I

Inequity Aversion provided humans with a key selective advantage that has enabled population growth into even Gillian L Vale and Sarah F Brosnan boreal climates (Boyd et al. 2011). In particular, National Center for Care, Michale E. such cooperative tendencies have been vital to our Keeling Center for Comparative Medicine and cultural development, facilitating our ability to Research, The University of Texas MD Anderson accumulate complex traits and knowledge over Cancer Center, Bastrop, TX, USA time (cumulative culture), a process that is widely Department of and Language believed to underpin humanity’s extraordinary Research Center, Georgia State University, biological success. Given the selective advantage Atlanta, GA, USA wide-scale cooperation has conferred to our spe- cies, it is important to address why this capability may have evolved and the processes that may Synonyms underpin it. In this entry, we consider one mech- anism that may be essential to maintaining the Equality, Equity, Fairness, success of cooperation, inequity aversion. When considering behaviors, evolutionary sci- entists typically focus on two features: its func- Definition tion, or why it evolved, and the mechanism(s) by which it operates. Regarding the first of these, Inequity aversion in animals is a response to out- the evolutionary pathway that leads to coopera- comes that are less than, or more than, what a tion is relatively well understood. It is typically social partner receives. accounted for by one of the three theories: kin selection, reciprocity, or mutualism. The kin selection account of cooperation is that coopera- Introduction tion can evolve when cooperative acts provide a fitness advantage to related individuals, despite a Humans are an exceptionally cooperative species, cost to helper organisms. The account of cooper- regularly engaging in cooperate acts with both ation that involves direct reciprocity indicates that related and unrelated individuals. Cooperation cooperation can arise when the short-term net in humans has played a fundamental role in losses are outweighed by the long-term benefits the success of our species, yielding unprecedented of cooperative interactions, often with specific collective action and a capacity to transmit infor- partners. Finally, mutualism describes coopera- mation through vast social networks, which has tion that can result in greater immediate payoffs

# Springer International Publishing AG 2017 J. Vonk, T.K. Shackelford (eds.), Encyclopedia of and Behavior, DOI 10.1007/978-3-319-47829-6_1084-1 2 Inequity Aversion resulting from working together as compared to cooperation, (ii) replace such unfair partners working alone, without necessarily requiring with more equitable partners, (iii) encourage future interaction. Central to all theories of coop- future interactions with valuable partners by dis- eration is the cost-benefit ratio resulting from tributing rewards in an equitable manner, and/or collaboration; cooperation requires average indi- (iv) manage their reputation as a fair partner (e.g., vidual inclusive fitness to be increased, directly or by rejecting unequal but beneficial gains). Such indirectly, when cooperating with genetically sim- strategies may play an important role in long-term ilar individuals. Herein lies the challenge; for cooperative actions (Brosnan 2011), particularly cooperation to evolve, it must provide more ben- when unrelated individuals stand to benefit from efit than the actor pays in costs. reciprocity and mutualism. Of course, not all of One way in which this might be done is these benefits will appear in every species, and through a sense of “fairness,” or more precisely, such behavior could evolve without individuals an ability to recognize inequitable outcomes (see understanding the function; all that is needed is an Brosnan 2011, for further discussion of the link evolved negative response to inequitable out- between inequity aversion and fairness). Specifi- comes that causes individuals to seek out other cally, sensitivity to inequity allows organisms to partners. identify when a partner unduly benefits from col- In this entry, we take an explicitly evolutionary laborative interactions versus when partners approach to explore how inequity aversion mutually and fairly benefit. Recognizing when evolved as an adaption to meet particular environ- payoffs are equitable or inequitable can inform mental and social challenges. We first provide actors whether it is beneficial to engage in future background on the study of inequity aversion in collaboration or whether it is time to find a new humans, following which we discuss the empiri- cooperative partner. Indeed, one advantage of cal evidence for inequity aversion in nonhuman having a sense of inequity is that it might allow species and, in particular, nonhuman . We individuals, in some cooperative contexts, to rec- then consider what this tells us about the phylo- ognize when they are being cheated by receiving genetic trajectory of the inequity response. These less than they should (rather than receiving noth- results indicate that inequity aversion and cooper- ing at all; Brosnan and Bshary 2016). ation coevolved, supporting the argument that In the past few decades, there has been increas- inequity aversion is a mechanism to promote suc- ing empirical and theoretical attention on whether cessful cooperation. Following this evolutionary species recognize inequitable outcomes and view, we look at the degree to which inequity may how this is (or is not) related to species’ coopera- impact cooperation between individual organ- tive tendencies. Empirically, researchers have isms. Inequity does lead to a breakdown in coop- established that many species are averse to ineq- eration, again indicating the degree to which uitable payoffs in at least some contexts. This is inequity responses are important in cooperation. described by the term “inequity aversion,” a neg- We end with a consideration of how inequity is ative response to an unequal payoff to oneself, related to the human sense of fairness and to what relative to others that can take the form of disad- degree studying responses to inequity in non- vantageous inequity aversion – a response to human species can help us understand the evolu- receiving less than others – or advantageous ineq- tion