Beyond “to Share or Not to Share”

Fairness Can Lead to Ungenerous Behavior

People sometimes prefer equal outcomes in which everyone gets less over unequal outcomes where everyone gets more (Dawes, Fowler, Johnson, McElreath, & Smirnov, 2007; for review, see Cooper & Kagel, in press). This is inconsistent with the idea that fairness functions to promote generosity. Proponents of these accounts could argue that much of this waste in the name of equality happens when the person who chooses equality has less than others and that these results may provide better evidence of envy than fairness (for extended discussion, see Shaw & Olson, 2012). Although this is true, recent research has demonstrated that people will also waste resources to keep things equal even in the absence of envy.

Shaw and Olson (2012) demonstrated that 6- to 8-year-old children endorse fairness over generosity; to avoid inequality, children were willing to discard a resource in the trash rather than generously giving it to others (see Fig. 1). In one study, children were asked to distribute five toy erasers to two recipients. Children were told that each recipient had received two erasers and were then asked whether they should give the remaining eraser to one of the recipients or throw it away. Children opted to throw the remaining eraser in the trash rather than share unequally (for a similar result in which children refused to share an additional resource to avoid inequality, see Kenward & Dahl, 2011). Shaw and Olson (2012) replicated these results in urban South Africa, suggesting they are not specific to relatively wealthy American societies; it is worth nothing, however, that South Africa is relatively Westernized. In another condition in which two extra erasers remained, children gave an eraser to each recipient, which demonstrates that children were not throwing the resource away because of personal envy (they also willingly created inequality when the recipient who was getting more had done more work). In addition, children threw erasers away to avoid inequality even when they were told that the two recipients did not know each other and would not know what the other received. This result suggests that children were not endorsing fairness because of explicit reasoning about upsetting the disadvantaged recipient. In another study in which five erasers were being divided between the child participant and another child, children were even willing to throw an eraser in the trash that could otherwise go to themselves to keep the distribution fair (see also Blake & McAuliffe, 2011). These laboratory dilemmas reveal that children will waste resources in the name of fairness.

OPEN IN VIEWER

Fig. 1.

Percentage of participants from Shaw and Olson (2012) who opted to give or throw away an eraser that could go to others (third party) or themselves (first party). In the third-party-equality condition, two nonpresent children had each received one eraser and participants were asked if they should give one additional eraser to each child or throw the two erasers away. In the third-party-inequality condition, two children had each received two erasers and participants were asked if they should give the additional eraser to one of the children or throw it away. The third-party-inequality-unknown condition was the same as the third-party-inequality condition, except that children were told explicitly that the two recipients did not know each other and would not know what the other received. The first-party-inequality condition was the same as the third-party-inequality condition except that the participant was now one of the two recipients and was asked if he or she should keep the additional eraser for themselves or throw it away.

Although adults may not throw resources away, they are similarly wasteful in the name of fairness. Shaw and Knobe (2013) demonstrated that adults value fairness over generosity when rewarding employees. Participants preferred not to give a raise of a dollar an hour to only one of two equally hard-working employees if there was no option of giving a raise to both (participants were told institutional rules require raises to be given in dollar increments), even when they were told that neither employee would know what the other received. In line with these data, adults seem to have similar intuitions in real world cases; in some contexts, they prefer income distributions in which the average earning is lower overall if there is less inequality (Lam, Schaubroeck, & Aryee, 2002; Mitchell, Tetlock, Mellers, & Ordóñez, 1993; Sweeney, 1990). If fairness concerns were designed to promote generosity, it is unclear why they would lead people to waste resources that could go to others.

