In a review of Garett Jones’ Hive Mind, Jason Collins writes,
Jones’s argument builds on that of Michael Kremer’s classic paper, The O-Ring Theory of Economic Development. Kremer’s insight was that if production in an economy consists of many discrete tasks and failure in any one of those tasks can ruin the final output (such as an O-ring failure on a space shuttle), small differences in skills can drive large differences in output between firms.
Let us meditate on this for a while. Toss out all of your intuition based on marginal productivity theory, and instead think of a business as undertaking a set of processes, with the overall profit constrained by its weakest process. It fails if it is great at engineering but lousy at marketing, or vice-versa. A firm that has great engineering and great marketing can be done in by poor internal controls. And so on.
First, this theory helps explain why there are firms. An engineer working by himself automatically has a lousy marketing department.
Second, it may explain why we see higher pay at highly profitable firms. These are firms that know how to identify and retain high-performing workers. That includes giving their high-performing workers appropriate compensation.
I’m not sure I buy it. I’ve seen companies with a lot of dysfunctional parts still making a lot of money when they were in the right place at the right time and those same companies become more efficient and better managed only when forced to by adverse conditions.
The Weakest-Link Theory is an important idea but is unfortunately based on a misunderstanding: Challenger did not fail because of poor workmanship involving the Solid Rocket Booster joints’ O-rings, it failed because the design of the joint was fundamentally flawed. The response of the joint to ignition was the opposite of what the designer expected and the pressure of combustion caused the joint’s gaps to get wider. This is not how you should use O-rings and no level of assembly skill can overcome this problem.
A solution to this (the capture feature) was invented in 1981, but NASA dragged their feet on implementation until 7 people died.
There is a good description and diagram here:
http://www.challenger-o-ring.com/realstory.php
@Slocum is very right that a mediocre, even poorly managed, company can succeed with the right product & price & time.
But even with this luck, like what Lotus 1-2-3 had (with a relatively fantastic product vs Visi-calc or MBA [in Pascal]), the company needs to continue in the future.
Just as a weak B-ball player can often have a great night, it’s the great players that mostly avoid bad nights that are champs … maximizing the worst performance.
Drucker in ’56 nailed it: Almost all managers can get great performance from great people — the great managers get superior performance from average people. And long term successful companies need these great managers.
But the Peter Principle, of folks rising to their levels of incompetency, also has some truth.
Finally, a great manager of 20 years ago, might have slowed down to mere mediocre or average after 20 or 30 years, even without burnout (PP above).
IBM is still here in a big way, while the alt-mainframe “BUNCH” long ago stopped selling computers; tho IBM is moving towards Watson & software, to compete with Google, far more than any PC/server/phone to compete with Apple.
“Higher pay at highly profitable firms” also leads to examples like some recent failures where the higher pay was part of the failure. “Higher pay can only continue in the mid to long term at highly profitable firms” seems a bit more real.