The Ultimatum Game is an experimental paradigm used in behavioral economics. In the Ultimatum Game, two anonymous participants (referred to as players), who typically have never met before, are randomly assigned to either the role of proposer or responder. The proposer receives a predetermined endowment from the experimenter and then has to decide how to split this endowment in any way they like between themselves and the responder. The responder then decides to accept or reject the proposed split. If the responder accepts the offer, the endowed amount is split according to the proposers’ suggested division; if the responder rejects the offer, neither the proposer nor the responder receives any money. Both players are informed about these rules prior to the start of the game, and thus, both players know the consequences when the responder accepts or rejects.
Bargaining situations such as those in the Ultimatum Game require the first mover (henceforth proposer) to make proposals to another party (henceforth responder) that the responder is most likely to accept in order to strike a deal while at the same time maximizing the payoff for the proposer. For this reason, the proposer has to form beliefs about the minimal acceptable payoff the responder is willing to accept. Simultaneously, the responder has to consider what cost they are willing to bear in order to reject offers that they consider unfair.
According to the game theoretic prediction, it is assumed that both players are only concerned with their own monetary payoffs and that monetary payoffs alone determine the utility of a given option. Given such rational players, the analytical solution of the game leads to the prediction that the proposer will split the endowment in such a way that only the smallest possible amount would be given to the responder. Furthermore, the responder has no incentive to decline the smallest amount offered, given that rejection would result in a payoff of zero – it is better to receive something rather than nothing. Based on backward induction, the proposer will derive at the insight that the responder should accept any offer that is greater than zero and would propose the smallest possible amount so as to maximize the proposer’s payoff. Whether these strategic assumptions of proposers in such situations, and the behavior of responders, are in line with the rational game-theoretic prediction was addressed in one of the first studies of the Ultimatum Game (Güth et al. 1982). The main conclusion from this study was that the game theoretic prediction was falsified. Responders regularly rejected non-zero offers, and proposers frequently offered more than the smallest possible amount to the responder. More specifically, on average proposals reflect 36.7% of the endowment, and responders rejected offers lower than 20% (Güth et al. 1982).
The general pattern of proposing more than the minimum amount and rejecting offers less than 20% has been replicated many times. Proposers most commonly offer a 50–50 split, and responders tend to reject proposals that are less than 20% of the total endowment (see Camerer 2003 for a review). Similar findings were also confirmed in a meta-analysis across 26 countries, which demonstrates that the average offer proposed is 39.4% of the total endowment, and that on avarage 16% of offers is rejected (Oosterbeek et al. 2004). These data suggest that people value other factors, such as equity over monetary compensation in the Ultimatum Game, and gave rise to a plethora of studies investigating theories accounting for the violation of the rational model based on game theory. To this end, the Ultimatum Game has been frequently used to investigate the influence of fairness considerations in situations without reciprocity, prosocial behavior and the influence of negative emotions on social interactive decision-making.
Ultimatum Game Theories: Fairness and Fear
Given that the typical behavior observed in the Ultimatum Game differentiates from game theoretical predictions on how people should behave, the obvious question why people would behave differently gained increasing attention. In particular, the concern for fairness exhibited by the players has influenced a number of theoretical models accounting for other regarding preferences. For example, the observed results informed the theory of fairness and reciprocity (Fehr and Schmidt 1999); the theory of equity, reciprocity, and completion (Bolton and Ockenfels 2000); and the theory of reciprocity (Falk and Fischbacher 2006). A common denominator in these theories is the explanation that proposer and responder behaviors are the results of responder’s preference for fair offers. Proposers anticipate this preference for fairness and as a result offer what they think responders consider fair.
From the perspective of the responder, rejection allows punishing the proposer for unfair offers (Bolton 1995). Consistent with this punishment viewpoint is that feelings of anger or spite in response to receiving unfair offers have been associated with the rejection of unfair offers (Pillutla and Murnighan 1996). The inclination to punish has been demonstrated to be mostly a social, interpersonal phenomenon, and rejection rates of unfair offers are higher when individuals interact with humans as opposed to computer partners (Sanfey et al. 2003; van ‘t Wout et al. 2006). The use of rejection as means of punishment might be explained by responders attributing intentional (unfair) action to the proposer that generates a negative emotional response. The role of emotions is also (indirectly) supported by psychophysiological data demonstrating that skin conductance responses, a measure of biological arousal, in response to unfair offers predicted subsequent rejection of these offers (van ‘t Wout et al. 2006). Moreover, neuroimaging studies have shown that brain regions associated with the processing and modulation of emotions such as disgust and internal responses such as heart rate are associated with the rejection of unfair offers by responders (Koenigs and Tranel 2007; Shamay-Tsoory et al. 2012).
This knowledge of an aversion to unfairness – and thus the anticipation of negative emotional reactions to unfair offers from responders – is shared by the proposer. Proposers who fear rejection of unfair offers will as a consequence propose (more) fair offers. However, while the behavior of the responders can be interpreted as reflecting a preference for fairness, it is difficult to distinguish whether motivations of fairness or fear of rejection due to envious responders underlies proposer behavior (Kirchsteiger 1994). It is probably both, but the degree to which both motives determine proposed offers likely differs for individual proposers and could be examined by contrasting Ultimatum Game behavior with that on the closely related Dictator Game in which a first mover unilaterally decides to split a given endowment that a second player can only accept (Kahneman et al. 1986). Hence, fair Dictator Game proposals cannot be explained by strategic motives to avoid rejection.
