New behavioural and neuroscientific evidence on the development of fairness behaviours demonstrates that the signatures of human fairness can be traced into childhood. Children make sacrifices for fairness (1) when they have less than others, (2) when others have been unfair and (3) when they have more than others. The latter two responses mark a critical departure from what is observed in other species because they enable fairness to be upheld even when doing so goes against self-interest. This new work can be fruitfully combined with insights from cognitive neuroscience to understand the mechanisms of developmental change.

Fairness is a complex concept that is applied across many social contexts. A central focus for both theoretical and empirical work on fairness has been the principles governing the allocation of resources, so-called distributive justice. Specifically, multiple experimental paradigms have been developed to explore how both adults and children divide up valuable resources and respond to unequal resource allocations (for example, see Box 1). Equality acts as an objective benchmark in investigations of fairness^4,5 and provides a foundation for mature conceptions of fairness⁶. Understanding how people react to unequal distributions, all else equal (that is, in the absence of richer contextual information like neediness and deservedness⁷), provides the key starting point for unlocking the psychology of fairness in humans. Indeed, this focus on equality as an objective benchmark of fairness has led to an explosion of studies across disciplines, including psychology, economics and anthropology^8,9,10. In our review of children's developing fairness behaviour, we adopt this same focus on equality and, more specifically, on children's responses to unequal outcomes. Additionally, we identify the motivations and mechanisms that lead to equal outcomes across development.

One striking feature of adult fairness is that people are sometimes willing to pay a cost to avoid inequality. This has been shown in four main contexts (Fig. 1). First, people pay to achieve equality when they receive less than others (disadvantageous inequity aversion^8,9,11). Second, adults not only make sacrifices to be fair themselves, they also punish those who are unfair. This punishment comes in two forms, as second-party punishment, where the punisher has been directly affected by unfair behaviour^12,13 and as third-party punishment, where the punisher witnesses unfairness happening to others but is not directly affected by it^14,15. Importantly, in both forms of punishment, the punisher punishes unfairness in others at personal cost or, at least, to no immediate personal benefit, thus often overriding self-interest to enforce fairness. Finally, people reject unfair distributions even when they receive more than others (advantageous inequity aversion^8,9,11), demonstrating costly conformity to a fairness norm.

Full size image

We organize our review of children's emerging fairness behaviour around these four signatures of human fairness. This structure is grounded in behavioural game theory¹⁰ and provides a taxonomy for exploring children's decisions about fairness in four key contexts in which they must reconcile their own self-interest with a desire for fairness. This taxonomy also represents, we argue, a developmental progression that is relevant for theories of fairness. Most broadly, children show an aversion to receiving less than others first — a focus on unfairness to self — and only later engage in decisions that correct or prevent unfairness to others. Theoretically, this progression marks a key shift towards understanding fairness as a norm that should be enforced, an important marker of a uniquely human sense of fairness.

Of the four contexts described above, only behaviour broadly consistent with disadvantageous inequity aversion has been found in non-human species^16,17,18. These data suggest that an aversion to getting less than others — or, at least, to getting less than you could be getting — has deep biological roots. However, there is no evidence that non-human animals enforce fairness through second- and third-party punishment^{19,​20,​21} and virtually no evidence that they reject advantageous allocations to prevent others from getting less than them¹⁸. Punishment and advantageous inequity aversion thus constitute clear signatures of the uniquely human sense of fairness. These reactions mark a critical departure from what is observed outside of humans because they enable fairness norms to be upheld even when doing so is not motivated by self-interest. In particular, the enforcement of norms by unaffected third parties is theorized to be a key means of stabilizing cooperative behaviour in humans²². Sanctioning norm violators in this context depends on the belief that ‘others ought to be fair generally’ not just that ‘others ought to be fair to me’. In thinking about the origins of this behaviour, a critical question emerges: namely, when do children shift from being the ones sanctioned for fairness norm violations to the ones sanctioning others? The answer to this question bears on deep theoretical issues about the origins of cooperation in development and, in particular, the origins of the uniquely human desire to promote good behaviour in others.

