On Two-Period Committee Voting: Why Straw Polls Should Have Consequences

We consider a committee voting setup with two rounds of voting where committee members, who possess private information about the state of the world, have to make a binary decision. We investigate incentives for truthful revelation of their information in the first voting period. Coughlan (2000) shows that members reveal their information in a straw poll only if their preferences are in fact homogeneous. By taking costs of time into account, we demonstrate that committees have strictly higher incentives to reveal information if a decision can be made for high levels of consensus in the straw poll already. In such scenarios, members of all homogeneous and some heterogeneous juries are strictly better off when the requirement for early decisions is chosen carefully.