of our sense of fairness. uity, a response to receiving more than others (Brosnan 2006). Theoretically, inequity aversion has been hypothesized to constitute one of the The Study of Inequity Aversion in mechanisms by which cooperation can be Humans maintained because it allows individuals to judge the value of their partners. Specifically, One challenge to an evolutionary approach to individuals can judge when they should fairness is that fairness is a social ideal, and it is (i) discontinue interactions with partners that impossible to study ideals in nonverbal species. take more than their share of the benefits of Therefore, in order to study this empirically, Inequity Aversion 3 researchers have operationalized the sense of fair- Comparing the to the “ultima- ness as an ability to recognize disadvantageous tum” game shows that contextual features can and advantageous outcomes relative to those of a influence how fairly humans act toward others. partner. One way researchers have empirically The mirrors the dictator game, approached this question has been to examine with the exception that the recipient can reject the the choices participants make when they are proposed reward allocation, in which case both faced with allocating resources between them- players receive nothing. This is both a form of selves and a partner. Various scenarios, or punishment and an inequity response (although “games,” borrowed from the field of rejecting means that the recipient now ends up have been used to investigate the contexts in with absolutely less than they could have had, which humans care about equity. In one such the outcome is relatively equal). In this context, game, called the “dictator game,” a resource average offers made in Western adults increase to (i.e., money) is provisioned to an individual (the approximately 30–40% (Camerer and Thaler proposer) who decides how to divide it between 1995). This indicates that when proposers can themselves and a “recipient.” In the dictator game, lose resources, proposals become more equitable, the recipient plays a passive role and has no option albeit typically still not equal. The ultimatum to influence the proposer’s offer. Thus, the pro- game mimics situations in which individuals, at poser can share resources at a cost (reduce advan- a cost, can defect from partners that cheat or tageous inequity or even create disadvantageous unduly benefit, a situation that encourages equity. inequity) or monopolize resources (maximize That human fairness judgments are linked to con- advantageous inequity). As the partner does not textual and social factors is further highlighted by have a say in the distribution, a situation is created cross-cultural studies. Interestingly, economic in which selfish behavior pays. This setup has games, such as the dictator and ultimatum been informative for establishing whether humans games, played with different societies, reveal make “fair” offers when there is no opportunity large between-group variations in the proposals for a partner to encourage fairness by reciprocity humans make (Henrich et al. 2004). Such findings or recourse. suggest that the human notion of fairness is very Dictator games have been played with children sensitive to differences in social and ecological and adults of different societies, yielding some environment as well. In particular, as is expected interesting developmental and cultural differ- if the hypothesized link between fairness and ences. In children, younger individuals act self- cooperation is correct, more generous offers are ishly, keeping all, or the majority, of rewards (e.g., proposed in societies that rely heavily on cooper- Gummerum et al. 2010). This picture begins to ation and trade (Henrich et al. 2004). change at around the age of 5, with a developmen- Researchers have also made use of the “ineq- tal shift toward acting more fairly, and children uity game” to explore humans’ reactions to ineq- will often begin to distribute the rewards equally. uitable outcomes. In this game, two participants This, and similar research, supports the proposi- take the role of the “actor” and “recipient” and the tion that children begin life acting in a rather self- experimenter provisions either equitable or ineq- centered manner before more egalitarian tenden- uitable rewards. The actor then makes a decision cies begin to set in at around 5–7 years of age. of whether to accept or reject the reward distribu- Turning to adults, the typical amount allotted to a tion, the latter of which results in both participants partner in dictator games averages around 20% receiving nothing. By including rewards that ben- (Forsythe et al. 1994) suggesting that, when a efit and disadvantage actors relative to the partner, cost is involved and repercussions are absent, investigators can establish whether subjects may humans do not maximize their absolute gains but be averse to advantageous, disadvantageous, or they are also not perfectly fair, either; instead, they both forms of inequity. The inequity game was make non-zero but nonetheless inequitable offers recently employed in a large-scale study that favor themselves. encompassing seven different societies and over 4 Inequity Aversion

800 pairs of children, finding that children in Bonobos (Pan paniscus), argued to be more all populations tested rejected disadvantage- socially tolerant than , tested on this ous rewards. Disadvantageous inequity aversion paradigm behaved more similarly to humans, appeared around the age of 4 but with variation in outperforming chimpanzees in the clumped the age of onset across the populations (Blake resource condition. That being said, more et al. 2015). Rejection of advantageous rewards, recently, chimpanzees have been shown to share in contrast, occurred only in some societies and at food more often and show greater tolerance in this a later age. Importantly, such findings have helped context than bonobos, leaving open the question to disentangle advantageous from disadvanta- of how our closest evolutionary ancestors shared geous inequity and emphasize that a self-serving food (Jaeggi et al. 2010). Nevertheless, such find- bias to avoid disadvantageous outcomes appears ings linking tolerance and cooperation have led to be a universal human trait, whereas aversion to researchers to hypothesize that