In an experiment designed to test for expressive voting, Tyran (JPubEc 2004) found a strong positive correlation between the participants' approval to a proposal to donate money for charity and their expected approval rate for fellow voters. This phenomenon can be due to a bandwagon effect or a false consensus effect. Both effects have been reported for voting decisions in the social science literature. Redoing Tyran's experiment and adding new treatments, we provide evidence for a false consensus effect. Following the experimental tradition in economics in not giving false feedback to participants, we are left with only weak tests for the impact of bandwagon motives and find none.