The Iron Law of Bureaucracy (2010) | Hacker News Reader

The Iron Law of Bureaucracy (2010)

jerrypournelle.com

67 pointsBluestein44 comments

At several big companies I worked at HR was at the center of the second group. They had no concept of what the organization "did" as far as product output and so on. As far as they were concerned they were the organization.

At one company we had a couple round of layoffs, thus there were more demands on everyone's time. HR then would drag us into "mandatory not required" meetings where they would waste a week of our time teaching us about "stress management" from people who as far as we could tell spent much of their time in the breakroom.

One such HR person even went so far as to file a complaint, with HR, that the team that sat next to them in the office ... weren't very social / didn't attend social events put on by HR.

I never know what to do with these kind of people... I laboriously explained to a few what other people did, but I mostly just insulate myself from them so it isn't as frustrating.

People make fun of the term "justice involved individual" but this is exactly the same dynamic. You are at your safest when you are as far as humanly possible, physically and organizationally, from their orbit. I'm safe from HR not because I'm innocent (like that matters) but because none of them even know who I am.

Interesting to learn about that Iron Law. Though what I find missing in the mini blog post is an explanation why the second group ("the pure institution for institution sake" mindset) tends to take over the organisation.

Is it because they are more invested in the organisational/administrative tasks versus the first group's actual goal archiving mindset ? As these goals of the first group lie outside the organisation, the simple distance of goals within vs without might be one significant part. E.g. easier to achieve something within then outside, hence more power over time.

> why the second group ("the pure institution for institution sake" mindset) tends to take over the organisation.

Because they can focus 100% of their time and effort on controlling the organization, whereas the first group can only focus the time and effort on that that they can spare from getting the actual work done that they are in the organization to do.

Makes sense. I'm wondering now how a "safeguard" from this behaviour could look like. E.g. make a person's promotion within an organisation depending on achieving also "first-order" goals of the organisation.

The problem is that the institutionalists define "first-order" safeguards. They also often control key relationships like access to the board of directors

The right and thorough answer to your question would involve psychology, psychology and evolution.

I would say that it’s because their job description and their KPIs are tied to it. HR is measured for doing HR things. HR’s sole purpose is to do HR things. The department becomes a sort of paperclip maximiser that is not concerned with things like the company’s goals.

This definition leaves out those in a bureaucracy who are only devoted to themselves. I assume since their personal devotions may or may not have to do with the goals of the organization or the organization itself that their impact cancels out.

> This definition leaves out those in a bureaucracy who are only devoted to themselves.

Indeed that's a frequent specimen, along with what I would call "devoted to nothing" - simple freeloaders, there for the ride, just getting by through the law of minimal effort ...

This is a good observation but one can argue that those devoted to themselves are a subset of those devoted to the organization (to further their goals, they act as being part of the second category, since that's the most profitable)

If you're being paid by the organization, you're being paid for something. Either for work that accretes to the goals of the organization, work that accretes to the organization itself, or some combination of the two

That's a fine interpretation, but the observable evidence lately would indicate otherwise. The two terms to look into are "institutions as platforms" from Yuval Levin at the American Enterprise Institute [0], describing congressional representatives who spend more staff budget on comms than policy and more of their time on cable news hits than on committee meetings. The second is "public choice theory" economics which is the branch of economics describing why policy doesn't align with stated goals on a regular basis. Robin Hanson had a good substack on the second one just today, in fact [1].

[0] - https://www.npr.org/2020/01/30/800922222/when-institutions-a... [1] - https://www.overcomingbias.com/p/the-big-econ-error

Yes, when I was doing work for the UN there were three categories of people in the organization and the lines between them were quite bright. I've described it this way to many people who are curious.

You had the true believers in the mission, which was a minority. You had an army of bureaucrats, who were solely motivated by a steady paycheck and were incentivized to be paperwork and meeting maximalists to justify their lifetime appointment. And then you had a large minority -- larger than the true believers -- of grifters and sinecures, typically a nephew of the president of some undeveloped country or similar, who were openly there to collect as much free money as possible without even the pretense of work.

