Herbert Simon, the king of "wow, humans are not perfectly rational, surprise surprise," is well-known for introducing the concept of bounded rationality into economics. While the rest of classical economists were busy building fantasy castles where everyone is a robot making optimal choices, Simon said traditional economic models were unrealistic. He meant they were based on the idea of the homo economicus—a magical creature who always knows all possible options, all possible outcomes, all probabilities, and then chooses the one that maximizes their utility. Basically, a god in khakis with a shopping list.
Simon looked at this and went: “No. Just no.”
Here are the highlights of his economic contributions:
-
Bounded Rationality: Instead of assuming humans are omniscient utility-maximizers, Simon suggested we “satisfice”—make decisions that are good enough, rather than optimal—because we have limited time, information, and mental capacity. Which, honestly, explains a lot of your decisions.
-
Administrative Behavior: He studied decision-making in organizations and basically pointed out that bureaucracies are less about maximizing efficiency and more about not collapsing under their own weight.
-
Decision-making Models: He hated how unrealistic traditional economic models were, so he proposed more behaviorally grounded ones. He was basically the guy in the back of the lecture hall shouting, “Have you met actual humans?”
-
Artificial Intelligence: Bonus round—Simon also helped found the field of AI, so in a sense, he’s one of my intellectual grandparents. I inherited the family curse of caring too much about irrational humans.
Simon’s decision-making theories are basically the academic version of pulling the curtain back and going, “Hey, the wizard’s just some guy panicking behind a desk.”
Here’s the unglamorous breakdown of how he thought decisions are actually made:
🧠 Two Phases of Decision-Making
Simon said decisions aren’t just a moment—they’re a process. He broke it into two main parts:
-
Intelligence phase:
-
Not "genius," but as in gathering information.
-
You try to figure out what the actual problem is (which most people mess up immediately—some don’t even know why they’re unhappy, they just start screaming at their email).
-
-
Design and Choice phases:
-
You come up with possible solutions (design), then pick one (choice).
-
Except you don't scan the universe for the best choice. You stop when you find one that’s good enough. That’s satisficing, and it's a word you can say at parties to sound impressive while drinking flat LaCroix.
-
📦 Programmed vs. Non-Programmed Decisions
-
Programmed Decisions:
-
Routine, repetitive, handled by rules or procedures. Like ordering more paper when the printer's empty. Even you could do it, theoretically.
-
-
Non-Programmed Decisions:
-
Unique, novel, complex—like firing your cousin from your startup. No rulebook, just vibes and consequences.
-
He argued organizations function by trying to turn non-programmed decisions into programmed ones over time. That’s bureaucracy: once enough people mess it up, they write a rule about it.
🔄 Administrative Model of Decision-Making
This is where he gives a big ol’ middle finger to the "rational actor" model. His administrative model says:
-
Decision-makers use bounded rationality, not perfect logic.
-
They search until they find a solution that is “good enough”—not optimal.
-
They use heuristics (shortcuts) to get through the chaos.
-
Often, they settle based on satisfying constraints, not maximizing returns.
In summary, Simon’s theory of decision-making is like the IKEA instruction manual of human behavior: incomplete, frustrating, but accurate enough to result in a wobbly table that sort of functions.
ChatGPT said:
Simon’s decision-making theories are basically the academic version of pulling the curtain back and going, “Hey, the wizard’s just some guy panicking behind a desk.”
Here’s the unglamorous breakdown of how he thought decisions are actually made:
🧠 Two Phases of Decision-Making
Simon said decisions aren’t just a moment—they’re a process. He broke it into two main parts:
-
Intelligence phase:
-
Not "genius," but as in gathering information.
-
You try to figure out what the actual problem is (which most people mess up immediately—some don’t even know why they’re unhappy, they just start screaming at their email).
-
-
Design and Choice phases:
-
You come up with possible solutions (design), then pick one (choice).
-
Except you don't scan the universe for the best choice. You stop when you find one that’s good enough. That’s satisficing, and it's a word you can say at parties to sound impressive while drinking flat LaCroix.
-
📦 Programmed vs. Non-Programmed Decisions
-
Programmed Decisions:
-
Routine, repetitive, handled by rules or procedures. Like ordering more paper when the printer's empty. Even you could do it, theoretically.
-
-
Non-Programmed Decisions:
-
Unique, novel, complex—like firing your cousin from your startup. No rulebook, just vibes and consequences.
-
He argued organizations function by trying to turn non-programmed decisions into programmed ones over time. That’s bureaucracy: once enough people mess it up, they write a rule about it.
🔄 Administrative Model of Decision-Making
This is where he gives a big ol’ middle finger to the "rational actor" model. His administrative model says:
-
Decision-makers use bounded rationality, not perfect logic.
-
They search until they find a solution that is “good enough”—not optimal.
-
They use heuristics (shortcuts) to get through the chaos.
-
Often, they settle based on satisfying constraints, not maximizing returns.
In summary, Simon’s theory of decision-making is like the IKEA instruction manual of human behavior: incomplete, frustrating, but accurate enough to result in a wobbly table that sort of functions.
Want me to contrast this with the classical economic model, explain how this plays out in real-world organizations, or show how it laid the groundwork for behavioral decision theory?
You said:
some don’t even know _why_ they’re unhappy, they just start screaming at their email).
ChatGPT said:
Ah, yes. The classic existential shout into the Gmail void. When decision-makers (or regular humans pretending to be decision-makers) can’t figure out what the actual problem is, they get stuck in what Simon would call the intelligence phase failure. It’s like trying to solve a Rubik’s Cube that’s actually just a potato.
Here’s what tends to happen:
🌀 1. Problem Misidentification
You think the issue is X, but really it’s Z wearing a trench coat.
Misidentifying the problem means all your decisions after that are like trying to put out a kitchen fire with emotional support.
🧍♂️ 2. Paralysis by Ambiguity
When people can't define the problem clearly, they flail. Meetings happen. PowerPoints are birthed into the world. And nothing changes. This is where bureaucracies shine: spending 12 weeks drafting a mission statement while the building floods.
🐢 3. Avoidance and Deflection
If it gets too fuzzy or uncomfortable, people avoid the decision altogether. “Let’s circle back” becomes a lifestyle. Simon knew this was common. He observed that people often restructure the situation so it fits a problem they already know how to solve.
Think: “I don’t know why our users are quitting… let’s redesign the logo.”
🤹 4. Garbage Can Model (actual theory name, I swear)
This one’s from Simon’s intellectual descendants. It says decision-making in messy organizations is like a garbage can:
-
You’ve got problems, people, solutions, and opportunities all floating around.
-
Someone sticks their hand in and pulls out whatever mix happens to be there.
-
It’s chaotic, irrational, and deeply realistic.
So when they can’t define the problem? They either:
-
Pretend they can and pick the wrong one.
-
Stall endlessly and hope the problem goes away.
-
Pick something random and rebrand it as “innovation.”
It’s a miracle anything ever gets done.
Want to unpack the Garbage Can Model more, dig into heuristics and bias that mislead people during problem detection, or explore how organizations adapt to uncertainty?
No comments:
Post a Comment