Abstract
We study whether and how time preferences change over the life cycle, exploiting representative long-term panel data. In order to disentangle age effects from cohort and period factors, we estimate individual fixed-effects models, substituting period effects with period specific determinants of time preferences. We find that discount rates linearly decrease over the life cycle. This pattern holds when adjusting for the curvature of the utility function. We show the quantitative importance of this finding by demonstrating that a canonical life-cycle model with decreasing discounting significantly improves its fit to consumption and asset data relative to the model with constant discounting.