Research

An ongoing debate regarding the evolution of morality is whether other species show precursory moral behavior. The veil of ignorance (VOI) paradigm is often used to elicit human moral judgment but has never been tested in other primates. We study the division of resources behind the VOI in Formosan macaques. Monkeys choose the equal division more often when a conspecific is present than when it is absent, suggesting a degree of impartiality. To better understand this impartiality, we measure a monkey’s reactions to two directions of inequity: one regarding inequity to its advantage and the other to its disadvantage. We find that disadvantageous inequity aversion correlates with the degree of impartiality behind the VOI. Therefore, seemingly impartial behavior could result from a primitive negative reaction to being disadvantaged. This suggests a mechanism to explain a tendency toward impartiality. 

This study investigates human behavior when individuals make choices that are uncertain in both realization timings and outcomes. Individuals’ choice patterns are explored in controlled experiments through a new type of lottery, “two stage compound time-lottery,” in which the first stage determines the realization timing and the second stage determines the realization outcome. We first examine the validity of the standard expected utility (EU) model and find that EU fails to account for the majority of our subjects. As the alternative, we examine three non-expected utility models developed from rank dependent utility: recursive rank dependent utility (RRDU), intertemporal certainty equivalent (ICE) and rank dependent utility on discounted payments (RDUDP). The results show that individuals are heterogeneous and two choice patterns emerge. Approximately two-thirds of subjects do not reduce compound time-lotteries and their choices can be rationalized by RRDU. For the rest one-third of subjects who employ reduction, their choices can be rationalized by RDUDP. Only fewer than 7% of subjects can be rationalized by ICE or discounted expected utility.

This paper introduces a new type of lottery, two stage compound time-lottery, to study choices which involve risk in both realization outcome and realization timing. The independence axiom and the reduction axiom are extended to the compound time-lottery framework. Based on these two axioms, this paper employs Rank Dependent Utility (RDU) to develop three different models to evaluate compound time-lotteries. An extension of stochastic dominance, two stage stochastic dominance, is introduced and the three models are demonstrated to be monotonic with respect to two stage stochastic dominance. Recently, DeJarnette et al. (2020) empirically document that risk aversion toward realization outcome is positively correlated with risk aversion toward realization timing. They also show that with the independence axiom, risk aversion over time lottery (RATL) cannot not be accommodated without violating stochastic impatience. This paper demonstrates that under RDU, the reduction axiom in compound time-lottery may provide a solution to simultaneously accommodate RATL and stochastic impatience. Further, the three models are shown to be able to account for the positive correlation between risk aversion over outcome and over time.