ریاکاری اخلاقی، قدرت و ترجیحات اجتماعی

Abstract We study how individuals adjust their judgment of fairness and unfairness when they are in the position of spectators before and after making real decisions, and how this adjustment depends on the actions they take in the game. We find that norms that appear universal instead take into account the players’ bargaining power. Also, individuals adjust their judgments after playing the game for real money, when they behaved more selfishly and only in games where choices have no strategic consequence. We interpret this possibly self-deceptive adjustment of judgments to actions as moral hypocrisy. This behavior appears produced by the attempt to strike a compromise between self-image and payoffs, so as to release oneself of one's responsibility for selfish behavior.

Introduction In public, most individuals promote social norms based on fairness and derive utility from being perceived as fair (Benabou and Tirole, 2006 and Andreoni and Bernheim, 2009). Parents encourage children to be generous; politicians emphasize dedication to serving others; businessmen promote corporate social responsibility. Real behavior may, however, reveal a different side of human nature. For example, individuals avoid beggars by changing the side of the street; they destroy the resources or production of others by envy (Mui, 1995 and Charness et al., 2014) or for the joy of destruction (Zizzo and Oswald, 2001); the power of public office leads some politicians to use it for their personal gain (Aidt, 2003). How do people reconcile their stated norms of fairness and the temptation of more selfish actions that may alter their perception as fair people? In this paper, we study how individuals try to maintain consistency between fairness judgments and real actions, and how they take into account the situation to adjust judgments on norms that appear universal. We test how much they maintain their image of fairness by adjusting their actions to suit their moral judgments, and instead how much they adjust their judgments and manipulate norms to justify their actions. Thus, the study of this interaction belongs to the study of moral hypocrisy (Batson et al., 1997 and Batson et al., 1999), defined as the motivation to appear moral to oneself and to others while avoiding the cost of acting morally. Studying moral hypocrisy from an economic perspective is important for several reasons. First, it helps understand how moral reflection influences economic behavior. We may assume that behavior follows given social norms. Here instead we investigate an aspect of the construction of norms by using a dynamic perspective. Studying moral hypocrisy means investigating one aspect of the co-construction of judgments and actions focusing on norm manipulation. Second, studying moral hypocrisy contributes to the understanding of the role of self-image in economic behavior (Bernheim, 1994, Bodner and Prelec, 2003, Benabou and Tirole, 2006, Ellingsen and Johannesson, 2008 and Ariely et al., 2009). In our experiment we analyze an intrapersonal game by which self-serving individuals may engage, possibly unconsciously, in self-deception to reconcile their judgments with their behavior and keep a good perception of the kind of person they are. Finally, understanding moral hypocrisy contributes to explain why pro-social behavior is not more developed and which incentives could encourage pro-sociality. If greedy individuals who care about looking fair use hypocrisy to keep a high self-esteem while reaping the monetary benefits of acting selfishly, increasing the cost of moral hypocrisy and norm manipulation may encourage them to behave less selfishly because of reputational concerns We adopt the methodological view that both actions and judgments constitute social behavior. We expect that individuals strike a compromise between monetary consequences and reputation, understanding that judgments are evaluated in view of actions, and actions are interpreted in view of judgments. Thus, to understand social behavior we need to measure both. Precisely, in our experiment, individuals participated in two consecutive sessions. In the first session we elicit their judgments regarding the fairness and unfairness of all possible shares between two hypothetical players in three different scenarios. The scenarios correspond to Dictator, Ultimatum and Trust games. Eliciting the individual's judgments of fairness and unfairness in terms of intervals rather than point estimates indicates the individual's moral vagueness and constitutes a novelty of our approach. A second novelty is that, instead of eliciting judgments in the position of impartial third parties who would have to make decisions for others (as in Konow, 2000 and Konow et al., 2009, for example), participants have to report their judgments in the position of partial spectators.1 They evaluate the fairness and unfairness of all possible shares in the perspective of an advantaged player and a disadvantaged player, successively, without knowing that they will have to make real decisions for themselves in the future. This information allows us to open a different perspective on the study of the effect of strategic environments on stated social norms: eliciting partial judgments helps understanding whether power influences the perception of justice, when we consider the weaker side. Thus realism, as well as selfishness and self-deception, may influence social norms. We find that this is the case. One week later, the same individuals are invited to play the Dictator, Ultimatum, and Trust games for real. In order to analyze whether actions motivate individuals to adjust their judgments, after making their decision subjects have to again report a judgment on the fairness and unfairness of all possible shares in the same three scenarios. This design allows us to measure both (i) how the actions in the second session diverge from the fairness judgments and hypothetical choices during the first session and (ii) whether the fairness judgments during the second session conform more to the initial statements or to the actions taken. We are also able to determine (iii) whether actions are more consistent with initial fairness judgments when the other player is not passive (i.e. in the Ultimatum game) than when he cannot react (i.e. in the Dictator and the Trust games). We observe clear evidence of moral hypocrisy: later actions violate fairness judgments made before individuals know their role, and norms stated later are manipulated in the direction of the actual decisions by more selfish players. The first findings alone would not characterize moral hypocrisy, as good faith in reporting judgments could simply be followed by a lack of willpower. But in most cases, individuals increase the range of shares declared fair after playing the game for real money compared with their initial statement. The discrepancy between hypothetical and real behaviors is larger in games where real behavior has no strategic consequence (Dictator and Trust) than in games where the other player can react to the decision (Ultimatum), so the subject has to take into account consequences of his choices. While the fairness statement in the Dictator scenario is based on an ethical judgment, the strategic dimension of the Ultimatum is immediately perceived (see also Charness and Gneezy, 2008). By using scenarios that vary the bargaining power we show that the adjustment of judgments is influenced by relative power and that both sides, the one in the advantageous and the one in the disadvantageous position, accept the fact that allocations are biased in favor of the powerful. Moral hypocrisy is not uniform among individuals: those who adjust their judgments to their action the most are also those who behave more selfishly, and whose hypothetical and actual decisions differ the most. This confirms that although it is rational to make selfish decisions from an economic point of view, individuals care about their self-image. By being hypocritical, they both pursue their self-interest and try to keep up appearances of pro-social motivations. However, the most selfish players do not adjust their judgments sufficiently to judge their decision as being fair, having to assume the violation of the fairness norm. In the remainder of this paper, Section 2 reviews related literature. Section 3 describes the experimental design and procedures. Section 4 presents the hypotheses we test, and Section 5 develops the results. Section 6 discusses our results and concludes.

