Linda: Less is More

Kahneman presents the most celebrated and controversial demonstration in the heuristics-and-biases literature: the Linda problem. Participants read that Linda is single, outspoken, very bright, a philosophy major deeply concerned with discrimination and social justice. Asked which is more probable — that Linda is a bank teller, or that Linda is a bank teller and active in the feminist movement — a large majority chose the second, even though it is a logical impossibility for a conjunction to be more probable than one of its components.

The mechanism is again representativeness overriding logic. 'Bank teller and feminist' fits the vivid story of Linda far better than 'bank teller' alone, so it feels more probable; the added detail makes the scenario more coherent and more representative even as it makes it, by the rules of probability, strictly less likely. The set of feminist bank tellers is necessarily smaller than the set of all bank tellers, but System 1 responds to the plausibility of the story, not to the logic of sets — a violation Kahneman and Tversky named the conjunction fallacy.

Kahneman draws the broader principle he calls 'less is more': adding plausible, representative detail to a scenario makes it feel more probable while making it actually less probable. A detailed, specific forecast — 'a flood will displace thousands in California next year' — is judged more likely than a vaguer, broader one — 'a natural disaster will displace thousands somewhere in North America' — even though the broader claim must be more probable because it includes the narrower one. Coherence and specificity buy believability at the cost of logical likelihood.

The robustness of the effect is striking: it persists among statistically sophisticated subjects, survives reformulation, and resists the objection that people merely misread 'probable' as 'plausible.' Kahneman concedes that wording and conversational norms play some role, but the core finding — that a more detailed conjunction can be judged more probable than its own component — held across many variations. The fallacy reveals how thoroughly the feeling of a good story can override the arithmetic of probability.

The applied takeaway is to be suspicious of detailed, coherent scenarios that feel highly likely. In forecasting, planning, and risk assessment, every added specific condition makes a story more vivid and persuasive but less probable, because each condition must also be satisfied. When a richly detailed prediction feels compelling, remember that its very richness lowers its true odds; the broader, blander version is the safer bet, however less satisfying it is to believe.

Kahneman's deeper point is that the Linda problem dramatizes the gap between the logic of probability and the psychology of plausibility. Our minds are built to evaluate stories for coherence and representativeness, not to compute the nested probabilities of sets, so a good narrative reliably defeats a logical truth. This is not a marginal error but a window into how intuition works: it favors the explanation that hangs together, and it is precisely this preference for coherence over correctness that makes detailed, convincing stories so persuasive and so frequently wrong.