Value Functions in Domains of Gains and Losses

Many medical situations incorporating gains and losses involve exchange of present-day costs for future benefits. In our era of chronic illness, with antecedents often distant in time, decision making by physicians or patients is frequently influenced in a significant manner by the principles and practice of discounting. Positive discounting is the term for diminishing value over time. A more diversified use becomes important as we attempt to refine decision models. Temporal effects in the medical preventive context include overeating, smoking, and drinking. Another example is investment in calcium intake or vitamin D to prevent fractures because of osteoporosis.

These are situations where there is a choice between immediate pleasures and the future benefits of good health. The benefits of behavioral change often occur so far in the future that they seem of little value relative to the immediate costs. If individuals prefer to live for the present rather than save for the future, such a preference may not yield the best lifetime outcomes for the individual.

Studies of gains and losses and time preference in medical decision making have dealt with either life saving or health change. Different methods have been used in the two areas to find individuals' time preference rate, however, and this complicates comparisons between these areas.

Research on decision making under uncertainty has focused largely on expected utility and behaviors that violate this axiom. Intertemporal choic-es—that is, choices between something now and something later—have been investigated less. Discounted utility, however, is in need of a model that takes into account anomalies and framing factors in medical decision making. These effects are a challenge to normative theory. In spite of the difference between economic and medical decisions, both domains have revealed discounting biases such as magnitude effect, dynamic inconsistency, instant endowment and status quo bias, sequence effect, sign effect, and loss aversion.