From The Jolly Contrarian
Jump to navigation Jump to search
The JC sounds off on Management

Click ᐅ to expand:

Comments? Questions? Suggestions? Requests? Insults? We’d love to 📧 hear from you.
Sign up for our newsletter.

A side-effect. An iatrogenic illness is not one that the cure is worse than, but that the cure actually causes.

So, Night Nurse™ may cause drowsiness and you shouldn’t drive tractors or operate photocopiers when dosed up on it, but at least it doesn’t make your head cold worse. Many forms of medical procedure can do this: antibiotics, for example, encourage bacteria to develop resistance to the antibiotics making your original problem harder to solve. Antibiotics are, in this way, iatrogenic. Long term, they make your problem worse.

This is why the “six second rule” isn’t quite the careless outrage the helicopter mums of North London imagine. It may be false, but allowing dear little Basil to ingest constant, small, amounts of bacteria off the sausage he just picked up off the lino — rather than nuking young sir’s entire theatre of operations with Dettol before he lays as much as a sticky finger on it — encourages his antifragile body to develop its own immunities to the bacteria, so you don’t need so much Dettol. This is cheaper, too. And makes the sausages tastier.[1]

Popularised by Nassim Nicholas Taleb in Antifragile: Things that Gain from Disorder, one can extrapolate iatrogenics to many other walks of life, in particular those involving service industries, and it is rather fun to do so.

So let’s have that fun!


Much of this effect derives from the rent-seeker’s fascination with the particular over the aggregate: as an abstract matter, it is better to address this peripheral risk here, than ignore it: the cost of addressing it now (say an hour of petulant argument) pails when compared with the consequences later should this risk come about. Let’s unitise this: say the expected cost of this catastrophe, should it happen is one thousand times worse than the cost of dealing with it up front, and its probability of ever happening is one in two thousand.

That those consequences are unlikely is rather beside the point: it’s only an hour, after all. That is a bearable cost well spent to avoid a small chance of a large loss later. At some level of abstraction it’s not a cost at all: it lives within the ebb and flow of human productivity in a day: we are not automata, we do glance up from our stalls and vainly peruse Expedia for last minute flights to Venice every now and then. What’s the odd hour, in the daily sludge?

This is the individual “time-series probability”. It is a trip down the spidery, meandering path of one mortal meat-sack. It is “just an hour” to fix it now against “God knows what” at some point down the line. It is to ask a legal eagle the question, do you feel lucky, punk?

Assume rent-seeking legal eagle, faced with an unchallenging hour earn his own keep, will not feel lucky. He has no interest in feeling lucky. to the contrary: all his personal incentives urge him to feel uncommonly cursed by ill fortune. He will spend that hour up-front every time. It is the prudent thing to do. It is insurance. It will evaporate, unmonitored, into the ether of the daily grind.

But the “ensemble probability” — the same risk spread across a population of meatsacks, plays out quite differently. It scales badly. Say we run this same risk two thousand times. Firstly, it’s harder to blow off two thousand hours as “part of the ebb and flow of the daily grind”. That’s fifty working weeks. It has its own daily grind —in each working day there’s a degree of googling and running off the the bathroom and so on — so call it more like sixty working weeks. And our catastrophic outcome: sure, it’s two thousand times more likely to happen, but since it was only a one-in two thousand shot, it’s still only likely to happen once. And even if it does happen, the loss is only half the cost of preventing it.


This is rather like insurance. For most purposes, insurance is a waste of money — realistically, you are never going to claim on your extended warranty if your toaster breaks down after 18 months because (a) you can’t find it and (b) the damn thing only cost twenty five quid — the bother of having to find the stupid warranty, read it — there is guaranteed to be some exclusion — and actually claim on it is more bother than just shelling out twenty five more quid on a new toaster — of a different brand: screw you, Morphy Richards — and being done with it.

But there are insurance policies that, by themselves, increase the likelihood of loss. Public liability insurance, for example. Directors’ and officers’ liability insurance.

See also


  1. Or is that just me?