Human decision making in black swan situations

Real-world decisions often involve “black swan” choices with extremely low probability chances of catastrophic loss, like riding a motorcycle or going on a dangerous trip. These have several characteristics that make them especially difficult from the perspective of decision theory. How do people assign utilities to losses like “go bankrupt” or “die”? Do people have the representational resolution to encode differences between extremely tiny probabilities? We address these questions in two experiments in which people make decisions involving very low probabilities (as low as 1 in 10,000) of losing all of their points (and monetary bonus). Our results indicate that people mostly appear not to encode differences between tiny probabilities and are indifferent to the magnitude of losses. These factors lead to a startling qualitative shift in behaviour between scenarios with the same expected value and very similar absolute risk levels: people are risk averse when only one option is a black swan but become strongly risk seeking when both are.