The bandwagon effect and less interest in election results due to poll exposure; an online experiment

We tested in an online experiment how pre-election polls affect voters. We found that votes for the most popular option increased on average by seven~(\%) when polls were shown, feeding from both minority options and options with intermediate popularity. This bandwagon-effect was consistent across different electoral systems and different political issues. Participants that voted for majority options after seeing the polls took more time to vote and showed more interest in the final results. In our experiment participants voted on the distribution of 1200\$ between politically active organization. Our experiment is unique in that it offers both, a highly controlled environment to study voting and voting options that are relevant to participants because of real-world political consequences.