social tolerance advantageous outcomes appears to be influenced over resources plays an important role in mediat- by society and culture. ing cooperation and that the low-tolerance levels Researchers have hypothesized that inequity observed in some species is a major constraint on responses evolved in tandem with cooperation their cooperative tendencies. Thus, ensuring (Brosnan 2006, 2011; Fehr and Schmidt 1999). equity, either actively by sharing or through high In particular, inequity is purported to act as a social tolerance for co-feeding, may be a key mechanism to compare one’s own outcome to factor in sustaining collaboration, whereas disad- those of others following collaborative effort. vantageous inequity aversion, coupled with com- Experimental support for this has been emerging petitive tendencies, may limit such opportunities. in the nonhuman literature (discussed below) but This supports the prediction that cooperation may several lines of evidence link fairness decisions to be maintained over the long term by distributing cooperation in the human literature as well, often rewards in an equitable manner or choosing to indicting that cooperation strengthens subjects’ work with partners who do so. motivations for fair payoffs. Similar cooperation studies show that by the Investigators have exposed children to experi- age of 3, children will correct resource inequity mental cooperation tasks very similar to those run after working together, sharing resources to a with nonhuman primates, allowing some direct greater extent after such collaboration than when comparisons to be made. In one such example, working in parallel or when simply gifted with Warneken et al. (2011) set 3-year-old children the resources (Hamann et al. 2011). Children are task of working together to gain otherwise also are more likely to share rewards evenly unavailable rewards that were either separated with those who collaborate than with free riders into two distributed piles, and thus could not be (Melis et al. 2013). These findings suggest that easily monopolized by one child, or were clumped children engage in equitable reward distribution into a single pile that could be easily monopo- based on collaboration and work effort (equity) lized. Importantly, children as young as 3 years of rather than simply ensuring equality irrespective age mostly split the resources evenly in the of collaboration. Comparative work with chim- clumped resource condition, suggesting children panzees shows no such flexible sharing according knew what was “fair.” This result contrasts with to whether others collaborate or not, indicating contexts that do not require children to engage in a potentially important species difference joint action, which reveal more selfish intentions (Hamann et al. 2011). In addition to correcting in children of this age (e.g., in the dictator game: resource inequities following cooperation, Gummerum et al. 2010). If we compare this to humans also anticipate who may serve as good cooperative studies with chimpanzees (Pan trog- collaborators. For example, at 15 months of age, lodytes), we find high levels of resource monopo- children already select partners that are fair over lization and a break down in collaboration in unfair ones (Burns and Sommerville 2014). By chimpanzees, but not children (Hare et al. 2007). adulthood, cooperative partners are selected Inequity Aversion 5 according to how generous they are, they will act treatment by others (e.g., fair or unfair) influences more generously when observed and will even whether we choose to engage with them or select compete to be more generous than each other someone else with whom to work. when future cooperation can lead to personal gain (Barclay and Willer 2007). Of course, indi- viduals need not always be aware of the future The Study of Inequity Aversion in benefit of selecting fair over unfair partners (e.g., Nonhumans 15 month olds almost certainly do not understand this) nor need they understand how such strategies Humans show strong responses to inequity, work. Instead, attraction to or a preference for fair but what about other species? Evidence accumu- or generous partners (in combination with a pref- lated over the past few decades shows that some erence to end unfair cooperative relationships) other species react negatively to receiving a less could maintain selectivity for partners that are desirable outcome than their partners. In a typical likely to provide equal or beneficial gains in future test of inequity aversion in nonhuman species, collaborations. two partnered individuals each complete a The study of inequity aversion in humans has task to receive a reward. The task is completed taken many forms, using tasks, joint activities, sequentially – with each individual taking turns – games, and vignettes, with various dependent and the rewards provisioned to each partner are measures ranging from self-reports or emotional either the same (equity condition) or different reactions to active resource allocations and the (inequity condition). If subjects are inequity rejection of absolute gains if they lead to relative averse, a pronounced reaction should occur in inequality. Key findings have emerged that pro- the inequity condition relative to the equity con- vide insight into which facets of fairness children dition. While different tasks have been used to may be biologically predisposed to pay attention study inequity aversion (and in some cases, no to and which may require some social experience task at all), the majority of studies have employed and culture. Concerning the former, that children an exchange paradigm in which subjects take, and from an early age (around 3–4 years old), and then return, an object (“token”) to the experi- from various cultures, respond to disadvantageous menter for a food reward, essentially bartering inequity across multiple experimental contexts, it. This method was originally conceived by suggests that disadvantageous inequity aversion Brosnan and de Waal (2003) and applied to is a robust response; humans, it seems, do not like brown capuchin monkeys (Cebus [Sapajus] to receive less than partners and are willing to pay apella). The authors documented higher levels of a price to avoid such circumstances. On the other rejection (unwillingness to exchange a token or