A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Bounded rationality, introduced by Herbert Simon (1955), holds that human decision-makers have limited cognitive resources — constrained information, computational capacity, and time — and therefore cannot achieve the perfect optimization assumed by standard economic models. Instead of maximizing utility by evaluating all alternatives, boundedly rational agents "satisfice" — they search through options sequentially and accept the first one that meets an aspiration level. This is not irrationality but a different kind of rationality adapted to real-world constraints. Bounded rationality is the foundational concept of behavioral economics because it motivates the entire research program: if people are not perfect optimizers, then understanding how they actually decide requires psychological investigation, not just mathematical axioms.
Standard economic theory rests on a powerful assumption: people are rational utility maximizers who evaluate all available options, correctly assess probabilities, and choose the alternative that maximizes their expected utility. This assumption makes models tractable and generates precise predictions. But Herbert Simon, an economist and cognitive scientist, recognized that this portrait of human cognition bears little resemblance to how people actually decide.
Simon's insight was not that people are stupid or erratic but that real decision-making operates under constraints the standard model ignores. Information is costly and incomplete — a shopper does not know the price of every product in every store. Computation is limited — a chess player cannot evaluate every possible sequence of moves. Time is scarce — decisions must be made before the opportunity passes. Under these constraints, exhaustive optimization is impossible, and decision-makers adopt strategies adapted to their limitations.
The key alternative strategy is satisficing. Rather than searching for the best option, a satisficing agent defines an aspiration level (a minimum acceptable outcome) and searches through alternatives until finding one that meets it. A job-seeker does not evaluate every available job in the economy; they search through opportunities and accept one that is good enough on salary, location, interest, and other criteria. The aspiration level itself may adjust over time — if the search is easy, aspirations rise; if it is difficult, they fall. This is a reasonable, adaptive strategy that economizes on cognitive resources while usually producing adequate outcomes.
Bounded rationality is foundational to behavioral economics because it motivates the central question: if people are not optimizing, what are they doing? This question led to two major research programs. The heuristics-and-biases program (Kahneman and Tversky) documented systematic departures from rational choice — people overweight vivid information, anchor on irrelevant numbers, and evaluate outcomes relative to reference points rather than in absolute terms. The ecological rationality program (Gigerenzer) argued that simple heuristics are often well-adapted to the structure of real environments and can outperform complex optimization when the environment is uncertain and information is limited.
The practical implications of bounded rationality extend far beyond academic economics. If consumers are not perfect optimizers, then market outcomes may differ from what standard models predict. If employees satisfice in job search, labor markets may not clear efficiently. If voters use heuristics rather than fully evaluating candidates, democratic outcomes reflect a different kind of "rationality" than political theory assumes. Bounded rationality does not tell us that people are making bad decisions — it tells us that understanding their decisions requires understanding their cognitive processes, not just their objectives.