News We Like to Share: How News Sharing on Social Networks Influences Voting Outcomes

We study the relationship between news sharing on social media and information aggregation by voting. Our context-neutral laboratory experimental treatments mimic the features of social networks in the presence of media bias to address concerns that voters getting political news via social media may become more polarized in their voting behavior. Our results suggest that these concerns are warranted: subjects selectively share news that is favorable to their party and do not account for biased news signals in their voting decisions. Overall, subjects behave as if news sharing and voting expresses their induced partisanship even though by design, their preferences have a common value component. Given these patterns of individual behavior, the welfare implications of social networks reflect the underlying quality of the shared news: with unbiased media, social networks raise collective decision making efficiency, but efficiency deteriorates markedly in the presence of media bias, as news signals become less reliable.