Abstract:
We use a field experiment and a within-subject design based on multiple Choice Lists (CLs) that integrate time and risk. Diminishing impatience with extended time horizons is studied by varying time horizons from one week to two years. Time-dated risky prospects are constant within CLs and are always compared with time-dated certain amounts to identify timedated Certainty Equivalents. Non-linear probability weighting is modeled with a 2-parameter Prelec function. First, we identify a strong diminishing impatience associated with longer time delay between prospects. Second, we test whether non-linear probability weighting can explain and reduce the observed diminishing impatience by replacing linear probability weighting with an estimated inverted S-shaped Prelec function. We find that this does not reduce the observed degree of diminishing impatience. We conclude that the observed diminishing impatience is neither explained by the combination of present bias and certainty bias nor by non-linear weighting of risk in future prospects.
