Before you read on: volunteers held one hand in painfully cold water for 60 seconds, and the other for the same 60 seconds plus 30 more as the water warmed slightly. Which trial did most choose to repeat?
A clear majority chose the longer trial. Kahneman explained that memory of an episode rests on its worst moment and its ending, and hardly at all on its length.
- Lived
- Born 5 March 1934 in Tel Aviv, then in British-ruled Palestine, to a family living in France; died 27 March 2024
- Nationality
- Israeli and American
- Work
- Psychologist at the Hebrew University of Jerusalem, the University of British Columbia, Berkeley and, from 1993, Princeton
- Key works
- Judgment under Uncertainty: Heuristics and Biases (1974) and Prospect Theory (1979), both with Amos Tversky; Thinking, Fast and Slow (2011); Noise (2021), with Olivier Sibony and Cass Sunstein
- Prize
- Nobel Prize in Economic Sciences, 2002, shared with Vernon Smith
- School
- Behavioural economics, which his work with Tversky helped to found
What Kahneman was reacting to
Economists modelled choice under risk with expected utility theory. A rational decision-maker was assumed to weigh the value to them (utility) of each outcome by its probability and pick the option with the highest expected utility.
Kahneman, a psychologist, began around 1969 to test the statistical instincts of trained researchers with his colleague Amos Tversky. Even experts misjudged how much a small sample can mislead. The pair set out to map such errors as clues to how the mind works.
Kahneman's key ideas
Can you name Kahneman's four key ideas?
Mental shortcuts for judging likelihood. The 1974 paper set out three: representativeness (resemblance to a type), availability (how easily examples come to mind) and anchoring (adjusting too little from a starting figure). Each produces biases: errors that point the same way again and again.
Can you think of an example?
People read a description that sounded like an engineer, of a man drawn from a group that was 70 per cent lawyers. Most judged him an engineer, and reversing the mix barely changed answers. They neglected the base rate: the share of each job in the group.
A 1979 account of real choice under risk. People judge outcomes as gains or losses from a reference point, usually where they stand now, and a loss weighs more than an equal gain: about twice as much, on their later 1992 estimate. They play safe with gains but gamble to avoid a sure loss.
Can you think of an example?
Seven in ten people preferred a 25 per cent chance of losing $6,000 to a 25 per cent chance of losing $4,000 plus a 25 per cent chance of losing $2,000. Both average a $1,500 loss, yet most chose the bigger single loss.
Choosing differently between two descriptions of the same options. In standard theory the wording should not matter.
Can you think of an example?
Facing a disease expected to kill 600, most preferred a plan that saves 200 for certain to a gamble. Told the same plan means 400 will die, most chose the gamble.
Experienced utility is how life feels moment by moment; remembered utility is how an episode is judged afterwards. Memory follows a peak-end rule: it rests on the most intense moment and the last one, while length barely counts.
Can you think of an example?
Colonoscopy patients rated their pain every minute. Their later verdicts tracked the peak and the ending; length, from 4 to 66 minutes, made almost no difference.