That last group, while obviously devoted to themselves, were devoted to the organization to the extent that they are parasites and the organization is the host. But they only need the organization to exist, not the organization's mission.

It should always be assumed that everyone participating in a bureaucracy is selfishly motivated.

The difference is some people see a path to personal success by creating shareholder value (better product, less cost, etc.), while other people focus on fortifying their current arrangement to ensure it can't be made worse.

Hand written html, no CSS, #FF0000 font color AND egregious use of gifs? And not a single piece of social media?! Now THIS is a what website should look like! Maybe I am just getting old but I miss the Good Old Days of the web.

Jerry Pournelle was a Sci-Fi writer, but he also consulter for the US Department of Defense and was a longtime columnist for Byte Magazine.

So I've been learning about expander graphs, and it seems like they preserve some of the good things about hierarchies (low fanout) while being very unlike hierarchies in other ways (higher # of paths, lower # of hops, between an arbitrary pair of nodes).

Have any HBR types done a case study of non-traditional firm organisation, along expander lines?

https://www.amazon.com/Team-Topologies-Organizing-Business-T...

Pace Wigner (at this point he deserves a “talisman” — FMD might agree now), those expanders are indeed unreasonably effective, just like bureaucracies (according to the ever optimistic Kafka). Emergent intelligence in governments and institutions, may not be just a Good Idea? (Abuse of Gandhi/de Maistre, sorry)

> Let me end on a more cheerful note. The miracle of the appropriateness of the language of mathematics for the formulation of the laws of physics is a wonderful gift which we neither understand nor deserve. We should be grateful for it and hope that it will remain valid in future research and that it will extend, for better or for worse, to our pleasure, even though perhaps also to our bafflement, to wide branches of learning.

EDIT: FrankWilhoit has some sayings regarding how, in order to reduce (Clausewitzian?) Friction, we should look at glue (interclique edges?)and not just syntax (cliques/juntas?) like what planners do. (The novelty of the idea seems to me to be in that planners usually dont attend to the strycture of egdes as much)

It just occurred to me that it's entirely fitting how often Kowalski refers to the Caudine Forks, for that battle is a metaphor for ping pong arguments:

Suppose we have two statements (Romans), A and B.

A priori, there are four possible situations:

    A & not B | A & B | not A & not B | not A & B

However, as mathematicians, let us play the Samnite. First we close one pass, in proving A → B, which is equivalent to ruling out A & not B. Then we close the other pass, in proving B → A (ruling out not A & B).

Now our romans, err, statements are confined to A & B | not A & not B, and (with no truth values having been harmed) must therefore pass under the yoke of equivalence: A ↔ B.

Note: algeria 1830s https://baheyeldin.com/writings/culture/how-the-dutch-got-th...

Thanks for application of irony to a thread about iron laws! Led me to the “unnamed” Ping-Pong Lemma, a great counterexample!

Vespasian was, apparently, more inconsistent: https://sci-hub.wf/10.2307/1192680

(Im guessing he was as aware as MAA of the burdens of office)

Hopefully we can return to the formalization of Nietzsche/Greco-Roman ethics/(New) systems of survival soon. State of my art, how to extend Boyd’s Ping-Pong Lemma* to Rao’s SoS (assuming it already works in JJ’s system)

Maybe the eternal Culture War is indeed the Russell one (Banks’ having his Culture win the Mandate of H) . Hey, JAW was concerned too, we are now on the same hunt as the Masters?

https://johnhorgan.org/cross-check/physicist-john-wheeler-an...

https://mindmatters.ai/2021/05/it-from-bit-what-did-john-arc...

