(When and how) do voters try to manipulate?

We study strategic voting in a laboratory experiment using a Borda mechanism. We find that manipulation rates are surprisingly low, even for individuals who know that they possess superior information about the other agents’ preferences. Exploring possible explanations, we find that manipulation rates rise significantly if individuals are not only informed about the other agents’ preferences but also about their actual votes. This suggests that uncertainty plays a key role in understanding strategic behavior in elections. By contrast, distributional concerns, e.g., in the form of inequality aversion, are found to play a negligible role in our context.