A monkey economy as irrational as ours

Laurie Santos, Yale psychologist, presents a series of experiments that taught capuchin monkeys to use money — and then documented, with rigorous controls, that the monkeys exhibit exactly the same irrational economic biases as humans: loss aversion, reference-dependent valuation, and irrational risk preferences. Her conclusion has profound implications: many of the cognitive biases that behavioral economists identify in humans are not cultural accidents but deep evolutionary inheritances.

Key Arguments

1. Monkeys invented money in the lab. Santos and colleagues taught capuchin monkeys to use coins as currency, establishing a functioning token economy where monkeys could “pay” for food. This created an experimental platform to test whether economic biases are learned culturally or inherited evolutionarily.
2. Monkeys show loss aversion. When presented with identical expected value gambles framed as gains vs. losses (receiving one grape or getting one extra vs. starting with two grapes and risking losing one), monkeys consistently preferred the gain-framed gamble — just like humans. Loss looms larger than gain.
3. Monkeys violate rational choice in the same ways humans do. The capuchins’ pattern of choices violated expected utility theory in ways that exactly mirror human irrationality — including the asymmetric treatment of equivalent outcomes depending on how they are framed.
4. If the bias is evolutionary, awareness alone won’t fix it. The practical implication Santos draws: if these biases are pre-cultural and pre-rational, knowing about them is necessary but not sufficient for overcoming them. Behavioral change requires environmental design, not just education.

Evidence Context

Santos’s capuchin market experiments are published in peer-reviewed behavioral economics journals and are among the most creative demonstrations of evolutionary roots of cognitive bias. The loss aversion finding in particular — one of the most robust in behavioral economics (documented in thousands of human experiments) — appears robustly in primate behavior, providing strong evidence for its evolutionary origin. The policy and design implications (that information alone cannot change bias-driven behavior) are consistent with decades of behavioral economics research.