Constant Discounting, Temporal Instability and Dynamic Inconsistency in Denmark: A Longitudinal Field Experiment

Glenn W. Harrison, Morten I. Lau, Hong Il Yoo

Research output: Working paperResearch


Evidence that individuals have dynamically inconsistent preferences is usually generated by studying individual discount rates over different horizons, but where those discount rates are elicited at a single point in time. If these elicited discount rates are constant over different time horizons the individual is claimed to have preferences that imply dynamic consistency. However, to test dynamic consistency one has to know if the same subject has the same discounting function at a later point in time. In turn, longitudinal tests of temporal stability and dynamic consistency require that one allow for endogenous sample selection and attrition processes that can directly confound inferences about constant discounting, temporal stability and dynamic consistency. We evaluate a longitudinal experiment undertaken with a nationally representative sample of the adult Danish population, making these corrections. We cannot reject the hypothesis of constant discounting at the population level, but we reject the hypotheses of temporal stability and dynamic consistency.
Original languageEnglish
Place of PublicationAtlanta, GA
PublisherCEAR, Georgia State University
Number of pages57
Publication statusPublished - 9 Jul 2020
SeriesWorking paper / Center for Economic Analysis of Risk (CEAR)


  • Time preferences
  • Dynamic consistency
  • Longitudinal experiments
  • Sample selection and attrition

Cite this