Neural basis of egalitarian behavior

Abstract

Individuals are willing to sacrifice their own resources to promote equality in groups. These costly choices promote equality and are associated with behavior that supports cooperation in humans, but little is known about the brain processes involved. We use functional MRI to study egalitarian preferences based on behavior observed in the "random income game." In this game, subjects decide whether to pay a cost to alter group members' randomly allocated incomes. We specifically examine whether egalitarian behavior is associated with neural activity in the ventromedial prefrontal cortex and the insular cortex, two regions that have been shown to be related to social preferences. Consistent with previous studies, we find significant activation in both regions; however, only the insular cortex activations are significantly associated with measures of revealed and expressed egalitarian preferences elicited outside the scanner. These results are consistent with the notion that brain mechanisms involved in experiencing the emotional states of others underlie egalitarian behavior in humans.

Fig. 1.

Sample initial screenshot (Upper Left) shows own income at bottom and income of each other group member at top. During decision phases, income beneath silhouette changes to reflect incomes that will result from the subject's actions. The timeline for each trial (Lower Left) included an initial phase where randomly drawn incomes for each group member were shown, three decision phases in which the subject sees a target group member's income and then chooses whether or not to give or take tokens from the target, and an outcome phase where resulting incomes for each group member are shown. To identify ROIs, we regressed voxel activations on the convolution of six time-varying covariates with heights depending on initial own income, initial group equality, target “richness” (target income minus mean group income among those who earn more than average), target “poorness” (mean group income minus target income among those who earn less than average), change in one's own income, and change in group equality. Consistent with previously published work (1), scanner subjects tend to give tokens to low earners (Upper Right) and take away tokens from high earners (Lower Right). Vertical lines indicate SEMs.

Fig. 2.

ROIs in the insular cortex where activations correlated significantly with change in group equality when resulting incomes were shown at the end of each trial. Between subject activations in Insula ROI 1 predict a self-reported measure of egalitarianism (first row, Center). They also predict egalitarian behavior as measured in a series of dictator games with different multipliers (first row, Right). Insula ROI 2 (second row) shows the same pattern, but the relationships are not significant. Two ROIs in the vmPFC (third and fourth rows) show no relationship with egalitarianism measured outside the scanner. Lines and P values based on Huber regression.