Theory Of Intertemporal Choice Research Articles

Self-control and loss aversion in intertemporal choice

The life-cycle theory of saving behavior (Modigliani, 1988) suggests that humans strive towards an equal intertemporal distribution of wealth. However, behavioral life-cycle theory (Shefrin & Thaler, 1988) proposes that people use self-control heuristics to postpone wealth until later in life. According to this theory, people use a system of cognitive budgeting known as mental accounting. In the present study it was found that mental accounts were used differently depending on if the income change was positive or negative. This was shown both in a representative nationwide sample of households and in a student sample. Respondents were more willing to cut down on their propensity to consume when faced with an income decrease than to raise it when the income increased. Furthermore, contrary to the predictions of behavioral life-cycle theory, it was found that the respondents adjusted their propensity to consume the most when the income increases or decreases took place immediately. Hence, it is suggested that theories of intertemporal choice (e.g., Loewenstein, 1988; Loewenstein & Prelec, 1992) provide a better account of the data than does the behavioral lifecycle theory.

Time preference for health gains versus health losses.

The purpose of this research was to determine whether people devalue further health gains differently from future health losses. 108 subjects in various states of health were randomised to groups that rated their preference for a hypothetical health gain or loss of variable duration and delay, in the condition of arthritis. Direction and duration of the hypothetical future health change had an interactive effect on time preference (p less than 0.001). For the health gain, devaluation due to delay was consistent across durations. For the health loss, devaluation depended on duration; the long-duration loss was devalued with delay while the fleeting loss was inflated. These findings cast doubt on the assumptions of positive time preference and constant rate discounting that underlie the classical model of discounting. They provide support for a theory of intertemporal choice which posits that vivid, fleeting, future events engender 'anticipation utility' which attenuates positive time preference. Our findings suggest that standard practices for discounting future health outcomes in economic evaluations of arthritis and possibly other conditions may need to be re-examined.

Impossibility results in the axiomatic theory of intertemporal choice

We have shown formally that for infinite generation, transitive, intertemporal choice procedures the familiar conditions of Pareto optimality and independence together with the equity condition of stationarity are incompatible unless one is willing to accept a dictatorship. As in the case of Arrow's impossibility theorem, the result is primarily conceptual rather than practical. Real world decisions are seldom made by a pre-specified rule that is broad enough to deal with a full domain of individual or generational preference profiles. Nevertheless, impossibility results give us important clues about inherent and surprising limitations in institutions we often accept unquestioningly. The appearance of stationarity, introduced in [2] and proposed as a social choice condition in [1], leads to broader and more practical questions involving discount rates. If we accept the full intertemporal choice model as presented, then we are led to question the validity of any debate about ‘fair’ discount rates. Indeed, no discount rate regardless of how close it may be to unity can be fair under the assumptions we have used. Such discounting implies stationarity and is tantamount to giving dictatorial power to the first generation. Since discount rates are in practice applied in much more specialized and varied ways, this criticism of discounting must be tempered somewhat. The results do suggest, however, that discounting should be employed discriminately and with caution. The dictatorship result of Arrow's theorem depends on the assumption that the social choice function is transitive. If transitivity is weakened to quasitransitivity or acyclicity, then one obtains, respectively, an oligarchy or a collegial polity (see [5] and [6]). It is not clear to what extent our results carry over to these situations, though it seems reasonable to conjecture that, in the quasitransitive case, the oligarchy must consist of consecutive initial segments of generations. In the acyclic case one would likewise expect the present generation to belong to the collegium (the intersection of all decisive sets) and thus to have, at least in the monotonic case, a power of veto.