ثبات یا بی ثباتی ترجیحات اجتماعی:استفاده از حساسیت عدالت برای پیش بینی زمانی که نوع دوستی فرو می ریزد

Recent research suggests that altruism can be surprisingly tenuous; minor situational variations can turn altruism on and off. For example, if provided with sufficient cover, “reluctant altruists” will often avoid situations that compel them to give, and they may even secretly renege on gifts they just made. This behavior puts pressure on classic explanations of altruism and raises many questions about its stability. Is everyone's altruism prone to such collapse? If not, how can one predict it? We show that some people exhibit more stable altruism, predicting who is who weeks prior to the task. We show that high degrees of justice sensitivity is associated with pro-social behavior across situations, while low degrees of justice sensitivity relate to the use of situational variables as excuse to display less altruistic behavior. Our findings contribute to recent research on altruism and give insight into how to predict it.

During the rise of social preference theories (e.g., Andreoni, 1999, Battigalli and Dufwenberg, 2009, Bolton and Ockenfels, 2000 and Fehr and Schmidt, 1999), it became even more clear that people had other-regarding preferences of some sort. For example, individuals were thought to have a preference for improving others’ welfare (Bolton and Ockenfels, 2000 and Fehr and Schmidt, 1999) or a desire to experience the “warm glow” associated with giving to others (Andreoni, 1999). More recently, the assumptions of these social preference theories have come into question by research suggesting that people are often “reluctant altruists”—they will give but will also avoid the situations that compel them to give (Cain et al., 2013); e.g., they will avoid giving if they can avoid appearing stingy (Andreoni and Bernheim, 2009, Dana et al., 2007, Levitt and List, 2007 and Ockenfels and Werner, 2012) or are provided opportunities to reconsider/renege when facing less social pressure (Dana et al., 2006 and Sah et al., 2013). Because would-be recipients are so often left with nothing, these results challenge the idea that people generally have other-regarding preferences. However, this all leaves open the possibility that some subsets of people are not reluctant in their altruism. In other words, are there not some people whose altruism is more stable and can be explained by the typical motivations, e.g., by preferences for warm-glows or distributive outcomes? Using personality differences in justice sensitivity, the present research suggests that predictable subsets of the population have stable other-regarding preferences, while (reluctant) others are more prone to finding ways of avoiding giving. Generally speaking, this personality trait serves as a detector of justice-relevant situations, and represents the importance that an individual places on justice in her daily life. We argue that allowing for such preference heterogeneity (i.e., individual differences) can deliver useful insights to ex ante predict, not just ex post explain both genuine as well as reluctant altruism. Importantly, this individual difference measure is well-grounded in personality psychology and cognitive psychology and thus provides a theoretical basis for understanding differences in behavior. While our main question is whether there are stable components of altruism—predictable by anything—perhaps using a known personality scale will open the door to better understanding the social-cognitive and motivational foundations of altruistic behavior. 1.1. Reluctant altruism in dictator games A prototypical lab measure of other-regarding preferences is the “dictator game” (Camerer, 2003, Forsythe et al., 1994 and Guala and Mittone, 2010). Player A is endowed with x dollars and can “dictate” how the money is allotted between players, giving any amount he or she desires, $0 to $x, to player B in an anonymous, one-shot task. A key feature of the game is that if player A sends $0, player B gets nothing, but is told that an anonymous dictator sent $0; i.e., player B would know that he or she was “stiffed,” but could do nothing about it. A substantial amount of studies suggest that the majority of individuals gives something to receivers and that the average gift equals 20% ( Forsythe et al., 1994 and Guala and Mittone, 2010). However, recent studies find that slight variations of the dictator game provide substantially different results, causing strong debates about what actually motivates social behavior. For example, Dana et al. (2006) show that many dictators took the opportunity to renege on a gift when they were surprised by a “dictator exit” option allowing them to pay a small price for leaving the receiver with no money and no knowledge that a dictator game had taken place. This acceptance of small costs to avoid (being seen) refusing a request is what Dana et al. (2006) call “crossing the street to avoid the beggar.” Similarly, Broberg et al. (2007) found that roughly two-thirds of participants were willing to accept the dictator-exit option, giving up part of their endowment to avoid sharing. Recent studies yield similar findings supporting instability of other-regarding preferences that merely focus on financial outcomes (DellaVigna et al., 2009 and Lazear et al., 2012). These observations rule out many classical explanations for the initial giving. Givers who vary their gift depending on the situation must not worry about receivers’ payoffs or the inequity of the situation (e.g., Bolton and Ockenfels, 2000 and Fehr and Schmidt, 1999); they must not value the warm, fuzzy feeling derived from giving (Andreoni, 1999)1; and they must not ultimately want to see themselves as altruistic because these players see themselves ultimately