3 ms·
> Doesn't this undermine the argument that CEOs need outrageous salaries to be incentivized to work, if you slash their salary, and they keep working? No, for
by smallgovt 10y ago
> Doesn't this undermine the argument that CEOs need outrageous salaries to be incentivized to work, if you slash their salary, and they keep working?
No, for a number of reasons.
1) His alternatives today could be substantially worse than when he started. If for no other reason, because of this mistake.
2) There may be large switching costs to changing jobs, and the reduction in salary may be non-substantial in comparison. This switching cost was a sunk cost when he initially joined, but will not be if he chooses to switch now.
- YPCrumble 10y agoBoth of your points are symptoms of the problem that CEOs are vastly overpaid. People are too picky choosing CEOs because they falsely believe that CEOs' high pay correlates to a large difference in performance from regular peasants. The truth is that many, many people could do a CEO's job. Everyone makes mistakes, this CEO's colossal mistakes are case-in-point. If we stopped falsely believing that CEO's are magical superhuman lords in a world of serfs we would reduce their pay. Switching costs would reduce dramatically. So would our intolerance for mistakes that are naturally part of the human predicament.
- smallgovt 10y ago> Both of your points are symptoms of the problem that CEOs are vastly overpaid. How so? I fail to see how even if he was overpaid, either of my points are caused by (i.e. are symptyoms of) him being overpaid. > People are too picky choosing CEOs because they falsely believe that CEOs' high pay correlates to a large difference in performance from regular peasants. The truth is that many, many people could do a CEO's job. Everyone makes mistakes, this CEO's colossal mistakes are case-in-point. Based on what evidence? To make this conclusion, you need to compare the quantity and quality of mistakes made by current CEO's and would-be "peasant" replacements. I find it hard to believe you have a factual basis for estimating the types of mistakes your supposed replacement would make. If you better understand what these CEO's do, I think you'll find the # of ppl who can ACTUALLY do their job is quite small, and it certainly doesn't include "regular peasants" or "many, many people". Furthermore, these companies have hug market caps. If the optimal CEO generates 1% more company value than his best alternative, it's more than worth the tens of millions he's being paid.