Abstract
Trust is essential for mutually beneficial human interactions in economic exchange and politics and people’s social identities notably have dramatic effects on trust behaviors toward others. Previous literature concerning social identities generally suggests that people tend to show in-group favoritism toward members who share the same identity. However, how our brains process signals of identity while facing uncertain situations in interpersonal interactions remains largely unclear. To address this issue, we performed an fMRI experiment with 54 healthy adults who belonged to two identity groups of opposing political orientations. The identity information of participants was extracted from a large-scale social survey on the 2012 Taiwan presidential election. Accordingly, participants were categorized as either the Kuomintang (KMT) or the Democratic Progressive Party (DPP) supporters. During the experiment, participants played trust games with computer agents with labels of the same or the opposing political identity. Interestingly, our results suggest that the behaviors of the two groups cannot be equally attributed to in-group favoritism. Behaviorally, only the DPP supporter group showed a significant trust preference toward in-group members, which did not hold for the KMT supporter group. Consistently, neurophysiological findings further revealed that only the DPP supporter group showed neuronal responses to both unexpected negative feedback from in-group members in anterior insula, temporoparietal junction, and dorsal lateral prefrontal cortex, as well as to unexpected rewards from out-group members in caudate. These findings therefore suggest that acquired identities play a more complex role in modulating people’s social expectation in interpersonal trust behaviors under identity-relevant contexts.
Keywords: social identity, trust, political orientations, decision making, neuroeconomics
Introduction
Trust is one of the most socially sophisticated skills that are essential for mutually beneficial human interactions (Fett et al., 2014), such as in economic exchange and political dealmaking (Kosfeld et al., 2005). Moreover, trust is specifically important when facing uncertain situations (Brann and Foddy, 1987). Social identities, which reflect a person’s shared characteristics and/or values associated with the group s/he belongs to (Tajfel and Turner, 1979), seems to be capable of affecting trust behaviors toward others (Huettel and Kranton, 2012). For example, when individuals recognize each other as coming from the same social group, both the interpersonal trust and trustworthiness increase; conversely, if the investors and the trustees came from different groups, levels of trust and trustworthiness decline (Stanley et al., 2012).
The social identity theory of intergroup behavior (Tajfel and Turner, 1986) provides a plausible model to explain the effect of social identity on people’s decision making. The theory posits that individuals strive to maintain a positive perception of their in-groups (i.e., with the same identity) while showing negative orientations toward their out-groups (i.e., with a different identity). Previous empirical evidence has revealed that individuals who have strong senses of identity to their group are more likely to have a stronger sense of group commitment (Huddy, 2003) and to discriminate against an out-group individual when they were requested to distribute resources between in-groups and out-groups (Perreault and Bourhis, 1999; Chen and Li, 2009; Gummerum et al., 2009). Such identity-based in-group/out-group interaction has been linked to both social cognitive (e.g., mentalizing) (Freeman et al., 2010) and affective processes (Cikara and Van Bavel, 2014). For example, people seem to engage more deliberate mentalizing processes when splitting an endowment with an out-group member than with an in-group member in a dictator game (Telzer et al., 2015). Psychologically, people tend to create a tendency of outcome expectation regarding in-group versus out-group interactions (Harris and Fiske, 2010), likely based on past experiences and knowledge (Olson et al., 1996). When in-group members behave in ways that violate rather than comply with socially stereotypical expectancies, perceivers tend to experience greater affective disturbance and involve more cognitive processing to resolve the unexpected uncertainty (Jost et al., 2014).
It seems to be difficult, however, to directly generalize the prediction of the social identity theory to acquired identities (e.g., Huddy, 2001, 2003). The theory mainly originated from empirical evidence with assigned identities that lack the identity choice freedom for people, a critical characteristic that is fundamentally different from acquired identities. Specifically, the group cohesion induced by acquired identities in which members have the freedom to voluntarily choose their groups could be qualitatively different from that induced by assigned identities in which members had no such freedom (Turner et al., 1984). In the context of trust-based interaction, in-group favoritism seems to occur only when participants believe other in-group members will reciprocate the favor (Yamagishi and Kiyonari, 2000). However, a few of recent studies indeed found significant differences in trust patterns between different political identity groups which cannot be fully explained by in-group favoritism alone (e.g., Carlin and Love, 2013; Yang et al., 2014; Hernandez and Minor, 2015). These data altogether suggest that shared contents and structures of beliefs about trust within a group can be critical in shaping group members’ trust behaviors (Yamagishi, 2003).
Although quite a few studies have investigated the distinct neural mechanisms underlying trust reciprocity and feedback learning (Delgado et al., 2005; King-Casas et al., 2005; Baumgartner et al., 2008; Krueger et al., 2008; Chang et al., 2011; Kang et al., 2011; Aimone et al., 2014; Fareri et al., 2015; Bellucci et al., 2017), the modulatory nature of social identity upon trust behaviors and its neurobiological underpinnings remains relatively unclear. Typically, negative feedback that violate pre-existing social expectations during economic decisions (e.g., from trustees with close social distance) results in increased activation in regions implicated in affective processing [anterior insula (AI) and amygdala], mental inferences [temporoparietal junction (TPJ)], and cognitive control [dorsolateral prefrontal cortex (DLPFC) and anterior cingulate cortex (ACC)] (Aimone and Houser, 2011; Cloutier et al., 2011; Aimone et al., 2014), whereas positive feedback that violate pre-existing social expectations (e.g., from trustees with bad reputation) results in increased activation in reward-related regions (caudate) (Delgado et al., 2005). It has been recently shown that some ascribed social identities (e.g., age, sex, and ethnicity) could modulate activations during trust-related interaction within brain regions including caudate/striatum, AI, amygdala, TPJ, ACC, and DLPFC (Riedl and Javor, 2012; Stanley et al., 2012). These regions were known to be involved in reward learning, emotional awareness, mentalizing, and social evaluation (Riedl and Javor, 2012). Up to date, however, how our brains process signals of acquired identities (e.g., political orientations in this case) while facing uncertain situations under the contexts of interpersonal trust remains largely unclear.