Contrary to the standard economic view of misreporting for material gain, recent evidence highlights robust preferences for truth-telling. Models consistent with aggregate behavior attribute these preferences to intrinsic honesty and social image concerns. We introduce a novel incentivized measure that independently identifies both motives at the individual level. It exhibits strong internal validity, predicts behavior in established honesty paradigms, and uncovers stable preference types across samples. We also propose a 2-minute survey module predicting behavior in a standard reporting task. Implemented in a large panel, the module allows us to shed light on how early-life experiences shape preferences for truth-telling.

This Version: December 2025