hand, an aversion to receiving more than consume a reward) in individuals that were partners – advantageous inequity – follows a dif- offered cucumber (a less preferred food reward) ferent developmental trajectory, appearing later in following their partner’s receipt of grape development (around the age of 8 in Western (a favored food) as compared to their rejection populations) and is subject to strong cultural influ- rate in the baseline equity condition, in which ence (Blake et al. 2015). Equity choices also both exchanged for cucumbers. The authors depend upon contextual factors, suggesting that interpreted this result as reflecting a response to fair behavior is not employed indiscriminately. inequitable outcomes based on social comparison, Rather, sharing depends upon factors such as that is, “inequity” based on the reward disparity who the partner is and how much effort they between oneself and others, rather than a reaction devoted to a task (cooperation) and whether the to personally observed or experienced reward partner can influence payoffs (e.g., ultimatum differences. games) or not (e.g., dictator games). Such selec- This interpretation did not go unchallenged. tivity begins with choosing with whom we prefer Two alternative explanations to the authors’ con- to interact or cooperate. This indicates that our clusion were based on nonsocial responses to 6 Inequity Aversion unequal “pay.” First was “food frustration effects” personal factors, not always in consistent ways. in which the mere presence of a better reward Investigations using very similar protocols have could enhance rejections if the favored food was shown that dominance, social affiliation, and present in the inequity, but not in the equity, con- gender can all influence inequity responses in dition. The second challenge was the potential some, but not all, situations. Brosnan and col- frustration brought on by a violation of an expec- leagues, for example, have documented a stronger tation to receive a favored, and previously expe- response in dominant individuals when they were rienced, food when the subject instead receives a undercompensated relative to a lower-ranking less preferred food (i.e., an expectation based on partner, but this dominance effect was not found past reward history). More recent studies of ineq- in other studies (see Price and Brosnan 2012 for a uity, however, have controlled for these alterna- review of these data). Similarly, there have been tive explanations by ensuring that both favored cases of chimpanzees with close social ties (i.e., and non-preferred foods are visible in all study housed together for long periods) being less conditions and by the inclusion of controls for responsive to a partner being overcompensated potential frustration, typically with a condition in as compared to chimpanzees that were housed which subjects are shown a favored food before together for much shorter periods, although offering a less preferred food, and by randomizing again, there have also been studies in which no the order in which conditions are presented. Even such variation was seen. Such inconsistencies with these controls, social comparison, or inequity have encouraged researchers to search for individ- aversion, still occurs. This suggests that refusals ual differences that may account for some of this are not simply (or only) due to seeing better variation. Personality, for example, was recently rewards or any potential frustration caused by found to correlate with social inequity responses, expecting a better reward than is received. although it seems likely that relationship quality, Brosnan and de Waal’s initial study of inequity which influences which individuals will voluntar- aversion in capuchin monkeys paved the way for ily pair together for the tasks themselves, affects new research exploring such capabilities across their responses as well (Brosnan et al. 2015). the order and, in a recently emerging Other factors not yet considered likely also trend, non-primate species. At least 30 studies underpin this variation. For example, the social have now been published, and these find substan- configuration and tolerance of partnered animals tial variation across species and contexts (see may influence responses to inequity (reviewed in Brosnan and de Waal (2014) for a table including Price and Brosnan 2012), as could variation in all studies published to that date). Here we briefly subjects’ daily food intake and any social change summarize this research, focusing on what species that may occur in the group at the time of testing. show a response to social inequity and, where There may also be interactions between factors. possible, the influence of procedural and contex- For example, dominant individuals may show tual variables on the manifestation of the inequity strong inequity responses simply because they response. are more satiated at the time of testing, through prior resource monopolization, relative to less dominant individuals. Satiation likely increases The Phylogeny of Inequity Aversion the likelihood of the costly behavior of forgoing an immediate, less preferred food in inequity con- Among the great apes, chimpanzees are the most ditions. Furthermore, dominants may be used to studied species. Chimpanzees generally show receiving the best foods from humans, strength- some degree of inequity aversion, much like the ening potential inequity responses. Depending brown capuchin monkeys (see Brosnan and de upon one’s group, and indeed housing facility, Waal 2014; Price and Brosnan 2012). However, dominant individuals’ propensity to monopolize they are notable because their responses are sub- food could vary, with implications for the food ject to high levels of group and individual varia- refusal rates. One open question, then, is whether tion that has been linked to various social and inequity aversion in the context of food rewards is Inequity Aversion 7 an artifact of captivity, as forgoing an immediate (Macaca ssp.), capuchins (Cebus ssp.), squirrel reward incurs little cost where there is reliable, monkeys (Saimiri ssp.), owl monkeys (Aotus and sufficient, food provisioning. This question spp.), marmosets (Callithrix spp.), and tamarins requires investigation in natural populations and (Saguinus oedipus; see Brosnan and de Waal 2014 could be examined