We conduct a theory-guided experiment where subjects are matched in groups of three and vote on a morally questionable decision. We find that the frequency of votes for this decision increases with the number of votes required for it (threshold effect). This effect persists even when only considering pivotal voters, who cannot rely on sufficiently many votes for the decision by other group members (thereby reaping a benefit without own moral costs). Our design allows us to attribute the threshold effect to guilt sharing. A series of novel treatments then identifies shared responsibility and a preference for group consensus as independent motives contributing to guilt sharing.
We conduct a theory-guided experiment where subjects are matched in groups of three and vote on a morally questionable decision. We find that the frequency of votes for this decision increases with the number of votes required for it (threshold effect). This effect persists even when only considering pivotal voters, who cannot rely on sufficiently many votes for the decision by other group members (thereby reaping a benefit without own moral costs). Our design allows us to attribute the threshold effect to guilt sharing. A series of novel treatments then identifies shared responsibility and a preference for group consensus as independent motives contributing to guilt sharing.