Sequential Agenda Setting with Strategic and Informative Voting

We present a two-period collective bargaining model with asymmetric information and a persistent agenda setter. Voters have private information about their policy preferences. Only upon the failure of the initial proposal does the setter have a chance to revise it and call for a re-vote. The status quo gets implemented if the revised proposal fails. Although the single-crossing condition does not necessarily apply, we establish the existence of an informative equilibrium for any number of voters and any q -rule, where the first-period voting serves as deliberation that enables the voters to signal their policy preferences. We apply the model to identify important sources that improve the political outcomes regarding individuals' welfare and policy progress. Specifically, we show that adding the first-period deliberation to the one-period model improves the setter's welfare and alleviates policy gridlock. Meanwhile, the revealed information through deliberation can also benefit the voters, provided that the setter is sufficiently constrained by the voting rule. In addition, voters' less sophistication can not only make a further Pareto improvement for themselves and the setter but can also avoid policy gridlock to a larger extent