by documenting participation for a review of this literature, including citations levels in coordinated activities following equita- for studies on each species). Capuchin monkeys, ble and inequitable gains, for instance, looking at like chimpanzees, have been highly studied, and whether coordinated hunt participation lowers fol- also like chimpanzees, show variation in their lowing disadvantageous inequitable food returns responses, with some studies finding evidence of and exploring how participation may interact with inequity aversion while others have not. Interest- food scarcity. It will also be important to deter- ingly, among the other New World monkeys that mine in which other contexts inequity may influ- have been tested (common marmosets, owl mon- ence behavior in natural conditions, particularly keys, squirrel monkeys, and tamarins), no species outside of the context of food. has been shown to respond to disadvantageous Turning to the other great apes, comparatively inequity. However, much like in the great apes, less work has been conducted on inequity aver- the focus primarily on one species – capuchins – sion. We know nothing of gorillas’ (Gorilla spp.) with very little data available concerning other capabilities, except that they do not respond to New World monkeys makes it premature to con- inequity when rewards are “gifted” by the exper- clude that other New World species lack a imenter (i.e., in the absence of a task; see Brosnan response to inequity, particularly given the varia- and de Waal 2014). However, no response in a tion in individuals’ responses in the best studied free-feeding context is common across species. In species. The situation is even less clear in Old every primate species in which inequity has been World monkeys, for which only macaques have documented, subjects only respond negatively been tested. Both species tested thus far (rhesus when the food rewards are received subsequent and long-tailed macaques) do respond negatively to completing some sort of task; when rewards are to social contrast. More recently, researchers have handed out “for free,” the same subjects do not studied non-primate species, and initial data indi- respond negatively even when they are unequal. cate that domestic dogs, rats, and corvids respond We know little more about the other great apes, to the presence of, or the potential for, inequity with only one published exchange study focusing (reviewed in Brosnan and de Waal 2014). on bonobos and two on orangutans (Bräuer et al. Considering variability within species, one 2009; Brosnan et al. 2011). For bonobos, evidence factor proposed to explain such discrepancies, at points toward some presence of inequity aversion; least in part, is the experimental procedures however, as only five subjects have been tested, employed. The combination of work effort pro- the results are inconclusive (Bräuer et al. 2009). vided by a task and subject’s proximity to their The two studies that have examined whether partner appear to be important for the expression orangutans respond to disadvantageous inequity of inequity aversion, with studies that have distant document no such trend in socially housed orang- partners and/or no task generally reporting a lack utans (Bräuer et al. 2009; Brosnan et al. 2011). of a response to inequity, even in species and This is interesting, should it hold, as orangutans groups that show inequity in other contexts are among the least social primates, particularly in (Brosnan and de Waal 2014). Other factors, such comparison to other great apes, and rarely coop- as the relative value of the rewards may also erate in the wild. However, given this paucity of impact inequity aversion; for instance, the prefer- data in bonobos and gorillas, it is difficult to ence discrepancy between some food pairings determine how inequity responses may have (e.g., pine nuts versus sunflower seeds) may differ evolved among the apes. to those used in other studies (e.g., cucumber Experimental studies of inequity have also versus grape). While the relative value of foods focused on monkeys, including macaques is always established within studies, these values 8 Inequity Aversion still differ between studies, and recent research in In considering the relationship between ineq- capuchins shows that relative food values strongly uity aversion and cooperation, factors that com- impact inequity responses (Talbot et al. 2017). plicate the picture should be noted. First, Finally, social and individual differences impact cooperation between non-kin alone does not inequity aversion in chimpanzees, which may also appear sufficient for the emergence of inequity play a role in other species’ unfairness judgments. aversion. Cooperative breeders (marmosets and For example, the relatedness of subjects and part- tamarins), species in which both parents and, ners are rarely reported but could have a profound sometimes, older offspring work together to raise effect on whether animals are inequity averse or infants, show no reaction to inequity. This may be not. Individuals’ propensity to use social over due to the high level of interdependence between personal information may also determine males and females that renders a negative responses to inequity. Many animals collect both response to inequitable outcomes more costly kinds of information but are reported to prioritize than beneficial (i.e., the cost of finding a new personal over social information (Rieucau and mate, and risking losing one’s current reproduc- Giraldeau 2011), meaning that animals may not tive investment, is higher than the cost of inequity always attend to social contrasts. in most contexts; Brosnan 2011). One caveat here Considering the variability across species, is whether cooperative breeders may be socially what does this phylogenetic dispersal of inequity tolerant in some, but not other, situations. Coop- aversion tell us about the of disadvan- erative breeders, for example, may be sensitive to tageous inequity aversion? Most apparent is the inequitable contributions in work effort but toler- inconsistent distribution of disadvantageous ineq- ant of inequitable food returns. Second, as many uity aversion across the primate order and its studies have been conducted in highly controlled presence in diverse taxa both within the primates laboratory