# Bounded Rationality **The model:** Herbert Simon's foundational claim that real human and organizational decision-making operates within strict cognitive, informational, and time constraints — and therefore looks nothing like the "rational actor" of classical economics. People do not optimize; they *satisfice* — they pick the first option that meets a threshold of acceptability and stop searching. The boundedness is not a flaw to be corrected; it is the structural condition under which all real decisions happen. The intellectual move was decisive for behavioral economics, organizational theory, and cognitive science. Where classical economics asked "what would a perfectly rational agent do?", Simon asked "what does an agent with limited time, limited information, limited memory, and limited computational power *actually* do?" — and then built a research program around that question. ## The Mechanism **Three boundaries.** Simon identified three: limits on *information* (you don't know everything you'd need to know), limits on *cognitive capacity* (you can't process everything you do know), and limits on *time* (you have to decide before you've finished thinking). All three are present simultaneously in essentially every real decision. **Satisficing replaces optimizing.** When perfect optimization is impossible, the rational strategy is to define a threshold ("good enough") and accept the first option that crosses it. Satisficing is not lazy; it is what optimization *looks like* under boundedness. **Heuristics are tools, not bugs.** Mental shortcuts (rules of thumb, stereotypes, snap judgments) are the operating system of bounded rationality. They produce fast decisions of acceptable quality, at the cost of occasional dramatic errors. Whether a given heuristic is good or bad depends on the environment — which is the [[Gerd Gigerenzer]] amendment to Simon (ecological rationality: heuristics that work in the right environment). **Organizations are bounded too.** Simon's work extended naturally to organizations and bureaucracies, which face their own information, cognitive, and time constraints — at scale. Standard operating procedures, organizational routines, and bureaucratic structures are all forms of organizational bounded rationality. They make collective action possible at the cost of inflexibility. ## KB Applications ### Voter Bounded Rationality is Not a Defect Most voters cannot read every candidate's policy platform, read every news story, weigh every consideration, and update on every revelation. Most voters use shortcuts — partisan affiliation, identity-group cues, single-issue salience, source trust, demographic signals. From a classical-rationality perspective, this looks like irrationality. From a bounded-rationality perspective, it's *what voting under cognitive constraint looks like*. The political question is not "are voters rational" but "which heuristics work better than others in the electoral environment, and which ones are exploitable?" ### The Right-Wing Media Ecosystem Optimizes for Bounded Rationality The contemporary partisan-media ecosystem has been engineered for the cognitive constraints of its audience. Headlines that fit one's existing frame; affective hooks that provide quick emotional sorting; repeated narratives that build availability without verification; in-group identity cues that signal trustworthiness. None of this is "irrational" — it's industrialized exploitation of bounded rationality at scale. The defense is not "voters should be more rational"; that's not a defense, it's a wish. The defense is environmental design that helps bounded-rational decision-making produce better outcomes. ### Why Information Density Doesn't Help A common naive policy proposal: provide more information to voters. This addresses one of Simon's three constraints (information access) while ignoring the other two (cognitive capacity, time). Adding information often makes decision quality *worse* — more options to evaluate, more dimensions to track, more opportunities for Anchoring Effect and Texas Sharpshooter. The right design is not "more"; it's "better-curated and better-presented." ### Bureaucratic Bounded Rationality and the Schedule F Problem Federal agencies operate under organizational bounded rationality: they develop standard operating procedures, expert specialization, and institutional memory precisely because no individual or even small team can re-derive optimal regulation from first principles every time. [[Schedule F]]'s threat to convert career civil servants to politically removable employees attacks one specific input to organizational bounded rationality (institutional memory and apolitical expertise) — without replacing it. The predictable result is that federal agencies decline in capacity to make even satisficing decisions; the [[DOGE]]-style chaos is what bounded organizational rationality looks like when it has been deliberately damaged. ### The Strategic Use of Information Overload "Flood the zone with shit" (Bannon's strategy) is information warfare that exploits bounded rationality at scale. It doesn't try to convince anyone of any particular thing; it overwhelms attention so that bounded-rational sorting is impossible. The result is that voters fall back on the simplest available cues — most commonly, partisan identification — because the cost of more careful analysis exceeds available cognitive budget. The defense requires either dramatically reducing the noise (hard) or providing curation that voters can trust without doing their own filtering (also hard). ## Implications 1. **Don't ask people to be more rational.** Bounded rationality is the structural condition; demanding super-rationality is asking for what doesn't exist. The honest design problem is helping bounded-rational decision-makers produce better outcomes. 2. **Heuristics are evaluated by environment.** A heuristic that works in one environment fails in another. The research program is not "are heuristics good or bad?" but "which environments make which heuristics work?" 3. **Choice architecture matters.** Defaults, framings, presentation orders, and option sets all affect bounded-rational choice. The [[Decision Science]] literature on nudge theory is the operational extension. 4. **Watch for adversarial choice architecture.** Bounded rationality is exploitable by actors who design environments to produce specific decisions. Recognizing the design intent is part of defending the decision. ## See Also - [[Herbert Simon]] — author - [[Decision Science]] — the discipline this framing founded - [[Wisdom of Crowds]] — bounded-rational individuals can collectively produce better decisions than any one of them - Action Bias — one of the bounded-rationality failure modes Simon's work helped name - [[Dysrationalia]] — Stanovich's refinement: capacity vs. disposition within bounded rationality - Superforecasting (Mental Model) — bounded rationality done well through disciplined cognitive dispositions - Principles of Influence — Cialdini's principles exploit bounded-rational shortcuts