What is a decision-making model? A decision-making model is a framework describing how decisions are or should be made. Normative models prescribe how you ought to decide; descriptive models describe how people actually decide.
The major decision-making models are: the rational or classical model (define the problem, evaluate all alternatives against complete information, choose the optimum — the idealized 'economic man'); bounded rationality and satisficing (Herbert Simon, Nobel 1978 — real deciders take the first good-enough option); the Carnegie model (Cyert, March and Simon — organizational decisions emerge from coalitions bargaining to a satisficing solution); the garbage can model (Cohen, March and Olsen, 1972 — in organized anarchies, problems, solutions, participants and opportunities are independent streams that collide); the Recognition-Primed Decision model (Gary Klein — experts recognize rather than compare); System 1 and System 2 thinking (Daniel Kahneman, Thinking Fast and Slow, 2011 — fast intuitive versus slow analytical); and the Cynefin framework (Dave Snowden and Mary Boone, Harvard Business Review 2007 — match approach to problem type across Clear, Complicated, Complex, Chaotic, and Confusion domains). Common biases that distort these models include anchoring, confirmation bias, sunk-cost fallacy, loss aversion, and groupthink. Argumentree operationalizes the useful core of these models with structured pro/con argument trees, multi-dimensional rating that aggregates into consensus scores, and a full audit trail.
Decision-Making Models
From the idealized rational model to the chaotic garbage can — the frameworks that explain how decisions really get made, who created them, and when to use each.
TL;DR
Decision-making models come in two flavors: normative (how you should decide — the rational model) and descriptive (how people actually decide — bounded rationality, recognition-primed decision). The history of the field is essentially one long argument with the rational model, starting the moment Herbert Simon pointed out that nobody has perfect information. Knowing the models tells you which one your situation actually calls for.
7 Key Models
Rational / Classical model — normative baseline
Define the problem, list all alternatives, evaluate against complete information, pick the optimum. Assumes a fully informed 'economic man' — the ideal every other model reacts to.
Bounded rationality & satisficing — Herbert Simon, Nobel 1978
Real deciders have limited information, time, and bandwidth, so they satisfice — take the first 'good enough' option instead of optimizing. The most influential correction to the rational model.
Carnegie model — Cyert, March & Simon
Applies bounded rationality to organizations: decisions emerge from coalitions of stakeholders bargaining to a satisficing solution everyone can accept — not from one rational optimizer.
Garbage can model — Cohen, March & Olsen, 1972
In 'organized anarchies,' problems, solutions, participants, and opportunities slosh as four independent streams; decisions happen when they collide. Solutions can exist before their problems.
Recognition-Primed Decision — Gary Klein
Experts under pressure don't compare options — they recognize a situation as familiar and act on the first workable course of action. The descriptive model of real-world expertise.
System 1 & System 2 — Daniel Kahneman, 2011
Two modes of thought: System 1 is fast, automatic, intuitive (and bias-prone); System 2 is slow, deliberate, analytical. Good decisions are partly knowing when to engage System 2.
Cynefin framework — Snowden & Boone, HBR 2007
Match your approach to the problem type across five domains — Clear, Complicated, Complex, Chaotic, Confusion. Using the wrong domain's logic is the classic failure.
The Turning Point: When Economics Admitted We're Not Robots
For a long time, economics ran on the "economic man" — a perfectly rational, fully informed optimizer. Then Herbert Simon argued that real people are limited by information, time, and cognition, so they satisfice rather than optimize. The idea was influential enough to earn him the 1978 Nobel Prize in Economics, and it spawned the Carnegie and administrative models. Nearly every modern model is a descendant of that one correction.
The other great descriptive turn came from Daniel Kahneman and Amos Tversky, whose work on heuristics and biases — and the System 1 / System 2 distinction — won Kahneman the 2002 Nobel Prize.
The Biases Every Model Has to Fight
Anchoring
The first number seen drags every later estimate. In Tversky & Kahneman's classic study, a rigged wheel landing on 10 vs 65 shifted people's guesses about UN membership from ~25% to ~45% — from a number they knew was random.
Groupthink
Irving Janis (1972) traced the Bay of Pigs fiasco to advisers suppressing doubt for the sake of consensus. The same team, after deliberately inviting dissent, navigated the Cuban Missile Crisis — same people, opposite outcomes, different process.
Loss aversion
Kahneman & Tversky's prospect theory (1979): losses hurt about twice as much as equivalent gains feel good — skewing 'rational' weighing.
Sunk-cost fallacy
Throwing good resources after bad because of what's already been spent.
How Argumentree Puts the Models to Work
You don't have to pick one model and live with its flaws. Argumentree gives you the rational model's discipline, bounded rationality's realism, and a defense against groupthink — built on argument mapping:
Rational-model structure
Options and their pros and cons laid out explicitly in a pro/con tree — the discipline of the classical model without pretending you have perfect information.
Measured 'good enough'
Net support scores let a group satisfice deliberately — set a threshold and stop — instead of optimizing into paralysis.
Anti-groupthink by design
Asynchronous contribution and per-argument rating surface dissent and quiet voices, instead of letting consensus pressure bury them.
System 2 on demand
When a decision is too important for gut feel, the structure forces the slow, analytical pass that System 1 skips.
See the broader decision making overview, the step-by-step decision-making process, and how groups apply these models in collaborative decision making.
Frequently Asked Questions
What is a decision-making model?
A decision-making model is a framework describing how decisions are or should be made. Normative models (like the rational model) prescribe how you ought to decide; descriptive models (like bounded rationality or recognition-primed decision) describe how people actually decide in practice.
What is the rational decision-making model?
The rational (classical) model is the idealized framework: define the problem, identify all alternatives, evaluate them against complete information, and choose the optimal option. It assumes unlimited information and cognitive capacity — the 'economic man' — and is the baseline that every later, more realistic model pushes back against.
What is bounded rationality?
Bounded rationality, from Nobel laureate Herbert Simon, holds that real decision-makers are limited by available information, time, and mental capacity, so perfect optimization is impossible. Instead they 'satisfice' — choose the first option that is good enough. It's the single most influential correction to the rational model.
What is the garbage can model?
The garbage can model (Cohen, March & Olsen, 1972) describes decision-making in 'organized anarchies' like universities. Problems, solutions, participants, and choice opportunities flow as four independent streams; a decision happens when they happen to collide. A striking implication is that solutions often exist before the problems they get attached to — decisions can be products of timing as much as logic.
What is the Cynefin framework?
Cynefin (Dave Snowden; popularized in a 2007 Harvard Business Review article with Mary Boone) sorts decisions into five domains — Clear, Complicated, Complex, Chaotic, and Confusion — and prescribes a different approach for each. Its core warning is that applying the wrong domain's logic, like treating a complex problem as merely complicated, leads to failure.
The best model is a visible one
Lay out the arguments, weigh them as a group, and keep the record — the discipline of the rational model with the realism of how people actually decide. Try Argumentree.
Start Free