A Measurement of Decreasing Impatience by DI-index in Money and Health Domain

This paper uses the DI index introduced by Rohde (2014) to measure the impatient condition and quantify deviations from constant discounting for health and money. Our measurements, which can easily be obtained from only two indifferences, make no assumptions about utility and also be computed for people with increasing or constant impatience. In an experiment, the group shows no deviation from constant discounting as a result of Wilcoxon test, which means increasing impatience is almost as prevalent as decreasing and constant impatience. But in individual level, a substantial minority was increasingly impatient. By comparing the deviation between monetary DI index and health DI index, we know that the intertemporal preferences are context dependent and findings for money outcomes cannot be simply transferred to health. The generalized hyperbolic discounting model is rejected by the average experiment data in health condition but accepted in money condition. The Quasi-hyperbolic model is rejected in both conditions. The curve for CADI model is discussed, and the CADI model is rejected. The CRDI model is also rejected by the same logic.