In recent years, research has begun to examine when, during development, children begin to respond negatively to unequal resource distributions and, critically, when they will pay to avoid them (Box 2). In what follows, we summarize evidence showing that the four signatures of human fairness are present in childhood, with disadvantageous inequity aversion emerging early, a willingness to punish unfairness in others (second- and third-party punishment) emerging next and advantageous inequity aversion appearing last. We then ground this description of behavioural phenomena in a framework of potential neurocognitive mechanisms underlying the emergence of these fairness behaviours.

Research on children's understanding of fairness has a long history in developmental psychology. Early methods focused on interviews and hypothetical scenarios, which have the advantage of removing immediate self-interest and allowing children to express their judgments about fair allocations. However, these methods do not assess children's fairness behaviour when actual resources are at stake. A recent integration of methods from behavioural economics into developmental research has allowed for tests of children's fairness behaviour when achieving fair outcomes is costly. Often the decisions children face are simplified to reduce the need for considering multiple alternative distributions and to contrast specific choices in order to reveal children's motivations and priorities. There are two main classes of paradigm that have been used to study fairness behaviour in children: paradigms in which children can select equality over other options and tasks that measure children's responses to unequal allocations.

Choosing equality over other options

One common choice paradigm for children involves selecting one of two possible distributions of resources between the child and a peer^29,91. For example, children can choose either one candy each (1–1) or two for the self and zero for the peer (2–0). Children receive other choice sets as well, typically a prosocial choice (1–1 versus 1–0) and an envy choice (1–1 versus 1–2). Across these different comparisons, children choose the equal option increasingly more often with age, which has been interpreted as evidence for the emergence of an egalitarian motive. Variations on these paradigms have also allowed researchers to identify a period in development during which children actively seek an advantage over peers, even when doing so is costly, as opposed to simply avoiding a disadvantage⁵¹. Children were presented with two distribution options, one of which was equal (2–2) and one of which provided them with an advantage (1–0). Researchers found that 5–6 year olds were willing to accept a lower payoff overall to achieve a relative advantage over their peer.

Measuring responses to unfairness

The choice of an equal outcome in the forced choice tasks could be supported by several mechanisms: altruistic motives (everyone should get at least one), egalitarian motives (everyone should get the same) or advantageous inequity aversion (I should not get more). A different paradigm is needed to examine the specific mechanisms that underlie the choice of equality. The Inequity Game achieves this by making the equal outcome both costly and not altruisic. Here, children face a forced choice decision that is either equal (0–0) or unequal. Each pair of children tested receives only one form of inequality: disadvantageous inequity (DI), 1–4; or advantageous inequity (AI), 4–1. Children can choose the equal option by rejecting the unequal allocation, so that both children receive nothing, or accept the inequality. Presenting the choices in this way isolates the mechanism at work. Rejecting either DI or AI denies rewards to the peer and is thus not altruistic. Rejections also deny rewards to the self and minimize the total rewards. By comparing children's rejections of DI or AI with rejections of equal allocations (1–1), researchers have identified unique mechanisms that motivate children to choose equal outcomes: disadvantageous inequity aversion (‘I should not get less than you’) and advantageous inequity aversion (‘I should not get more than you’).

While the Inequity Game isolates children's responses to unequal resource distributions, it does not test how children react to intentional unfairness. To test these reactions, the Ultimatum Game and Third-Party Punishment Game (Box 1) have been adapted for children. For instance, in one commonly used version of the Ultimatum Game, children are told that they will be paired with a partner who will either accept or reject their offer of resources. They are then presented with all possible resource divisions and asked to select what they would like to do. Using the same method, children can act as responders and either accept or reject offers made by a partner. This method has been successfully implemented in the scanner with children and adolescents, allowing researchers to measure brain activity associated with different fairness behaviours (see figure, panel c).

Three examples of paradigms used to test fairness behaviour in children.

a, A choice task in which the child has a choice between the equal option, two for self and two for partner, or the advantageous — but lower overall — option, one for self and zero for partner. b, The Inequity Game used to test children's responses to disadvantageous (upper) and advantageous (lower) unfairness. In this task the actor has access to two handles and decides whether to accept (green handle) or reject (red handle) different allocations. c, The Ultimatum Game as presented to children in the scanner.

