Partisan sortedness is used to signal trustworthiness

Americans overestimate the extremity of differences between, and homogeneity among, the (non-political) characteristics and preferences of Republicans versus Democrats. We propose that everyday self-presentation contributes to these misperceptions: signaling motives lead people to represent themselves as more ``sorted'', i.e. as having preferences and characteristics that are closely aligned with what is perceived as typical of their party. This leads observers to have miscalibrated (and exaggerated) perceptions of differences between the groups. We present two studies testing the hypothesis that signaling motives exaggerate public expressions of partisan sortedness. Study 1 examines how social media users' profile descriptions vary in sortedness across social media platforms using two multi-platform datasets (N=740,055; 10,224). Study 2 (N=3,013) demonstrates a causal effect of signaling opportunities on sortedness using a pre-registered economic trust game experiment where participants write a self-description. We randomize whether this description will be shared with a co-partisan who is deciding how much to trust the participant. Across both studies, we found that people present themselves as more sorted when returns to signaling in-party status are larger. Furthermore, people perceive sortedness of fellow partisans as a signal of trustworthiness - and sortedness is indeed a signal of trustworthiness, but only in the absence of signaling motives. Together, these findings shed light on how misperceptions of the extent of differences between political groups may be rooted in strategic self-presentation practices. This work has implications for disentangling performative and substantive polarization, and designing practical and effective depolarization interventions.