Accountability and Representation in Repeated Elections with Uncertain Policy Constraints

We consider repeated elections in which two long lived parties have fixed preferences over a two dimensional policy space. In each period the government is constrained in its policy selection decisions. The public never observes this constraint making control of the government quite difficult. In pure strategies the voter’s preferred equilibria have the following features: (i) parties

favored by the current constraint enact the voter’s constrained optimum and are reelected while parties not favored by the current constraint distort policy in their preferred direction, (ii) reelection may follow the enactment of policies that seem biased in favor of the governing party (iii) the sets of polices that will result in reelection differ across parties. Voters prefer equilibria involving partial control to repeatedly selecting the best ex-ante party. In addition, if the voters

can randomize (as in mixed strategies) then full control is possible.

For more information, contact Antonio Merlo.