Research on the development of fairness has undergone a shift in recent years from a focus on children's explicit verbal reasoning about hypothetical situations to a focus on their behavioural responses in real situations (see Box 2). Drawing on paradigms from behavioural economics (Box 1), new evidence shows that children engage in sophisticated fairness behaviours much younger than previously believed.

Children avoid getting less than others (disadvantageous inequity aversion). Several recent studies have found that infants are surprised when witnessing resources distributed unequally for no apparent reason^23,24,25, and by three years of age children will state explicitly that equality is a norm that should be followed in distribution tasks²⁶. However, when young children face inequality themselves, they are measurably upset at receiving less, but quite happy to receive more²⁷. By four years of age, children will pay a cost to prevent a peer from getting more, showing the first signs of disadvantageous inequity aversion. In one task, the Inequity Game (Box 1), children could accept or reject different allocations of sweets²⁸ (Box 2). Children as young as four years of age rejected allocations that placed them at a disadvantage relative to a peer, sacrificing a small amount of candy to prevent their partner from receiving more. Nevertheless, these rejections were not automatic — children took longer to make their decisions when facing a disadvantage compared with when they faced equal trials. This finding suggests that children experience a conflict between self-interest and fairness to the self when confronted with disadvantageous inequity. Results from the Inequity Game and other experimental paradigms show that children's aversion to disadvantageous outcomes tends to grow stronger with age^28,29. Interestingly, recent evidence suggests that children are more tolerant of a disadvantage they create themselves, a prosocial act^30,31.

Subsequent work has shown that disadvantageous inequity aversion is even observed in non-social contexts (that is, when a child receives the lesser of two alternative payoffs), although the response is stronger in social situations when the larger payoff goes to a peer³². Additional work has shown that children's rejections of disadvantageous allocations are motivated by spite — the desire to deprive a peer of the better payoff³³. Most recently, this task was conducted across seven societies to characterize cross-cultural variation in the development of disadvantageous inequity aversion. Results showed that children across all societies rejected disadvantageous allocations, providing evidence that disadvantageous inequity aversion may be a universal response in humans³⁴.

Children punish those who have been unfair to them (second-party punishment). While in the Inequity Game, children make decisions about pre-existing distributions, in bargaining tasks such as the Ultimatum Game (Box 1) these distributions are actual offers made by a peer. In this context, disadvantageous (selfish) offers can be viewed as an intentional slight and the responder has the opportunity to punish the proposer by rejecting all of the rewards. Research consistently shows that with age, children become more likely to reject unfair offers. While few studies have tested young children, existing studies show that children at the ages of 4–5 years tend to accept everything, even highly unfair offers^{35,​36,​37}. By contrast, starting around 6–8 years of age, children begin to systematically reject unfair offers in which they would obtain less than half of the resource^{37,​38,​39,​40,​41,​42,​43}, a response that becomes even stronger with age^{38,39,42,44,45} (but see ref. 43). These studies point to important age differences, with the first clear signs of children being willing to pay a cost to punish unfair treatment at around 6–8 years of age.

When in the role of the proposer in the Ultimatum Game, older, but not younger children, make mostly fair offers, perhaps because they realize that the responder will not accept a disadvantageous proposal. To assess whether this change is truly strategic or represents a shift to more generous behaviour, researchers have compared children's offers in Ultimatum Games with their offers in Dictator Games (Box 1). If children are able to anticipate that the recipient will punish them for unfair offers, they should offer the recipient significantly more in Ultimatum Games (where the recipient can punish) than Dictator Games (where the recipient has no such power). Studies show that in the age range of 6 to 8 years, children offer significantly more as proposers in Ultimatum Games than in Dictator Games^{37,40,43,​44,​45}. Few studies have tested younger children, but existing studies find that younger children do not offer significantly more in the Ultimatum Game than the Dictator Game^36,37, suggesting that more strategic behaviour emerges with age.

Taken together, both children's behaviour as responders and proposers show that starting around 6 to 8 years of age, children engage in and expect second-party punishment: they will pay a cost to prevent unfair treatment towards themselves and anticipate that others will engage in this behaviour as well.