It should be noted that these data and theoretical explanations only apply to “standard” single-shot Ultimatum Games, i.e., games in which an anonymous proposer and responder only interact once. Methodological changes of the game and its structure, including providing (social) information about players, framing effects such as referring to players as “partners,” or interacting multiple times with the same player, all affect the behavior observed in the game.
Individual Differences in Ultimatum Game Behavior
Given the abovementioned motives explaining deviations from the rational game theoretical predictions, it is not surprising that there are individual differences in how a player in either role behaves. While some proposers in Ultimatum Games consistently opt for a 50–50 split, others propose offers that are slightly more beneficial to themselves but that come with some cost to the responder (e.g., 60%:40% or 70%:30%). Similarly, there are responders who will reject each offer less than 40–50% of the total endowment as well as responders who will accept every offer. These individual differences may provide further insight into more detailed explanations for behavior and the tension between strategic motives and social equity norms which will first be discussed from the perspective of the proposer and then from the responder.
From the perspective of the proposer, benevolence of the proposer is associated with higher offers when responders cannot reject offers as is the case in the Dictator Game. However, benevolence is not associated with the size of the offer when the respondent can reject the offer (the standard Ultimatum Game). This indicates that fairness considerations appear strategic rather than being explained by benevolence motives to account for Ultimatum Game proposer behavior (Brandstätter and Güth 2002). Besides benevolence, behaving fair and receiving payoffs are two potential sources of pleasure for proposers, and offering fairly in the Ultimatum Game can be explained by both motives, i.e., fair offers are pleasurable in terms of fairness and increase the likelihood of monetary rewards. Proposers who experience fairness as pleasurable tend to propose fair offers when responders can reject offers (Ultimatum Games) as well as when responders cannot reject offers (Dictator Games). On the other hand, proposers, who derive pleasure from larger rewards, only offer fair splits in Ultimatum Games, but not Dictator Games where they maximize pleasure from larger payoffs (Haselhuhn and Mellers 2005). This finding is mirrored by Scheres and Sanfey (2006) who show that personality traits reflecting drive and reward responsiveness were associated with proposing higher offers in Ultimatum Games but lower offers in Dictator Games, indicating a strategy focused on maximizing the likelihood of reward over maximizing the monetary amount of reward in Ultimatum Games. Furthermore, looking at physiology, testosterone and oxytocin seem to affect proposer behavior. Increased testosterone – whether freely available in the body or artificially increased– may result in less generous offers (Zak et al. 2009), whereas increased oxytocin, a hormone associated with bonding, resulted in more generous offers (Zak et al. 2007).
Individual differences in responder behavior may be associated with a propensity to experience anger and willingness to punish. For example, emotionally stable introverts and emotionally instable extraverts (indicative of having a high reciprocity orientation) rejected more unfair offers, while emotionally stable extraverts and emotionally instable introverts (indicating a low reciprocity orientation) accept more offers (Brandstätter and Güth 2002; Brandstätter and Königstein 2001). Likewise, individual differences in the general ability to detect internal bodily sensations, and which is associated with being able to detect one’s own internal negative emotional reactions, influenced rejection of unfair offers. More specifically, psychophysiological arousal in response to unfair offers predicted rejection of these offers only in people who were better in detecting internal bodily sensations (Dunn et al. 2012). Individual differences in physiology also impact responder behavior. Higher levels of the hormone testosterone and lower levels of the neurotransmitter serotonin have been associated with (impulsive) aggression, and increased rejection rates of low, unfair offers (see Montoya et al. 2012 for a review). Given the role of serotonin in the experience of anxiety and depression, research exploring the influence of depression on Ultimatum Game behavior demonstrates that individuals with depression report perceive unfair offers more negative, yet accept these offers to a higher degree compared to individuals without depression (Harlé et al. 2010).
The merit of the Ultimatum Game lies in showing now well-known deviations from rational game theory which point to strategic motives often underlying fairness and social norms. For current research the Ultimatum Game provides an interesting experimental paradigm to examine individual differences in strategic interpersonal decision-making behavior. Investigating proposer behavior sheds light on how beliefs about the others’ social preferences affect strategic social behavior. When focusing on responders, examining conditions influencing rejections allow insight into factors affecting the willingness to enforce norms of fairness. Such research will lead to a deeper understanding of motivations underlying strategic economic behavior and demonstrates the usefulness of game theoretical tasks to better understand social, interactive decision-making behavior and potential aberrations from rational predictions.
- Brandstätter, H., & Güth, W. (2002). Personality in dictator and ultimatum games. Central European Journal of Operations Research, 10(3), 191–215.Google Scholar
- Camerer, C. F. (2003). Behavioral game theory: Experiments in strategic interaction. New York: Russell Sage Foundation.Google Scholar
- Dunn, B. D., Evans, D., Makarova, D., White, J., & Clark, L. (2012). Gut feelings and the reaction to perceived inequity: The interplay between bodily responses, regulation, and perception shapes the rejection of unfair offers on the ultimatum game. Cognitive, Affective, & Behavioral Neuroscience, 12(3), 419–429.CrossRefGoogle Scholar
- Shamay-Tsoory, S., Suleiman, R., Aharon-Peretz, J., Gohary, R., & Hirschberger, G. (2012). Sensitivity to fairness and intentions of others in the ultimatum game in patients with ventromedial prefontal lesions. Journal of the International Neuropsychological Society, 18(6), 952–961.CrossRefPubMedGoogle Scholar