Discussion and conclusion Our results suggest the general rule that the practical implications of actions affect the fairness judgments, and vice versa. This intrusion of strategic evaluation into normative settings occurs in two ways. First, norms that appear abstract take into account the bargaining power of the two sides. Players in the advantageous position anticipate how future behavior might deviate from tight moral standards, and thus choose to make them less stringent beforehand; for example, their judgments deviate from the norm of equal sharing. Power works the other way too: individuals in weaker positions also anticipate that the stronger side will take advantage of the position, and they adjust the moral judgment to this prediction, declaring as fair unequal allocations that favor the opposite side. Thus, power intrudes in ethical judgments, in the mind of the weak and the strong, and bends the norm in favor of the powerful. Second, people employ principles of justice in self-serving ways (Konow, 2000, Cappelen et al., 2007, Cappelen et al., 2011, Babcock et al., 1995 and Rodriguez-Lara and Moreno-Garrido, 2012). Controlling for their beliefs about others’ behavior, individuals adjust the range of fair shares and unfair shares after playing the game for real money compared with the initial statement they gave when the criteria of fairness and unfairness were elicited as universal but inconsequential norms. Moral hypocrisy is used as a tool to manage the tradeoff between the immediate convenience of the actions and the conflict these actions create with judgments. It balances the need to maintain a positive self-image and the convenience of a present choice. There is experimental evidence that people not only judge intentions, independently of outcomes, but also anticipate others doing so (Rabin, 1993 and Charness and Rabin, 2002). But then, if people understand the importance of the evaluation of intentions given by others, it is all the more important for them to adjust their stated judgments to their actions to send a signal that reassures the others of the fairness of their intentions. The size of the adjustment is reduced when another player has behaved selfishly (i.e. in the Trust game when the sender expressed low trust). The adjustment of judgments to actions is considerably smaller when there are strategic reasons that dictate prudence and fairness in deciding transfers. The discrepancy between hypothetical and real behavior is larger when the action being judged has no further consequence (as the first move in the Dictator and the second move in the Trust game) than when it does (as the first move in the Ultimatum game). If an allocation has the strategic value of affecting future actions, then it is perceived differently already in the first session. This difference in perception is reflected in the lack of adjustment of judgments and choices in the second session, since the action was already evaluated as strategic in the first session, the fact that it has real consequences of affecting payoffs of individuals is not a novelty. The fact that individuals adjust their statements when they are the last movers but not when strategic motives are present shows that the adjustment of judgments in the direction of selfishness cannot be simply interpreted as a learning effect. Similarly, learning cannot explain the diverging changes between hypothetical and actual choices according to the level of transfers relative to the median. We show that more power facilitates the enunciation of generous ethical judgments that may be distant from real actions because individuals do not examine the situation as a strategic one when they express them. The discrepancy between judgments and acts and the subsequent adjustment to actions of fairness and unfairness statements pose the question of a systematic study of moral hypocrisy in strategic behavior. Beyond the intrusion of expectations about what others think and do in the moral reasoning, occurrence of hypocrisy is due to the fact that people build an identity when stating their initial principles and hypothetical choices. Since there is no cost for looking pro-social when roles are not assigned yet, most of them build an excessively generous image of themselves. However, our design forces people to make decisions in the second session, when acting in accord with pro-social judgments is costly; at that moment they cannot remain ignorant of their true identity as there is no opt-out option. After making choices that are usually less pro-social than the initial judgments, especially if they get the highest bargaining power, they deceive themselves. They adjust their judgments in the direction of a better alignment with their actions to keep up the appearance of being pro-social and to maintain a positive self-image. By reducing the distance between their fairness judgments and their actions, people may convince themselves that their actions do not hurt their morale since they are closer or belong to a more permissive fairness interval. Moral hypocrisy is not systematic, however, hypocritical players are those who reported initially the most generous statements and who then behaved more selfishly. In contrast, the individuals who behave more generously do not feel the need to reevaluate their judgments. Many extensions are required to better understand the role of power and moral hypocrisy in ethical judgment. One of these extensions would put participants in the position of impartial spectators making decisions for others, as third parties, after having made their partial judgments from the perspectives of each party. In this environment, eliciting what individuals think one should do would help differentiate judgments on what is right and judgments on what is fair. Finally, identifying separately the conscious and unconscious mechanisms behind moral hypocrisy and the maintenance of self-image would help to better understand its deep motivation and how incentives should be redesigned to encourage pro-social behavior. An important direction of inquiry should clarify how conscious the misrepresentation is. Our design does not allow us to test whether moral hypocrisy is conscious or unconscious. Hypocrisy may be a conscious attempt to claim morality while acting selfishly but it may also be unconscious, based on self-deception (Von Hippel and Trivers, 2011).