The Performance of Time-Preference and Risk-Preference Measures in Surveys

Abstract

Time preferences and risk preferences play an important role in a wide range of behavior, including financial decisions, entrepreneurship, and the proper incentivizing of agents. Numerous methods have been developed to measure these preferences hypothetically in surveys, but they have yielded inconsistent results. We analyze a panel data set in which subjects have collectively answered more than 400 surveys including 15 time-preference and 36 risk-preference elicitations. We evaluate the performance of these measures using the criteria of (1) ability to predict economically important behavior and (2) distinctness from other observables. We find substantial heterogeneity in the predictiveness of the measures. The best performing measure for time-preference is a titration method, in which a sequence of adaptive binary-choice questions narrows in on a subject’s indifference point, and for risk-preference it is a self-report measure of risk aversion. Using factor analysis, we find that time preferences are well explained by a single factor, but risk preferences load on multiple factors. However, the first factor loads almost entirely on self-reported risk-preference measures, and this factor explains much of the variation. The evidence can help inform researchers about which elicitation methods to include in their surveys. This paper was accepted by Yan Chen, decision analysis.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 28 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.