Quicksand or Bedrock for Behavioral Economics? Assessing Foundational Empirical Questions

Published: 2018


Behavioral economics lacks empirical evidence on some foundational questions. We adapt standard elicitation methods to measure multiple behavioral factors per person in a representative U.S. sample, along with financial condition, cognitive skills, financial literacy, classical preferences, and demographics. Individually, behavioral factors are prevalent, distinct from other decision inputs, and correlate negatively with financial outcomes in richly-conditioned regressions. Conditioning further on other B-factors does not change the results, validating common practice of modeling B-factors separately. Corrections for low task/survey effort modestly strengthen the results. Our findings provide bedrock empirical foundations for behavioral economics, and offer methodological guidance for research designs.

Key Findings

    • “Behavioral factors”—psychology-based deviations from classical economic preferences, beliefs, and problem-solving approaches—are quite prevalent in a representative sample of U.S. individuals
    • A “B-count” measuring how many B-factors an individual displays is strongly predictive of self-assessed financial well-being, as well as “hard” measures of retirement preparedness, such as wealth and stock market participation
    • The findings can help financial service providers and policymakers design products and communicate information in ways that improve decision-making associated with savings and retirement planning.


Stango, Victor, Joanne Yoong and Jonathan Zinman. 2018. “Quicksand or Bedrock for Behavioral Economics? Assessing Foundational Empirical Questions.”Ann Arbor MI: University of Michigan Retirement Research Center (MRRC) Working Paper, WP 2018-378. https://mrdrc.isr.umich.edu/publications/papers/pdf/wp378.pdf