ACT II: THE MACHINE — How the system works (or doesn't)
To discuss this, I'd like to establish certain prior conditions as given.
Imagine the company ran a selection process according to its standards, and during the worker's probationary period there was proper follow-up, with appropriate feedback sessions that resulted in the worker's definitive and satisfactory incorporation.
With these conditions assumed, let's move to the matter at hand.
For an infinity of reasons and variables, there are workplace situations that result in deficient performance from certain workers. In some cases, these situations repeat exponentially, generating ever-greater impact on both the operation and the work climate. Let's also assume at this point that the company has monitoring mechanisms in place and takes measures to improve performance or modify behaviours harmful to the team.
At this point I ask myself: what more can be done?
Faced with this question, many companies seem to make a mathematical-financial-accounting calculation that concludes a dismissal is nothing but loss, and this reasoning typically ends up eliminating this option from the possibilities. The result of this decision—or rather, this lack of decision—generates a series of consequences, none of them positive.
"Since I don't want to fire you, I'm going to make you leave on your own..."
If this phrase doesn't represent certain companies, it comes bloody close.
In these cases there seems to be a psychological battle between the worker and their superior(s). As a first measure, they try to separate the worker from the operation—and it makes sense (eh?), since they're not doing the job well. The more they separate them from the operation, the more inoperative they become, and the more unsustainable working with that person becomes. Yet they keep them in the position, because the objective is for them to leave voluntarily... (sounds pretty sick, no? You don't know any cases like this?).
I ask myself: what do they expect to happen other than everything going to shit and the situation getting worse instead of better?
What image does this create for the rest of the workers when these situations arise?
What's the operational cost of a worker, in whatever position, performing at the minimum possible level?
How much work gets distributed among other workers because of such performance?
Can't these consequences, and many others, be translated into an economic cost for the operation?
Finally... is it really possible that after running a serious selection process, proper probationary follow-up, the necessary coaching or feedback to correct possible deviations—you arrive at the point where a worker is good for nothing?
Could it be that all those previously assumed conditions are false?
And if the answer is that they're true, wouldn't it be better—for the good of both parties—to pay the dismissal and put an end to all this madness?
The Taleb Connection
Every "let's not fire them, it's expensive" calculation is the same trade inverted: the visible cost of severance is small and immediate; the invisible cost of keeping a non-performer is large and accumulating. The article walks through the latter — operational drag, work redistributed onto others, the slow corrosion of the climate, the cynical theatre of trying to make someone "leave on their own."
"Iatrogenics, literally, 'caused by the healer,' iatros being a healer in Greek." — Nassim Nicholas Taleb, Antifragile (2012), Book VII, Ch. 21
The pattern is iatrogenic. The manager avoids the one visible, traceable harm (a dismissal that shows up on a balance sheet) and prefers the distributed, invisible, unblameable harm (the team carrying a corpse for months). Taleb's repeated point in Antifragile's Chapter 21 is that the healer who fears the visible side-effect more than the actual disease is the one who should worry you. The same applies here: the dismissal is the side-effect; the situation is the disease.
The honest question at the end of the article is the test for managers: would you tolerate this if you had to work alongside it every day?
This article was translated from the 2012 Spanish original and revised in 2026 through human-AI collaboration — clarifying transitions and adding connections to Nassim Taleb's framework.