Social information and bandwagon behavior in voting: An economic experiment

Abstract

We present the first economic experiment on bandwagon behavior in voting. Subjects are given an individual endowment and vote by majority rule to either keep the endowment for private use or to donate it to a non-governmental organization. In our experiment, participants are divided at random in two treatments. Individuals in both treatments are provided with true yet diverging information regarding the approval rates observed in previous experimental sessions. Our results show that the voting behavior differs across treatments in a way that is consistent with bandwagon voting. They also confirm the importance of instrumental motives and – unlike the majority of previous experiments – find support for expressive voting motives. Finally, we assess the impact of the false-consensus effect from a theoretical perspective. We show that it has the potential to change voter behavior if it results from an anchoring bias and voters follow non-instrumental motives.