Abstract:

In social economic decisions, people not only care about their own payoffs but also the payoffs of others, a tendency termed altruistic preference. Numerous studies have shown that the sheer sense of being observed is sufficient to augment subjects’ altruistic choices. However, whether subjects’ altruistic behavior can be modulated by revealing their decision processes to other stake-holders (i.e., the receiver) remains unclear. In this study, we provide experimental evidence about the effects of both decision visibility to others and social norms on participants’ altruistic preference in two studies respectively. First, we confirmed that receiver visibility affected deciders’ altruistic preference in Study 1. In Study 2, we further showed that social norms modulated the effect of behavior visibility on deciders’ altruistic preference, suggesting a potential avenue via which social norms influences the relationship between behavioral visibility and altruistic preference.
Study 1 implemented a 2 (Visibility: Visible vs. Invisible) × 2 (Reaction type: Choice vs. Rating) × 2 (Inequity type: AI vs. DI) within-subject design. We recruited 38 participants and they were required to either choose from two reward allocation options (Choice task) with another partner, or rate how satisfied concerning a particular allocation (Rating task) in a dictator game (DG). Participants’ behavior was either observed by their “partners” (Visible condition) or remained private (Invisible condition) (Figure 1). We provided both model-free and model-based evidence for the effects of visibility on altruistic preference. Specifically, we calculated the mean absolute differences between self- and other-payoffs in the chosen trials in Choice condition, and the mean rating scores in Rating condition, as model-free indicators (indicators were calculated in advantageous domain and disadvantageous domain separately, Table 1). As for model-based measures, we utilized the inequity aversion model $(U={{M}_{s}}-\alpha \cdot \max \{{{M}_{s}}-{{M}_{o}},0\}-$ $\beta \cdot \max \{{{M}_{o}}-{{M}_{s}},0\})$ to quantify social preferences (Table 2). Repeated measures ANOVA on model-free and model-based measures revealed similar patterns: Compared to the Invisible condition, participants exhibited greater altruistic preference when their behavior were visible to the receivers (partners), indicated by increased advantageous inequity aversion (AIA) and decreased disadvantageous inequity aversion (DIA) in the Visible condition. This tendency was significant across both Choice and Rating tasks. In addition, participants cared more about allocation efficiency in the Choice task than in the Rating task, indicated by decreased AIA and increased DIA. Finally, Visibility alleviated the behavioral discrepancies between the Rating and Choice tasks, indicating that social preference and choice strategies tend to converge in the Visible condition (Table 3 & 4).
Study 2 implemented a 2 (Visibility: Visible vs. Invisible) × 2 (Social norm: Altruistic vs. Non-altruistic) within-subject design. 53 participants took part in the choice task with altruistic or non-altruistic social norms. Different social norms were manipulated by the proportions of unfair options chosen by supposedly previous participants (Altruistic social norm vs. Non-altruistic norm) (Figure 2). Measures remained the same as the choice condition in Study 1, except that we calculated the mean other-payoffs in the chosen trials as an additional measure of altruistic behavior (Table 5 & 6). ANOVA results exhibited an interaction between Visibility and Social norm: In the Altruistic social norm condition, visibility significantly increased participants’ altruistic preference; However, such effect diminished in the Non-altruistic social norm condition (Table 7 & 8).
Our study revealed that deciders’ behavioral visibility to receivers increased altruistic preference and promoted altruistic behavior. Furthermore, altruistic social norm played a modulatory role on the visibility effect, supporting the signaling hypothesis of altruistic preference.

Key words: social preference, altruism, visibility, reaction type, social norm