The effect of voter control on public policy

Abstract

In democracies, the public make decisions that affect policy. In some situations, these decisions are only indirectly related to policy: voters choose an elected executive, who then appoints an unelected policy-maker, who in turn decides policy. In other situations, these decisions are more directly related to policy: voters bypass the executive and elect the policy-maker directly. In still other situations, voters bypass the electoral process altogether, deciding policy for themselves. Do these different configurations matter? While centuries of debate over the merits of democracy have been premised on the assumption they do, there is still limited evidence that voter control affects policy. In this dissertation, I provide three empirical tests of the claim that voter control institutions matter for public policy. The first empirical chapter examines what happens when voters lose control over property tax policy in New York towns. Consistent with expectations, voter control has large impacts on property tax policy. The second empirical chapter examines what happens when voters gain control over local education policy in Virginia school districts. In this case, policy is unaffected when voter power is increased. The third and final empirical chapter examines what happens when voters gain control over fire protection policy in Illinois special district governments. In this case, the increase in voter control happens via two channels: elections and referendums. While elections have no effect on policy, referendums cause significant changes in both policy and performance. The final chapter concludes by considering several outstanding questions raised by the results, including the precise conditions under which voter control will matter, the implications of these results for debates over citizen competence, and the degree to which the results may be driven by elites capturing the democratic process.