contexts, the extent to which the appar- and beyond (e.g., corvids and rats). From this, it is ent sensitivity to inequitable outcomes generalizes evident that convergent evolution is at play to the natural environment remains unknown. (Brosnan 2011); disadvantageous inequity has Third, given the large variation in the inequity evolved independently in various organisms, response documented within those species most likely in response to similar ecological and/or studied (chimpanzees and capuchins), caution social niches. Such pressures have been hypothe- must be used when assuming an absence of ineq- sized to include species’ sociality and degree of uity aversion in species based on just one or two cooperation with non-kin. Sociality alone does studies. Finally, it is important to keep in mind that not appear critical here as some gregarious spe- ecologically valid measurements of cooperation cies, living in large social groups, are nonetheless in natural populations could have more relevance insensitive to inequity (e.g., squirrel monkeys). to interpreting the possible relationship between Instead, recent theory has favored a link between inequity aversion and cooperation than those inequity aversion and cooperation with non-kin; derived from captive studies, which may be of course, regular exposure to non-kin relates to more artificial. This is highlighted by orangutans, species sociality/social structure, but the coopera- a semi-solitary species, that in two independent tion component seems to be important. Wild studies did not respond to inequity nor do they chimpanzees engage in coordinated hunts, terri- show high cooperative tendencies in the wild, yet tory defense, and coalitions, suggesting coopera- they will work together to gain a reward in a tion or at least behavioral coordination between captive setting, cooperating much like chimpan- group members. Likewise, the social behavior of zees and capuchins (Chalmeau et al. 1997). capuchins and macaques hints at cooperation with Overall, while it is currently the best-supported non-kin. A number of experimental studies also hypothesis, additional data are required to test the show a degree of cooperation in these species, and hypothesis that inequity aversion and cooperation even that inequitable outcomes will disrupt coop- coevolved. In particular, contrasts between eration, providing further evidence for this link. closely related species known to cooperate with Inequity Aversion 9 non-kin in the wild and those that do not, and inequitable. Key to this study was that partici- social versus solitary species, would be useful in pants, who were co-housed in the same enclosure, this regard. Another means to examine the link had to decide who worked for each reward. Thus, between cooperation and inequity aversion has conflict could arise over the unequal resources. been to establish whether inequitable outcomes The authors found that when a dominant partner directly influence the success of cooperation. To “unfairly” benefited by monopolizing the better do so requires studies in which resource distribu- reward, cooperation tended to break down. tions can be manipulated or controlled; thus, Indeed, cooperation rates when partners took while ecological valid measurements in the wild turns in gaining the better reward were two to are important, the available evidence derived from three times higher than when dominants monop- controlled experimental tests of cooperation must olized the better outcomes, and cooperation with also be considered. unfair partners broke down even in control tests in which rewards were equal. What this research suggests is that capuchins will cooperate over The Influence of Inequity on long periods even when short-term rewards are Cooperation in Nonhumans unequal as long as each partner is willing to “play fair” over the long term, yielding (roughly) equi- As discussed above, inequity aversion is hypoth- table gains to both partners overall. esized to help maintain cooperation by enabling Similar findings have been documented in organisms to do one or more of the following: chimpanzees; they select partners based on their (i) disengage from collaborating with those from previous history of tolerant sharing (Melis et al. whom they receive unfair returns, (ii) encourage 2006), reduce their cooperation when rewards can more equitable future gains by choosing equitable be monopolized by dominant individuals (Hare partners, (iii) encourage future cooperation by et al. 2007), and, when not cooperating with kin, distributing rewards in an equitable manner, and prefer to work with partners of similar rank, who (iv) manage their reputation as a fair partner by are presumably more socially tolerant partners rejecting unequal but beneficial gains. We have (Suchak et al. 2014). These findings indicate that previously described how humans readily distrib- some primates remedy inequity though partner ute rewards equitably following collaboration, selectivity, that is, by not working collaboratively select fair or generous partners, and will some- when resources are or can be monopolized and/or times reject advantageous inequitable rewards. In by taking turns gaining the beneficial but inequi- this section, we consider whether inequity aver- table outcome of collaboration. Thus, there is sion constitutes a key mechanism supporting some support for the predictions that inequity long-term cooperation in nonhumans by focusing aversion may result in (i) leaving collaborators on each of these predictions. that take more than their share, (ii) using partner Studies of inequity aversion show that some choice to manage future gains, and possibly even individuals are less likely to participate in tasks (iii) distributing rewards in an equitable manner to when they receive unfavorable outcomes relative promote cooperation between individuals (i.e., to a social partner. Inequity tasks, however, do not capuchins taking turns to gain the best reward). require cooperation between partners, which is a The last prediction indicates that inequity aver- key component of this hypothesis. As a result, sion may aid cooperation when individuals reject additional work has been done focusing explicitly inequitable, but beneficial, outcomes (i.e., advan- on cooperation and reciprocity. In one such study, tageous inequity