Children punish those who have been unfair to others (third-party punishment). Costly third-party punishment of unfairness requires that people combine a reaction to unfairness with a willingness to intervene against bad behaviour, even when they are unaffected bystanders. Developmental data from infancy suggest that the building blocks for third-party punishment can be found early in ontogeny: infants and toddlers dislike antisocial agents who, for example, slam a box shut that another agent is trying to open and like individuals who punish these antisocial agents^46,47. By mid-childhood, children will punish individuals for violating fairness norms, marking the emergence of costly third-party punishment^{38,48,​49,​50}. For instance, when children are presented with scenarios in which they learned of a selfish child who had refused to share candies with a partner, six-year-olds will sacrifice their own resources to take candies away from the selfish individual⁴⁸. Another study testing older children and adolescents showed that 8-, 12- and 15-year-olds punished others for unfair resource divisions, with the degree of punishment mapping onto the degree of unfairness³⁸.

Children's incipient willingness to engage in third-party punishment is part of an important shift away from a myopic focus on promoting fairness for self to a more generalized focus on promoting fairness for others. This shift requires an integration of a willingness to punish at a cost to oneself with recognition of fairness as a norm that must be followed by all. Importantly, this behaviour places children in a more active role as enforcers of cooperative norms. The emergence of costly third-party punishment in children suggests that the motivation and skills needed to promote cooperation in others are present in mid-childhood.

Children sometimes avoid getting more than others (advantageous inequity aversion). Although young children refuse inequity when they receive less than others, they seem perfectly willing — even eager— to receive more than others^28,51. However, by 8 years of age children begin to reject advantageous allocations in the Inequity Game, incurring a relatively high cost to ensure that a peer does not receive less than them^28,32,34. Children in this age group take longer to decide what to do when faced with an advantageous allocation compared with an equal allocation, indicating that children experience a conflict when faced with this decision²⁸. Although the nature of the conflict is unclear, the large reward to oneself creates a tension between self-interest and a desire either to enact fairness norms or, at least in face-to-face contexts like the Inequity Game, a desire to appear fair to the partner. Additionally, unlike rejections of disadvantageous inequality, which occur both in social and non-social contexts (that is, when the child is not paired with a partner), children exclusively reject advantageous inequity when they are paired with a social partner, suggesting that advantageous inequity aversion is an inherently social response³².

Other experimental paradigms have found a similar shift at this age in which children now pay a cost to reduce their advantage over others^29,52. Interestingly, one recent cross-cultural study suggests that this behaviour may not be universal during childhood³⁴. Experiments with the Inequity Game found advantageous inequity aversion in late childhood in the United States, Canada and Uganda, but not in India, Mexico, Peru and Senegal. This study offers further evidence that advantageous inequity aversion is tied to local norms of fairness and expectations of child behaviour that vary by culture.

These findings show that an important developmental shift occurs in late childhood: namely, a focus on ensuring that ‘I get treated well’ is matched by a desire to ensure that ‘others also get treated well’, resulting in a more generalized preference for fairness. Existing cross-cultural comparisons suggest that the exact time at which this developmental shift occurs may vary, perhaps due to the acquisition of local norms.

Synthesis of the emergence of fairness behaviours in children. Children show the signatures of our uniquely human sense of fairness: not only do they react to unfairness when they get less than others, they also reject unfairness when it is advantageous and pay to punish others for unfair behaviour. These data paint an interesting picture with regards to the time course of changes in fairness-related behaviours, with concerns for fairness to self emerging first, a willingness to punish unfairness in others emerging next and, finally, a willingness to reject advantageous unfairness emerging last (Fig. 2). This trajectory provides a clear example of why developmental studies are so useful: they demonstrate that these behaviours are dissociated across development, despite appearing to be united in adulthood. In doing so, this evidence provides a theoretical motivation for considering potentially distinct neural mechanisms that underlie these behaviours. In the next section, we review evidence for the cognitive and neural mechanisms that underlie fairness behaviours in children. We suggest candidate neurocognitive processes that may help shift children from an early-emerging self-regarding form of fairness to a later-emerging willingness to both conform to and enforce fairness norms even at cost to the self.