*the same old, but generalizing boss to adversaries, also noting in passing parallelism of your justice-prudence to guardian/commerce, so temperance-fortitude maps to Blue/Red hypernormal/weirdo??? https://en.wikiquote.org/wiki/John_Boyd

Front page activity indicates that cpython is becoming hypernormal ;) (even if GvR as a concept remains a Wojak)

> becoming hypernormal

becoming -> being, typo, or rather, forgot my own earlier classification

that reminds me, you might be interested in Prigogine, From Being to Becoming: Time and Complexity in the Physical Sciences (1980).

Lagniappe: https://xkcd.com/2904/

I did a cyrsory examination of the genealogy a while back due to the events.. Ilya was probably quite related to.. Yevgeniy (verify?)

on a quick Boyd tangent, the japanese three pwns (三殺法) easily map into Boyd's OODA:

  kill the sword - you are observing your opponent's actions
  kill the technique - you are observing your opponent's decisions
  kill the spirit - you are observing your opponent's orientations

Not enough spare cycles to chew on that yet, just pointing:

To buy time:

weirdo -> peripherally oriented

hypernormal -> centroidally oriented

(Neither really observing the states of opposing camp beyond 1st pwn level, though the hypernormal may regard the other hypernormals as memetic rivals [the individuation in the Apolline])

Linux kernel devs —> … ?

(Those customs inherited by crypto coredevs, hence also Yarvin, by diffusion, ..)

Edit: something reminds me of the threeway struggle between the RI, the klepts, & the Met in Gibson’s trilogy+adaptation, which is not cleanly an indo-aryan division, as the Met is clearly the 3rd estate yet a warrior caste, while the klepts are merchant princes..

https://www.penguinrandomhouse.com/series/TJ3/the-jackpot-tr...

Re: "it from bit", the MGM view is that most "it"s arise because the universe is full of interactions with negligible energy but decohering phase.

(which physicist said "better to be wrong in an interesting way than right in a boring one"? I guess that was an early weirdo>>hypernormal slogan?)

I'm closer to Apache devs than Linux devs, so using them as a proxy I'd guess the Linux devs to be closest to our earlier-proposed third ("hacker") SoS (though not as close as Knuth's squad, who are probably the nonfictional crew having the highest dot product with Castalia of which I am aware).

In particular, the geek ethic (a peripheral center?) values prowess and openness simultaneously...

[I'll have to read the Jackpots; tarnishing my geek cred I'm only up on Gibson through Neuromancer. On the jock front it's a tournament weekend for us, so my HN b/w will be limited.]

For quick catchup, watch the amazon prime video at 2x. some elements not in the original but imho the screenwriters surpassed Gibson in exposition. casting almost perfect, exactly how i had imagined it. my guess is in the upcoming WG himself will retcon the RI in some form, heh. (If anything, channelling some of the klept’s villainy to academe makes more sense today)

“In contrast for China, nature means spontaneity, therefore the future is not completely predictable.” —Ilya Prigogine.

Reportedly (could be in the 1980 book? To check later) he also said that the best way to cope with the future is to create it, but he didnt say that you could be thoroughly relaxed about it.

Physicist — its tegmark or weinsteins advsior, again, to verify later..

Pardon the interruption of riverian activities, also i understand the b/w concern, yet this is kinda impt (so will repeat later)

[One might have to selectively apply MGM amnesia to MGM’s own teachings.. ]

Aaronson-Carroll’s quest (and indeed Kolmogorov, Prigogine, AND MGM’s ultimate quest (ergodicity econ))

  is to move BEYOND what you just quoted of MGM.

(So that old teaching which you had for decades kept in cache is an instance of MGM being correct in a boring way)

(you might have been paying attention to my attempts to draw out experts recently?)

[i get why experts on quests prefer to stay silent on HN, so, dabblers, blowhards, and cranks are acceptable]

Makes sense — I shan't feel so bad about lack of b/w then, because if the Santa Fe Institute hasn't cracked it in four decades, the chances I'll have anything worthwhile to add in a weekend are probably on the order of IEEE epsilon?

(Is Stuart Kauffman's "poised realm" in scope?)

Is this right: sophistication is proportional to the smallest proper essence, where an essence is proper iff K (essence x + accident x) <= K x ?

