Self-signaling in voting

This paper presents a two-wave survey experiment to examine the impact of self-image concerns on voting behavior. We elicit votes on a ballot initiative on animal welfare in Switzerland that spurred campaigns involving widely shared normative values. We send a message to voters about scientific evidence supporting the claim that “good-hearted people tend to be good to animals.” We interpret this message as a factor that may alter the self-signaling value linked to voting in favor of the initiative. We investigate how this message affects selection and processing of information, as well as reported voting behavior. We find that the message is effective in several ways: voters agree more with arguments in favor of the initiative, are more likely to anticipate voting in favor, and do report having voted in favor of the initiative more often.