CONICET | Buscador de Institutos y Recursos Humanos

Introduction. Inequality aversion is a largely inquired phenomenon. Numerous experimental studies show evidence of people?s preference for reducing inequality even at a personal cost and without personal gains involved. Nonetheless, there is also survey evidence of people?s preference for unequal distributions in real life. When asked about large-scale distributions, people seem to favor unequal ones up to a certain level. In fact, a novel series of experimental studies recently found that uninterested third parties preferred to reduce inequality but only if it did not jeopardize the relative relation between subjects´ payoffs (rank order; Xie et al., 2017). Authors called this ?rank reversal aversion?. However, those experiments focused only on the decision of the third party after the initial distribution was randomly assigned. We aim to examine this effect in light of the substantial literature about fairness perception and deservedness. Specifically, we wonder about the moderating role of justice perceptions on rank reversal aversion. Many studies show varying fairness considerations and preferences due to different allocation mechanisms, such as the acceptance of unfair offers in the Ultimatum Game when randomly generated, in contrast with its rejection if chosen by a person. Also, formally identical situations have been found to generate different beliefs due to changes in their framing. Judgments are not made in a vacuum: according to the available elements and the most salient information, people tend to change their beliefs about what is appropriate in a particular situation, which, in turn, influences preferences and behavioral intentions. Measured by such beliefs, social norms have been found to possess high explanatory power on the variation of preferences and decisions. Goals. In this study, we aim to investigate the effect of information about performance in a real effort task (and therefore, deservingness) which does or does not match the initial random allocation between two subjects, on the redistributive preferences of an uninterested third party. We also want to inquire about the social norms present in those settings. Methodology. We propose an experiment based on the original design of Variant 1 by Xie et al. but with a 3x3 within-subject design in which we vary the type of monetary transference (whether it reduces inequality without reversing the ranking, it equalizes payoffs, or reduces inequality and reverts participants´ original ranking) and the information of performance in a real effort task and whether relative performance and relative initial distribution matched or not (i.e., whether the participant that gets the initial higher payoff is the one that performed better in the effort task). In this experiment, participants play a Disinterested Dictator Game (participant C decides on the payoffs of participants A and B) indicating their preference on every possible redistributive scenario (strategy method). Also, to estimate prescriptive and descriptive social norms in each condition, an independent group of participants state their beliefs about what others would consider appropriate and what others would do on each experimental scenario. Redistributive decisions have economic consequences for the participants involved except for the decision-maker, whereas the estimation of social norms also is monetarily incentivized. Results. The experiment will be carried out in June 2019 and ready to be communicated on August 2019 during the AACC meeting. We expect variations on people?s beliefs and preferences as a function of the experimental condition. Specifically, we hypothesize that people will be prone to reverse the ranking when the initial allocation is perceived as unfair due to a mismatch with the performance on the real effort task. Discussion. A better understanding of redistributive preferences would be beneficial both for the field of research on behavioral sciences and for the design of interventions and public policies.