Abstract
An increase in political polarization in the USA has been reported by many researchers using different kinds of data (e.g., attitudinal, affective and behavioral measures). Here, we report the results of three incentivized experimental studies (with a total of 1842 participants) in which participants had to choose how to divide money given to them by a political opponent (decisions were made for actual money of up to $11). In Study 1, participants could choose whether to act trustworthily (i.e., to share the money evenly or keep it for themselves). In Study 2, participants had no financial incentives not to reciprocate the trust placed in them (i.e., they did not earn their own benefits from harming their interaction partner). In both Study 1 and Study 2, more participants actively harmed a political opponent than a political co-partisan or a person with an unknown political affiliation. Most strikingly, such discrimination was not only governed by disliking members of the other side, but also perceived as justified moral aggression (i.e., it was regarded as the behavior that should be chosen). In Study 3, a previously well-established intervention to mitigate affective political polarization increased the likability of opponents but did not reduce discriminatory reciprocity. In all three studies, when compared with an anonymous interaction partner participants only slightly favored political affiliates but strongly discriminated against political opponents. In this, our results were highly symmetrical: Democrats and Republicans did not systematically differ in their willingness to act fairly towards each other.