Attacks without consequence? Candidates, parties, groups, and the changing face of negative advertising

Prior work finds that voters punish candidates for sponsoring attack ads. What remains unknown is the extent to which a negative ad is more effective if it is sponsored by a party or an independent group instead. We conducted three experiments in which we randomly assigned participants to view a negative ad that was identical except for its sponsor. We find that candidates can benefit from having a party or group "do their dirty work," but particularly if a group does, and that the most likely explanation for why this is the case is that many voters simply do not connect candidates to the ads sponsored by parties and groups. We also find that in some circumstances, a group-sponsored attack ad produces less polarization than one sponsored by a party. We conclude by discussing the implications our research has for current debates about the proper role of independent groups in electoral politics.