The current studies investigated how a manipulation in joint outcome influenced individuals’ responses to pro-equality/individualistic decision makers. In Study 1 (N = 175), we examined the impact of whether equal distribution led to maximum joint outcome or not on individuals’ evaluations of, and reactions to, partners choosing either equal or individualistic distributions. In Study 2 (N = 164), we further examined the moderating roles of individual differences in general social value orientation (SVO) and preferences for joint outcome (vs. equality). Important findings include: a) individuals evaluated a pro-equality partner as less warm when equal distribution did not afford maximum joint outcome than when it did; b) individuals, especially those who scored high on preferences for joint outcome (relative to equality), were less likely to chose equal distribution when equality did not maximize joint outcome than when it did; and c) individuals who preferred joint outcome to equality evaluated individualistic partners as warmer when equal distribution did not yield maximum joint outcome than when it did. Theoretical and practical implications of these findings are discussed.

Decades of research show that (i) social value orientation (SVO)is related to important behavioral outcomes such as cooperation and charitablegiving, and (ii) individuals differ in terms of SVO. Aprominent scale to measure SVO is the social value orientation slider measure(SVOSM). The central premise is that SVOSM captures a stable trait. But it isunknown how reliable the SVOSM is over repeated measurements more than one weekapart. To fill this knowledge gap, we followed a sample of N =495 over 6 months with monthly SVO measurements. We find that continuous SVOscores are similarly distributed (Anderson-Darling k-sample p =0.57) and highly correlated (r ≥ 0.66) across waves. Theintra-class correlation coefficient of 0.78 attests to a high test-retestreliability. Using multilevel modeling and multiple visualizations, wefurthermore find that one’s prior SVO score is highly indicative of SVOin future waves, suggesting that the slider measure consistently capturesone’s SVO. Our analyses validate the slider measure as a reliable SVOscale.

Most prior research on the external validity of mixed-motive games has studied only one single game version and/or one specific type of real-life prosocial behavior. The present study employs a different approach. We used multiple game trials — with different payoff structures — to measure participants’ behavior in the Prisoner’s Dilemma, the Commons Dilemma, and the Public Goods Dilemma. We then examined the associations between these aggregated game behaviors and a wide set of self-reported prosocial behaviors such as donations, commuting, and environmental behaviors. We also related these prosocial behavior measures to a dispositional measure of prosociality, social value orientation. We report evidence that the weak statistical relationships routinely observed in prior studies are at least partially a consequence of failures to aggregate. More specifically, our results show that aggregation over multiple game trials was especially effective for the Prisoner’s Dilemma, whereas it was somewhat effective for the Public Goods Dilemma. Yet, aggregation on the side of the prosocial behaviors was effective for both these games, as well as for social value orientation. The Commons Dilemma, however, turned out to yield invariably poor relationships with prosocial behavior, regardless of the level of aggregation. Based on these findings, we conclude that the use of multiple instances of game behavior and prosocial behavior is preferable to the use of only a single measurement.

This paper introduces the Hierarchical Mindreading Model (HMM), a new model of mindreading in two-person, mixed-motive games such as the Prisoners’ Dilemma. The HMM proposes that the strategies available to decision makers in these games can be classified on a hierarchy according to the type of mindreading involved. At Level 0 of the HMM, there is no attempt to infer the intentions of the other player from any of the context-specific information (i.e., signals, payoffs, or partner reliability). At Level 1, decision makers rely on signals to infer the other’s intention, without considering the possibility that those signals might not reflect the other’s true intention. Finally, in Level 2 strategies, decision makers infer the other player’s intended choice by integrating information contained in their signals with the apparent reliability of the other participant and/or the game’s payoffs. The implications of the HMM were tested across four studies involving 962 participants, with results consistently indicating the presence of strategies from all three levels of the HMM’s hierarchy.

Social Value Orientation (SVO) refers to an individual’s preference for the division of resources between the self and another person. Since evidence suggests that hormones influence several facets of human social behavior, we asked whether SVO might fluctuate across the female menstrual cycle. Using self-report data obtained in two independent online studies, we show that cooperative preferences, as indexed by SVO, are indeed significantly more prosocial in the early follicular compared to the midluteal phase in naturally ovulating women. Furthermore, when estimating hormonal variations from norm data, we found estradiol, but not progesterone or testosterone, to be a significant predictor of SVO across the menstrual cycle in both studies, with a negative correlation. Our findings provide evidence that the willingness to cooperate varies across the natural female menstrual cycle and highlight the potential of investigating psychological effects of ovarian sex hormones.

How individuals tend to evaluate the combination of their own and other’s payoffs—social value orientations—is likely to be a potential target of future moral enhancers. However, the stability of cooperation in human societies has been buttressed by evolved mildly prosocial orientations. If they could be changed, would this destabilize the cooperative structure of society? We simulate a model of moral enhancement in which agents play games with each other and can enhance their orientations based on maximizing personal satisfaction. We find that given the assumption that very low payoffs lead agents to be removed from the population, there is a broadly stable prosocial attractor state. However, the balance between prosociality and individual payoff-maximization is affected by different factors. Agents maximizing their own satisfaction can produce emergent shifts in society that reduce everybody’s satisfaction. Moral enhancement considerations should take the issues of social emergence into account.