In this article, the authors investigate the relationship between local social networks and voting behavior, in the case where the voter chooses between a larger (“safe”) party and a smaller, unsafe party (the “underdog”).
Using a lab experiment with fictive parties and preferences induced by monetary incentives, the authors randomly assigned voters to a treatment where the information on others’ preferences was very similar to or different from one’s own, to a random social network to no social network signaling.
Our results showed a clear and significant difference in behavior in voting behavior between treatment groups: when participants were in the treatment with a network that consistently supported the underdog, the chance to make the risky choice was significantly higher and the vote share for the underdog increased, compared with if the network was random or non-existent. This has implications for e mechanisms of echo chambers and potential local influence on global election results.
Read more about the article on Cambridge University Press’ website