Presentation (CTI)

OBJECTIVES: To demonstrate why compound discounting is wrong, how it affects many health economics studies, and how to fix the problem.
METHODS: Historical evidence reveals that compound discounting as practiced in all economics, including health economics, arose from then-available medieval mathematical methods (arithmetic) and became embedded in economic literature. However, numerous behavioral economics studies show that people’s true inter-temporal preferences do not follow compound discounting, but rather, “hyperbolic discounting.” The best functional form to estimate such preferences remains unknown. I propose five plausible criteria to evaluate the properties of various hyperbolic discounting functions from behavioral economics and those associated with standard economic utility functions.
RESULTS: All proposed hyperbolic discount functions imply infinitely large perpetuity values, discrediting them from HEOR and HTA analyses for various reasons. Discounting functions associated with all standard economic utility functions also fail on at least one criterion. A new discount function, offset exponential power (OEP) discounting meets all five criteria. Discrete Choice Experiments provide clear ways to choose between alternative discount functions and to estimate relevant parameters. Two viable options arise that require only two parameters to estimate.
CONCLUSIONS: A simple question illuminates the problem: “Which do you prefer, $1000 today or $2000 in X days?” This cannot be properly analyzed without simultaneously assessing the discount function and the function describing the utility of income. The entire economics and health economics literature ignores this problem by assuming that compound discounting properly describes people’s true time preferences, but numerous studies show this to be false. To solve this problem, health economists should lead the economics discipline by estimating proper discount functions, using methods proposed herein, and then repairing defects in the health economics literature by applying appropriate discounting functions to replace compound discounting. Once the best discounting function is known, re-evaluation of most health economics and HEOR studies is relatively simple.