aversion), perhaps encouraging Brosnan et al. (2006) designed methods to directly future reciprocity or gaining a good reputation to test the assumption that unequal rewards influence attract future partners by establishing oneself as a levels of cooperation. Pairs of capuchins could fair partner. Interestingly, while in no study to date perform a joint action (pull in a counterweighted have monkeys minded options that reward them- tray) to gain rewards that were either equitable or selves more than a partner, one study documented 10 Inequity Aversion chimpanzees rejecting a preferred food when a a few occasions, overall it is worth their while to partner received a non-preferred food (i.e., advan- continue the interaction (Proctor et al. 2013). tageous inequity aversion), albeit at a much lower Further indication that chimpanzees may be rate than they rejected a less preferred food when a sensitive to gaining more than conspecifics is a partner received a more preferred food (reviewed finding that chimpanzees, in a group situation, in Brosnan and de Waal 2014). counterintuitively opt to follow the low-paying Researchers have also tested nonhuman pri- behavior of a dominant chimpanzee, despite the mates on the ultimatum game. As in the human presence of a better-paying alternative (Hopper ultimatum game described above, two individuals et al. 2011). The authors interpreted this as chim- are assigned the “proposer” and “responder” panzees “conformity” to the low-paying behavior roles. The proposer offers some of the resources of the dominant chimpanzee (Hopper et al. 2011). to the responder, who can accept the offer, in An alternative explanation is that the selection of which case each receives the rewards set out by the low-paying behavior could be due to an aver- the proposer, or reject the offer, in which case sion to advantageous inequity: perhaps the costs neither receives anything. Interestingly, as would of receiving more than the model were too high to be predicted by game theory, chimpanzees appear make it worthwhile to use the better option. to accept any non-zero reward offers, suggesting Research over the past two decades highlights they are rational maximizers (Jensen et al. 2007; the link between cooperation and inequity aver- Proctor et al. 2013). This is unlike humans, who sion. Further research is required to tease apart in will reject offers below around 20% of the endow- which situations primates are indifferent versus ment. Chimpanzee proposers, on the other hand, being sensitive to others’ payoffs and how this is make both fair (Proctor et al. 2013) and selfish mediated by opportunities to cooperate, recipro- offers (Jensen et al. 2007). In one study, proposers cate rewards, and, in some cases, punish others or did not make the fair offer to their partners (Jensen manage one’s reputation. Nonetheless, several et al. 2007); however, these chimpanzees also studies now support the hypotheses that primates showed the costly behavior of delivering zero manage cooperation by leaving unfair collabora- rewards to their partner, which had high rejection tions and by being selective about which partners rates (44%), despite the availability of a second they choose. Less clear is whether nonhuman self-serving option that was only rejected by their primates actively distribute rewards fairly partners 14% of the time. This, coupled with (as could be the case of capuchins taking turns to responders’ rejection of zero offers on only 44% gain the best reward) and, related to this, the of trials, suggests that the chimpanzees only had a degree to which any nonhuman species routinely limited understanding of this task. In a more shows refusal of unfair but beneficial personal recent version, chimpanzees did prefer fair offers gains. These qualities might constitute notable in the ultimatum game more so than in a prefer- species differences between humans and other ence test that was similar to a dictator game primates, as may other cognitive abilities (i.e., (Proctor et al. 2013). Key to this study was that inhibiting one’s desire to accept the better reward fair reward options were favored only when part- or planning for future interactions). ner chimpanzees had recourse to reject offers (i.e., the ultimatum condition), whereas selfish offers were favored when partners had no such ability The Evolution of Fairness (i.e., the dictator analogue), suggesting that the chimpanzees understood the procedure. Indeed, Inequity aversion has been found in diverse spe- the lack of refusals by the responder may be cies, including rats, dogs and wolves, some birds, logical in a repeated game with a known partner; and primates. The comparative literature has pro- rather than rejecting and ending the relationship, vided important insights into the evolution of the chimpanzees could protest and encourage the pro- human sense of fairness. Knowing that inequity posers to choose the more equal outcome. There- aversion has evolved in these diverse taxa indi- fore, even though the recipient will receive less on cates that convergence, rather than evolution Inequity Aversion 11 through common descent, has been at play. References Researchers have also now established that a sense of fairness can take different forms, and Barclay, P., & Willer, R. (2007). Partner choice creates while disadvantageous inequity appears more competitive in humans. Proceedings of the Royal Society of London B: Biological Sciences, widespread across taxa, advantageous inequity 274(1610), 749–753. aversion is extremely rare. Indeed, even among Blake, P. R., McAuliffe, K., Corbit, J., Callaghan, T. C, humans, disadvantageous inequity aversion Barry, O., Bowie, A., ...