This research project formulates a normative theory of learning both preferences and probabilities that explains a broad spectrum of economic behavior heretofore judged irrational.
Economists have historically assumed that economic agents have direct perceptual access to the utility of goods and to probabilities in the world. Convergent streams of evidence from behavioral economics and neuroscience suggest such assumptions are incorrect: preference and probability are both learned from experience. The normative theory of learning developed by this project remains amenable to standard economic analyses and may lead to an enriched representation of agency for primitive units of economic behavior in both micro- and macroeconomics.
