When Voters Matter: The Limits of Local Government Responsiveness
I use over-time variation resulting from geographic sorting to estimate the impact of local preferences on local policy. Building a new panel data set of county-level voting behavior and county-level policy choices, I find a causal impact of changes in preferences on changes in policy. However, this responsiveness is not unlimited. While public safety and infrastructure spending respond, policies more constrained by state laws do not; and while general purpose government spending is responsive, school and special district spending are not. These results corroborate recent evidence of local responsiveness using a more extensive sample, and a research design that rules out many alternative explanations. They also reveal many local policies to be unresponsive due to state-imposed constraints.