Dynamic adjustments of cognitive control during economic decision making

Highlights

• Ultimatum game measures conflicts between fairness and selfishness.

• We tested for conflict adaptation effects in the Ultimatum game.

• Focus on self-interest was increased following conflicts.

• Cognitive control resolves conflicts between fairness and self-interest.

Abstract

Decision making in the Ultimatum game requires the resolution of conflicts between economic self-interest and fairness intuitions. Since cognitive control processes play an important role in conflict resolution, the present study examined how control processes that are triggered by conflicts between fairness and self-interest in unfair offers affect subsequent decisions in the Ultimatum game. Our results revealed that more unfair offers were accepted following previously unfair, compared to previously fair offers. Interestingly, the magnitude of this conflict adaptation effect correlated with the individual subjects' focus on economic self-interest. We concluded that conflicts between fairness and self-interest trigger cognitive control processes, which reinforce the focus on the current task goal.

Introduction

Humans often have to make decisions under conflict between their self-interest and social norms like fairness. An experimental paradigm that allows investigating such decisions under conflict is the Ultimatum game (Fehr and Camerer, 2007, Guth et al., 1982). In the Ultimatum game, the first player (proposer) has to split a certain amount of money between himself and the second player (receiver), who can accept or reject the offer. An important aspect of this game is that both the proposer and the receiver only receive money if the receiver accepts, but not if she rejects the offer. According to the homo economicus model, the receiver should accept every offer to maximise the monetary outcome (Nash, 1950). However, the typical empirical finding is that receivers often reject offers they consider as unfair (Knoch et al., 2006, Sanfey et al., 2003). This result can be explained by the assumption that humans integrate social norms into their decisions and therefore punish unfair offers even if this clashes with their economic self-interest. This interpretation is supported by the finding that decision times are longer for unfair than for fair offers (Knoch et al., 2006); the prolonged decision time for unfair offers is thought to reflect the time required for solving the conflict between the competing goals fairness and self-interest.

A previous study of De Neys, Novitskiy, Geeraerts, Ramautar, and Wagemans (2011) already provided evidence for a contribution of cognitive control to decision making in the Ultimatum game. This study found (a) a positive correlation between the acceptance rate for unfair offers and performance in a cognitive task requiring cognitive control as well as (b) higher rejection rates for unfair offers under conditions of high cognitive load (i.e., when participants had to perform a demanding memory task while playing the Ultimatum game) relative to low cognitive load. These findings suggest that cognitive control contributes to resolving the conflict between self-interest and fairness when participants are confronted with unfair offers.

A general model describing the role of cognitive control processes in conflict processing is the conflict monitoring account (Botvinick, Braver, Barch, Carter, & Cohen, 2001). According to this account, the occurrence of a conflict in information processing (e.g., between incompatible response alternatives or goals) leads to the activation of enhanced cognitive control processes, which allow resolving subsequent conflicts more efficiently. Conflict processing costs have often been found to be reduced following conflict trials compared to following non-conflict trials in standard interference paradigms like the Stroop task or the Simon task (Egner, 2007, Kerns et al., 2004, Notebaert et al., 2006). This facilitation of conflict resolution following conflict trials is considered to be an indicator of cognitive control and is often referred to as the “conflict adaptation effect”. In the current study we hypothesized that if unfair offers in the Ultimatum game evoke conflicts between self-interest and fairness, conflict adaptation should also occur in the Ultimatum game. The finding of a conflict adaptation effect in the Ultimatum game would provide further support for the importance of cognitive control processes in the Ultimatum game. Based on the assumptions of the conflict monitoring account, we predicted that the decision time costs for unfair offers (as an indicator of conflict processing) should be reduced following previously unfair compared to previously fair offers (as it is found in standard interference paradigms) because unfair offers should activate enhanced cognitive control processes in the subsequent trial. In addition, we hypothesized that conflict-triggered control processes should modulate the acceptance of unfair offers too: enhanced cognitive control following conflicts should either increase or decrease the acceptance rates for unfair offers, depending on whether cognitive control suppresses the fair or the selfish response alternative in the Ultimatum game (note that both an increase and a decrease of acceptance rates following conflicts would provide evidence for conflict adaptation).

At the moment there is a debate on how cognitive control affects decision-making in the Ultimatum game. The assumption of control processes that allow acting according to self-interests by suppressing the negative emotional response to unfair offers is supported by findings showing a positive correlation between the acceptance rate of unfair offers and response inhibition processes (De Neys et al., 2011). However, this hypothesis is challenged by results showing that disrupting DLPFC activity with transcranial magnetic stimulation increases the acceptance rates of unfair offers (Baumgartner et al., 2011, Knoch et al., 2006). Because the DLPFC is often thought to be involved in conflict resolution (Kerns et al., 2004), this finding may indicate that decisions are strongly guided by self-interest when the exertion of cognitive control is impaired. The data of the current study may allow shedding light on this debate. If cognitive control increases the focus on self-interested goals (as the findings of De Neys et al. suggest), then enhanced control processes should result in higher acceptance rates for unfair offers. In contrast, if cognitive control inhibits the selfish impulse to accept unfair offers (Knoch et al., 2006), we expect more unfair offers to be rejected following previously unfair compared to fair offers.