The World of Corporate Managers. Updated edition. Oxford ; New York: Oxford University Press.
Lorenz, Chris. 2014. Fixing the Facts: The Rise of New Public Management, the Metrification of ‘Quality’ and the Fall of the Academic Professions.” Moving the Social 52: 5–26.
Prasad, Monica. 2022. Proto-Bureaucracies.” Sociological Science 9 (September): 374–405.
Taberner, Andrea Mary. 2018. The Marketisation of the English Higher Education Sector and Its Impact on Academic Staff and the Nature of Their Work.” International Journal of Organizational Analysis 26 (1): 129–52.

No comments yet. Why not leave one?

GitHub-flavored Markdown & a sane subset of HTML is supported.