Can Adaptively Rational Voters Control Public Policies?

Computational Extension of Outcome - Oriented Voting

抄録

According to the Downsian Median Voter Theorem, voters can control two competing parties in order to achieve an optimal policy outcome. However, compared with real party politics, there are several restrictive conditions in the classical Downsian spatial model. In this paper, we investigate whether voters can control public policies when the outcomes are weighted sum of positions of two parties. In addition, voters are adaptively rational in the sense that they mainly determine their vote just by observing policy outcomes. Our experiments reveal that voters can control public policies if they can abstain and show the frustration toward both parties.