Fairness Undermines Favoritism, Which Is Important for Cooperative Reciprocity

Shaw, DeScioli, and Olson (2012) investigated the second prediction of the generosity accounts: Fairness concerns should cause a person to like those who are generous, especially those who are selectively generous to the person him- or herself. Six- to eight-year-old children were presented with two distributors who shared toy erasers between the child participant and another nonpresent child. The children were asked to choose which distributor they liked better. The two distributors were equally generous (each giving four erasers), but one of them showed favoritism by sharing more erasers with the participant (giving all four to the participant) and the other shared equally between the two children (giving two to the participant and two to the other child). If a sense of fairness was designed to promote cooperation, then children should overwhelmingly choose the distributor who showed them favoritism, because even without a sense of fairness, other accounts of cooperation (Trivers, 1971), as well as pure self-interest, predict liking those who give one more. However, children were less likely to choose the person who favored them when favoritism conflicted with fairness (for replications and controls that suggest children understood the procedure, see Shaw et al., 2012). In another condition where no one shared equally, children overwhelmingly selected the person who gave them more.

The authors further found that children sometimes value equality over generosity. They presented another group of children with two distributors, one who shared equally between two third parties (giving two erasers to each recipient) and one who was more generous but shared unequally (giving two erasers to one recipient and six to the other). Here, children preferred the distributor who shared equally to the one who was more generous (Shaw et al., 2012). In the real world, people respond very negatively to favoritism in varied forms such as lobbying, cronyism, and quid pro quo transactions (Kwon, 2006; Loewe, Blume, & Speer, 2008; Pelletier & Bligh, 2008; Reich, 2012). Relationships maintained through favoritism are a great instance of reciprocal generosity because both parties gain, but these relationships are still called unfair. Accusations of unfairness may lead people to avoid engaging in favoritism such as reciprocity and cronyism.

Proponents of the generosity accounts could argue that although fairness does undermine generosity in some contexts, these are rare exceptions; over the long term, fairness does function to make people generous to others. One could accommodate the data presented here, which are inconsistent with the generosity explanation, if fairness uniquely explained generosity. However, several models demonstrate that generosity can be explained through the benefits of reciprocity, in which people avoid noncooperators and selectively associate with cooperators (Delton, Krasnow, Cosmides, & Tooby, 2011; Panchanathan & Boyd, 2004; Trivers, 1971). Yet these models include no tendency toward fairness—an equal division of resources. The fact that generosity is already explained by these other models makes it more difficult to accommodate data suggesting that fairness concerns sometimes interfere with generosity and instead lead to destruction of resources and undermine discriminative favoritism. Because fairness does not uniquely predict generosity, and what it does uniquely explain does not seem well designed for promoting cooperation through generosity, it becomes difficult to argue that fairness is for promoting generosity. So then, what is the function of fairness?

The Impartiality Account of Fairness

One recent suggestion is that fairness functions as a way to signal impartiality to others, in order to avoid third-party condemnation (Shaw & Olson, 2012). This impartiality account is consistent with the data reviewed thus far. An account of fairness based in signaling impartiality can explain why children throw away usable resources to uphold fairness and why they choose fair distributors over those who might be better cooperators toward them. Although fairness can work against discriminative generosity, being unfair could be construed by others as demonstrating partiality and thus may incite condemnation and fighting between group members. If fairness is about avoiding condemnation from others, then people should continue to be fair even at the expense of cooperative opportunities in some contexts.

What predictions does the impartiality account make? First, if fairness is for signaling to others, then people should be more willing to be unfair when others will not know they have been unfair. Data with both children and adults confirm this prediction (Andreoni & Bernheim, 2009; Dana, Weber, & Kuang, 2007; Levitt & List, 2007). In addition, Shaw et al. (in press) recently found that children are much less willing to discard resources in the name of fairness if the experimenter will be unaware that they have been unfair. However, the importance of signaling to others is also important in cooperative accounts of fairness, so what unique predictions does the impartiality account make?

