Noise

In Noise, Daniel Kahneman teams with Olivier Sibony and Cass Sunstein to study a flaw in human judgment that bias alone cannot explain: the random variability in decisions that should come out the same. They show that when different experts evaluate identical cases, whether medical scans, court sentences, insurance risks, or job applicants, their judgments scatter far more than anyone assumes, and that scatter carries a steep hidden cost. The authors distinguish this noise from bias, explain how to measure it through a noise audit, and propose ways to reduce it, including structured protocols, independent judgments aggregated together, and what they call decision hygiene. The result extends the project Kahneman began in Thinking, Fast and Slow into new ground.