Full size image

Although few developmental studies exist on neurocognitive mechanisms of fairness specifically, much is known about brain development more generally (Box 3). Further, research on neurocognitive mechanisms of fairness in adults can inform our understanding of what neural changes are required to reach a mature sense of fairness. In this section, we highlight the role of neurocognitive mechanisms — both common and unique — for each of our four signature fairness behaviours.

Shared components of fairness. The four signatures of human fairness vary along different psychological dimensions but also have certain key elements in common. Children must first engage in social comparison to note the difference in quantities allocated between players and then judge the difference to be unfair or undesirable. While this ability appears to already be in place in infancy^{23,​24,​25}, it does not in and of itself explain the emergence of fair behaviours. A critical aspect of fair behaviour is the willingness to incur a cost to create equality. Such costs create a conflict between self-interest and a fair outcome. Resolving this conflict and implementing the fair outcome thus also requires behavioural control. It is well documented that behavioural control improves considerably throughout childhood, becoming increasingly proactive, flexible and self-guided^53,54. Behavioural control therefore constitutes a primary candidate to account for developmental shifts in fairness behaviours as the costs to create fair outcomes increase.

Research on neurocognitive mechanisms of behavioural responses to unfairness in adults has identified a network of brain regions that is activated during fairness decisions. These regions can be tied to the cognitive processes shared across the fairness behaviours and includes the anterior insula, the anterior cingulate cortex (ACC), ventromedial prefrontal cortex (VMPFC) as well as dorso- and ventrolateral prefrontal cortices (DLPFC and VLPFC, respectively) and striatal brain regions^{55,​56,​57} (see Fig. 3). In the context of fair decisions, these brain regions subserve specific functions, such as social comparison (ventral striatum), sensitivity to fairness and emotional responses to inequality (anterior insula), conflict between following norms and one's own desires (ACC), complying with social norms (DLPFC) and implementing fair decisions (VMPFC). Thus, fairness behaviours are undergirded by a whole network of brain regions, supporting a multitude of cognitive and affective processes that interact with one another to bring about fair decisions^{56,​57,​58}. This outline of the neural systems involved provides a starting point for identifying the role of specific mechanisms needed to implement the four signature fairness behaviours and how they change with age.

Full size image

The role of negative emotions in creating fairness for oneself. Developmental studies report that preschool children display negative emotional reactions when they receive less than a peer²⁷. Additionally, the subjective experience of envy has been shown to relate to disadvantageous inequity aversion in children as young as 6 years⁵⁹. Other work suggests that disadvantageous inequity aversion in children is motivated by spite, since children only reject a small reward when doing so will reduce the partner's larger reward³³. Having less than others thus elicits a negative emotional state, which drives the response to reject such divisions. This links well with findings in adults, where activity of the amygdala, a brain region implicated in various emotional states⁶⁰, particularly negative ones^61,62, correlated positively with subjects' willingness to reduce others' outcomes to create equality⁶³. The experience of negative emotions is thus a key component for the emergence of disadvantageous inequity aversion and second-party punishment, the earliest developmental signatures of fairness. Importantly, however, these aspects of fairness are still quite limited, as they focus on fairness towards oneself.

The role of conflict resolution, behavioural control and social norm compliance in creating fairness for oneself and others. Creating fairness entails incurring a cost, which in turn creates a conflict with self-interest. Evidence of such a conflict can be detected in children's reaction times. For instance, in the Inequity Game (see Box 2), children of all ages take longer to decide when faced with disadvantageous inequity than equity.

Resolving the conflict between self-interest and fairness can be achieved by complying with social norms, which requires behavioural control. Such conflict is evident in all four signatures of human fairness. However, the neural mechanisms of conflict resolution and, more importantly, behavioural control and social norm compliance have been most extensively investigated in the context of punishing others' fairness violations. Several studies in adults report a consistently activated system of brain regions implicated in resolving conflicts during both second-party^{57,64,​65,​66} and third-party punishment tasks^{56,67,​68,​69}. In addition to the anterior insula, this network includes the ACC and the DLPFC. Activity of the ACC is considered to reflect conflict detection and resolution⁷⁰ and, in the context of punishing fairness violations, is argued to reflect the conflict between self-interest and restoring fairness⁶⁵.