Your conjecture, even if incorrect, certainly reminds one of Jensen’s inequality does it not :) ?

Compare with Aaronson-Carroll-et-al p5 properties of c-sophistication..

Prigogine 1980 Ch5, peddling copium in the form of Lyapounov functions to the Chinese ? (Wouldnt have found that without your tip!)

(Personally, i increased my weights for IRP having stolen Dennis Gabor’s quote, especially since it was re-liberated by an actuary)

All mutterings from SFI are in scope, haha

How have kernel devs (lkml data) or apache devs (your anecdata) been (ab)using the Fehmi principles? Any new insights from your taykh (fishing?) escapades?

[Swiss hydrogen storage using iron as catalyst keeping kosher, but that stainless steel tank isnt lol]

hmm, I'd interpreted the jade taijitu as book-ending the Brusselator pics at the start of the chapter?

(are sine waves nothing but imaginary spirals?)

Ahh i wanted to refer you to the region in Ch7 where the word “cope” was located.. but Ch5 seems to be a better targeted ad.

Its a barbarian taijitu though.. so nominally Han the sword would have been north-made

>imaginary

yepp you can substitute virtual for imaginary

Ahh apologies.. addresser was not actually an actuary but uh a Librarian ???

(Lagniappe: attendee could be a wiktionary-approved contronym)

https://www.soa.org/4934ee/globalassets/assets/library/proce...

Possibly Hessean (Nietzschean, even), that swimming in Der Teich is quite contrary to survival..

(But not “immortality”?)

If immortality were an available option, I'd expect to see Sumerian hedge funds still trading; at some point any particular coarse-grained institution winds up stuck in an oxbow lake?

(sure, conservation holds at a fine-grained level, but is traversing a final-initial edge distinguishable from absorption?)

Sorry im kinda still rate limited ..

10)virtual particles have a real effect (re conservation laws, noether, etc)

01) note that mints and mines are pretty much immortal (on the human, if not geologic scale, not even needing basterdized primogeniture?

[ Emmanuel Todd has some theories on how inheritance culture is why old country needs to choose between slavics and teutonics.

https://en.wikipedia.org/wiki/After_the_Empire ])

i seriously doubt anyone has figured out to study how coarse grained institutions should behave, if no professional has even thought of seriously applying expanders to study how they already behave..

TIL the inspiration for Zazie at the pot de thèse: https://en.wikipedia.org/wiki/Zazie_dans_le_Métro_(novel)#Pl...

(with pots de thèses being mentioned in https://www.lix.polytechnique.fr/Labo/Arnaud.Dartois/these/P... : « Et après, il y aura un super goûter avec tous mes amis. Génial ! »)

[Will reply to the other one abt soph after i get more b/w myself]

Well, if your not being facetious— you should be familiar with the recurring institutional failures in the old country? MGM, Leggett, Geroch might have been fine in their salad days but as they got on, nothing really diffused to their squads.. As it is, Ole Peters is having difficulties defending himself against serious charges (from the orthodox econs) of unfamiliarity with Jensen’s inequality..

The TCS horde is routinely cleaning out the institutionalized physicists these days, from what i can see its due to effective horizontal virtue transfers from the european/pure math community. Wont be long before the econs suffer the same, seems like they are walled up in the same old way.

it does look like UK physics education today is in pretty ok shape compared to the others. i will assume from now on that the Research Institute (featured in the amazon adaptation) was created by american refugees.

All Bureaucratic Structures eventually become self supporting. Religion, business, government, non-profits, industry groups, standards bodies, everything.

Bureaucracy isn't inherently bad, but it is a dragon that everyone should keep a very watchful eye on.

I wonder whether the iron law of bureaucracy could be a special case, or maybe related to, the iron law of oligarchy by Robert Michels.

Not sure this counts as 'a law' in scientific terms.

Why would it have to?

As soon as I finish my scientific investigation into Murphy's Law, I'll get right on it...

Reply on news.ycombinator.com