Efficient Coding and Risky Choice
82 Pages Posted: 5 Nov 2018 Last revised: 7 Aug 2020
Date Written: August 7, 2020
We experimentally test a theory of risky choice in which the perception of a lottery payoﬀ is noisy due to information processing constraints in the brain. We model perception using the principle of eﬃcient coding, which implies that perception is most accurate for those payoﬀs that occur most frequently. Across two pre-registered laboratory experiments, we manipulate the distribution from which payoﬀs in the choice set are drawn. In our ﬁrst experiment, we ﬁnd that risk taking is more sensitive to payoﬀs that are presented more frequently. In a follow-up task, we incentivize subjects to classify which of two symbolic numbers is larger. Subjects exhibit higher accuracy and faster response times for numbers they have observed more frequently. In our second experiment, we manipulate the payoﬀ distribution so that eﬃcient coding induces the decision maker’s perceived value function to switch from concave to convex. We ﬁnd that demand for risk is signiﬁcantly higher when eﬃcient coding induces a convex value function. Together, our experimental results suggest that risk taking depends systematically on the payoﬀ distribution to which the decision maker's perceptual system has recently adapted. More broadly, we provide novel evidence of the importance of imprecise and eﬃcient coding in economic decision-making.
Keywords: efficient coding, perception, risky choice, neuroeconomics
JEL Classification: G02, G41, D81, D87
Suggested Citation: Suggested Citation