Disentangling self- and fairness-related neural mechanisms involved in the ultimatum game: an fMRI study

Abstract

Rejections of unfair offers in the ultimatum game (UG) are commonly assumed to reflect negative emotional arousal mediated by the anterior insula and medial prefrontal cortex. We aimed to disentangle those neural mechanisms associated with direct personal involvement ('I have been treated unfairly') from those associated with fairness considerations, such as the wish to discourage unfair behavior or social norm violations ('this person has been treated unfairly'). For this purpose, we used fMRI and asked participants to play the UG as responders either for themselves (myself) or on behalf of another person (third party). Unfair offers were equally often rejected in both conditions. Neuroimaging data revealed a dissociation between the medial prefrontal cortex, specifically associated with rejections in the myself condition, thus confirming its role in self-related emotional responses, and the left anterior insula, associated with rejections in both myself and third-party conditions, suggesting a role in promoting fair behavior also toward third parties. Our data extend the current understanding of the neural substrate of social decision making, by disentangling the structures sensitive to direct emotional involvement of the self from those implicated in pure fairness considerations.

Fig. 1

(A) Four conditions were employed: the first two conditions refer to the UG, whereas the last two conditions refer to the FW. In the first and third conditions, participants' decisions were related to themselves (myself trials), whereas in the second and the fourth condition, decisions were related to another person (third-party trials). (B) Trial structure.

Fig. 2

Behavioral results. Rejection rates are plotted as a function of gain in both UG (black circles) and FW (white triangles) tasks and myself and third-party conditions.

Fig. 3

Surface renderings of the functional contrasts testing regions exhibiting a larger neural activity when subjects were engaged in UG rather than FW.

Fig. 4

(A) Sections displaying the functional contrast testing ultimatum game > free win (blue activation) and the contrast testing rejections > acceptances (yellow activation). (B) Sections displaying the functional contrast testing the interaction term (green activations) and myself > third party (violet activation). (C) The parameter estimates associated with representative voxels of the activated areas are displayed together with 95% confidence intervals (for AI, we choose the local maxima obtained when testing rejection > acceptances). Red bars refer to offers addressing oneself, whereas cyan bars refer to offers addressing a third party.