giving nothing. Indeed, one might quip, if these people truly have other-regarding preferences, not only should they not avoid the beggar but even sign up for beggar-delivery services, bringing the needy to their doorsteps and allowing them to fulfill their putative desire to improve others’ welfare or reduce inequality in the world. If many people prefer to avoid requests for help, why do they also help when confronted? It is suggested that giving not only reflects preferences over outcomes but also something like “a desire not to violate other's expectations” (Dana et al., 2006, p. 193), as concluded from givers’ willingness to pay to leave receivers in the dark as to what is actually going on—what one does not know about cannot disappoint. Specifically, less-generous behavior occurs as soon as transparency between (beliefs about) dictator-actions and (beliefs about) receiver-reactions is clouded (Dana et al., 2007). A variety of experimental evidence points to the importance of such belief-dependent preferences, for example, in trust games (Charness and Dufwenberg, 2006), in field experiments (Andreoni et al., 2011), and in a large-scale newspaper dictator game (Ockenfels and Werner, 2012).2 As intriguing as these results are, however, they do not entail that nobody has other-regarding preferences over outcomes. Surely some altruism is genuine and is not merely disappointment aversion; and surely, the warm glow that people obtain from giving does drive some altruism. After all, contrary to our prior quip about beggar delivery, people sometimes do leave their house in search of ways of making the world a better place; such altruism may not be reluctant at all. Perhaps there are stable individual differences along these lines. Given additional experimental and theoretical evidence about people's diverging motivations to be perceived as “fair” ( Andreoni and Bernheim, 2009 and Fetchenhauer and Dunning, 2006), or their aversion to falling short of others’ expectations (Battigali and Dufwenberg, 2007; Dana et al., 2006), we posit that there are stable, measureable interpersonal differences with regard to these justice-related perceptions and motivations. These differences can be captured by measuring people's personality prior to engaging in these economic games. 1.2. Measuring personality by means of justice sensitivity Justice sensitivity (JS) is a personality trait that people possess in various degrees. It captures stable and consistent inter-personal differences in people's inclination to perceive situations as justice-relevant, as well as in the strength of emotional and behavioral responses to such situations ( Huseman et al., 1987 and Lovas and Wolt, 2002). Schmitt et al. (1995) suggested that people who score highly in terms of JS differ from those whose JS is low in four respects: the frequency at which they interpret a given situation as justice-relevant, the intensity of their emotional reactions to such situations, the degree of their cognitive reflections in these situations, and finally, their observed behavior. Because justice situations can be experienced in different roles (e.g., victim/helper) and because reactions differ widely depending on one's perspective on a situation ( Mikula et al., 1990 and Weiss et al., 1999), JS has been decomposed into several dimensions: the sensitivity to become a victim of injustice (JSvictim), the sensitivity of a neutral observer (JSobserver), the sensitivity of an active perpetrator (JSperpetrator), and the sensitivity of a passive beneficiary (JSbeneficiary), the latter three being often combined to JSothers. As a measurement, Schmitt et al. (2010) suggested simple self-report scales that cover each perspective. 3 In experimental games, JS has repeatedly proven to be a strong predictor of behavior. In the dictator game, JSothers (but not JSvictim) predicted other-regarding behavior toward an anonymous and powerless other (Fetchenhauer and Huang, 2004). In the ultimatum game (Camerer, 2003 and Güth, 1995), JSothers predicted rejection rates of unequal offers, despite the fact that rejecting an “unfair” offer implied getting nothing instead (Fetchenhauer and Huang, 2004). Furthermore, JSothers determined punishment behavior in a third-party punishment game (Brandstätter et al., 1999, Fetchenhauer and Huang, 2004 and Lotz et al., 2011), as well as a third-party intervention game allowing for punishment and compensation (Lotz et al., 2011). We will use JSothers to predict reluctant and stable altruism and show that variations in other-regarding behavior across three different economic games can be explained by personality, measured several weeks before behavior. As will be explained, the game variations differed in terms of how psychologically difficult it was to exploit the situation for the sake of selfishness. 1.3. Hypotheses We expect the following pattern of behavior in our experiment, with prediction #3 being our main interest: 1. Low scores on JSothers will correspond to a lower degree of altruism (i.e., lower allocations of money to recipients in a dictator game). 2. There will be a main treatment effect, such that the “easier” it becomes to be selfish (as defined by the treatment variations in the economic games and as explained below), the less money receivers will get on average. 3. Capturing individual differences in reluctant altruism: JS and type of game will interact, such that predictable subsets of dictators (low JS) will take advantage of the “ease” of being selfish (from #2), while others (high JS) will remain altruistic. Therefore, we expect people high in JSothers to be more immune to situational variations. Thus, the relationship hypothesized above (#2) is enhanced for people low in JSothers and eliminated for people high in JSothers.