The Bureaucratically Boiled Frog

[jducoeur]

In the context of an organization I'm involved with (not going to say which one -- really isn't relevant), I got into a discussion of some rules that had recently been put in place, that IMO are unnecessary: they're good guidelines, but shouldn't be rules. After making a characteristically impassioned argument against them, this postscript occurred to me. Since it's really quite general, and I haven't posted a good rant on this subject in a few years, it seems worth posting here, with the serial numbers filed off.

---

To use a metaphor that had never occurred to me before now (but which I will probably use going forward): bureaucracy is like climate change.

Each new rule is, in and of itself, tiny and tolerable, and there's a natural reaction of, "why are you objecting so hard to this?" The problem is that, once put in place and well-established, it tends to be nearly impossible to dislodge those rules: weird though it may seem, it is much, much easier, in every organization I've ever observed, to add rules than to remove them. Adding them tends to be fairly easy; trying to remove them later gets catastrophized, and requires extreme efforts -- many people go to great lengths to rationalize why they are deathly important once they're entrenched.

The result is that they build up in the "atmosphere" of the organization, step by step over the course of years. It's very incremental, but over the span of time, organizations that start out vibrant, cooperative and ground-up inevitably turn slow, stodgy, and top-down, buried under the weight of their own processes.

(None of this is unique to our situation, BTW. These observations were originally formulated from the various clubs and activities I'm involved with, but turn out to apply almost identically to companies, governments, and all other organizations past a certain size. Far as I can tell, any group large enough that consensus is impossible winds up prone to this effect.)

In my experience, this process is inevitable in the long run, but the collective attitude affects how fast it happens. If you can foster a general outlook of "every inflexible rule must be viewed with an acid eye, and must fight for its life", you can slow it down a good deal. That preserves organizational health and vigor for longer, and generally keeps things more fun.

Hence, the passion displayed above -- I like X a lot, and want to keep liking it for as long as possible.

---

Please take this as food for thought, any time that you are in a position of power in an organization. Each rule is small; the collective of them can be crushing. And the more slowly you boil that frog, the longer the organization will tend to remain fun...

Comments

[alexxkay]

Alternately, have the law come with an explicit purpose, a metric for measuring whether it is achieving that purpose, and a regular schedule of checks where, if the metric falls too low, the law ends.

[ckd]

Label all the Chesterton's Fences, and write unit tests? Sounds good to me.

[jducoeur]

Mmm -- I like that. At least for organizations that are already used to metrics, that could be a fairly compelling argument. (And crafting good metrics is enough of a pain in the butt that it may give people pause before writing a rule.)

The downside, of course, is that measuring the metric is itself even more overhead, so one shouldn't go too far overboard...

[squirrelitude]

I like that. :-)

At work we use Architecture Decision Records, which are fairly lightweight documents where we describe why a change is being made. Later, when someone wants to change the decision (or some related code), they can look back and see if the context is still relevant.

[jducoeur]

Yeah, I've been having discussions at work about the optimal use of ADRs. (And how not to turn them into excessively heavy bureaucracy themselves.)