Simon (1978): Bounded Rationality and Decision-Making in Organizations

Humans Don’t Optimize — They Satisfice

Classical economics rests on a heroic assumption: people are perfectly rational. They gather all available information, consider every possible option, calculate the expected outcome of each, and choose the one that maximizes their utility. It’s a beautiful model — and Herbert Simon thought it was nonsense.

Not because people are stupid, but because the world is too complex. Information is incomplete. Time is limited. Our brains have finite processing power. In the real world — especially inside organizations where most economic decisions actually happen — people don’t optimize. They look for solutions that are good enough. Simon called this satisficing (a blend of “satisfy” and “suffice”), and the broader framework bounded rationality.

This wasn’t just a minor correction to economic theory. It was a fundamental challenge to its foundations — and it opened the door to behavioral economics, organizational theory, and artificial intelligence.

The Problem with “Economic Man”

Standard economic theory models humans as homo economicus — perfectly rational agents with unlimited computational ability, complete information, and clear preferences. This “economic man” always finds the optimal solution.

Simon identified three reasons why this is unrealistic:

• Information limits: You can’t know everything. Gathering information is costly and time-consuming. Most decisions must be made with incomplete data
• Cognitive limits: Even with perfect information, the human brain can’t process it all. We can’t evaluate millions of options simultaneously or calculate complex probabilities in our heads
• Time limits: Decisions must be made under deadlines. The world doesn’t wait while you compute the optimal answer

The result: real decision-makers use heuristics — rules of thumb, shortcuts, and simplified models — not global optimization.

Satisficing: Good Enough Is Good Enough

Simon’s alternative to optimization was satisficing. Instead of searching for the best possible option, a satisficer sets an aspiration level — a threshold of acceptability — and chooses the first option that meets it.

Looking for an apartment? The optimizer would visit every listing in the city, rank them all, and pick the best. The satisficer decides: “I need two bedrooms, under $1,500, within 30 minutes of work” — and takes the first place that qualifies. In a complex world with search costs, satisficing is often more rational than optimizing.

This insight applies powerfully to organizations:

• Firms don’t maximize profit in the textbook sense — they set targets and adjust behavior when targets aren’t met
• Managers use routines and rules rather than recalculating from scratch for every decision
• Organizations develop standard operating procedures that work well enough most of the time, and only change them when performance falls below acceptable levels

Administrative Behavior: How Organizations Really Decide

Simon’s 1947 book Administrative Behavior — written when he was just 30 — revolutionized organizational theory. He argued that to understand economic decisions, you must understand the organizations in which they’re made.

Key insights:

• Organizations shape rationality: The structure of an organization — its hierarchy, information channels, division of labor — determines what information reaches decision-makers and what options they consider. Rationality is not just bounded by individual cognition but by organizational design
• The authority relationship: People in organizations don’t just follow their own preferences — they accept the authority of the organization, following instructions and procedures. This is fundamentally different from the autonomous agents of economic theory
• Identification with organizational goals: Employees internalize the goals of their department or unit, which simplifies decision-making but can create conflicts between subunits — each optimizing locally while the whole organization suffers
• The distinction between programmed and non-programmed decisions: Routine decisions can be handled by rules and procedures; novel decisions require judgment, creativity, and search. Organizations must design different processes for each

Beyond Economics: AI and Cognitive Science

Simon’s intellectual range was extraordinary. His work on bounded rationality led him directly into two other fields:

• Artificial intelligence: With Allen Newell, Simon created the Logic Theorist (1956) and the General Problem Solver — among the first AI programs. Their approach modeled AI on human problem-solving: heuristic search, means-ends analysis, and satisficing rather than brute-force optimization. Simon received the Turing Award in 1975 for these contributions
• Cognitive psychology: Simon studied how chess players, scientists, and managers actually think — using protocol analysis (having people think aloud while solving problems). He showed that expertise consists largely of pattern recognition built from experience, not raw computational power
• The science of design: Simon argued that the proper study of human-made systems — organizations, computers, economies — requires a “science of the artificial” distinct from natural science, focused on how things ought to be rather than how they are

Why It Matters

Simon’s bounded rationality didn’t just critique economic theory — it opened entirely new research programs:

• Behavioral economics (Kahneman, Tversky, Thaler) built directly on Simon’s foundation, studying systematic biases in human decision-making
• Organizational economics took seriously the idea that firms are not black boxes but complex decision-making structures
• Mechanism design asks how to structure institutions so that boundedly rational agents make good decisions
• AI and machine learning continue to grapple with the trade-off between optimization and satisficing that Simon identified

His 1978 Nobel Prize was awarded “for his pioneering research into the decision-making process within economic organizations.”

Explain It to a Child

Classical economics says people are like supercomputers — they always find the perfect answer. Simon said: no, people are more like someone shopping in a huge store with no map and only 10 minutes. You don’t check every single item to find the absolute best deal. You walk down a few aisles, find something that looks good and fits your budget, and you buy it. That’s not irrational — it’s the smartest thing you can do when you don’t have unlimited time and brainpower. Simon called it “satisficing” — finding what’s good enough.