& Warneken, F. (2015). The appears to be both more widespread and to ontogeny of fairness in seven societies. Nature, 528(7581), 258–261. develop at an earlier age, indicating a strong bio- Boyd, R., Richerson, P. J., & Henrich, J. (2011). The logical influence. Species found to have both cultural niche: Why social learning is essential for forms of inequity aversion come close to having human adaptation. Proceedings of the National Acad- – the “complete” sense of fairness seen in humans. emy of Sciences, USA, 108, 10918 10925. ’ Bräuer, J., Call, J., & Tomasello, M. (2009). Are apes Of course, humans sense of fairness entails other inequity averse? New data on the token-exchange par- aspects that are likely not present in other species, adigm. American Journal of Primatology, 71(2), such as explicit group norms of behavior. How 175–181. ’ humans’ advanced sense of fairness, which Brosnan, S. F. (2006). Nonhuman species reactions to inequity and their implications for fairness. Social Jus- requires advanced cognition and emotional con- tice Research, 19, 153–185. trol, arose during the course of our evolution Brosnan, S. F. (2011). A hypothesis of the co-evolution of remains a significant question for science. None- cooperation and responses to inequity. Frontiers in theless, one explanation consistent with the cur- Neuroscience, 5, 43. Brosnan, S. F., & Bshary, R. (2016). On potential links rent evidence is that fairness coevolved with between inequity aversion and the structure of interac- cooperation with non-kin, and functions to help tions for the evolution of cooperation. Behaviour, individuals maintain beneficial cooperative rela- 153(9–11), 1267–1292. tionships. It is clear that individuals benefit from Brosnan, S. F., & de Waal, F. B. M. (2003). Monkeys reject unequal pay. Nature, 425, 297–299. leaving partnerships in which they routinely get Brosnan, S. F., & de Waal, F. B. (2014). Evolution of less than the partner, but they may also benefit responses to (un) fairness. Science, 346(6207), from declining an unfair advantage. By paying the 1251776. short-term costs of rejecting an unfairly advanta- Brosnan, S. F., Freeman, C., & de Waal, F. B. M. (2006). Partner’s behavior, not reward distribution, determines geous outcome, individuals signal to their partner success in an unequal cooperative task in capuchin (and potentially to others) that they value the monkeys. American Journal of Primatology, 68, partner. If they can ameliorate their partner’s frus- 713–724. tration, the long-term relationship can continue Brosnan, S. F., Flemming, T., Talbot, C. F., Mayo, L., & Stoinski, T. (2011). Orangutans (Pongo pygmaeus)do (see Brosnan and de Waal 2014). Humans, with not form expectations based on their partner’s out- our exceptional abilities at delay of gratification comes. Folia Primatologica, 82(1), 56–70. and planning for the future, inhibition, and under- Brosnan, S. F., Hopper, L. M., Richey, S., Freeman, H. D., standing of others’ thoughts and feelings (i.e., Talbot, C. F., Gosling, S. D., ... & Schapiro, S. J. (2015). Personality influences responses to inequity theory of mind), were able to take these roots and contrast in chimpanzees. Animal Behaviour, 101, and develop the full-blow sense of fairness seen 75–87. in humans today. Burns, M. P., & Sommerville, J. A. (2014). “I pick you”: The impact of fairness and race on infants’ selection of social partners. Frontiers in Psychology, 5, 93. Camerer, C., & Thaler, R. H. (1995). Anomolies: Ultima- Cross-References tums, dictators and manners. The Journal of Economic Perspectives, 9(2), 201–219. ▶ Co-operation Chalmeau, R., Lardeux, K., Brandibas, P., & Gallo, A. ▶ (1997). Cooperative problem solving by orangutans Dictator Game (Pongo pygmaeus). International Journal of Primatol- ▶ Economics ogy, 18(1), 23–32. ▶ Game Theory ▶ Ultimatum Game 12 Inequity Aversion

Fehr, E., & Schmidt, K. M. (1999). A theory of fairness, Melis, A. P., Hare, B., & Tomasello, M. (2006). Chimpan- competition, and cooperation. The Quarterly Journal zees recruit the best collaborators. Science, 311(5765), of Economics, 114(3), 817–868. 1297–1300. Forsythe, R., Horowitz, J. L., Savin, N. E., & Sefton, M. Melis, A. P., Altrichter, K., & Tomasello, M. (2013). (1994). Fairness in simple bargaining experiments. Allocation of resources to collaborators and Games and Economic Behavior, 6(3), 347–369. free-riders in 3-year-olds. Journal of Experimental Gummerum, M., Hanoch, Y., Keller, M., Parsons, K., & Child Psychology, 114(2), 364–370. Hummel, A. (2010). Preschoolers’ allocations in the Price, S. A., & Brosnan, S. F. (2012). To each according to dictator game: The role of moral emotions. Journal of his need? Variability in the responses to inequity in Economic Psychology, 31(1), 25–34. non-human primates. Social Justice Research, 25(2), Hamann, K., Warneken, F., Greenberg, J. R., & 140–169. Tomasello, M. (2011). Collaboration encourages Proctor, D., Williamson, R. A., de Waal, F. B., & equal sharing in children but not in chimpanzees. Brosnan, S. F. (2013). Chimpanzees play the ultimatum Nature, 476(7360), 328–331. game. Proceedings of the National Academy of Hare, B., Melis, A. P., Woods, V., Hastings, S., & Sciences, 110(6), 2070–2075. Wrangham, R. (2007). Tolerance allows bonobos to Rieucau, G., & Giraldeau, L. A. (2011). Exploring the outperform chimpanzees on a cooperative task. Current costs and benefits of social information use: An , 17(7), 619–623. appraisal of current experimental evidence. Philosoph- Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., & ical Transactions of the Royal Society of London B: Gintis, H. (2004). Foundations of human sociality: Biological Sciences, 366(1567), 949–957. Economic experiments and ethnographic evidence Suchak, M., Eppley, T. M., Campbell, M. W., & from fifteen small-scale societies. Oxford, UK: Oxford de Waal, F. B. (2014). Ape duos and trios: Spontaneous University Press. cooperation with free partner choice in chimpanzees. Hopper, L. M., Schapiro, S. J., Lambeth, S. P., & PeerJ, 2, e417. Brosnan, S. F. (2011). Chimpanzees’ socially Talbot, C.F., Parrish, A.E., Watzek, J., Essler, J.L., maintained food preferences indicate both conserva- Leverett, K.L., Paukner, A., & Brosnan, S.F. (2017). tism and conformity. Animal Behaviour, 81(6), The influence of reward quality and quantity and spatial 1195–1202. proximity on the responses to inequity and contrast in Jaeggi, A. V., Stevens, J. M., & Van Schaik, C. P. (2010). capuchin monkeys (Cebus [Sapajus] apella). Journal of Tolerant food sharing and reciprocity is precluded by Comparative Psychology. despotism among bonobos but not chimpanzees. Warneken, F., Lohse, K., Melis, A. P., & Tomasello, M. American Journal of Physical , 143(1), (2011). Young children share the spoils after collabo- 41–51. ration. Psychological Science, 22(2), 267–273. Jensen, K., Call, J., & Tomasello, M. (2007). Chimpanzees are rational maximizers in an ultimatum game. Science, 318(5847), 107–109.