Both fMRI studies and studies using non-invasive brain stimulation such as transcranial magnetic stimulation (TMS) and direct current stimulation (tDCS) suggest that activity of the DLPFC and VMPFC is crucial for social norm compliance^{65,71,​72,​73,​74,​75}. These findings suggest that DLPFC is critical in implementing norm-related behaviour by means of exerting a modulating effect on VMPFC, a brain region implicated in integrating costs and benefits of multiple decision options⁷³. Such an interpretation of DLPFC is supported by studies of the neurocognitive underpinnings of other types of norm violations (that is, criminal acts) in a third-party context, which also suggest that DLPFC is involved in norm enforcement^76,77.

The development of behavioural control has been shown to explain the emergence of punishing norm violations. For instance, in one study, 6-year-olds claimed that they would reject an unfair offer when asked hypothetically, yet when faced with real situations they accepted such offers readily⁴⁴. This stands in contrast to older children who do as they say more often. Such a developmental dissociation between judgments of fairness and actual behaviour has been found in other contexts as well. Most notably, in Dictator Games children as young as 3 years recognize an equal split as a norm but keep more for themselves, suggesting that young children do not view fairness norms as obligatory²⁶. By about 6–8 years of age, children are more likely to follow the stated norm in this task, but compliance with the norm appears to be linked to behavioural control at least until 12 years of age⁷⁸. Finally, a recent study provides direct evidence for the role of developmental changes in DLPFC and associated behavioural control in accounting for changes in social norm compliance. Proposer offers in the Ultimatum Game compared with the Dictator Game in children aged 7 to 13 years were strongly correlated with behavioural control, and in turn supported by both the function and structure of the left DLPFC⁴⁴. Behavioural control is thus likely to be an important mechanism that supports the emergence of norm compliance required for punishment of norm violations.

Behavioural control is also required when rejecting advantageous distributions, given that a relatively large sum is sacrificed for the sake of equality. Notably, once advantageous inequity aversion emerges around 8 years of age, children take longer to make their decisions, suggesting a conflict between self-interest and norm compliance²⁸. Sacrificing an advantage and resolving this conflict in favour of norm compliance depends on an improvement or shift in the quality of behavioural control. One such developmental shift is that behavioural control becomes increasingly more proactive, flexible and internalized⁵³. This leads to the hypothesis that as behavioural control increases, fairness norms will be followed in a wider set of contexts and also applied not just to others but also to oneself⁷⁹.

The role of valuing equal outcomes in creating fairness for others. Finally, it is necessary to consider the mechanisms that allow children to incur costs on behalf of another, as is the case both in third-party punishment and in rejections of advantageous distributions. In addition to an increased ability to comply with norms, we argue that this shift can be partially explained by an increasing value placed on fair outcomes per se. This is evidenced in some countries by children around 8 years of age who show clear signs of advantageous inequity aversion and also explicitly claim that they do so because it did not seem fair to have more than someone else²⁸. The anterior insula has long been noted to be particularly sensitive to fairness violations⁶⁴. It has also been implicated in responding to the emotional experiences of others⁸⁰ and thus constitutes a region highly suited to respond to others falling victim to fairness violations^81,82. It is therefore noteworthy that, in adults, both in the context of third-party punishment and advantageous inequity aversion, the anterior insula plays a critical role, particularly with regards to its connectivity to other brain regions, such as the striatum and the VMPFC^56,63,69. We propose that studying the developmental changes in the functional recruitment and integration of the anterior insula into the neural circuitry underlying fairness behaviours is a promising avenue to explain the emergence and development of fairness in childhood.

Synthesis of the mechanisms supporting the emergence of fairness behaviours in children. In sum, a set of circumscribed mechanisms supports the emergence of distinct signatures of human fairness (Fig. 4). When taken together with behavioural evidence, neural data begin to paint a more complete picture of the emergence of fairness in childhood and provide critical insight into the mechanisms that support this process. Strikingly, the dissociation between disadvantageous and advantageous inequity aversion observed across development is mirrored by unique supporting mechanisms. To synthesize what we know so far: first, the appearance of disadvantageous inequity aversion is driven by self-centered motives of not wanting to have less than others. Second, punishment of unfairness in others appears to occur as a function of increased behavioural control, which supports norm compliance. Finally, since incurring costs to forgo an unfair advantage and punishing on behalf of others seem to rely on a heightened sense of fairness, we propose that both behaviours are likely to be coded, at least in part, in the insula. Based on adult data, connectivity between the insula, striatum and VMPFC appears to be a critical neural mechanism for both advantageous inequity aversion and third-party punishment. Connectivity between distinct neural systems (Box 3) is a protracted developmental process. This may help explain why advantageous inequity aversion, which we hypothesize to be reliant on this mechanism, emerges later in development. However, before putting too much weight on this claim, it is important to test this hypothesis with developmental samples.

