Approval voting with costly information

In many approval voting settings voters are a priori uncertain regarding their true preferences, yet can obtain this information if willing to incur some cost. This paper provides a comprehensive analysis of such model focusing in simultaneous and sequential voting. The analysis enables demonstrating that costly preference-related information acquisition changes some inherent model properties. In particular, the introduction of such cost may lead to all sorts of manipulations in the sequential case, resulting in an assortment of examples where the latter is dominated by simultaneous voting and vice versa. This, as opposed to the case where such information is freely available, where it can be proved that the two variants are truthful and equivalent. These findings suggest important implications to policy makers and the designers of voting systems.