Under this view, factors that make unequal distributions seem less partial should correspondingly be viewed as less unfair. Consistent with the notion that fairness is tied to impartiality, adults think inequality is acceptable if it is achieved using an impartial procedure (Kimbrough, Sheremeta, & Shields, 2012; Tyler, 2000). In addition, children demonstrate a willingness to pay costs in order to uphold fairness at approximately the same time developmentally (by approximately age 6) that they understand that one’s alliances and partiality can bias decision-making (Mills & Grant, 2009). The impartiality account predicts that inequality per se should not be seen as unfair, only inequality that could seem to be predicated on some form of partiality, such as inequality based on individual allegiances. If the inequality is based on some factor that is commonly acknowledged in one’s society as a reason for unequal distributions (e.g., need or work), then the action should be viewed as less partial and thus less unfair. Essentially, once one knows which actions evidence partiality, one will likely consider these actions unfair. Future work should investigate these possibilities.

Why Would People Care About Appearing Impartial?

If the function of fairness is to avoid the appearance of impartiality, then one question remains: Why should people care about being impartial? Partiality toward one’s allies and those who show generosity to one has a straightforward explanation (Tooby & Cosmides, 1996; Trivers, 1971); what is mysterious is why one would ever be impartial (DeScioli & Kurzban, 2009b, 2013). On the surface, impartiality is a bad strategy; it seems a person would do best by delivering benefits only to allies and expecting the same from them (Tooby & Cosmides, 1996). However, such open displays of favoritism may cause conflict with nonallies or less highly ranked allies who may see such new alliances as a threat. This might explain why, although people rank their friends, they do not want their low-ranked friends to know where they are ranked (DeScioli & Kurzban, 2009a). Indeed, if people become envious because others get more resources, or see unequal resource distribution as a way for others to strengthen or initiate new alliances, they may attempt to curtail such behavior by condemning it. Avoiding such condemnation is important because it can lead to hostility, ostracism, and even violence (DeScioli & Kurzban, 2013). In addition, people probably maintain some public impartiality to avoid creating aggression between in-group members because such in-group fighting may be antithetical to the pursuit of individual goals. Perhaps fairness evolved through natural or cultural selection, or a mixture of the two (Richerson & Boyd, 2004), as a mechanism for avoiding others’ negative reactions to partiality.

Of course, environmental circumstances and culture influence how much people endorse fairness. We know from previous work that Westerners tend to value fairness more highly than people from other cultures (for review, see Henrich, Heine, & Norenzayan, 2010). This makes sense because cultural circumstances such as scarcity, conflict, and competition are likely to undermine the importance of signaling one’s impartiality to others. Because fairness specifically undermines alliance building, factors that increase the importance of alliance building should cause people to place less value on signaling their impartiality, because shoring up alliances will be more important. In support of this idea, when children are in a competitive context, they value favoritism more strongly than fairness (Shaw et al., 2012). Although there has been little research on the specific role of impartiality in resource sharing cross-culturally, there has been considerable evidence to suggest that even people in small-scale societies value impartiality when it comes to morality and punishment (Beckett, 1967; for review, see DeScioli & Kurzban, 2009b). One’s culture may influence judgments about the situations in which one can give unequally without indicating partiality. Future work should investigate how people’s concerns with fairness are calibrated by environmental and cultural factors.

In this article, I have suggested two things. First, fairness concerns function to signal impartiality to others; this account explains current data and makes testable predictions that can guide future research. Second, generosity and fairness are each a unique puzzle. Recognizing this, and keeping things separate, should make it easier to see how the pieces of each puzzle fit together.

Recommended Reading

Cooper, D., & Kagel, J. (in press). (See References). A review of economic research on fairness, or what economists call inequity aversion.

DeScioli, P., & Kurzban, R. (2013). (See References). A theoretical paper that discusses the connection between alliances and impartiality.

Shaw, A., & Olson, K. R. (2012). (See References). A review of recent work on fairness in children with an empirical demonstration that children will discard resources to be fair.

West, S. A., El Mouden, C., & Gardner, A. (2011). 16 common misconceptions about the evolution of cooperation in humans. Evolution & Human Behavior, 32, 231–262. An excellent introduction to the evolutionary study of cooperation that corrects many common misunderstandings about how to interpret results from psychological and economic studies of resource sharing.