The standard Median Voter Theorem says that parties announce a moderate policy. But, in the real world, this does not necessarily hold. In some cases, a party announces an extremist policy and wins the election.
In this paper, we try to explain such cases.In our model, the voters do not know the ability of parties. That is, some parties may be less able. If the less able party wins the election, then the voters' utility is reduced. We assume that each party has his own ideological position. The party, announcing the policy that is distant from his ideological position, needs coordination and so on that are costly. So the less able party hardly announces such a policy.Consequently an extremist policy has signaling effect.
In this paper, we show that there are the cases where the moderate parties with high ability announce an extremist policy. On the other hand, if each party is an extremist, he announces the moderate policy regardless of his ability. Moreover, we analyze the policies and the welfare in the equilibrium. Some derived results are counterintuitive. For example, if the less able party's ability improves, the policy in the equilibrium goes to more extreme one and has negative effect on voters' welfare.

View full abstract

This paper focuses on an optimal provision of public services for child care, considering them as local public goods. According to the fiscal competition theory, policy competition among local governments may not induce an optimal provision of local public goods. On the basis of this theoretical consequence, it examines, using an empirical method, whether or not public services for child care of local governments deviate from a level of optimal provision.
More specifically, I verify whether or not there exits strategic interaction on public expenses for child care among local governments, by using cross section data of Japanese prefectures and estimating linear reaction functions, based on the welfare competition model adopted by Wildasin (1991). From the positive analysis, it can be seen that there is “strategic complement" in the decisions making of local governments. I therefore point out that public expenses for child care can actually deviate from an optimal level.

View full abstract