Negative emotions such as anger, envy and spite drive reactions towards unfairness to self, presumably mediated by the amygdala (purple) and the anterior insula (green). Incurring costs to follow and enforce social norms and achieve fair outcomes is related to the development of behavioural control, mediated by the development of connectivity between DLPFC (blue) and VMPFC (orange). Finally, we hypothesize that an increased value placed on fairness drives reactions towards unfairness to others, which is linked to increased connectivity of the anterior insula with the striatum (red) and, more importantly, VMPFC.

Full size image

The evidence reviewed here offers an exciting look at the emergence of fairness behaviour in human ontogeny and the mechanisms supporting it. However, our working definition of fairness is admittedly a lean one: namely, costly responses to unequal resource distributions. Of course, fairness judgments and behaviours are not restricted to these cases and an important direction for future work is to understand how children make decisions about other forms of fairness such as procedural⁸³ and retributive justice⁸⁴. Additionally, we know that a suite of factors influences fairness behaviour in adults and children, including reputational concerns⁸⁵, perspective taking⁸⁶, in-group favouritism⁵⁰, as well as cues of deservedness such as merit⁸⁷ and need⁷. Integrating evidence about the extent to which these factors affect fairness behaviour with burgeoning neuroscientific data on these same aspects of human cognition (for example, reputation⁸⁸, perspective taking⁸⁹ and parochialism⁹⁰) is an exciting avenue for future work. Moreover, we know little about the extent to which inter-individual variation in different aspects of children's developing cognition (for example, empathy, numerical knowledge and working memory) influences children's fairness behaviour. Characterizing influences on fairness at the individual level will provide detailed insight into the cognitive mechanisms that support its development.

The human sense of fairness has its roots in ontogeny, yet shows striking developmental changes with age. In early childhood, children avoid having less than others (disadvantageous inequity aversion), by mid-childhood they punish unfairness in both second- and third-party contexts and by late childhood they avoid having more than others (advantageous inequity aversion). Recent neuroscientific evidence offers possible explanations for this pattern of development. Specifically, improvements in behavioural control and a growing value placed on fair outcomes may help explain how children shift from a seemingly myopic focus on ensuring fairness for self towards a more generalized sense of fairness which entails the costly enforcement of and compliance to fairness norms. More broadly, we show that the integration of evidence from developmental psychology and cognitive neuroscience can offer fundamental insights into the mechanisms supporting the human sense of fairness.

How to cite this article: McAuliffe, K., Blake, P. R., Steinbeis, N. & Warneken, F. The developmental foundations of human fairness. Nat. Hum. Behav. 1, 0042 (2017).

The authors would like to thank J. Greene, B. Güroğlu, D. Knoch, L. Somerville, and L. Young for their feedback on this review. N.S. is supported by an Early Career Research Fellowship from the Jacobs Foundation. F.W. is supported by an NSF CAREER grant.

Affiliations

1. Department of Psychology, Boston College, Chestnut Hill, Massachusetts 02467, USA.

   • Katherine McAuliffe

2. Department of Psychological and Brain Sciences, Boston University, Boston, Massachusetts 02215, USA.

   • Peter R. Blake

3. Department of Developmental Psychology, Leiden University, Leiden 2333AK, The Netherlands.

   • Nikolaus Steinbeis

4. Department of Social Neuroscience, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig 04103, Germany.

   • Nikolaus Steinbeis

5. Department of Psychology, Harvard University, Cambridge, Massachusetts 02138, USA.

   • Felix Warneken

Competing interests

The authors declare no competing interests.

Corresponding author

Correspondence to Katherine McAuliffe.